article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
a basic question in random matrix theory is to ask for the probability distribution of the number of real eigenvalues for an ensemble of @xmath2 random matrices with real entries . with the ensemble made up of standard gaussian random matrices , i.e. in the circumstance that each element is independently chosen as a real standard gaussian , edelman @xcite was the first person to obtain results on this problem .
the approach taken centered on knowledge of the explicit functional form of the probability density function ( pdf ) for the event that there are @xmath0 real eigenvalues denoted @xmath3 , and @xmath4 complex eigenvalues denoted @xmath5 with @xmath6 ( the fact that the complex eigenvalues occur in complex conjugate pairs implies @xmath0 must have the same parity as @xmath7 ) .
thus it was shown that this is equal to @xmath8 where @xmath9 denotes the vandermonde determinant and @xmath10 ( see also @xcite ) .
integrating ( [ 3.1 ] ) over @xmath11 gives the probability @xmath12 that there are exactly @xmath0 real eigenvalues .
the simplest case to compute is when @xmath13 and thus all eigenvalues are real , for which the probability was found to equal @xmath14 .
questions relating to the probability that all eigenvalues are real for random matrices with real entries occur in applications .
consider first the tensor structure @xmath15 , represented as the column vector @xmath16 .
as reviewed in @xcite , it is of interest to find matrices @xmath17
\in \mathbb r^{p \times r}$ ] , @xmath18 \in \mathbb r^{p \times r}$ ] , @xmath19 \in \mathbb r^{2 \times r}$ ] such that @xmath20 for @xmath21 referred to as the rank as small as possible .
it turns out that with both @xmath22 and @xmath23 random matrices , entries chosen from a continuous distribution , one has that @xmath24 if all the eigenvalues of @xmath25 are real , and @xmath26 otherwise @xcite . in the gaussian case
these probabilities have been computed in @xcite and @xcite as equal to @xmath27 , where @xmath28 denotes the barnes @xmath1-function , and the corresponding large @xmath21 asymptotic form has been computed in @xcite .
a second example comes from quantum entanglement .
lakshminarayan @xcite considered the problem of quantifying when two - qubits @xmath29 and @xmath30 are an optimal pair , in the case that the states are chosen from a uniform distribution on the unit 3sphere .
the condition of being an optimal pair is known @xcite as particular inequalities for certain weighted inner products between the qubits .
it was shown in @xcite that these can be interpreted as the condition for the probability that the random matrix product @xmath31 , with each @xmath32 a @xmath33 real gaussian matrix , having all eigenvalues real , which was furthermore shown to be equal to @xmath34 .
an intriguing effect was observed in the study @xcite , which seems to have escaped early notice .
thus , noting from the result of edelman cited above that for a single real gaussian @xmath33 random matrix the probability of all eigenvalues being real is equal to @xmath35 , while for a product of two independent gaussian @xmath33 random matrices it is @xmath34 , the fact that @xmath36 led lakshminarayan to investigate if the probability of all eigenvalues being real was an increasing function of the number of matrices in the product .
numerical simulation indicated that this is indeed the case , and further the probability that all eigenvalues are real tends to unity as the number of random matrices in the product tends to infinity .
evidence that this is also true for products of @xmath37 real gaussian matrices was given in @xcite , while the follow up work @xcite provided similar evidence for random matrices with independent non - gaussian entries . a proof in the instance of the latter circumstance that the entries are all independent and identically distributed with a pdf containing an atom has recently been given in @xcite .
the appearance of the work @xcite coincided with the appearance of works containing other surprising advances relating to the eigenvalues of products of random matrices .
consider the random matrix product @xmath38 where each @xmath32 is an @xmath2 standard gaussian matrix . in the case of complex entries , akemann and
burda @xcite showed that the eigenvalues form a determinantal point process in the complex plane .
this means that the @xmath0-point correlation function for the eigenvalues @xmath39 is fully determined by a single function @xmath40 , referred to as the correlation kernel , according to @xmath41_{j , l=1,\dots , k}.\ ] ] in the case of real entries , forrester @xcite found a closed form expression for the probability that all eigenvalues are real . to specify
this latter result requires introducing the meijer @xmath1-function @xmath42 where @xmath43 is an appropriate contour relating to the validity of the inverse mellin transform formula . with @xmath44 denoting the probability that the random matrix product @xmath45 ( [ yx ] ) has exactly @xmath0 real eigenvalues it was shown in @xcite that , for each @xmath32 a real ginibre matrix , we have @xmath46_{k=1,\ldots , n/2}^{j=1,\ldots , n/2}\big ] , & n\ \text{even}\\[2 mm ] \det\big [ [ a_{j , k}]_{k=1,\ldots,(n-1)/2}^{j=1,\ldots,(n+1)/2}\quad [ \tilde a_j]_{j=1,\ldots,(n+1)/2}\big ] , & n\ \text{odd } \end{cases}\ ] ] with @xmath47 a simple identity for the meijer @xmath1-function evident from the definition ( [ gg ] ) shows that @xmath48 is equal to the meijer @xmath1-function occurring in @xcite .
moreover , these explicit formulas were used to prove that @xmath49 as @xmath50 .
extension to rectangular matrices where given in @xcite , while special arithmetic properties were shown to be present in the case @xmath51 @xcite .
a primary aim of the present paper is to extend this result to the calculation of @xmath44 , for general @xmath52 with the same parity as @xmath7 .
the following theorem will be proved in section [ sec : k - real ] .
[ t1 ] consider the random matrix product ( [ yx ] ) , in which each @xmath32 is a real ginibre matrix .
let @xmath53 with @xmath54 , @xmath55 given by and @xmath56 .
for @xmath7 even , the probability @xmath57 that exactly @xmath58 eigenvalues are real is given by @xmath59 \det \big [ b_{j , l}(\zeta ) \big ] _ { j , l=1,\dots , n/2},\ ] ] while for @xmath7 odd we have @xmath60 \det \big [ [ b_{j , l}(\zeta ) ] ^{j=1,\dots,(n+1)/2}_{k=1,\dots,(n-1)/2 } \quad
[ \tilde a_j ] _ { j=1,\dots,(n+1)/2 } \big]\ ] ] with @xmath61 from . in both and @xmath62
is a generating function parameter for the probabilities and @xmath63 f(\zeta ) $ ] denotes the coefficient of @xmath64 in the power series expansion of @xmath65 , i.e. for @xmath7 even @xmath66_{j , l=1,\dots , n/2}\ ] ] and similarly for the @xmath7 odd case in terms of .
a formula closely related to theorem [ t1 ] in the case @xmath67 was derived by akemann and kanzieper @xcite , and this working was soon after refined @xcite to obtain a formula equivalent to ( [ a1 ] ) .
also for this case forrester and nagao @xcite gave a result more general than ( [ a1 ] ) , applying to a real random matrix formed from a general linear combination of gaussian symmetric and anti - symmetric matrices .
we now turn our attention to the other primary aim of our work .
this relates to the statistical state formed by the eigenvalues of the product ( [ yx ] ) . in the complex case
, it has been remarked that the statistical state is a determinantal point process . in the real case
, it is known from the work of ipsen and kieburg @xcite that the eigenvalue correlations form instead a pfaffian point process .
thus , considering for definiteness the real eigenvalues , one now has @xmath68_{j , l=1,\dots , k } , \qquad \mathbf k^{\rm rr } ( x , y ) = \begin{bmatrix } d(x , y ) & s(x , y ) \\
- s ( y , x ) & \tilde{i}(x , y ) \end{bmatrix},\ ] ] where @xmath69 and @xmath70 are antisymmetric functions of @xmath71 and @xmath72 .
a concern of the present paper is to compute the explicit form of the correlation kernel in ( [ k0 ] ) in the case of the real eigenvalues of ( [ yx ] ) for real standard gaussian matrices , and also for the case of the complex eigenvalues . in this paper
, we will see that these correlation kernels possess many similarities with other results for product of random matrices .
for example , the kernel for the pfaffian point process specifying the scaled statistical state about the origin of the real eigenvalues of products of real ginibre matrices is given in terms of meijer @xmath1-functions .
in the simplest case of the one point function @xmath73 the resulting functional form is very succinct .
[ t2 ] define @xmath74 \
, \delta ( \lambda - \lambda^{(1 ) } \cdots \lambda^{(m)}).\ ] ] we have @xmath75 for singular values of products of complex ginibre matrices , it is similarly the case that the kernel for the scaled determinantal point process in the neighbourhood of the origin can be expressed in terms of meijer @xmath1-functions @xcite ; see also the recent review @xcite .
moreover , for fixed @xmath7 , knowledge of the real - to - real eigenvalue correlations gives information about the moments of the distribution function for the probability that there are @xmath0 real eigenvalues .
in particular , integration of the spectral density ( one - point function ) gives the expected number of real eigenvalues .
the rest of this paper is organised as follows . in section [ sec : jpdf ] we find the joint eigenvalue pdf for a gaussian product matrix with a given number of real eigenvalues . in section [ sec :
skew ] we introduce the generalised partition function and find the skew - orthogonal polynomials ; we combine these results with the joint eigenvalues pdf to prove theorem [ t1 ] .
section [ sec : corr ] focuses on the real - to - real and the complex - to - complex eigenvalue correlations . in
particularly , we study local and global scaling limits for the spectral densities and use the real spectral density to compute the expected number of real eigenvalues . the final section briefly sketches how all these results may be extended to products of rectangular matrices .
our first task is to find the explicit functional form for the eigenvalue pdf of the random matrix product ( [ yx ] ) in the case that each @xmath32 is an independent @xmath2 standard real gaussian matrix . with this specification
the joint probability measure for @xmath76 is equal to @xmath77 actually this task , extended to the general bi - orthogonal invariant ensembles , has already been addressed by ipsen and kieburg @xcite .
however the workings therein are not sufficient for all our purposes . in particular proportionality constants
are ignored , meaning that it is not possible to proceed to derive the formulas of theorem [ t1 ] for the probabilities @xmath44 .
these normalisation constants were included in the thesis @xcite but the pdf were given in terms of @xmath78 matrices , which is impractical for our purpose .
furthermore , the case that the working of @xcite which is a generalisation of the strategies used in @xcite and @xcite in the cases @xmath67 and @xmath79 respectively treats the real and complex eigenvalues on an equal footing , whereas we prefer to proceed in the way used in @xcite for @xmath67 which distinguishes the real and complex eigenvalues from the outset .
below we give a more practical formulation of the joint eigenvalue pdf .
[ t3 ] let @xmath80 \
, \delta ( \lambda - \lambda^{(1 ) } \cdots \lambda^{(m)}),\ ] ] referred to as the real ( or one - point ) weight function and let @xmath81^{1/2 } \big ) \ ] ] with @xmath82 \delta(g - g^{(1 ) } \cdots g^{(m)}),\ ] ] referred to as the complex ( or two - point ) weight function .
consider the product .
given that there are @xmath0 real eigenvalues ( @xmath0 of the same parity as the matrix dimension @xmath7 ) , the joint eigenvalue pdf is @xmath83 { \delta\big(\{\lambda_l\}_{l=1}^{k } \cup \ { x_j \pm i y_j \}_{j=1}^{(n - k)/2}\big ) } \prod_{j=1}^k w_r(\lambda_j)\prod_{j = k+1}^{(n+k)/2 } w_c(x_j , y_j)\ ] ] with @xmath84 given by and @xmath85 as above .
the starting point is to use a generalised real schur decomposition to triangularise the matrices @xmath86 which appear in the product . assuming that the product matrix has @xmath0 real eigenvalues , the decompositions states that for invertible matrices ( gaussian matrices are invertible almost surely )
we may write ( * ? ? ?
a.26 ) @xmath87 with @xmath88 .
here each @xmath89 is a real orthogonal matrix in @xmath90 with @xmath91 defined to be the set of matrices in @xmath92 with the first entry in each column positive .
each @xmath93 is a ( block ) diagonal matrix with the first @xmath0 diagonal entries scalars @xmath94 and the next @xmath95 block entries @xmath33 matrices @xmath96 , while each @xmath97 is a strictly upper triangular matrix consisting of @xmath98 independent gaussian random variables .
the generalised schur decomposition may be verified by applying an ordinary schur decomposition on the product matrix itself and then using ( partial ) qr decompositions on @xmath99 , recursively ( see ( * ? ? ? *
appendix a ) for details ) . we stress that while it is possible to choose @xmath100 of the matrices @xmath93 in to be strictly diagonal rather than block diagonal ( due to the @xmath100 qr decompositions )
, we do not do so as it would complicate the derivation of the jacobian . for the following
, it will be convenient to introduce the product @xmath101 which again is a block diagonal matrix .
the first @xmath0 diagonal entries are scalars , @xmath102 , while the latter @xmath95 entries are @xmath33 matrices , @xmath103 . with this notation , the jacobian for the above given change of variables reads ( * ? ? ?
a.26 ) @xmath104 where @xmath105 refers to the eigenvalue(s ) of the @xmath106-th entry of the block diagonal matrix @xmath107 , i.e. @xmath108 for @xmath109 and @xmath105 denotes the two ( complex ) eigenvalues of the @xmath33 block @xmath110 for @xmath111 .
thus , using the notation @xmath112 with @xmath113 for the complex eigenvalues , we have @xmath114 this notation is the same as used by edelman ( * ? ? ?
* eq . ( 6 ) ) .
more compactly , we may write @xmath115 where the vandermonde determinant is defined as in . for the weight function in ( [ z1 ] )
we have @xmath116 where we can integrate out the dependence on @xmath117 and @xmath118 according to @xmath119 the latter is equal to @xmath120 . using all the above results
, it follows that , for a given @xmath0 , the joint probability measure for the eigenvalues is equal to @xmath121.\end{gathered}\ ] ] we have , at this point , not yet explicitly introduced the constraint that the eigenvalues of each @xmath122 are not real and thus are consequently a complex conjugate pair . for this reason
, we have a similarity with ( * ? ? ?
4.26 ) . in order to explicitly impose our constraint that the product matrix has exactly @xmath0 real eigenvalues ,
we suppose an orthogonal similarity transformation has been used to bring each matrix @xmath123 into the form @xmath124 with @xmath125 .
the eigenvalues are then @xmath126 with @xmath127 , and we know too ( see e.g. ( * ? ? ?
* proof of prop .
15.10.1 and prop .
15.10.2 ) ) that changing variables from the elements of @xmath123 to @xmath128 , where @xmath129 parametrises the orthogonal similarity transformation and @xmath130 introduces the jacobian @xmath131 the fact that the integrand in ( [ l3 ] ) is invariant under real orthogonal transformations allow us to simplify further .
firstly , we may integrate out @xmath129 , which contributes with an extra factor of @xmath132 .
secondly , we may replace the matrix @xmath123 by the diagonal matrix of its singular values , @xmath133 and @xmath134 say . in terms of the variables
@xmath135 it is straightforward to compute that the singular values are given by ( [ l3a ] ) .
combining these results completes the proof . due to the relatively involved expression for the two - point weight ,
it might be beneficial to briefly expand on the simplest cases , @xmath67 and @xmath79 , where explicit expressions are known .
for @xmath67 , the joint pdf ( [ 3.1a ] ) must , of course , reduce to the classical result ( [ 3.1 ] ) . inspection of and shows that the integration therein are immediate for @xmath67 due to the delta functions . in the real case we then read off that @xmath136 . in the complex case , substituting in ( [ l3a ] ) gives @xmath137 where the second equality first appeared in @xcite , albeit out by a factor of 2 as remarked in @xcite . substituting these evaluations in
indeed reproduces .
returning now to the case @xmath79 , the meijer @xmath1-function ( [ c1a ] ) is a modified bessel function , @xmath138 to simplify ( [ l3a ] ) requires simplifying ( [ l3 ] ) . for this purpose , and without yet restricting @xmath139
, we introduce @xmath33 real matrices @xmath140 and set @xmath141 equal to the @xmath78 identity matrix .
we note that @xmath142 which allows the integration over @xmath143 to be carried out in using the delta function , showing that @xmath144 \frac{e^{-\frac12 \operatorname{tr}((m^{(m-1)}m^{(m-1)t})^{-1}gg^t)}}{\sqrt{2\pi^3}}.\ ] ] this is the two - by - two matrix version of ( * ? ? ?
* eq . ( 2.20 ) ) .
a further change variables @xmath145 for each @xmath146 shows @xmath147
\frac{e^{-\frac12 \operatorname{tr}((a^{(m-1)})^{-1}gg^t)}}{\sqrt{2\pi^3 } } , \ ] ] where the integration is over positive - definite real symmetric matrices @xmath148 , @xmath146 . in the case
@xmath79 we can also express the integral in terms of modified bessel functions .
[ le1 ] we have @xmath149 \big ) } = 8 \int_1^\infty \frac{s}{(s^2 - 1)^{1/2 } } k_0(s \mu_+ ) k_0(s \mu_- ) \ , ds.\ ] ] write for the @xmath33 positive definite matrix @xmath150 @xmath151 then @xmath152 is equivalent to @xmath153 using the notation @xmath154 for the determinant , and expressing this equation as a delta function constraint allows us to write @xmath155 the working now is elementary .
we first integral over @xmath156 , change variables @xmath157 , @xmath158 , and integrate over @xmath159 , then @xmath160 and @xmath161 , using the fact that @xmath162 the last step is to change variables @xmath163 .
alternative expressions for @xmath164 are known .
one , which involves not the @xmath165 bessel function but rather the @xmath166 bessel function is based on changing variables to the eigenvalues and eigenvectors of @xmath150 , @xmath167 , and using the matrix integration formula for the integral over haar measure of the @xmath33 orthogonal group restricted to matrices with elements in the first entry of each column positive , @xmath168 implying that @xcite @xmath169 another , which is based on working similar to that used in the proof of lemma [ le1 ] , but starting from ( [ l3b ] ) rather than ( [ l3c ] ) tells us that @xcite @xmath170 there is some advantage in the form ( [ 2.17 ] ) , due to its functional dependence on @xmath171 and @xmath172 , which according to ( [ l3a ] ) are given in terms of @xmath173 by @xmath174 recalling the definition of @xmath164 in ( [ l3c ] ) , this tells us that @xmath175 substituting in ( [ l3a ] ) and using the integral in ( [ wc1 ] ) to integrate over @xmath176 we obtain @xcite @xmath177 in the following , we will see that it is possible to calculate the probability finding exactly @xmath0 eigenvalues without such explicit knowledge of the two - point weight function . here , we make note of them to make contact with the existing literature and as a reference for a comment in section [ sec : c - eigenvalues ] .
finally , we note that an important difference compared to the result presented in ( * ? ? ?
4.26 ) is the shift from the two - by - two matrix weight function to which will be essential in the remaining sections .
let us denote the joint pdf ( [ 3.1a ] ) by @xmath178 , and define the generalised partition function for @xmath0 real and @xmath95 complex conjugate pairs of eigenvalues by @xmath179 = \prod_{j=1}^k\int_{\mathbb{r}}d\lambda_j \ , u(\lambda_j ) \prod_{l=1}^{(n - k)/2 } \int_{{\mathbb{r}}\times{\mathbb{r}}_+ } dx_l dy_l \ , v(x_l , y_l ) \mathcal q(p_m).\ ] ] we have that with @xmath180 the generalised partition function is the probability of finding @xmath0 real eigenvalues and @xmath95 complex conjugate pairs of eigenvalues .
functional differentiation of @xmath181 : = \sum_{k=0}^n z_{k,(n - k)/2}[u , v],\ ] ] where the sum is restricted to @xmath0 of the same parity of @xmath7 allows the correlation functions to be computed ; see e.g. @xcite .
independent of the specific functional form of @xmath182 and @xmath183 in ( [ 3.1a ] ) , an observation of sinclair @xcite tells us that due to the product of difference @xmath184 , the method of integration over alternative variables implies that @xmath185 $ ] can be written as a pfaffian .
the details of the necessary working can be found in e.g. ( * ? ? ?
15.10.3 , @xmath7 even ) and ( * ? ? ?
* ( @xmath7 even ) and 4.3.2 @xmath7 odd ) ) .
we report the final result only .
[ u1 ] let @xmath186 be a set of monic polynomials , with @xmath187 of degree @xmath188 .
let @xmath189 and @xmath190 for @xmath191 even we have @xmath192 = \big ( \frac1{z_{k , n } } \big ) ^m [ \zeta^{k/2 } ] { \rm pf } [ \zeta \alpha_{j , l } + \beta_{j , l}]_{j , l=1,\dots , n},\ ] ] while for @xmath191 odd we have @xmath193 = \big ( \frac1{z_{k , n } } \big ) ^m [ \zeta^{(k-1)/2 } ] { \rm pf } \begin{bmatrix } [ \zeta \alpha_{j , l } + \beta_{j ,
l } ] & [ \mu_j ] \\ [ -\mu_l ] & 0 \end{bmatrix } ] _ { j , l=1,\dots , n}\ ] ] with @xmath194f(\xi)$ ] defined as in theorem [ t1 ] and @xmath195 given by .
the matrix @xmath196 $ ] is antisymmetric . for @xmath197 and @xmath198 ,
it is possible to choose the monic polynomials @xmath186 so that this anti - symmetric matrix is block diagonal , with the blocks @xmath33 anti - symmetric matrices @xmath199 @xmath200 , @xmath7 even , and @xmath201 , @xmath7 odd , with the last diagonal entry 0 in this latter case .
in fact , from a theoretical perspective this is also true for general @xmath62 , however our method for these polynomials ( given below ) is only valid if @xmath202 .
the use of skew - orthogonal polynomials is standard in random matrix theory ; see e.g. ( * ? ? ?
* ch . 6 ) .
schmidt procedure shows that the construction of such polynomials is always possible , and that they are unique up to the mapping @xmath203 this mapping , for @xmath204 an arbitrary constant , leaves the skew - orthogonality property unchanged . with @xmath205 and @xmath206 specified by define the skew - product @xmath207 in the case @xmath67 , when the underlying point eigenvalues pdf is given by ( [ 3.1 ] ) , the corresponding skew - orthogonal polynomials were first determined by forrester and nagao @xcite .
they were found to be @xmath208 with normalisation @xmath209 the case @xmath79 has been considered by akemann and collaborators @xcite .
in fact these authors determined the skew - orthogonal polynomials for a more general model , in which the matrices @xmath210 and @xmath211 in ( [ yx ] ) with @xmath79 are a general linear combination of gaussian symmetric and anti - symmetric matrices , as already noted below theorem [ t1 ] . specialising to the case that @xmath210 and @xmath211 are both standard gaussian matrices , we read off the skew - orthogonal polynomials @xmath212 with normalisation @xmath213 comparing the skew - orthogonal polynomials and as well as the normalisations and a simple pattern seems apparent : the coefficient in the odd skew - orthogonal polynomials as well as the normalisations constants are raised to powers of @xmath139 , @xmath67 and @xmath79 , respectively .
this pattern indeed persists in the general case .
[ prop : skew ] for the skew - product and @xmath214 , the polynomials @xmath215 form a skew - orthogonal set with normalisation @xmath216 in @xcite the method used to find the skew - orthogonal polynomials was to first establish that with @xmath67 @xmath217 which in turn made essential use of knowledge of the explicit functional form of @xmath218 and @xmath219 .
some of the details of the working are given in ( * ? ? ?
* proof of prop .
15.10.4 ) .
soon after sommers @xcite noted that knowledge of the functional form of the averages of the product of two characteristic polynomials @xmath220 , summed over @xmath0 contains sufficient information to fully determine the skew - orthogonal polynomials .
subsequently , akemann , kieburg and phillips ( * ? ? ?
( 4.6)(4.7 ) ) gave the explicit matrix averages formulas @xmath221 for the skew - orthogonal polynomials , where in the present setting the average over @xmath1 is over the @xmath139 standard gaussian matrices @xmath222 of size @xmath223 .
forrester @xcite gave a systematic way to compute averages of the form ( [ f1 ] ) in the cases that @xmath1 is drawn from an ensemble invariant under real orthogonal transformations .
this method was based on the use of zonal polynomials , and was built on ideas contained in @xcite . here
we will show that elementary methods suffice to evaluate ( [ f1 ] ) . according to ( [ yx ] )
, @xmath45 is the product of @xmath139 independent standard gaussian matrices @xmath222 .
moreover , from the rule for matrix multiplication , and this specification of the @xmath32 , we see that elements taken from distinct rows @xmath224 and columns @xmath225 , each @xmath226 are uncorrelated , so that with @xmath227_{j , k=1,\dots , n}$ ] @xmath228 from the definition of a determinant we have @xmath229 where @xmath230 denotes the parity of the permutation @xmath231 and @xmath232 denotes the kronecker delta .
averaging over @xmath222 using ( [ f2 ] ) shows that the only non - zero term comes from the identity permutation and furthermore this average is equal to @xmath233 .
this establishes @xmath234 in . for the odd polynomials , multiplication by @xmath235 shows that a non - zero value will appear after averaging when there is a single monomial @xmath236 in the expansion of @xmath237 .
we see from ( [ f1b ] ) that this is only possible in the case of the identity permutation , and that we require the coefficient of @xmath238 therein .
thus @xmath239 here the final equality follows by noting that @xmath236 consists of a sum of @xmath240 terms which are monomials in the elements of the @xmath32 , and due to ( [ f2 ] ) the only terms that survives this averaging after squaring are the @xmath240 perfect squares , which contribute unity .
substituting this result into establishes @xmath241 in .
it remains to establish ( [ 2 m ] ) . on this point
, we first note that from the meaning of @xmath185|_{u = v=1}$ ] as the probability that there are exactly @xmath0 real eigenvalues , it follows that @xmath242|_{u , v=1 } = 1 $ ] , where @xmath242 $ ] is specified by ( [ l3d1 ] ) . on the other hand ,
it follows from ( [ 3.33 ] ) that for @xmath243 even @xmath244 = \big ( \frac{1 } { 2^{n(n+1)/4 } \prod_{l=1}^n \gamma(l/2 ) } \big ) ^m { \rm pf } [ \alpha_{j , l } + \beta_{j , l}]_{j , l=1,\dots , n}.\ ] ] setting @xmath198 , and using the skew - orthogonal polynomials , the rhs can be evaluated to give @xmath245 and ( [ 2 m ] ) follows .
[ mario - remark ] examination of the above proof shows that invariance of a single matrix entry under the reflection @xmath246 implies @xmath247 it has already been remarked below the definition of the generalised partition function that with @xmath198 this quantity can be interpreted as the probability @xmath44 that for the ensemble of matrices specified by ( [ yx ] ) , with each @xmath32 therein an @xmath2 real standard gaussian , there are exactly @xmath0 real eigenvalues .
this assumes @xmath0 and @xmath7 have the same parity ; if not the probability is zero .
according to proposition [ u1 ] these probabilities can be written as pfaffians .
let us suppose the polynomials therein are furthermore even ( odd ) when there degree is even ( odd ) .
we then know , by the symmetry of the integrands , that each @xmath248 unless the parity of @xmath249 and @xmath250 is opposite . furthermore making use of the fact that the @xmath251 is anti - symmetric in @xmath252
allows the pfaffian to be written as a determinant of half the size , telling us that for @xmath243 even @xmath253 \det [ ( \zeta \alpha_{2j-1,2l } + \beta_{2j-1,2l } ) |_{u = v=1 } ] _ { j , l=1,\dots , n/2},\ ] ] and for @xmath243 odd @xmath254 \det\big [ ( \zeta \alpha_{2j-1,2l } + \beta_{2j-1,2l})|_{u = v=1 } ] \quad [ \mu_{2j-1 } ] \big]_{\substack{j=1,\dots,(n+1)/2 \\
l=1,\dots,(n-1)/2}}.\ ] ] we are now well placed to establish ( [ a1 ] ) and ( [ 14a ] ) .
we choose the polynomials @xmath255 as the skew - orthogonal polynomials ( [ 1 m ] ) so we have @xmath256 with the explicit value of @xmath257 being given by ( [ 2 m ] ) .
thus we have been able to eliminate the dependence on @xmath258 , which from the definition ( [ be1 ] ) involves the weight @xmath219 a quantity which from ( [ l3a ] ) is not in general known in terms of explicit special functions .
the remaining quantity @xmath259 is specified by ( [ be1 ] ) , and the weight therein @xmath218 is given as a meijer @xmath1-function according to ( [ c1a ] ) .
in fact this very same quantity , up to a proportionality has appeared in the earlier study ( * ? ? ?
* proposition 3 ) and we read off the evaluation @xmath260 where we use the definition ( [ aa ] ) with @xmath261 since the lowest order odd skew - orthogonal polynomial is a monomial . substituting ( [ s1 ] ) in ( [ s0 ] ) , and substituting the result in turn in ( [ s5 ] ) we obtain after minor manipulation the formula ( [ a1 ] ) . to deduce ( [ 14a ] ) we require the additional evaluation , also contained in ( * ? ? ?
* proposition 3 ) , @xmath262 , and similarly substitute in ( [ s6 ] ) .
for @xmath67 the probabilities @xmath263 have been known since the late nineties and they are all of the form @xmath264 where @xmath265 and @xmath266 are rational numbers @xcite .
tabulations can be found in ( * ? ? ?
* table 5 ) and ( * ? ? ?
* table 2 ) .
recently , an evaluation of the meijer @xmath1-function @xmath267 as a summation over a linear combination of @xmath268 has been given by kumar @xcite , and this was used to show @xmath269 which allows us to get explicit expressions for the probabilities @xmath270 ( i.e. @xmath79 ) .
note in particular that this is of the form @xmath271 times a rational number , a feature which was conjectured in @xcite .
substituting in ( [ a1 ] ) and ( [ 14a ] ) in the case @xmath79 makes the structure of the probabilities explicit for @xmath270 .
these are all polynomials of degree @xmath272 in @xmath132 with rational coefficients ; probabilities for low values of @xmath7 are tabulated in table [ table : k - real ] .
it is worth noting that similar probabilities for the real spherical and the truncated orthogonal ensembles are also given as polynomials in @xmath132 and @xmath273 ; see @xcite and references therein for an extensive summary . beyond the cases @xmath67 and @xmath79 ,
evaluation formulas for the meijer @xmath1-function in ( [ aa ] ) are challenging .
in addition to the contour integral representation , we may also write the meijer @xmath1-function as an @xmath139-fold integral on the real line , @xmath274 \frac{x_1^k}{(1+x_1)^{j+k-1/2}},\ ] ] which may be checked to agree with for @xmath79 .
such @xmath139-fold integral representations give a relation to product of random scalars .
however , explicit expressions in terms of elementary functions remain unknown for @xmath275 . with an explicit method for calculating the probability of finding @xmath0 real eigenvalues , it seems natural to ask for different types of number statistics
. a prime example would be the expected number of real eigenvalues .
albeit such expectation values may be calculated using theorem [ t1 ] , we will see in section [ sec : r - eigenvalues ] that the spectral density for the eigenvalues can be used to obtain a more efficient formula .
the interest in real eigenvalue statistics , of course , extends beyond the expected number of real eigenvalues .
another common question is to ask for extreme value statistics , i.e. the probability that there are abnormally many ( or few ) real eigenvalues . as mentioned in the introduction , the probability that all eigenvalues are real
has already be studied in @xcite , which led to the remarkable conclusion that this probability tends to unity for @xmath276 .
it is more challenging to ask for the probability of finding only a few real eigenvalues in the large-@xmath7 limit , say the probability that an even dimensional product matrix has no real eigenvalues .
a step in this direction was taken in @xcite , where using the relation to the brownian annihilation process @xmath277 , the first two terms of the large @xmath266 asymptotics of the probability that there are no real eigenvalues in an interval of size @xmath266 near the origin for @xmath278 real ginibre ( @xmath67 ) was computed .
it was realized kanzieper et al .
@xcite that heuristic at least this result implies for large @xmath7 @xmath279 with @xmath280 denotes the riemann zeta function and @xmath281 and moreover these authors gave a rigorous proof of the leading term .
it is not known how to generalize the workings of @xcite , which are based on theorem [ t1 ] , beyond @xmath67 .
however , our theorem [ t1 ] at least allows us to establish numerical estimates , e.g. fitting @xmath282 to @xmath283 for @xmath284 suggest that @xmath285 for @xmath7 even .
we note that @xmath286 , which is in the agreement with the expectation that @xmath287 decreases with increasing @xmath139 .
the pfaffian formulae of proposition [ u1 ] for the generalised partition function , combined with the simplification inherent in the use of skew - orthogonal polynomials , @xmath289 , allow the @xmath0-point correlation to be expressed in the form ( [ k0 ] ) with entries given in terms of @xmath289 . while ( [ k0 ] ) refers to the real - to - real eigenvalue correlations , this same structure remains true for the general correlation functions .
in fact , the entries of the correlation kernel also have the same structure ; see e.g. ( * ? ? ? * and 4.6 ) .
let @xmath290 denote the upper complex half - plane , and specify @xmath218 and @xmath219 according to ( [ c1a ] ) and ( [ l3a ] ) .
define @xmath291 \tau_j(\mu ) & = \left \ { \begin{array}{ll } \displaystyle - \frac12 \int_{-\infty}^\infty { \rm sgn } \ , ( x - y ) q_j(y ) \ , dy & \mu = x
\in \mathbb r \\
i q_j(x+iy ) , & \mu = x + i y \in \mathbb c_+ \end{array } \right .
\nonumber \\[.5em ] \epsilon(\mu,\eta ) & = \left \ { \begin{array}{ll } \frac12 { \rm sgn } \ , ( \mu - \eta ) , & \mu,\eta \in \mathbb r , \\ 0 , & { \rm otherwise } \end{array } \right .
\end{aligned}\ ] ] in this notation , the entries of the correlation kernel ( [ k0 ] ) in the case of the correlation between real eigenvalues only , or the correlation between complex conjugate pairs of eigenvalues are given by @xmath292 for @xmath7 odd these expressions require modification ; see e.g. @xcite , @xcite .
for efficiency of presentation , we will restrict attention to the @xmath7 even case .
our main interest in section [ sec : c - eigenvalues ] and [ sec : r - eigenvalues ] will be spectral densities ( one - point correlation functions ) and quantities derivable from these .
for this reason , we focus on the complex - to - complex and the real - to - real eigenvalue correlations , but real - to - complex correlations can be treated in a similar manner .
we see from ( [ ab ] ) and ( [ ab1 ] ) that in the case of the correlation between complex eigenvalues , up to factors involving @xmath219 all the quantities are polynomials , and are related by @xmath293 thus it suffices to consider @xmath294 , where @xmath295 and @xmath296 . for this , ( [ ab ] ) and ( [ ab1 ] )
tell us that @xmath297 upon use of the skew - orthogonal polynomials given by proposition [ prop : poly ] this simplifies to @xmath298 we are typically interested in either a global scaling regime ( where the eigenvalues are concentrated within a region with compact support ) or local scaling regimes ( where the eigenvalue interspacing is order unity ) . for simplicity ,
let us focus on the one - point function ( i.e. the spectral density ) which for complex eigenvalues is given by @xmath299 .
the global scaling regime for the spectral density is known from free probability @xcite , @xmath300 where @xmath301 if @xmath150 is true , @xmath302 otherwise .
this holds because the full spectral density ( i.e. including complex as well as real eigenvalues ) is dominated by the complex spectrum in the global scaling regime .
we note that there also exists a global scaling regime for the real spectrum , albeit sub - dominant .
we will return to this limit in section [ sec : r - eigenvalues ] . on the local scale ,
the region near the origin is of greatest interest since it gives rise to new types scalings ( i.e. different than the ordinary ginibre case ) .
the local density near the origin is given by @xmath303 since @xmath304 we recall from section [ sec : jpdf ] that the weight function @xmath219 has an explicit and concise expression for @xmath305 but not for @xmath306 .
we note that if @xmath305 then the meijer @xmath1-function in evaluates as @xmath307 with the latter being a modified bessel function . combining this with the weight functions from section [ sec : jpdf ] reproduces known formulae for the density ( the @xmath79 case was given in @xcite ) . in this section
we focus on the part of the spectrum which is located on the real axis . similarly to the complex spectrum described above , all correlations may be expressed in terms of the pre - kernel @xmath308 .
we see from ( [ ab ] ) and ( [ ab1 ] ) that @xmath309 which produce the correlation functions by insertion in .
we note that the relations between the pre - kernels are more complicated for the real - to - real correlations than for the complex - to - complex correlations where the pre - kernels are related according to . on the other hand ,
the weight functions are simpler in the real case than in the complex case . using ( [ ab ] ) and ( [ ab1 ] ) and the skew - orthogonal polynomials ( proposition [ prop : poly ] ) , we write the pre - kernel as @xmath310 for @xmath67 ( i.e. the ordinary ginibre ensemble )
, the sum may be rewritten as an incomplete gamma function times an exponential and the integral over @xmath311 can be performed , which yields @xcite @xmath312 this formulation of the pre - kernel is extremely useful in the study of large-@xmath7 asymptotics .
unfortunately there are no direct generalisation of this result to @xmath313 , which makes asymptotic analysis more challenging .
however , it is possible to perform the integral over @xmath311 in for arbitrary @xmath139 .
to do so , we rewrite ( [ sxy ] ) as @xmath314 now , standard identities for the meijer @xmath1-function give @xmath315 \displaystyle y^{1+j } { g^ { { m } , { 1 } } _ { { 1 } , { m+1 } } \big ( \begin{matrix } -(j-1)/2 \\
0,\dots,0,-(j+1)/2 \end{matrix}\ , \big\vert\ , \frac{y^2}{2^m } \big ) } , & j \
\textup{even}. \end{array } \right.\ ] ] this form of the pre - kernel @xmath308 is useful if we are interested in the expected number of real eigenvalues .
we recall that the expected number of real eigenvalues , @xmath316 , can be found by integration over the real spectral density @xmath317 .
thus , it follows from that @xmath318 where @xmath319 the quantity @xmath320 is precisely the same quantity appearing in the study @xcite , which evaluates to @xmath321 with @xmath48 given by the meijer @xmath1-function ( [ aa ] ) . in the case where the first index of @xmath322 is even and the second index odd ,
we use the anti - symmetric property @xmath323 .
this gives @xmath324 where @xmath325 and @xmath326 denote the ceiling and floor function , respectively .
we recall that the formulae above assume that @xmath7 is even ( for odd @xmath7 the expression is altered by the addition of unity ) .
as already mentioned , an evaluation of @xmath327 in terms of arithmetic constants is only known for @xmath305 ; consequently the same holds for .
the @xmath67 case is known since the mid nineties @xcite , while the @xmath79 case is evaluated using ; the results for small @xmath7 are tabulated in table [ table : e(real ) ] .
as anticipated , table [ table : e(real ) ] reveals that the expected value of real eigenvalues are consistingly larger for @xmath79 than for @xmath67 . for @xmath306 a computation of the expectation value requires numerical evaluation of the meijer @xmath1-functions .
the expected number of real eigenvalues can , of course , also be obtained using the probabilities given by theorem [ t1 ] .
in fact , for @xmath79 and small @xmath7 the expected number of real eigenvalues follows immediately from table [ table : k - real ] , e.g. for @xmath328 we see that @xmath329 which agrees with table [ table : e(real ) ] .
@xmath330_{m=2 } & n=2 & n=3 & n=4 & n=5 & n=6 & n=7\\ \hline\hline \text{exact } & \frac12\pi & 1+\frac{5}{16}\pi & \frac{755}{1024}\pi & 1+\frac{4185}{8192}\pi & \frac{3821355}{4194304}\pi & 1+\frac{22392747}{33554432}\pi \\ \text{aprox . } & 1.5708 & 1.9817 & 2.3163 & 2.6049 & 2.8622 & 3.0966 \\ \text{simul . } & 1.5704 & 1.9813 & 2.3168 & 2.6030 & 2.8607 & 3.0948 \\ ( \mathbb e[\#\text{reals}]_{m=1 } ) & ( 1.4142 ) & ( 1.7071 ) & ( 1.9445 ) & ( 2.1490 ) & ( 2.3312 ) & ( 2.4971 ) \end{array}\ ] ] let us return to the pre - kernel and consider large-@xmath7 asymptotics for the real spectral density . similarly to section [ sec : c - eigenvalues ] we focus on the local density near the origin and the global density . using , it is immediately seen that the local scaling regime near the origin gives ( [ r - local ] ) announced in theorem [ t2 ] .
compared to the same result for the complex density , the real density has the advantage that the weight function @xmath218 has a known expression as a meijer @xmath1-function for all @xmath139 while @xmath219 does not .
we note that for @xmath67 the meijer @xmath1-functions in are all simple exponentials ; this allows integration over @xmath311 and confirms that the local spectral density is constant for @xmath67 .
moreover , for @xmath67 the corresponding @xmath0-point correlation takes on the explicit form @xmath331 as obtained in @xcite .
we remark that it has been argued by beenakker and co - workers @xcite that the statistical state implied by ( [ k1 ] ) is realised by the level crossings of so - called majorana zero modes for a disordered semiconducting wire at a josephson junction , in a weak magnetic field . and
this same correlation kernel appears in the seemingly unrelated problem of the annihilation process @xmath332 in the limit @xmath333 @xcite .
a study of the global scaling regime for the real spectrum is more challenging . unlike the complex spectral density ( section [ sec : c - eigenvalues ] )
, we have no help from free probability . a qualified guess for this spectral density
might be obtained by looking at the @xmath139-th power of a real ginibre matrix rather than at the product of @xmath139 independent matrices .
it is immediate that the @xmath139-th power and the @xmath139-th product share the same complex macroscopic spectral density , thus assuming that this extends to the real spectrum we expect that @xmath334 where @xmath335 is defined as in .
for @xmath67 the density is well - known @xcite ; a verification follows from using known asymptotics for the incomplete gamma functions .
moreover , we see that the real spectrum develops a non - integrable singularity at the origin when @xmath139 tends to infinity similarly to as we would expect . for @xmath336
we have no rigorous derivation of but the form is supported by ( i ) a heuristic saddle point analysis and ( ii ) numerical data .
let us first look at the saddle point analysis , which takes as the starting point .
the first step is to introduce an approximation for the sum in .
we know from ( * ? ? ?
* appendix c ) that @xmath337 for @xmath338 while exponentially suppressed in @xmath7 for @xmath339 .
an approximation for the weight function is known from the literature on special functions @xcite , and we have @xmath340 we insert these approximations into and want to evaluate the integral over @xmath311 using a saddle point approximation . note that there are two maxima of the integrand symmetrically distributed around @xmath341 ( the integrand is equal to zero at @xmath341 ) .
these two maxima tend to @xmath71 from left or right , respectively , as @xmath7 tends to infinity .
thus , we will use an ansatz @xmath342 for our saddle points where @xmath343 is sub - dominant in @xmath7 . with this ansatz and expanding to lowest order
, the saddle points are found to be @xmath344 evaluation at either of these saddle points yields the conjectured form up to a normalisation .
finally , let us compare the density with a simulation of the random matrix product .
figure [ fig : histogram ] shows the visual similarity between the density for @xmath79 and numerical data stemming from a simulation of @xmath345 matrix products with @xmath346 .
it should be noted that convergence is expected to be exponentially fast in the bulk but considerably slower near the edges .
similar numerical tests have been performed for @xmath347 and it has been verified that the difference between the analytic formula and the numerical data decreases with increasing @xmath7 .
furthermore , we expect that the real global density is universal in the sense that the gaussian entries may be replaced by other independent variables under suitable assumptions on their moments .
this type of universality is known to hold for the complex spectra @xcite and the expectation that such results extend the real spectra is strengthend by numerical comparison generated from random sign ( @xmath348 ) matrices .
although it seems a very natural problem , this type of universality for the real global spectrum has received little attention in the literature ; this is true even for the classical ginibre ensemble ( @xmath67 ) . , while the histogram shows the distribution of the real eigenvalues from @xmath345 realisations of a product of two @xmath349 real asymmetric gaussian matrices ( @xmath350 eigenvalues in total ) . ]
a generalisation to the case of rectangular matrices is also available and we briefly treat it here . the main idea when dealing with a product of random matrices is to reformulate problem as a product of square random matrices with the same eigenvalue properties ; this is possible due to a general reduction procedure presented in @xcite ( see also ( * ? ? ?
2.4 ) ) . after this reformulation
, the approach is similar to the previous sections because proposition [ u1 ] as well as the formulae and are completely general . due to this similarity
we will only sketch the main ideas here .
we consider a product matrix , @xmath351 where each matrix @xmath32 has dimensions @xmath352 with @xmath353 denoting non - negative integers such that @xmath354 . here
the constraint is introduced to ensure that the product matrix is square and has @xmath7 non - trivial eigenvalues .
we note that if @xmath355 but @xmath356 for some @xmath357 ( i.e. the smallest matrix dimension is still @xmath7 ) then there will be @xmath358 eigenvalues which are trivially equal to zero ( and therefore real ) but the joint pdf otherwise remains the same except for an obvious change in normalisation .
consequently , all formulae given below may effortlessly be extended to the @xmath359 case if desired .
the generalisation of the probabilities with for a purely real spectrum have already appeared in the thesis ( * ? ? ?
* prop . 4.29 ) .
they are given by @xmath360_{k=1,\ldots , n/2}^{j=1,\ldots , n/2}\big ] , & n\ \text{even}\\[2 mm ] \det\big [ [ a_{j , k}^\nu]_{k=1,\ldots,(n-1)/2}^{j=1,\ldots,(n+1)/2}\quad [ \tilde a_j^\nu]_{j=1,\ldots,(n+1)/2}\big ] , & n\ \text{odd } \end{cases}\ ] ] with @xmath361 these formulae allow us to make some straightforward generalisations of the exact expressions presented by kumar @xcite in the @xmath79 case . following @xcite , we have @xmath362 the next step would be to rewrite gamma functions with a non - integer argument using gauss duplication formula .
the right - hand side of evaluates as @xmath363 for even @xmath364 and @xmath365 for odd @xmath364 where @xmath265 denotes some rational constant ( depending on both @xmath7 and @xmath364 ) .
this difference in the power of @xmath132 for even and odd @xmath364 has a remarkable consequence : for even @xmath364 the probabilities are given as a rational number times @xmath366 but for odd @xmath364 these constants are simple rational constants ( i.e. there is no powers of @xmath132 ) .
the probabilities of a purely real spectrum are tabulated in table [ table : rect-1 ] for small values of @xmath7 and @xmath364 .
@xmath367 as we have seen in previous sections , to extend the probabilities for a purely real spectrum to the probabilities @xmath368 we need a formula for the joint pdf of the eigenvalues and a formula for the skew - orthogonal polynomials , i.e. generalisations of theorem [ t3 ] and proposition [ prop : skew ] .
given such generalisations the rest of the results presented in previous sections may be extended as well due to the generality of proposition [ u1 ] .
[ prop : jpdf - rect ] given a gaussian product matrix of dimension @xmath7 with @xmath0 real eigenvalues , @xmath3 , and @xmath95 complex conjugate pairs of a eigenvalues , @xmath369 , the joint pdf for these eigenvalues is given by @xmath370 { \delta\big(\{\lambda_l\}_{l=1}^{k } \cup \ { x_j \pm i y_j \}_{j=1}^{(n - k)/2}\big ) } \prod_{j=1}^k w_r^\nu(\lambda_j)\prod_{j = k+1}^{(n+k)/2 } w_c^\nu(x_j , y_j)\ ] ] where @xmath371 \ , \delta ( \lambda - \lambda^{(1 ) } \cdots \lambda^{(m)}).\ ] ] @xmath372 with @xmath373 as in and @xmath374 \delta(g - g^{(1 ) } \cdots g^{(m)}).\ ] ] the normalisation is given by @xmath375 the proof follows the same lines as the proof of theorem [ t3 ] .
we use generalised real schur decomposition to get an expression for the joint pdf in terms of real eigenvalues and @xmath78 matrices , see ( * ? ? ?
finally , changing variables in this expression from the general @xmath78 matrix @xmath1 to a matrix using an orthogonal similarity transformation and introducing the singular values , @xmath373 , completes the proof .
[ prop : poly ] for the skew - product defined in accordance with the joint pdf given by proposition [ prop : jpdf - rect ] , the polynomials @xmath376 form a skew - orthogonal set with normalisation @xmath377 for a product square matrices , we found the skew - orthogonal polynomials by exploiting that elements taken of different rows and columns are uncorrelated .
this property is still true for rectangular matrices , thus skew - orthogonal polynomials are obtained following the exact same steps .
likewise for the normalisation where we evaluate the generalised partition function at @xmath198 and use . with these two propositions established , it is straightforward to extend the rest of our results from square to rectangular matrices . in
particularly , we have that the probability of finding exactly @xmath58 eigenvalues are real is given by @xmath378\det \big [ b_{j , l}^\nu(\zeta ) \big ] _ { j , l=1,\dots , n/2},\ ] ] for @xmath7 even , while the probability of finding @xmath379 real eigenvalues is @xmath380\det \big [ [ b_{j , l}^\nu(\zeta ) ] ^{j=1,\dots,(n+1)/2}_{k=1,\dots,(n-1)/2 } \quad
[ \tilde a_j^\nu ] _ { j=1,\dots,(n+1)/2 } \big],\ ] ] for @xmath7 odd . here , we have defined @xmath381 with @xmath382 given by , @xmath383 and @xmath384 ( @xmath385 ) given by and @xmath386 .
the similarity with theorem [ t1 ] is immediate .
moreover , the local densities at the origin is given by @xmath387 for the complex eigenvalues and @xmath388 for the real eigenvalues .
this generalises and , respectively .
the generalised formulae and follows from the derivations in section [ sec : c - eigenvalues ] and [ sec : r - eigenvalues ] now using the weights and polynomials from proposition [ prop : jpdf - rect ] and [ prop : poly ] . the global densities remains unaltered as long as @xmath389 are kept fixed in the large-@xmath7 limit .
we would like to thank mario kieburg and oleg zaboronski comments on this manuscript .
remark [ mario - remark ] on page was given to us by mario kieburg .
the work of pjf was supported by the australian research council grant dp140102613 , and that of jri by the arc centre of excellence for mathematical and statistical frontiers .
beenakker , j.m .
edge , j.p .
dahlhaus , d.i .
pikulin , shuo mi , and m. wimmer , _ wigner - poisson statistics of topological transitions in a josephson junction _ , phys .
* 111 * ( 2013 ) , 037001 [ arxiv:1305.2924 ] . | random matrices formed from i.i.d .
standard real gaussian entries have the feature that the expected number of real eigenvalues is non - zero .
this property persists for products of such matrices , independently chosen , and moreover it is known that as the number of matrices in the product tends to infinity , the probability that all eigenvalues are real tends to unity .
we quantify the distribution of the number of real eigenvalues for products of finite size real gaussian matrices by giving an explicit pfaffian formula for the probability that there are exactly @xmath0 real eigenvalues as a determinant with entries involving particular meijer @xmath1-functions .
we also compute the explicit form of the pfaffian correlation kernel for the correlation between real eigenvalues , and the correlation between complex eigenvalues .
the simplest example of these the eigenvalue density of the real eigenvalues gives by integration the expected number of real eigenvalues .
our ability to perform these calculations relies on the construction of certain skew - orthogonal polynomials in the complex plane , the computation of which is carried out using their relationship to particular random matrix averages .
_ department of mathematics and statistics , + arc centre of excellence for mathematical and statistical frontiers , + the university of melbourne , victoria 3010 , australia _ |
eighteen years of doppler shift measurements of the primary star in the nearby binary system 55 cancri ( hd 75732 ; hereafter 55 cnc ) have gradually raised the number of orbiting planets from 1 ( butler et al .
1997 ) , to 3 ( marcy et al . 2002 ) , to 4 ( mcarthur et al . 2004 ) , and recently to 5 ( fischer et al . 2008 ) .
although the 4 inner planets are closely packed together ( their semimajor axes @xmath2 au ) , the orbits of all 5 planets are remarkably circular ( see table 1 ) and presumably coplanar .
the system exhibits astonishing regularity , reminiscent only of the planetary order in our solar system and unmatched by any other currently known multiple
planet extrasolar system ( for reviews of orbital and dynamical characteristics of exoplanets , see marcy et al .
[ 2005 ] and butler et al . [ 2006 ] ) .
the most massive planet , 55 cnc d with @xmath3 au and a minimum mass ( mcarthur et al .
2004 ) , which increases all masses by 25% . ] of 3.835 jupiter masses ( table 1 ) , is the most distant gaseous giant known among exoplanets with well defined orbits and , like jupiter , it orbits beyond 5 au from the star .
furthermore , there are no giant planets in the area roughly between 1 and 5 au , just like in our solar system .
in fact , no planets have been found to date in the 5 au gap between the two outermost planets , f and d , with semimajor axes @xmath4 au and @xmath5 au , respectively .
cccccc + + index & planet & orbital & semimajor & minimum & orbital + @xmath6 & designation & period & axis & mass & eccentricity + & & @xmath7 ( days ) & @xmath8 ( au ) & @xmath9 ( @xmath10)@xmath11 & @xmath12 + 1 & e & 2.817 & 0.038 & 0.034 & 0.070 + 2 & b & 14.65 & 0.115 & 0.824 & 0.014 + 3 & c & 44.34 & 0.240 & 0.169 & 0.086 + 4 & f & 260.0 & 0.781 & 0.144 & . . .
@xmath13 + 5 & d & 5218 & 5.77 & 3.835 & 0.025 + + + + despite the similarities in orbital characteristics , the planetary system of 55 cnc exhibits some conspicuous differences when compared to our solar system ( fischer et al .
[ 2008 ] and table 1 ) : ( a ) three of the planets orbit too close to the star ( within 0.24 au ) , that is closer than the orbit of mercury around the sun .
( b ) the second planet , 55 cnc b ( @xmath14 au ) , is also the second largest with a mass of @xmath15 jupiter masses .
this planet also shows the lowest orbital eccentricity ( @xmath16 ) among all 5 planets .
( c ) the 4 inner planets are all quite massive , with the innermost planet , 55 cnc e ( @xmath17 au ) , having the smallest mass ( @xmath18 earth masses ) , a value roughly comparable to the mass of uranus .
most , if not all , of these planets are expected to be gaseous giants in contrast to the terrestrial planets in our solar system .
the central star of 55 cnc and the dynamics of its system have kept observers occupied for other reasons as well : 1 .
this star is the primary in a visual binary ( separation @xmath1910@xmath20 au ) and belongs to a group of @xmath1930 multiple stellar systems in which giant planets are seen orbiting very close to the primary stars ( eggenberger et al .
2004 ; raghavan et al .
in fact , the most massive inner planets tend to be found in multiple stellar systems and those with the shorter periods ( @xmath21 days ) tend to also have very low eccentricities .
the three inner planets of 55 cnc , with orbital periods @xmath22 days and eccentricities @xmath23 , all fit precisely this trend .
2 . the primary star of 55 cnc has very high metallicity compared to the sun ( @xmath24 ; gonzalez & vanture 1998 ; valenti & fischer 2005 ) , a property that has been associated with a higher probability of forming massive protoplanetary cores and gaseous giant planets in the surrounding , also metal rich , disk by the process of core accretion ( ida & lin 2004b ; bodenheimer & pollack 1986 ; pollack et al .
it seemed for a time that an extended disk of dust was visible in the infrared , but this result was negated when additional infrared observations failed to detect a disk ( schneider et al .
2001 ; bryden et al . 2006 ) and submillimeter observations indicated that the excess flux could be attributed to background sources ( jayawardhana et al .
4 . the orbits of the two more massive of the inner planets , 55 cnc b and c , appear to be close to but not quite in mean motion resonance with a period ratio of nearly 3:1 ( marcy et al .
2002 ; fischer et al .
this is the only known 3:1 apparent resonant pair , whereas in all the other known cases the planets appear to have been captured in 2:1 resonances ( butler et al .
2006 ) .
the picture that emerges from the observations of 55 cnc is enigmatic .
the most important questions can be summarized as follows ( fischer et al .
2008 ) : why did this star end up with so many massive planets when 90% of the observed stars do not have any giants at all ?
the similarities to our solar system seem to indicate that multiple
planet systems like 55 cnc ( and @xmath25 ara ara ( hd 160691 , gj 691 ) also has a gaseous giant orbiting beyond 5 au ( @xmath26 au ) and 3 additional giants orbiting close to the star ( santos et al .
2004 ; pepe et al . 2007 ; and table 5 below ) .
their distribution shows a large gap between 1.5 and 5.2 au ; but the orbital eccentricities of the inner planets are comparatively larger ( @xmath27 ) than those in 55 cnc ( see also [ muara ] ) . ] ) should be common , but why have they not been observed in larger numbers ? and if such multi planet systems around metal rich stars represent the new norm of planet evolution ( ida & lin 2004b ) , where are the systems with central stars chemically similar to our sun ?
one would certainly be reluctant to admit that our solar system is special because of the inherent bias associated with such a proposition .
in addition , were the inner planets of 55 cnc ( and @xmath25 ara ) formed `` in situ '' on the observed orbits ( lissauer 1995 ; bodenheimer et al .
2000 ) , or are they the result of inward migration ( ward & hahn 2000 ; lin et al .
2000 ; armitage et al . 2002 ; alibert et al . 2004 , 2005a , 2006 ) that terminated conveniently ( marcy et al .
2000 ; trilling et al . 2002 ; eggenberger et al .
2004 ) to allow for the planets to survive and circularize their orbits ? and why are all the planetary orbits in 55 cnc so nearly circular when large eccentricities are quite common in exoplanets ( marcy et al .
2000 , 2005 ) and it appears that resonant tidal interactions did not take hold since an actual 3:1 commensurability was not established ? finally , could there be smaller planets or an asteroid belt in the 5 au gap and , in any case , why was matter acquisition so disparate as to produce a series of smaller planets surrounded by two very massive planets , 55 cnc b and d ? finding answers to even some of these questions can be a giant step toward understanding planet formation and evolution .
we will revisit the underlying issues in
[ questions ] below .
the observations of 55 cnc have motivated several empirical modeling efforts that attempted to explore the various compelling issues that arise from the impressive regularity of this planetary system , the existence of the 5 au gap , and the unusual chemical properties of the parent star : 1 .
_ metal rich environment for planet formation_. ida & lin ( 2004a , b , 2005 ) carried out monte carlo simulations of protoplanetary core growth , orbital migration , and gas acquisition .
their model predicts that the probability of forming gas giants increases rapidly with increasing metal abundances in the host stars and the surrounding disks , which is consistent with the multitude of planets observed in 55 cnc and @xmath25 ara . on the other hand ,
the physical and orbital properties of these planets are in disagreement with the main features of this model , namely that there should not be planets with @xmath28 earth masses at @xmath29 au and that short period , neptune size planets should not be common around such g - type stars as opposed to m dwarfs .
this led ida & lin ( 2005 ) to speculate that the planets seen in 55 cnc and @xmath25 ara have not migrated to their present locations but they were instead formed in
situ with the help of sweeping mean motion or secular resonances .
however , there is no evidence that gas - giant migration or resonances ever played a role in the formation of these planets , and this adds to the mystery surrounding such well ordered planetary systems . 2 .
_ long term orbital stability_. the stability of the planetary orbits in 55 cnc has been investigated independently by several groups : raymond & barnes ( 2005 ) integrated the orbits of the 3 most massive planets ( b d ) for 100 myr and found that the system could afford to have more planets in stable orbits between planets c and d. fischer et al .
( 2008 ) integrated the orbits of all 5 planets for 1 myr and found that planets b and c are not locked into a 3:1 mean motion resonance .
gayon et al .
( 2008 ) integrated the orbits of all 5 planets for 400 myr and showed that the system represents a case of stable chaos in which , however , close encounters between planets are avoided .
all of these models indicate clearly that the system is dynamically stable over long time scales ( @xmath30 0.1 gyr ) . 3 . _
apparent 3:1 mean
motion resonance_. the apparent 3:1 mean motion resonance between the orbits of planets b and c in 55 cnc was initially hailed as a new example of a low order commensurability in a `` resonant system '' ( barnes & quinn 2004 ) , with 2:1 resonances providing the only previous example in observed systems .
although barnes & quinn ( 2004 ) did not study 55 cnc explicitly , their models showed that planets in resonant systems have very narrow zones of stability and raised the possibility that such systems are only marginally stable .
modeling of 55 cnc was undertaken by zhou et al .
( 2008 ) who showed that planets b and c can be captured in a 3:1 resonance if they start with modest eccentricities ( @xmath31 ) and migrate slowly ( for @xmath190.1 myr ) .
the above models would indicate that the arrangement of the inner planets in 55 cnc is the result of special circumstances , an uncomfortable notion . since we now know that planets b and c are not in resonance ( fischer et al .
2008 ; gayon et al .
2008 ) , a more palatable conjecture is that no special migration has ever occurred in 55 cnc and that its inner planets form a typical `` interacting system '' ( barnes & quinn 2004 ) in which the planets are able to perturb one another and have broad zones of stability available to them .
modeling efforts and related conjectures concerning the apparent 3:1 resonance in 55 cnc demonstrate how difficult it is to simulate such complex systems because we do not have a handle on the initial conditions ; numerical models that start from arbitrary but convenient initial conditions are capable of producing `` easily '' just about all of the observed results .
so it is not surprising that more recent modeling efforts ( although they did not address 55 cnc directly ) now seem to provide evidence against 3:1 resonant captures of planets in general : pierens & nelson ( 2008 ) found that massive planets captured by a dominant , jupiter size planet tend to end up exclusively in 2:1 or 3:2 mean motion resonances , depending on their mass . furthermore , adams et al .
( 2008 ) and lecoanet et al .
( 2008 ) found that turbulence in the host protoplanetary disk can upset mean motion resonances and thus resonant systems should be rarely observed .
_ packed planetary systems hypothesis_. the orbital stability calculations of barnes & quinn ( 2004 ) and related simulations by barnes & raymond ( 2004 ) used massless test particles in models of several known extrasolar multiplanetary systems in order to identify broad regions full of dynamically stable orbits and to predict the possible existence of additional , yet unseen planets .
these calculations were extended by raymond & barnes ( 2005 ) , raymond et al .
( 2006 ) , and barnes et al .
( 2008 ) to include various saturn mass and terrestrial mass particles in between the orbits of the known planets in each simulated system .
although it was easy to find stable orbits for additional planets with terrestrial or smaller masses in some of these systems , the main result of these numerical experiments came as a surprise : many systems are on the edge of stability and small changes to the values of @xmath32 or @xmath33 push them to become unstable .
following the ideas of laskar ( 1996 ) , barnes & raymond ( 2004 ) conjectured that the remaining apparently stable systems are also packed with enough planets to be marginally stable .
this is the packed planetary systems ( pps ) hypothesis and it predicts that stable systems may contain unseen planets that bring them to the edge of stability .
the pps models successfully predicted the existence of hd 74156 d ( barnes et al .
2008 ) before its discovery ( bean et al . 2008 ) and proved to be consistent with the presence of 55 cnc f that was found ( fischer et al .
2008 ) orbiting at the inner edge of the predicted stable zone ( @xmath34 au @xmath35 3.2 au , @xmath36 ; barnes & raymond 2004 ) .
the stable zone in 55 cnc is too wide , as a result the 5planet system is still not packed and there could be more undetected planets between the planets f and d ( raymond et al . 2008 ) . specifically , the pps models suggest the following interesting possibilities : ( a ) a giant planet with up to saturn mass may exist near 2 au ( the center of the stable zone ) with @xmath37 ( raymond & barnes 2005 ) .
( b ) alternatively , several terrestrial planets with up to 0.63 earth masses may exist at 1.13.6 au with @xmath38 ( raymond et al .
although the possibility of having earth like planets in the `` habitable zone '' around 1 au is intriguing , our own interest in this work is focused entirely on the possibility of having a gas giant orbiting at @xmath192 au for the reasons that we explain in [ newmodel ] below .
the current observations of 55 cnc do not rule out such a planet : at a distance of @xmath192 au , a giant planet could avoid detection if its minimum mass is smaller than 100 earth masses ( fischer et al .
2008 ) , i.e. , approximately equal to or less than the mass of saturn . 5 . _ empirical titius bode relation_. poveda & lara ( 2008 ) presented an exponential fit of the semimajor axes of the 5 known planets in 55 cnc .
this empirical relation is akin to the titius
bode ( tb ) rule for the solar system , but it works only if there exists a hypothetical sixth planet at @xmath39 au . based on this result , poveda & lara ( 2008 ) proposed that an unseen planet with @xmath40 au should be present in 55 cnc .
there is no physical basis for this prediction , yet it is quite interesting that the empirical tb relation of 55 cnc is in complete agreement with the prediction of the pps model for a gas giant at the same orbital distance .
since the pps model and the tb relation converge to the same prediction , they can both be tested by additional doppler shift measurements of the primary star of 55 cnc . by extending their baseline
, future observations should be able to search for an additional periodicity in the 5 au gap of 55 cnc and , irrespective of the outcome , the result can be pivotal for planet formation theories ( see also [ newmodel].5 ) .
6 . _ the protoplanetary disk of 55 cnc_. fischer et al .
( 2008 ) made an attempt to estimate some key parameters of the minimum mass protoplanetary disk of 55 cnc . using a typical surface density profile for the disk that varied with distance @xmath0 as @xmath1 and an estimate of the core masses of the 5 observed planets that took into account the excessive metallicity of the system
, they found a total mass of 0.031 @xmath41 out to 6 au and a surface density for the dust of @xmath42 at 5 au .
the dust - to - gas mass ratio was set to the nominal value of 1% and the mass estimate also took into account the intermediate inclination of the orbital plane of the system ( @xmath43 , as derived by mcarthur et al .
[ 2004 ] for 55 cnc d ) .
the assumed @xmath44 profile was first established for the minimum mass solar nebula ( mmsn ) by weidenschilling ( 1977 ) and hayashi ( 1981 ) and it has been commonly used ever since in calculations of this type .
recently , this profile was investigated again in our solar system ( davis 2005 ) , in extrasolar systems ( kuchner 2004 ) , and in dusty circumstellar disks around t tauri stars ( kitamura et al .
2002 ; andrews & williams 2007 ) . due to large systematic uncertainties ,
none of these investigations has been able to directly confirm or reject the original @xmath44 relation , although all of them provide useful hints about possible improvements of the `` typical '' surface density profile of the solar nebula and extrasolar protoplanetary disks . for what follows , it is interesting to note that the the model of davis ( 2005 ) indicates that the original power
law @xmath45 with @xmath46 tends to overestimate the surface density within the central 1 au of the solar nebula .
also , kuchner ( 2004 ) finds a steeper profile with @xmath47 for the surface density of the typical minimum mass `` extrasolar nebula '' ( mmen ) constructed by weidenschilling s ( 1977 ) method from 26 exoplanets found in multi planet systems . the @xmath47 power law index results , again , in an overestimate of the surface density within the central 1 au of the mmen .
the lack of a central core region , such as that suggested for the mmsn ( lissauer 1987 ; davis 2005 ) and for circumstellar disks ( garaud & lin 2007 ) , is the obvious reason for the excessive central densities in models that adopt a single power law density profile for the entire protoplanetary disk .
we are interested in 55 cnc because this is the only extrasolar system with so many planets in well ordered , long lived , circular orbits .
we believe that the high degree of regularity seen in this system makes it an ideal laboratory for exploring the above discussed issues and for testing the leading current hypotheses and models about planet formation in protostellar disks .
our work starts with a new physical model that we formulated rigorously based on exact solutions of the isothermal lane
emden equation with rotation and that relates the observed planets in a well ordered , multi planet system to key physical properties of the protoplanetary disk that hosted their formation .
first we applied this model successfully to the 11 largest planets of our solar system ( christodoulou & kazanas 2008 , hereafter ck ) and we proceed here to apply it to the 5 known planets of 55 cnc .
the results obtained from our modeling are tied to the theoretical models of [ th ] in the following ways : 1 .
our model uses all the planets in regular orbits irrespective of size , mass , and metal abundances . as such
, it works equally well for solar type systems like our solar system and for systems with different metallicities or with gas giants in short period orbits like 55 cnc .
we do expect however that the modeling results will have to be interpreted consistently with the physical and the chemical properties of the detected planets and their parent stars .
it is important that the planetary orbits of 55 cnc were found to be dynamically stable over long time scales because our model associates such long lived orbits with local minima of the gravitational potential well in the midplane of the original protoplanetary disk .
this association does not hold for many odd systems in which one or two planets have migrated inward and are now seen very close to their parent stars .
however , the regular spacing of the planets and the stability of 55 cnc argue against migration as a process that dominated its evolution .
it is now clear that 55 cnc b and c are not caught in a 3:1 mean motion resonance .
this result also suggests that no significant migration has taken place in the disk of 55 cnc and it supports our picture of a well - ordered system in which the planets formed in
situ by core accretion . in turn ,
such regular planetary orbits readily provide information about the locations of density enhancements ( the local potential minima ) in the disk during the planet formation stage .
the pps models predict that there must exist at least one more planet in the 5 au gap of 55 cnc and they will be tested by future observations for the first time in such a well ordered , multi planet , extrasolar system . our model is also not impervious to the possible existence of an additional planet , but it does not necessarily need a 6@xmath48 planet in order to work . as it predicts entirely different physical characteristics for the disk of 55 cnc depending on whether 5 or 6 planets are used ( see [ properties ] ) , our model too needs an observational resolution of this dilemma .
5 . based on the work of ck
, we now understand that the tb relation can be accommodated within our approach , but the rule has no inherent physical significance for our solar system or for 55 cnc .
since the tb relation of 55 cnc necessarily predicts a 6@xmath48 planet at @xmath40 au , the absence of such a planet from the 5 au gap will settle this issue without a doubt .
such an observational result will however be detrimental to the pps hypothesis as well .
alternatively , if an additional planet is found in the 5 au gap , then the doubts will persist , the pps hypothesis will persevere , and our model ( and presumably models such as those of weidenschilling [ 1977 ] and davis [ 2005 ] ) will then show a clear preference for a different radial density profile of the protoplanetary disk of 55 cnc ( see [ properties ] for details ) .
the attempt of fischer et al .
( 2008 ) to estimate some of the properties of the minimum mass protoplanetary nebula ( mmpn ) of 55 cnc is noteworthy by insufficient .
the main drawback of these estimates is that the authors assumed a @xmath49 profile for the disk without actually using weidenschilling s ( 1977 ) method or any of the competing methods cited in
[ th].6 above .
consequently , the particular density profile produced by our disk model of 55 cnc can not be compared to this previous work . some of the parameter values obtained by fischer et al .
( 2008 ) and by kuchner ( 2004 ) , who did use weidenschilling s method , can however be compared to the results of our modeling , and we do so in [ surface ] below . in [ model ] , we describe in detail the new physical model of the midplane of the protoplanetary disk of 55 cnc and we derive the fundamental dynamical parameters of this disk ( minimum mass , density profile , specific angular momentum , rotation frequency , and equation of state ) .
we also examine the dynamical stability of the model and we describe the effects that a possible 6@xmath48 planet will have if it is discovered orbiting at @xmath40 au , as expected by the pps and tb models of the system ( [ th ] ) . in [ discussion ] , we interpret the results from our model and we discuss the issues and the questions mentioned above that pertain to the impressive regular structure of this planetary system and its intriguing 5 au gap .
we also report briefly on our modeling effort of the related multi
planet system @xmath25 ara .
we are interested in the equilibrium structure of a rotating , self gravitating , isothermal disk of angular velocity @xmath50 , mass density @xmath51 , and sound speed @xmath52 . following ck , we adopt cylindrical coordinates ( @xmath53 ) , the assumption of axisymmetry ( @xmath54 ) , and the assumption of cylindrical symmetry ( @xmath55 ) that is valid on the midplane of any disk .
we cast all quantities in dimensionless form as follows : we use the central density @xmath56 and the length scale of the disk @xmath57 where @xmath58 is the gravitational constant , and we define the normalized radius as @xmath59 and the normalized density as @xmath60 .
in addition , we adopt a rotation profile of the form @xmath61 where @xmath62 and the dimensionless function @xmath63 is to be determined .
finally , we use @xmath64 and the jeans frequency @xmath65 to define a dimensionless rotation parameter of the form @xmath66 the equilibrium structure of the midplane of the disk is then described by the isothermal lane emden equation with rotation that takes the dimensionless form @xmath67 in ck , we showed that any power
law density profile is an exact analytic solution of this equation , provided that the rotation profile @xmath63 is also also determined self consistently from eq .
( [ main1 ] ) .
such `` baseline '' power
law profiles are particular solutions of the lane emden equation , but they are unable to satisfy the central boundary conditions , namely that @xmath68 they do however determine analytically the appropriate rotation profile @xmath63 and the mean density profile irrespective of the applicable ( physical ) boundary conditions .
the exact solutions can then be obtained by a simple numerical integration that obeys the boundary conditions ( [ bc ] ) , and they are bound to oscillate permanently about the corresponding baseline solutions without ever settling on to them ( that would be a violation of the boundary conditions ) .
it is important to note that , although any monotonic power law profile can be chosen as a baseline solution and it will produce a monotonic rotation profile , the corresponding exact solution subject to the boundary conditions ( [ bc ] ) will still have to be oscillatory , creating thus a series of local maxima and minima across the entire density profile . for our baseline equilibrium model of the midplane of the mmpn of 55 cnc
, we adopt a composite analytic solution in which the mean density profile is a combination of a flat inner region ( a `` core '' ) followed by a declining power
law : @xmath69 where @xmath70 is the radius of the constant density core region and @xmath71 . compared to the model of the solar nebula in ck , this model does not have a flat outer region because no outer equidistant planets are observed in 55 cnc .
the rotation pofile is determined next by substituting eq .
( [ den ] ) into eq .
( [ main1 ] ) and by solving the resulting differential equation : @xmath72 } \ , \ \ \ { \rm if } \ \ \ x > x_1 \end{array } \right . .
\label{den31}\ ] ] it is easy to show that this rotation law obeys the physical requirements that @xmath73 and @xmath74 everywhere for any choice of @xmath71 . the functions @xmath75 and @xmath63 are plotted in figure [ fig1 ] for @xmath76 and for various choices of the power law index @xmath71 .
we see that , as @xmath77 is increased , the rotation profile is not affected as much as the density profile .
the above composite equilibrium model is characterized by three free parameters : the core radius @xmath78 , the rotation parameter @xmath79 , and the power
law index @xmath71 . the density profile ( eq . [ [ den ] ] ) of this baseline solution of the lane emden equation ( [ main1 ] )
is not capable of satisfying the physical boundary conditions ( [ bc ] ) at the center and it serves only as a mean approximation to the mean density of the corresponding physical model .
the general form of the rotation law of the baseline ( eq . [ [ den31 ] ] ) can , however , be adopted for the differential rotation @xmath63 of the physical model as well .
then eq .
( [ den ] ) provides a prescription for the rhs of the lane
emden equation ( [ main1 ] ) which can thus be written as @xmath80 this differential equation can be integrated numerically subject to the physical boundary conditions ( [ bc ] ) . for all choices of the free parameters \{@xmath81 , @xmath79 , @xmath71 } , the numerical solutions oscillate about the baseline solutions ( [ den ] ) .
the local maxima of the density profile correspond to minima of the gravitational potential well in the midplane of the disk and they are ideal locations for the formation of planetary cores from the dust grains that also have to aggregate inside these wells .
we construct a model of the mmpn of 55 cnc by choosing the 3@xmath82 planet ( @xmath83 au ) from the star to set the physical scale of the model .
( similar results are obtained when any of the inner 4 planetary orbits is used for scaling . )
then the third density peak of the model ( not counting the central peak at @xmath84 ) is associated with a distance of 0.24 au from the center , and the exact solutions of the lane
emden equation ( [ main2 ] ) are optimized by varying the free parameters until the remaining density enhancements match as closely as possible the observed semimajor axes of the remaining planetary orbits of 55 cnc .
fit model is shown in figure [ fig2 ] and its density peaks are listed in table 2 along with the observed semimajor axes given by fischer et al .
the third density peak occurs at @xmath85 in the best
fit model of 55 cnc , implying that the length scale of the protoplanetary disk in the isothermal phase of its evolution was extremely small ( @xmath86 au ) .
this value is @xmath1930 times smaller than the length scale of our solar system ( @xmath87 au ; ck ) and indicates that the inner disk of 55 cnc was denser than the solar nebula ( see [ properties ] below ) .
furthermore , there is no other density peak out to 100 au , so this model predicts no more planets beyond the orbit of 55 cnc d. cccccc + + + index & planet & semimajor & peak & relative & peak + @xmath6 & designation & axis & location & deviation & density + & & @xmath8 ( au ) & @xmath88 ( au ) & ( % ) & @xmath89 + 1 & e & 0.038 & 0.042 & 10.5 & @xmath90 + 2 & b & 0.115 & 0.091 & @xmath91 & @xmath92 + 3 & c & 0.240 & 0.240 & . . .
& @xmath93 + 4 & f & 0.781 & 0.862 & 10.4 & @xmath94 + 5 & d & 5.77 & 5.61 & @xmath95 & @xmath96 + the relative deviations between the locations of the model peaks and the planetary orbits are also listed in table 2 . the largest relative deviation ( @xmath1921% )
is observed for the second peak that corresponds to the orbit of planet b. this difference is still quite small ( @xmath190.02 au ) in absolute terms but it appears disproportionately large because our method of optimization is skewed toward the inner planets for which it produces larger relative deviations ( see ck for more details ) .
the values listed in table 2 can not be improved individually because of the extreme nonlinear nature of the problem .
only the value of @xmath97 can be optimized independently of the other parameters by minimizing the deviation of the innermost planet from the corresponding density peak .
the strong coupling of the remaining two parameters then demonstrates that fitting 5 planetary orbits with 3 free parameters is not a simple matter , as nonlinearities do not generally allow for one - to - one relations between parameters and planetary positions .
the parameters of the best
fit model determined by the optimizing algorithm along with the scaling assumption that @xmath98 au ( see table 2 ) constitute a set of important dynamical parameters of the mmpn of 55 cnc : @xmath99 the value of the power
law index @xmath100 indicates that the density profile of the differentially rotating region of the disk declined with radius , on average , as @xmath101 .
this region extended outward beyond the core radius of @xmath102 au .
the low value of @xmath97 indicates that the rotation of the isothermal nebula was sufficiently slow to avoid nonaxisymmetric instabilities ( see [ stability ] below ) .
the unusually small value of @xmath103 , along with eq .
( [ length ] ) and a typical low gas temperature , indicates that the nebula was very dense , about @xmath1910@xmath20 times denser than the solar nebula ( see table 3 below ) . so the picture that emerges from the best
fit model of 55 cnc is that of a relatively heavy and slowly rotating disk that is long lived and that can comfortably build massive planets over time by accumulating the solids and then accreting the gas contained within each local potential well .
the optimization procedure also finds some models of high quality in which somewhat larger magnitudes of @xmath97 are compensated by smaller values of @xmath70 .
the most different model gives us an idea about how shallow the area around the true minimum is in the three
parameter space of the model .
this model is characterized by the following values : @xmath104 , @xmath105 , @xmath106 au , @xmath107 , @xmath108 au , and the relative deviations are smaller than 22% .
however , the power law index does not differ from @xmath109 by more than 1.2% in any of the high quality models .
similarly , the variations of @xmath103 and @xmath70 always produce core radii to within 1% of @xmath1100.052 au in all of these models , so there is no doubt that the core region of the mmpn of 55 cnc was extremely small ( @xmath1910 solar radii ) .
emden equation that we solved can serve as a model of a differentially rotating disk supported by thermal pressure in the @xmath111direction , so we expect that the scale height from pressure support will be @xmath112 down the radial density gradient . ] where @xmath113 is the rotation speed . in our models
, @xmath113 becomes asymptotically flat when the density exhibits a power
law profile with @xmath71 and @xmath52 is constant , leading to the approximate relation that @xmath112 . ] in this case , the value of @xmath100 implies that the corresponding power law index in the surface density profile ( @xmath114 ) of the nebular disk was @xmath115 .
this value is virtually identical to that obtained for the solar nebula by weidenschilling ( 1977 ) who used an entirely different method and by ck who used the same optimization technique as in this work .
the conformance of these results does not hold if a 6@xmath116 planet is assumed to exist in the 5 au gap of 55 cnc .
we have also optimized models that included a hypothetical planet at @xmath117 au ( as predicted by the pps and tb models of 55 cnc ) , and the best
fit model of the six planetary orbits has the following parameter values : @xmath118 , @xmath119 , @xmath120 au , @xmath121 , @xmath122 au , and the relative deviations are smaller than 19% .
these values are comparable to those in the best
fit model of the 5 observed planets , except for @xmath77 whose smaller value produces a softer density gradient to accommodate the hypothetical planet at 2 au .
this softer gradient can also accommodate one additional planet farther out , but the outermost density peak occurs at 20.56 au which differs from @xmath123 au predicted for a hypothetical 7@xmath116 planet by the empirical tb relation of poveda & lara ( 2008 ) .
this peak is also not consistent with the recent pps prediction ( raymond et al .
2008 ) that an unseen outer planet could exist close to 9 or 10 au , i.e. near the inner edge of a wide zone of stability found beyond 55 cnc d. thus , our model with hypothetical planets is still markedly different than the pps and tb models of 55 cnc .
our modeling also delineates some of the fundamental physical characteristics behind the mean density profile of the midplane of the disk of 55 cnc . with the aid of our best
fit model , we can deduce substantial information concerning the structure and the dynamics of the nebular disk .
in addition to the structural and rotational parameters discussed above , we can use our analytic baseline model in order to probe the dynamical state of the protoplanetary disk of 55 cnc in its isothermal phase .
following ck , we calculate the relevant physical quantities for the mmpn of 55 cnc and we summarize our results in table 3 .
ccccc + + description & quantity & & 55 cnc value @xmath11 & solar system value @xmath11 + equation of state & @xmath124 & @xmath125 & @xmath126 & @xmath127 + & @xmath128 & @xmath125 & @xmath129 & @xmath130 + + central density @xmath13 & @xmath131 & @xmath125 & @xmath132 & @xmath133 + jeans frequency & @xmath134 & @xmath125 & @xmath135 & @xmath136 + rotation frequency & @xmath64 & @xmath125 & @xmath137 & @xmath138 + rotation period & @xmath139 & @xmath125 & 1.4 yr & 12 yr + + + + surface density & @xmath140 & @xmath125 & @xmath141 & @xmath142 + mass & @xmath143 & @xmath125 & @xmath144 & @xmath145 + specific angular momentum & @xmath146 & @xmath125 & @xmath147 & @xmath148 + + + + mass & m & @xmath125 & @xmath149 & @xmath150 + + + + + the corresponding results obtained by ck for the solar nebula are also listed in table 3 for comparison purposes .
compared to the core of the solar nebula , the mmpn core of 55 cnc was smaller ( @xmath151 ratio @xmath1916 ) , denser ( @xmath131 ratio @xmath1910@xmath20 ) , heavier ( @xmath152 ratio @xmath1930 ) , and it was rotating @xmath1910 times faster .
the disk contained @xmath19100 times less mass ) and @xmath103 is extremely small in both cases .
the @xmath111integration needs to be extended for about 1 au away from the midplane in order to produce masses @xmath190.01 @xmath41 . ] within a 6-au radius and only 2/5 of the angular momentum per unit mass .
there is no doubt that these two disks were different and their differences are reflected in the distributions and the physical properties of the various planets that they formed over time . using the @xmath153 value of 55 cnc from table 3 , the core radius @xmath154 au , and the power
law index @xmath155 , we can write the surface density profile of the mmpn of 55 cnc for @xmath156 in the form @xmath157 where @xmath0 is measured in au .
this equation shows that the surface density of the gas is @xmath158 at @xmath159 au and @xmath160 at @xmath161 au .
these values are 48% of the corresponding values in the solar nebula .
the 1-au value is in agreement with kuchner s ( 2004 ) result of @xmath162 , although he used just the 3 planets b
d , the only ones known at the time . in particular , planet f was not included in the model and this is the reason that the 1-au value was underestimated and the density profile turned out to be too steep with @xmath163 . in general , undetected planets cause the main problem in this method , and if such planets are included , they tend to soften the density gradients and push kuchner s average mmen value of @xmath164 closer toward @xmath165 .
the 5-au value corresponds to a typical surface density of solids of @xmath166 for a nominal value of 1% for the dust - to - gas mass ratio .
this value is comparable to the low
end values calculated by fischer et al .
( 2008 ) using only the planetary masses inferred from observations . some additional estimates by the same authors reach as high as @xmath42 .
such estimates rely on uncertain assumptions about the masses of the solid planetary cores and they are obviously too high , as they predict much larger densities in the disk of 55 cnc compared to those in the solar nebula . but we believe that the opposite is true since our model and kuchner s ( 2004 ) model , following entirely different methodologies , both predict significantly lower densities at 5 au for the mmpn of 55 cnc than the mmsn .
we note that such low densities in the outer disk of 55 cnc do not pose a problem for the formation of the massive planet d. as fig .
[ fig2 ] shows , the half width of the corresponding density peak is about 3 au , and all the solids within this emormous potential well were bound to aggregate at the bottom and form the massive core of this planet .
all composite power
law models with @xmath71 described above are stable to axisymmetric perturbations because they satisfy the rayleigh criterion : the specific angular momentum @xmath167 is an increasing function of @xmath0 at all radii . this can be shown easily by using eq .
( [ den31 ] ) to provide the function @xmath168 and then by proving that @xmath169 for all @xmath170 and @xmath171 .
the above models are also stable to nonaxisymmetric perturbations because they do not rotate too fast . for sufficiently slow rotation ,
the nonaxisymmetric modes are neutral and incapable of merging to produce instability .
this behavior is captured in the @xmath172parameter stability criterion ( christodoulou et al .
1995 ) which for gaseous disk models can be written as @xmath173 in our notation , this is equivalent to requiring that @xmath174 for stability , thus our best
fit model with @xmath175 is far from the threshold of nonaxisymmetric instability .
a disk of dust would also be stable with this rotation law .
for an unisotropic disk of particles , the corresponding stability threshold switches from 0.35 to 0.25 , and the requirement for stability then becomes @xmath176 .
in [ intro ] , we placed in proper context all the available information about the multi planetary system of 55 cnc .
our knowledge of this system comes from 18 years of doppler shift observations of the central star ( fischer et al . 2008 ; mcarthur et al . 2004 ; marcy et al . 2002 ; butler et al .
1997 ) and from theoretical modeling of the inferred planetary orbits ( fischer et al . 2008 ; raymond et al . 2008 ; poveda & lara 2008 ; gayon et al .
2008 ; zhou et al . 2008 ; raymond & barnes 2005 ; ida & lin 2005 ; kuchner 2004 ) .
the metal rich primary star of 55 cnc has 5 massive planets in remarkably circular orbits and , outside of our sun , this is the only star with so many planets in well ordered , perfectly stable , nonresonant orbits ( table 1 ) .
the system does however exhibit a large 5 au gap between its two outermost planets and this vast empty region has sparked speculation that one or more additional planets may still remain undetected between planets f ( @xmath4 au ) and d ( @xmath5 au ) .
our contribution to the investigation of 55 cnc is a new physical model of the midplane of the protoplanetary disk in which the planets were formed .
this model was introduced in [ newmodel ] and it was described in detail in [ model ] .
it is based on exact solutions of the isothermal lane
emden equation with rotation ( eq . [ [ main1 ] ] ) subject to the appropriate central boundary conditions ( eq . [
[ bc ] ] ) .
these solutions describe radial density profiles that are oscillatory by nature , despite the fact that the corresponding rotation profiles are strictly monotonic and well - behaved .
the oscillations occur about an average ( baseline ) density ( eq . [ [ den ] ] ) that is a combination of two power laws , a flat uniform core and a decreasing outer section ( [ baseline ] ) .
such baseline density profiles are also exact intrinsic solutions of the lane - emden equation but they are incapable of satisfying the central boundary conditions ( see ck for more details and for an application of the model to our solar system )
. the baselines do however obey the same rotation law ( eq .
[ [ den31 ] ] ) as the corresponding oscillatory solutions .
some typical baselines and their rotation profiles are shown in fig .
[ fig1 ] , while a typical oscillatory solution and its baseline are shown in fig .
[ fig2 ] . the oscillatory density profile shown in fig .
[ fig2 ] represents the best
fit model for the midplane of the protoplanetary disk of 55 cnc .
the fit was obtained by nonlinear unconstrained optimization that matches the orbital semimajor axes of the observed planets to consecutive density maxima ( or gravitational potential minima ) of the oscillatory profile ( see [ exact ] and table 2 ) and that uses the semimajor axis of the third planet from the star to set the physical scale of the model .
the dynamical parameters of the best
fit model are listed in eq .
( [ neb ] ) and its physical properties are summarized in [ properties][surface ] and in table 3 .
the midplane of the disk of 55 cnc was composed of a very small ( @xmath190.05 au ) and very dense ( @xmath1910@xmath177 ) uniform core in slow rotation ( period @xmath191 yr ) , followed by a power
law gradient such that the surface density varied with distance as @xmath178 ( eq . [ [ sigma ] ] ) out to @xmath196 au .
this configuration is stable to both axisymmetric and nonaxisymmetric perturbations ( [ stability ] ) .
the @xmath178 dependence of the surface density profile of 55 cnc is effectively identical to that found by ck for the solar nebula . despite this resemblance
, the detailed comparison between the two nebular disks shown in table 3 indicates that the two models are generally dissimilar , and the differences seen in table 3 are also reflected in the differing characteristics of the planetary systems that were created around these two stars .
specifically , the inner disk of 55 cnc was denser and heavier , and this explains why 4 massive planets were formed within the inner 0.8 au .
these high densities are also responsible for the large growth of the second planet from the star , 55 cnc b , which is the most massive of the inner 4 planets .
this planet was free to grow in a high density environment , while the area around the orbit of the first planet could have been depleted by the protostar ( just as mercury s growth was severely limited by the protosun ) .
on the other hand , the same disk showed much smaller densities beyond 2 au , and this explains why only one massive planet managed to form at the bottom of the single , very wide potential well outside of 5 au .
ccc + physical & large light disks & small heavy disks + property & ( solar system ) @xmath179 & ( 55 cancri a ) @xmath180 + & & + disk & & + & & + metallicity ( @xmath181 ) & @xmath182 & @xmath183 + radial scale ( @xmath103 ) & 0.02 au & 0.001 au + radial size & 4070 au & 610 au + core size ( @xmath151 ) & 1 au & 0.05 au + core density ( @xmath131 ) & @xmath184 & @xmath185 + & & + rotation & & + & & + @xmath97 & 0.4 & 0.1 + rotation period ( @xmath139 ) & 12 yr & 1 yr + jeans period ( @xmath186 ) & 5 yr & 0.1 yr + & & + gas surface density & & + & & + @xmath153 & 2500 & 75000 + @xmath187 & 2000 & 1000 + @xmath188 & 200 & 100 + @xmath189 , @xmath190 & @xmath165 & @xmath165 + + + + + the planetary system of 55 cnc and the solar system are both remarkably well ordered despite the differences in structure and dynamical characteristics and the differences between their host stars .
this comparison seems to indicate that there exist two modes of planet formation that lead to stable , well organized , multi planet systems . based on our modeling and on the results of ck
, we can describe these two extremes of multiple planet formation in the isothermal regime of protostellar collapse as follows ( see also table 4 ) : 1 .
_ large light disks._solar type protoplanetary disks with @xmath191 au cores and moderate rotational support ( no more than @xmath1940%50% of maximum rotation ) that have core densities a few times above those required to form planets ( @xmath1910@xmath191 ; lissauer 1993 ) are capable of forming rocky planets in the first 2 au and , due to the comparatively high densities in their outer regions , several gas giants beyond 5 au as well .
such disks can extend for at least 4070 au in the radial direction . because their core densities are barely supercritical
, these disks can form only terrestrial planets in the vicinity of the core radius .
2 . _ small heavy disks._metal rich protoplanetary disks with @xmath190.05 au cores and little rotational support ( no more than @xmath1910%15% of maximum rotation ) that have core densities @xmath191000 times above the critical value are capable of forming giant planets in the first 2 au and , due to the substantial density drop in their outer regions , only one gas giant beyond 5 au .
such disks can not extend for more than 610 au ( and if they do , their densities are too low to form giant planets ) . because their core densities are strongly supercritical , these disks can easily form gas giants around 1 au , so terrestrial planets are not expected to be present in the habitable zone in this mode .
returning now to the questions posed in
[ obs ] , we can outline some feasible answers suggested by the above results : most of the observed stars ( @xmath1990% ) do not show any giant planets at all . some of them , especially those of solar type , may have planetary systems formed in large light disks ( llds ) , in which case their planets are not observable yet with the current resolution because the inner planets have very low masses and the outer planets are too distant from their stars .
this is probably the reason why the solar system is currently the only example of a system formed in an lld . on the other hand ,
the inner giant planets formed in small heavy disks ( shds ) can be presently detected .
besides the standard example of 55 cnc , all other planets found within 23 au from their stars in nearly circular orbits are good candidates for the shd mode .
this also includes @xmath25 ara , hd 37124 , and the massive , short period ( @xmath21 days ) planets with low eccentricities found in multiple stellar systems ( see [ obs ] ) . on the question of in situ formation versus migration , the slow rotation found in our models ( table 4 ) and the structure of the disks that show several local potential minima strongly support the in situ formation of planets by planitesimal aggregation and gas accretion inside these potential wells .
this does not mean that planetary migration can not occur at all , it just indicates that the multiple planets of well ordered systems do not migrate .
migration may still be the dominant mechanism in systems with planets in highly eccentric and/or irregularly spaced orbits .
but even in such cases , our modeling adds a new dimension to the migration problem that has not been previously explored : the ability of a planet to migrate can no longer be determined solely by the disk torques that redistribute the angular momentum of a planetary orbit . that was done in the past ( e.g. , lin et al .
[ 2000 ] , ward & hahn [ 2000 ] , and references therein ) while the density profile of the underlying disk was assumed to be monotonically decreasing without any localized potential wells . in the present
setting in which the planets are trapped inside gravitational potential minima , the energy content of each planet needs to be considered as well .
a planet orbiting inside a local gravitational potential well can migrate only if the interaction with the disk provides enough energy for it to leap out of its well otherwise the torqing of the planetary orbit will simply lead to stable oscillations and the development of some eccentricity insde this potential minimum . on the question of the existence of additional planets in 55 cnc
, our model boldly predicts that no other planet will be found in the future .
in particular , we do not believe that there is room in the 5 au gap even for a small rocky terrestrial planet or an asteroid belt because the potential minima of planets f and d reach into the gap ( see fig .
[ fig2 ] ) , and all debris must have been swept by these two massive planets .
this prediction is in contrast to the predictions of the pps models ( raymond et al .
2008 ) and the tb conjecture ( poveda & lara 2008 ) for the system , and the disagreement can be resolved by future observations of 55 cnc ( see also [ properties ] above ) .
the question of the impressive growth of inner planet 55 cnc b ( @xmath14 au , mass @xmath192 jupiter masses ) was discussed briefly in [ modes ] above .
the density of the peak where this planet was formed ( table 2 ) is quite high ( @xmath193 ) .
only planet e was formed in a higher density environment in the shd of 55 cnc , but this planet resides deep inside the core of the nebula and its gaseous envelope could have easily been ablated by the protostar at a later time .
the situation is similar in the inner solar system , although venus is slightly less massive than the earth .
the emerging paradigm , valid for both llds and shds , is that of a depleted innermost planet followed by planets that are free to grow as much as the local conditions in the disk would allow .
this picture appears to be borne out also in another shd candidate , @xmath25 ara , but with some caveats that we discuss below .
recent observations ( pepe et al . [ 2007 ] and table 5 ) of the metal rich star @xmath25 ara ( @xmath24 ; santos et al . 2004 ) indicate 4 orbiting planets in a configuration very similar to that of 55 cnc : a single massive outer planet ,
@xmath25 ara e , is separated from the 3 inner planets ( c , d , and b ) by a large empty region between 1.5 and 5.2 au .
the minimum mass of the innermost planet is comparable to that of the innermost planet in 55 cnc , so this planet is also a low mass ( depleted ) giant while its two neighbors have minimum masses comparable to the mass of jupiter .
cccccc + + index & planet & orbital & semimajor & minimum & orbital + @xmath6 & designation & period & axis & mass & eccentricity + & & @xmath7 ( days ) & @xmath8 ( au ) & @xmath9 ( @xmath10)@xmath11 & @xmath12 + 1 & c & 9.6386 & 0.091 & 0.03321 & 0.172 + 2 & d & 310.55 & 0.921 & 0.5219 & 0.0666 + 3 & b & 643.25 & 1.497 & 1.676 & 0.128 + 4 & e & 4205.8 & 5.235 & 1.814 & 0.0985 + + + although the eccentricities of its planetary orbits are larger than those of 55 cnc ( see tables 1 and 5 ) , @xmath25 ara appears to be a good example of an shd system .
the size of the system is also comparable to that of 55 cnc ( @xmath196 au ) , so we expected our best
fit model to produce parameter values similar to those of 55 cnc .
we also thought that , by analogy , a pps model of the system would find a stable zone in the 3.7 au gap between planets b and e , and a tb model would predict a planet in the same gap .
such models are not yet available , but the tb relation of poveda & lara ( 2008 ) can be easily applied to the observed semimajor axes .
it turns out that the tb models are not particularly good fits to the data with or without a hypothetical 5@xmath116 planet in the 3.7 au gap .
the planetary orbits of @xmath25 ara shown in table 5 can not be matched to the density maxima in any of our models .
the reason for this failure is that planet d is too close to planet b , and this leaves an inconspicuous 0.8 au gap in the distribution between planets c and d. a hypothetical planet is needed in this gap to restore the regular spacing of the inner planets .
this , of course , does not mean that an undetected planet necessarily exists in the 0.8 au gap , it just implies that a density maximum is needed in that area of the model for the distribution to be typical of a well organized shd system . by the same token , some models allow for yet another density maximum interior to the orbit of planet c , at @xmath190.05 au . such experimental models are characterized by different parameter sets , so a `` best fit model '' can not be determined , unless additional planets are discovered in @xmath25 ara by future observations .
a surface density profile of the form @xmath194 is applicable outside the central core region of the protoplanetary disk of 55 cnc and the solar nebula ( table 4 ) .
the presence of a roughly uniform core is strongly suggested by both theory and observations ( see [ th].6 ) , and our models suggest that llds and shds can be distinguished by their core sizes ( @xmath191 au and @xmath190.05 au , respectively ) . on the other hand ,
we do not know why both of the nebular disks in table 3 show the @xmath178 trend , although this fact that we independently obtained the same result ( as opposed to two different power laws ) from two different planetary systems argues in favor of our modeling assumptions . at this point
, we do not have any more evidence of the ubiquity of the @xmath178 profile among llds and shds , but we can certainly argue that this density power law is more relevant to planet formation than the profiles from steady state viscous and irradiated accretion disks ( kitamura et al . [
2002 ] , andrews & williams [ 2007 ] , garaud & lin [ 2007 ] , and references therein ) .
the isothermal phase of protostellar collapse that our models address is an extended intermediate phase of the whole process during which the magnetic fields have diffused away , the central core mass is a minute fraction of the solar mass , matter continues being deposited onto the disk from above and below , most of the gas in the disk is molecular , and all the energy produced is efficiently radiated away ( see tohline [ 2002 ] for a recent review ) . at this stage ,
the evolution of the gas is not driven by viscosity or by ionization because the gas is transparent to radiation ; the gas responds fully to the effects of self gravity and ( non keplerian ) rotation , and this is why it is very encouraging to know that our modeled disks are dynamically stable ( [ stability ] )
. theoretical calculations of viscous , magnetized , or irradiated accretion disks are not applicable in this regime of protostellar collapse , while observations of thermal radiation from t tauri stars probe the disks long after the isothermal phase is over , the gas has become opaque to its own radiation , and the grown protostars have imposed nearly keplerian rotation profiles onto the disks . therefore , there is no reason to expect that the surface density profiles should have power
law indices of @xmath195 or softer ( seen in t tauri disks ; kitamura et al .
2002 ; andrews & williams 2007 ) during the isothermal phase of collapse ; and if the planetary results of kuchner ( 2004 ) give any indication about extrasolar systems , we should expect instead @xmath196indices between @xmath165 and @xmath197 .
adams , f. c. , laughlin , g. , & bloch , a. m. 2008 , , 683 , 1117 alibert , y. , baraffe , i. , benz , w. , et al .
2006 , a&a , 455 , l25 alibert , y. , mordasini , c. , & benz , w. 2004 , a&a , 417 , l25 alibert , y. , mordasini , c. , benz , w. , & winisdoerffer , c. 2005a , a&a , 434 , 343 alibert , y. , mousis , o. , mordasini , c. , & benz , w. 2005b , apj , 626 , l57 andrews , s. m. , & williams , j. p. 2007
, , 659 , 705 armitage , p. j. 2003 , , 582 , l47 armitage , p. j. , livio , m. , lubow , s. h. , & pringle , j. e. 2002 , mnras , 334 , 248 barnes , r. , gozdiewski , k. , & raymond , s. n. 2008 , , 680 , l57 barnes , r. , & quinn , t. 2004 , , 611 , 494 barnes , r. , & raymond , s. n. 2004 , , 617 , 569 bean , j. l. , mcarthur , b. e. , benedict , g. f. & armstrong , a. 2008 , , 672 , 1202 bodenheimer , p. , hubickyj , o. , & lissauer , j. j. 2000 , icarus , 143 , 2 bodenheimer , p. , & pollack , j. p. 1986 , icarus , 67 , 391 bryden , g. , beichman , c. a. , trilling , d. e. , et al .
2006 , , 636 , 1098 butler , r. p. , marcy , g. w. , williams , e. , hauser , h. , & shirts , p. 1997
, , 474 , l115 butler , r. p. , wright , j. t. , marcy , g. w. , et al .
2006 , , 646 , 505 christodoulou , d. m. , & kazanas , d. 2008 , , submitted ( astro - ph/0706.3205 ) ( ck ) christodoulou , d. m. , shlosman , i. , & tohline , j. e. 1995 , , 443 , 551 davis , s. s. 2005 , , 627 , l153 desort , m. , lagrange , a .- m . , galland , f. , et al .
2008 , a&a , preprint ( astro - ph/0809.3862 ) eggenberger , a. , udry , s. , & mayor , m. 2004 , a&a , 417 , 353 fischer , d. a. , marcy , g. w. , butler , r. p. , et al .
2008 , , 675 , 790 garaud , p. , & lin , d. n. c. 2007 , , 654 , 606 gayon , j. , marzari , f. , & scholl , h. 2008 , , 389 , l1 gonzalez , g. , & vanture , a. d. 1998 , a&a , 339 , l29 hayashi , c. 1981 , prog .
suppl . , 70 , 35 ida , s. , & lin , d. n. c. 2004a , apj , 604 , 388 ida , s. , & lin , d. n. c. 2004b , apj , 616 , 567 ida , s. , & lin , d. n. c. 2005 , apj , 626 , 1045 jayawardhana , r. , holland , w. s. , kalas , p. , et al .
2002 , , 570 , l93 kitamura , y. , momose , m. , yokogawa , s. , et al . 2002 , , 581 , 357 kuchner , m. j. 2004 , apj , 612 , 1147 laskar , j. 1996 , celest .
astron . , 64 , 115 lecoanet , d. , adams , f. c. , & bloch , a. m. 2008 , , preprint ( astro - ph/0810.4076 ) lin , d. n. c. , papaloizou , j. c. b. , terquem , c. , bryden , g. , & ida , s. 2000 , protostars and planets iv , ed . v. mannings , a. p. boss , & s. s. russell ( univ . of arizona press : tucson ) ,
1111 lissauer , j. j. 1987 , icarus , 69 , 249 lissauer , j. j. 1993 , , 31 , 129 lissauer , j. j. 1995 , icarus , 114 , 217 marcy , g. w. , butler , r. p. , fischer , d. a. , et al .
2002 , , 581 , 1375 marcy , g. , butler , r. p. , fischer , d. , et al .
2005 , prog .
suppl . , 158 , 24 marcy , g. w. , cochran , w. d. , & mayor , m. 2000 , protostars and planets iv , ed . v. mannings , a. p. boss , & s. s. russell ( univ . of arizona press : tucson ) , 1285 mcarthur , b. e. , endl , m. , cochran , w. d. , et al .
2004 , , 614 , l81 pepe , f. , correia , a. c. m. , mayor , m. 2007 , a&a , 462 , 769 pierens , a. , & nelson , r. p. 2008 , a&a , 482 , 333 pollack , j. b. , hubickyj , o. , bodenheimer , p. , et al .
1996 , icarus , 124 , 62 poveda , a. , & lara , p. 2008 , rmxaa , 44 , 243 raghavan , d. , henry , t. j. , mason , b. d. , et al .
2006 , apj , 646 , 523 raymond , s. n. , & barnes , r. 2005 , , 619 , 549 raymond , s. n. , barnes , r. , & gorelick , n. 2008 , , , preprint ( astro - ph/0808.3295 ) raymond , s. n. , barnes , r. , & kaib , n. a. 2006 , , 644 , 1223 santos , n. c. , bouchy , f. , mayor , m. , et al .
2004 , a&a , 426 , l19 schneider , g. becklin , e. e. , smith , b. a. , et al .
2001 , , 121 , 525 tohline , j. e. 2002 , , 40 , 349 trilling , d. e. , lunine , j. i. , & benz , w. 2002 , a&a , 394 , 241 valenti , j. a. , & fischer , d. a. 2005 , apjs , 159 , 141 vogt , s. s. , butler , r. p. , marcy , g. w. , et al .
2005 , , 632 , 638 ward , w. r. , & hahn , j. m. 2000 , protostars and planets iv , ed . v. mannings , a. p. boss , & s. s. russell ( univ . of arizona press : tucson ) , 1135 weidenschilling , s. j. 1977 , a&ss , 51 , 153 zhou , l .- y . , ferraz - mello , s. , & sun , y .- s .
2008 , exoplanets : detection , formation and dynamics , proc .
iau symposium no .
sun , s. ferraz - mello , & j .-
zhou ( cambridge univ . press : cambridge ) , 119 0.5truein | five planets are presently believed to orbit the primary star of 55 cnc , but there exists a large 5 au gap in their distribution between the two outermost planets .
this gap has attracted considerable interest because it may contain one or more lower mass planets whose existence is not contradicted by long - term orbit stability analyses , in fact it is expected according to the `` packed planetary systems '' hypothesis and an empirical titius
bode relation recently proposed for 55 cnc .
furthermore , the second largest planet is just the second farthest and it orbits very close to the star .
its orbit , the most circular of all , appears to be nearly but not quite commensurable with the orbit of the third planet , casting doubt that any migration or resonant capture of the inner planets has ever occurred and lending support to the idea of `` in situ '' giant planet formation by the process of core accretion .
all of the above ideas will be tested in the coming years in this natural laboratory as more observations will become available .
this opportunity presents itself in conjunction with a physical model that relates the orbits of the observed planets to the structure of the original protoplanetary disk that harbored their formation at the early stages of protostellar collapse . using only the 5 observed planets of 55 cnc
, this model predicts that the surface density profile of its protoplanetary disk varied with distance @xmath0 precisely as @xmath1 , as was also found for the minimum mass solar nebula . despite this similarity ,
the disk of 55 cnc was smaller , heavier , and less rotationally supported than the solar nebula , so this system represents a different mode of multi planet formation compared to our own solar system . |
frustrated ( zigzag ) spin-@xmath14 chains with competing nearest - neighbor ferromagnetic ( @xmath10 ) and next - nearest - neighbor antiferromagnetic ( @xmath11 ) couplings reveal rich physics at low temperatures and in applied magnetic fields .
when the chains are coupled in three dimensions , helical order arises in zero field for while the helical order itself gives rise to a very unusual phenomenon of magnetic - field - induced ferroelectricity @xcite , further interesting effects occur when stronger magnetic fields break this order down .
licuvo@xmath1 , one of the best studied frustrated - chain materials , undergoes a first - order transition around 8.5 t from the helically - ordered phase toward a spin - density - wave ( sdw ) phase , where magnetic moments align with the field , and the length of the moment is modulated @xcite .
detailed nature of this phase is , however , debated @xcite , along with the putative nematic phase appearing around 40 t right before saturation @xcite .
additionally , different types of multipolar order are expected for the @xmath8 chains in the applied magnetic field @xcite .
@xmath0-tevo@xmath1 is a candidate frustrated - chain material with spin-@xmath14 ( fig .
[ fig : structure ] ) . its magnetic behavior was initially described within the model of a uniform spin-@xmath14 chain @xcite , although the presence of three low - temperature transitions at @xmath15k , @xmath16k , and @xmath17k in an applied magnetic field as weak as 0.02 t clearly indicates a more complex interaction topology .
recently , sal and radtke @xcite performed a microscopic analysis of isotropic exchange couplings and concluded that @xmath0-tevo@xmath1 is a good realization of the @xmath8 chain model with ferromagnetic ( fm ) @xmath10 and antiferromagnetic ( afm ) @xmath11 .
subsequently , magnetic susceptibility of @xmath0-tevo@xmath1 was re - analyzed in the framework of the @xmath8 model @xcite .
et al . _
@xcite reported non - monotonic evolution of phonon frequencies and speculated on possible structural changes around 150k and even on a change in the orbital state of v@xmath13 at low temperatures .
finally , pregelj _ et al . _
@xcite performed detailed neutron - scattering experiments in zero field and observed a helical magnetic structure with the propagation vector @xmath18 below @xmath7 . between @xmath7 and @xmath6
, @xmath0-tevo@xmath1 reveals an enigmatic stripe - like spin texture , whereas between @xmath6 and @xmath5 a spin - density - wave ( sdw ) phase akin to the field - induced sdw phase in licuvo@xmath1 @xcite has been proposed . the crystal structure of @xmath0-tevo@xmath1 @xcite features chains of vo@xmath19 polyhedra ( fig .
[ fig : structure ] ) .
these chains are directed along the crystallographic @xmath4-axis and linked via asymmetric teo@xmath1 units .
the overall structure is centrosymmetric ( space group @xmath20 ) , but inversion centers are located between the chains , so that dzyaloshinsky - moriya ( dm ) anisotropy terms are allowed for both @xmath10 and @xmath11 , which is different from any of the frustrated - chain compounds previously reported .
magnetic anisotropy can have strong effect on field - induced phase transitions and on new phases induced by the magnetic field .
for example ,
linarite pbcu(oh)@xmath21so@xmath1 , which symmetry is lower than in licuvo@xmath1 @xcite , shows very complex and still poorly understood temperature - field phase diagrams @xcite that are weakly reminiscent of those for licuvo@xmath1 . in the following , we report a combined experimental and microscopic study of @xmath0-tevo@xmath1 and address several pending questions concerning this interesting material . in sec .
[ sec : diagram ] , we report temperature - field phase diagrams for different directions of the applied magnetic field and thus probe magnetic anisotropy in @xmath0-tevo@xmath1 experimentally . we show that @xmath0-tevo@xmath1 reveals very unusual phase diagrams with the first - order transition toward the helically ordered phase and second - order transitions between other phases . in secs .
[ sec : structure ] and [ sec : nmr ] , we discuss possible structural changes happening in the paramagnetic state and conclude that the overall symmetry of the structure as well as the orbital state of v@xmath13 are essentially unchanged down to at least @xmath5 . finally , in sec .
[ sec : microscopic ] we derive the microscopic spin hamiltonian of @xmath0-tevo@xmath1 , and briefly discuss its implications .
our results are summarized in sec .
[ sec : summary ] .
a slab - shaped single crystal of approximate dimensions @xmath22mm@xmath23 was oriented using x - ray scattering , and found to have the longest dimension along the crystallographic @xmath4-axis and the shortest along the @xmath2-axis , while the intermediate dimension is parallel to the @xmath3-axis .
it was experimentally found that while the @xmath24-plane cleaves easily , exposing the chain - like underlying structure , neither the @xmath25- nor the @xmath26-plane do so .
thermal expansion at constant magnetic fields and magnetostriction in pulsed magnetic field measurements were accomplished using an optical fiber bragg grating ( fbg ) technique described before @xcite . here ,
light reflected at the bragg wavelength @xmath27 by a grating inscribed in a telecom 125 @xmath28 m diameter optical fiber is recorded with a spectrometer furbished with fast ingaas line - array camera working at 46khz @xcite and used to follow the sample dilation as the temperature and/or external magnetic field are changed . for these experiments
the fiber was attached to the sample @xmath26-plane when studying the strain along the @xmath2 and @xmath3 crystallographic direction , and to the @xmath25-plane when studying the strain along the @xmath4 direction .
the magnetization in pulsed magnetic fields to 60 t at constant temperatures was obtained with a sample - extraction magnetometer working to @xmath23he temperatures , and calibrated with measurements in a quantum design^^ ppms system to 14 t .
specific heat measurements at constant magnetic fields were completed in the same ppms system .
high - resolution x - ray diffraction ( xrd ) data were collected on the id31 beamline of the european synchrotron radiation facility ( esrf , grenoble ) using the wavelength of 0.4a .
a small crystal of @xmath0-tevo@xmath1 was crushed , ground into fine powder and placed into a thin - wall borosilicate glass capillary that was spun during the measurement .
the signal was collected by eight si(111 ) analyzer crystals .
structure refinements were performed in the ` jana2006 ` program @xcite , and the resulting crystal structure was visualized using ` vesta ` @xcite .
all @xmath12 te nmr spectra were recorded on bruker avance - ii nmr spectrometer at 14.1 t magnetic field using home - built probe with a single - axis goniometer and he - flow cryostat from janis research inc .
the temperature was monitored and regulated by a lakeshore-332 temperature controller .
each data point was obtained by recording the signal with a spin - echo sequence .
the magnetic shift scale was referenced to the @xmath12te resonance frequency of me@xmath21te 189.349mhz .
the microscopic analysis of @xmath0-tevo@xmath1 is based on density - functional - theory ( dft ) band - structure calculations performed in the ` fplo ` code @xcite .
experimental crystal structure obtained from low - temperature xrd has been used , and either the local - density approximation ( lda ) @xcite or generalized - gradient - approximation ( gga ) @xcite exchange - correlation potentials were chosen .
isotropic exchange couplings were obtained from two complementary procedures , the lda - based model analysis and the dft+@xmath29 supercell calculations , as further explained in sec .
[ sec : microscopic ] . for the dft+@xmath29 calculations , we used supercells doubled along either @xmath2 or @xmath4 directions .
t in a wide temperature range .
inset : inverse magnetic susceptibility vs temperature showing curie - weiss behavior and fit .
bottom panel : magnetic susceptibility vs temperature in the low - temperature region at constant magnetic fields .
anomalies in @xmath30 are indicated as @xmath5 , @xmath6 and @xmath7 , and dashed lines are guides to the eye . ]
inset : magnetization vs magnetic field showing the full field range to 60 t . ]
-axis as a function of the increasing temperature in zero field , shown here to @xmath31k .
] magnetic susceptibility for the @xmath8 chain was obtained by combining transfer - matrix renormalization group ( tmrg ) @xcite and exact - diagonalization ( ed ) simulations for the low- and high - temperature parts of the data , respectively .
exact diagonalization was performed for a finite lattice with @xmath32 sites and periodic boundary conditions .
the use of tmrg helps to eliminate finite - size effects that manifest themselves at low temperatures .
field - dependent magnetization for the @xmath8 chain was calculated at zero temperature using density - matrix renormalization group ( dmrg ) method .
the ed and dmrg simulations were performed in the ` alps ` simulations package @xcite .
the magnetic susceptibility ( @xmath34 ) of a single - crystal sample was measured as a function of the temperature at constant magnetic fields between 0.5 t and 14 t .
the results are displayed in fig .
[ fig : chi ] . at high temperatures ,
the magnetic susceptibility follows a curie - weiss dependence with a clear maximum centered at @xmath35k ( fig .
[ fig : chi]a ) , followed by several smaller anomalies as the temperature is reduced
. a curie - weiss fit of the data in the @xmath36k range ( fig .
[ fig : chi]a , inset ) gives a curie constant @xmath37emukmol@xmath38 and @xmath39k indicating a nearly perfect balance of fm and afm couplings , in agreement with earlier work @xcite
. the low - temperature anomalies are more clearly visible when the low temperature range is expanded ( fig .
[ fig : chi]b ) .
three phase transitions were identified as @xmath5 , @xmath6 and @xmath7 , and followed as the magnetic field was increased .
while @xmath5 and @xmath6 share characteristics of second - order - like transitions , @xmath7 instead involves a rapid drop in @xmath34 suggesting a different , possibly first - order - like , process .
dashed lines indicate the evolution of them with applied field .
[ fig : expansion - c2 ] direction in pulsed magnetic fields .
anomalies are indicated by arrows . as seen here ,
contrary to @xmath3- an @xmath4-axis , the @xmath2-axis length contracts with applied fields .
the lack of saturation at fields @xmath40 t could be an artifact caused by magnetic torque on the sample . ]
[ fig : diagram - c ] the magnetization vs field @xmath41 measured in a powder sample , displayed in figure [ fig : mvsh ] , shows @xmath41 increasing gradually with field .
two small anomalies at 2.5 t and 5.8 t , not visible to the naked eye , are seen at @xmath42k ( fig .
[ fig : mvsh ] , inset ) and two large steps at 19 and 21 t .
these anomalies , likely associated to the suppression of low - temperature magnetic order , wash out and eventually vanish as the temperature is increased to 4k .
a saturation magnetization of 0.86@xmath43/f.u . persists to the maximum field of 60 t .
note that the saturation magnetization is somewhat below 1.0@xmath44 expected for a spin-@xmath14 system .
this might be due to ambiguities in scaling the pulsed - field data against the data collected in static fields .
the specific heat of a 6.2 mg slab - shaped sample was measured in magnetic fields applied along the crystallographic @xmath2- and @xmath3-axes .
representative data is displayed in fig .
[ fig : heat1 ] for different field directions .
particularly interesting are the @xmath45-axis data : here , the transition @xmath46k in zero field looks different from the other @xmath47-like transitions , much as in the case of @xmath48 data , reinforcing the support for a different type of transition . also , along this direction the transition @xmath5 is observed to split for magnetic fields @xmath49 t .
the nature of low - temperature phase transitions in @xmath0-tevo@xmath1 can be tested with probes that are sensitive to the crystal lattice , such as thermal expansion and magneto - strain @xmath50 , to gather information on the relevance and strength of spin - lattice correlations .
figure [ fig : expansion - c ] shows a negative thermal expansion along the @xmath4 direction ( @xmath28strain @xmath51 10@xmath52 ) , @xmath53 the sample expands as the temperature drops .
similar behavior was observed for the @xmath3 direction .
figure [ fig : expansion - b]a displays the thermal expansion at constant magnetic field @xmath54 t , in the low temperature range where the transitions at @xmath5 , @xmath6 , and @xmath7 are clearly visible . while @xmath5 and @xmath6 look like conventional @xmath47-like second - order transitions , @xmath7 , again , looks different . the sharp drop in lattice parameter on cooling resembles a first - order like transition .
it is often the case that phase transitions displaying important lattice involvement become first order and develop hysteresis . in this case , however , hysteresis was not significant
. we must point out , however , that our optical fiber technique is not the most appropriate for detection of thermal hysteresis as the fiber ( the only contact between sample and sample holder ) is a very poor thermal conductor and some thermal lag between sample and thermometer is always present .
figures [ fig : expansion - b]b and [ fig : expansion - b]c display the magnetostriction obtained in a pulsed magnet to @xmath55 t .
the transition @xmath7 is visible as an increase in the lattice parameter with field , while @xmath5 appears split first as a decrease of the lattice parameter at 20 t then as a larger increase at 23 t .
figure [ fig : expansion - c2]a ( right ) shows the thermal expansion measured along the chains direction @xmath4 at constant magnetic fields in the low temperature region .
@xmath5 , not visible in zero field , becomes clear as the magnetic field is increased to 14 t .
@xmath6 and @xmath7 , on the other hand , look very similar to the @xmath45 case .
figures [ fig : expansion - c2]b and [ fig : expansion - c2]c show the magnetostriction measured in pulsed magnetic fields , where the transitions @xmath7 , @xmath6 and @xmath5 are clearly identified .
while these transitions maintain similar characteristics for @xmath56 and @xmath57 , the critical fields are significantly different .
figure [ fig : expansion - a ] displays the magnetostriction in pulsed fields measured along the @xmath2-axis , which shows anomalies at approximately 19 t and 22.5 t . at these anomalies
the @xmath2-axis behaves opposite to the @xmath3- and @xmath4-axes , pointing to a partial compensation of the expansion trend shown by those .
the unit cell volume , hence , appears to expand as the sample magnetization is saturated by external magnetic fields .
critical temperatures and critical fields identified in the above discussed magnetization @xmath58 , specific heat @xmath59 , thermal expansion @xmath28strain(@xmath60 ) , and magnetostriction @xmath28strain(@xmath61 ) were extracted and plotted in two phase diagram displayed in figure [ fig : diagram - a ] .
these plots reveal the previously unknown and intricate anisotropy , and the possibility of a tricritical point at @xmath62 t , and @xmath63k . using zero - field neutron data @xcite
, we identify the phase below @xmath7 as the helical order ( h ) , the phase between @xmath6 and @xmath7 as the long - period stripe order ( s ) , and the phase between @xmath5 and @xmath6 as the sdw order .
the low - temperature helical phase is thus relatively stable in magnetic fields applied along @xmath2 and @xmath3 and becomes fragile when the field is applied along @xmath4 , where , on the other hand , a broader region of the stripe phase is observed .
our observations are consistent with the results of ref . , where temperature - field phase diagrams for two different field directions ( @xmath2 and @xmath4 not resolved ) were reported based on magnetic susceptibility measurements and below 5 t , only .
.[tab : coordinates ] atomic positions in @xmath0-tevo@xmath1 refined from high - resolution synchrotron powder xrd .
@xmath64 are isotropic atomic displacements parameters ( adps ) given in 10@xmath65a@xmath66 .
for each atom , the first line refers to the 10k data , and the second line refers to the room - temperature data .
the adps of oxygen atoms were constrained during the refinement .
lattice parameters are @xmath67a , @xmath68a , @xmath69a , @xmath70 at 10k and @xmath71a , @xmath72a , @xmath73a , @xmath74 at room temperature .
the space group is @xmath20 .
[ cols="^,^,^,^,^ " , ] ( color online ) formation of the helical magnetic structure in @xmath0-tevo@xmath1 . left panel : the competition between the helical structure along the chain and the fm interchain couplings ( @xmath75 , @xmath76 , @xmath77 ) , resulting in a phase shift @xmath78 between the neighboring helices along the @xmath2 directions .
right panel : same direction of rotation ( counter - clock - wise in this case ) in the chains 12 and 34 is imposed by the afm interchain coupling @xmath79 .
it is , however , incompatible with the dm vectors @xmath80 ( shown by orange arrows ) that change sign from ( @xmath81 to @xmath82 upon going from the chain 12 to the chain 34 .
note that the @xmath9 plane is arbitrarily chosen as the helix plane for the sake of better visualization . ]
we can now rationalize the incommensurate magnetic order that has been observed in @xmath0-tevo@xmath1 experimentally @xcite .
the propagation vector @xmath18 below @xmath7 implies the pitch angle @xmath83 for the helix propagating along the @xmath4 direction .
this is in very good agreement with @xmath84 that yields the classical pitch angle @xmath85 , whereas quantum corrections to the pitch angle are small in this range of @xmath86 @xcite .
the leading interchain coupling along the @xmath3 direction , @xmath79 , is compatible with the helical order and leads to the same spin arrangement within every second chain , so that , e.g. , both atoms 3 in fig .
[ fig : structure ] ( middle ) feature parallel spins resulting in @xmath87 . on the other hand ,
the fm interchain couplings are incompatible with the helical order ( fig .
[ fig : dm ] , left ) .
the couplings @xmath10 , @xmath75 , and @xmath76 ( @xmath77 ) build triangles , where the non - collinear order of the two spins coupled by @xmath10 can not be combined with the fm order imposed by @xmath75 and @xmath76 ( @xmath77 ) .
this frustration is alleviated by introducing a non - collinear spin arrangement on @xmath75 and @xmath76 ( @xmath77 ) too . by a classical energy minimization for the magnetic model defined by the exchange couplings from table [ tab : exchanges ] .
we checked that the simultaneous variation of @xmath78 and @xmath88 leads to a nearly indistinguishable result , because the value of @xmath88 is fixed by a much larger energy scale of @xmath89k .
] , we find that the spins on atoms 1 and @xmath90 should be turned by an angle @xmath91\simeq -53.9^{\circ}\ ] ] corresponding to @xmath92 , which is somewhat lower than @xmath93 in the experiment , but shows the correct sign and explains the opposite directions of rotation along @xmath2 and @xmath4 .
we find that the incommensurability along the @xmath2 direction is controlled by a rather subtle difference between the diagonal interactions @xmath76 and @xmath77 that correspond to the bond vectors @xmath94 and @xmath95 , respectively , where @xmath96 is the difference between the @xmath97-coordinates of two neighboring v atoms along the chain .
the interaction @xmath76 is more fm than @xmath77 .
therefore , the angle @xmath78 has to counteract the rotation introduced by @xmath88 in order to bring spins 1 and 2 * connected by @xmath76 closer to the parallel configuration . in the following ,
we extend our microscopic analysis to the anisotropy parameters entering eq . .
technically , these parameters could be evaluated in the same spirit of perturbation theory as in eq . , following refs .
however , the evaluation of magnetic anisotropy requires that hoppings to all four unoccupied @xmath98-orbitals , including @xmath99 and @xmath100 , are defined , which is not the case in @xmath0-tevo@xmath1 , where the strong hybridization between te @xmath101 and v @xmath102 orbitals prevents unambiguous modeling of the higher - lying @xmath98-bands .
therefore , we restrict ourselves to a qualitative analysis that provides useful insight into the role of magnetic anisotropy in @xmath0-tevo@xmath1 .
the dm terms are allowed by symmetry for both intrachain couplings .
symmetry relations between different components of the @xmath103 and @xmath80 on individual lattice bonds are summarized in table [ tab : dm ] .
helical ground state implies that same rotations occur between every two contiguous spins in a chain ( fig .
[ fig : dm ] ) . since all three components of @xmath104 change sign within the chain
, the nearest - neighbor dm couplings do not gain any energy from the helical order .
the same argument shows that @xmath105 can not stabilize the helical order , because it changes sign between @xmath106 and @xmath107 . on the other hand , @xmath108 and @xmath109 gain energy from the helical order and
should thus stabilize the helix in the plane perpendicular to @xmath110 within the @xmath111 chain .
however , the relevant dm components in the neighboring @xmath112 chains are @xmath113 implying that the helix with the opposite sense of rotation will be stabilized . on the other hand ,
the afm interchain coupling @xmath79 imposes same sense of rotation in the helices @xmath111 and @xmath112 ( fig .
[ fig : dm ] , right ) .
the sign change of @xmath80 between the two contiguous chains implies that the helix plane in @xmath0-tevo@xmath1 can not be uniquely defined when the dm anisotropy is considered .
this striking observation should be parallel to the following .
first , the helical order is stabilized only below @xmath114 upon a first - order phase transition that entails an abrupt shrinkage of the unit cell along the @xmath3- and @xmath4-directions implying a structural effect that probably alleviates this frustration .
second , neutron - scattering data indicate two different helix planes for the chains @xmath111 and @xmath112 , respectively @xcite . this observation is hard to reconcile with the monoclinic crystal structure of @xmath0-tevo@xmath1 , where these two chains are crystallographically equivalent ( related by an inversion symmetry ) , and thus a structural distortion must be involved .
the incompatibility of the helical order with the dm anisotropy in @xmath0-tevo@xmath1 is also a plausible reason behind the delayed formation of the helical phase upon cooling in zero field .
other @xmath8 frustrated - chain compounds reported so far @xcite undergo a direct ( and second - order ) transition from the paramagnetic phase to the helically - ordered phase in zero field . in @xmath0-tevo@xmath1 , however , the paramagnetic and helically - ordered phases are separated by the sdw phase ( between @xmath5 and @xmath6 ) and by the long - period stripe order ( between @xmath6 and @xmath7 ) , see fig .
[ fig : diagram - c ] .
we can also conclude that the helical order is largely destabilized in the magnetic field applied along the @xmath4-direction ( compare the two panels of fig .
[ fig : diagram - a ] ) , which should then be the common direction of the two helices , i.e. , the direction that is most vulnerable to the application of the magnetic field .
@xmath0-tevo@xmath1 is a structurally perfect material prototype of the @xmath8 frustrated spin chain model .
our data rule out any structural distortions preceding magnetic transitions in this compound , and clearly exclude any change in the orbital state of v@xmath13 upon cooling . on the other hand , magnetic anisotropy triggered by
the low crystallographic symmetry introduces a very complex behavior , especially in low magnetic fields . by comparing temperature - field phase diagrams obtained for different field directions ,
we conclude that the transitions at @xmath6 and @xmath7 are strongly direction - dependent and should be influenced or even triggered by the presence of magnetic anisotropy .
the field evolution of @xmath5 is , at first glance , reminiscent of a conventional long - range afm ordering and reveals no appreciable anisotropy . finally , the high - field phase emerging above 18 t turns out to be weakly dependent on the field direction and thus nearly isotropic .
the emergence of the high - field phase for different field directions indicates its relation to the physics of the isotropic @xmath8 spin chain .
this phase can be ascribed to the multipolar order or nematic state envisaged in recent theoretical studies @xcite .
the multipolar order should be robust with respect to the interchain couplings at @xmath115 @xcite , which renders @xmath0-tevo@xmath1 a good model material for studying high - field physics of the frustrated @xmath8 spin chain .
further investigation of this high - field phase is highly desirable . in lower fields
, we observed that the helical phase is destabilized by the magnetic field @xmath57 and gives way to a larger region of the long - period stripe order , while for other field directions the stripe phase shrinks to a small pocket visible in low magnetic fields only .
both helical and spin - density - wave phases are incommensurate along both @xmath2 and @xmath4 @xcite .
while the incommensurability along the @xmath4-direction is a natural result of the intrachain frustration , the incommensurability along @xmath2 can be understood as a competition of ferromagnetic interchain couplings with the helical ( or spin - density - wave ) order within the chain . regarding magnetic anisotropy
, @xmath0-tevo@xmath1 is different from any other @xmath8 frustrated - spin - chain compound reported so far .
materials like cugeo@xmath116 and nacu@xmath21o@xmath21 lack dm couplings completely , because inversion centers are found in the middle of both nearest - neighbor and next - nearest - neighbor bonds . in licuvo@xmath1 and linarite , inversion symmetry forbids @xmath80 , whereas @xmath103 may be non - zero , although its exact magnitude is still unknown . finally , @xmath0-tevo@xmath1 has inversion centers between the chains only .
therefore , all dm couplings are non - zero , and the symmetry of @xmath80 is compatible with the helical order , thus providing additional stabilization energy for each helix , but impeding the order between the helices .
this frustration of anisotropic exchange couplings may be the crux of the @xmath0-tevo@xmath1 physics in low magnetic fields that awaits further investigation with direct methods , such as neutron scattering in applied magnetic field .
we are grateful to peter lemmens for initiating this work .
we also acknowledge fruitful discussions with andrs sal , myron b. salamon , and johannes richter , and the provision of the beamtime by the esrf .
the national high magnetic field laboratory pulsed - field facility is supported by the national science foundation ( nsf ) , the us department of energy ( doe ) , and the state of florida through nsf cooperative grant dmr-1157490 . at
was supported by the federal ministry for education and research via the sofja kovalevskaya award of alexander von humboldt foundation .
the work in tallinn was supported by the estonian research council grants mtt77 , put733 , put210 and iut23 - 7 .
47ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ]
+ 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) , @noop * * , ( ) , @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) , @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop `` , '' ( ) , @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop _ _ ( , ) @noop * * , ( ) @noop * * , ( ) , @noop * * , ( ) , @noop * * , ( ) , @noop * * , ( ) , @noop * * , ( ) @noop * * , ( ) , @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) , @noop * * , ( ) | isotropic and anisotropic magnetic behavior of the frustrated spin chain compound @xmath0-tevo@xmath1 is reported .
three magnetic transitions observed in zero magnetic field are tracked in fields applied along different crystallographic directions using magnetization , heat capacity , and magnetostriction measurements .
qualitatively different temperature - field diagrams are obtained below 10 t for the field applied along @xmath2 or @xmath3 and along @xmath4 , respectively .
in contrast , a nearly isotropic high - field phase emerges above 18 t and persists up to the saturation that occurs around 22.5 t . upon cooling in low fields ,
the transitions at @xmath5 and @xmath6 toward the spin - density - wave and stripe phases are of the second order , whereas the transition at @xmath7 toward the helical state is of the first order and entails a lattice component .
our microscopic analysis identifies frustrated @xmath8 spin chains with a sizable antiferromagnetic interchain coupling in the @xmath9 plane and ferromagnetic couplings along the @xmath2 direction .
the competition between these ferromagnetic interchain couplings and the helical order within the chain underlies the incommensurate order along the @xmath2-direction , as observed experimentally .
while a helical state is triggered by the competition between @xmath10 and @xmath11 within the chain , the plane of the helix is not uniquely defined because of competing magnetic anisotropies . using high - resolution synchrotron diffraction and @xmath12te nuclear magnetic resonance
, we also demonstrate that the crystal structure of @xmath0-tevo@xmath1 does not change down to 10k , and the orbital state of v@xmath13 is preserved . |
one of the simplest cases to study the effects of impurities in magnetic systems occurs in some doped uniaxial antiferromagnets such as @xmath0 @xcite , where the @xmath1 impurities do not induce competing interactions and monte carlo simulations with the three - dimensional random - site ising model reproduce accurately the behavior close to the curie temperature @xcite .
nevertheless , the successful application of renormalization group techniques @xcite to the calculation of critical exponents for the pure ising model makes perturbative expansions a very appealing alternative to the monte carlo calculations mentioned above .
the appropriate field theory for this problem is the usual @xmath2 theory in three dimensions modified with a random mass term , and the analytic difficulties raised by the quenched impurities are usually bypassed with the replica trick : the system is temporarily replaced by @xmath3 non - interacting copies for which the disorder is now annealed , yielding an @xmath4 theory with coupling constant proportional to minus the variance of the quenched disorder @xmath5 , with an additional cubic anisotropy term with coupling constant @xmath6 .
this theory can be studied perturbatively and the results corresponding to the original system recovered ( in principle ) in the limit @xmath7 . indeed , there are rather accurate analytic calculations of the critical exponents of the random - site ising model @xcite , but the situation is less satisfactory than for pure systems @xcite .
for instance , the asymptotic parameter in disordered systems is @xmath8 instead of @xmath9 ( where @xmath10 , @xmath11 being the dimensionality of the space ) , and the @xmath12-function computed at three loops shows no stable fixed point @xcite . in an attempt to understand disordered systems in a simpler setting , bray @xcite and mckane @xcite studied the asymptotic expansion for the free energy in the zero - dimensional case , with discouraging results : the toy - model was found non borel - summable , in sharp contrast with the ( ordered ) zero - dimensional @xmath2 theory , where borel - summability has been rigorously proved @xcite .
( indeed , borel summability in three dimensions has been also proved @xcite . )
we consider in this paper the same zero - dimensional problem of references @xcite , and show that a slightly more elaborated procedure can recover the free energy from the perturbative expansion : the borel - summation has to be carried out first in the strength of the original coupling of the ising model @xmath6 and subsequently in the variance of the quenched disorder @xmath5 .
the layout of the rest of the paper is as follows .
we first review the main results of bray and mckane in section [ sec : pre ] ; the derivation of the borel - summable double asymptotic expansion for the free energy is carried out in the third section ; in section [ sec : num ] we give the details of our numerical procedure and some illustrative examples , and the paper ends with a brief summary .
the free energy of the zero - dimensional ising model with quenched dilution can be studied with a @xmath2 theory with a random mass term @xmath13 where @xmath14 is the partition function .
bray @xcite fix @xmath15 and write the asymptotic expansion for the free energy as @xmath16 where the @xmath17 are polynomials in @xmath18 with rational coefficients and degree @xmath19 .
then they use the replica trick an a saddle point argument to infer that the asymptotic behavior of @xmath17 as @xmath20 is given by two terms : @xmath21 the term @xmath22 is dominant for @xmath23 and borel - summable , while @xmath24 with coefficients @xmath25 , @xmath26 , @xmath27 and @xmath28 that they ultimately adjust by comparison between this asymptotic formula and the numerical values of the @xmath17 for @xmath19 up to 200 and several values of @xmath18 greater than @xmath29 .
the increasing - period cosine oscillations suggest them that there is an essential singularity of the borel - transform in the positive real @xmath6 axis , and therefore the series is not borel - summable for any finite disorder ( although the term @xmath30 is subdominant in the region @xmath23 ) . despite the non - summability of the series ( [ eq : fak ] ) , i.e. acknowledging that there would be some error even if they could compute with all the terms , bray set @xmath31 and try to extract an accurate answer from ( [ eq : fak ] ) using a conformal transformation suitable for oscillatory series , but they find that as soon as @xmath32 the method fails .
mckane @xcite avoids the replica trick in favor of a more direct analysis that clearly identifies the two sources of non - analytic behavior of the free energy leading to the two contributions is equation ( [ eq : akasy ] ) : the branch cut along the negative @xmath6 axis and the zeros of the partition function ( [ eq : z ] ) .
moreover , he obtains the exact values of @xmath25 , @xmath26 , @xmath27 and @xmath28 in terms of the zeros of the partition function and guesses the form of the nonperturbative contributions to the free energy that yield the asymptotic behavior ( [ eq : akasy ] ) .
the possibility of alternative summation methods , however , remained unaddressed in these works .
for analyticity considerations , we remark that in terms of the new variable @xmath33 the partition function ( [ eq : z ] ) can be written as @xmath34 which in turn @xcite can be expressed in terms of a parabolic cylinder function or a modified bessel function : @xmath35 there is , however , another expression particularly suitable for later summability considerations : via a mellin - barnes integral @xcite , the partition function can be written as a laplace transform in @xmath36 , @xmath37^{3/4 } } \qquad ( \re ( z^2 ) > 0 ) \label{eq : z2lt}\\ & = & \frac{z}{(4u)^{1/4}\gamma(\case14 ) } \int_{0}^{\infty } \frac{\rmd t\,\e^{- t}}{[t(t+z^2)]^{3/4 } } \qquad ( \re z > 0 ) .
\label{eq : zlt}\end{aligned}\ ] ] equations ( [ eq : zz ] ) or ( [ eq : zd ] ) show that @xmath38 is an entire function of @xmath39 , and from equation ( [ eq : zk ] ) and reference @xcite it follows that the only zeros of @xmath40 appear in complex - conjugate pairs @xmath41 in the @xmath42 half - plane , and are given asymptotically by @xmath43 to stay close to the notation of bray and mckane and yet show that we are dealing with a two - variable problem , we set @xmath44 @xmath45 @xmath46 in terms of these variables , the free energy can be written as @xmath47 \label{eq : phi}\ ] ] were the integral in the right - hand side is taken along the real axis , and is well defined for any complex @xmath27 and @xmath48 .
note that as @xmath39 varies along the real axis in the increasing sense , @xmath40 traces the positive axis in the decreasing sense , but as we shift ( say ) upwards the integration path and @xmath39 picks up a constant ( positive ) imaginary part [ i.e. @xmath49 , equation ( [ eq : zd ] ) shows that @xmath40 traces a spiral ending at the origin for @xmath50 .
in particular , if @xmath51 the spiral passes again through the origin ( see figure 1 ) .
therefore , if we shift vertically the integration path to @xmath52 ( with @xmath53 ) , we pick up a finite number of contributions from the zeros of @xmath40 whose evaluation in terms of the complementary error function is given by the following self - explanatory chain of equations : @xmath54\nonumber\\ { } + \sum_{0<\im z_{k}<\im\mu } \frac{1}{(\lambda\pi)^{1/2 } } \int_{-\infty , z_{k}^{(+ ) } } \rmd z\ , \e^{- ( z - \mu)^2/\lambda } \log[f(z)]\\ \lo= \case12 \log(2\mu ) + \frac{1}{(\lambda\pi)^{1/2 } } \int_{-\infty}^{\infty } \rmd t\ , \e^{- t^2/\lambda } \log[f(t+\mu)]\nonumber\\ { } + \sum_{0<\im z_{k}<\im\mu } \frac{2\pi\rmi}{(\lambda\pi)^{1/2 } } \int_0^\infty \rmd t\ , \e^{- ( z_k - t - \mu)^2/\lambda}\\ \lo= \case12 \log(2\mu ) + \frac{1}{(\lambda\pi)^{1/2 } } \int_{-\infty}^{\infty } \rmd t\ , \e^{- t^2/\lambda } \log[f(t+\mu)]\nonumber\\ { } + \rmi\pi\sum_{0<\im z_{k}<\im\mu } { \rm erfc}\left(\frac{\mu - z_{k}}{\lambda^{1/2}}\right ) .
\label{eq : erfc}\end{aligned}\ ] ] there are three key points to note in these expressions .
first , the determination of the logarithms : they are real for real argument ( as @xmath55 ) and must be followed by continuity according to figure 1 .
second , for @xmath56 there are no contributions from the zeros and we do have @xmath57 .
\label{eq : fzm}\ ] ] lccc & @xmath58 & @xmath59 & @xmath60 + partial sum & @xmath61 & @xmath62 & @xmath63 + equation ( [ eq : fzm ] ) & @xmath64 & @xmath65 & @xmath66 + equation ( [ eq : erfc ] ) & @xmath64 & @xmath67 & @xmath68 + equation ( [ eq : phi ] ) & @xmath64 & @xmath67 & @xmath68 + we stress that the right - hand side of equation ( [ eq : fzm ] ) is well defined whenever @xmath69 , but it is equal to the free energy only for @xmath56 . as a numerical illustration of this point , in the last three rows of table 1 we show numerical calculations of the shifted integral ( [ eq : fzm ] ) ,
the shifted integral plus corrections ( [ eq : erfc ] ) , and the unshifted integral ( [ eq : phi ] ) for @xmath70 and three values of @xmath27 with @xmath71 , @xmath72 and @xmath73 respectively .
third , equation ( [ eq : fzm ] ) can be rewritten as a laplace transform in @xmath74 @xmath75 where @xmath76 .
\label{eq : bdef}\ ] ] the singularities of @xmath77 as a function of @xmath78 are readily calculated in terms of the @xmath79 and ( apart from the trivial singularity at the origin ) stay at a finite distance of the positive real axis whenever @xmath69 .
equations ( [ eq : zlt ] ) and ( [ eq : fdef ] ) imply that we can obtain a borel - summable asymptotic expansion for @xmath40 and its derivatives , and equations ( [ eq : bmt ] ) and ( [ eq : bdef ] ) that , as far as we stay away from the zeros of @xmath80 , we can also obtain a borel - summable series in @xmath18
. our final result will be , therefore , a borel - summable series ( in @xmath18 ) with each of its coefficients given by a borel - summable series ( in @xmath81 ) .
now we proceed to the details of the derivation .
we first expand @xmath82 $ ] in equation ( [ eq : fzm ] ) in ( convergent ) taylor series around @xmath83or equivalently @xmath84 in equation ( [ eq : bmt ] ) around @xmath85and integrate term by term to get @xmath86 then , we use the borel - summable asymptotic expansion for @xmath40 that follows directly from equation ( [ eq : zlt ] ) @xmath87 to obtain asymptotic expansions for the derivatives @xmath88 .
incidentally , we mention that the summability sector of @xmath40 in terms of @xmath6 is just @xmath89 , i.e. we are rederiving the well - known summability of the ( ordered ) zero - dimensional @xmath2 theory as explained , for example , in reference @xcite .
note that for the zero - th derivative we have @xmath90 & \sim & - \case12\log(2 \mu ) + \log\left[{}_{2}f_{0}(\case14,\case34;;-\mu^{-2})\right]\\ & = & - \case12\log(2 \mu ) + \sum_{k=1}^{\infty } \frac{(-1)^k b_{k}}{\mu^{2k } } \label{eq : logfmu } \end{aligned}\ ] ] and the logarithm cancels the first term in the right - hand side of equation ( [ eq : fzm ] ) .
( equation ( [ eq : logfmu ] ) is in fact the definition of the coefficients @xmath91 ) .
higher derivatives have pure asymptotic power series that can be computed easily in terms of the zero - th derivative series @xmath92
. \label{eq : f2n}\ ] ] the final result is the double series @xmath93 \label{eq : dsf}\ ] ] whose structure in terms of the original variables is ( @xmath94 ) @xmath95 this double series is formally equivalent to equations ( b10 ) , ( b11 ) and ( b13 ) in reference @xcite ( the only difference is that we have transposed the indices in @xmath96 ) , but with a definite ordering that can not be altered without further considerations : to take advantage of the summability we should first borel - sum the series in @xmath6 ( see the precise definition in the next section ) @xmath97 and then in @xmath5 @xmath98 instead of performing first the finite sum @xmath99 followed by some kind of summation in @xmath6 . although this last reordering preserves the asymptotic nature of the series to the function defined by equation ( [ eq : fzm ] ) ( as we also illustrate in the first row of table 1 ) , the numerical and analytic evidence presented by bray shows that it spoils the summability of equation ( [ eq : dsf ] ) .
of the two standard methods to achieve the analytic continuation implicit in the borel - summation ( conformal mapping or pad approximants ) we have chosen the second because it does not require a precise knowledge of the singularities of the function , although some pad approximants may introduce spurious singularities close to the integration path .
equation ( [ eq : f2n ] ) suggests that an appropriate numerical borel - sum of the @xmath6 series is @xmath100}(t ) \,\rmd t \label{eq : nbt}\ ] ] where @xmath101}(t)$ ] is the @xmath102$]-pad approximant for @xmath103 equation ( [ eq : nbt ] ) has the additional advantage over other equivalent forms that the pad approximants and their zeros ( which we use later ) need to be calculated only once .
furthermore , to avoid problems with the numerical integration we expand the pad approximant as a polynomial plus partial fractions , i.e. assuming that all the poles are simple @xmath104}(t ) = \sum_{k=0}^{p - q } p_{k } t^{k } + \frac{r(t)}{s(t ) } = \sum_{k=0}^{p - q } p_{k } t^{k } + \sum_{k=1}^{q } \frac{r(t_{k})}{s'(t_{k})(t - t_{k})}\ ] ] and the integration in equation ( [ eq : nbt ] ) can be carried out in terms of complete and incomplete gamma functions evaluated at the @xmath105 poles @xmath106 of @xmath107}(t)$ ] : @xmath108 we have performed our calculations with the same 200 terms of the expansion ( [ eq : logfmu ] ) used in reference @xcite , using the arbitrary precision numerical capabilities of _ mathematica_. figure 2 shows some typical results of this first stage of our double - summation method . for four values of @xmath6 and the lowest eleven coefficients @xmath109 we plot a measure @xmath110 of the number of correct digits to the right of the decimal point , @xmath111 as a function of the order @xmath112 of the diagonal pad approximants @xmath113}(t)$ ] in equation ( [ eq : nbt ] ) . as a general trend ,
the precision of the summed coefficients increases with the order of the approximant , although with smaller slope for higher values of @xmath6 ( note that we have used the same scales in the four plots ) .
the irregularities in figure 2 are often due to ill - conditioned linear systems of equations in the calculation of the pad approximants , and are usually corrected using higher precision or higher order approximants .
we have also checked that neighboring off - diagonal pad approximants give similar results . the second stage , the summation in @xmath5 , is equivalent to the @xmath114 summation of the first stage .
now the borel - transformed coefficients @xmath115 are precisely the coefficients of the convergent taylor expansion of @xmath116 , and to illustrate how well the pad approximants continue the taylor series beyond the radius of convergence , we plot in figure 3 the exact integrand @xmath117 corresponding to @xmath31 in equation ( [ eq : bmt ] ) , and its approximations @xmath118}(t)$ ] for @xmath119 , @xmath120 and @xmath121 .
the radius of convergence of the taylor series is marked in the plot by the vertical line at @xmath122 . in figure 4 ( again with @xmath31 and for several values of @xmath18 ) we plot the number of correct digits to the right of the decimal point in the borel - summed free energy as a function of the order of the diagonal pad approximant .
we have performed the calculations with all the @xmath123 of the same precision , so that the numerical error is due only to this second - stage summation .
the first feature worth mentioning is the steady increase in the precision of the summed free energy ( with a few exceptions that are related again to the pad approximants ) , although the convergence is slower for larger values of @xmath18 .
note that the plateau in the summation of bray for @xmath124 ends approximately at fifty terms of the series , while at @xmath125 ( i.e. with 81 terms of the series ) we have steady numerical convergence ; @xmath126 corresponds to the switch between asymptotic behaviors in equation ( [ eq : akasy ] ) , while in our reordering it is not a distinguished point ; finally , for @xmath127 bray are already unable to get a good result , while our summation still gives four correct digits for @xmath128 .
we have analyzed the borel - summability of the perturbative expansion for the free energy in one of the simplest disordered systems , the zero dimensional quenched diluted ising model .
this expansion is a double asymptotic series in the original coupling @xmath6 of the ising model and in the variance @xmath5 of the quenched disorder , and we have shown that the expansion is in fact borel - summable by a sequential method : first we sum in the coupling @xmath6 and subsequently in the variance @xmath5 .
this summability results are valid at least for @xmath89 and @xmath129 , with the understanding that if we want to recover the free energy and not the integral ( [ eq : fzm ] ) , additional @xmath130 terms are needed in the domains specified in section [ sec : asy ] . a straightforward numerical implementation of our double summation method has permitted us to obtain ( with the same number of terms of the original series ) accurate free energies for values of the parameters well beyond the region where bray were already unable to get an estimate .
we expect that the insight provided by these zero dimensional results will be helpful in the understanding of the more complicated higher - dimensional case , despite the significant qualitative differences motivated by griffiths singularities @xcite in the generalized sense of dotsenko @xcite .
in fact we feel that a clear understanding of the griffiths phase in the field - theoretical setting will be needed before the problem will be finally solved .
finally , we would like to mention as open problems the study of the negative @xmath18 region , and a more thorough understanding of the solution of this problem using replicas .
we thank professors g parisi for useful discussions and a j bray for providing references .
the financial support of the universidad complutense under project pr156/97 - 7100 , the comisin interministerial de ciencia y tecnologa under project aen96 - 1708 , and a postdoctoral grant from the mec are also gratefully acknowledged .
00 slani z , belanger d p and fernandez - baca j a 1999 _ phys .
lett . _ * 82 * 426 ballesteros h g , fernandez l a , martn - mayor v , muoz sudupe a , parisi g and ruiz - lorenzo j j 1998 _ phys .
b _ * 58 * 274 zinn - justin j 1990 _ quantum field theory and critical phenomena _ ( oxford : oxford science publications ) newman k e and rieded e k 1982 _ phys .
b _ * 25 * 264 mayer i o 1989 _ j. phys . a : math .
* 22 * 2815 folk r , holovatch y and yavorskii t 1999 _ pisma v zhetf _ * 69 * 698 [ _ jetp lett . _ * 69 * 747 ] bray a j , mccarthy t , moore m a , reger j d and young a p 1987 _ phys . rev .
b _ * 36 * 2212 mckane a j 1994 _ phys . rev . b _ * 49
* 12003 eckmann j p , magnen j and snor r 1975 _ comm . math . phys .
_ * 39 * 251 slater l j 1960 _ complex hypergeometric functions _ ( cambridge : cambridge university press ) watson g n 1944 _ a treatise on the theory of bessel functions _
( cambridge : cambridge university press ) griffiths r 1969 _ phys .
* 23 * 17 dotsenko v 1999 _ j. phys . a : math .
gen . _ * 32 * 2949 | we show analytically that the perturbative expansion for the free energy of the zero dimensional ( quenched ) disordered ising model is borel - summable in a certain range of parameters , provided that the summation is carried out in two steps : first , in the strength of the original coupling of the ising model and subsequently in the variance of the quenched disorder .
this result is illustrated by some high - precision calculations of the free energy obtained by a straightforward numerical implementation of our sequential summation method . |
the @xmath0 meson , the heaviest of the ground pseudoscalar meson with explicit both bottom and charm quantum numbers , was observed for the first time via the decay @xmath0
@xmath1 @xmath6 in @xmath7 tev @xmath8 collisions using the cdf detector at the fermilab tevatron in 1998 @xcite .
the mass is currently measured at the @xmath9 level @xcite . however , there is about @xmath10 uncertainty in the present @xmath0 lifetime measurement @xcite .
the @xmath0 meson , laying below @xmath11 threshold , is stable for both the strong and electromagnetic annihilation interactions
. it can decay only via the weak interaction .
the decay modes can be divided into three classes ( taking @xmath12 as an illustration ) @xcite : ( 1 ) the @xmath13 quark decays while the @xmath14 quark as a spectator , that is , @xmath13
@xmath1 @xmath15 @xmath16 @xmath17 ( or @xmath18 ) ; ( 2 ) the @xmath14 quark decays while the @xmath13 quark as a spectator , that is , @xmath14
@xmath1 @xmath15 @xmath16 @xmath19 ( or @xmath20 ) ; ( 3 ) the annihilation channel , that is , @xmath13
@xmath16 @xmath14 @xmath1 @xmath15 ; where the virtual @xmath15 boson materializes either into a pair of leptons @xmath21 , or into a pair of quarks which then hadronizes . both the @xmath13 and @xmath14 quarks in @xmath12 meson can decay individually , resulting in that the @xmath0 lifetime is about three times less than other ground pseudoscalar @xmath22-flavored mesons , that is , @xmath23
@xmath24 @xmath25 @xcite . enhanced by the hierarchy of the cabibbo - kobayashi - maskawa ( ckm ) quark - mixing matrix elements @xmath26
@xmath27 @xmath28 , the class ( 1 ) decay contributes @xmath29 @xmath30 to the @xmath0 decay width @xcite , which shows the critical contribution of the charm quark to the @xmath0 lifetime . in this paper , we study the @xmath0 @xmath1 @xmath2 decay which belongs to class ( 1 ) .
our motivations are listed below .
firstly , from an experimental consideration .
it is estimated @xcite that with large production cross section @xmath31 @xmath29 @xmath32 ( taking into account the cascade decays of the excited states @xmath33 below @xmath11 threshold ) and high luminosity of @xmath34 @xmath35 @xmath36 , one could expect @xmath37 of @xmath0 meson per year at lhc .
in addition , lhcb , a dedicated @xmath22-physics precision experiment at lhc , allows proper time resolution of @xmath29 @xmath38 fs @xcite which is one order of magnitude shorter than lifetime of the @xmath39 meson .
this property facilitates the separation between primary and decay vertices .
the rate for the charged pion identification is larger than @xmath40 @xcite .
there seems to be a golden opportunity and a real possibility to investigate the @xmath0 weak decay .
the reconstructed events of the @xmath0 @xmath1 @xmath2 decay is estimated to be @xmath29 @xmath41 per year @xcite .
recently , the lhcb collaboration has presented research that the @xmath0 @xmath1 @xmath2 decay is observed with significance in excess of five standard deviations based on 3 fb@xmath42 data sample @xcite . secondly , from phenomenological considerations . with abundant measurements
, we can carefully test various theoretical models .
accordingly , accurate theoretical prediction of @xmath0 decay is essential , although the @xmath0 weak decay is complicated because of strong interaction effects .
based on an expansion in @xmath43 and @xmath44 ( where @xmath45 , @xmath46 and @xmath47 are the strong coupling constant , qcd characteristic scale and mass of heavy quark @xmath48 , respectively ) , several attractive methods have been proposed to evaluate the hadronic matrix elements , such as the qcd factorization @xcite , perturbative qcd method ( pqcd ) @xcite , and soft and collinear effective theory @xcite .
these approaches have been applied to @xmath49 hadronic decays with reasonable explanation for measurements . whether or not these approach is suitable for @xmath0 @xmath1 @xmath2 decay needs to be examined .
herein we investigate @xmath0 @xmath1 @xmath2 decay with the pqcd approach to give an estimate of the associated branching ratio .
for hadronic decays , typically the effective hamiltonian calculations with operator product expansion scheme is used .
the effective hamiltonian for @xmath0 @xmath1 @xmath2 decay can be written as @xcite : @xmath50 where @xmath51 is fermi coupling constant . the ckm factor @xmath52 @xmath29 @xmath53 .
@xmath54 and @xmath55 are color indices .
@xmath56 @xmath35 @xmath57 .
the wilson coefficients @xmath58 , incorporating the physics contributions from heavy particles such as @xmath59 and @xmath60 bosons , top quark , and scales higher than @xmath61 , have been calculated to the next - to - leading order with perturbation theory @xcite .
there most difficult problem remaining in calculation is how to evaluate the hadronic matrix elements of the local operators properly and accurately .
using brodsky - lepage approach @xcite , the hadronic matrix element is commonly written as a convolution of hard - scattering kernels and hadron wave functions ( wfs ) .
it is shown that @xcite due to appearance of the endpoint divergences , the qcdf formalisms based on the collinear approximation fail to give the contributions of the spectator and annihilation interactions satisfactorily .
the pqcd approach advocates that @xcite the endpoint singularity in collinear approximation could be smeared by retaining the transverse momentum @xmath62 of quarks and by introducing the sudakov factor @xmath63 . using the pqcd formalisms ,
one decay amplitude is factorized into three factors : the `` harder '' effects incorporated into the wilson coefficients @xmath64 , the heavy quark decay subamplitude @xmath65 , and the universal hadron wfs @xmath66 , such that : @xmath67 where @xmath68 and @xmath69 are the corresponding kinematic variables and characteristic scale , respectively . in the terms of the light cone
coordinate , the momenta of the valence quarks and hadrons in the rest frame of the @xmath0 meson are defined as : @xmath70 where the subscript @xmath71 @xmath35 @xmath72 , @xmath73 , @xmath74 refer to @xmath0 , @xmath3 , @xmath75 meson , respectively .
variables @xmath76 , @xmath77 , @xmath78 are the four - dimensional momentum , transverse momentum and longitudinal momentum fraction of light quark , respectively .
@xmath79 @xmath35 @xmath80 .
hadron wave functions are basic input in eq.([am01 ] ) . adopting the notation in @xcite , the two - valence - particle wfs for double - light pion , heavy - light @xmath3 meson and double - heavy @xmath0 meson are decomposed into : @xmath81\big\}_{{\beta}{\alpha } } , \label{wf - pi-01 } \\ { \langle}b_{s}(p_{2}){\vert }
\bar{s}_{\alpha}(z)b_{\beta}(0 ) { \vert}0{\rangle } & = & \frac{-i}{\sqrt{2n_{c } } } { \int}{\bf d}^{4}k_{2}\,e^{+ik_{2}{\cdot}z } \big\ { { \gamma}_{5 } \big(\!\!\not{\!p}_{2}\!+\!m_{b_{s}}\!\big ) \big[\frac{\not{\!n}_{+}}{\sqrt{2}}{\phi}_{b_{s}}^{-}\ ! + \frac{\not{\!n}_{-}}{\sqrt{2}}{\phi}_{b_{s}}^{+ } \big]\big\}_{{\beta}{\alpha } } , \label{wf - bs-01 } \\ { \langle}0{\vert }
\bar{b}_{\alpha}(0)c_{\beta}(z ) { \vert}b_{c}(p_{1}){\rangle } & = & \frac{-i}{\sqrt{2n_{c } } } { \int}{\bf d}^{4}k_{1}\,e^{-ik_{1}{\cdot}z } \big\{\big [ \frac{\not{\!n}_{+}}{\sqrt{2}}{\phi}_{b_{c}}^{+}\ ! + \frac{\not{\!n}_{-}}{\sqrt{2}}{\phi}_{b_{c}}^{- } \big]\big(\!\!\not{\!p}_{1}\!+\!m_{b_{c}}\!\big ) { \gamma}_{5}\big\}_{{\beta}{\alpha } } , \label{wf - bc-01 } \end{aligned}\ ] ] where @xmath82 @xmath35 @xmath74 is the color number . @xmath83 and @xmath84 are the minus and plus null vectors , respectively .
@xmath85 @xmath35 @xmath72
. for the heavy - light @xmath3 meson , there are two scalar wfs @xmath86 and @xmath87 .
the equation of motion for wfs @xmath88 is @xcite @xmath89 the relationship is helpful in constraining models for the @xmath3 wfs , which leads to @xmath90 vanished at the endpoint and @xmath91 @xmath35 @xmath53 for @xmath92 @xmath1 @xmath93 @xcite .
here we will investigate three candidates of wfs for @xmath3 meson .
the first candidate is the exponential type ( gn ) suggested in @xcite , @xmath94 \frac{1}{1+(b\;{\omega}_{\rm gn})^{2 } } \label{wf - bsp-01 } , \\
{ \phi}_{b_{s}}^{-}(x , b)&= & \frac{f_{b_{s}}}{2\sqrt{2n_{c}}}n_{\rm gn}^{- } { \exp}\big[-\frac{x\,m_{b_{s}}}{{\omega}_{\rm gn}}\big ] \frac{1}{1+(b\;{\omega}_{\rm gn})^{2 } } \label{wf - bsm-01}. \end{aligned}\ ] ] the third candidate is the kkqt type derived form the qcd equations of motion and heavy - quark symmetry constraint @xcite , @xmath100 where @xmath101 @xmath35 @xmath102 @xmath97 @xmath92 . in the above equations eq.([wf - bsp-01][wf - bsm-03 ] )
, @xmath22 denotes the conjugate variables of the transverse momentum of @xmath17 quark in @xmath3 meson .
there is only one parameter @xmath103 for each kind of wfs candidate .
the normalization constants @xmath104 is related to the decay constant @xmath105 through the relation : @xmath106 for the double - light pion , the expression of @xmath107 can be found in @xcite .
for the double - heavy @xmath0 meson , it can be described approximatively by nonrelativistic dynamics . at tree level and in leading order of the expansion in the relative velocities , @xmath22 and @xmath13 quarks in the @xmath0 meson just share the total momentum according to their masses @xcite , @xmath108 where @xmath109 is the decay constant , and @xmath110 @xmath35 @xmath111 .
the @xmath0 @xmath1 @xmath3 transition form factors are defined as follows @xcite : @xmath112 f_{1}^{b_{c}{\to}b_{s}}(q^{2 } ) \nonumber \\ & + & \frac{m_{b_{c}}^{2}-m_{b_{s}}^{2}}{q^{2}}q^{\mu } f_{0}^{b_{c}{\to}b_{s}}(q^{2 } ) \label{ff-01 } , \end{aligned}\ ] ] where @xmath113 @xmath35 @xmath114 @xmath97 @xmath115 is the momentum transfer .
usually , the longitudinal form factor @xmath116 is compulsorily equal to the transverse form factor @xmath117 in the largest recoil limit to cancel singularities appearing at the pole @xmath118 @xmath35 @xmath93 , i.e. , @xmath119 @xmath35 @xmath120 . within the pqcd framework
, the one - gluon - exchange diagrams contributing to the @xmath0 @xmath1 @xmath3 transition form factors are displayed in fig.[fig01 ] .
it has been shown that the pqcd approach works ideally in the large recoil region @xcite .
the expression of form factors is : @xmath121 \nonumber \\ & & + e_{b}(t_{b}){\alpha}_{s}(t_{b } ) h_{b}({\alpha},{\beta}_{b},b_{2})\ , x_{1}{\phi}_{b_{c}}^{-}(x_{1 } ) \nonumber \\ & & { \times } \big[{\phi}_{b_{s}}^{-}(x_{2})\big(1-r_{b_{s}}\big ) + { \phi}_{b_{s}}^{+}(x_{2})r_{b_{s}}^{2}\big]\big\ } \label{ff-11 } , \end{aligned}\ ] ] where @xmath122 @xmath35 @xmath123 is color factor .
@xmath124 @xmath35 @xmath125 is the sudakov factor .
@xmath126 where @xmath127 @xmath35 @xmath128 is the quark anomalous dimension .
the expression of @xmath129 can be found in @xcite .
the hard - scattering kernel function @xmath130 in eq.([ff-11 ] ) is defined as : @xmath131 where @xmath132 and @xmath133 are the modified bessel functions ; @xmath54 and @xmath55 are the virtualities of internal gluons and quarks , respectively .
@xmath134 \label{ff-21 } , \\ { \beta}_{a}&=&-m_{b_{c}}^{2}\big[1+r_{b_{s}}^{2}\bar{x}_{2}^{2 }
-\bar{x}_{2}\big(1+r_{b_{s}}^{2}\big)-r_{c}^{2}\big ] \label{ff-22 } , \\ { \beta}_{b}&=&-m_{b_{c}}^{2}\big[r_{b_{s}}^{2}+\bar{x}_{1}^{2 }
-\bar{x}_{1}\big(1+r_{b_{s}}^{2}\big)\big ] \label{ff-23 } , \\
t_{i}&=&{\max}(\sqrt{{\vert}{\alpha}{\vert } } , \sqrt{{\vert}{\beta}_{i}{\vert}},1/b_{2 } ) \label{ff-24}. \end{aligned}\ ] ]
within the pqcd framework , the feynman diagrams for @xmath0 @xmath1 @xmath2 decay are shown in fig.[fig02 ] , where ( a ) and ( b ) are factorizable topology , ( c ) and ( d ) are nonfactorizable topology .
after a straightforward calculation , the decay amplitudes is written as : @xmath135 where the explicit expressions of @xmath136 are collected in appendix [ app01 ] . from these expressions , we see that only the twist-2 light - cone distribution amplitude ( lcda ) @xmath137 of pion contributes to nonfactorizable decay amplitude .
the expression of the twist-2 pion lcda is defined in gegenbauer polynomials @xcite as : @xmath138 where the nonperturbative parameter @xmath139 is the gegenbauer moment .
the branching ratio in the @xmath0 meson rest frame can be written as : @xmath140 where the common momentum @xmath141 @xmath35 @xmath142 .
the input parameters in our numerical calculation are collected in table [ tab01 ] .
if not specified explicitly , we shall take their central values as the default input . our results on form factor @xmath143(0 ) and
branching ratio @xmath5 are listed in table [ tab02 ] , where the first uncertainty comes from the mass of charm quark ; the second uncertainty comes from the shape parameter of @xmath3 meson wfs , that is , @xmath144
@xmath35 @xmath145 gev in eq.([wf - bsp-01]-[wf - bsm-01 ] ) , @xmath146 @xmath35 @xmath147 gev in eq.([wf - bsp-02]-[wf - bsm-02 ] ) and @xmath102 @xmath35 @xmath148 in eq.([wf - bsp-03]-[wf - bsm-03 ] ) ; the third uncertainty comes from the choice of hard scales @xmath149 in eq.([ff-24 ] ) .
we can see that the large uncertainty come from the @xmath3 wfs .
in addition , the decay constants @xmath109 and @xmath105 will bring some @xmath150 uncertainty to the form factor . for the form factor @xmath4 , our study show that : firstly , the interference between fig.[fig01](a ) and ( b ) is constructive . compared with fig.[fig01](b ) , the contribution of fig.[fig01](a ) is @xmath151 @xmath152 ; secondly , both @xmath153 and @xmath154 contribute to the form factor , and their interference is constructive . compared with @xmath154 ,
the contribution of @xmath153 is @xmath151 @xmath155 ; thirdly , the interference between @xmath156 and @xmath157 is constructive .
compared with @xmath156 , the contribution of @xmath157 is @xmath24 @xmath155 for both gn and kls type ; fourthly , the form factor @xmath4 is sensitive to the shape parameter of @xmath3 wfs ; fifthly , by keeping the parton transverse momentum @xmath62 , and employing the sudakov factors to suppress the long distance contribution in large @xmath22 region @xcite , the form factor @xmath4 is perturbatively calculable with the pqcd approach .
the contribution to form factor comes completely from @xmath43 @xmath24 @xmath158 region with the scale of eq.([ff-24 ] ) ; lastly , considering the uncertainties , our results is reasonable agreement with the previous results listed in table [ tab03 ] . for the branching ratio @xmath5 , our study show that : firstly , the dominated contribution comes from the factorizable topology fig.[fig02](a , b ) .
the interference between nonfactorizable diagrams fig.[fig02](c ) and ( d ) is destructive ; secondly , the main uncertainty is from @xmath3 wfs . considering the uncertainties ,
our results is basically comparable with the previous results listed in table [ tab03 ] , and is also agreement with recent lhcb estimate @xmath5 @xmath29 10% @xcite .
herein we consider three models ( exponential , gaussian , and kkqt type ) of @xmath3 meson wfs , and study the @xmath0 @xmath1 @xmath3 transition form factor and the branching ratio for the @xmath0 @xmath1 @xmath2 decay with the pqcd approach .
we find that ( 1 ) by keeping the parton transverse momentum @xmath62 and employing the sudakov factors , @xmath4 and @xmath5 are perturbatively calculable within the pqcd framework . the contribution to form factor comes completely from @xmath43 @xmath24 @xmath158 region .
( 2 ) @xmath4 and @xmath5 are sensitive to the model and the shape parameter @xmath103 of the @xmath3 meson wave functions . ( 3 ) with appropriate inputs , our estimates on @xmath4 and @xmath5 are comparable with the previous results .
the branching ratio for the @xmath0 @xmath1 @xmath2 decay is about a few per cent , which is agreement with the recent lhcb measurement . with the running of lhc
, the more data will be accumulated at lhcb , the more precise branching ratio for the @xmath0 @xmath1 @xmath2 decay will be obtained , and various models for hadron wave function and factorization treatments for hadron matrix element will be more stringently tested .
this work was supported by _ national natural science foundation of china _ under grant nos . 11147008 , u1232101 and 11275057 ) .
we thank prof .
hsiangnan li , prof .
caidian l , prof .
zhenjun xiao for their helpful discussion .
abe f , et al , ( cdf collaboration ) .
observation of @xmath0 meson in @xmath171 collisions at @xmath172 @xmath35 @xmath7 tev .
phys rev d , 1998 , 58 : 112004 ; phys rev lett , 1998 , 81 : 2432 - 2437 .
aaltonen t , et al , ( cdf collaboration ) .
observation of the decay @xmath173 @xmath1 @xmath174 and measurement of the @xmath173 mass .
phys rev lett , 2008 , 100 : 182002 .
aaij r , et al,(lhcb collaboration )
. measurements of @xmath12 production and mass with the @xmath12 @xmath1 @xmath175 decay .
phys rev lett , 2012 , 109 : 232001 ; aaij r , et al , ( lhcb collaboration ) .
observation of @xmath12 @xmath1 @xmath176 and @xmath12 @xmath1 @xmath177 decays .
phys rev d , 2013 , 87 : 112012 .
abazov v , et al , ( d0 collaboration ) .
measurement of the lifetime of the @xmath178 meson in the semileptonic decay channel .
phys rev lett , 2009 , 102 : 092001 .
abulencia a , et al , ( cdf collaboration ) .
measuerment of @xmath178 meson lifetime using the decay mode @xmath12 @xmath1 @xmath179 .
phys rev lett , 2006 , 97 : 012002 .
lusignoli m , masetti m. @xmath0 decays .
z phys c , 1991 , 51 : 549 - 555 .
chang c , chen y. decays of the @xmath0 meson .
phys rev d , 1994 , 49 : 3399 - 3411 .
beringer j , et al , ( particle data group ) .
review of particle physics .
phys rev d , 2012 , 86 : 010001 .
beneke m , g. buchalla g. @xmath0 meson lifetime .
phys rev d , 1996 , 53 : 4991 - 5000 ; chang c , et al .
lifetime of the @xmath0 meson and some relevant problems .
phys rev d , 2001 , 64 : 014003 ; chang c , et al .
study of the @xmath0 meson lifetime .
commun theor phys , 2001 , 35 : 57 - 60 .
gouz i p , et al
. prospects for the @xmath0 studies at lhcb .
phys atom nucl , 2004 , 67 : 1559 - 1570 .
altarelli m. lhcb : detector performance and first physics results . arxiv:1105.5330 [ hep - ex ] .
aaij r , et al , ( lhcb collaboration ) .
observation of the decay @xmath12
@xmath1 @xmath180 .
phys rev lett , 2013 , 111 : 181801 .
beneke m , et al .
qcd factorization for @xmath181 @xmath1 @xmath182 decays : strong phase and @xmath183 violation in the heavy quark limit .
phys rev lett , 1999 , 83 : 1914 - 1917 ; beneke m , et al .
qcd factorization for exclusive non - leptonic @xmath181 meson decay : general argumengs and the case of hevay - light final states .
nucl phys b , 2000 , 591 : 313 - 418 ; beneke m , et al .
qcd factorization in @xmath181 @xmath1 @xmath184 , @xmath182 decays and extraction of wolfenstein parameters .
nucl phys b , 2001 , 606 : 245 - 321 ; du d , yang d , zhu g. analysis of the decay @xmath181 @xmath1 @xmath182 and @xmath184 with qcd factorization in the heavy quark limit .
phys lett b , 2000 , 488 : 46 - 54 ; du d , yang d , zhu g. infrared divergence and twist-3 distribution amplitudes in qcd factorization for @xmath181 @xmath1 @xmath185 .
phys lett b , 2001 , 509 : 263 - 272 ; du d , yang d , zhu g. qcd factorization for @xmath181 @xmath1 @xmath185 .
phys rev d , 2001 , 64 : 014036 .
chang c , li h. three - scale factoriztion theorem and effective field theory : analysis of nonleptonic heavy meson decay .
phys rev d , 1997 , 55 : 5577 - 5580 ; yeh t , li h. factorization theorems , effective field theory , and nonleptonic heavy meson decays .
phys rev d , 1997 , 56 : 1615 - 1631 ; keum y , li h , sanda a. fat penguins and imagiary penguins in perturbative qcd .
phys lett b , 2001 , 504 : 6 - 14 ; keum y , li h , sanda a. penguin enhancement and @xmath181 @xmath1 @xmath186 decays in perturbative qcd .
phys rev d , 2001 , 63 : 054008 ; keum y , li
h. nonleptonic charmless @xmath181 decays : factorization versus perturbative qcd .
phys rev d , 2001 , 63 : 074006 ; l c , ukai k , yang m. branching ratio and @xmath183 violation of @xmath181 @xmath1 @xmath182 decays in the perturbative qcd appraoch .
phys rev d , 2001 , 63 : 074009 ; l c , yang m. @xmath181 @xmath1 @xmath187 , @xmath188 decays in perturbative qcd approach .
eur phys j c , 2002 , 23 : 275 - 287 ; hong b , l , c. direct cp violation in hadronic @xmath181 decays .
sci china g , 2006 , 49 : 357 - 366 .
bauer c , fleming s , luke m. summing sudakov logarithms in @xmath181 @xmath1 @xmath189 in effective field theory .
phys rev d , 2001 , 63 : 014006 ; bauer c , et al .
an effective field theory for collinear and soft gluons : heavy to light decays .
phys rev d , 2001 , 63 : 114020 ; bauer c , stewart i. invariant operators in collinear effective theory .
phys lett b , 2001 , 516 : 134 - 142 ; bauer c , pirjol d , stewart i. soft - collinear factorization in effective field theory .
phys rev d , 2001 , 65 : 054022 ; bauer c , et al .
hard scattering factorization from effective field theory .
phys rev d , 2002 , 66 : 014017 ; beneke m , et al .
soft - collinear effective theory and heavy - to - light currents beyond leading power .
nucl phys b , 2002 , 643 : 431 - 476 ; beneke m , feldmann t. multipole - expanded soft - collinear effective theory with non - abelian gauge symmetry
. phys lett b , 2003 , 553 : 267 - 276 ; beneke m , feldmann t. factorization of heavy - to - light form factors in soft - collinear effective theory .
nucl phys b , 2004 , 685 : 249 - 296 .
buchalla g , buras a , lautenbacher m. weak decays beyond leading logarithms .
rev mod phys , 1996 , 68 : 1125 - 1244 ; buras a. weak hamiltonian , @xmath183 violation and rare decays .
arxiv : hep - ph/9806471 .
lepage g , brodsky s. exclusive processes in perturbative quantum chromodynamics .
phys rev d , 1980 , 22 : 2157 - 2198 .
beneke m , feldmann t. symmetry - breaking corrections to heavy - to - light @xmath181 meson form factors at large recoil .
nucl phys b , 2001 , 592 : 3 - 34 .
kawamura h , et al . @xmath181-meson
light - cone distribution amplitudes in the heavy quark limit .
phys lett b , 2001 , 523 : 111 - 116 .
descotes - genon s , sachrajda c. sudakov effects in @xmath181 @xmath1 @xmath190 form factors .
nucl phys b , 2003 , 625 : 239 - 278 .
grozin a , neubert m. asymptotics of heavy - meson form factors .
phys rev d , 1997 , 55 : 272 - 290 .
kurimoto t , li h , sanda a. leading power contributions to @xmath181 @xmath1 @xmath75 , @xmath191 transtion form factos .
phys rev d , 2001 , 65 : 014007 .
kurimoto t. uucertainty in the leading - order perturbative qcd calculations of @xmath181-meson decays .
phys rev d , 2006 , 74 : 014027 .
l c , yang m. @xmath181 to light meson transition form factors calculated in perturbative qcd approach .
eur phys j c , 2003 , 28 : 515 - 523 .
ball p. theoretical update of pseudoscalar meson distribution amplitudes of higher twist : the nonsinglet case .
jhep , 1999 , 01 : 010 .
khodjamirian a , et al .
@xmath181 @xmath1 @xmath190 width and @xmath192 from qcd light - cone sum rules .
phys rev d , 2011 , 83 : 094031 .
bell g , feldmann t. modelling light - cone distribution amplitudes from non - relativistic bound states .
jhep , 2008 , 04 : 061 .
wirbel m , stech b , bauer m. exclusive semileptonic decays of heavy mesons .
z phys c , 1985 , 29 : 637 - 642 .
li h. applicability of perturbative qcd to @xmath181 @xmath1 @xmath193 decays .
phys rev d , 1995 , 52 : 3958 - 3965 .
wei z , yang m. the systematic study of @xmath181 @xmath1 @xmath75 form factors in pqcd approach and its reliablity .
nucl phys b , 2002 , 642 : 263 - 289 .
chiu t , et al , ( twqcd collaboration ) .
beauty meson in lattice qcd with exact chiral symmetry .
phys lett b , 2007 , 651 : 171 - 176 .
khodjamirian a. hadronic form factors : combining qcd calculations with analyticity .
int j mod phys a , 2010 , 25 : 513 - 523 .
sharma n , verma r. predictions of @xmath0 meson decay emitting pseudoscalar and heavy scalar mesons using isgw ii model .
phys rev d , 2010 , 82 : 094014 ; sharma n , dhir r , verma r. branching ratios of the @xmath0 meson decaying to pseudoscalar and axial - vector mesons .
j phys g , 2010 , 37 : 075013 .
naimuddin s , et al .
nonleptonic two - body @xmath0 meson decays .
phys rev d , 2012 , 86 : 094028 .
isgur n , et al .
semileptonic @xmath181 and @xmath193 decays in the quark model .
phys rev d , 1989 , 39 : 799 - 818 .
sun j , et al .
study of @xmath0 @xmath1 @xmath194 , @xmath195 decays with qcd factorization .
phys rev d , 2008 , 77 : 114004 .
barik n , et al .
semileptonic decays of the @xmath0 meson .
phys rev d , 2009 , 80 : 074005 .
choi h , ji c. nonleptonic two - body decays of the @xmath0 meson in the light front quark model and the qcd factorization approach .
phys rev d , 2009 , 80 : 114003 ; choi h , ji c. semileptonic and radiative decays of the @xmath0 meson in the light cone quark model .
phys rev d , 2009 , 80 : 054016 .
ivanov m , korner j , santorelli p.
exclusive semileptonic and nonleptonic decays of the @xmath0 meson .
phys rev d , 2006 , 73 : 054024 .
wang w , shen y , l c. covariant light front approach for @xmath0 transition form factors .
phys rev d , 2009 , 79 : 054012 .
dhir r , sharma n , verma r. flavor dependence of @xmath0 meson form factors and @xmath0 @xmath1 @xmath185 decays .
j phys g , 2008 , 35 : 085002 ; dhir r , verma r. @xmath0 meson form factors and @xmath0 @xmath1 @xmath196 decays involving flavor dependence of thransverse quark momentum .
phys rev d , 2009 , 79 : 034004 .
el - hady a , muoz j , vary j. semileptonic and nonleptonic @xmath0 decays .
phys rev d , 2000 , 62 : 014019 .
huang t , zuo , f. semileptonic @xmath0 decays and charmonium distribution amplitude .
eur phys j c , 2007 , 51 : 833 - 839 ; huang t , et al .
semileptonic @xmath197 decays in the light cone qcd sum rules .
int j mod phys a , 2008 , 23 : 3237 - 3245 .
hernndez e , nieves j , verde - velasco j. study of exclusive semileptonic decays of @xmath198 in a nonrelativistic quark model .
phys rev d , 2006 , 74 : 074008 ; study of semileptonic and nonleptonic decays of the @xmath198 meson .
eur phys j a , 2007 , 31 : 714 - 717 .
liu j , chao k. @xmath0 meson weak decays and @xmath183 violation .
phys rev d , 1997 , 56 : 4133 - 4145 .
choudhury d , kundu a , mukhopadhyaya b. testable signals of charm decays in the @xmath0 meson .
mod phys lett a , 2001 , 16 : 1439 - 1448 .
ebert d , faustov r , galkin v. weak decays of the @xmath0 meson to @xmath3 and @xmath181 mesons in the relativistic quark model .
eur phys j c , 2003 , 32 : 29 - 43 .
gershtein s , et al .
theoretical status of the @xmath0 meson .
arxiv : hep - ph/9803433 ; berezhnoy a , et al .
@xmath0 meson at lhc .
phys atom nucl , 1997 , 60 : 1729 - 1740 .
ivanov m , krner j , santorelli p. semileptonic decays of the @xmath0 meson .
phys rev d , 2001 , 63 : 074010 .
gouz i , et al .
prospects for the @xmath0 studies at lhcb .
phys atom nucl , 2004 , 67 : 1559 - 1570 .
choudhury d , kundu a , mukhopadhyaya b , semileptonic charm decay as a test for the sepctator model in the @xmath0 meson .
arxiv : hep - ph/9810339 .
kiselev v , kovalsky a , likhoded a. @xmath0 decays and lifetime in qcd sum rules .
nucl phys b , 2000 , 585 : 353 - 382 ; phys atom nucl , 2001 , 64 : 1860 - 1875 .
bagan e , et al .
hadrons with charm and beauty . z phys
c , 1994 , 64 : 57 - 71 .
colangelo p , fazio f. using heavy quark spin symmetry in semileptonic @xmath0 decays .
phys rev d , 2000 , 61 : 034012 .
colangelo p , nardulli g , paver n. qcd sum rules calculation of @xmath0 decays .
z phys c , 1993 , 57 : 43 - 50 .
gershtein s , et al .
physics of @xmath0 mesons .
phys usp , 1995 , 38 : 1 - 37 .
nobes m , woloshyn r. decays of the @xmath0 meson in a relativistic quark - meson model . j phys g , 2000 , 26 : 1079 - 1094 .
du d , wang z. predictions of the standard model for @xmath178 weak decays .
phys rev d , 1989 , 39 : 1342 - 1348 .
anisimov a , et al .
exclusive and inclusive decays of the @xmath0 meson in the light front isgw model .
phys atom nucl , 1999 , 62 : 1739 - 1753 . | the @xmath0 @xmath1 @xmath2 decay is studied with the perturbative qcd approach .
three types of wave functions for @xmath3 meson are considered . the transition form factor @xmath4 and the branching ratio @xmath5 are sensitive to the model of the @xmath3 meson wave functions . with appropriate inputs ,
our estimate on @xmath5 is comparable with the recent lhcb measurement .
a clear signal of @xmath0 @xmath1 @xmath2 decay should be easily observed at the large hadron collider .
keywords : @xmath0 meson , weak decay , the perturbative qcd approach , branching ratio . |
it has recently been proposed that a new type of superconductivity exists in the heavy fermion cecu@xmath1si@xmath1 under high pressure@xcite . in this compound ,
the superconducting transition temperature @xmath2 is enhanced near 3@xmath0gpa from its ambient pressure value of 0.7@xmath0k to around 2.5@xmath0k .
the proposed pairing mechanism at high pressure is based on the exchange of critical valence fluctuations close to a ( nearly ) first order valence transition of the ce ion at a pressure @xmath5 4.5@xmath0gpa . at ambient pressure in contrast ,
critical magnetic fluctuations are believed to mediate the superconductivity .
the magnetically ordered state , which in pure cecu@xmath1si@xmath1 competes with the superconducting state is thought to disappear at a magnetic quantum critical point at a small positive pressure @xmath6 of approximately 0.1@xmath0gpa @xcite . this quantum critical point is masked by the superconducting state in pure samples , but can be directly observed by substituting ge for si which leads to a suppression of superconductivity @xcite . the necessary background to this work can be found in ref .
@xcite , which provides the context for the results and discussion reported below , and should be read in conjunction with this paper .
the most important point to remember is that there is a second critical pressure @xmath3 where the occupation number @xmath7 of the electronic @xmath8-orbitals on the ce atom changes abruptly .
this can be thought of as analogous to the ce @xmath9 transition , though with the critical endpoint at very low , or perhaps negative temperature ( the latter leading to a crossover rather than a first order transition ) .
the system goes from a kondo regime ( @xmath10 ) to that with characteristics of an intermediate valence state ( where @xmath11 ) .
the enhancement of @xmath2 near @xmath3 , mediated by critical valence fluctuations , was predicted by an extended anderson model including an extra coulomb repulsion term @xmath12 between the conduction @xmath4- and @xmath8-electrons @xcite .
a series of anomalies in the normal state electronic properties was also predicted , and these have been observed to occur around this pressure .
they include a large enhancement of the residual resistivity @xmath13 , an enhancement of the sommerfeld coefficient @xmath14 , and a resistivity linear in temperature , @xmath15 @xcite .
this simple model is surprisingly successful in describing the observed behaviour in cecu@xmath1si@xmath1 , and can help to provide the basis for a more complete description of the behaviour of related systems .
however , there are further experimental facts beyond its scope , which may prove useful guidance towards a more complete understanding of the microscopic physics involved .
the superconducting and magnetic properties of cecu@xmath1si@xmath1 at ambient pressure are highly sensitive to small changes in composition and disorder @xcite , but there has been little systematic investigation into similar effects at high pressure .
the ground state at ambient pressure depends strongly on the exact stoichiometry of the sample , giving so - called types a , a / s and s , where the magnetically ordered ` a phase ' competes with the superconducting state ` s ' .
ishikawa @xcite recently proposed a further subdivision of cecu@xmath1si@xmath1 properties , into so - called ` low @xmath2 ' and ` high @xmath2 ' types , with differing signs in the pressure dependence of the superconducting transition temperature @xmath2 .
in addition , the isostructural compound cepd@xmath1si@xmath1 has been shown to be extremely sensitive to uniaxial stress under pressure .
dramatic variations in @xmath2 in cepd@xmath1si@xmath1 result from a change in crystalline orientation with respect to small non - hydrostatic components in a bridgman anvil pressure cell with a quasi - hydrostatic steatite medium @xcite . with these facts in mind
, we wished to systematically explore , via resistance measurements under pressure , how the electronic properties around @xmath3 depended on the sample , and on the pressure conditions ; especially to see how valence - fluctuation mediated superconductivity in cecu@xmath1si@xmath1 is affected by the presence of disorder and uniaxial stress .
the results previously obtained in a hydrostatic helium pressure medium @xcite provide a baseline for comparison .
resistivity measurements under pressure on a total of six cecu@xmath1si@xmath1samples from two different sources are reported below .
there were two polycrystalline samples , of type ` low-@xmath2 ' and ` high-@xmath2 ' , respectively labelled # 50 and # 57 and prepared by ishikawa s group by a levitation method ; the remaining four samples were a / s type single crystals from the same original source crystal prepared at the mpi dresden .
the latter crystal were grown in an al@xmath1o@xmath16 crucible by a modified bridgman technique using cu excess as flux medium @xcite .
x - ray powder diffraction confirmed the tetragonal thcr@xmath1si@xmath1 structure with lattice parameters @xmath17 and @xmath18 .
specific heat measurements display a large mean field like anomaly at @xmath19k followed by a large peak at @xmath20k . according to the strong difference in their magnetic field dependence
, these anomalies can be attributed to the transition into the magnetic a - phase and into the superconducting state , respectively .
the large size of the anomalies proves that both transitions are bulk ones .
two different pressure runs are reported in this paper , referred to below as cell # 1 ( containing four samples from all sources ) and # 2 ( with two a / s samples ) using a bridgman anvil cell technique with a quasi - hydrostatic steatite solid pressure medium .
pressure gradients are believed to be less than 5% of the total at the highest pressures , determined from the width of the superconducting transition of pb manometer .
the first cell ( # 1 ) contained the two polycrystals # 50 and # 57 and two a / s type single crystals , oriented with the measurement current parallel to the @xmath4- and @xmath21-axes , and labelled ` a / s @xmath22 ' and ` a / s @xmath23 ' respectively .
the current direction was perpendicular to the axis of the pressure cell , and we believe though unfortunately are not certain , that the @xmath4-axis of sample ` a / s @xmath23 ' was oriented parallel to the cell axis . cell # 2 , containing two a / s samples cut from neighbouring positions on the same source crystal .
they are oriented with their @xmath4-axes parallel and perpendicular to the cell axis , and the current flows along the @xmath21 direction . ]
any uniaxial stresses due to an imperfectly hydrostatic pressure medium are likely to be aligned with the cell axis , and the second pressure cell ( # 2 ) , fig . [
fig : strainphoto ] , was designed to exploit this systematically .
the cell contained two samples , both cut from neighbouring positions on the same a / s type single crystal source .
they were both oriented so that the current flowed along the @xmath21 direction , while the @xmath4-axis was oriented parallel and perpendicular to the cell axis .
the samples have been labelled ( @xmath24 ) and ( @xmath25 ) respectively .
multiple contacts on each sample enabled the resistivity within different regions of the same sample to be compared . as they shared their longest side in the original source crystal , variations in composition along the length of one sample
should be comparable to that between samples .
the resistivity was first measured at p=0 ; excellent agreement was found with previous results .
apart from the superconducting transition temperature @xmath2 , several features of the normal state resistivity under pressure will be highlighted .
these are the residual resistivity @xmath13 and @xmath26 coefficient of the temperature dependence in the power law behaviour @xmath27 of the normal state at low temperature , and the position of the two crystal - field split resistivity maxima @xmath28 and @xmath29 .
all of these show distinctive features at the valence instability pressure @xmath3 .
the resistance has been normalized by the geometric factor of each sample measured on construction of the cell , which usually gives an absolute value within 10% for proper four - point measurements .
contrary to cell # 2 , which provided straightforward results , the gasket of cell # 1 formed under too high a load , and due to migration of the contact wires , full four - point resistance measurements were possible in only one of the samples ( # 50 ) .
the resistance measured for the other samples contained an additional series contribution , of varying size .
it would be possible to subtract this , for example assuming a linear additional term and/or an adjustment of the geometric factor used to obtain the absolute resistivity , but the results obtained will then depend on these assumptions .
we should add that obtaining even semi - quantitative values for the absolute resistivity at such high pressures is a challenging task , therefore we make no apology for trying to extract the maximum amount of information from what might be seen as an imperfect experiment .
normal state resistivity at @xmath2 in cell # 1 , showing the enhancement of @xmath13 at @xmath3 .
note that only sample # 50 had full four - point contacts , so the others contain a monotonically decreasing additional contribution to the resistance , but the sample contribution is visible in all . ] in fig .
[ fig : rho0 ] , one can easily see the enhancement of the residual resistivity at @xmath3 in all samples in cell # 1 , superimposed in three cases on a monotonically decreasing additional series resistance .
it is clear that the enhancement of the residual resistivity in sample # 50 ( the sample with four electrical contacts ) is much larger than in the others , and reaches almost 200@xmath30 cm close to 5@xmath0gpa .
this sample also has a small negative magnetoresistance at 5.5@xmath0gpa , around 0.5% at 8@xmath0 t and 4.2@xmath0k , in contrast to the other samples which show a positive magnetoresistance .
this is typical of cecu@xmath1si@xmath1samples with very high residual resistivities at @xmath3 , and perhaps due to the zeeman shift of the @xmath8-level @xmath31 , moving the system away from the valence instability .
resistivity of the ` low @xmath2 ' ( # 50 , unadjusted ) and ` high @xmath2 ' ( # 57 , adjusted , see text ) samples in cell # 1 at 4.34 gpa .
note the extremely large residual resistivity of # 50 , yet a nearly complete superconducting transition is observed . ] in fig .
[ fig:5057p4 ] , samples # 50 and # 57 are compared at 4.34@xmath0gpa , close to the maximum of @xmath13 . to compensate for experimental difficulties ,
the resistivity of # 57 has been adjusted slightly to give zero resistance below @xmath2 , and to match that of # 50 at room temperature .
this was done by subtracting a small constant contact resistance and slightly adjusting the geometric factor , each corresponding to about 10% of the unadjusted data .
the resulting resistivity curve is typical of cecu@xmath1si@xmath1 at this pressure . in sample # 50 however , there is evidently an enormously enhanced impurity contribution to the total resistivity , with a negative temperature dependence reminiscent of a kondo impurity system , in addition to the usual scattering which increases up to @xmath32 .
a negative slope becomes apparent at higher pressure where the @xmath26 coefficient of the resistivity has collapsed and the impurity contribution dominates the resistivity .
the two samples regain very similar behaviour on approaching room temperature .
we should add that this sort of behaviour , where superconductivity coexists with very high residual resistivity , has also been observed in single crystals @xcite .
the existence of samples which superconduct at high pressure despite such enormous residual resistivities , of the order of the ioffe - regel limit ( which is around 100@xmath30 cm at ambient pressure ) , is strong evidence in favour of valence fluctuation mediated pairing .
the common view is that unconventional superconductivity is incompatible with large electronic scattering rates .
however , in this case both the superconductivity and the enhanced impurity scattering may share a common origin , as we believe that valence fluctuations are responsible for both the pairing interaction and the renormalization of impurity potentials @xcite .
the superconducting pressure range is extended considerably in three out of four samples .
only sample ` a / s @xmath23 ' showed the expected total disappearance of superconductivity above 7@xmath0gpa . ]
figure [ fig : tconset ] shows the onset temperature of the superconducting transition @xmath33 for the samples in cell # 1 .
the most remarkable feature is that in three out of four samples , the superconducting region extends up to a much higher pressure than seen in hydrostatic conditions . only in sample ` a / s @xmath23 ' was @xmath33 suppressed at high pressures similar to the result observed in the helium cell .
the transitions at the highest pressures are partial , but there is a clear drop in resistance below @xmath33 .
very broad resistive transitions are intrinsic to cecu@xmath1si@xmath1 at high pressure , and have been identified as due to filamentary superconductivity rather than to pressure gradients @xcite .
the superconducting state at very high pressure may therefore be solely of a filamentary nature .
the enhanced @xmath33 at very high pressure is a clear indication of the sensitivity of the high pressure superconducting state to anisotropy , either via the current direction , or anisotropic pressure conditions .
cell # 2 was designed to test the effect of anisotropic stress , in analogy to similar work on cepd@xmath1si@xmath1 @xcite .
the two samples were aligned at right angles with respect to the cell axis .
we might expect the slight uniaxial stress associated with the solid pressure medium to be oriented along this axis .
the superconducting and normal state properties did indeed differ between the two samples in a way which corresponded to more than just a shift of the pressure scale , which might be expected if each sample was simply sampling part of an inhomogeneous pressure distribution .
onset @xmath33(dotted lines ) , and completion @xmath34 ( continuous lines ) temperatures in cell # 2 compared with helium cell results .
two results are shown for samples ( @xmath24 ) ( open circles ) and ( @xmath25 ) ( filled squares ) , taken from measurements using different contacts on the same sample .
the differences in @xmath2 between samples are significantly larger than local variations within a single sample . ]
figure [ fig : straintc ] shows the onset @xmath33 , and completion @xmath34 temperatures of the resistive superconducting transition for samples ( @xmath24 ) and ( @xmath25 ) .
@xmath34 has been previously identified with the bulk superconducting transition @xcite .
it is clear that the extension of @xmath33above 6@xmath0gpa seen in cell # 1 was not reproduced , indeed both samples in cell # 2 follow the pressure dependence of ` a / s @xmath23 ' fairly closely .
however , the difference in @xmath2 between the two samples is an order of magnitude larger than within each .
there is a slight enhancement of @xmath2 in sample ( @xmath24 ) , and a very clear difference in the shape of the resistive transitions ( see @xcite for details ) .
the lack of extension of the superconducting region to very high pressure in cell # 2 does not entirely rule out uniaxial stress as the explanation for such behaviour in cell # 1 ; the gasket instability in the latter probably lead to higher non - hydrostatic stresses inside the pressure cell .
the normal state properties provide a much clearer picture of the effect of uniaxial stress on cecu@xmath1si@xmath1 under pressure .
this can be most simply described as a shift of the valence instability @xmath3 to higher pressure in sample ( @xmath25 ) than in ( @xmath24 ) .
we have previously identified @xmath3 with several anomalies in the normal state resistivity , including : * merging of the crystal field split kondo resistivity maxima @xmath28 and @xmath29 . * a maximum in the residual resistivity @xmath13 . * a sudden change in the @xmath26 coefficient of the resistivity @xmath35 , and a change of the @xmath36 scaling .
the shift of @xmath3 is evident in all three properties , as shown in figs .
[ fig : tmaxrho0 ] & [ fig : strainavstmax ] . in fig .
[ fig : tmaxrho0](a ) , one can see that @xmath28 , proportional to the kondo temperature @xmath37 , rises faster in sample ( @xmath24 ) than in ( @xmath25 ) .
@xmath29 , which reflects the crystal field splitting @xmath38 , is difficult to distinguish above the lowest pressure , but it has been shown to remain more or less constant @xcite .
the two crystal - field split resistivity peaks merge into a single maximum at different pressures in the two samples , and this corresponds approximately to the @xmath13 maximum .
the latter is shown in fig . [ fig : tmaxrho0](b ) , where a lorentzian peak can be fitted to the variation of @xmath13 . note that the values of @xmath13 plotted in fig .
[ fig : tmaxrho0](b ) are extracted from a power law fit to the resistivity above @xmath2 , so the very low values at 0.6@xmath0gpa may be slightly anomalous .
the presence of the a phase at ambient pressure also makes direct comparison difficult .
( a ) the crystal field split kondo resistivity maxima @xmath28 and @xmath29 ( defined in inset ) merge at a higher pressure in sample ( @xmath25 ) ( filled symbols ) than in ( @xmath24 ) ( empty symbols ) .
the uncertainty on @xmath28 is smaller than the symbol size .
( b ) the residual resistivity maxima in the two samples corresponds approximately to the pressures determined in ( a ) .
@xmath3 is therefore at a higher pressure in ( @xmath25 ) than in ( @xmath24 ) .
the same symbols at a given pressure correspond to measurements from different contacts on the same sample .
] there is a drop in @xmath39 of more than an order of magnitude at the valence transition , at different pressures for each sample .
the inset shows the corresponding change of the @xmath40 scaling expected due to the kadowaki - woods relation ( solid lines ) .
he cell results ( open circles ) using a quadratic fit are included in the inset . ]
the shift of @xmath3 determined above corresponds to a difference of around 0.5@xmath0gpa between ( @xmath25 ) and ( @xmath24 ) .
this is confirmed when the @xmath26 coefficient of the resistivity is examined ( fig .
[ fig : strainavstmax ] ) .
the analysis is slightly complicated by the non - fermi liquid behaviour of the resistivity found in the entire region around @xmath6 and @xmath3 , where exponents @xmath41 are found in power law resistivity fits @xmath42 .
however in this case a reasonable comparison is still possible on a logarithmic scale , since there is a drop in @xmath39 of over an order of magnitude at @xmath3 , and even without forcing a quadratic fit , the variations in @xmath43 can only contribute a maximum factor of two to the result . at higher pressure ( above 6@xmath0gpa ) , the temperature dependent impurity scattering dominated the resistivity and made it impossible to obtain a reliable fit . the inset of fig .
[ fig : strainavstmax ] , where @xmath39 is plotted against @xmath28 , indicates the crossover from the strongly to weakly correlated branches of the kadowaki - woods plot @xcite . a comparison with the helium cell results from ref .
@xcite ( where a quadratic fit was forced ) shows that the drop in @xmath39 around 5@xmath0gpa corresponds to the transition between the two regions where the expected @xmath40 scaling is followed ( indicated by solid lines ) .
the result is clear .
the two samples track each other up to 4.5@xmath0gpa , after which the drop in @xmath39 occurs more quickly in ( @xmath24 ) than ( @xmath25 ) , confirming that ( @xmath24 ) has reached the valence transition at a lower pressure .
ishikawa attributed the differences at ambient pressure between ` low @xmath2 ' and ` high @xmath2 ' type cecu@xmath1si@xmath1 to different pairing symmetries @xcite .
normal state and superconducting properties at ambient pressure differed considerably between the two categories of sample , which were located in regions side by side the homogeneity range @xcite .
the ` low @xmath2 ' samples were characterized by a slight excess of copper , and a higher residual resistivity .
however , specific heat measurements showed a more robust superconducting transition in the ` low @xmath2 ' than the ` high @xmath2 ' samples . despite the difference in @xmath44 previously observed close to @xmath45 , the ` low @xmath2 ' and ` high @xmath2 ' samples both displayed the usual enhancement of @xmath2 at high pressure .
the most striking difference was seen in the residual resistivity , which reached huge levels in sample # 50 around @xmath3 .
given that both samples had very similar but not identical compositions , we might speculate on the origin of this . according to refs .
@xcite , there appeared to be no significant difference in crystal structure between the two categories of sample .
the exact location of any atomic disorder may be highly significant , as the enhancement of impurity scattering is due to critical fluctuations @xcite and the effect may well depend significantly on the nature and location of the impurity itself , not only at @xmath3 , but also @xmath6 . the effect of disorder on @xmath2 noted by yuan et al .
@xcite , suggests that the initial reduction in @xmath2with pressure in the ` low @xmath2 ' samples is due to such ( possibly enhanced ) impurity scattering , suppressing superconductivity between the two critical points .
the a / s type cecu@xmath1si@xmath1 shows a similar enhancement of @xmath2 at high pressure to the sample measured in hydrostatic conditions .
all other samples , including those of other types , have shown the enhancement of @xmath2 when pressurized to over 3@xmath0gpa .
regarding both the properties at ambient pressure and around @xmath3 , the evidence so far seems to suggest only that the a@xmath46a / s@xmath46s series represents a slight shift of the pressure scale .
thus any individual cecu@xmath1si@xmath1 sample can be classified by two or maybe three variables , corresponding to the aforementioned pressure shift , and to the impurity concentration ( and perhaps also to the nature of the disorder ) . regarding the effect of uniaxial stress on cecu@xmath1si@xmath1 , monthoux and lonzarich
have shown that a more anisotropic structure leads to an increase in @xmath2 for both magnetic and density fluctuation mediated pairing @xcite . in the ce115 systems , and related pu compounds , it has also been shown that @xmath2 is strongly dependent on the ratio of the tetragonal lattice parameters @xmath47 @xcite .
while the exact effect of the quasihydrostatic medium is hard to quantify , our observations are not inconsistent with these scenarios .
this does not necessarily help to distinguish between spin and valence - fluctuation mediated superconductivity , but it provides a constraint which must be satisfied by any more complete theoretical model .
further avenues for exploration from both a theoretical and experimental point of view come from the highly anisotropic resistivity , including some evidence that @xmath2 , as defined by zero resistivity , may depend on current direction !
one clear result not so far mentioned is that the height of the resistivity peak at @xmath28 is significantly larger for @xmath23 than @xmath22 .
this is not the case for the magnitude of @xmath48 , reflecting the higher symmetry of the higher - lying crystal field split @xmath8-states .
the effect of the cef on the anisotropy of the resistivity in cecu@xmath1si@xmath1was analysed in ref .
@xcite .
finally , we would like to emphasize again the importance of the coexistence of superconductivity and very high residual resistivity shown in fig .
[ fig:5057p4 ] .
this is perhaps the most compelling evidence so far for a novel pairing mechanism in cecu@xmath1si@xmath1 at high pressure .
resistivity measurements were carried out at high pressure on several different cecu@xmath1si@xmath1 samples , in the presence of small non - hydrostatic stress components .
all samples showed an enhancement of superconductivity around 23@xmath0gpa , and also an enhancement of residual resistivity with a maximum around 5@xmath0gpa .
complete resistive superconducting transitions can coexist with residual resistivities of the order of 150200@xmath30 cm around the valence instability close to 5@xmath0gpa .
individual cecu@xmath1si@xmath1 samples , usually classified at ambient pressure by the presence or otherwise of an ordered ` a ' phase and/or superconductivity , belong to the same continuum , which can be traversed by pressure and/or ge substitution .
disorder can also be used to classify samples , and the enhancement of residual resistivity under pressure probes this .
it may be a fruitful avenue for future research to examine the effect of changing the nature of the impurities .
uniaxial stress was found to have a significant effect on both the normal and superconducting properties around the valence instability , which can be summarized by a shift of the valence instability pressure @xmath3 .
ath wishes to thank the japan society for the promotion of science for financial support during part of the preparation of this manuscript .
mi thanks dr .
p. ahmet , now at national institute for materials science for his help in preparing samples by a levitation method . | resistivity measurements were carried out up to 8@xmath0gpa on single crystal and polycrystalline samples of cecu@xmath1si@xmath1 from differing sources in the homogeneity range . the anisotropic response to current direction and small uniaxial stresses was explored , taking advantage of the quasi - hydrostatic environment of the bridgman anvil cell .
it was found that both the superconducting transition temperature @xmath2 and the normal state properties are very sensitive to uniaxial stress , which leads to a shift of the valence instability pressure @xmath3 and a small but significant change in @xmath2 for different orientations with respect to the tetragonal @xmath4-axis .
coexistence of superconductivity and residual resistivity close to the ioffe - regel limit around 5@xmath0gpa provides a compelling argument for the existence of a valence - fluctuation mediated pairing interaction at high pressure in cecu@xmath1si@xmath1 . |
the strong degeneracy between the curvature of the universe and the dark energy equation of state causes difficulties for constraining the two parameters simultaneously .
the curvature is commonly left out in a dark energy analysis , or conversely , a dark energy constant is assumed in a determination of the curvature .
however , a simple flatness assumption may result in erroneously reconstructing the dark energy equation of state even if the true curvature is very small @xcite , and a cosmological constant assumption may arise confusions between a dynamical dark energy non - flat model and the flat @xmath7cdm model @xcite . in @xcite , when arguing the defects of a zero curvature assumption , a direct curvature determination by combining measurements of the hubble parameter @xmath1 and the comoving angular diameter distance @xmath8 was proposed : @xmath9 ^ 2-c^2}{[h_0d(z)]^2 } \label{eq : curv}\ ] ] where @xmath10 denotes derivative with respect to redshift @xmath11 .
this formula benefits from the baryon acoustic oscillation ( bao ) measurements which can provide @xmath1 and @xmath2 simultaneously at the same redshift . here
@xmath2 is the angular diameter distance , which is correlated to the comoving angular diameter distance by @xmath12 . since the derivation of @xmath1 and @xmath2 pairs from bao measurements are purely geometrical , eq.([eq : curv ] )
can be evaluated without any assumption of dynamic evolution of universe , therefore breaks the degeneracy between curvature and dark energy equation of state .
several works have already derived the @xmath1 and @xmath2 pairs from the data of wigglez dark energy survey at @xmath13 @xcite , and the third generation sloan digital sky survey ( sdss - iii ) at @xmath14 @xcite and @xmath15 @xcite . with the quasar - lyman @xmath3 forest in sdss - iii
, the measurement of @xmath1 and @xmath2 has extended to high redshift such as @xmath4 @xcite .
these data can afford us to directly determine the curvature parameter via eq.([eq : curv ] ) .
the remaining issue is to estimate the @xmath16 reasonably .
@xcite has applied a novel method to take the derivative of the luminosity distance @xmath17 with respect to the scale factor in their model - independent reconstruction of hubble parameter .
they decomposed the observables into a suitable basis functions , then recombined the derivatives of the basis functions to yield the @xmath18 .
the basis system was whereafter optimized by @xcite to be capable to describe cosmologies independently of their background physics and improve the quality of the estimation of @xmath18 .
this method is independent to any cosmology model and can be employed by us to estimate the @xmath16 . by combining the data and method described above
, we can determine the curvature parameter in a model - independent manner .
this approach is different from previous works employing smoothing procedures in redshift bins or reconstruction of both hubble parameter and comoving angular diameter distance ( e.g. * ? ? ?
the property of eq.([eq : curv ] ) has determined that the error on the measured curvature parameter decreases as the redshift increases , thus we can benefit from the bao lyman @xmath3 forest measurement which can provide @xmath1 and @xmath2 pair at high redshift .
the structure of this paper is as follows . in section
[ sec : method ] , we review the essential parts of the model - independent method . the description of data and application of the method are shown in section [ sec : application ] .
the discussions are presented in section [ sec : discussion ] and the conclusions are drawn in section [ sec : conclusion ] .
we follow the idea in @xcite , and use the method proposed in @xcite and further developed by @xcite , to determine the curvature parameter in a model - independent manner . in friedmann - robertson - walker metric
, the comoving angular diameter distance is written as @xmath19 where the hubble parameter @xmath1 at late time is given by @xmath20^\frac{1}{2 } \label{eq : h(z)}\ ] ] the parameters @xmath21 , @xmath0 and @xmath22 are the current density of matter , curvature and dark energy in unit of the critical density respectively .
the function @xmath23 depends on the ratio @xmath24 between the pressure and the density of the dark energy . by combining eq.([eq : h(z ) ] ) and eq.([eq : f(z ) ] ) , the degeneracy between curvature and dark energy can be easily seen , as the non - zero curvature can be mimicked by the model with zero curvature and a dark energy component with @xmath25 , therefore results in difficulties in fitting eq.([eq : d(z ) ] ) to observables .
in fact , by solving the integral part from eq.([eq : d(z ) ] ) and taking derivative with respect to redshift on both side of the integral , one can derive eq.([eq : curv ] ) .
the dynamic assumption of the universe thus has no place in the determination of the curvature .
the key issue here is to estimate @xmath16 the derivative of the comoving angular diameter distances with respect to the redshift in eq.([eq : curv ] ) .
we suppose there is an underlying function @xmath8 describing the behaviors of the comoving angular diameter distances . the @xmath8 could be expanded into a series of suitable orthonormal functions @xmath26 , @xmath27 by fitting eq.([eq : decomp ] ) to the obsevables @xmath28 , the @xmath29 coefficients @xmath30 are determined .
the number of the terms to be included in the expansion depends on the choice of the orthonormal basis and the quality of the data .
the derivative of eq.([eq : decomp ] ) is then taken as the estimation of the derivative in eq.([eq : curv ] ) .
the basis @xmath31 , in principle , could be arbitrary with ideal data , but it is not in practice .
@xcite used an arbitrary orthonormal basis generated by gram - schmidt orthonormalization of @xmath32 to decompose the luminosity distances .
they observed a systematic trend on the slope of the reconstructed @xmath1 at intermediate redshifts when compared with the predictions of a @xmath7cdm cosmology , although they were consistent within the error bars .
it indicated that a randomly chosen system of orthonormal basis functions may not be well adapted to the behavior of the measured data .
@xcite has proposed an optimal basis system derived from principal component analysis ( pca ) . under this optimal basis system ,
the number of coefficients @xmath29 in eq.([eq : decomp ] ) reaches minimal and the possible bias introduced by the choice of the basis is removed @xcite .
the derivation of the optimal basis starts with writing the comoving angular diameter distances and their redshifts in column vectors @xmath33 and @xmath34 respectively .
the data set @xmath33 is therefore considered as a single point in an @xmath35-dimensional space , where @xmath35 is the number of the data points .
then we select a group of models that are believed to span the viable cosmologies , and calculate the comoving angular diameter distances @xmath8 for each model at the redshifts in @xmath34 to generate a set of vectors @xmath36 @xmath37 , where @xmath29 is the number of the models .
these new vectors correspond to a cluster of points in the @xmath35-dimensional space and should meet the condition that the data set @xmath33 is tightly enclosed in the distribution of the cluster @xcite .
this ensemble of models , referred as the _ training set _
, samples the possible behaviors of the comoving angular diameter distances , and initializes the pca .
once the training set models @xmath38 @xmath39 are defined , we build the so - called _ scatter matrix _ @xmath40 with @xmath41 .
@xmath42 contains the differences between each training vector @xmath36 .
@xmath43 is the reference model that defines the origin of the @xmath35-dimensional space .
@xmath44 is actually the deviation of the training set from the reference model .
@xmath43 could in principle be any combination of \{@xmath36 } , and usually be set the average of the training set @xmath45 .
any other choice of a reference model only affects the number of principal components ( pcs ) in decomposing the comoving angular diameter distances , e.g. @xmath29 in eq.([eq : decomp ] ) .
it does not affect the final reconstruction of @xmath8 .
the pcs are derived by solving the eigenvalue problem @xmath46 , where @xmath47 and @xmath48 are the @xmath49-th eigenvalue and eigenvector respectively .
the eigenvector of the largest eigenvalue is the first pc .
it corresponds to the direction on which the projection of @xmath44 has the largest variance , which means that the cluster of points mainly align in this direction .
the second pc is the eigenvector of the second largest eigenvalue corresponding to the direction of the secondary largest variance , and so on .
since pca aims at reducing the dimensionality of the training set substantially while retaining almost all the variation , an important issue arises that how many pcs should be employed in the reconstruction of @xmath8 .
the selection criterion is based on the cumulative percentage of total variation @xcite defined as : @xmath50 where @xmath51 are the eigenvalues corresponding to the pcs and sorted in a descendent sequence @xmath52 , @xmath35 is the number of total pcs , and @xmath53 is the number of pcs we shall employ in the reconstruction .
the @xmath54 , varying between 0 and 100 , quantifies how much percentage of variance in the training set is preserved in the first @xmath53 pcs . after setting a threshold ( e.g. @xmath55 )
, @xmath54 returns the suitable number of pcs .
then the deviation of the @xmath8 from the reference model is decomposed by the @xmath53 pcs @xmath56 eq.([eq : decomp_pca ] ) is similar with eq.([eq : decomp ] ) , except that the reference model here is not zero and the basis @xmath48 is optimized via pca .
the coefficients @xmath57 are determined by fitting eq.([eq : decomp_pca ] ) to @xmath33 through @xmath58 minimization .
the optimal basis for decomposition of @xmath8 is also the optimal basis for decomposition of the @xmath59-independent angular comoving distances @xmath60 .
when build the training set @xmath61 with @xmath59-independent angular comoving distances @xmath62 , eq.([eq : decomp_pca ] ) can be rewritten as , @xmath63 where the @xmath64 is the @xmath59-independent reference model . in the @xmath58 minimization step ,
we use eq.([eq : decomp_pca2 ] ) to leave the hubble constant @xmath59 as a free parameter .
the @xmath16 in eq.([eq : curv ] ) is derived by taking derivative on both side of eq.([eq : decomp_pca2 ] ) with respect to redshift .
the @xmath1 and @xmath2 pairs are collected from current literatures of bao measurements .
since the sdss and boss cmass samples have been analyzed multiple times , we tend to use the most recent published results to avoid overlap .
the @xmath65 pairs and the bao surveys are listed in table [ tab : bao ] .
we consider the covariance between distances constrained by different bao samples at different redshift ranges to have negligible effect on our determination of derivative of distance with respect to redshift .
llll 0.44 & @xmath66 & @xmath67 & wigglez + 0.6 & @xmath68 & @xmath69 & + 0.73 & @xmath70 & @xmath71 & + 0.35 & @xmath72 & @xmath73 & sdss dr7 + 0.57 & @xmath74 & @xmath75 & boss dr11 cmass + 2.36 & @xmath76 & @xmath77 & boss dr11 ly-@xmath3 forest to lower the statistical errors in @xmath16 estimation , we enlarge the distance samples with the luminosity distances @xmath78 from the currently largest homogeneously reduced compilation of sn ia , the union2.1 @xcite , which contains 580 sne ia .
the data are in form of distance modulus @xmath79 which should be converted to the luminosity distance by @xmath80 the corresponding comoving angular diameter distances are derived via the etherington s relation which holds for any space - time @xmath81 . for simplicity , we use the covariance matrix of distance modules of sne ia with statistical errors only , which is in diagonal form , during the @xmath58 minimization procedure .
fig.[fig : pcs ] has depicted the first four pcs for the comoving angular diameter distances .
the training set @xmath61 is built from 1000 non - flat @xmath7cdm models with the parameters uniformly sampled in the cubic parameter space with boundaries @xmath82 and @xmath83 , the sum of @xmath21 , @xmath22 and @xmath0 can deviate from 1 . here
the reference model takes the average of the training set @xmath84 .
a different reference model would not change the reconstruction of @xmath8 as long as the distribution of the training set tightly enclosing the data @xcite .
the first pc retains 98.8% of the total variance in the sample , which means it has already considered the major properties in the expansion of the training set .
we therefore use the first pc to decompose the @xmath8 . here
we emphasis that , although the pcs are determined by the training set sampled from different non - flat @xmath7cdm models , they are able to constrain other cosmologies that are not explicitly contained in the training set ( see * ? ? ?
fitting eq.([eq : decomp_pca2 ] ) to the observables @xmath33 yields the coefficients @xmath85 and @xmath86 km s@xmath87 mpc@xmath87 . by substituting the coefficients and bao data into eq.([eq : decomp_pca2 ] ) and eq.([eq : curv ] ) , each @xmath65 pair has derived a curvature parameter .
the upper panel of fig.[fig : curv ] has shown the @xmath0 with @xmath88 errors .
the curvature measurement using high redshift data at @xmath4 has the best constrain that @xmath89 , which is consistent with the curvature @xmath90 constrained by the nine year wmap data only @xcite . at low and medium redshifts ,
the deviations of @xmath0 from zero are although nearly of order unit , the measurements are still consistent with a flat universe .
the lower panel of fig.[fig : curv ] has shown a trend that the errors in @xmath0 decrease with increasing redshift . by comparing the errors of the input @xmath91 data with the errors of the output @xmath0 , we have noticed that even if the bao data at higher redshift may have less precisions than the lower ones , the curvatures determined by higher redshift bao data still have smaller errors , such as 3.7% for @xmath92 and 3.5% for @xmath93 while 1.6% for @xmath94 and 3.2% for @xmath95 .
consider the error propagation formula of eq.([eq : curv ] ) @xmath96 ^ 2\left[\left(\frac{\delta h}{h}\right)^2+\left(\frac{\delta d'}{d'}\right)^2\right]\\ & ~+ 4\omega_k^2\left[\left(\frac{\delta d}{d}\right)^2+\left(\frac{\delta h_0}{h_0}\right)^2\right ] \end{split } \label{eq : errors}\ ] ] we can find that the behavior of @xmath97 is dominated by @xmath98 at low redshift . this feature results in large errors in the determination of curvature when using low redshift data .
nevertheless , eq.([eq : errors ] ) tends to be a constant which only depends on the errors of @xmath99 when using very high redshift data , revealing that eq.([eq : curv ] ) could tightly constrain the curvature parameter via precise measurements of @xmath1 and @xmath2 at high redshifts .
we repeat the procedure in previous sections to a synthetic sample simulated by a standard flat @xmath7cdm model with parameters @xmath100 , @xmath101 , @xmath102 from @xcite .
the uncertainties of @xmath1 and @xmath8 measurements resemble a theoretical bao cosmic variance forecast for a full - sky bao survey in table 2 of @xcite , in which case the @xmath1 and @xmath2 measurements can reach a precision of 0.2% at @xmath103 .
+ fig.[fig : curv_mock ] has shown the results of the curvature measurements on the mock data .
the tendency that @xmath104 changes with @xmath11 is in agreement with those derived from real bao data . with precise bao measurements in the future
, we could expect this method to constrain the curvature parameter within @xmath105 error limit or better .
based on the work of @xcite , @xcite , @xcite and @xcite , we present a model - independent method to determine the curvature parameter .
the @xmath1 and @xmath2 pairs involved in this method are derived by bao measurements which only depend on the space - time geometry , thus can afford us to measure the curvature without any assumptions of the dynamic evolution of the universe .
the luminosity distances @xmath78 from union2.1 sn ia compilation are included to have a better constrain on estimation of the derivative of comoving angular diameter distance with respect to redshift .
the curvature parameters measured in this work are in agreement of a flat universe within error limits .
the feature of eq.([eq : curv ] ) leads to the fact that the accuracy of the curvature measurement improves with increasing redshift of @xmath106 and @xmath107 and will reach a limit primarily determined by the data quality . in this work ,
the errors of curvature measurements at low redshift are nearly of order unit , while the curvature measurement at a high redshift @xmath4 has derived a much better constrain @xmath89 , which is consistent with the nine year wmap - only results .
we use the density parameters from @xcite and theoretical bao cosmic variance forecast from @xcite to generate a small synthetic sample to test the curvature measurement .
at least one order of magnitude improvement of this curvature measurement could be expected .
this work is supported by the national science foundation of china ( grant no .
11033003 ) , the national science foundation of china ( grant no .
11173006 ) , the ministry of science and technology national basic science program ( project 973 ) under grant no .
2012cb821804 .
anderson , l. , aubourg , e. , bailey , s. , et al .
2014 , , 439 , 83 benitez - herrera , s. , rpke , f. , hillebrandt , w. , et al .
2012 , , 419 , 513 bennett , c. l. , larson , d. , weiland , j. l. , et al .
2013 , , 208 , 20 benitez - herrera , s. , ishida , e. e. o. , maturi , m. , et al .
2013 , , 436 , 854 blake , c. , brough , s. , colless , m. , et al .
2012 , , 425 , 405 chuang , c .- h .
, prada , f. , cuesta , a. j. , et al .
2013 , , 433 , 3559 chuang , c .- h . &
wang , y. 2012 , , 426 , 226 clarkson , c. , corts , m. & bassett , b. 2007 , j. cosmology astropart .
phys . , 8 , 11 farooq , o. and ratra , b. 2013 , , 766 , 7 font - ribera , a. , kirkby , d. , busca , n. , et al .
arxiv:1311.1767 [ astro-ph.co ] hemantha , m. d. p. , wang , y. & chuang , c .- h . arxiv:1310.6468 [ astro-ph.co ] jolliffe , i. t. 2002 , principal component analysis ( 2nd ed.;new york : springer - verlag ) kazin , e. a. , snchez , a. g. , cuesta , a. j. , et al .
2013 , , 435 , 64 maturi , m. and mignone , c. 2009 , , 508 , 45 mignone , c. and bartelmann , m. 2008 , a&a , 481 , 295 mortsell , e. and jonsson , j. arxiv:1102.4485 [ astro-ph.co ] planck collaboration , ade , p. a. r. , et .
arxiv:1303.5076 [ astro-ph.co ] reid , b. a. , samushia , l. , white , m. , et al .
2012 , , 426 , 2719 samushia , l. , reid , b. a. , white , m. , et al .
arxiv:1312.4899 [ astro-ph.co ] sapone , d. , majerotto , e. & nesseris , s. arxiv:1402.2236 [ astro-ph.co ] suzuki , n. , rubin , d. , lidman , c. , et al .
2012 , , 746 , 85 virey , j .- m . , talon - esmieu , d. , ealet , a. , et al .
2008 , j. cosmology astropart .
, 12 , 8 weinberg , d. , mortonson , m. , eisenstein , d. , et al .
arxiv:1201.2434 [ astro-ph.co ] xu , x. , cuesta , a. j. , padmanabhan , n. , et al .
2013 , , 431 , 2834 | we present a model - independent determination of the curvature parameter @xmath0 by using the hubble parameter @xmath1 and angular diameter distance @xmath2 from the recent baryon acoustic oscillation ( bao ) measurements .
each @xmath1 and @xmath2 pair from a bao measurement can constrain a curvature parameter .
the accuracy of the curvature measurement improves with increased redshift of @xmath1 and @xmath2 data . by using the @xmath1 and @xmath2 pair derived from bao lyman @xmath3 forest measurement at @xmath4 , the @xmath0
is confined to be [email protected] , which is consistent with the curvature @xmath6 constrained by the nine - year wmap data only .
considering future bao meausurements , at least one order of magnitude improvement of this curvature measurement could be expected . |
in past attempts to investigate the relationship between nuclear multifragmentation and a liquid to vapor phase transition @xcite various studies have sought to determine one or more critical exponents @xcite , other studies have examined caloric curves @xcite , and still others have reported the observation of negative heat capacities @xcite .
these studies suffer from the lack of knowledge of the system s location in pressure - density - temperature @xmath3 space .
for example , interpretations of caloric curves and negative heat capacities depend on assumptions of either constant pressure or constant density @xcite . in the case of determining critical exponents
, it was assumed that the fragmenting system is at coexistence and the dominant factor in fragment production was the surface energy .
the analysis presented below makes no assumptions about the location of the system in @xmath4 space and allows for other energetic considerations with regards to fragment production
. in this paper the analysis technique recently used on multifragmentation data collected by the isis collaboration @xcite is applied to the data sets for the multifragmentation of gold , lanthanum and krypton nuclei collected by the eos collaboration .
all three eos experimental data sets are shown to contain the signature of a liquid to vapor phase transition manifested by the scaling behavior predicted by fisher s droplet formalism , and the liquid - vapor coexistence line is determined over a large temperature interval extending up to and including the critical point .
the critical exponents @xmath1 and @xmath0 , as well as the critical temperature @xmath5 , the surface energy coefficient @xmath2 and the compressibility factor @xmath6 are directly extracted . from the behavior of the fragment yields the @xmath7-@xmath8 and
@xmath8-@xmath9 coexistence curves are determined and the critical pressure @xmath10 and critical density @xmath11 are estimated .
the paper is organized as follows : section [ eos - experiment ] reviews the eos data sets ; section [ fishers - droplet - formalism ] reviews fisher s droplet formalism and its connection to nuclear evaporation ; section [ fitting - the - data ] discusses the details of the data analysis ; section [ results ] reports the results of the data analysis ; section [ phase - diagram ] shows the physical implications of these results ; and finally , in section [ conclusion ] a brief discussion of the results is made . in order to demonstrate the efficacy of the analysis performed on the eos data sets ,
an appendix shows the results of this analysis performed on percolation cluster distributions .
the eos collaboration has collected data for the reverse kinematics reactions @xmath12 agev au@xmath13c , @xmath12 agev la@xmath13c and @xmath12 agev kr@xmath13c @xcite .
there were @xmath14 , @xmath15 and @xmath16 fully reconstructed events recorded for the au@xmath13c , la@xmath13c and kr@xmath13c reactions , respectively .
the term `` fully reconstructed '' means that the total measured charge in each event was within three units of the charge of the projectile . for every event ,
the charge and mass of the projectile remnant ( @xmath17 , @xmath18 ) were determined by subtracting the charge and mass of the particles knocked out of the projectile from the charge and mass of the projectile @xcite .
the thermal component of the excitation energy per nucleon of the remnant @xmath19 was determined as follows . first , the total excitation energy per nucleon @xmath20 was reconstructed based on an energy balance between the initial stage of the excited remnant and the final stage of the noninteracting fragments .
the prescription @xcite for calculating @xmath20 is then @xmath21 / a_0 \label{energy - balance}\ ] ] where @xmath22 is the multiplicity of neutrons produced via fragmentation , @xmath23 is the kinetic energy of the @xmath24 fragment in the reference frame of the remnant and @xmath25 is the removal energy and @xmath8 is the temperature of a fermi gas
. for further details see reference @xcite .
the thermal component of the excitation energy per nucleon of the remnant @xmath19 was then determined by subtracting the expansion energy @xmath26 from @xmath20 , where the quantity @xmath26 is given by @xmath27 with @xmath28 the total kinetic energy ; @xmath29 the sum of the translational thermal contribution to the fragment spectra ; and @xmath30 the coulomb contribution .
the translational energy is given by @xmath31 where @xmath32 is the multiplicity of fragments and @xmath33 is the temperature calculated from the isotopic yields @xcite .
this form follows that outlined in reference @xcite .
the coulomb contribution is given by @xmath34 \label{coulomb - component}\ ] ] where @xmath35 is the radius of the excited remnant , @xmath36 is the number of fragments with charge @xmath37 , @xmath38 is the radius ( at normal density ) of a fragment with charge @xmath37 . the volumes ( and radii )
were @xmath39 the volume of the remnant at normal density and @xmath40 the volume of the excited remnant that isentropically expands from the normal volume of the projectile @xmath41 @xcite ; @xmath35 is then determined from @xmath42 assuming a spherical volume .
this form of @xmath30 follows reference @xcite and takes into account the changing volume of the excited remnant as a function of excitation energy .
previous estimates of @xmath30 did not account for the changing volume of the fragmenting remnant @xcite .
this difference leads to a few amev difference in @xmath19 in the most violent collisions . for the analysis in this paper ,
the data for each system was binned in terms of @xmath19 in units of @xmath43 amev ; i.e. @xmath44 bins covered the excitation energy range @xmath45 .
figure [ systematics ] shows some of the systematics of the eos data binned in this manner .
these results are consistent with other eos publications @xcite .
the systematics of the eos data sets shown in fig .
[ systematics ] demonstrates the similarity in behavior exhibited by the data sets when their differing sizes are taken into account by normalizing the quantity in question to the projectile charge @xmath46 or the charge of the fragmenting system @xmath47 .
the exception is seen in fig .
[ systematics]d where only below @xmath48 amev do all three systems behave similarly . above @xmath48 amev , the size of the fragmenting systems
dominate .
this is reflected in the ordering of @xmath49 ; from lowest to highest : krypton , lanthanum and gold .
as with several other analyses @xcite , the basis of the present effort lies in an examination of the fragment yield distribution in the context of fisher s droplet formalism @xcite .
thus , a brief review of fisher s formalism is given in the following section , together with a justification for its applicability to nuclear decay rates .
fisher s droplet formalism and its forerunners @xcite are based on an equilibrium description of physical clusters or droplets that condense in a low density vapor . while fisher s formalism has long been applied to nuclear multifragmentation yields
@xcite , the question arises as to the validity of a picture of clusters in equilibrium within a low density vapor to experiments in which excited nuclei undergo multifragmentation in vacuum . specifically : `` in which sense is there an equilibrium between liquid and vapor in the free ( vacuum ) decay of a ( multifragmenting ) hot intermediate ( nucleus ) ? '' or more to the point : `` where is the vapor ? '' if one assumes , as in a compound nucleus reaction , that the initial collision entity relaxes _ quickly _ to a hot thermalized blob , which proceeds _
slowly _ to emit particles stochastically , the answers to this question is simple .
the hot blob is the liquid which is evaporating in free space according to standard evaporation theories . to establish coexistence
, the vapor need not be present .
all that is necessary is to appreciate that : in fact the concentration of species @xmath50 is completely defined by @xmath51 where @xmath52 is the emission flux of @xmath50 , @xmath53 is the concentration of species @xmath50 and @xmath54 is the average velocity of @xmath50 which is of order of @xmath55 . in other words , the outward flux , at equilibrium , is the same as the inward flux . 1 . at equilibrium ,
the evaporated particle is replaced by the back flux from the vapor . however , since the back flux is absent in the case of nuclear multifragmentation , this analysis is limited to particles with low emission probability ( first chance ) and must avoid particles which are emitted with high multiplicity .
this is approximately achieved by eliminating fragments with @xmath56 from the ensuing analysis .
2 . in the same spirit as above
, the pertinent temperature is that of the blob as it evaporates low probability particles .
thus , rather than worrying about the role of high multiplicity particles and their associated cooling on the energy - temperature relationship , the fermi gas relation ship @xmath57 can be assumed with good confidence .
the basic idea is that in a non - ideal vapor of particles interacting with repulsive cores and short range attractive forces , can be approximated at low densities and temperature by an ideal gas consisting of noninteracting monomers , dimers , trimers at equilibrium .
the ( free ) energy of sufficiently large clusters can be estimated in terms of their volume and surface energy .
these clusters are in equilibrium with each other and the relative abundances of differently sized clusters changes with temperature and pressure @xcite .
the relative abundances of clusters with @xmath50 constituents is given by : @xmath58 .
\label{droplet_distribution}\ ] ] here @xmath59 is a constant of proportionality which is fixed by the critical density @xcite .
the power law @xmath60 arises from a combinatorial factor that depends on the fact that the surface of the cluster must be closed @xcite .
the distance from coexistence is : @xmath61 where @xmath62 is the chemical potential the liquid at coexistence and @xmath63 is the chemical potential of the system .
for @xmath64 ( a super - heated vapor ) and @xmath65 ( liquid - vapor coexistence ) the above sum always converges . while for @xmath66 ( a super - saturated vapor ) , the sum diverges .
the `` classical '' part of the surface energy is parameterized by @xmath67 , where @xmath2 is the zero temperature surface energy coefficient , @xmath68 and @xmath69 relates the number of constituents of a cluster to the most probable surface area .
fisher s critical exponents @xmath0 and @xmath1 depend on the euclidean dimensionality and universality class of the system .
the total pressure of the entire cluster distribution is given by summing all of the partial pressures @xmath70 : @xmath71 and the density is @xmath72 thus the pressure and density of the system can be inferred from the knowledge of the cluster distributions . at the critical point
the system is at coexistence ( @xmath65 ) and the classical part of the surface energy cost vanishes ( @xmath73 ) .
thus both exponential factors are unity leaving only the temperature independent power law : @xmath74 away from the critical point , but along the coexistence curve so that @xmath65 , the cluster distribution is given by : @xmath75 equation ( [ droplet_distribution_at_mucoex ] ) can be rewritten as : @xmath76 thus the cluster distribution along the coexistence curve is given by a boltzmann factor with @xmath77 and @xmath78 this boltzmann factor manifests itself in arrhenius plots for the fragment yields where a linear relation between @xmath79 and and @xmath80 is observed .
this behavior has long been observed in many nuclear fragment yield distributions @xcite and has recently been observed in the cluster distributions of percolation ( with the bond breaking probability playing the role of temperature ) @xcite and the ising model @xcite .
as discussed above , fisher s formalism relates directly to a reaction rate picture . in this picture , the heavy fragments ( e.g. @xmath81 ) are the product of first chance emission from the excited remnant .
the first chance emission from a compound nucleus can be written as : @xmath82 thus the fragment yields , parameterized via fisher , can be related to the decay rates ( widths @xmath83 ) .
furthermore , @xmath83 , which controls the first chance emission yields , is the same decay width which controls the mean emission times @xmath84 since : @xmath85 thus @xmath86 the mean time for fragment emission reported by the isis collaboration @xcite is well described as a boltzmann factor .
it was also noted that the boltzmann factor describing the emission times is the same as that describing the fragment yields @xcite .
this indicates that the thermal reaction rate picture is valid for multifragmentation ; fragments can be viewed as being the result of the evaporation of an excited nucleus .
preliminary fits of the gold , lanthanum and krypton data with eq .
( [ droplet_distribution ] ) led to puzzlingly large results for @xmath87 ( @xmath88 amev ) which could be interpreted as a substantial degree of super - saturation .
a much more plausible alternative explanation is the lack of account of the coulomb effects in fisher s formalism .
equations ( [ boltzmann_distribution ] ) and ( [ barrier ] ) support the presence of a barrier controlling the flux from liquid to vapor and vice versa .
this barrier should depend not only on the surface energy of the fragment but should reflect the entire energy necessary to remove a fragment from the liquid and place it into the vapor . at the least
, the energy necessary to relocate a charge @xmath37 from the bulk to `` near '' the surface of the `` residual '' nucleus should be evaluated .
this energy is negative and counteracts the effects of the surface energy .
furthermore , it is to leading order , linear in @xmath37 and thus in @xmath50 .
thus the large values of @xmath89 .
an attempt to include the coulomb energy explicitly is then @xmath90 with @xmath91 with @xmath92 fm . in eq .
( [ coulomb ] ) there is a recognizable coulomb interaction energy of two touching spheres modified by a factor @xmath93 .
the parameter @xmath94 ( left as a fit parameter ) takes into account the numerical coefficients of the linear term in @xmath37 plus polarization effects , and @xmath95 takes care of the need for the vanishing difference between the liquid and vapor near the critical point .
note that the coulomb energy discussed in eq .
( [ coulomb - component ] ) is different from the coulomb energy discussed in eq .
( [ coulomb ] ) .
equation ( [ coulomb - component ] ) describes the total coulomb energy present in the fragmentation process , while eq .
( [ coulomb ] ) describes the cost in moving a fragment from the nuclear liquid to the nuclear vapor .
the mass of a fragment @xmath50 prior to secondary decay was estimated by multiplying the measured fragment charge @xmath37 by @xmath96 and then by a factor of @xmath97 where @xmath98 is the binding energy of the fragment and @xmath99 is a fit parameter to allow for an increase or decrease in the amount of secondary decay .
the temperature was determined by assuming a degenerate fermi gas , @xmath100 the parameter @xmath101 was taken to be @xcite : @xmath102 in order to accommodate the empirically observed change in @xmath101 with excitation energy @xcite .
here @xmath103 is the binding energy of the fragmenting system . using the fermi gas approximation to relate @xmath19 and @xmath8 gives a reasonable estimate of the temperature of the excited remnant at the time of first chance emission .
it has been observed that even the isotope ratio thermometer follows the fermi gas approximation quite well so long as the average number of imfs is less than one @xcite . to obtain the concentration of fragments of a given mass ,
the total number of fragments @xmath104 of a given size @xmath50 was normalized to the size of the fragmenting system @xmath105 so that @xmath106 .
the location of the critical point , in terms of excitation energy , was determined from an examination of measured fluctuations . in general , as the critical point of a system
is approached from the two phase region , the difference between phases diminishes and the system fluctuates from one phase to the other . at the critical point
the fluctuations are maximal .
however , while the maximum in the fluctuations occurs at the critical point , the presence of a peak in the fluctuations is a necessary , but not sufficient , condition for a existence of a phase transition @xcite .
the fluctuations measured in the eos data are : ( 1 ) in the charge of the largest fragment normalized to the charge of the fragmenting system , and ( 2 ) related to the average mass number of a fragment as measured by the quantity @xmath107 @xcite , where @xmath108 with @xmath109 as the @xmath110 moments of the fragment distributions @xmath111 these fluctuations are shown in fig .
[ fluctuations1 ] .
the peak in the fluctuations was found by smoothing the data ( solid lines in fig .
[ fluctuations1 ] ) , taking the numerical derivative of the smoothed data and finding the value of @xmath19 where the the derivative passed through zero , see fig . [ fluctuations2 ] .
finally , the value of the excitation energy at the critical point @xmath112 was determined by averaging the results from both measures of the fluctuations .
table [ critical - point - table ] lists the results . for this analysis
the values determined for the excitation energy at the critical point for the au@xmath13c reaction are in proximity of other values observed in previous eos analyses ( @xmath113 amev ) @xcite .
differences in the values of @xmath112 arise from the different method of constructing the thermal portion of the excitation energy described above @xcite .
estimates of the critical temperature @xmath5 are made by using the values of @xmath112 in eq .
( [ fermi - gas - temp ] ) and lead to values , shown in table [ critical - point - table ] , that are comparable to theoretical estimates for small nuclear systems @xcite . as an aside , as shown in table [ critical - point - table ] the value of @xmath5 increases with decreasing projectile ( and thus remnant ) mass .
this is opposite of the trend assumed in a prior analysis of the eos gold multifragmentation data where the coulomb energy was neglected @xcite but in agreement with the trend reported in other analysis of the eos data sets @xcite . in fig .
[ fluctuations1 ] the value of @xmath107 for the kr system attains a peak value of only @xmath114 .
it has been suggested that the magnitude of the peak in @xmath107 could distinguish between the presence of a power law with @xmath115 ( @xmath116 ) and an exponential distribution ( @xmath117 ) in the cluster yields @xcite .
however , it was seen that this is not the case @xcite and it will be seen in the appendix that small percolation lattices have values of @xmath107 with peak magnitudes of less than two yet still exhibit a continuous phase transition with an exponent of @xmath118 in the power law describing the cluster yields at the critical point .
thus , the height of the peak in @xmath107 can not be used to rule out the presence of a critical point and the associated power law in the cluster distribution or provide information about the value of the power law exponent . to determine the `` energy of vaporization '' of a cluster @xmath119 the ideal gas approximation @xmath120
is followed so that @xmath121 , where @xmath8 is replaced by @xmath122 in keeping with standard practice in percolation work and @xmath122 is the average bond breaking probability considered ; @xmath123 , @xmath124 and @xmath125 for @xmath126 , @xmath127 and @xmath128 respectively .
the @xmath129 values listed in table [ perc - thermo - table ] were found by dividing @xmath130 by the most probable cluster size ( @xmath131 , @xmath132 and @xmath133 for @xmath126 , @xmath127 and @xmath128 respectively ) , this puts @xmath119 on a `` per site '' basis .
the values of @xmath129 shown in table [ perc - thermo - table ] are nearly identical to the values of the surface energy coefficient @xmath2 , which is not surprising since for percolation on a simple cubic lattice @xmath2 arises from the bonds broken to form the surface .
furthermore , the `` energy of vaporization '' is approximately equal to the number of bonds per lattice site ( also shown in table [ perc - thermo - table ] ) , a strong indication that the @xmath134 calculated here is the `` bulk binding energy '' of the lattice in question .
the value of @xmath129 decreases with the size of the lattice because the percolation calculations were performed for open boundary conditions .
the compressibility factor at the critical point was determined via eq .
( [ comp - fac - eq ] ) , the results are shown in table [ perc - thermo - table ] . from @xmath135 , @xmath136 and @xmath11 ( determined below ) the `` pressure '' at the critical point can be found .
the resulting values of @xmath10 are shown in table [ perc - critical - point ] , but the interpretation of these values is an open question . following the thermodynamic treatment of the percolation results , the reduced probability versus `` reduced density '' phase diagram is produced via eq .
( [ reduced_density ] ) .
this leads to the points shown in fig .
[ perc - prob - dens ] .
these points are then fit to guggenheim s empirical formula , eq .
( [ gugg - eq ] ) , with @xmath137 ( in good agreement with text book values @xmath138
@xcite ) from eq . ( [ beta ] ) .
these results are shown for each lattice by the solid lines in fig .
[ perc - prob - dens ] . while it is not clear what density this plot describes , some insight can be gained by noting that the `` liquid '' branch reaches @xmath139 to @xmath140 at @xmath141 . assuming that at @xmath141 @xmath142 , since no bonds are broken , then @xmath143 to @xmath144 , which is approximately the percentage of bonds broken at the critical point .
thus it seems that the density in fig .
[ perc - prob - dens ] is related to the number of broken bonds .
it is also noted that for @xmath145 the vapor branch of the coexistence curve shows @xmath146 , this serves as an illustration of the magnitude of the error associated with this procedure .
it is also possible to directly explore the behavior of the reduced density of the `` liquid , '' at least in the larger system .
this is done by normalizing the size of the largest cluster at a given value of @xmath122 to the size of the largest cluster at the critical point @xmath147 .
figure [ perc - prob - dens ] shows that for the @xmath148 lattice , the measured normalized density of the liquid tracks along the coexistence curve predicted by guggenheim s empirical formula and the reduced density of the vapor . for @xmath149 the effects of the finite size of the lattice
are observed and the measured reduced density of the liquid deviates from the coexistence curve .
the effects of finite size are more evident in the smaller lattices where there is little or no agreement between the measured reduced density of the liquid and the coexistence curves .
effects of finite size on the largest cluster such as these have been observed previously @xcite . | the fragment yields from the multifragmentation of gold , lanthanum and krypton nuclei obtained by the eos collaboration are examined in terms of fisher s droplet formalism modified to account for coulomb energy .
the critical exponents @xmath0 and @xmath1 and the surface energy coefficient @xmath2 are obtained .
estimates are made of the pressure - temperature and temperature - density coexistence curve of finite neutral nuclear matter as well as the location of the critical point . |
@xmath11 : , ( b ) the canonical potential - energy probability distribution at @xmath12 , and ( c ) the microcanonical inverse temperature for @xmath13 lennard - jones particles for reaction coordinate @xmath9 . the black solid line in ( a ) shows the minimal free - energy path . the equal - area construction in ( c ) yields .
the panels ( d ) and ( e ) show the analogous quantities reevaluated for reaction coordinate @xmath8 [ here ] . ]
@xmath14 : , ( b ) the canonical potential - energy probability distribution at @xmath12 , and ( c ) the microcanonical inverse temperature for @xmath13 lennard - jones particles for reaction coordinate @xmath9 . the black solid line in ( a ) shows the minimal free - energy path . the equal - area construction in ( c ) yields .
the panels ( d ) and ( e ) show the analogous quantities reevaluated for reaction coordinate @xmath8 [ here ] . ]
a paradigm of nucleation and dissolution is the equilibrium droplet formation in a supersaturated gas @xcite .
the first - order condensation - evaporation transition separates a homogeneous gas phase from an inhomogeneous phase , where a single macroscopic droplet of size @xmath15 is in equilibrium with the remaining vapor @xcite .
in fact , the probability for intermediate - sized droplets was shown to vanish @xcite . in the vicinity of the transition ,
the energy - dominated inhomogeneous condensed phase coexists with the entropy - dominated homogeneous gas phase .
a transition between both phases may only occur by energy variation upon nucleation or dissolution . in reality
, this of course refers to the total energy @xmath8 . for systems where the momentum phase space may be integrated out explicitly ( see methods ) , the problem simplifies in terms of computability . here , we start by presenting a qualitative picture in the potential - energy formulation ( denoted by a hat , e.g. , @xmath16 ) as a direct result of computer simulations before we generalize the results and compare both reaction - coordinate formulations . figure [ figillustration ] ( a ) shows the free - energy landscape of lennard - jones droplet condensation - evaporation at the finite - size transition temperature defined as , where @xmath17 is a generalization of the ( conformational ) density of states to the two - dimensional @xmath9@xmath15 reaction - coordinate space . for fixed @xmath9
this resembles a parabola with a single local minimum .
the resulting transition path is shown as a black line and its relative maximum along this path ( around @xmath18 ) is an estimate of the transition free - energy barrier . instead
, however , one may consider the projection along the droplet size @xcite or equivalently along the energy , connected to the corresponding probability distributions @xmath19 and @xmath20 .
we follow the latter approach and derive the free - energy barrier from the suppression of transition states in the canonical energy probability distributions ( see methods ) for both reaction coordinates @xmath8 and @xmath9 .
the probability distributions are shown in figs .
[ figillustration ] ( b ) and ( d ) and are seemingly asymmetric , with a narrow peak for the gas phase and a broad peak for the droplet phase .
methodologically , both ensembles are analogous so that we limit in the following the notation to the case of total energy @xmath8 . at the equal - height inverse temperature @xmath21 , @xmath22 has two peaks at @xmath23 of equal height and in between a minimum at @xmath24 .
the resulting free - energy barrier is .
equivalently , one may perform the analysis entirely in the microcanonical frame @xcite and consider the enclosed area by the microcanonical inverse temperature @xmath25 and the canonical inverse temperature @xmath26 , \label{eqdeltafbeta}\ ] ] shown in figs .
[ figillustration ] ( c ) and ( e ) . demanding areas of equal size yields the equal - area inverse temperature @xmath27 , which is in fact identical to @xmath21 @xcite .
we notice that the energy probability distributions @xmath28 and @xmath29 are related by a convolution involving the maxwell - boltzmann distribution @xmath30 as ( see methods ) .
this in turn corresponds to a physical smoothing , which diminishes the ratio between maxima and minimum . as a consequence
we expect a lower barrier in the total - energy formulation due to the kinetic contribution .
the free - energy barrier is commonly assumed to be proportional to the occurring interface . here the surface of the droplet @xmath31 separates the liquid droplet from the surrounding gas and consequently @xmath32 , with the interface tension @xmath33 .
the surface should scale as @xmath34 .
however , the droplet size itself at the condensation - evaporation transition does not scale trivially with system size . exploiting the equivalence to the ising model for equilibrium droplet formation at fixed temperature
it was shown that directly at the finite - size transition density a single macroscopic droplet of size @xmath35 coexists with the surrounding vapor @xcite .
this result may be translated to a fixed - density setup using taylor expansions , where directly at the finite - size transition temperature the analogue scaling @xmath36 was verified for both lattice and off - lattice models @xcite . introducing an effective interfacial free energy @xmath37
then yields @xmath38 .
it is common for the study of interface tensions to consider logarithmic corrections , dating back to early field - theoretic results @xcite .
altogether this yields our final scaling ansatz @xmath39 where @xmath40 and @xmath41 are constants , in accordance with previous results for the two - dimensional lattice gas @xcite .
this is the desired leading - order exponent in eq . .
neglecting the prefactor @xmath5 for now , we obtain from eq . to leading order the rate of equilibrium droplet formation as @xmath42 .
thus , for increasing system size the probability that a single macroscopic droplet forms decreases exponentially . and @xmath8 .
the inset shows the finite - size scaling of @xmath37 according to eq . .
] figure [ figbarrier ] shows the free - energy barrier as a function of system size for both reaction coordinates @xmath8 and @xmath9 , obtained via eq . .
both estimates yield barriers up to about @xmath43 , showing at close sight an almost constant shift .
only the total energy @xmath8 includes the kinetic contribution which is here reflected in a smaller barrier , whereas the interfacial free energies are expected to be identical , @xmath44 . in fact , fits to eq .
yield @xmath45 and @xmath46 , each for @xmath47 with goodness - of - fit parameter @xmath48 , for reaction coordinate @xmath8 and @xmath9 , respectively .
this remains consistent within error bars under variation of @xmath49 $ ] .
the accordance of data and fit is demonstrated in the inset . in order to test the significance of the logarithmic corrections
, we considered in addition a restricted fit to @xmath50 .
we obtain @xmath51 and @xmath52 for @xmath53 with @xmath54 and @xmath55 , respectively
. however , the estimate of @xmath37 gradually decreases with increasing @xmath56 .
thus , the most probable scenario remain small logarithmic corrections competing with a constant shift .
the leading scaling behaviour was also observed in ref .
@xcite directly as a linear function of the droplet surface .
the advantage of the present approach is that it avoids difficulties and uncertainties coming from the `` correct '' identification of the nucleus interface .
we observe that considering the kinetic contribution results in a shifted barrier , for the considered system sizes this difference is consistently of @xmath57 .
this additive contribution , while much smaller than the leading behaviour , leads to a multiplicative relation between the nucleation rates @xmath58 . neglecting
the momentum phase space thus underestimates the rates . in common situations , however ,
the deviations between experiment and simulations are of the order of several magnitudes , such that the effect of the kinetic contribution may be considered subleading . and @xmath8 .
the inset shows a vanishing finite - size difference between both temperatures . ]
the evaluation of the free - energy barrier via equal areas provides us with a definition of the finite - size transition temperature . at fixed density , we showed in ref .
@xcite that the transition temperature , obtained from specific - heat peaks , scales as a power of the critical droplet radius , @xmath59 . the same is expected for all other transition temperature definitions .
figure [ figfsstemp ] shows the equal - area definition together with a fit including higher - order corrections of the form @xmath60 least - square fits for @xmath61 yield @xmath62 and @xmath63 with @xmath64 , for reaction coordinate @xmath8 and @xmath9 , respectively .
the excellent fit results show that the empirical , yet physically motivated , higher - order corrections describe the finite - size scaling very well . in addition , the strong finite - size deviations open a possibility to study finite - size scaling in experiments on the nanoscale ( for a conversion see methods ) .
the finite - size transition temperature of the largest system ( @xmath65 ) is @xmath66 , which still deviates from the thermodynamic limit by @xmath67 .
it is worth noting that typical canonical finite - size transition temperatures , e.g. , the peak location of the specific heat , do not depend on the kinetic contribution to the partition function . in the canonical expectation values
, the kinetic prefactor simply cancels . here
, however , we observe a finite - size difference in the transition temperature depending on whether we consider the kinetic contribution or not .
of course , the thermodynamic limit coincides .
this is illustrated in the inset of fig .
[ figfsstemp ] with a finite - size scaling of the transition temperature difference @xmath68 .
it shows a prominent power - law scaling of the form @xmath69 , which interestingly is the same scaling as the inverse transition droplet volume .
the finite - size difference arises from the convolution of an asymmetric energy probability distribution with the asymmetric maxwell - boltzmann distribution , compare fig .
[ figillustration ] and eq . , which manifests in the geometric differences of the microcanonical inverse temperature and the enclosed areas in eq . .
it appears that correlations between the ensemble definitions account for compatible leading - order scaling corrections in eq . , which further supports this ansatz and explains the observed difference .
we have presented a shape - free approach to the estimation of canonical free - energy barriers in equilibrium droplet formation . the finite - size scaling is dominated by the predicted @xmath70 behaviour but we identified additional logarithmic corrections from precise numerical estimates . somewhat surprisingly , the absolute free - energy barrier is sensitive to the consideration of the kinetic contribution .
it is well - known that the restriction to the conformational phase space does not influence finite - size transition points determined from canonical expectation values .
these are evaluated on the level of the canonical partition function .
the free - energy barrier and the associated equal - height or equal - area transition temperature , however , are determined from the energy probability distribution . here
, the two formulations are related by a convolution with the maxwell - boltzmann distribution which explains the finite - size differences . at the same time , the probability distributions are the integrands of the respective partition functions . in the end , it boils down to the observation that equality of integrals does not imply equality of the integrands .
this may become relevant once theoretical predictions and experimental measurements becomes precise enough .
still , a restriction to the conformational phase space retains intensive parameters in the thermodynamic limit . as a numerical advantage , considering the total energy as reaction coordinate shows less fluctuations in the microcanonical partition function and canonical probability distribution , since the underlying convolution is a smoothing procedure of physical origin .
it is expected that the energy remains a suitable reaction coordinate for general nucleation - like mechanisms . in this case
, generalized - ensemble methods may unfold their full power .
moreover , our approach at fixed density provides the possibility for experiments to perform heating - cooling studies in order to probe transition rates .
this has potential application for temperature - driven self - assembly @xcite , crystallization in so - called phase - change materials @xcite and glassy solids @xcite , cluster formation in protein solutions @xcite and polymer aggregation @xcite , dislocation nucleation @xcite , or the study of surface nanobubbles and nanodroplets @xcite .
of course , experimental observations commonly include the formation of multiple clusters .
reasons for this include impurities acting as nucleation seeds .
we suppose that this leads to a local quasi - equilibrium on the respective length scales . here
, a proper combination of the canonical droplet formation with the effect of nucleation seeds @xcite seems to be a fruitful approach to a systematic understanding . with further developments , simulations may provide reliable estimates for finite - size systems and meet experiments on the nanometer scale .
recently , there has been some ambiguity with the definition of a `` microcanonical ensemble '' in computer simulations @xcite .
this is a crucial aspect relevant for physical interpretations that appears to be unwittingly softened in the last decade .
the _ microcanonical ensemble _ ( @xmath71 ) describes an isolated system in which the number of constituents @xmath72 , the volume @xmath73 , and the total energy @xmath8 are conserved . here
, the transfer of potential energy @xmath9 into kinetic energy @xmath74 and vice versa is a valid and relevant mechanism , where @xmath75 .
the microcanonical ( boltzmann ) entropy is defined as @xmath76 , with the partition function @xmath77)$ ] , where @xmath78 denotes the integration over state space and @xmath79 over momentum space .
the other common definition is the _ conformational microcanonical ensemble _ ( @xmath80 ) , describing instead a system with fixed potential energy @xmath9 .
the conformational microcanonical entropy is @xmath81 , where @xmath82 is the density of states or the conformational microcanonical partition function .
the consequences are drastic : a ( physical ) interpretation of energy transfer from potential to kinetic energy is no longer valid .
this is natural for spin systems , where a kinetic contribution is not defined in the first place ( but may be exploited for numerical purposes @xcite ) . on the contrary
, it is particularly relevant for situations in soft condensed matter , e.g. , for particles and polymers , where interpretations of energy transfer become natural .
however , there are good reasons for this choice : @xmath83 is a fundamental property of statistical mechanics .
it encodes the full information about the conformational space and allows for identification of ( structural ) phase transitions @xcite .
furthermore , it may be exploited for reweighting techniques and flat - histogram monte carlo methods @xcite .
the relation between @xmath84 and @xmath83 is given by a convolution with the kinetic - energy contribution @xcite : if momenta and positions are independent one may separate the kinetic - energy contribution @xmath85 , explicitly perform the momentum integration , and obtain for @xmath72 particles in three dimensions @xcite @xmath86 we define in order to extend the integral over the full energy range @xmath87 .
in this way the total - energy - surface entropy @xmath88 appears as a ( weighted ) potential - energy - volume entropy .
it is interesting to notice that , since all @xmath89 , the @xmath71 entropy increases with @xmath8 and the microcanonical inverse temperature can not become negative , opposed to its conformational counterpart .
this may be an interesting aspect for a recent debate on the correct definition of entropy when connected with the phenomenological thermodynamic entropy , e.g. , in refs .
@xcite and references therein .
the _ canonical ensemble _ is defined in terms of the partition function @xmath90 , where each phase - space point is weighted with the boltzmann factor according to the total energy .
again , the kinetic part may be explicitly integrated for generic systems .
each degree of freedom contributes with a gaussian integral , and one obtains for @xmath72 particles , @xmath91 where @xmath92 is the partition function of the _ conformational canonical ensemble_. both partition functions may be expressed as integrals in terms of the respective energies , namely and .
the corresponding canonical energy probability distributions are defined as and .
we may now relate the two energy probability distributions by starting with the definition of @xmath28 and inserting eq .
: @xmath93 we identify the latter part of the integrand as the @xmath72-particle maxwell - boltzmann distribution @xmath30 and may write eq . as a convolution @xmath94 .
we consider a system of lennard - jones particles in a dimensionless periodic box of length @xmath95 with fixed density .
mutual avoidance and short - range attraction are modeled by the 12 - 6 lennard - jones potential with @xmath96 and @xmath97 . for computational efficiency , the potential is cutoff at @xmath98 and shifted by @xmath99 .
system sizes range up to @xmath65 , which is competitive with state - of - the - art molecular dynamics simulations like well - tempered metadynamics @xcite . for the chosen density @xmath100 a system with @xmath101 particles
requires a box of linear dimension @xmath102 . for argon ,
@xmath103 such that @xmath104 is on the nanoscale .
of course , for a comparison to an experimental setup one should include both the explicit geometric constraints and the full lennard - jones potential .
parallel multicanonical monte carlo simulations @xcite allow us to efficiently sample the suppressed states , by iteratively adapting an auxiliary weight function @xmath105 to yield a flat histogram @xmath106 .
the final weight function is related to the density of states up to a multiplicative factor : .
this gives direct access to microcanonical estimates and canonical expectation values at any temperature .
using eq .
this even provides an estimate of @xmath84 .
monte carlo updates include short - range and long - range particle displacements .
we measure the conformational ( potential ) energy @xmath9 and the number of particles in the largest cluster @xmath15 as in ref .
error bars are obtained by the jackknife method @xcite .
10 j. feder , k. c. russell , j. lothe , and g. m. pound , _ homogeneous nucleation and growth of droplets in vapours _ , adv .
phys . * 15 * , 111 ( 1966 ) .
d. w. oxtoby , _ homogeneous nucleation : theory and experiment _ , j. phys : condens .
matter * 4 * , 7627 ( 1992 ) .
d. kashchiev , _ nucleation : basic theory with applications _
( butterworth - heinemann , oxford , 2000 ) .
s. auer and d. frenkel , _ prediction of absolute crystal - nucleation rate in hard - sphere colloids _ , nature ( london ) * 409 * , 1020 ( 2001 ) .
r. p. sear , _ nucleation : theory and applications to protein solutions and colloidal suspensions _ ,
j. phys . : condens .
matter * 19 * , 033101 ( 2007 ) .
a. stradner , h. sedgwick , f. cardinaux , w. c. k. poon , s. u. egelhaaf , and p. schurtenberger , _ equilibrium cluster formation in concentrated protein solutions and colloids _ ,
nature ( london ) * 432 * , 492 ( 2004 ) .
j. a. kalb , f. spaepen , and m. wuttig , _ kinetics of crystal nucleation in undercooled droplets of sb- and te - based alloys used for phase change recording _ , j. appl .
phys * 98 * , 054910 ( 2005 ) .
m. wuttig and n. yamada , _ phase - change materials for rewriteable data storage _ , nat . mat . * 6 * , 824 ( 2007 ) . t. h. lee and s. r. elliott , _ ab initio computer simulation of the early stages of crystallization : application to ge@xmath107 sb@xmath107 te@xmath108 phase - change materials _ , phys .
lett . * 107 * , 145702 ( 2011 ) .
lee , g. w. burr , r. m. shelby , s. raoux , c. t. rettner , s. n. bogle , k. darmawikarta , s. g. bishop , and j. r. abelson , _ observation of the role of subcritical nuclei in crystallization of a glassy solid _ , science * 326 * , 980 ( 2009 ) .
j. s. langer , _ theory of the condensation point _ ,
* 41 * , 108 ( 1967 ) .
a. l. goodmann , j. i. kapusta , and a. z. mekjian , _ liquid - gas phase instabilities and droplet formation in nuclear reactions _ ,
c * 30 * , 851 ( 1984 ) .
n. hansen and w. f. van gunsteren , _ practical aspects of free - energy calculations : a review _ , j. chem .
theory comput .
* 10 * , 2632 ( 2014 ) .
m. schrader , p. virnau , and k. binder , _ simulation of vapor - liquid coexistence in finite volumes : a method to compute the surface free energy of droplets _ , phys .
e * 79 * , 061104 ( 2009 ) .
s. ryu and w. cai , _ validity of classical nucleation theory for ising models _ ,
e * 81 * , 030601r ( 2010 ) .
a. nubaumer , e. bittner , and w. janke , _ free - energy barrier at droplet condensation _ ,
suppl . * 184 * , 400 ( 2010 ) .
b. n. hale and m. thomason , _ scaled vapor - to - liquid nucleation in a lennard - jones system _ , phys .
lett . * 105 * , 046101 ( 2010 ) .
k. k. tanaka , j. diemand , r. anglil , and h. tanaka , _ free energy of cluster formation and a new scaling relation for the nucleation rate _
, j. chem . phys . * 140 * , 194310 ( 2014 ) .
a. statt , p. virnau , and k. binder , _ finite - size effects on liquid - solid phase coexistence and the estimation of crystal nucleation barriers _ ,
. lett . * 114 * , 026101 ( 2015 ) .
k. binder and m. h. kalos , _ `` critical clusters '' in a supersaturated vapor : theory and monte carlo simulation _ , j. stat .
* 22 * , 363 ( 1980 )
. t. neuhaus and j. s. hager , _
2d crystal shapes , droplet condensation , and exponential slowing down in simulations of first - order phase transitions _ , j. stat .
* 113 * , 47 ( 2003 ) .
m. biskup , l. chayes , and r. koteck , _ on the formation / dissolution of equilibrium droplets _ , europhys .
60 * , 21 ( 2002 ) .
m. biskup , l. chayes , and r. koteck , _ critical region for droplet formation in the two - dimensional ising model _ , commun .
phys . * 242 * , 137 ( 2003 ) .
k. binder , _ theory of the evaporation / condensation transition of equilibrium droplets in finite volumes _ , physica a * 319 * , 99 ( 2003 ) .
a. nubaumer , e. bittner , t. neuhaus , and w. janke , _ monte carlo study of the evaporation / condensation transition of ising droplets _ , europhys . lett .
* 75 * , 716 ( 2006 ) a. nubaumer , e. bittner , and w. janke , _ monte carlo study of the droplet formation - dissolution transition on different two - dimensional lattices _ ,
e * 77 * , 041109 ( 2008 ) .
j. zierenberg and w. janke , _ exploring different regimes in finite - size scaling of the droplet condensation - evaporation transition _ , phys .
e * 92 * , 012134 ( 2015 ) .
d. h. e. gross , _ microcanonical thermodynamics _
( world scientific , singapore , 2001 ) .
w. janke , _ canonical versus microcanonical analysis of first - order phase transitions _ , nucl .
b ( proc . suppl . )
* 63 * a - c , 631 ( 1998 ) .
p. jonkheijm , p. van der schoot , a. p. h. j. schenning , and e. w. meijer , _ probing the solvent - assisted nucleation pathway in chemical self - assembly _ , science * 313 * , 80 ( 2006 ) .
b. a. h. huisman , p. g. bolhuis , and a. fasolino , _ phase transition to bundles of flexible supramolecular polymers _ ,
lett . * 100 * , 188301 ( 2008 ) .
j. zierenberg , m. mueller , p. schierz , m. marenz , and w. janke , _ aggregation of theta - polymers in spherical confinement _ , j. chem
141 * , 114908 ( 2014 ) .
j. zierenberg and w. janke , _ from amorphous aggregates to polymer bundles : the role of stiffness on structural phases in polymer aggregation _ ,
* 109 * , 28002 ( 2015 ) .
s. ryu , k. kang , and w. cai , _ entropic effect on the rate of dislocation nucleation _ ,
u. s. a. * 108 * , 5174 ( 2011 ) .
d. lohse and x. zhang , _ surface nanobubbles and nanodroplets _ , rev .
phys . * 87 * , 981 ( 2015 ) .
a. cacciuto , s. auer , and d. frenkel , _ onset of heterogeneous crystal nucleation in colloidal suspensions _ ,
nature * 428 * , 404 ( 2004 ) .
p. schierz , j. zierenberg , and w. janke , _ molecular dynamics and monte carlo simulations in the microcanonical ensemble : quantitative comparison and reweighting techniques _ , j. chem
. phys . * 143 * , 134114 ( 2015 ) . v. martin - mayor , _ microcanonical approach to the simulation of first - order phase transitions _ ,
lett . * 98 * , 137207 ( 2007 ) .
m. kastner and m. pleimling , _ microcanonical phase diagrams of short - range ferromagnets _ , phys .
* 102 * , 240604 ( 2009 ) .
s. schnabel , d. t. seaton , d. p. landau , and m. bachmann , _ microcanonical entropy inflection points : key to systematic understanding of transitions in finite systems _
, phys . rev .
e * 84 * , 011127 ( 2011 ) .
w. janke , _ monte carlo simulations in statistical physics from basic principles to advanced applications _ , in : _
order , disorder and criticality : advanced problems of phase transition theory _
y. holovatch ( world scientific , singapore , 2012 ) , pp .
93166 . w. janke and w. paul , _ thermodynamics and structure of macromolecules from flat - histogram monte carlo simulations _ , soft matter * 12 * , 642 ( 2016 ) .
see , e.g. , p. labastie and r. l. whetten , _ statistical thermodynamics of the cluster solid - liquid transition _ , phys .
lett . * 65 * , 1567 ( 1990 ) ; and references therein . f. calvo , j. p. neirotti , d. l. freeman , and j. d. doll , _ phase changes in 38-atom lennard - jones clusters .
ii . a parallel tempering study of equilibrium and dynamic properties in the molecular dynamics and microcanonical ensembles _ , j. chem . phys . * 112 * , 10350 ( 2000 ) . j. dunkel and s. hilbert , _ consistent thermostatistics forbids negative absolute temperatures _ , nat .
* 10 * , 67 ( 2014 ) .
s. hilbert , p. hnggi , and j. dunkel , _ thermodynamic laws in isolated systems _ , phys .
e * 90 * , 062116 ( 2014 ) .
r. h. swendsen and j. s. wang , _ gibbs volume entropy is incorrect _ , phys .
e * 92 * , 020103(r ) ( 2015 ) .
m. salvalaglio , p. tiwary , g. m. maggioni , m. mazzotti , and m. parrinello , _ overcoming timescale and finite - size limitations to compute nucleation rates from small scale well tempered metadynamics simulations _ , arxiv:1508.01642 ( 2015 ) .
b. a. berg and t. neuhaus , _ multicanonical algorithms for first order phase transitions _
b * 267 * , 249 ( 1991 ) .
b. a. berg and t. neuhaus , _ multicanonical ensemble : a new approach to simulate first - order phase transitions _
rev . lett . * 68 * , 9 ( 1992 ) .
w. janke , _ multicanonical simulation of the two - dimensional 7-state potts model _
c * 03 * , 1137 ( 1992 ) .
w. janke , _ multicanonical monte carlo simulations _
, physica a * 254 * , 164 ( 1998 ) . j. zierenberg , m. marenz , and w. janke , _ scaling properties of a parallel implementation of the multicanonical algorithm _ , comput .
. commun . *
184 * , 1155 ( 2013 ) .
j. zierenberg , m. wiedenmann , and w. janke , _ application of the parallel multicanonical method to lattice gas condensation _ , j. phys .
* 510 * , 012017 ( 2014 ) .
b. efron , _ the jackknife , the bootstrap and other resampling plans _ ( society for industrial and applied mathematics , philadelphia , 1982 ) .
young , _ everything you wanted to know about data analysis and fitting but were afraid to ask _ , springerbriefs in physics ( springer international publishing , 2015 ) .
the project was funded by the european union , the free state of saxony and the deutsche forschungsgemeinschaft ( dfg ) under grant no .
ja 483/31 - 1 .
the authors gratefully acknowledge the computing time provided by the john von neumann institute for computing ( nic ) on the supercomputer jureca at jlich supercomputing centre ( jsc ) under grant no .
part of this work has been financially supported by the leipzig graduate school of natural sciences `` buildmona '' and by the deutsch - franzsische hochschule ( dfh - ufa ) through the doctoral college `` @xmath109 '' under grant no .
cdfa-02 - 07 .
all authors contributed equally to the theory and j.z . performed the simulations .
j.z . and w.j
analyzed the data and wrote the paper .
all authors discussed the results and commented on the manuscript .
the authors declare no competing financial interests . | a common approach to study nucleation rates is the estimation of free - energy barriers .
this usually requires knowledge about the shape of the forming droplet , a task that becomes notoriously difficult in macromolecular setups starting with a proper definition of the cluster boundary or a proper ensemble choice . here , we demonstrate that a shape - free determination of temperature - driven droplet formation is directly accessible in the canonical ensemble . combined with rigorous results on canonical equilibrium droplet formation ,
this allows for a well - defined finite - size scaling analysis of the effective interfacial free energy at fixed density .
we verify the theoretical predictions for the formulation of a liquid droplet in a supersaturated gas by generalized - ensemble monte carlo simulations of a lennard - jones system . because a standard approach in monte carlo simulations is to work in the conformational ensemble ,
we show that excluding the kinetic energy from the partition function leads to finite - size differences in the free energy but retains intensive parameters in the thermodynamic limit .
our study of nucleation rates at fixed density corresponds to a heating - cooling framework where simulation and experiment may meet at the nanometer scale .
the formation of equilibrium droplets from a supersaturated gas is a long - standing subject of interest , being an essential phase transition in nature @xcite .
more importantly , the underlying mechanism is relevant for a multitude of nucleation - like processes from statistical mechanics to material science . these include crystallization in colloidal suspensions @xcite , cluster formation in protein solutions @xcite , as well as domain formation in so - called phase - change materials @xcite and glassy solids @xcite .
it is even connected to field theory @xcite and nuclear reactions @xcite .
the formal framework of free - energy calculations is straightforward , e.g. , in terms of reaction coordinates in phase space , but the application in computer simulations is diverse with an ongoing demand for further methodological developments @xcite .
a seminal application was the parameter - free estimate of crystal nucleation rates from equilibrium free - energy barriers @xcite .
it seems natural that the estimation of nucleation barriers becomes increasingly difficult when considering more complex systems like polymer or protein solutions @xcite .
the rate of nucleation @xmath0 is related by classical nucleation theory @xcite to the free - energy cost @xmath1 of a nucleus on top of the nucleation barrier : @xmath2 with the inverse temperature @xmath3 and the boltzmann constant @xmath4 .
the kinetic prefactor @xmath5 includes the kinetic details of the nucleation process , such as diffusion and nucleus - attachment rates .
the free - energy barrier may be related to the suppression in the equilibrium probability distribution .
physically relevant barriers for liquid - vapor condensation are supposed to be in the range of @xmath6 to @xmath7 ( see , e.g. , ref .
@xcite ) .
the typical setup for the study of free - energy barriers is at fixed temperature by variation of the density or degree of supersaturation , or directly in the grand - canonical ensemble .
the barrier is then associated to the suppression in the droplet - size probability distribution @xcite and is clearly temperature dependent @xcite .
this usually requires estimating the droplet size and interface , a task that introduces systematic uncertainties and strongly depends on the droplet definition . instead
, the free - energy barrier may be directly related to the volume of the critical nucleus and the pressure difference @xcite , exploiting a thorough understanding of the underlying phenomenon in a clever way . in this context
, the problem can be reduced to conformational phase space , knowing that canonical expectation values typically do not depend on the kinetic energy . in the following ,
we address the question of how to easily obtain dependable results on nucleation barriers without invoking elaborate thermodynamic reasoning or estimating nucleus shapes .
this opens the door also to more complex systems with nucleation - like mechanisms such as self - assembly and aggregation .
importantly , we consider a setup at fixed density with varying temperature an intuitive approach from a condensed matter perspective .
analyzing a two - dimensional free - energy landscape , we identify the energy as a suitable reaction coordinate .
this allows us to formulate the problem in the microcanonical ensemble of either fixed total energy @xmath8 or fixed potential energy @xmath9 .
the first is the usual textbook definition , while the latter has been frequently applied in recent computer simulation studies .
this enables us to directly discuss the effect of kinetic energy when changing between the two formulations @xmath10 .
if kinetic energy matters , only the first one allows a direct physical interpretation . |
the hopf - laplace equation arises in the study of the dirichlet energy integral @xmath0= \iint_{{\mathbb{x } } } { \lvertdh\rvert}^2 = 2 \iint_{{\mathbb{x } } } \left ( { \lverth_z\rvert}^2 + { \lverth_{\bar z}\rvert}^2 \right)\ , { \textnormal d}z\ ] ] for homeomorphisms @xmath1 between two designated domains @xmath2 and @xmath3 in the complex plane @xmath4 . here and throughout this text we take advantage of the complex partial derivatives @xmath5 in fact complex notation will be indispensable for advancing this work , especially when quadratic differentials will enter the stage .
let us commence with the variational formulation of the classical dirichlet problem .
one asks for the energy - minimal mapping @xmath6 of the sobolev class @xmath7 whose boundary values are prescribed by means of a given mapping @xmath8 .
the _ first variation _
, @xmath9 , in which @xmath10 can be any test function and @xmath11 , leads to the laplace equation @xmath12 however , this approach is invalid when one seeks to minimize @xmath13 $ ] among homeomorphisms ; the injectivity of @xmath14 is usually lost .
the difficulty is circumvented by performing the _ inner variation _
@xmath15{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{x}}\ ] ] where @xmath16 are @xmath17-smooth automorphisms of @xmath2 onto itself , defined for parameters @xmath18 , and @xmath19{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{x}}$ ] .
this is simply a change of the independent variable in the domain of definition of @xmath20 . here
each @xmath16 may be the identity on @xmath21 , but it need not be .
the latter situation is called slipping along the boundary . the first derivative test @xmath22=0 $ ] , at @xmath23 , yields ( * ? ? ?
* lemma 1.2.4 ) what we call the _ hopf - laplace equation _ @xmath24
we shall investigate the hopf - laplace equation independently of its roots .
nevertheless , because of its affiliation to the dirichlet integral the natural domain of definition , that we shall always assume here , sometimes implicitly , is the sobolev space @xmath25 .
thus the differential expression @xmath26 will represent an integrable function . by weyl s lemma
the equation reduces equivalently to the first order nonlinear pde , @xmath27 we view @xmath28 also as unknown quantity . in this formulation
one can speak of generalized solutions in the sobolev space @xmath29 .
but we shall study , predominantly , the _ orientation preserving mappings _ , meaning that the jacobian determinant is nonnegative almost everywhere @xmath30 within the class of orientation preserving mappings we do not capture generalized solutions .
indeed , in this case @xmath31 , thus @xmath20 is ( locally ) a _ qc - deformation _ in the sense of ahlfors @xcite .
since the complex derivatives @xmath32 and @xmath33 are intertwined by the beurling - ahlfors transform , it follows that @xmath34 .
in particular , for all @xmath35 we have @xmath36 , hence @xmath37 for all @xmath38 .
the lipschitz regularity is much harder to handle ; this is the intricate part of our paper .
note that the qc - deformations of ahlfors are not necessarily lipschitz .
returning to the variational approach to the hopf - laplace equation , if one admits slipping along the boundary , then the inner variations of the dirichlet energy should include diffeomorphisms @xmath39{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{x}}$ ] that are free on @xmath21 . consequently , supplementary equations on @xmath40 emerge , which are best stated in terms of the _ holomorphic quadratic differential_. @xmath41 it is called the _ hopf differential _ of @xmath20 in recognition of the related work of h. hopf @xcite .
stated informally , the additional boundary equations ( * ? ? ?
* lemma 1.2.5 ) say that @xmath42 is real along @xmath21 , see definition [ realqd ] .
this additional boundary equation will be satisfied by _ minimal deformations _ , see .
clearly , conformal automorphisms @xmath39{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{x}}$ ] do not change the energy .
more generally , any conformal transformation of the domain @xmath43 does not affect the hopf - laplace equation ( [ hopfeq ] ) .
in fact , because of that , it is the shape of the target @xmath44 and its closure , called _ deformed configuration _ , that will really matter in questions to follow .
naturally , complex harmonic functions solve the hopf - laplace equation .
worth noting , is that real - valued solutions in @xmath45 must be harmonic .
if @xmath46 is @xmath47-smooth then the hopf - laplace equation yields @xmath48 thus @xmath49-solutions are harmonic in the region where the jacobian determinant @xmath50 .
there are other situations where the hopf - laplace equation implies harmonicity , see theorem [ thharm ] .
nevertheless , nonharmonic solutions arise naturally in global analysis of minimal surfaces ( * ? ? ?
2 ) , @xcite and in the calculus of variations ( * ? ? ?
21 ) , @xcite . in mechanics ,
in particular in elasticity theory , the hopf - laplace equation appears under the name of the energy - momentum equations .
let us look at some elementary though critical examples .
one might expect from that all @xmath51-solutions whose jacobian determinant does not vanish ( almost everywhere ) are harmonic maps .
this , however , is easily seen to be false , for in the complex plane the piece - wise linear mapping @xmath52 is not harmonic .
observe that the jacobian determinant changes sign .
more elaborate examples are provided in section [ examples : section ] , which reveal that : [ exnoimp ] for each @xmath53 there exists a generalized solution @xmath54 ( in the unit disk @xmath55 ) to the hopf - laplace equation @xmath56 whose first derivatives belong to the marcinkiewicz weak space @xmath57 but not to @xmath58 .
the message from this example is that without supplementary conditions of a topological nature the hopf - laplace equation will not guarantee any substantial improvement of the regularity of the solutions .
this is in marked contrast to the case of elliptic pdes , like the laplacian .
consequently , we focus on solutions in a suitable closure of homeomorphisms . mathematical models of nonlinear elasticity motivate our calling such solutions _ hopf deformations_. the obvious question to ask is what topological preconditions do we really need ?
partial answers are given in a few already known regularity results .
hlein @xcite proved that quasiconformal solutions are harmonic .
this is actually true @xcite for any homeomorphic solution in the sobolev space @xmath59 .
even more , injectivity is not necessary as long as the solution represents an open discrete map , such are the mappings of integrable distortion @xcite .
in particular , @xmath60-solutions ( not necessarily injective ) with jacobian @xmath61 are also harmonic . however , in example [ exsqueeze ] , we give a lipschitz solution with @xmath62 , almost everywhere in the unit disk @xmath55 , which is not @xmath63-smooth .
even more , this solution is a homeomorphism in @xmath64 except for a tiny crack along the radius which is squeezed into a point .
such a failure of injectivity along the cracks ( reminiscent of elastic deformations ) typically occurs when we pass to a weak limit of a minimizing sequence of homeomorphisms . with the loss of injectivity
the extremal mappings will not be harmonic .
the examples , like the one mentioned above , demonstrate that the solutions can be lipschitz at best .
our main results can be described as follows * lipschitz regularity ( theorem [ mainth ] ) : any hopf deformation is locally lipschitz .
* partial harmonicity ( theorem [ thharm ] ) : any hopf deformation @xmath20 restricts to a harmonic diffeomorphism of @xmath65 onto @xmath3 . * improved regularity ( theorem [ smooth ] ) : if @xmath3 is a polygon - type domain , then any @xmath66-smooth minimal deformation is a harmonic diffeomorphism . in a summary , we confirm the following principle . 0.2 cm _ topologically well behaved solutions to the hopf - laplace equation enjoy lipschitz regularity . _
0.2 cm our lipschitz regularity results apply to the minimal deformations ; that is , to minimizers of the dirichlet energy in the class of suitably generalized homeomorphisms .
an explicit example of a nonharmonic minimal deformations is the following @xcite : [ exhammering ] consider two annuli @xmath67 and @xmath68 where @xmath69 and @xmath70 .
the map @xmath71 takes @xmath72 into @xmath73 .
it satisfies the hopf - laplace equation @xmath74 the quadratic differential @xmath75 is real and positive along the boundary circles .
the regularity theory of energy minimizing mappings has a long history .
the monographs @xcite provide an overview of the subjects .
note that in @xcite the solutions of the hopf - laplace equation are called _ weakly noether harmonic maps_. numerous recent studies concern the lipschitz continuity of energy minimizers in the setting of metric spaces , where higher degrees of smoothness are not available @xcite .
however , in these studies minimization is performed among all mappings that are either prescribed on the boundary or belong to a given homotopy class .
this minimization problem allows one to use the _ first variation_. our approach is different in that we obtain the lipschitz continuity using only the inner variation .
thus we have in our disposal only the hopf - laplace equation instead of the laplacian . before rigorous statements we need to review some notation and basic definitions .
we shall be concerned with mappings @xmath76 between bounded planar domains of finite connectivity @xmath77 thus each boundary @xmath78 and @xmath79 consists of @xmath80 disjoint continua .
we reserve the notation , @xmath81 every homeomorphism @xmath82{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] gives rise to a one - to - one correspondence between boundary components of @xmath2 and boundary components of @xmath3 .
it will involve no loss of generality in assuming ( by re - arranging the indices , if necessary ) that the correspondence is : @xmath83 this simply means that @xmath84 as @xmath85 .
the above definition of the boundary correspondence is also pertinent to more general maps @xmath86{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , specifically to those which are proper and monotone in the sense defined below .
@xcite a continuous mapping @xmath87 between metric spaces @xmath2 and @xmath3 is _ monotone _ if for each @xmath88 the set @xmath89 is compact and connected .
it is _ proper _ if for each compact set @xmath90 the set @xmath91 is also compact .
the idea of the @xmath94-limit is a useful compromise between the concepts of @xmath95-uniform and uniform convergence . a sequence of mappings @xmath96
is said to converge _
@xmath94-uniformly _ to a mapping @xmath97 if * @xmath98 @xmath95-uniformly ( uniformly on compact subsets of @xmath2 ) * @xmath99 uniformly in @xmath100 we shall write it as @xmath101{{}_{cd}}}h\,$ ] and denote the class of @xmath94-limits of homeomorphisms @xmath102{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}\,$ ] satisfying by @xmath103 .
we emphasize that the boundary points of @xmath3 may be in the range of @xmath104 .
this fact is crucial for several results that follow .
precisely , we have @xmath105 if the range of a mapping @xmath106 equals @xmath107 then @xmath46 is both monotone and proper .
actually , we have an even more precise statement . [ uniappthm ] a continuous mapping @xmath108 between bounded @xmath109-connected domains belongs to @xmath93 if and only if it is monotone proper and surjective .
let us introduce the class @xmath110 of continuous mappings @xmath111 between bounded domains , which are monotone proper and surjective .
note that such mappings , as opposed to @xmath93 , do not take values in @xmath112 .
our notation is meant to emphasize this distinction .
the following is the extension of our earlier result @xcite on approximation of homeomorphisms to the setting of monotone proper mappings .
[ thm : diffeoapprox ] let @xmath113{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] be a continuous monotone proper mapping of sobolev class @xmath114 between bounded @xmath109-connected domains @xmath2 and @xmath3 .
then there exist diffeomorphisms @xmath115{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] such that * @xmath116 * @xmath117 .
hereafter @xmath118 is the royden algebra equipped with the norm @xmath119 and @xmath120 is a subalgebra obtained by completing the space @xmath121 with respect to this norm .
the following corollary is immediate from theorem [ thm : diffeoapprox ] .
[ jacdncs ] if @xmath122 , then the jacobian of @xmath20 does not change sign ; that is , either @xmath123 a.e . or @xmath124 a.e .
the concept of _ deformation _ seems to differ only a little from the routinely used @xmath126weak limit of homeomorphisms .
however , when the topological features of the minimal - energy solutions are of major concern the _ deformations _ become better suited than the weak limits of homeomorphisms .
for example , in @xcite the concept of _ deformations _ was critical in establishing existence of harmonic homeomorphisms between planar doubly connected domains . before making the precise definition let us look at some examples of both geometric and analytical nature .
a sequence of homeomorphisms converging weakly in @xmath126 actually converges @xmath95-uniformly , so the limit is a continuous map .
but that is all , in general , what the weak limit of a minimizing sequence can receive from homeomorphisms .
consider the classical example : conformal automorphisms @xmath127{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d\ , $ ] of the unit disk , @xmath128 are the minimal energy maps .
they converge weakly in @xmath129 and @xmath130-uniformly to a constant map .
such a trivial loss of topological distinctions is due to the failure of @xmath131-convergence . on the other hand , when sobolev mappings come into play , we find that a @xmath95-uniform limit of homeomorphisms @xmath132{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}\,$ ] that are uniformly bounded in @xmath133 satisfies the measure theoretical condition of _ non - overlapping _
@xmath134 due to the @xmath135-weak subconvergence of jacobians ( * ? ? ?
* theorem 8.4.2 ) . in nonlinear elasticity theory
this may be interpreted as saying that _ interpenetration of matter _ does not occur .
however , we will be forced to take into consideration more general minimizing sequences @xmath136 in which the mappings are neither homeomorphisms nor they have uniformly bounded energy .
we still impose the nonoverlapping condition . in view of corollary [ jacdncs ] the jacobian of @xmath20 does not change sign , and thus it entails no loss of generality to assume that @xmath137 a.e .
these observations drive us to the following [ deformations ] let @xmath2 and @xmath3 be bounded @xmath109-connected domains .
a deformation is a mapping @xmath138 such that * @xmath139 * the jacobian determinant of @xmath20 is nonnegative a.e . and satisfies the _ non - overlapping condition _
@xmath140 the class of all deformations will be denoted by @xmath141 . throughout what follows ,
if no confusion can arise , we shall freely assume without explicit mention that the class @xmath142 is nonempty .
we again strongly emphasize that a deformation @xmath143 may take points of @xmath2 into @xmath144 , a key point that will affect the forthcoming arguments .
the structure of the preimage @xmath145 of a point in @xmath146 is a delicate issue that we address in sections [ secpreimages ] and [ secsmo ] . in @xcite
we have already established the essential properties of deformations . in particular ,
the class @xmath147 is sequentially weakly closed in @xmath148 , provided that @xmath2 has @xmath149 boundary components none of which are points ( * ? ? ?
* lemma 3.13 ) . under these hypotheses
we have the following .
the infimum energy of the dirichlet energy within the class @xmath147 is attained .
a mapping @xmath150 such that @xmath151 will be called a _ minimal deformation_. [ hopfdefdef ] let @xmath2 and @xmath3 be bounded @xmath109-connected domains .
a deformation @xmath152 which satisfies the hopf - laplace equation @xmath153 is called a _
hopf deformation_. thus hopf deformations are among stationary solutions ( with respect to the inner variation ) of the dirichlet energy integral .
we shall prove that , [ thharm ] any hopf deformation @xmath154 is a harmonic diffeomorphism of @xmath155 onto @xmath3 .
observe that in example [ exhammering ] the mapping @xmath20 fails to be harmonic exactly in the part of @xmath2 that is squeezed into a boundary component of the target annulus @xmath3 .
more precisely , this particular component of @xmath156 is the inner boundary circle at which @xmath44 is not convex .
in particular , when the range of @xmath20 is @xmath44 , we can combine theorem [ thharm ] with proposition [ uniappthm ] to obtain let @xmath2 and @xmath3 be bounded @xmath157-connected domains and @xmath158 a monotone proper mapping of sobolev class @xmath159 that satisfies the hopf - laplace equation .
then @xmath20 is a harmonic diffeomorphism .
this corollary represents another advance on the eells - lemaire problem @xcite : under what conditions does the holomorphicity of the hopf differential imply that the mapping is harmonic ? the foremost interesting and much harder task is to determine the regularity of a hopf deformation when its image @xmath160 goes over the designated target @xmath3 ; that is , into its closure .
there are minimal deformations , in particular hopf deformations , which are not @xmath66-smooth in @xmath2 , see corollary [ cor45 ] . our main result in this paper is the following best possible regularity of hopf deformations .
[ mainth ] every hopf deformation @xmath161 is locally lipschitz continuous . in particular ,
a minimal deformation of planar @xmath157-connected domains is locally lipschitz continuous .
[ smooth ] suppose that @xmath162 is a minimal deformation and the nonconvex part of @xmath163 is at most countable .
then @xmath20 is @xmath66-smooth if and only if it is a harmonic homeomorphism . the term _ nonconvex part of _ @xmath163 refers to a set of points in @xmath163 where @xmath3 is not convex in any neighborhood .
precisely , we say that @xmath3 is convex at @xmath146 if for some @xmath164 the set @xmath165 is convex .
a natural example of a domain with finite nonconvex part of the boundary is an @xmath109-connected polygonal domain . according to ( * ? ? ?
* theorem 2.4 ) there exists a nondecreasing function @xmath166 such that the following holds .
whenever two bounded doubly connected domains @xmath2 and @xmath3 admit an energy minimizing diffeomorphism @xmath113{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , we have @xmath167 where @xmath168 stands for the conformal modulus .
now combining theorem [ smooth ] with the nonexistence of energy minimizing diffeomorphisms , we arrive at the following corollary .
[ cor45 ] let @xmath2 and @xmath3 be bounded doubly - connected domains such that the nonconvex part of @xmath163 is at most countable and fails .
then any minimal deformation in @xmath142 is not @xmath66-smooth .
-smooth.,scaledwidth=80.0% ] we conclude this introduction with a brief outline of the proofs of the main results . to prove the partial harmonicity of a hopf deformation @xmath20 we show that @xmath20 has a stronger energy minimization property in @xmath65 than in all of @xmath2 , namely holds .
the three main ingredients of this proof are : diffeomorphic approximation of monotone sobolev mappings ( [ secapp ] ) , a reich - strebel - type inequality ( lemma [ reich - strebel ] ) , and the structure of preimages of points under @xmath20 ( [ secpreimages ] ) .
the partial harmonicity of @xmath20 is essential in the proof of its lipschitz continuity , however a different approach is required for the set @xmath169 . to this end
we use a bonnesen - type inequality , i.e. , a stability result for the isoperimetric inequality ( [ sectionlip ] ) .
the proof of theorem [ smooth ] invokes besicovitch s removability theorem for holomorphic functions , further analysis of preimages of points under @xmath20 , and the boundary point lemma of e. hopf .
one method of proving lipschitz continuity of @xmath20 proceeds through the subharmonicity of the energy density @xmath170 , as in @xcite and @xcite .
this method relies on the first variation of @xmath20 , which is not available to us .
although the hopf - laplace equation appears to imply the subharmonicity of @xmath170 on the formal level , our example [ exnoimp ] shows that this is not the case .
the uniform limit of self - homeomorphisms of the sphere @xmath171 is a monotone surjective mapping ( * ? ? ?
* ix.3.11 ) . in the converse direction , a monotone mapping of @xmath171 onto itself
can be refined in some jordan subdomains of @xmath171 in which it becomes a homeomorphism .
let @xmath2 and @xmath3 be @xmath109-connected domains and @xmath172 . recall that @xmath173 is a jordan domain if it is the interior of a homeomorphic image of the closed unit disk .
[ homext ] for every jordan domain @xmath173 and @xmath172 the preimage @xmath174 is a simply connected domain in @xmath2 . furthermore , there is @xmath175 , _ the youngs refinement _ , such that * @xmath176 in @xmath177 .
* @xmath178 restricted to @xmath179 is a homeomorphism of @xmath179 onto @xmath180 .
this is a consequence of the youngs modification theorem ( * ? ? ?
* theorem 10.1 ) or ( * ? ? ?
* ii.1.47 ) about continuous monotone mappings @xmath181{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 $ ] .
it says that to every jordan domain @xmath182 there corresponds a continuous monotone mapping @xmath183{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 $ ] such that * @xmath184 in @xmath185 * @xmath186 maps @xmath187 homeomorphically onto @xmath188 .
we prove proposition [ homext ] by applying the youngs theorem to the @xmath109-point compactifications of @xmath2 and @xmath3 .
indeed , both @xmath2 and @xmath3 are homeomorphic to the sphere @xmath171 with @xmath109 punctures , say via the maps @xmath189 and @xmath190 where the punctures are enumerated in the same way as the boundary components . for any mapping
@xmath172 the composition @xmath191 extends to a continuous mapping @xmath192{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 , \quad \mbox { with } f \big(\phi ( x_\nu)\big ) = \psi ( y_\nu ) , \quad \nu=1 , \dots , \ell.\ ] ] any simply connected domain @xmath193 is mapped via @xmath194 onto a simply connected domain @xmath195 which stays away from the punctures . by the youngs theorem @xmath187 is a simply connected domain in @xmath171 .
it follows that @xmath196 is a simply connected domain compactly contained in @xmath2 .
the homeomorphism @xmath197{{}_{\!\!\textnormal{onto}\!\!}}}\delta$ ] ( the youngs refinement ) yields a homeomorphism @xmath198{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] .
first , suppose that @xmath199{{}_{\!\!\textnormal{into}\!\!}}}{\mathbb{y}}$ ] is a @xmath94-limit of homeomorphisms @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] .
therefore , @xmath201 , by .
the preimage of a compact set in @xmath3 , under @xmath202 , is closed and stays away from @xmath21 because @xmath202 satisfies .
thus @xmath202 is proper . to show monotonicity we appeal to the induced mappings @xmath203{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 $ ] , @xmath204 as in .
these are homeomorphisms of @xmath171 onto itself converging uniformly to @xmath205 .
thus @xmath206 is monotone ( * ? ? ?
* theorem 11.1 ) . since @xmath199{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ]
we see that @xmath202 is monotone as well . in the converse direction
we must show that every mapping @xmath172 belongs to @xmath207 . in the proof of proposition [ uniappthm ] we shall appeal to the classical youngs approximation theorem ( * ? ? ?
* theorem 11.1 ) ( or ( * ? ? ?
* ii.1.57 ) ) .
it asserts that 0.2 cm _ a continuous mapping @xmath208{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 $ ] is monotone if and only if it is a uniform limit of homeomorphisms of @xmath171 onto @xmath171 .
_ 0.2 cm let us compactify @xmath2 and @xmath3 as in the proof of proposition [ homext ] , via the homeomorphisms @xmath209{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2\setminus\{x_1,\dots , x_{\ell}\}$ ] and @xmath210{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2\setminus\{y_1,\dots , y_{\ell}\}$ ] .
for any @xmath172 the induced mapping @xmath211{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 $ ] , being monotone , can be uniformly approximated by homeomorphisms @xmath203{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{s}}^2 $ ] . by altering @xmath212 near punctures
we can ensure that @xmath213 for @xmath214 .
we then return to the domains @xmath2 and @xmath3 via @xmath215 and @xmath216 to define homeomorphisms @xmath217{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}\ ] ] and observe that @xmath218{{}_{cd}}}f$ ] by construction .
the youngs refinement provides a homeomorphic replacement of a monotone mapping .
the following proposition shows that such a replacement can be chosen to be harmonic .
it combines classical results of potential theory @xcite with a recent extension of the rad - kneser - choquet theorem .
[ poisson ] let @xmath219 be a domain .
to every bounded simply connected domain @xmath220 there corresponds a unique linear operator @xmath221 such that for every @xmath222 1 .
@xmath223 in @xmath177 .
@xmath224 is harmonic in @xmath179 .
such an operator has the additional properties @xcite 1 .
@xmath225 , whenever @xmath226 .
moreover , @xmath227 \le { \mathcal{e}}_{\mathbb u } [ f ] , \qquad \mbox{provided } \quad { \mathcal{e}}_{\mathbb u } [ f ] < \infty .\ ] ] 2 .
suppose that the restriction @xmath228 of @xmath229 is a homeomorphism of @xmath179 onto a convex domain @xmath230 , then @xmath231{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] is a harmonic diffeomorphism .
we call @xmath232 the _ harmonic replacement _ of @xmath202 .
in this section we recall some basic facts about holomorphic quadratic differentials .
the general reference for these topics is @xcite .
let @xmath2 will be a bounded @xmath109-connected domain and @xmath233 a holomorphic function with isolated zeros , called _
critical points_. denote @xmath234 . in a neighborhood of every point @xmath235
one can introduce a local conformal mapping @xmath236 , called a natural parameter near @xmath237 . through every regular point there pass two @xmath17-smooth orthogonal arcs , called horizontal and vertical arcs .
a vertical arc is a @xmath17-smooth curve @xmath238 , @xmath239 , along which @xmath240 ^ 2 \varphi \big(\gamma ( t)\big)<0 , \quad a < t < b.\ ] ] a horizontal arc is a @xmath17-smooth curve @xmath241 , @xmath242 , along which @xmath243 ^ 2 \varphi \big(\beta ( t)\big)>0 , \quad c < t < b.\ ] ] we emphasize that this yields , in particular , that such arcs only contain regular points of @xmath28 .
a _ vertical trajectory _ of @xmath28 in @xmath2 is a maximal vertical arc ; that is , not properly contained in any other vertical arc .
hereafter , with the customary abuse of notation , the same symbol @xmath244 will be used for both the parametrization @xmath245 and its range .
similarly , a _ horizontal trajectory _ is a maximal horizontal arc in @xmath2 . through every regular point of @xmath28
there passes a unique vertical ( horizontal ) trajectory . a trajectory whose closure contains a critical point of @xmath28
is called a _
critical trajectory_. there are at most a countable number of critical trajectories , so they cover a set in @xmath2 of measure zero
. we will be largely concerned with noncritical trajectories .
( @xmath28-rectangle ) a @xmath28-rectangle of a quadratic differential @xmath42 is any simply connected domain @xmath246 on which the natural parameter @xmath247 has a univalent branch which takes @xmath248 onto a euclidean rectangle @xmath249 note that @xmath248 contains no zeros of @xmath28 .
we will be concerned with @xmath28-rectangles which are compactly contained in @xmath250 , so @xmath215 defines a diffeomorphism of a neighborhood of @xmath251 onto a neighborhood of @xmath252 .
then we define the horizontal edges of @xmath248 , @xmath253 \times \{a\}\right)$ ] and @xmath254 \times \{b\}\right)$ ] and similar for the vertical edges . _
every noncritical vertical trajectory @xmath255 in a simply connected domain @xmath256 is a cross cut _ , see theorem 15.1 in @xcite .
thus in the maximal interval @xmath239 of the existence of @xmath245 both limit sets at the end - points of @xmath244 , denoted by @xmath257 and @xmath258 , lie in @xmath259 .
let @xmath260 be any closed vertical subarc of @xmath244 , defined by @xmath261 for @xmath262 , where @xmath263 .
then the @xmath28-length of @xmath264 is equal or smaller than @xmath28-length of any rectifiable curve @xmath265 which connects @xmath266 with @xmath267 .
this means that @xmath268 see ( * ? ? ?
* theorem 16.1 ) .
note that @xmath269 consists of two components ( two disjoint vertical arcs ) .
inequality can be slightly generalized ; it is not necessary to assume that the end - points of @xmath270 coincide with the endpoints of @xmath264 .
[ lengthineq ] let @xmath265 be a locally rectifiable arc in a simply connected region whose closure intersects both components of @xmath269 , then @xmath271 let @xmath272 be points in different components of @xmath269 that are approachable through the arc @xmath270 ; that is , @xmath273 let @xmath274_\gamma$ ] denote subarc of @xmath244 that connects @xmath275 and @xmath276 .
we certainly have @xmath277_\gamma$ ] , so @xmath278_\gamma } { \lvert\varphi\rvert}^{\nicefrac{1}{2}}\ , { \lvert{\textnormal d}z\rvert}.\ ] ] similarly , we denote by @xmath279_\beta \subset \beta$ ] the closed ( rectifiable ) subarc of @xmath270 which connects @xmath280 and @xmath281 . since @xmath282 and @xmath283 for sufficiently large @xmath284 the straight segments @xmath285
$ ] and @xmath286 $ ] lie in @xmath256 .
we now have a rectifiable curve @xmath285 \cup [ a_n , b_n]_\beta \cup [ b_n , b]$ ] in @xmath256 which connects the end - points of @xmath274_\gamma$ ] .
therefore , we have @xmath287_\gamma } { \lvert\varphi\rvert}^{\nicefrac{1}{2 } } \le \int_{[a_n , b_n]_\gamma } { \lvert\varphi\rvert}^{\nicefrac{1}{2 } } + \int_{[a , a_n ] } { \lvert\varphi\rvert}^{\nicefrac{1}{2 } } + \int_{[b_n , b ] } { \lvert\varphi\rvert}^{\nicefrac{1}{2 } } \\ & \le \int_{\beta } { \lvert\varphi\rvert}^{\nicefrac{1}{2}}+ \left ( { \lverta_n -a\rvert } + { \lvertb_n - b\rvert } \right){\lvert\varphi\rvert}_{\mathscr c ( \omega)}^{\nicefrac{1}{2 } } \longrightarrow \int_{\beta } { \lvert\varphi\rvert}^{\nicefrac{1}{2 } } , \end{split}\ ] ] as desired . the next lemma deals with a holomorphic quadratic differential @xmath288 which is real on the boundary of a @xmath66-smooth domain @xmath2 , no single points as components of @xmath21 . [ realqd ]
a quadratic differential @xmath288 is said to be real on the boundary of a @xmath66-smooth @xmath109-connected domain @xmath2 if @xmath28 is smooth up to @xmath289 and each component of @xmath21 is either a horizontal or a vertical trajectory of @xmath288 .
[ callsomething ] let @xmath2 be a finitely connected domain with @xmath66-smooth boundary .
let @xmath288 be a holomorphic quadratic differential in @xmath2 which is real on @xmath21 .
suppose that @xmath290 is a vertical trajectory of @xmath288 with both ends approaching the same boundary component of @xmath2 .
then the components of @xmath291 are not simply connected .
suppose that the set @xmath291 has a simply connected component @xmath292 .
there are no closed trajectories in @xmath292 , for such a trajectory must enclose a pole of @xmath28 .
the global structure of trajectories of a holomorphic quadratic differential with finite norm @xcite is inconsistent with @xmath293 being a union of a vertical trajectory and another ( vertical or horizontal ) trajectory .
[ fublem ] let @xmath42 be a holomorphic quadratic differential in a simply connected domain @xmath294 , @xmath295 .
suppose that @xmath206 and @xmath296 are measurable functions in @xmath256 such that @xmath297 then for almost every vertical trajectory @xmath244 of @xmath42 we have @xmath298 if , in addition , @xmath299 then @xmath300 according to @xcite @xmath256 can be covered , up to a set of measure zero , by a countable number of disjoint @xmath28-strips . these are open connected subsets of @xmath256 with no critical points , such that a locally defined analytic function @xmath301 is actually a univalent conformal mapping of the @xmath28-strip onto a euclidean vertical strip @xmath302 in the @xmath303-plane , @xmath304 .
@xmath305 where @xmath306 are measurable functions . making a substitution @xmath307
the problem reduces equivalently to the usual fubini s theorem in euclidean vertical strip , for functions @xmath308 and @xmath309 .
[ fubcor ] assume , instead of condition in lemma [ fublem ] that @xmath310 for almost every noncritical vertical trajectory of @xmath311 .
then @xmath312 replace @xmath206 and @xmath296 in lemma [ fublem ] , with @xmath313 and @xmath314 , where @xmath315 given a quadratic holomorphic differential @xmath288 we define two partial differential operators , called the _ horizontal _ and _ vertical derivatives _
@xmath316 if @xmath20 satisfies the hopf - laplace equation @xmath317 , then the horizontal and vertical trajectories of @xmath288 are the lines of maximal and minimal stretch for @xmath20 .
precisely , the following identities hold . @xmath318 as a consequence @xmath319
actually such solutions can be defined in the entire plane .
first we define @xmath20 in the upper half plane , @xmath320 , where one can settle the analytic branches of power functions and the logarithm .
@xmath321 we have @xmath322 thus the hopf - laplace equation @xmath323 holds in the upper half plane
. then we extend @xmath20 to the lower half of the plane by setting @xmath324 for @xmath325 .
it is a general fact , and easy to see , that such an extension gives a sobolev function in the entire plane .
the hopf - laplace equation remains true in the lower half of the plane as well .
-solution to the hopf - laplace equation , scaledwidth=95.0% ] figure [ fig1 ] illustrates the case @xmath326 .
thus @xmath327 belongs to @xmath328 for every @xmath329 , but is not locally lipschitz continuous .
[ exsqueeze ] we use the polar coordinates for @xmath330 in the closed unit disk @xmath331 , @xmath332 , @xmath333 and @xmath334 .
define @xmath335 @xmath336 = z- \bar z - i \left [ z^{\nicefrac{3}{2}}- \bar z^{\nicefrac{3}{2 } } \right].\ ] ] this mapping is lipschitz continuous , since it has bounded derivatives @xmath337 moreover , its hopf differential is holomorphic , @xmath338 . thus @xmath20 solves the hopf - laplace equation @xmath339 .
formulas show that @xmath20 fails to be @xmath66-smooth in any neighborhood of the ray @xmath340 . concerning topological behavior
, @xmath20 turns out to be a harmonic diffeomorphism of @xmath341 onto the butterfly domain @xmath342 .
figure [ fig2 ] shows the grid of horizontal and vertical trajectories in @xmath2 as well as their images in @xmath3 .
hopf deformation , scaledwidth=80.0% ] the radius @xmath343 is squeezed into the origin , which is a boundary point of @xmath3 .
figure [ fig2 ] illustrates that @xmath343 is an arc of a critical vertical trajectory of the quadratic differential @xmath344 .
let us notice that the functions @xmath345 and @xmath346 are actually continuous .
indeed , we have @xmath347 in particular , the jacobian determinant and the energy density function are also continuous @xmath348 it is easy to see that @xmath20 is a @xmath94-limit of homeomorphisms of @xmath180 onto @xmath3 .
thus @xmath20 is a hopf deformation . in section [ secsmo ]
we demonstrate through nonexplicit examples , that even minimal deformations need not be @xmath66-smooth .
let us begin by recalling the following approximation of homeomorphisms , established in @xcite .
let @xmath349{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] be a homeomorphism of sobolev class @xmath350 .
then there exist diffeomorphisms @xmath351{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , @xmath352 , such that * @xmath353 . *
@xmath354 as @xmath355 . in view of this result
we need only construct , for every @xmath356 , a homeomorphism @xmath349{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] such that a. @xmath357 .
b. [ ii ] @xmath358 . the construction of @xmath359 proceeds in three steps .
we construct @xmath110 mappings @xmath360 , @xmath361 , @xmath362 and @xmath363 , in which @xmath364 will turn out to be a desired homeomorphism of @xmath2 onto @xmath3 . in each step
we make suitable harmonic replacements to gain more points of injectivity .
moreover , estimate will follow from : @xmath365 we shall go into the construction of @xmath366 in detail in step 1 . for @xmath367
we follow the construction from step 1 , but with @xmath366 in place of @xmath20 . in much the same way
@xmath364 will be obtained as a refinement of the mapping @xmath367 . before passing to the actual construction of @xmath366
we need some geometric considerations .
an open dyadic square in @xmath368 is the set @xmath369 hereafter , the number @xmath370 is the size of the square .
note that : 1 .
two different squares of the same size are disjoint .
each square of size @xmath370 is contained in exactly one square of size @xmath371 , namely @xmath372 we refer to @xmath373 as the dyadic square _
next _ to @xmath374 . 1 .
every two dyadic squares are either disjoint or one contains the other .
a _ dyadic mesh _ in @xmath368 is a family @xmath375 of open dyadic squares .
let @xmath3 be a bounded domain .
we will be interested only in those dyadic squares which are compactly contained in @xmath3 .
call such a dyadic square @xmath376 maximal if the next dyadic square to @xmath377 is not compactly contained in @xmath3 .
denote by @xmath378 clearly @xmath379 is a disjoint family and @xmath380 * claim 1 . *
_ every compact subset @xmath381 intersects at most a finite number of closed squares @xmath382 , where @xmath383 . _ for ,
if not , we would find an arbitrarily small square @xmath383 whose closure intersects @xmath384 , because one can accommodate only a finite number of large squares in @xmath3 .
but then the next dyadic square , being small enough , would be compactly contained in @xmath3 .
this contradicts maximality of @xmath383 . in what follows
we shall subdivide each @xmath383 into @xmath385 congruent dyadic subsquares , later referred to as fine squares .
the numbers @xmath386 will be chosen and fixed according to the needs for the construction of @xmath366 . in the meantime , let us reserve a notation and point out basic features of fine squares .
@xmath387 these are disjoint open dyadic squares of size @xmath388 , where @xmath370 is the size of @xmath377 . for each @xmath389
we have @xmath390 once a subdivision of each @xmath383 is made the family of fine squares will be denoted by @xmath391 this is a disjoint family of squares whose closures cover the entire domain @xmath3 .
as in claim 1 , we have * claim 2 .
* _ every compact set @xmath381 intersects at most a finite number of closed fine squares .
_ we now proceed to the construction of the mapping @xmath366
. * step 1 .
* let @xmath376 be a generic square in @xmath379 , and @xmath392 , @xmath393 , the corresponding fine squares , with @xmath386 to be determined later .
since @xmath394 , by proposition [ homext ] , each preimage @xmath395 is a simply connected domain compactly contained in @xmath2 .
we refer to @xmath396 as cells in @xmath2 .
caveat lector the closed cell @xmath397 can be substantially smaller than @xmath398 ; it may even lie in the interior of @xmath398 .
the youngs refinement , proposition [ homext ] , tells us that @xmath399{{}_{\!\!\textnormal{onto}\!\!}}}q_\alpha$ ] admits a homeomorphism @xmath400{{}_{\!\!\textnormal{onto}\!\!}}}q_\alpha$ ] with continuous extension to @xmath2 .
the extended mapping , still denoted by @xmath401{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , coincides with @xmath20 on @xmath402 and belongs to @xmath110 .
however , the youngs refinement does not guarantee that @xmath403 belongs to @xmath350 . at this point , since @xmath404 , the poisson operator in proposition [ poisson ] comes to the rescue .
we simply replace each @xmath400{{}_{\!\!\textnormal{onto}\!\!}}}q_\alpha$ ] with a harmonic diffeomorphism @xmath405{{}_{\!\!\textnormal{onto}\!\!}}}q_\alpha , \quad h_\alpha \in h+ \mathscr a_\circ ( \mathbb u_\alpha)\ ] ] and extend to @xmath2 by setting @xmath406 on @xmath407 .
this yields a mapping @xmath408{{}_{\!\!\textnormal{onto}\!\!}}}q , \quad \mbox{defined by } h_q^n = h+ \sum_{\alpha = 1}^{4^n } [ h_\alpha -h]_\circ\ ] ] where @xmath409_\circ$ ] stands for the function in @xmath2 that equals @xmath410 in @xmath411 and vanishes outside @xmath411 .
the energy of each @xmath412 does not exceed that of @xmath20 , namely @xmath413\le { \mathcal{e}}_{{\mathbb{u}}_\alpha } [ h]$ ] .
outside of the cells @xmath411 the mapping @xmath414 coincides with @xmath20 , thus has the same energy as @xmath20 .
therefore , 1 .
@xmath416 \le { \mathcal{e}}_{\mathbb{u}}[h]$ ] .
we also have @xmath417 when @xmath284 increases to @xmath418 the mappings @xmath419 converge uniformly to @xmath20 on @xmath420 .
furthermore , they are bounded in @xmath421 .
thus @xmath419 converge weakly to @xmath20 in @xmath421 . by weak lower semicontinuity
, we have @xmath422 \le \liminf_{n \to \infty } { \mathcal{e}}_{\mathbb{u}}[h_q^n ] \le { \mathcal{e}}_{\mathbb{u}}[h]\ ] ] so @xmath423 \to { \mathcal{e}}_{\mathbb{u}}[h ] $ ] .
we now recall the well known fact that if functions in @xmath424 converge weakly and their norms converge to the norm of the weak limit then such functions actually converge strongly .
it is at this stage that we choose and fix number @xmath386 , which will also depend on @xmath425 , to be large enough to satisfy @xmath426 & \le \frac{{\lvertq\rvert } \ , \epsilon^2 } { { \lvert{\mathbb{y}}\rvert}}. \end{split}\ ] ] finally , we conjoin all mappings @xmath427 , with @xmath383 and @xmath428 .
we obtain the desired mapping @xmath429{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , @xmath430_\circ \in \mathscr a_\circ ( { \mathbb{x } } ) .\ ] ] clearly , we have @xmath431 & = \sum_{q\in \mathcal m ( { \mathbb{y } } ) } { \mathcal{e}}_{{\mathbb{u } } } [ h_q^n -h ] \le \epsilon^2 \sum_{q \in \mathcal m ( { \mathbb{y } } ) } \frac{{\lvertq\rvert}}{{\lvert{\mathbb{y}}\rvert } } = \epsilon^2 . \end{split}\ ] ] hence the estimate , @xmath432
what we gained , as compared to @xmath20 , is that the mapping @xmath429{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] is a harmonic diffeomorphism on every cell @xmath433 thus @xmath434 for other preimages , we have @xmath435 in either case the preimage of a point in @xmath3 is connected .
thus @xmath366 is a monotone mapping .
similarly we argue that @xmath366 is a proper mapping .
indeed , let @xmath384 be compact in @xmath3 .
there are only finite number of closed fine squares in @xmath436 which intersect @xmath384 .
therefore the preimage of @xmath384 under @xmath366 is contained in the union of a finite number of closed cells in @xmath2 and in @xmath437 .
thus @xmath438 stays away from @xmath21 and , being relatively closed in @xmath2 , is indeed compact .
step 1 is completed .
* steps 2 and 3 . * in step 1 the construction of @xmath429{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] started with a mesh @xmath375 of dyadic squares in @xmath368 ; let us now redenote this mesh as @xmath439 .
such a mesh actually depends on the choice of the orthogonal coordinates for @xmath368 .
translating the origin of the coordinate system leads to new meshes .
these meshes would work for the constructions of @xmath366 just as @xmath439 .
it would , however , lead us to different dyadic squares in @xmath3 , different family @xmath440 , and different cells in @xmath2 .
we shall take advantage of this observation by considering three _ incommensurate _ meshes in @xmath368 .
one way to construct incommensurate meshes is by shifting the squares in @xmath375 through a vector with irrational coordinates , say @xmath441 .
specifically , let @xmath442 the key observation is that no three squares from different meshes have a common boundary point ; that is , @xmath443 whenever @xmath444 , @xmath445 and @xmath446 .
recall the corresponding families of open fine squares @xmath447 , @xmath448 and @xmath449 . in each family
the closures of fine squares cover the entire domain @xmath3 .
but the essential feature of these families is that the open fine squares all together cover @xmath3 , in symbols @xmath450 we are now ready for the construction of @xmath367 and @xmath364 .
following the construction of @xmath366 in step 1 , but with @xmath366 in place of @xmath20 and with mesh @xmath451 in place of @xmath452 we obtain a mapping @xmath453{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] which is continuous monotone and proper .
moreover , @xmath454 then , in the same fashion , we refine @xmath367 by using the mesh @xmath455 .
we arrive at the desired mapping @xmath456{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] such that @xmath457 . to see that @xmath364 is a homeomorphism we need only verify injectivity .
let @xmath458 and suppose , to the contrary , that @xmath459 is not a singleton .
this means that @xmath460 so @xmath461 lies in the boundary of some square @xmath462 .
recall that for such a boundary point we have @xmath463 .
this in turn means that @xmath464 , so @xmath461 lies in the boundary of a square @xmath465 .
for such a point we have @xmath466 . as before
, this means that @xmath467 , so @xmath461 lies in the boundary of a square @xmath468 .
in conclusion , @xmath461 belongs to @xmath469 .
this contradicts .
the proof of theorem [ thm : diffeoapprox ] is complete .
next we apply this theorem to harmonic @xmath470 mappings .
[ harmdiff ] any harmonic mapping of class @xmath471 is a diffeomorphism . indeed , suppose @xmath472 , where we note that @xmath32 and @xmath473 are holomorphic functions .
since @xmath20 is surjective , @xmath474 , which means that @xmath32 admits only isolated zeros . then we obtain a meromorphic function @xmath475 which is bounded by @xmath476 .
the zeros of @xmath32 are removable singularities .
the maximum principle yields @xmath477 in @xmath2 ; because it can not be that @xmath478 .
thus @xmath479 everywhere in @xmath2 and , therefore , @xmath20 is a local diffeomorphism .
the monotonicity implies that @xmath20 is actually injective .
such deformations can now be completely characterized as follows .
[ thmequiv2 ] let @xmath480 be a continuous mapping between @xmath157-connected bounded domains in the sobolev class @xmath481 with nonnegative jacobian .
then the following seven statements are equivalent : * @xmath202 is monotone proper and surjective .
* @xmath202 is a uniform and strong @xmath482limit of homeomorphisms @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , in which @xmath483 , for all @xmath484 .
* @xmath202 is a uniform and strong @xmath482limit of @xmath485diffeomorphisms @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , in which @xmath483 , for all @xmath484 .
* @xmath202 is a uniform and weak @xmath486limit of @xmath487diffeomorphisms @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , in which @xmath483 , for all @xmath484 .
* @xmath202 is a uniform and weak @xmath486limit of homeomorphisms @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , in which @xmath483 , for all @xmath484 .
* @xmath202 is a weak @xmath486limit of homeomorphisms @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] , in which @xmath483 , for all @xmath484 .
* @xmath202 is a deformation .
the assumption @xmath488 a.e .
does not impose an essential restriction on the mapping @xmath202 , by virtue of corollary [ jacdncs ] .
the implications @xmath489 @xmath489 are just a restatement of theorem [ thm : diffeoapprox ] .
the implications @xmath489 @xmath489 @xmath489 are obvious . for the proof of @xmath489 we argue as follows . since @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] are orientation preserving homeomorphisms of sobolev class @xmath490 converging weakly in @xmath481 to @xmath202 , it follows that @xmath491 hence there is a compact subset @xmath492 such that @xmath493 by weak @xmath135-convergence of nonnegative jacobians ( * ? ? ?
* theorem 8.4.2 ) we have @xmath494 . therefore , @xmath495 it then follows that there is @xmath356 , such that @xmath496 let @xmath497 be a point for which @xmath498 has distance at least @xmath425 from @xmath163 .
we may assume , by passing to a subsequence if necessary , that @xmath499 let @xmath500{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{x}}$ ] and @xmath501{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] be local perturbations ( arbitrarily small ) of the identity mapping near @xmath502 and @xmath503 , respectively , to satisfy , @xmath504 now the homeomorphisms @xmath505{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] take @xmath502 into @xmath503 .
we consider @xmath212 as a mapping of the punctured domain @xmath506 onto @xmath507 , each of which has @xmath508 boundary components .
these mappings coincide with @xmath509 outside a compact subset of @xmath2 . at this point
we appeal to the following uniform estimate of the distance to @xmath510 ( * ? ? ?
* theorem 1.1 ) .
@xmath511 where @xmath512 is a continuous function in @xmath513 vanishing on @xmath514 .
we emphasize that this function depends only on the domains @xmath250 and @xmath515 .
since @xmath516 near @xmath21 the estimate yields , for each @xmath509 , @xmath517 where @xmath518 is controlled from above by the energy of @xmath509 , so is independent of @xmath519 . finally , since @xmath520 @xmath95-uniformly we conclude that @xmath521 uniformly in @xmath2 .
this shows that @xmath202 is a deformation .
the implication @xmath489 is a part of proposition [ uniappthm ] .
the observant reader may notice that the conditions tell us something about the boundary behavior of a deformation with range @xmath3 . in a way every deformation @xmath199{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] must agree on @xmath21 with a homeomorphism @xmath200{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{y}}$ ] .
this is understood in the sense of sobolev boundary data @xmath522 .
let @xmath523 be a continuous mapping .
the multiplicity function of @xmath20 defined by @xmath524 , @xmath525 , is measurable , so one can speak of the essential supremum of @xmath526 .
we are concerned with mappings such that @xmath527 note that hopf deformations enjoy the property ( * ? ? ?
* lemma 3.8 ) @xmath528 the following proposition deals with more general solutions to the hopf equation .
[ lily ] let @xmath529 be a continuous @xmath530-solution to the hopf - laplace equation @xmath531 where @xmath28 is a holomorphic function in a domain @xmath219 .
assume that the multiplicity function @xmath524 is essentially bounded .
then for each @xmath532 the union of all vertical trajectories of the quadratic differential @xmath42 in @xmath2 that intersect @xmath533 has zero measure . to simplify writing we assume that @xmath534 .
let @xmath535 denote the family of all vertical trajectories of @xmath42 in @xmath2 .
these are disjoint open @xmath17-smooth curves without self - intersections whose union covers @xmath536 .
every point in @xmath250 has a neighborhood in which a single valued branch of the analytic function @xmath301 can be chosen .
this is a local conformal mapping which takes the arcs of vertical trajectories into open vertical intervals in the @xmath303-plane , @xmath537 . in general
it may not be possible to perform analytic continuation of @xmath215 along the entire trajectory ; the local branches of @xmath538 may not coincide if their domains of definition are overlapping .
this difficulty is usually overcome by performing analytic continuation of the inverse @xmath539 along the straight vertical lines in the @xmath303-plane , see @xcite for a thorough discussion .
such a procedure leads to the concept of a vertical strip .
a vertical strip in the @xmath303-plane associated with @xmath311 is a simply connected domain of the form @xmath540 where @xmath541 are measurable functions in @xmath542 .
moreover , there is a single valued analytic function @xmath543{{}_{\!\!\textnormal{into}\!\!}}}{\mathbb{x}}$ ] which takes every vertical interval @xmath544 onto a complete vertical trajectory in @xmath2 .
this mapping @xmath216 is locally conformal and its inverse , locally defined , is a branch of @xmath545 .
thus the image @xmath546 is an open subset of @xmath2 .
each vertical trajectory in @xmath2 either lies entirely in @xmath547 or otherwise is disjoint from @xmath547 .
the point is that the whole domain @xmath250 can be covered by a countable number of domains such as @xmath547
. denote @xmath548 the family of vertical trajectories in @xmath2 which intersect the set @xmath549 and assume , to derive a contradiction , that the union @xmath550 has positive measure .
we shall confine ourselves to one particular subdomain @xmath546 and trajectories selected from @xmath551 that lie in @xmath547 . with a suitable choice of @xmath547
we ensure that the union of the selected trajectories still has positive measure . rather than discuss this subdomain , let us assume that @xmath552 .
further simplification comes by considering the mapping @xmath553 .
this simplifies not only the domain of definition but also the hopf - laplace equation translates into the somewhat easier form @xmath554 let @xmath555 denote the family of all vertical intervals in @xmath302 ; these are vertical trajectories of @xmath556 , @xmath557 among them there are intervals that pass through the set @xmath558 which we designate by @xmath559 in this way we are reduced , equivalently , to showing that the union @xmath560 has positive measure . that this is indeed an equivalent problem follows from the observation that @xmath216 , being a local diffeomorphism , takes a null family of vertical intervals in @xmath302 into a null family of vertical trajectories in @xmath547 . furthermore , the strip @xmath302 , possible infinite , can be exhausted with an increasing sequence of vertical strips compactly contained in @xmath302 , @xmath561 let the family @xmath562 consist of those vertical intervals in @xmath563 which pass through the set @xmath558 . clearly , we have @xmath564 ; the latter is a subset of @xmath302 with positive measure .
thus for some large @xmath284 we still have @xmath565 therefore , we may and do assume , instead of introducing new notation that @xmath302 is bounded and @xmath543{{}_{\!\!\textnormal{onto}\!\!}}}\psi ( \mathcal s)$ ] extends as a local conformal mapping to a neighborhood of @xmath566 . in particular , @xmath567 and the multiplicity function of @xmath568 is also bounded . for almost every @xmath569 such that @xmath570 , we have @xmath571 let @xmath572 denote the set of parameters @xmath573 such that @xmath574 and @xmath570 .
this means that @xmath202 is a constant mapping on each interval @xmath575 , for @xmath576 . since @xmath577
we conclude that @xmath578 on @xmath579 . on the other hand , in view of the hopf - laplace equation , @xmath202 can not vanish on a set of positive measure , so @xmath580 .
hence @xmath581 has zero linear measure , as claimed .
we now choose and denote by @xmath582 a set of positive linear measure such that @xmath583 where @xmath584 is a sufficiently small positive number .
consider a sequence of concentric annuli centered at @xmath585 , @xmath586 it follows from that @xmath587 , with @xmath588 , is a connected set which joins @xmath585 with a point outside the outer boundary of @xmath589 .
elementary geometric arguments give an estimate of @xmath476-dimensional hausdorff measure of the set @xmath590 , namely @xmath591 for almost every @xmath588 , the function @xmath202 is absolutely continuous on @xmath575 , because @xmath592 .
consider a subset @xmath593 of the interval @xmath594 .
we have @xmath595 where we used the inclusion @xmath596 . integrating with respect to @xmath588 , by fubini s theorem
, we obtain @xmath597 next we apply hlder s inequality @xmath598 it is at this point that we shall appeal to the hopf - laplace equation and formula , which gives us a pointwise inequality in terms of the jacobian determinant of @xmath202 , @xmath599 now recall that the multiplicity function of @xmath20 is essentially bounded and @xmath568 has finite multiplicity .
therefore , the function @xmath600 is essentially bounded as well , say @xmath601 for almost every @xmath525 .
we have @xmath602 where the second inequality follows from ( * ? ? ?
* theorem 6.3.2 ) . substituting into yields @xmath603 .
finally we add these inequalities for @xmath604 to obtain @xmath605 where @xmath109 can be any positive number we wish .
thus @xmath606 , completing the proof of proposition [ lily ] .
[ c1lily ] under the assumptions of proposition [ lily ] , suppose @xmath607 and @xmath608 does not contain any critical points of @xmath609 .
then @xmath608 is a closed vertical arc . by lemma 3.7 in
@xcite the set @xmath610 is connected and compact .
then the union of vertical trajectories which intersect @xmath611 is connected and , by proposition [ lily ] , has zero measure .
this is possible only when @xmath611 is contained in exactly one vertical trajectory .
the outline of the proof is as follows .
we may assume that the hopf differential of @xmath20 does not vanish identically , for otherwise @xmath20 is holomorphic in @xmath2 .
using the notation of theorem [ thharm ] let @xmath180 be an open disk compactly contained in @xmath3 and @xmath612 .
we will prove that @xmath20 is harmonic in @xmath613 by showing that the energy of @xmath20 does not exceed the energy of @xmath359 , @xmath614 \le { \mathcal{e}}_{{\mathbb{u } } } [ h]\ ] ] where @xmath359 is the poisson refinement of @xmath20 in @xmath613 .
indeed , shows that @xmath615 in @xmath613 and therefore @xmath20 is harmonic diffeomorphism of @xmath613 onto @xmath180 .
since @xmath20 is also monotone , it is a global diffeomorphism .
thus is all we need to prove theorem [ thharm ] .
the first step toward proving is the following computation .
[ energylemma ] let @xmath613 and @xmath180 be bounded simply connected domains .
suppose that @xmath616{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] , of sobolev class @xmath617 , is monotone and proper and has a continuous extension to @xmath420 .
furthermore , let @xmath618{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] be a @xmath17-diffeomorphism of sobolev class @xmath619 which extends continuously to @xmath420 with @xmath620 for @xmath621 .
then for @xmath622{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] and @xmath623 we have @xmath624- { \mathcal{e}}_{\mathbb{u}}[h ] & \ge \frac{4}{{\lvert\varphi\rvert}_{\mathscr l^1({\mathbb{u}})}}\ , \left[\iint_{\mathbb{u}}{\big\lvert\chi_z- \frac{\varphi}{{\lvert\varphi\rvert } } \chi_{\bar z}\big\rvert } \sqrt{{\lvert\varphi(z)\rvert}}\sqrt{{\lvert\varphi \big(\chi ( z ) \big)\rvert } } \ , dz \right]^2 \\ & - 4 \iint_{{\mathbb{u } } } { \lvert\varphi\rvert}. \end{split}\ ] ] here
we assume that @xmath28 is continuous , @xmath625 , and the term @xmath626 is understood as equal to zero whenever @xmath28 vanishes .
first assume , in addition to the above hypotheses that @xmath616{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] is a diffeomorphism . the chain rule can be applied to the composition @xmath627{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{u}}$ ] @xmath628 where @xmath629 .
the partial derivatives of @xmath630 at @xmath303 can be expressed in terms @xmath631 and @xmath632 at @xmath633 by the rules @xmath634 where the jacobian determinant @xmath635 is strictly positive .
this yields @xmath636 let @xmath637 be a compactly contained subdomain of @xmath613 .
we compute the energy of @xmath359 over the set @xmath638 by substitution @xmath639 , @xmath640 \ge\mathcal e_{\chi({\mathbb{u}}')}[h]&=2\iint_{\chi({\mathbb{u } } ' ) } \left({\lverth_w\rvert}^2+{\lverth_{\bar w}\rvert}^2\right)\,{\textnormal d}w \\ & = 2\iint_{{\mathbb{u}}'}\frac{{\lverth_z \overline{\chi_z } - h_{\bar z } \overline{\chi_{\bar z}}\rvert}^2 + { \lverth_{\bar z } \chi_z - h_{z } \chi_{\bar z}\rvert}^2}{{\lvert\chi_z\rvert}^2-{\lvert\chi_{\bar z}\rvert}^2}\,{\textnormal d}z . \end{split}\ ] ] on the other hand , the energy of @xmath20 over the set @xmath641 is @xmath642=2\iint_{{\mathbb{u}}'}\left({\lverth_z\rvert}^2+{\lverth_{\bar z}\rvert}^2\right)\,{\textnormal d}z.\ ] ] subtract these two integral expressions to obtain @xmath643-\mathcal e_{{\mathbb{u}}'}[h ] & \ge 4\iint_{{\mathbb{u } } ' } \frac{\left({\lverth_z\rvert}^2+{\lverth_{\bar z}\rvert}^2\right)\cdot { \lvert\chi_{\bar z}\rvert}^2 -2\operatorname{re}\left[h_z\overline{h_{\bar z } } \overline{\chi_z}\chi_{\bar z}\right]}{{\lvert\chi_z\rvert}^2-{\lvert\chi_{\bar z}\rvert}^2}\,{\textnormal d}z \\ & \ge 4 \iint_{{\mathbb{u } } ' } \frac { 2{\lverth_zh_{\bar z}\rvert } \cdot { \lvert\chi_{\bar z}\rvert}^2 -2\operatorname{re}\left[h_z\overline{h_{\bar z } } \overline{\chi_z}\chi_{\bar z}\right]}{{\lvert\chi_z\rvert}^2-{\lvert\chi_{\bar z}\rvert}^2}\,{\textnormal d}z \\ & = 4 \iint_{{\mathbb{u } } ' } \left[\frac{{\lvert\chi_z-\sigma(z ) \chi_{\bar z}\rvert}^2}{{\lvert\chi_z\rvert}^2-{\lvert\chi_{\bar z}\rvert}^2 } -1 \right ] \ , { \lverth_zh_{\bar z}\rvert } \,{\textnormal d}z \end{split}\ ] ] where @xmath644 using hlder s inequality we continue the above chain of estimates as follows @xmath645 ^ 2}{\iint_{{\mathbb{u } } ' } j(z,\chi ) { \lvert\psi \big ( \chi(z ) \big)\rvert } \,{\textnormal d}z } -4\iint_{{\mathbb{u } } ' } { \lverth_zh_{\bar z}\rvert}.\ ] ] where @xmath646 can be any continuous function , provided @xmath647 on @xmath638
. the denominator in is uniformly bounded from above @xmath648 hence @xmath649-\mathcal e_{{\mathbb{u}}'}[h]\ge 4\ , \frac{\left[\iint_{{\mathbb{u}}'}\left| \chi_z-\sigma\chi_{\bar z}\right| \,\sqrt{{\lvert\varphi ( z ) \rvert } } \sqrt{{\lvert\psi\big(\chi ( z ) \big)\rvert } } \,{\textnormal d}z\right]^2 } { \iint_{{\mathbb{u } } } { \lvert\psi ( z)\rvert } \,{\textnormal d}z } -4\iint_{{\mathbb{u}}'}{\lvert\varphi\rvert}. \ ] ] this inequality can now be generalized by an approximation argument . by theorem [ thm : diffeoapprox ] , we have a sequence of diffeomorphisms @xmath650{{}_{\!\!\textnormal{onto}\!\!}}}\mathbb d$ ] , converging to @xmath20 uniformly and strongly in @xmath617 .
moreover , each @xmath651 , so @xmath652 extends continuously to @xmath420 with @xmath653 on @xmath654 .
we may and do assume , by passing to a subsequence if necessary , that @xmath655 and @xmath656 converge almost everywhere to @xmath32 and @xmath33 , respectively .
let @xmath657 .
since the sequence @xmath658{{}_{\!\!\textnormal{onto}\!\!}}}{\mathbb{u}}$ ] of self - diffeomorphisms of @xmath613 is converging to @xmath659 uniformly and strongly in @xmath51 on subdomains @xmath637 , it follows that @xmath649-\mathcal e_{{\mathbb{u}}'}[h^j]\ge 4\frac{\left[\iint_{{\mathbb{u}}'}\left| \chi^j_z-\sigma^j \chi^j_{\bar z}\right| \,\sqrt{{\lvert\varphi^j ( z ) \rvert } } \sqrt{{\lvert\psi\big(\chi ( z ) \big)\rvert } } \,{\textnormal d}z\right]^2 } { \iint_{{\mathbb{u } } } { \lvert\psi ( z)\rvert } \,{\textnormal d}z } -4\iint_{{\mathbb{u}}'}{\lvert\varphi^j\rvert}.\ ] ] passing to the limit as @xmath660 yields @xmath661-\mathcal e_{{\mathbb{u}}'}[h]&\ge 4\frac{\left[\iint_{{\mathbb{u}}'}\left| \chi_z-\sigma\chi_{\bar z}\right| \,\sqrt{{\lvert\varphi ( z ) \rvert } } \sqrt{{\lvert\psi\big(\chi ( z ) \big)\rvert } } \,{\textnormal d}z\right]^2 } { \iint_{{\mathbb{u } } } { \lvert\psi ( z)\rvert } \,{\textnormal d}z}\\ & -4\iint_{{\mathbb{u}}'}{\lvert\varphi\rvert}. \end{split}\ ] ] to see this we simply note that @xmath662 while @xmath663 everywhere .
here we recall that @xmath28 is assumed to be continuous , so we can take @xmath664 in . finally ,
since @xmath641 was an arbitrary compact subset of @xmath613 , we conclude from with the desired estimate , completing the proof of lemma [ energylemma ] . for
the proof of it remains to show that the right hand side of is nonnegative .
this requires a careful analysis of the boundary behavior of @xmath659 , as this mapping is not necessarily continuous up to @xmath665 .
we need a definition and a lemma .
[ needit ] let @xmath219 be a domain and @xmath179 be a simply connected domain compactly contained in @xmath2 .
let @xmath233 be a holomorphic function such that @xmath625 .
we say that a mapping @xmath666 is _ compatible with @xmath28 _ if the following holds for any vertical arc @xmath244 of @xmath667 that intersects @xmath179 and has endpoints in @xmath177 .
let @xmath264 be a maximal subarc of @xmath244 contained in @xmath179 , and denote its endpoints by @xmath237 and @xmath668 .
the connected components of @xmath669 are naturally denoted as @xmath670 and @xmath671 .
the condition we impose on @xmath659 is @xmath672 where @xmath673 and @xmath674 are cluster sets @xcite .
[ compatible ] the mapping @xmath659 in lemma [ energylemma ] is compatible with @xmath28 provided that @xmath675 contains no zeros of @xmath28 .
we use the notation of definition [ needit ] . according to corollary [ c1lily ] , the sets @xmath676 and @xmath677 are vertical arcs , hence subarcs of @xmath244 .
the continuity of @xmath20 and @xmath359 implies that @xmath673 is a subset of @xmath676 , and similarly for @xmath674 .
the claim follows .
the restriction concerning zeros of @xmath28 in lemma [ compatible ] is easily fulfilled by choosing a generic radius for the disk @xmath180 .
we are finally ready to handle the expression in , thus completing the proof of theorem [ thharm ] .
the following result is related to the reich - strebel inequality @xcite , see also @xcite for a recent extension . however ,
in our lemma [ reich - strebel ] the assumptions on @xmath659 are different from those in @xcite .
[ reich - strebel ] let @xmath219 be a domain and @xmath179 a simply connected domain compactly contained in @xmath2 .
let @xmath233 be a holomorphic function such that @xmath678 .
suppose that @xmath679 is continuous proper and compatible with @xmath28 .
then @xmath680 for almost every vertical noncritical trajectory @xmath244 the mapping @xmath659 is locally absolutely continuous on @xmath244 .
let @xmath264 be a maximal subarc of @xmath244 in @xmath179 .
denote by @xmath237 and @xmath668 the endpoints of @xmath264 . by the compatibility condition the curve @xmath681 connects two different components of @xmath269 .
by lemma [ lengthineq ] we have @xmath682 because @xmath683 is the magnitude of directional derivative of @xmath659 along @xmath244 . in view of corollary [ fubcor ] inequality follows .
combining with yields , completing the proof of theorem [ thharm ] .
in this section @xmath2 and @xmath3 are bounded @xmath109-connected domains .
suppose @xmath684 is a hopf deformation , that is , @xmath685 is a holomorphic function in @xmath2 .
we shall actually prove the following explicit bound @xmath686 let @xmath687 and note the pointwise inequality @xmath688 which is a consequence of the subharmonicity of @xmath689 in @xmath2 .
since the jacobian @xmath690 is nonnegative , it follows from that @xmath33 is locally bounded , specifically @xmath691 the boundedness of @xmath33 implies that @xmath345 is locally in @xmath692 , and consequently @xmath693 for every @xmath694 .
a similar argument was carried out in @xcite in a somewhat different context .
however , it does not yield the lipschitz continuity of @xmath20 , which we will prove by an entirely different method .
the proof of theorem [ mainth ] is preceded by several lemmas .
we denote the average value of a function by an integral sign with a dash . the normal and tangential derivatives of @xmath20 are defined as @xmath695 [ avelip ] suppose that @xmath696 and @xmath697 .
let @xmath698 .
then the circular mean @xmath699 is a locally lipschitz function of @xmath584 .
specifically we have @xmath700 as a consequence , @xmath701 note that @xmath702 is an absolutely continuous function of @xmath584 , even more @xmath703 .
therefore , we can differentiate with respect to @xmath584 for a.e .
@xmath704 to obtain @xmath705 combining and the identity @xmath706 yields @xmath707 this together with implies and the integration yields .
[ equi ] suppose @xmath696 . for almost every @xmath584 , @xmath708 , we have @xmath709 using the identities we find that @xmath710 and @xmath711 integration of over @xmath712 shows that @xmath713 from and we obtain . [ connected ] suppose @xmath714 and @xmath715 . if the set @xmath611 intersects some domain @xmath716 , then it also intersects @xmath259 .
let @xmath716 be a domain such that @xmath717 contains a point @xmath237 .
consider a @xmath94-convergent sequence of homeomorphisms @xmath718 .
for each @xmath719 @xmath720 since the convergence is uniform on compact sets , letting @xmath721 , we conclude @xmath722 next we require a fourier series lemma that can be viewed as a bonnesen - type inequality @xcite . [ fs ] let @xmath723 be a function in @xmath724 . expand it into the fourier series @xmath725 . if @xmath726 then @xmath727 indeed , inequality is apart from the better factor @xmath728 , a form of isoperimetric inequality in the plane .
we gain this better factor because of the assumption which can be interpreted as saying that the image of @xmath202 is far from being a circle centered at @xmath729 .
we may assume that @xmath202 is nonconstant .
normalize @xmath202 so that @xmath730 and @xmath731 . clearly , @xmath732 we need a lower bound for @xmath733 as well . to this end , consider @xmath734 and observe that @xmath735 by virtue of . on the other hand , @xmath736 which together with yield @xmath737 using and we arrive at as follows .
@xmath738 let the `` good '' part of @xmath2 be @xmath739 and the `` bad '' part be @xmath740 . by theorem [ thharm ]
the restriction of @xmath20 to @xmath296 is a harmonic diffeomorphism onto @xmath3 .
thus the case @xmath741 is trivial . from now on we assume that @xmath276 is nonempty .
this is only possible if @xmath625 . [ mpro ]
suppose that @xmath696 and @xmath742 .
then for a.e .
@xmath708 , we have @xmath743 fix @xmath584 such that the restriction of @xmath20 to @xmath712 is in @xmath744 .
let @xmath745 .
the change of variables formula implies @xmath746 thus we may assume @xmath747 ; otherwise is immediate from . by lemma
[ connected ] the mapping @xmath20 assumes the value @xmath585 on @xmath712 , hence @xmath748 thus , lemma [ fs ] applies to the restriction of @xmath20 onto @xmath712 .
the estimate reads as @xmath749 which implies .
we will show that for every point @xmath750 @xmath751 we may assume @xmath752 .
let @xmath698 .
* case 1 . * @xmath753 .
the first step is to rewrite as a differential inequality for the function @xmath754 here we restrict ourselves to @xmath755 which yields @xmath756 for @xmath757 . since @xmath758 ,
applying the inequality we obtain @xmath759 next , the integral on the right is estimated using , and lemma [ equi ] : @xmath760 combining and we obtain @xmath761 for notational simplicity we introduce the constant @xmath762 .
inequality yields @xmath763 integrate over the interval @xmath764 to obtain @xmath765 finally , multiply by @xmath766 and rewrite it as @xmath767 we further simplify using the pointwise inequality @xmath768 .
@xmath769 this yields even with a better constant .
* @xmath770 ; that is , @xmath771 .
let @xmath772 .
since @xmath20 is harmonic in @xmath296 , the subharmonicity of @xmath170 yields @xmath773 if @xmath774 , then already implies .
otherwise pick @xmath775 such that @xmath776 .
clearly @xmath777 and the righthand side of can be estimated by applying to @xmath778 . since @xmath779 , we may apply inequality with @xmath780 and @xmath781 in place of @xmath782 .
this gives the estimate @xmath783 combining , , and gives the inequality .
this estimate yields the pointwise inequality at the lebesgue points of @xmath170 , completing the proof of theorem [ mainth ] .
throughout this section the following standing assumptions are made on the mappings under considerations : @xmath2 and @xmath3 are @xmath109-connected bounded domains and @xmath784 is a @xmath66-smooth hopf deformation , @xmath714 ; that is , @xmath785 where @xmath28 is a holomorphic function .
recall that the _ convex part _ of the boundary of a domain @xmath3 is @xmath786 we designate the domain of regular points of the quadratic differential @xmath42 by @xmath787 additional assumptions on @xmath20 will be explicitly stated when needed .
the proof of theorem [ smooth ] proceeds by a number of claims and lemma [ bes ] . by virtue of we
have a lower bound for the horizontal derivative @xmath790 if @xmath791 had an infinite number of zeros there would be an accumulation point of zeros .
since @xmath20 is @xmath66-smooth along @xmath792 , its horizontal derivative would vanish at the accumulation point of zeros , in contradiction with .
the domain @xmath250 can be expressed as countable union of @xmath28-rectangles compactly contained in @xmath250 , see [ secqd ] for the definition and consideration of @xmath28-rectangles .
let @xmath248 be one such @xmath28-rectangle .
it suffices to show that the set @xmath793 can be covered by a countable number of vertical trajectories .
let @xmath794 and @xmath270 denote the horizontal edges of @xmath248 .
these are horizontal arcs of @xmath311 compactly contained in @xmath250 . in view of claim 1 , we may and do assume that @xmath20 does not vanish on the horizontal edges of @xmath248 , for otherwise we can replace @xmath248 by a finite number of parallel subrectangles of @xmath248 ; simply cut @xmath248 along vertical trajectories crossing the horizontal edges @xmath794 and @xmath270 at the zeros of @xmath20 . with this assumption
the proof of claim 2 will be completed by showing that inside of @xmath248 there are no zeros of @xmath20 . indeed , suppose otherwise .
then , by lemma [ connected ] , the set @xmath549 intersects at least one of two vertical edges of the rectangle @xmath248 . by the same reasoning @xmath549 intersects one of two vertical edges of any subrectangle obtained by continuously compressing @xmath248 with vertical trajectories .
this procedure results in a family of vertical arcs intersecting @xmath549 whose union has positive measure , in contradiction with lemma [ lily ] .
recall that theorem [ smooth ] deals with minimal deformations .
these are special mappings for which the hopf differential @xmath795 is real along @xmath21 , provided @xmath21 is @xmath66-smooth .
it involves no loss of generality in the proof of theorem [ smooth ] to assume that @xmath21 is indeed @xmath66-smooth . for , if necessary , we could transform @xmath2 conformally onto , say , a circular schottky domain .
our next lemma tells us that no point in @xmath2 can be mapped into a convex part @xmath796 of @xmath163 , see and figure [ fig4 ] .
assume , to the contrary , that the set @xmath798 consider the `` good '' subdomain @xmath799 .
it is indeed connected by theorem [ thharm ] .
let @xmath800 be the set of points @xmath801 for which there is a disk @xmath802 such that @xmath803 .
note that @xmath800 is dense in @xmath804 .
our first step is to prove that suppose to the contrary that @xmath807 for some @xmath808 .
since @xmath3 is convex at @xmath303 the function @xmath809 , for some constant @xmath810 , attains a local maximum at @xmath330 , which yields @xmath811 on the other hand , there is a disk @xmath812 such that @xmath813 , @xmath814 and @xmath815 for all @xmath816 .
the hopf boundary point lemma ( ( * ? ? ?
* lemma 3.4 ) , @xcite ) implies that the inner normal derivative of @xmath817 at @xmath813 is strictly negative .
this contradicts .
finally , since @xmath818 is open in @xmath163 and @xmath800 is dense in @xmath804 , the claim follows . the intersection of @xmath290 with @xmath820 is open in @xmath290 . on the other hand the vertical derivative @xmath823 vanishes a.e . on @xmath821 .
indeed , by ( * ? ? ?
* lemma 3.8 ) @xmath824 vanishes a.e . on @xmath821 and by the vertical derivative satisfies @xmath825
therefore , @xmath20 is constant along @xmath290 , say equal to @xmath822 . on the other hand , by claim 2 ,
the preimage @xmath611 is covered by a countable union of vertical trajectories and critical points .
we find that @xmath290 is indeed one of the connected components of @xmath611 .
recall from the introduction , [ sec101 ] , the components @xmath826 of @xmath827 .
suppose otherwise , that both ends of @xmath290 approach the same boundary component , say @xmath828 .
by lemma [ callsomething ] the vertical trajectory @xmath290 separates two other boundary components of @xmath2 , say @xmath829 and @xmath830 .
let @xmath831 be a continuum connecting the components @xmath832 and @xmath833 of @xmath163 , in accordance with the correspondence in .
then @xmath834 is a continuum ( * ? ? ?
* lemma 3.7 ) that connects @xmath829 and @xmath830 in @xmath291 , a contradiction .
now we have the required contradiction that proves lemma [ bes ] .
indeed @xmath290 , being a connected component of @xmath835 , can not approach different components of @xmath21 ; otherwise its image @xmath836 would approach different components of @xmath163 by virtue of .
the function @xmath32 is continuous in @xmath2 and holomorphic in @xmath65 . by a theorem of besicovitch (
* theorem 2 ) sets of @xmath837-finite @xmath476-dimensional hausdorff measure are removable for holomorphic continuous functions . therefore , @xmath32 is holomorphic in @xmath2 .
it follows that @xmath20 is harmonic in @xmath2 , which by proposition [ harmdiff ] implies that @xmath20 is a diffeomorphism .
the reader may notice that the domain @xmath3 in theorem [ smooth ] can not be @xmath66-smooth . also , the butterfly domain @xmath3 of example [ exsqueeze ] has nonsmooth boundary .
these observations lead to the following question .
a. s. besicovitch , _ on sufficient conditions for a function to be analytic , and on behaviour of analytic functions in the neighbourhood of non - isolated singular points _
london math .
( 2 ) * 32 * ( 1931 ) , no . 1 , 19 . y. w. chen , _ discontinuity and representations of minimal surface solutions _ , proceedings of the conference on differential equations ( dedicated to a. weinstein ) , pp . 115138 .
university of maryland , college park , md ( 1956 ) . | the central theme in this paper is the hopf - laplace equation , which represents stationary solutions with respect to the inner variation of the dirichlet integral . among such solutions
are harmonic maps .
nevertheless , minimization of the dirichlet energy among homeomorphisms often leads to nonharmonic solutions .
we investigate the hopf - laplace equation for a certain class of topologically well behaved mappings which are almost homeomorphisms , called _
hopf deformations_. we establish lipschitz continuity of hopf deformations , the best possible regularity one can get . thus in particular we show that the minimal - energy deformations are lipschitz continuous , a result of considerable interest in the theory of minimal surfaces , calculus of variations , and pdes , with potential applications to elastic plates . |
during the past 15 years an increasing interest in two - dimensional ( 2d ) magnets has developed .
this is a result of ( i ) the investigation of a wide class of well - characterized quasi-2d magnetic materials which allow a detailed experimental study of their properties ( inelastic neutron scattering , nuclear magnetic resonance , etc . ) , and ( ii ) the availability of high - speed computers with the capabilities for simulations on large lattices .
examples of these materials are ( 1 ) layered magnets @xcite , like @xmath1 , @xmath2 , @xmath3 and @xmath4 with @xmath5 and @xmath6 ; ( 2 ) @xmath7 graphite intercalation compounds @xcite , ( 3 ) magnetic lipid layers @xcite , like @xmath8 . for the first class of these examples
the ratio of inter- to intraplane magnetic coupling constants is typically @xmath9 to @xmath10 .
this means that the behavior is nearly two - dimensional as concerns the magnetic properties . for the second class ,
the above ratio of the coupling constants can be tuned by choosing the number of intercalated graphite layers . for the third class ,
even monolayers can be produced , using the langmuir - blodgett method .
many of the above materials have an easy - plane or @xmath0-symmetry .
this means that the spins prefer to be oriented in the @xmath0-plane which is defined as the plane in which the magnetic ions are situated .
the simplest model for this symmetry is the 2d classical anisotropic heisenberg hamiltonian ( see ref .
@xcite ) @xmath11\,.\ ] ] @xmath12 labels nearest neighbors of a 2d lattice ; usually a square lattice is used .
the subscripts @xmath13 and @xmath14 stand for the components of the classical spin vector @xmath15 ; the spin length @xmath16 is set to unity by the redefinition @xmath17 .
@xmath18 is the magnetic coupling constant , both ferro- and antiferromagnetic materials were investigated .
the anisotropy parameter @xmath19 lies in the range @xmath20 ; note that @xmath21 corresponds to the @xmath0-model , not to the so - called planar model where the spins are strictly confined to the @xmath0-plane ; this confinement is possible only if one is not interested in the dynamics . instead of
the exchange anisotropy in the hamiltonian ( [ h = j ] ) , one can also use an on - site anisotropy term @xmath22 , which yields similar results .
there are two kinds of excitations : spin waves which are solutions of the linearized equations of motion , and vortices which are topological collective structures .
the vortices are responsible for a topological phase transition @xcite at the kosterlitz - thouless transition temperature @xmath23 .
below @xmath24 , vortex - antivortex pairs are thermally excited and destroyed ; above @xmath23 these bound pairs dissociate and the density of free vortices increases with temperature .
it must be noted that an order - disorder phase transition is not possible according to the mermin - wagner - theorem @xcite ; the reason is that all long - range order is destroyed by the long - wave linear excitations in all 1d and 2d models with continuous symmetries .
both theory @xcite and monte carlo simulations @xcite showed that @xmath23 is only very weakly dependent on the anisotropy @xmath19 , except for @xmath19 extremely close to @xmath25 , when @xmath26 .
for the materials mentioned above , coupling constants were estimated from fits to spin - wave theory , and @xmath19 values are in the range @xmath27 , where @xmath23 is still close to its value for @xmath21 .
however , the out - of - plane structure of the vortices ( i. e. , the structure of the @xmath28 components ) depends crucially on @xmath19 , while the in - plane structure ( @xmath29 and @xmath30 components ) remains the same @xcite . for @xmath31 for a square lattice @xcite ) ,
static vortices have null @xmath28 components , they are termed in - plane or planar vortices in the literature . for @xmath32
there are out - of - plane or non - planar vortices which exhibit a localized structure of the @xmath28 components around the vortex center .
this structure falls off exponentially with a characteristic length , the vortex core radius @xcite @xmath33 in units of the lattice constant @xmath34 . the core radius increases with diminishing @xmath19 , allowing a continuous crossover to the isotropic heisenberg model ( @xmath35 ) where the topological excitations are merons and instantons @xcite , rather than vortices .
compared to the vortices in classical fluids @xcite and superfluids @xcite , there is an important difference : the vortices of the easy - plane heisenberg model carry _ two _ topological charges , instead of one : ( 1 ) the vorticity @xmath36 which is defined in the usual way : the sum of changes of the azimuthal angle @xmath37 on an arbitrary closed contour around the vortex center yields @xmath38 ; if the center is not inside the contour the sum is zero . in the following only the cases @xmath39 ( vortex ) and @xmath40 ( antivortex ) will be considered , the cases @xmath41 would be relevant only for high temperatures .
( 2 ) the polarization or polarity @xmath42 : for the non - planar vortices @xmath43 indicates to which side of the @xmath0-plane the out - of - plane structure points , while @xmath44 for the planar vortices ( fig .
[ vortexstruct ] ) . both topological charges and @xmath45
are @xmath46- and @xmath47-topological charges , respectively ; see ref .
@xcite . ] are crucial for the vortex dynamics , which is governed by two different forces : ( 1 ) 2d coulomb - type forces @xmath48 , which are proportional to the product of the vorticities of two vortices and inversely proportional to their distance ( assuming that the distance is larger than @xmath49 , such that the out - of - plane structures do not overlap ) , ( 2 ) a gyrocoupling force @xcite @xmath50 where @xmath51 is the vortex velocity , and @xmath52 is the trajectory of the vortex center . the force ( [ fg = x ] ) is formally equivalent to the magnus force in fluid dynamics and to the lorentz force on an electric charge .
the gyrocoupling vector @xcite @xmath53 where @xmath54 is the unit vector in @xmath14-direction , does not represent an external field but is an intrinsic quantity , namely a kind of self - induced magnetic field which is produced by the localized @xmath28-structure and carried along by the vortex . for antiferromagnets
@xmath55 vanishes .
since the gyrovector ( [ g=2 ] ) contains the product of both topological charges , the dynamics of the two kinds of vortices is completely different : + ( 1 ) for planar vortices @xmath56 , therefore they have a newtonian dynamics @xmath57 where the vortex mass @xmath58 will be defined in section 2.2 .
however , in the simulations the trajectories are not smooth due to strong discreteness effects @xcite .
+ ( 2 ) non - planar vortices have smooth trajectories if the diameter @xmath59 of the out - of - plane structure is considerably larger than the lattice constant @xcite ; i. e. , if @xmath19 is not close to @xmath60 . for steady state motion ,
the dynamics is described by the thiele equation @xcite @xmath61 which was derived from the landau - lifshitz equation ( section 2 ) for the spin vector @xmath62 .
this equation is identical to hamilton s equations with hamiltonian ( 1.1 ) .
however , for arbitrary motion the thiele equation ( 1.5 ) is only an approximation because a rigid vortex shape was assumed in the derivation .
if a velocity dependence of the shape is allowed @xcite , an inertial term @xmath63 appears on the l. h. s. of eq .
( [ xxg ] ) .
however , this @xmath64-order equation could not be confirmed by computer simulations @xcite .
the reason will be discussed in section 2 ; in fact , a collective variable theory reveals that the dynamics of non - planar vortices can only be described by odd - order equations of motion .
therefore the dynamics is _ non - newtonian_. another interesting topic is the dynamics under the influence of thermal fluctuations ( section 3 ) . here
white noise and damping are implemented in the landau - lifshitz equation .
the same type of collective variable theory as in section 2 then yields stochastic equations of motion for the vortices .
the stochastic forces on the vortices can be calculated and the solutions of the equations of motion are compared with langevin dynamics simulations . in this way , the diffusive vortex motion can be well understood . for somewhat higher temperatures ( still below @xmath23 ) the thermal noise can induce transitions of non - planar vortices from one polarization to the opposite one .
theoretical estimates of the transition rate are compared with langevin dynamics simulations ( section [ polartrans ] ) .
for the temperature range above @xmath23 , so far only a phenomenological theory exists , the vortex - gas approach ( section [ tempgtkt ] ) .
it contains only two free parameters ( the density of free vortices and their r.m.s .
velocity ) .
the theory predicts central peaks in the dynamic form factors for both in - plane and out - of - plane spin correlations .
such peaks are observed both in computer simulations and in inelastic neutron scattering experiments , and many of the predicted features are confirmed .
however , a recent _ direct _ observation of the vortex motion in monte carlo simulations has revealed that a basic assumption of the theory , namely a _
ballistic _ vortex motion , might not be valid .
( in section [ tempgtkt ] we confine ourselves to the ferromagnetic case , although easy - plane antiferromagnets were also investigated ) .
the spin dynamics is given by the landau - lifshitz equation , which is the classical limit of the heisenberg equation of motion for the spin operator @xmath65 , @xmath66 where @xmath67 is the hamiltonian , in our case that of the anisotropic heisenberg model ( [ h = j ] ) .
the meaning of eq .
( [ ds ] ) is that the spin vector @xmath65 precesses in a local magnetic field @xmath68 , with cartesian components @xmath69 . in spin dynamics simulations ( see below ) , the landau - lifshitz equation is integrated numerically for a large square lattice , typically with @xmath70 lattice points .
as initial condition a vortex is placed on the lattice and the trajectory @xmath71 of the vortex center is monitored .
this trajectory is then compared with theory , i. e. with the solution of an equation of motion for the vortex .
the standard procedure to obtain this equation consists in taking the continuum limit @xmath72 where @xmath73 is a vector in the @xmath0 plane , and in developing a _ collective variable theory_. the simplest version makes the travelling wave ansatz @xcite @xmath74 where the functions @xmath75 on the r. h. s. describe the vortex shape .
( strictly speaking , these functions should bear an index to distinguish them from the functions on the l. h. s. ) as the equation of motion is expected to contain a force , the following operations are performed with eq .
( [ ds ] ) which yield force densities @xmath76 } \right)}= -s^{2}\frac{\delta h}{\delta { \bf s}}{\frac{\partial { \bf s}}{\partial x_i}}= -s^{2}{\frac{\partial { \cal h}}{\partial x_i}}\ ] ] with @xmath77 and hamiltonian density @xmath78 . according to the ansatz ( [ srt ] ) ,
@xmath79 is inserted , with summation over repeated indices .
integration over @xmath73 then yields the equation of motion @xmath80 with the force @xmath81 and the gyrocoupling tensor @xmath82 the expression on the right was obtained by introducing the two canonical fields @xmath83 instead of the three spin components @xmath75 with the constraint @xmath84 .
the static vortex structure is @xcite @xmath85}\quad\mbox{for $ r'\ll r_v$ } \\ \label{psi = r_gg_rv } \psi_0 & = & p a_2\sqrt{\frac{r_v}{r ' } } { \rm e}^{-r'/r_v}\quad\mbox{for $ r'\gg r_v$ } \end{aligned}\ ] ] where the constants @xmath86 and @xmath87 can be used for a matching , and @xmath88 the integral then yields @xmath89 where @xmath90 is the antisymmetric tensor .
interestingly , only the value @xmath42 of the @xmath28 component at the vortex _ center _ enters the final result @xcite , i. e. , the out - of - plane vortex structure needs not to be explicitly known here .
as @xmath55 is antisymmetric , eq . ( [ gx = f ] ) is identical to the thiele equation ( [ xxg ] ) with gyrovector @xmath91 .
since @xmath44 and thus @xmath92 for planar vortices , the thiele equation is incomplete in this case . obviously there must be a non - vanishing term on the l. h. s. , this will be an inertial term ( section [ vortexmass ] ) .
moreover , the ansatz ( [ srt ] ) , and thus the thiele equation , is only valid for steady - state motion when the vortex shape is rigid ( in the moving frame ) .
this includes not only translational motion with constant velocity @xmath93 but also rotational motion with constant angular velocity @xmath94 .
both types of motion can be obtained by considering two non - planar vortices at a distance @xmath95 which drive each other by their coulomb force @xmath96 . for a certain velocity
this force is compensated by the gyrocoupling force ( [ fg = x ] ) .
in fact , since each vortex carries two charges , there are four physically different scenarios which represent stationary solutions of the thiele equation .
they fall into two classes : if the gyrovectors of the two vortices are parallel ( i. e. , @xmath97 ) , a vortex - vortex or vortex - antivortex pair rotates with @xmath94 on a circle with radius @xmath98 around each other , where @xmath99 if the gyrovectors are antiparallel , the pair performs a parallel translational motion with velocity @xmath93 and distance @xmath95 , where @xmath100
. both kinds of scenarios also appear for vortices in ( super)fluids @xcite .
however , in these contexts there are only two physically different situations : vortex - vortex rotation and vortex - antivortex translation .
the above scenarios for magnetic vortices were not tested by computer simulations until 1994 , and the results were very surprising @xcite : using a square or circular system with free boundary conditions , two of the four scenarios ( vortex - vortex rotation and vortex - antivortex translation ) were confirmed , but not the two other ones : for vortex - antivortex rotation and vortex - vortex translation the observed trajectories showed pronounced oscillations around the trajectories predicted by the thiele equation ( [ xxg ] ) or ( [ gx = f ] ) , see fig . [ trajectory ] .
in fact , such oscillations had already been predicted @xcite by assuming that the vortex shape in the travelling wave ansatz ( [ srt ] ) depends explicitely on the velocity @xmath101 .
this leads to an additional term @xmath102 in eq .
( [ ds / dt ] ) and thus to an inertial term in the thiele equation @xmath103 where @xmath104 is the mass tensor @xmath105 this integral can be easily evaluated if the vortex is placed at the center of a circular system with radius @xmath106 and free boundary conditions .
the dominant contribution stems from the outer region @xmath107 , with @xmath108 , where the velocity dependent parts of the vortex structure are @xcite @xmath109 @xmath110 together with the static parts @xmath111 from eq .
( [ phi = q ] ) and @xmath112 , which falls off exponentially outside the core ( [ rv=1/2 ] ) , the rest mass @xmath58 is obtained @xmath113 the constant @xmath114 stems from the inner region @xmath115 , where the vortex structure is not well known due to the discreteness , but @xmath114 is not important for discussion .
there is also a velocity dependent contribution to the mass which is negligible because the vortex velocities in the simulations are always much smaller than the spin wave velocity , which is the only characteristic velocity of the system .
the above vortex mass is consistent with other results in the literature : generalizing the momentum of solitons in 1d magnets @xcite , the vortex momentum @xmath116 is defined and can be shown to be a generator of translations @xcite
. then @xmath117 results , but @xmath118 is not a canonical momentum because the poisson bracket @xmath119 does not vanish . for the kinetic energy @xmath120
is obtained @xcite , therefore this energy and the rest energy @xmath121 both show the same logarithmic dependence on the system size @xmath106 . for a continuum model of a 2d ferromagnet with uniaxial symmetry and a magnetostatic field ,
a slightly different vortex momentum was defined @xcite , namely @xmath122 where @xmath123 represents the gyrovector density , cf .
( [ eq : gij ] ) which is related to the gyrovector as described below eq .
( [ gij2 ] ) .
the definition ( [ ppt ] ) depends on the choice of origin of the system and is not a generator of translations .
nevertheless , if the time derivative of @xmath124 is set equal to the force @xmath48 on the vortex , the generalized thiele equation ( [ mx = gx ] ) is again obtained @xcite . therefore the vortex dynamics is qualitatively the same , as will be discussed now .
( [ mx = gx ] ) has the same form as that for an electric charge @xmath125 in a plane with a perpendicular magnetic field @xmath126 and an in - plane electric field @xmath127 .
i. e. , the vortex motion is completely analogous to the hall effect : the trajectory is a cycloid with frequency @xmath128 cf .
the cyclotron frequency @xmath129 , where @xmath130 is the speed of light .
it is possible to transform to a frame where the force @xmath131 vanishes and the vortex rotates ( i. e. , guiding center coordinates @xcite ) .
the cycloidal trajectories can explain _ qualitatively _ the oscillations observed in the simulations ( fig .
[ trajectory ] ) . moreover
, the fact that oscillations are observed only for two of the four scenarios for two vortices ( see above ) can also be explained : a generalization of the ansatz ( [ srt ] ) to the case of two vortices yields two coupled equations of motion with a 2-vortex mass tensor @xcite . in the case of vortex - vortex rotation and vortex - antivortex translation
there are cancellations in this mass tensor which prevent oscillations , in agreement with the simulations . however , a _ quantitative _ comparison of eq .
( [ mx = gx ] ) and the simulations reveals two severe discrepancies @xcite : + ( 1 ) the mass @xmath132 which is obtained from eq .
( [ om = g ] ) by inserting the observed frequency , turns out to be much larger than predicted by eq .
( [ eq : m ] ) . moreover , the dependence on the system size @xmath106 is linear , while eq .
( [ eq : m ] ) predicts a logarithmic dependence . + ( 2 ) instead of the _ one _ frequency @xmath133 of eq .
( [ om = g ] ) , _ two _ frequencies @xmath134 are observed in the spectrum of the oscillations ( fig .
[ fourierspec ] ) . as @xmath135 and @xmath136
are close to each other , a pronounced beat is observed in the trajectories ( fig .
[ trajectory ] ) : the cycloidal frequency is @xmath137 , but the shape of the cycloids changes slowly with @xmath138 .
both discrepancies have nothing to do with 2-vortex effects , because they also occur in 1-vortex simulations @xcite . here
the vortex is driven by image forces , in analogy to electrostatics .
the most convenient geometry is a circular system with radius @xmath106 and free boundaries . in this case
there is only one image vortex which has opposite vorticity , but the same polarization @xcite . the radial coordinate of the image is @xmath139 , where @xmath140 is the vortex coordinate .
the vortex trajectory can be fitted very well to a superposition of two cycloids .
this observation will be explained in the next section .
recently a very general collective variable theory was developed for nonlinear coherent excitations in classical systems with _
hamiltonians @xcite .
in this theory the dynamics of a single excitation is governed by a hierarchy of equations of motion for the excitation center @xmath52 .
the _ type _ of the excitation determines on which levels the hierarchy can be truncated consistently :
gyrotropic
excitations are governed by odd - order equations and thus do not have newtonian dynamics , e. g. galileo s law is not valid . non - gyrotropic excitations are so - to - speak the normal case , because they are described by even - order equations , i. e. by newton s equation in the first approximation
. examples of the latter class are kinks in 1d models like the nonlinear klein - gordon family and the planar vortices of the 2d anisotropic heisenberg model ( [ h = j ] ) .
the non - planar vortices of this model represent the simplest gyrotropic example .
3d models have not been considered so far .
the basis fo the above collective variable theory is a _ generalized travelling wave ansatz _ for the canonical fields in the hamiltonian . for a spin system , as considered in this review , this ansatz reads @xmath141 this generalization of the standard travelling wave ansatz ( [ srt ] ) yields an @xmath142-order equation of motion for @xmath143 , because eq . ( [ ds / dt ] )
is replaced by @xmath144 the same procedure as described below eq .
( [ ds / dt ] ) then yields the @xmath142-order equation .
the case @xmath145 corresponds to the @xmath64-order equation ( [ mx = gx ] ) . in one dimension
the antisymmetric tensor @xmath55 from eq .
( [ eq : gij ] ) vanishes and ( [ mx = gx ] ) reduces to a newtonian equation .
the same is true for the 2d planar vortices where @xmath146 because @xmath44 in eq .
( [ gij2 ] ) . + the case @xmath147
yields the @xmath148-order equation @xmath149[c]{.}\makebox[0.45ex][c]{.}\makebox[0.45ex][c]{. } $ } } } { x}}}$ } + { \bf m } { \bf \ddot{x } } + { \bf g } { \bf \dot{x } } = { \bf f}({\bf x})\ ] ] with the @xmath148-order gyrotensor @xmath150 for the above 1d models and the 2d planar vortices nothing changes because @xmath151 turns out to be zero ( below ) .
but the 2d non - planar vortices are the first gyrotropic example . for the calculation of the integral ( [ eq : aij ] ) the acceleration dependence of the outer region of the vortex
is needed @xcite @xmath152 @xmath153 together with the static parts @xmath111 and @xmath112 this yields @xmath154 where the constant @xmath155 stems from the inner region . in the simulations
the dynamic in - plane structures @xmath156 and @xmath157 can not be clearly observed because it is difficult to subtract the static structure @xmath111 which varies drastically with the vortex position .
however , the dynamic out - of - plane structures @xmath158 and @xmath159 can be observed and distinguished by looking at specific points of the trajectory : e. g. , at the turning points in fig .
[ trajdetail ] the acceleration is maximal while the velocity is small .
[ psistruct1 ] , [ psistruct2 ] confirm the structure of @xmath159 in the outer region , they also show that this structure oscillates as a whole , with frequency @xmath160 .
an alternative interpretation of this oscillating structure in terms of certain quasilocal spinwaves will be given in section 2.4 .
as @xmath159 and @xmath157 depend linearly on @xmath161 and as @xmath151 contains derivatives with respect to @xmath161 , the leading contribution to @xmath151 is independent of velocity and acceleration .
there are higher - order terms which are negligible , though ; cf .
the discussion of the mass . for these reasons
the l. h. s. of the equation of motion ( [ axmx ] ) is linear .
the radial coulomb force @xmath48 on the r. h. s. can be linerarized by expanding in the small radial displacement @xmath162 from the mean trajectory which is a circle of radius @xmath98 @xmath163 the @xmath148-order equation of motion ( [ axmx ] ) then has the following solutions @xcite : a stationary solution and a superposition of two cycloids @xmath164 where @xmath165 is the azimuthal displacement .
the results for @xmath134 can be considerably simplified for @xmath166 , which is the case in the simulations .
the frequencies @xmath134 form a weakly - split doublet .
the mean frequency depends on @xmath167 , but not on @xmath58 : @xmath168 for large @xmath106 .
contrary to this , the splitting of the doublet @xmath169 is proportional to the mass . as @xmath170 is known from eq .
( [ gij2 ] ) , the last two equations are sufficient to determine @xmath58 and @xmath167 by using the frequencies @xmath171 observed in the simulations . in ref .
@xcite the data for lattice sizes @xmath172 were extrapolated for @xmath173 by using several runs for small @xmath174 , because the theoretical predictions ( [ eq : m ] ) and ( [ eq : a ] ) were made for @xmath175 .
for the anisotropy @xmath176 the data for @xmath167 are well represented by @xmath177 with @xmath178 and @xmath179 .
this agrees rather well with the @xmath180-term in ( [ eq : a ] ) , the constant @xmath155 can not be tested .
however , @xmath181 is nearly independent of @xmath106 , in contrast to the logarithmic dependence in eq .
( [ eq : m ] ) .
this can be connected to the observation that the velocity dependent part @xmath158 seems to approach an @xmath106-independent constant at the boundary , in contrast to the predicted @xmath182-decay in eq .
( [ psi = q ] ) .
a very interesting point is the discussion of the size dependence of the different terms in the equation of motion ( [ axmx ] ) .
as every time derivative of @xmath183 contributes a factor of @xmath184 , the orders of the terms are @xmath185[c]{.}\makebox[0.45ex][c]{.}\makebox[0.45ex][c]{. } $ } } } { x}}_i$ } = { \cal o } ( 1/l ) , m\ddot{x}_i = { \cal o } ( \ln l / l^2)$ ] and @xmath186 .
therefore the strong @xmath148-order term can not be neglected when the weak @xmath64-order term is retained .
this is the reason for the two severe discrepancies between the simulations and the predictions of the @xmath64-order equation ( [ mx = gx ] ) , which were discussed in section 2.2 .
however , the neglection of both the @xmath64- and @xmath148-order terms represents a consistent first approximation , namely , the thiele equation ( [ gx = f ] ) . the next consistent approximation in the hierarchy is the @xmath148-order equation ( [ axmx ] ) , as discussed above .
an investigation of even higher - order terms in ref .
@xcite confirms the conjecture that only the odd - order equations of the hierarchy represent valid approximations for gyrotropic excitations ( for non - gyrotropic ones all members of the hierarchy are even ) .
the odd higher - order equations predict additional frequency doublets @xmath187 , @xmath188 etc .
these frequencies normally vanish in the background of the spectrum .
however , they become visible if the simulations are specially designed : for a vortex in the center of a quadratic system with antiperiodic boundary conditions , all image forces cancel exactly and only the small pinning forces remain . for this configuration two additional doublets with strongly decaying amplitudes
were indeed observed @xcite .
finally we mention that cycloidal vortex trajectories have been found not only in field - theoretic models for magnets @xcite , but also for ( 1 ) neutral and ( 2 ) charged superfluids : + ( 1 ) in the ginzburg - landau theory , which describes vortex motion in thin , neutral , superfluid films , the usual assumption of incompressibility was dropped @xcite .
a moving vortex then exhibits a density profile in the region outside the vortex core .
this profile is similar to the velocity dependent out - of - plane structure ( [ psi = q ] ) of the magnetic vortices , and the consequences are also similar : there are small - amplitude cycloids @xcite superimposed on the trajectories in both 2-vortex scenarios ( vortex - vortex rotation and vortex - antivortex translation , see end of section [ thieleeq ] ) .
+ ( 2 ) for the dynamics of vortices in a charged superfluid the same kind of superimposed cycloids was found , again for both 2-vortex scenarios @xcite . here
a field - theoretic model was used , where a charged scalar field is minimally coupled to an electromagnetic field and a @xmath189-potential is included ; this is proposed as a phenomenological model for a superconductor .
however , compared to the dynamics of magnetic vortices as reviewed in this article , there is an important difference : for all the models discussed above , there is only _
one _ cycloidal frequency ; correspondingly the vortex dynamics is governed by @xmath64-order equations of motion
. it will be interesting to study many other physical contexts in which topological vortex structures appear ( e. g. , dislocations in solids and flux lattices , vortex filaments in compressable fluids and complex ginzburg - landau models ) .
we expect that the details of the vortex dynamics differ , depending on the order of the equations of motion , for instance ( see beginning of this section ) .
the spirit of the _ generalized travelling wave ansatz _ ( [ s = s ] ) differs considerably from the well - known _ standard ansatz _ , which reads for a spin system @xmath190 here @xmath191 represents the _
static _ structure of a single nonlinear coherent excitation ( a vortex in our case ) and @xmath192 represents a magnon field ( or meson field for other kinds of systems ) . since @xmath191 is static , but the shape of the excitation usually depends on the velocity , a part of the dynamics obviously
must be taken over by the spin waves .
based on earlier work , this concept has very recently been carried through @xcite .
first , the magnon modes in the presence of a single static vortex were obtained by a numerical diagonalization for relatively small , discrete systems with fixed ( dirichlet ) boundary conditions ( bc ) @xcite .
analytical investigations were done for planar vortices in antiferromagnets @xcite and ferromagnets @xcite .
these articles demonstrated nontrivial properties of the eigenmodes , e. g. , the presence of quasi - local ( resonance type ) @xcite or truly local @xcite modes .
moreover , the relevance of these modes for the vortex dynamics was shown , in particular the transition between planar and non - planar vortices was investigated @xcite .
however , for non - planar vortices all this was based on numerical diagonalization .
but very recently , a general theory of vortex - magnon coupling was developed for arbitrarily large systems with circular symmetry and general bc @xcite .
the s - matrix for vortex - magnon scattering was calculated and expressed by bessel and neumann functions .
using the s - matrix , general formulas for the eigenfrequencies were obtained , as a function of the parameters and size of the system , and for different bc , namely for dirichlet ( fixed ) and von neumann ( free ) bc .
there is a very good agreement between the frequencies of the two lowest quasi - local modes and the frequencies @xmath134 , which were observed for the vortex trajectories in spin dynamics simulations for the discrete system ( section ( [ vortexmass ] ) .
the error is only 0.8 % for the mean frequency @xmath193 and 4 % for the splitting @xmath194 .
this demonstrates very clearly that the _ cycloidal vortex motion _ is accompanied by certain magnon modes , namely by _ quasi - local modes_. these modes are both extended ( like the other magnon modes ) and local , because they exhibit a localized structure around the vortex center ( fig .
[ spinwave ] ) .
the structure of these modes is qualitatively similar to the dynamic vortex structure , which oscillates with the rythm of the cycloidal motion ( figs .
[ psistruct1 ] , [ psistruct2 ] ) .
however , only the out - of - plane part @xmath195 of this structure can be clearly observed , and only in the outer region ( see section [ hierarchy ] ) .
as the frequencies @xmath134 in the cycloidal vortex motion are the same as the frequencies of the lowest quasi - local modes , and as the latter ones were calculated analytically , the parameters @xmath167 and @xmath58 in the @xmath148-order equation of motion ( [ axmx ] ) can also be calculated analytically @xcite .
the results can be expressed by the first root @xmath196 of the equation @xmath197 , where @xmath198 is a bessel function . the values for @xmath34 and
@xmath199 define the boundary conditions : @xmath200 for free bc and vice versa for fixed bc .
the general result @xmath201 in the limit of large @xmath106 specializes to @xmath202 for @xmath176 and free bc , which agrees perfectly with the result @xmath203 from the simulations ( section 2.3 ) . for fixed bc @xmath204
is about four times smaller .
the general result for the vortex mass @xmath58 is much more complicated than eq .
( [ a / pi ] ) and can be found in ref .
the numerical value @xmath205 for @xmath206 and free bc agrees well with the value @xmath207 from the simulations ( section 2.3 ) .
however , both results are independent of the system size @xmath106 , in contrast to the logarithmic @xmath106-dependence ( [ eq : m ] ) predicted by the integral ( [ eq : mij ] ) in the collective variable theory , cf . the discussion in section 2.3 .
for fixed bc , @xmath208 is obtained , about half of the above value for free bc . the fact
that both the mass @xmath58 and the factor @xmath167 in the @xmath148-order gyrocoupling term depend strongly on the boundary conditions appears very natural because the vortices are not localized excitations like solitons in 1d ; the in - plane vortex structure falls off with @xmath182 .
it is important to note that the higher - lying quasi - local modes could be calculated by the same methods , and are expected to agree with the higher - order doublets @xmath209 appearing in the spectrum of the trajectories .
each additional doublet corresponds to taking into account two additional orders in the generalized travelling wave ansatz ( [ s = s ] ) , as discussed in section [ hierarchy ] . finally , we mention the recent development of a very general collective variable theory for an arbitrary hamiltonian @xmath210 $ ] supporting nonlinear coherent excitations , without making any approximations @xcite .
this theory starts with the standard ansatz for @xmath211 and @xmath212 , cf .
( [ srt36 ] ) , but imposes constraints between the meson fields and the functions @xmath213 and @xmath214 , which describe the shape of the excitation . here
the @xmath215 are @xmath216 collective variables for the position and the internal modes of the excitation ( if there are any ) .
there are then @xmath217 constraints in order to preserve the correct number of degrees of freedom .
consequently , the mathematical formalism is based on the classical limit of dirac s quantum mechanics for constraint systems .
the equations of motion for the collective variables are a generalization of thiele s equation ( [ gx = f ] ) .
the _ rank _ of the gyrocoupling matrix @xmath55 is used to classify the excitations : in the case of vanishing @xmath55 the excitations have an effective mass and exhibit newtonian dynamics ; in the case of regular @xmath55 , the excitations behave like charged , massless particles in an external magnetic field , similar to the gyrotropic excitations defined in section [ hierarchy ] .
the above theory is a generalization of earlier work of tomboulis @xcite , boesch et al .
@xcite , and willis et al .
@xcite , which applies only to `` standard '' hamiltonians ( i. e. , consisting of the sum of kinetic and potential energy terms ) .
this generalization is necessary , e. g. , in the case of magnetic systems which can not be modelled by standard hamiltonians .
as already mentioned in the introduction , below @xmath23 vortex - antivortex pairs appear and vanish spontaneously due to thermal fluctuations .
but these pairs do not move and therefore give no contribution to the dynamic correlation functions . above @xmath23 ,
some of the pairs unbind and the free vortices can move . in section 4
, this situation will be investigated by a phenomenological theory , namely the vortex gas approach . on the other hand ,
the effect of thermal fluctuations on _ single _ vortices can be studied by putting a vortex in a thermal bath .
this is a non - equilibrium situation , in fact it takes a long time until equilibrium is reached : the vortex very slowly approaches the boundary where it annihilates together with an image antivortex ; during this process spin waves are radiated which are eventually thermalized .
the vortex motion with thermal noise is a random walk process , where a _ vortex diffusion constant _ can be defined .
this offers the possibility to develop an ab - initio theory for the dynamic form factors .
this can be compared with the phenomenological vortex gas approach , which assumes , however , a ballistic motion ( section [ tempgtkt ] ) . in principle , the generalization of a collective variable theory to finite temperatures is a straightforward procedure consisting of 4 steps : + ( 1 ) thermal noise and damping ( because of the fluctuation - dissipation theorem ) are introduced into the microscopic equations .
+ ( 2 ) a travelling wave ansatz for a nonlinear coherent excitation is made which yields equations of motion with stochastic forces acting on the excitation .
+ ( 3 ) the solutions of these stochastic o. d. e.s are used to calculate the variances of the trajectory , which contain as a factor an effective diffusion constant for this excitation .
the dependences of this constant on the temperature and other parameters can be discussed .
+ ( 4 ) the predicted variances and the effective diffusion constant are compared with the same quantities as observed in langevin dynamics simulations .
this step is very problematical , although there are many papers in which the problems are either not appreciated or hidden .
there are basically two major problems : + ( a ) the microscopic equations can often be written in different ways which are equivalent . in the case of spin systems , the landau - lifshitz equation for the spin @xmath15 is equivalent to the hamilton equations for @xmath218 and @xmath195 .
however , if noise is implemented , e. g. , by an additive term , the results may be _ qualitatively different _ ( see below ) .
+ ( b ) typically , either _ additive _ or _ multiplicative _ noise can be used , but many papers do not give a reason why one of the two types was chosen . however , the results are usually _ qualitatively different _ for the two types of noise . in the case of the vortex dynamics ,
both major problems have been investigated in a preprint @xcite : additive noise in the hamilton equations yields a vortex diffusion constant @xmath219 which depends _ logarithmically _ on the system size , while in the multiplicative case @xmath219 is _ independent _ of the system size .
additive noise in the landau - lifshitz equation @xcite yields a diffusion constant with the same logarithmic term as in the case of additive noise in the hamilton equations , but the small constants that appear in addition to the logarithmic term differ for the two cases because the vortex core gives different contributions .
unfortunately , the additive noise in the landau - lifshitz equation has an unphysical feature , namely the spin length is not conserved . in the collective variable theory
this problem is overcome by a constraint , while in the simulations a renormalization of the spin length is necessary after every time step .
however , taking additive noise and using these little tricks is not really necessary , because there is a better noise term which eventually leads to the same results , but which is well motivated on physical grounds and which conserves the spin length @xcite : in the landau - lifshitz equation ( [ ds ] ) each spin @xmath65 precesses in a local magnetic field @xmath126 which is the gradient of the energy with respect to the spin compononts .
this local field is the only way through which the spin @xmath65 can feel any changes in its environment . adding a thermal noise term to the local field thus accounts for the interaction of the spin with magnons , phonons and any other thermally generated excitations .
the stochastic landau - lifshitz equation then reads because of the cross product with @xmath65 this noise is _
multiplicative_. in ref .
@xcite gaussian white noise is used : @xmath221 where @xmath222 is the diffusion constant and @xmath223 denote cartesian components . following refs .
@xcite , a gilbert damping with damping parameter @xmath224 was chosen in eq .
( [ ds = sm ] ) , chiefly because it is isotropic , in contrast to the landau - lifshitz damping @xcite . strictly speaking
, the three equations ( [ ds = sm ] ) do not really represent langevin equations , because _ all _ the components of @xmath225 appear in each equation due to the cross product . to properly introduce the noise ,
one first has to group all the time derivatives on the l. h. s. of the equation , and only then one can introduce independent white noise terms for each spin component . but
this procedure eventually produces only a renormalization of the damping parameter @xmath224 in the order of @xmath226 . in the simulations values of @xmath224 in the order of @xmath9
were used , and therefore the correction factor can even be neglected .
another , even more important issue is the interpretation of the stochastic differential equation ( [ ds = sm ] ) . as the noise is _ multiplicative _ , ito and stratonovic interpretations do not coincide , in contrast to the additive noise case @xcite . in principle , when thinking of thermal excitations interacting with the spins , there would be a finite correlation time which would lead to a colored noise term .
taking white noise means taking the limit of zero correlation time , and therefore the stochastic landau - lifshitz equation ( [ ds = sm ] ) must be interpreted in the _
stratonovic sense_. the generalized travelling wave ansatz ( [ s = s ] ) up to order @xmath147 yields the @xmath148-order stochastic o. d. e.@xmath227[c]{.}\makebox[0.45ex][c]{.}\makebox[0.45ex][c]{. } $ } } } { x}}}$ } + { \bf m } { \bf \ddot{x } } + { \bf g } { \bf \dot{x } } = { \bf f}({\bf x } ) + { \bf f}^{\rm mult } ( t)\,\ ] ] the tensors @xmath228 and @xmath55 are the same as in eqs .
( [ gij2 ] ) , ( [ eq : m ] ) and ( [ eq : a ] ) , except that all the vanishing components in these three expressions are now replaced by damping terms proportional to @xmath224
. thus the damping appears at every order in eq .
( [ ax ] ) . without @xmath229 ,
the solution is obtained in complete analogy to section [ hierarchy ] : two cycloids are superimposed on an outward spiral on a circle ( fig . [ trajsketch ] ) . in the simulations
the purpose of the damping is to dissipate the energy which is supplied to the system by the kicks of the noise . therefore the range of @xmath224 ( for a given system size @xmath106 ) must be determined in which the cycloidal frequencies @xmath134 in the trajectories are not influenced by the damping @xcite .
the result is the condition @xmath230 . the stochastic force in ( [ ax ] ) @xmath231 which stems from the multiplicative noise in eq .
( [ ds = sm ] ) , has zero mean and the correlation function @xcite @xmath232 putting the vortex on the center of a circular system of radius @xmath106 with free bc , the leading contribution to the variance is obtained @xcite by using the static vortex structure ( [ phi = q ] ) to ( [ psi = r_gg_rv ] ) @xmath233 with the vortex diffusion constant @xmath234 here @xmath235 stems from the inner vortex region @xmath115 and is obtained by numerical integration @xcite .
the variance ( [ var ] ) is a remarkable result for two reasons : + ( a ) _ the stochastic forces acting on the vortex represent additive white noise _ with an effective diffusion constant .
the point is that this was shown by going from the microscopic level , where _ multiplicative _ white noise was implemented , to the level of the collective variables .
this approach is much more satisfying than assuming _ ad hoc _ a noise term on the collective variable level .
+ ( b ) the mean and variance of the stochastic force ( [ fist ] ) turn out to be the _ same _ as those of the stochastic force @xmath236 which resulted from starting with additive white noise @xmath237 in the landau - lifshitz equation ; this noise has already been discussed in step 1 .
this equivalence of the two forces means that in the case of the vortices additive noise and the multiplicative noise in eq .
( [ ds = sm ] ) have the _ same _ effect on the vortex dynamics .
this is a nontrivial result , because normally a qualitative difference is expected ( cf . the discussion in step 1 ) .
qualitatively , this result can be understood by noting that the renormalization of the spin length in the case of additive noise introduces effectively a multiplicative noise term , in addition to the additive one ( cf.footnote in step 1 ) .
the inhomogeneous stochastic @xmath148-order o. d. e. ( [ ax ] ) can be solved @xcite by a standard green s function formalism , because the l. h. s. is linear and the deterministic force @xmath238 on the r. h. s. can be linearized by expanding in the displacement @xmath239 from the mean trajectory @xmath240 on which the vortex is driven by @xmath48 , cf .
( [ fr ] ) . knowing @xmath241 , the variance matrix @xmath242 can be calculated ; here @xmath243 denote polar coordinates @xmath140 and @xmath218 .
each element of the matrix ( [ defvar ] ) turns out to consist of 36 terms . in order to facilitate the discussion
several simplifications can be made which give the following results @xcite : @xmath244\ ] ] with @xmath245 and @xmath246 is the damping constant for the vortex motion . for large times ,
@xmath247 , only the first term remains and the variance is linear in time , the standard random walk result .
interestingly , this result is identical to the one which is obtained by omitting the @xmath64- and @xmath148-order terms in the stochastic equation of motion ( [ ax ] ) .
this obviously means that these two terms have the following effects : + ( a ) they produce the oscillatory parts in eq .
( [ sigma2 ] ) , which are naturally connected to the cycloidal oscillations in the vortex trajectories , + ( b ) for small times , @xmath247 , the slope of @xmath248 , averaged over the oscillations , is larger by a factor of 3/2 compared to late times .
however , the first effect can not be observed in the simulations because the oscillations are hidden in strong discreteness effects ( see next step ) .
for this reason we discuss only the long - time behavior of the other elements of the variance matrix @xmath249\end{aligned}\ ] ] with @xmath250 and @xmath251 , @xmath252 .
the quadratic and cubic time dependences are _ non - standard _ results which appear in addition to the standard linear dependence because the deterministic force field @xmath253 is _
@xmath254 is a radial force which drives the vortex in the azimuthal direction , due do the gyrocoupling force ( [ fg = x ] ) .
therefore , only the @xmath218-components of @xmath255 are affected : @xmath256 acquires a factor @xmath257 , while @xmath258 acquires it twice .
the stochastic landau - lifshitz equation ( [ ds = sm ] ) was solved numerically for a large circular lattice with free bc and one vortex driven by a radial image force @xcite .
the mean trajectory is an outward spiral .
therefore a small damping parameter @xmath224 was chosen in order to allow long integration times .
there are several qualitatively distinct temperature regimes . for @xmath259 ( in dimensionless units , where @xmath260 for the xy - model @xcite ) , two frequencies are observed in the oscillations around the mean trajectory which can be identified with the cycloidal frequencies
@xmath134 in section [ hierarchy ] . as these frequencies are constant in the whole regime , the @xmath148-order equation of motion ( [ ax ] ) with _ temperature independent _ parameters can in fact describe the vortex dynamics . for @xmath261 the above two frequencies
can not be observed any longer due to large fluctuations , thus the @xmath262-order equation of motion is sufficient here .
however , in some of the runs the vortex suddenly changed its direction of motion ; this will be explained in the next subsection .
finally , for @xmath263 , a single vortex theory is no longer adequate because here the probability for the spontaneous appearance of vortex - antivortex pairs in the neighborhood of the single vortex becomes too large . the linear , quadratic , and cubic time dependences in eqs .
( [ sigma2 ] ) - ( [ sigma_phi ] ) are well confirmed by the simulations .
therefore the factor @xmath219 can be fitted and turns out to differ from the predicted vortex diffusion constant ( [ dv ] ) only by a _
nearly temperature independent _ factor of about @xmath264 , see fig .
[ d_effplot ] ( the constant @xmath235 in ( [ dv ] ) was obtained by a numerical integration over the vortex core @xcite ) .
this agreement is amazingly good , taking into account that the simulations were performed for a discrete lattice while the theory works in the continuum limit and uses additional approximations , like the expansion ( [ fr ] ) of the coulomb force .
although discreteness effects are hardly visible in the trajectories of non - planar vortices at zero temperature @xcite , see also fig .
[ trajectory ] , these effects are very important at finite temperatures : fig .
[ discrete ] clearly demonstrates that the _ variance _ of the trajectory shows a pronounced peak whenever the vortex center moves over a ridge of the peierls - nabarro potential . due to this effect , the predicted cycloidal oscillations ( [ sigma2 ] ) in the variance can not be observed . as already mentioned above , a vortex in a thermal bath can suddenly change its direction of motion on its outward spiral .
a closer inspection shows that this occurs because , opposite to the vorticity @xmath265 , the polarization @xmath42 is not a constant of motion for a _ discrete _ system : the out - of - plane vortex structure can flip to the other side of the lattice plane due to the stochastic forces acting on the core spins .
then the direction of the gyrovector ( [ g=2 ] ) is reversed and according to the thiele equation ( [ xxg ] ) the direction of motion is reversed as well ( the same holds for the @xmath148-order equation ( [ axmx ] ) because @xmath266 in eq .
( [ eq : a ] ) ) . in a preprint @xcite the cores of both planar and non - planar vortices
are described by a discrete hamiltonian , similar to the one which was used for the study of the instability at @xmath267 @xcite . using the stochastic landau - lifshitz equation ( [ ds = sm ] ) the fokker - planck equation
is derived .
its stationary solution exhibits two maxima for the two possible polarizations of the non - planar vortex and a saddle point for the planar vortex , if the anisotropy parameter @xmath19 lies in a certain temperature dependent range .
the rate @xmath268 for the transition from one polarization to the opposite one is calculated in analogy to langer s instanton theory @xcite , using the fact that for @xmath269 there is a soft mode among the normal modes which were obtained numerically for a system with one vortex @xcite .
taking into account only the four innermost spins of the core , a very simple result is obtained @xmath270 where @xmath271 is the energy difference between the planar and the non - planar vortex . for @xmath272 , the average transition times @xmath273 are 100440 , 4837 , and 1060 for @xmath274 , @xmath275 , and @xmath276 respectively . despite of the crude model for the vortex core
, these values agree rather well with the transition times from langevin dynamics simulations : @xmath277 , @xmath278 , and @xmath279 for the above temperatures , with statistical errors of @xmath280 , @xmath281 , and @xmath282 , respectively .
this is a phenomenological theory which is based on the following assumptions : above @xmath23 , the _ free _ vortices form a _ dilute _ gas and move either _ diffusively _ @xcite or _ ballistically _ @xcite . in the former case ,
only spin _ autocorrelation _ functions were calculated which lead to dynamic form factors without wavevector dependence .
therefore we review only the latter case @xcite : the density @xmath283 and the r. m. s. velocity @xmath284 are the only free parameters . according to kosterlitz and thouless @xcite @xmath285 where @xmath286 is the static spin correlation length which diverges with an essential singularity for @xmath287 from above .
therefore the vortex gas is in fact dilute if @xmath288 is close enough to @xmath23 .
the vortex density is homogeneous only on the average , locally the distribution is expected to be _
random_. therefore the net force is zero on the average and the distribution around zero is gaussian , which yields a _ maxwellian _ velocity distribution .
this also holds for the non - planar vortices because the velocity is proportional to the cross product of the net force and the gyrovector , due to the thiele equation ( [ xxg ] ) . however , the assumption of a ballistic vortex motion is problematic ( section [ mcmotion ] ) . under this assumption
the dynamic spin - spin correlation functions @xmath289 and their space - time fourier transforms , namely the dynamic form factors @xmath290 , can be calculated analytically @xcite .
there is an important difference between in - plane correlations ( @xmath291 ) and out - of - plane ones ( @xmath292 ) : as the in - plane vortex structure is _ not _ localized ( it has no spatial fourier transform ) , @xmath293 are only _ globally sensitive _ to the presence of the vortices and the characteristic length scale is their average distance @xmath294 .
the dynamic form factor exhibits a `` _ central peak _ '' , i. e. a peak around @xmath295 , with the ( squared ) _ lorentzian _ form @xmath296}^2 } \,,\ ] ] where @xmath297 . here
the core structure did not enter the result , because the theory was worked out on a length scale much larger than the core radius @xmath298 .
the integrated intensity @xmath299 of ( [ sxx ] ) is _ inversely _ proportional to the density @xmath283 , therefore the motion of the vortices actually _ destroys _ correlations . for the out - of - plane correlations @xmath300
the situation is completely different : the non - planar vortices have statically the localized @xmath28 structure ( [ psi = r_ll_rv ] ) , ( [ psi = r_gg_rv ] ) which has a spatial fourier transform , namely the static form factor @xmath301 .
therefore @xmath300 is _ locally sensitive _ to the vortices , i. e. , to their size and shape .
consequently @xmath302}\ ] ] contains @xmath301 and is proportional to the density .
this _ gaussian _
central peak simply reflects the maxwellian velocity distribution . for the planar
vortices the situation is more complicated because they do not have a static @xmath28 structure but only a dynamic one , namely eqs .
( [ psi = q ] ) , ( [ eq : phi ] ) .
therefore the form factor is velocity dependent which yields a more complicated result for @xmath300 @xcite , containing the same gausssian as in eq .
( [ szz ] ) .
however , the intensity of this peak is much smaller than that of the peak ( [ szz ] ) , because the dynamic @xmath28 components are about two orders of magnitude smaller than the static ones @xcite . in combined monte - carlo ( mc ) and spin dynamics ( sd ) simulations many features of the predicted dynamic form factors ( [ sxx ] ) and ( [ szz ] )
were confirmed : + ( 1 ) _ in - plane correlations _ : both for the xy - model ( @xmath303 ) @xcite and for the weakly anisotropic case ( @xmath304 ) @xcite , the observed @xmath305 exhibits a central peak for @xmath306 ( but not for @xmath307 , as expected ) .
the statistical errors were too large to decide about the shape .
however , the width @xmath308 and the integrated intensity @xmath299 of the lorentzian ( [ sxx ] ) can be fitted to the data , in this way the free parameters @xmath286 and @xmath284 are determined .
@xmath309 agrees rather well with the static correlation length @xcite .
@xmath310 increases with @xmath288 and then saturates ; except of a factor of about 2 , this agrees with a result of huber @xcite who assumed a diffusive vortex motion and calculated the velocity autocorrelation function .
central peaks were also observed in inelastic neutron scattering experiments : for @xmath311 @xcite and @xmath2 @xcite the measurement was performed for only a few @xmath265-values .
the reported widths of the peaks differ from the predicted @xmath308 by factors of about @xmath312 and @xmath313 , resp .. however , in this comparison one has to take into account that the theory has neglected many features of the real quasi-2d materials : e. g. , the lattice structure , a pronounced in - plane anisotropy in the case of @xmath314 , and quantum effects .
\(2 ) _ out - of - plane correlations _ : in mc - sd simulations , @xmath315 exhibits a central peak only together with a spin wave peak which sits on its shoulder ( in contrast to @xmath316 where the spin waves are strongly softened for small @xmath265 ) . for the xy - model ( @xmath303 ) , which bears planar vortices ,
the intensity of the central peak is expected to be small ( see end of section [ vortexgas ] ) .
early simulations with large statistical errors @xcite had difficulties to substract the spin wave peak and reported only upper bounds for the width @xmath317 and the intensity @xmath318 of the central peak .
recent simulations with much better statistics investigated the range @xmath319 .
some papers did not find a central peak @xcite others did @xcite .
however , for the weakly anisotropic case ( @xmath304 ) with non - planar vortices a central peak was indeed found @xcite .
experimentally , @xmath315 was first not accessible because of intensity problems .
only the use of spin - polarized neutron beams made it possible to observe a central peak in @xmath320 and to distinguish it from the spin wave peak and from other contributions @xcite .
the measured width is practically equal to the width of the gaussian ( [ szz ] ) .
however , such a good agreement seems to be accidental because this material exhibits a breaking of the rotational symmetry in the xy - plane which is described by a more complicated hamiltonian @xcite .
for the same reason it is not clear whether this material actually belongs to the case @xmath304 .
the vortex gas approach assumes for single free vortices a diffusive @xcite or ballistic @xcite motion .
very recently this was tested in mc - simulations by monitoring the position of each vortex in the system ( free or bound in a pair ) as a function of time @xcite .
the surprising result is that a single vortex very seldom moves freely over a larger distance .
normally the vortex travels only one or a few lattice spacings until it annihilates with the antivortex of a pair which meanwhile appeared spontaneously in the neighborhood .
another possibility is that the single vortex docks on a pair or a cluster of pairs and after a while one of the vortices leaves the cluster .
these results confirm suggestions that vortices can not move freely for more than a few lattice spacings , which were made by computing the vortex density - density correlation function @xcite . however , the interpretation of these results is not clear at all .
one possibility is that the vortex gas approach is not valid @xcite , but then the striking qualitative agreement with the central peaks in the simulations and experiments is not understood .
we favour another possibility : probably only the _ effective _ vortex motion is important for the dynamic correlation function @xmath321 ( [ spinspincorr ] ) .
this would mean that it does not matter if a vortex is annihilated with the antivortex of a pair in the considered time interval @xmath322 $ ] , because the vortex of that pair continues the travel instead of the original vortex .
only effective lifetimes are seriously affected .
this picture is strongly supported by the mechanism @xcite which yields the central peak @xmath323 in eq .
( [ sxx ] ) : looking at the orientation of one particular spin in the xy - plane , one sees that its direction is _ reversed _ after the vortex center has gone over this spin or its neighborhood .
thus @xmath29 and @xmath30 have _ changed their signs _ after the vortex is gone , and therefore @xmath324 here @xmath325 is the number of vortices which pass an arbitrary nonintersecting contour connecting @xmath326 and @xmath327 .
obviously , it does not matter if those vortices are replaced by other vortices during their travel ( except maybe if the replacement happens just when the contour is passed ) .
the evaluation of eq .
( [ sxx=-1^n ] ) leads to the central peak ( [ sxx ] ) @xcite .
the vortex dynamics at zero temperature can be well understood by a collective variable theory : a generalized travelling wave ansatz which allows for deformations of the vortex shape due to velocity , acceleration etc .
, leads to equations of motion . in the case of non - planar vortices ,
the trajectories exhibit a superposition of cycloids which is fully confirmed by spin dynamics simulations .
this collective variable theory can be generalized to finite temperatures which yields stochastic equations of motion .
the vortex motion is diffusive and agrees well with langevin dynamics simulations .
moreover , the rate of noise - induced transitions between vortex states with opposite polarization is calculated and agrees with the simulations .
however , above the kosterlitz - thouless transition temperature there is so far no theory which can fully explain the central peaks which were observed both in inelastic neutron scattering experiments and in combined monte - carlo and spin dynamics simulations .
a qualitative agreement is achieved by a phenomenological vortex gas theory , but one of its assumptions , namely ballistic vortex motion , is questionable .
probably , both the diffusive character of the vortex motion and annihilation and creation processes must be incorporated into a theory which can fully explain the above facts .
it is likely that the situation is similar for all kosterlitz - thouless ( and many other ) phase transitions .
many thanks for a very fruitful collaboration go to ( in historical order ) : gary wysin ( kansas state university , usa ) , elizabeth gouva ( belo horizonte , brazil ) , armin vlkel ( xerox , palo alto , usa ) , hans - jrgen schnitzer ( aachen , germany ) , angel snchez , francisco domnguez - adame , esteban moro ( madrid , spain ) , niels grnbech - jensen ( los alamos , usa ) , boris ivanov , yuri gaididei ( kiev , ukraine ) , alex kovalev ( kharkov , ukraine ) and till kamppeter ( bayreuth , germany ) .
we are grateful to frank ghmann ( stony brook , usa ) , dimitre dimitrov , grant lythe , and roman sasik ( los alamos , usa ) for critical readings of the manuscript and constructive comments .
k. hirakawa , h. yoshizawa , j. d. axe , and g. shirane , suppl .
. jpn . * 52 * , 19 ( 1983 ) ; l. p. regnault , j. p. boucher , j. rossat - mignod , j. bouillot , r. pynn , j. y. henry , and j. p. renard , physica b+c * 136b * , 329 ( 1986 ) ; m. t. hutchings , p. day , e. janke , and r. pynn , j. magn .
magn . mater . *
54 - 57 * , 673 ( 1986 ) ; s. t. bramwell , m. t. hutchings j. norman , r. pynn , and p. day , j. de phys . * 49 * , c8 - 1435 ( 1988 ) ; l. p. regnault , c. lartigue , j. f. legrand , b. farago , j. rossat - mignod , and j. y. henry , physica b * 156 - 7 * , 298 ( 1989 ) t. kamppeter , f. g. mertens , a. snchez , n. grnbech - jensen , and a. r. bishop , in _ theory of spin lattices and lattice gauge models _ , eds . j. w. clark and m. l. ristig , lecture notes in physics , springer ( 1997 ) l. p. regnault , j. p. boucher , j. rossat - mignod , j. bouillot , r. pynn , j. y. henry , j. p. renard , physica b+c * 136b * , 329 ( 1986 ) ; l. p. regnault and j. rossat - mignod , in _ magnetic properties of layered transition metal compounds _ , ed . by l. j. de jongh and r. d. willet , kluwer academic ( 1990 ) b. v. costa , d. p. landau , j. e. r. costa , and k. chen in _ computer simulation studies in condensed matter physics _
viii , ed . by d. p. landau , k. k. mon , and h. b. schuettler , springer verlag , berlin ( 1995 ) | theories , simulations and experiments on vortex dynamics in quasi - two - dimensional magnetic materials are reviewed .
these materials can be modelled by the classical two - dimensional anisotropic heisenberg model with @xmath0 ( easy - plane ) symmetry .
there are two types of vortices , characterized by their polarization ( a second topological charge in addition to the vorticity ) : planar vortices have newtonian dynamics ( even - order equations of motion ) and exhibit strong discreteness effects , while non - planar vortices have non - newtonian dynamics ( odd - order equations of motion ) and smooth trajectories .
these results are obtained by a collective variable theory based on a generalized travelling wave ansatz which allows a dependence of the vortex shape on velocity , acceleration etc .. an alternative approach is also reviewed and compared , namely the coupling of the vortex motion to certain quasi - local spinwave modes .
the influence of thermal fluctuations on single vortices is investigated .
different types of noise and damping are discussed and implemented into the microscopic equations which yields stochastic equations of motion for the vortices .
the stochastic forces can be explicitly calculated and a vortex diffusion constant is defined .
the solutions of the stochastic equations are compared with langevin dynamics simulations .
moreover , noise - induced transitions between opposite polarizations of a vortex are investigated . for temperatures above the kosterlitz - thouless vortex - antivortex unbinding transition , a phenomenological theory , namely the vortex gas approach ,
yields central peaks in the dynamic form factors for the spin correlations .
such peaks are observed both in combined monte carlo- and spin dynamics - simulations and in inelastic neutron scattering experiments .
however , the assumption of ballistic vortex motion appears questionable . |
the aharonov - bergmann - lebowitz ( abl ) rule@xcite was proposed in 1964 to compute measurement result probabilities of systems that are both preselected and postselected .
it was meant to provide a time - reversal invariant formulation of nonrelativistic quantum mechanics . as long as a measurement
is actually carried out between pre- and postselection , the abl rule is a straightforward consequence of the born probability rule and usual assumptions on the quantum state following a measurement .
attempts have been made , however , to interpret the abl rule counterfactually , that is , to cases where no intermediate measurement is made .
the resulting controversy is , i believe , partly due to the fact that proposed definitions of counterfactuals in the present quantum - mechanical context do not adequately capture the intuitive meaning of the notion .
in this paper i first review the derivation of the abl rule and some of its properties , in particular contextuality .
next i analyze alleged proofs of the impossibility of interpreting the rule counterfactually , as well as various arguments supporting or opposing them .
i attempt to clarify the meaning of a counterfactual interpretation , and point out circumstances in which the counterfactual assertion of the abl rule is or is not correct .
consider a quantum system @xmath3 and three observables @xmath4 , @xmath5 and @xmath6 pertaining to it . for simplicity
, we shall first assume that these observables are discrete and nondegenerate .
suppose that at time @xmath7 , @xmath3 is prepared in an eigenstate @xmath8 of @xmath4 . at @xmath9 ,
the observable @xmath5 is measured . according to the born rule , the probability of obtaining result @xmath10 ( where @xmath11 is the set of eigenvalues of @xmath5 ) is given by @xmath12 .
suppose the result @xmath10 is indeed obtained upon measurement at @xmath0 .
we assume the measurement interaction is such that immediately after @xmath0 , the system is in state @xmath13 . at a later time
@xmath14 , the observable @xmath6 is measured .
the probability of obtaining result @xmath15 is then given by @xmath16 .
hence the probability of result @xmath10 and result @xmath15 , conditional on preparation @xmath8 , is equal to @xmath17 we are interested here in the probability of obtaining @xmath10 at @xmath0 , on the condition that @xmath3 is prepared in @xmath8 at @xmath1 and found in @xmath18 at @xmath2 .
this we shall denote by @xmath19 . from the definition of conditional probability
, we have ( for @xmath20 ) @xmath21 here @xmath22 is the total probability of @xmath15 ( given preparation @xmath8 ) , equal to the sum of @xmath23 over all possible results @xmath24 . from ( [ joint ] )
we obtain @xmath25 this is the abl rule@xcite . in the right - hand side ,
@xmath26 and @xmath27 are projectors on states @xmath8 and @xmath18 .
the operator @xmath28 projects on @xmath13 or , if the eigenvalue @xmath10 is degenerate , on the associated subspace .
is then carried out on these subspaces . ]
the abl rule ( expressed in terms of projection operators ) also holds in that case , provided the intermediate measurement obeys the lders rule @xmath29 ( where @xmath30 is a normalization constant ) .
the abl rule can also be written for multiple intermediate measurements or for selection by means of density matrices rather than pure states , but we wo nt need these generalizations here .
a much discussed property of the abl rule is its contextuality .
let @xmath31 , @xmath32 and @xmath33 be an orthonormal basis in a three - dimensional hilbert space .
each of these kets can represent a state wherein a particle is in one of three disjoint boxes .
take the initial and final states as let the intermediate observable @xmath5 be so chosen as to have @xmath31 as one of its eigenvectors , corresponding to a nondegenerate eigenvalue @xmath34 .
the abl rule then yields @xmath35 it turns out that the right - hand side of this equation is not well - defined unless @xmath5 itself is well - defined .
let @xmath34 , @xmath36 and @xmath37 be three different real numbers and define @xmath38 elementary algebra shows that @xmath39 .
but if @xmath40 then @xmath41 .
hence the abl probability of @xmath34 depends not only on the eigenspace associated with that eigenvalue , but also on the structure of the observable in the orthogonal eigenspace .
this is contextuality . in terms of boxes , this means the following . if an observable ( @xmath42 ) distinguishes the first box from the other two taken together , then the abl probability of being in the first box is 1 .
if another observable ( @xmath5 ) distinguishes the three boxes one from another , the abl probability of being in the first box is 1/3 .
it is easy to check that if @xmath43 , @xmath44 likewise if @xmath45 , @xmath46 from ( [ pa ] ) and ( [ pb ] ) , albert , aharonov and damato have argued that noncommuting observables @xmath4 and @xmath6 must be simultaneously well - defined at @xmath0@xcite .
this conclusion hinges on a counterfactual use of the abl rule .
it was indeed intended to apply specifically to the case where neither @xmath4 nor @xmath6 are measured between pre- and postselection .
can the abl rule be interpreted counterfactually ?
that question was answered in the negative by sharp and shanks@xcite , cohen@xcite and miller@xcite , and much debated afterwards .
the crux of the sharp and shanks argument ( as well as others ) can be stated rather succinctly .
let @xmath47 and @xmath48 be the possible final states of a two - state system .
suppose @xmath5 is not measured , and assume that the abl rule correctly gives the probability of nondegenerate result @xmath34 , conditional on pre- and postselection , had @xmath5 been measured .
the total probability of @xmath34 should then be given as a weighted sum on the possible final states , that is , @xmath49 here @xmath50 is the probability of final state @xmath18 when no intermediate measurement is made , and @xmath51 is the conditional probability of @xmath34 based on the counterfactual interpretation of the abl rule . but according to standard quantum mechanics , @xmath52 is given by @xmath53 .
hence we should have @xmath54 since this is not true in general ( counterexamples are easily found ) , sharp and shanks conclude that the counterfactual use is invalid .
the validity of the counterfactual use and the relevance of the sharp and shanks argument were debated between vaidman@xcite and kastner@xcite .
vaidman s objection to the proof consists in pointing out that the weight @xmath50 in ( [ ss ] ) is the probability of @xmath15 if no intermediate measurement occurs .
since we are asking for the total probability of result @xmath34 , we must use the expression for the probability of @xmath15 if @xmath5 is measured at @xmath0 . that probability
is given by @xmath55 .
but then eq . ( [ ss ] ) becomes p(c_1 |a ) & = & _ j ( p_b p_c_j p_a p_c_j ) p(c_1 |a , b ) + _ j ( p_b p_c_j p_a p_c_j ) p(c_1 |a , b ) + & = & ( p_b p_c_1 p_a p_c_1 ) + ( p_b p_c_1 p_a p_c_1 ) + & = & ( p_c_1 p_a p_c_1 ) + & = & |c_1 |a |^2 , [ ss1 ] in accordance with standard quantum mechanics .
the significance of this calculation of @xmath52 is best brought out by quoting vaidman s definition of the counterfactual meaning of the abl rule@xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ if a measurement of an observable @xmath5 were performed at time @xmath0 , then the probability for @xmath56 would equal @xmath57 , provided that the results of measurements performed on the system at times @xmath7 and @xmath14 are fixed . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this is a statement about the statistical distribution of results of unperformed experiments , made on the basis of a law derived from a large number of performed experiments _ identical to the former in all relevant respects_. it is like saying , on the basis of numerous tosses of a die yielding essentially uniform outcomes , that had the die been tossed one additional time , the probability of obtaining `` 3 '' would have been 1/6 .
both this statement and vaidman s definition express the regularity of nature .
both assert , on the basis of a general rule drawn from numerous experimental trials , that if the experiment were done over again in the same relevant conditions , the results would fall under the same general rule .
hence vaidman s statement is ( presumably ) true , genuinely counterfactual , but not particularly illuminating , at least as far as specificities of quantum mechanics are concerned .
in contrast with vaidman , kastner defended the sharp and shanks proof and argued that a nontrivial counterfactual assertion of the abl rule ( i.e. one that provides information about specific systems ) is false .
kastner s analysis draws from both goodman s and lewis s theories of counterfactuals@xcite .
her argument can be summarized as follows .
a counterfactual @xmath58 ( read `` if it were that @xmath59 , it would be that @xmath60 . '' ) is true if @xmath59 is not true and the conjunction of @xmath59 with laws of nature @xmath61 and suitable background conditions @xmath3 implies @xmath60 , that is , @xmath62 now clearly , there must be restrictions on @xmath3 for , if @xmath3 includes the statement @xmath63 ( `` not @xmath59 '' ) , implication ( [ pls ] ) will hold trivially .
these restrictions usually state that @xmath3 should be `` cotenable '' with @xmath59 , a notion not so easy to define but essentially meaning that @xmath3 is independent of the truth of @xmath59 .
suppose that in the real world , preparation @xmath8 is followed by postselection @xmath18 , with no intermediate measurement .
then had there been an intermediate measurement , the intermediate state would have changed , and the final measurement result @xmath15 could no longer be expected to obtain . in other words ,
result @xmath15 is not cotenable with measuring @xmath5 .
does it follow from this argument that the nontrivial counterfactual assertion of the abl rule is false ?
the answer is affirmative if the background conditions must include the result @xmath15 .
but this is not necessarily the case .
the background conditions can be construed as encompassing everything that characterizes the state of the system at @xmath0 , including whatever might induce it to yield result @xmath15 upon measurement of @xmath6 at @xmath14 . with such a definition ,
the result @xmath15 itself is not part of the background conditions .
and although cotenability might be analyzed in that context , i shall use a different approach .
i submit that the proper way of investigating the counterfactual validity of the abl rule is to enquire about the `` true value '' of @xmath5 at @xmath0 .
strict empiricists will no doubt quit reading right here , since it is absolutely impossible , in a situation where @xmath5 has not been measured between @xmath7 and @xmath14 , to empirically ascertain what the true value of @xmath5 was at the intermediate time @xmath0 .
but that does nt prevent different theories or interpretations to make claims on what the true value is , as we will presently see .
it is instructive to analyze the sharp and shanks argument in terms of true values .
if , following standard quantum mechanics , @xmath52 is equal to @xmath64 , then the first equality in ( [ ss1 ] ) holds identically .
but sharp and shanks claim that the counterfactual meaning of the abl rule is encapsulated in eq .
( [ ss ] ) .
it is easy to see that ( [ ss ] ) and ( [ ss1 ] ) will coincide if @xmath65 together with a similar equation with @xmath15 replaced by @xmath66 . in quantum mechanics , eq .
( [ bcac ] ) does not hold in general .
the left - hand side of ( [ bcac ] ) is the uncontroversial probability of @xmath15 when no intermediate measurement is performed , while the right - hand side is the uncontroversial probability of @xmath15 when @xmath5 is measured at the intermediate time . that the equality is false means that whichever true value @xmath5 does or does not have at the intermediate time @xmath0 , it can not be one that is simply revealed in and unaffected by an eventual measurement .
it was pointed out in@xcite that ( [ bcac ] ) holds if @xmath67 makes up a consistent family of histories . indeed in that case one can maintain that @xmath5 has a well - defined value ( equal to one of its eigenvalues ) without running into contradictions .
but of course one is not compelled to do so .
associating or not associating true values to observables belonging to a consistent family of histories contributes in defining the interpretation of the theory .
it is instructive to recall the example of sect . 2 , with @xmath8 and @xmath18 defined as in eqs . ( [ ar ] ) and ( [ br ] )
let @xmath5 and @xmath42 be defined as in ( [ obsc ] ) and ( [ obsc1 ] ) and let @xmath68 it is easy to check that @xmath69 and @xmath70 make up distinct consistent families of histories .
the abl probability that @xmath71 is one , and so is the abl probability that @xmath72 .
we wo nt run into contradictions if we maintain that the particle was surely in box 1 at @xmath0 .
likewise we wo nt run into contradictions if we maintain that it was in box 2 .
of course both statements can not be held at once , since they are logically contradictory .
note that the family of histories @xmath73 more refined than either ( [ h1 ] ) or ( [ h2 ] ) , is not consistent .
the families @xmath74 and @xmath75 , where @xmath76 and @xmath77 stand for the set of eigenvalues of @xmath4 and @xmath6 , are always consistent .
hence it can be maintained that @xmath4 has the true value @xmath78 at intermediate times , or that @xmath6 has the true value @xmath15 .
we have assumed that @xmath20 .
therefore @xmath18 and @xmath8 are not orthogonal .
so it is not a priori logically inconsistent to assume that @xmath4 and @xmath6 both have true values at intermediate times .
denotes the projector on the subspace spanned by @xmath8 and @xmath18 , then the family of histories @xmath79 is consistent . ]
it is not clear whether a full - fledged interpretation can be implemented along these lines , for arbitrary observables @xmath4 and @xmath6 .
the abl rule is symmetric under permutation of preselection and postselection .
this means that for the purpose of making probabilistic statements about intermediate measurement outcomes , the initial and final states @xmath8 and @xmath18 have exactly the same utility .
this does not necessarily entail that they are equally useful for the purpose of making ontological statements . in von neumann s measurement theory@xcite , for instance ,
a measurement is an interaction between a quantum system and an apparatus , followed by a collapse . from the time @xmath7 when the system is prepared in state @xmath8 to the time @xmath0 when an ontological statement is to be made ,
no physical action occurs on the system .
such is not the case , however , between @xmath0 and @xmath14 . indeed
a physical interaction of the quantum system with an apparatus has to occur sometime before @xmath14 , for the system to collapse to @xmath18 at @xmath14 . in that context
, it may be more natural to hold that @xmath8 , rather than @xmath18 , is the correct state at @xmath0 , and that @xmath80 , rather than @xmath81 , expresses a true value . in some interpretations of quantum mechanics ,
true values of observables can be assigned outside the context of consistent families of histories .
their statistical distributions , however , will not obey the abl rule .
an example is bohmian mechanics , where the true value of position is defined at any intermediate time@xcite .
but in general , @xmath82 does not make up a consistent family . in this context , the meaning of background conditions proposed at the end of the last section is particularly clear .
suppose that @xmath83 and @xmath47 correspond to one - dimensional wave functions @xmath84 and @xmath85 , with @xmath86 a gaussian and @xmath87 a function uniform over some interval @xmath88 and zero elsewhere ( i.e. @xmath47 postselects through a slit of width @xmath88 ) .
all bohmian trajectories going through @xmath88 at @xmath14 have gone through some other interval @xmath89 at @xmath0 .
hence the background conditions associated with preselection @xmath83 and postselection @xmath47 are the wave function at @xmath0 together with the interval @xmath89 of true positions .
of course if position is measured at @xmath0 , the wave function will change accordingly , and so will the measurement result probabilities at @xmath14 .
it was pointed out by vaidman@xcite that no hidden variable theory can reproduce the abl rule in all situations .
indeed take an ensemble of spin 1/2 particles prepared in the state @xmath90 .
if there is no intermediate measurement , postselection in the state @xmath91 will in fact introduce no additional selection .
hence if no backward causality is assumed , true values at @xmath0 must be the same whether postselection does or does not occur , irrespective of any hidden variables .
but then the probability that a measurement at @xmath0 of the spin along @xmath92 yields @xmath93 must be equal to @xmath94 , which differs from the abl value . in the copenhagen interpretation of quantum mechanics ( or at least in the most instrumentalistic versions of it ) , an observable @xmath5 has a value only when a measurement of @xmath5 indicates that value .
this is also the case in mohrhoff s more recent interpretation@xcite , which incorporates the abl rule explicitly .
the fact that the abl rule predicts a statistical distribution of values of @xmath5 implies , according to mohrhoff , an objective fuzziness of @xmath5 in the case where the experiment is not performed .
i believe that statement is genuinely and nontrivially counterfactual .
it asserts that no unmeasured observable of no individual system whatsoever has a true value in the interval between pre- and postselection .
the noncounterfactual meaning of the abl rule is not controversial .
the validity of the rule is then a straightforward consequence of standard quantum mechanics and usual hypotheses on the state of a quantum system immediately after measurement .
the rule is also true counterfactually if it simply expresses the reproducibility of experiments .
a counterfactual meaning more relevant to the specificities of quantum mechanics refers to the true value of an observable at an intermediate time , when no observable is actually measured between pre- and postselection . in general
the abl rule is then counterfactually false .
it can be true , however , when associated with a consistent family of histories or when asserting objective fuzziness or nonvaluedness of observables not being measured .
although statements about true values between pre- and postselection may not have definite empirical meaning , they can fruitfully be viewed as contributing to define the interpretation of the quantum - mechanical theory .
99 y. aharonov , p. g. bergmann and j. l. lebowitz , `` time symmetry in the quantum process of measurement , '' phys .
b * 134 * , 141016 ( 1964 ) .
d. z. albert , y. aharonov and s. damato , `` curious new statistical predictions of quantum mechanics , '' phys . rev .
lett . * 54 * , 57 ( 1985 ) .
w. d. sharp and n. shanks , `` the rise and fall of time - symmetrized quantum mechanics , '' philos .
* 60 * , 48899 ( 1993 ) .
o. cohen , `` pre- and postselected quantum systems , counterfactual measurements , and consistent histories , '' phys .
a * 51 * , 437380 ( 1995 ) .
d. j. miller , `` realism and time symmetry in quantum mechanics , '' phys .
a * 222 * , 3136 ( 1996 ) .
l. vaidman , `` time - symmetrized quantum theory , '' fortschr .
phys . * 46 * , 72939 ( 1998 ) .
l. vaidman , `` time - symmetrized counterfactuals in quantum theory , '' found .
* 29 * , 75565 ( 1999 ) .
l. vaidman , `` defending time - symmetrised quantum counterfactuals , '' stud . hist .
. mod .
phys . * 30 * , 37397 ( 1999 ) .
l. vaidman , `` the meaning of elements of reality and quantum counterfactuals : reply to kastner , '' found .
* 29 * , 86576 ( 1999 ) .
l. vaidman , `` discussion : time - symmetric quantum counterfactuals , '' ` http://philsci-archive.pitt.edu/archive/00001108/ ` ( 2003 ) .
r. e. kastner , `` time - symmetrised quantum theory , counterfactuals and ` advanced action ' , '' stud
. hist .
phys . * 30 * , 23759 ( 1999 ) .
r. e. kastner , `` tsqt ` elements of possibility ' ? '' stud . hist .
* 30 * , 399402 ( 1999 ) .
r. e. kastner , `` the three - box ` paradox ' and other reasons to reject the counterfactual usage of the abl rule , '' found .
* 29 * , 85163 ( 1999 ) .
r. e. kastner , `` the nature of the controversy over time - symmetric quantum counterfactuals , '' philos .
sci . * 70 * , 14563 ( 2003 ) .
n. goodman , `` the problem of counterfactual conditionals , '' j. philos . *
44 * , 11328 ( 1947 ) .
d. lewis , _ counterfactuals _ ( harvard university press , cambridge ma , 1973 ) .
p. horwich , _ asymmetries in time _
( mit press , cambridge , ma , 1987 ) , chap .
l. marchildon , _ quantum mechanics . from basic principles
to numerical methods and applications _ ( springer , berlin , 2002 ) .
j. von neumann , _ mathematical foundations of quantum mechanics _ ( princeton university press , princeton , 1955 ) .
y. aharonov and d. z. albert , `` the issue of retrodiction in bohm s theory , '' in _ quantum implications _ , b. j. hiley and f. d. peat , eds.(routledge and kegan paul , london , 1987 ) .
u. mohrhoff , `` what quantum mechanics is trying to tell us , '' am
. j. phys .
* 68 * , 72845 ( 2000 ) .
u. mohrhoff , `` objective probabilities , quantum counterfactuals , and the abl rule a response to r. e. kastner , '' am .
j. phys .
* 69 * , 86473 ( 2001 ) .
r. e. kastner , `` comment on ` what quantum mechanics is trying to tell us , ' by ulrich mohrhoff , '' am .
j. phys .
* 69 * , 86063 ( 2001 ) . | the aharonov - bergmann - lebowitz rule assigns probabilities to quantum measurement results at time @xmath0 on the condition that the system is prepared in a given way at @xmath1 and found in a given state at @xmath2 .
the question whether the rule can also be applied counterfactually to the case where no measurement is performed at the intermediate time @xmath0 has recently been the subject of controversy .
i argue that the counterfactual meaning may be understood in terms of the true value of an observable at @xmath0 . such a value can not be empirically determined for , by stipulation
, the measurement that would yield it is not performed . nevertheless , it may or may not be well - defined depending on one s proposed interpretation of quantum mechanics .
various examples are discussed illustrating what can be asserted at the intermediate time without running into contradictions . |
the crossover from bcs superconductors to a bose - einstein condensate of tightly bound fermion pairs was first studied many years ago in refs .
@xcite@xcite in a model with @xmath3-function attraction .
this crossover has recently raised renewed interest @xcite- @xcite , especially after the doping dependence of the critical temperature ( @xmath1 ) and non - fermi - liquid behavior above @xmath1 were combined in a phenomenological discussion of high-@xmath1 cuprates @xcite , @xcite . in this paper
we present a detailed study of the crossover based on a nonperturbative procedure in which derivative expansion is used to set up an effective xy - model whose well - known nonperturbative properties render information on the entire crossover regime of the above model . physically , the most important distinctions between conventional ( bcs ) and strong - coupling ( bose - einstein ) regime lies in the fact that in the former only a small fraction of the conduction electrons is paired with the superfluid density involving all pairs , whereas in the latter practically all carriers are paired below a certain temperature @xmath4 , although not condensed .
the temperature has to be lowered further below some critical temperature @xmath5 to make these pairs condensed and establish phase coherence , which leads to superconductive behavior . in the model to be investigated in this paper , the crossover from bcs - type to bose - type superconductivity
will take place either by varying the coupling strength , or by decreasing the carrier density .
since in the bcs theory pair coupling is weak , it can well be described by mean - field theory for the pair fields . in the opposite limit of strong pair binding , on the other hand , superconductivity set in as a macroscopic occupationis of @xmath6 level and we are obliged to go beyond mean field level to describe such an effect . in three dimensions ,
crossover from bcs superconductivity to the bose - einstein condensation of tightly bound fermion pairs was first investigated in ref .
@xcite by summing particle - particle ladder diagrams which correspond to gaussian fluctuations around the mean field , in the functional integral formalism it was studied in ref
. @xcite . in both papers ,
fluctuation corrections were retained in the number equation which was solved together with mean - field gap equation . in these papers , starting from fermionic system , gas of electron pairs was maped in the strong - coupling limit to the ideal bose gas and critical temperature was shown to becomes asymptoticaly temperature of bose - einstein condensation of ideal bose gas of particles with mass @xmath7 and density @xmath8 , where @xmath9 and @xmath10 are fermion mass and density . in this aproximation @xmath1 has an arificial maximum at intermediate coupling strength thus approaching limiting value in the strong - coupling limit from above .
this artifact was removed in the generalized self - consistent green s function numerical approach in @xcite . in this paper
we shall study the properties of collective modes with help of the lowest gradient terms governing the gaussian fluctuations around the mean - field solution .
these are most violent in the phase of the order parameter .
phase transitions in a system with these fluctuations are well understood in two and three dimensions from extensive studies of the xy - model . by setting up an equivalent xy - model we are therefore able to describe well the onset and disappearance of superconductivity in the entire crossover regime . in this way
, we shall obtain simple formulas for the critical temperature @xmath1 as a monotonously increasing function of the coupling strength and carrier density in both two and three dimensions . in the weak - coupling limit
, we give simple explicit formulas which show how the temperature of the @xmath2-model transition converges to the transition temperature in the bcs theory .
the weak- to strong - coupling superconductivity crossover in two dimensions was studied via the kosterlitz - thouless theory in refs .
@xcite , @xcite . in ref .
@xcite it was investigated within the same model as ours at a fixed carrier density , but only numerically and @xmath4 in the weak - coupling region is missing - the effect that we show analyticaly in our paper . ] .
the different properties of size and phase fluctuations was also exploited in ref . @xcite .
the hamiltonian of our model is the typical bcs hamiltonian in @xmath11 dimensions ( @xmath12 ) @xmath13 where @xmath14 is the fermi field operator , @xmath15 denotes the spin components , @xmath9 is the effective fermionic mass , and @xmath16 the strength of an attractive potential @xmath17 .
the mean - field equations for the gap parameter @xmath18 and the chemical potential @xmath19 are obtained in the standard way ( see for example @xcite ) : @xmath20 where the sum runs over all wave vectors @xmath21 , @xmath22 is the total number of fermions , @xmath23 the volume of the system , and @xmath24 are the energies of single - particle excitations .
the @xmath3-function potential produces an artificial divergence and requires regularization .
a bcs superconductor possesses a natural cutoff supplied by the debye frequency @xmath25 . for the crossover problem to be treated here
this is no longer a useful quantity , since in the strong - coupling limit all fermions participate in the interaction , not only those in a thin shell of width @xmath25 around the fermi surface . to be applicable in this regime ,
we renormalize the gap equation in three dimensions with the help of the experimentally observable @xmath26-wave scattering length @xmath27 , for which the low - energy limit of the two - body scattering process gives an equally divergent expression @xcite@xcite : @xmath28 eliminating @xmath29 from ( [ 1.4 ] ) and ( [ 1.1 ] ) we obtain a renormalized gap equation @xmath30 , \label{1.5}\ ] ] in which @xmath31 plays the role of a dimensionless coupling constant which monotonically increases from @xmath32 to @xmath33 as the bare coupling constant @xmath34 runs from small ( bcs limit ) to large values ( be limit ) .
this equation is to be solved simultaneously with ( [ 1.2 ] ) .
these mean - field equations were first analyzed at a fixed carrier density in refs .
@xcite and @xcite . here
we shall first reproduce some of the obtained estimates for @xmath4 and @xmath19 . in the bcs limit
, the chemical potential @xmath35 does not differ much from the fermi energy @xmath36 , whereas with increasing interaction strength , the distribution function @xmath37 broadens and @xmath35 decreases . in the be limit
we have tightly bound pairs and nondegenerate fermions with a large negative chemical potential @xmath38 . in the strong coupling limit eq .
( [ 1.5 ] ) provide us an estimation for @xmath4 - characteristic temperature of the thermal pair breaking @xcite , whereas eq .
( [ 1.2 ] ) determines @xmath19 . from eq .
( [ 1.2 ] ) we obtain that in the be limit @xmath39 , where @xmath40 is the binding energy of the bound pairs . in the be limit , we can estimate that the pseudogap sets in at @xmath41 . a simple `` chemical '' equilibrium estimate @xmath42 yields for the temperature of pair dissociation : @xmath43 which shows at strong couplings @xmath4 is indeed related to pair formation @xcite , @xcite ( which in the strong - coupling regime lies above the temperature of the onset of phase coherence @xcite@xcite ) .
the gap in the spectrum of single - particle excitations has a special feature @xcite , @xcite , @xcite when the chemical potential changes its sign .
the sign change occurs at the minimum of the bogoliubov quasiparticle energy @xmath44 where this energy defines the gap energy in the quasiparticle spectrum : @xmath45 thus , for positive chemical potential , the gap energy is given directly by the gap function @xmath46 , whereas for negative chemical potential , it is larger than that : @xmath47 in two dimensions , a nonzero bound state energy @xmath48 exists for any coupling strength .
the cutoff can therefore be eliminated by subtracting from the two - dimensional zero - temperature gap equation @xcite-@xcite @xmath49 where @xmath50 , @xmath51 the bound - state equation @xmath52 after performing the elementary integrals , one finds : @xmath53 in the next sectioins we will work at finite tempetrature , in doing so , we do not fix the carrier density but assume the presence of a reservoir which provides us with a temperature - independent chemical potential @xmath54 .
such a fixed @xmath19 will be most convenient for deriving simple analytic results for the finite - temperature behavior of the system . in this fixed @xmath19 model carrier density
becomes temperature dependent . for experimental measurements of @xmath19
in this section we make use of derivative expansion to determine the relevant stiffness parameter for the study of phase fluctuations , which in two dimensions determines the temperature of the kosterlitz - thouless transition . in two - dimensional system ,
the phase fluctuations are most violent causing the strongest modifications of the mean - field properties .
the coleman - mermin - wagner - hohenberg theorem @xcite forbids the existence of a strict long - range order , but there is quasi - long - range order manifesting itself in a power behavior of the correlation functions at all temperatures below @xmath55 .
we shall neglect the coupling to the magnetic vector potential throughout the upcoming discussion , so that the phase coherence below @xmath1 can be of long or quasi - long range , unspoiled by the meissner effect which would reduce the range to a finite penetration depth .
the effective hamiltonian from which we deduce the stiffness of the phase fluctuations was derived in @xcite , @xcite . in this section
we summarize a few important aspects of it , with a reminder of its derivation given below . writing the spacetime - dependent order parameter as @xmath56 , where @xmath57 denotes the four - vector @xmath58 formed from imaginary time and position vector , the partition function
may be written as a functional integral @xcite @xmath59 } , \label{bk1}\ ] ] where @xmath60 is the one - loop effective action , containing the inverse green function of the fermions in the collective pair field @xmath61 + \hat{i } \left[\frac{i \nabla^{2 } \theta(\tau , \mbox{\bf x})}{4 m } + \frac{i \nabla \theta(\tau , \mbox{\bf x } ) \nabla}{2 m } \right ] .
\label{bk3}\end{aligned}\ ] ] here @xmath62 are the usual pauli matrices , and @xmath63 is added for regularization .
let us now assume that phase gradients are small .
then @xmath64 can be approximated as follows : @xmath65 with the kinetic " term ( see @xcite ) @xmath66 and the potential " term @xmath67 the latter coincides with the mean - field energy , determining the modulus of @xmath68 and thus the stiffness of phase fluctuations .
the kinetic part @xmath69 contains gradient terms whose size is determined by the modulus of @xmath68 .
given the stiffness , one may immediately set up an equivalent xy - model . both @xmath69 and @xmath70 are expressed in terms of the green function of the fermions , which solves the equation @xmath71 { \cal g}(\tau , \mbox{\bf x } ) = \delta(\tau ) \delta(\mbox{\bf x } ) \label{bkt6}\ ] ] and @xmath72 - \hat{i } \left[\frac{i \nabla^{2 } \theta}{4 m } + \frac{i \nabla \theta(\tau , \mbox{\bf x } ) \nabla}{2 m } \right ] .
\label{bkt7}\ ] ] the gradient expansion that we use to determine stiffness was first made in ref .
@xcite at zero temperature . in ref .
@xcite , the kinetic term @xmath69 was calculated in two dimensions at finite temperature for arbitrary chemical potential retaining terms with @xmath73 in the expansion ( [ bkt5 ] ) .
the result is @xmath74 , \label{bk8}\ ] ] where @xmath75 is the stiffness coefficient whose explicit form is @xmath76 the other coefficents are : @xmath77 and @xmath78 is the density of fermions ( [ 1.2 ] ) which varies with temperature in our model . at the temperature @xmath4 where the modulus of @xmath18 vanishes , also the stiffness disappears .
we are now ready to specify the effective xy - model governing the phase fluctuations .
the model hamiltonian corresponding to the gradient term in @xmath79 is @xcite , @xcite : @xmath80 ^ 2.\end{aligned}\ ] ] in contrast to the standard xy - model , the stiffness parameter is not a constant but depends on temperature via the solution of gap and number equations ( [ 1.1 ] ) and ( [ 1.2 ] ) .
clearly , kosterlitz - thouless transition always take place below @xmath4 . for vortices of a high fugacity
, the temperature of the phase transition is determined by the well - known formula @xcite @xmath81 which follows from the divergence of the average square size of a vortex - antivortex pair .
since these attract each other by a coulomb potential @xmath82 , the average square distance is @xmath83 which diverges indeed at the temperature ( [ kt ] ) . in our case
@xmath55 should be determined self - consistently : @xmath84 from ( [ bk9 ] ) , ( [ 1.2 ] ) and ( [ kt ] ) it is easily seen that @xmath85 indeed tends to zero when the pair attraction vanishes in which case @xmath86 . in general , the behavior of @xmath85 for strong and weak couplings is found by the following considerations .
we observe that the particle number @xmath10 does not vary appreciably in these limits with temperature in the range @xmath87 , so that weak - coupling estimates for @xmath85 derived within the model with temperature - independent chemical potential ( i.e. , when the system is coupled to a large reservoir ) practically coincide with those derived from a fixed fermion density .
further it is immediately realized that in the weak - coupling limit @xmath88 is a small parameter . at zero coupling ,
the stiffness @xmath89 vanishes identically , such that an estimate of @xmath90 at weak couplings requires calculating a lowest - order correction to the second term of eq.([bk9 ] ) proportional to @xmath91 .
thus weak - coupling expression for stiffness reads : @xmath92 equating this with the stiffness in ( [ @jint ] ) we obtain the weak - coupling equation for @xmath85 : @xmath93 where @xmath94 is the fermi energy of free fermions .
it is useful to introduce reduced dimensionless temperatures @xmath95 and @xmath96 which are small in the weak - coupling limit
. then we rewrite eq .
( [ e2 ] ) as @xmath97 for small @xmath98 we may expand @xmath99 this equation shows explicitly how for decreasing coupling strength @xmath85 merges with @xmath4 . for weak coupling strengths , @xmath85 behaves like @xmath100 the merging of the two temperatures in the weak - coupling regime is displayed in fig .
[ f7 ] . 2d.tps + consider now the opposite limit of strong couplings . there
we see from eqs .
( [ e1 ] ) , ( [ 1.1 ] ) ( [ 1.2 ] ) , and ( [ bk9 ] ) for @xmath85 , @xmath101 , and @xmath102 that @xmath85 tends to a constant value .
we can observe that in the strong - coupling limit @xmath103 is always situated close to the zero - temperature value of @xmath104 . taking this into the account we derive an estimate for the second term in ( [ bk9 ] ) , thus obtaining the strong - coupling equation for @xmath85 : @xmath105 \right\ } \label{e30}\end{aligned}\ ] ] with the approximation @xmath106
we find that the first term in the exponent tends in the strong - coupling limit to a constant , @xmath107 , whereas the first term in the brackets tends to @xmath108 , so that eq .
( [ e30 ] ) has the limiting form @xmath109 \right\}. \label{e32}\end{aligned}\ ] ] thus for increasing coupling strength , the phase - decoherence temperature @xmath85 tends very quickly towards a constant : @xmath110 in this limit we know from eq .
( [ 1.2 ] ) that the difference in the carrier density at zero temperature , @xmath111 , becomes equal to @xmath112 , so that our limiting result coincides with that obtained in the `` fixed carrier density model '' : @xmath113 where we have inserted @xmath114 for the fermi energy of free fermions at the carrier density @xmath115 .
from the above asymtotical formulas for weak- and strong - coupling limits we see that the temperature of the kosterlitz - thouless transition is a monotonous function of coupling strength and carrier density .
the crossover takes place in a very narrow region @xmath116 .
it is also observed in the behavior of three - dimensional condensation temperature @xmath1 of the gas of tightly bound , almost free composite bosons . in refs .
@xcite , @xcite which include only quadratic fluctuations around the mean field ( corresponding to ladder diagrams ) , @xmath1 was shown to tend to a constant free bose gas value @xmath117^{2/3 } \pi /m$ ] , with no dependence on the internal structure of the boson . here
we find a similar result in two dimensions , where @xmath55 tends to a constant depending only on the mass @xmath7 and the density @xmath8 of the pairs .
no dependence on the coupling strength is left .
the only difference with respect to the three - dimensional case is that here the transition temperature @xmath118 is linear in the carrier density @xmath10 , while growing like @xmath119 in three dimensions .
our limiting result ( [ e4 ] ) agrees with ref .
@xcite and @xcite .
there exists a corresponding equation for the temperature @xmath4 in the strong - coupling limit @xmath120 : @xmath121
in this section we discuss in a completely analogous way the fluctuations in three dimensions , where stiffness is , for small temperatures where @xmath122 is close to @xmath123 : @xmath124 governing the phase fluctuations via an effective xy - model @xmath125 ^ 2.\ ] ] the temperature of the phase transition in this model can reasonably be estimated using mean - field methods for the lattice 3d xy - model @xcite : @xmath126 @xmath127 is the lattice spacing of the theory @xcite where @xmath128 is number of pairs . in the weak - coupling limit
, the stiffness coefficent may be derived with the help of gorkov s well - known method ( setting @xmath129 ) : @xmath130 this is precisely the coefficient of the gradient term in the ginzburg - landau expansion , in the weak - coupling limit the two temperatures merge according to the formula : @xmath131 which contains a larger power of @xmath132 in the second term as well as a smaller prefactor @xmath133 , as compared with the two - dimensional separation formula ( [ s01 ] ) . the merging behavior is displayed in fig .
( [ 3d ] ) . 3d.tps + in the strong - coupling limit of the theory where we have tightly bound composite bosons , the phase stiffness tends asymptotically to : @xmath134 , \label{@stiff@}\ ] ] it obviously tends in this limit quickly to @xmath135 an estimate for the critical temperature , obtained via the mean - field treatment of the 3d xy - model on the lattice reads in this limit : @xmath136 \label{chaoslabs}\ ] ] this quickly tends from below to the value : @xmath137 this result is very close to the temperature of the condensation of bosons of mass @xmath7 and density @xmath8 , which , as it was discussed in the introduction was obtained including the effect of gaussian fluctuations into the mean - field equation for the particle number @xcite yielding - model approach . ]
@xmath138^{2/3 } \pi / m = 0.218 \epsilon_f .
\label{@bosons}\ ] ]
we have studied the crossover from bcs - type to bose - type superconductivity . for this purpose
we have used the gradient expansion of the effective energy functional to set up an equivalent xy - model which allows us to investigate the onset of long - range order in the phase fluctuations . in two dimensions ,
we have given a simple analytic expression which shows how the resulting kosterlitz - thouless temperature @xmath85 at which quasi - long - range order sets in moves towards the pair - binding temperature @xmath4 , and merges with it in the weak - coupling limit .
a similar expression was found in three dimensions . in the strong coupling limit
we find that critical temperature tends in both two and three dimensions to a constant value as the chemical potential changes it sign .
let us finally remark that the separation of @xmath4 and @xmath1 has an analogy in the ferroelectrics and magnets which also contain two separate characteristic temperatures , for example in the latter case the stoner- and the curie - temperature .
it also can be studied more precisely in a simple field theoretic model in 2+@xmath139 dimensions with an o(@xmath10 ) symmetry for large @xmath10 .
in such a model , the existence of two small paramters @xmath139 and @xmath140 has permitted us recently to _ prove _ the existence of two transitions , and to exhibit clearly their different physical origins @xcite .
we thank profs .
k. bennemann , k. maki , and v.emery for explaining to us some aspects of @xmath4 crossover in superconductive cuprates and to proffessor y. uemura for useful points .
one of us ( e.b . ) is grateful to all members of prof .
kleinert s group at the institut fr theoretische physik freie universitt berlin for their kind hospitality , and dr .
sharapov for useful discussions . + + * note added in proof * : + here and in our earlier preprint cond - mat/9804206 we criticize the discussion of the weak - coupling behavior of superconductive phase transition in two dimensions given in ref .
however , after our paper went to print in phys .
b * 59 * , 12083 ( 1999 ) ) , the authors of @xcite updated their discussion in their preprint cond - mat/9709034 version 2 , which appeared in jetp * 88 * , 685 ( 1999 ) , and our criticism no longer applies .
uemura , in _ proceedings of the workshop in polarons and bipolarons in high-@xmath141 superconductors and related materials , cambridge , 1994 _ ed . by e. salje _
( cambridge univ .
press , 1995 ) pp . 453 .
uemura , in _ proceedings of the ccast symposium on high-@xmath141 superconductivity and the c@xmath142 family , beijing , 1994 _ ed . by s. feng and h.c .
ren ( gordon and breach , new york , 1995 ) pp . 113 .
gusynin , v.m .
loktev , s.g .
sharapov preprint cond - mat/9709034 ; jetp lett .
* 65 * 182 ( 1997 ) ; low . temp . phys .
* 23 * 612 ( 1997 ) + v.m .
loktev , s.g .
sharapov ; cond .
( lviv ) * 11 * 131 ( 1997 ) ; preprint cond - mat/9706285 .
+ v.m.loktev,r.m.quick , s.g .
sharapov preprint cond - mat/9804026 .
r.micnas and t.kostyrko in `` recent developments in high temperature superconductivity '' proceedings af the first polish - us conference .
wroclaw 1995 .
+ r.micnas _ et .
b * 52 * 16223 ( 1995 ) . | for an electron gas with @xmath0-function attraction we investigate the crossover from weak- to strong - coupling supercoductivity in two and three dimensions .
we derive analytic expressions for the stiffness of phase fluctuations and set up effective xy - models which serve to determine nonperturbatively the temperature of phase decoherence where superconductivity breaks down .
we find the transition temperature @xmath1 as a monotonous function of the coupling strength and carrier density both in two and three dimensions , and give analytic formulas for the merging of the temperature of phase decoherence with the temperature of pair formation in the weak - coupling limit .
# 1 # 1 # 1 * # 1 # 1 # 1 0.8 cm # 1#2#1/#2 0.001 in journal ref . :
phys . rev .
* b59 * 12083 ( 1999 ) * nonperturbative @xmath2-model approach to strong coupling superconductivity in two and three dimensions . * * egor babaev * 0.3 cm _
institut fr theoretische physik + freie universitt berlin , arnimallee 14 , 1000 berlin 33 , germany _ _ and institute for theoretical physics , uppsala university + box 803 , s-75108 uppsala , sweden _ * hagen kleinert * 0.3 cm _
institut fr theoretische physik + freie universitt berlin , arnimallee 14 , 1000 berlin 33 , germany _ 0.5 cm 0.15 in 0.6 cm |
the _ loop approach _ to quantum gravity @xcite has reached a mature status as a physical theory and it is one of the most likely candidate to be a complete consistent quantum theory of gravitational phenomena ( for a detailed review and update bibliography we refer to @xcite for quantum gravity in general and to @xcite for the loop approach in particular ) . in this approach it was possible : _
i ) _ to construct the auxiliary hilbert space @xcite appropriate to the discussion of the kinematics _ ii ) _ to find an explicit basis ( the spin - networks basis ) on this hilbert space @xcite ; _ iii _ ) to rigorously solve the 3-diffeomorphism constraints @xcite ; _ iv ) _ to construct operators corresponding to volume and area and determine their spectrum @xcite . _ v _ ) to derive the beckenstein - hawking formula for the entropy of black holes @xcite the elusive feature of the loop approach to quantum gravity is that in the _ frozen - time _ canonical picture the relation between operator and physical observable is cumbersome .
a space time picture was lacking until m. reisenberger a c. rovelli @xcite derived a spin foam model ( using the terminology introduced by j. baez @xcite ) , to which they refer as _ world sheet model _
, from the exponential of the super - hamiltonian constraint .
they also show that this is formally related to the construction of the 4 dimensional path integral for gravity .
this work was a first step in the direction of developing a sum over surfaces framework in `` loop quantum gravity '' whoose importance was first claimed by reisenberger in @xcite .
this recent developments showed surprising formal analogy between loop quantum gravity and other approaches to the quantization of gravity .
the model of dynamical triangulation @xcite and the ponzano - regge model @xcite are exactly models of this kind .
the analogy is even deeper if one considers the ponzano - regge model [ the physicist version of the celebrated turaev - viro state sum model for 3-manifold invariant @xcite ) ] that can be formulated exactly as a sum over branched surfaces with colored graph observables ( spin networks ) on the boundary . the previous consideration , together with the fact that the topological bf field theory in 4 dim @xcite , @xcite ( see also @xcite ) admits an analogous formulation as a spin foam model [ in this case , the spin foam dual to a simplicial decomposition ( triangulation ) ] , strongly suggest that quantum gravity should be naturally formulated using the language of spin foam models .
now , two explicit proposals for the realization of this idea have been put forward .
the first one
( by barret and crane @xcite ) is related to the discretization of the path integral of the euclidean plebanski action @xcite ( see @xcite ) and the second ( by m. reisenberger @xcite ) of self dual euclidean gravity @xcite .
the last model is the explicit quantization of the lattice model @xcite that has self dual euclidean gravity as its continuum limit .
this research line is deeply connected to the idea of express classical gravity as a deformation ( by the imposition of additional lagrange constraints ) of a topological bf field theory 4 dimension .
the idea is that @xcite topological bf theory , in the context of diff - invariant theory , can play a role analogue to the one of free theory in standard poincar invariant quantum fields theories .
this idea is at the origin of the systematic proposal for deriving _
spin foam weight factors _ from a classical action principle that has been recently proposed by freidel and krasnov @xcite .
the discussion here it is in framework of piecewise linear topology and we refer the reader to @xcite for details . intuitively , this setting is equivalent to assume that cells , surfaces and edges that we use to decompose the manifold are finite unions of piecewise analytic sub - manifolds . in particular , a graph is a finite collection of analytic path .
this framework as certain advantages and the resulting theory is simpler than the ones in the analytic or in the smooth setting .
the utility of the pl setting in loop quantum gravity was first pointed out by zapata @xcite . all the topological objects ( graph , manifold , cellular decomposition , .... ) are assumed orientable . for the wigner nj - symbols we adopt the non normalized graphical convention of @xcite and we restrict to the case of the group @xmath11 and use penrose s binor formalism @xcite .
most of the formulas remain valid in the quantum group case , but , in this case , factors related to the braiding can appear .
loop quantum gravity is based on the hamiltonian formulation of classical general relativity in terms of the canonically conjugated variables @xmath12 ( densitized triad fields ) and @xmath13 ( ashtekar s connection ) @xcite .
the dynamics is given , as it happens for any diff - invariant theory , only in terms of constraints . in our case : @xmath14 the loop program does the quantization of gravity in a straightforward way .
it is based on the assumption that a particular class of classical functions ( the @xmath15 observables , i.e. , the _ traces _ of the _ holonomy _ of ashtekar s connection ) become well defined quantum operators .
i.e. , it is assumed that , for any choice of the close loop @xmath16 , to the classical functional of the ashtekar connection : @xmath17 = { \rm tr}\left [ \exp\left ( \int_\gamma a_a(x ) dx^a\right ) \right]\ ] ] corresponds a well defined quantum operator on a hilbert space @xmath18 .
all the other properties should be essentialy derived from this assumption .
a consequence of this choice is that the canonical variables @xmath12,@xmath13 do not become well defined operator on the hilbert space @xmath18 and the previous definition of the constraint ( unless properly regularized ) are meaningless on @xmath18 .
this is one of the main technical reasons behind the difficulty of dealing with the super - hamiltonian constraint .
given this assumption , there are two possible ways to define the theory . explicitly construct the vector space on which these operators are well defined ( loop representations , see @xcite ) , in a algebraic way , and then determine the correct scalar product . alternatively , construct a hilbert space @xmath18 ( the connection representation ) and show that , on this hilbert space , these operators are well defined @xcite .
these two approaches are equivalent @xcite .
however , two technical but important decision must be taken .
the first one is connected to the properties of the loops @xmath16 .
the request of piecewise analyticity brings to a different setting with respect to the smooth case since two piecewise analytic loop can have only a finite number of isolated intersection points .
the second is that these constraction are well defined only in the case of a real connection .
now , the standard ashtekar connection is naturaly real only in the case of euclidean gravity while in the case of lorentzian gravity is naturally complex and indeed what follows naturally applies only to euclidean gravity . however , it is important to note that , it is possible to use a real connection in the lorentzian case too ( the barbero connection @xcite ) at the prize of a more complicated form for the super - hamiltonian constraint .
a full discussion of the kinematical setting is not important here .
we refer to @xcite for the technical details and for complete references . for the following discusson it is sufficient to recall the basic structure of the hibert space @xmath18 .
since a connection associates a group element @xmath19 to each edge ( segment embedded in @xmath20 ) of the graph @xmath16 .
it is natural to consider for each graph @xmath16 the hilbert space ( @xmath21 $ ] ) of square integrable function , with respect to the unique normalized haar measure @xmath22 , on @xmath23 ( the number of edges @xmath24 in @xmath16 ) copies of the group [ @xmath25 @xmath26 .
@xmath18 can be characterized as the projective limit of the previous family of hilbert spaces @xmath27 associated to all possible graphs @xmath16 embedded into @xmath20 .
a rigorous definition of the diffeomorphism invariant hilbert space can be obtained @xcite using the gauge averaging procedure .
the essential idea is that on @xmath39 there is a natural definition of the action of a diffeomorphism @xmath40 ( @xmath41 ) .
in fact , in the spin - networks basis we can associated to each diffeomorphism @xmath42 the operator @xmath43 where @xmath44 is the imagine under @xmath42 of the spin network @xmath45 . at this point
one can define the class of spin networks knots @xmath46 as the equivalence class of spin networks under diffeomorphism , i.e. @xmath47 it it exists a @xmath40 such that @xmath44 . in this way
, we can define @xmath48 with the scalar product : @xmath49 \langle s|\hat{u}_\phi s'\rangle = n \int [ d\phi ] \langle \phi(s)| s'\rangle \label{sp : diff}\ ] ] where @xmath50 is an arbitrary , not yet fixed , normalization factor and @xmath51 are two arbitrary spin networks in the @xmath46-knot class of @xmath46 and @xmath52 , respectively .
the integration in eq .
( [ sp : diff ] ) is meaningful because the scalar product of two spin networks of @xmath39 is different from zero only if they have support on the same graph . in the previous derivation of the hilbert space @xmath53
we made an implicit assumption on the kind of diffeomorphism that are allowed .
having previously assumed that the loops are in the class of the piecewise analytics ones , it is natural to assume piecewise analytics diffeomorphism . in this case , the quotient space @xmath53 is a separable hilbert space .
this leads directly to the pl framework .
we refer to the work of zapata for the discussion of the subtleties involved @xcite . in the first presentation of _ loop quantum gravity _
@xcite was emphasized that loops without intersection , ( i.e. , spin network states whose graph @xmath16 with only bivalent vertices ) it is automatically in the kernel of the super - hamiltonian constraint .
this gives the knowledge of an infinite dimensional subspace of the * physical * hilbert space @xmath54 .
unfortunately , with the introduction of the volume operator @xcite , it was realized that all the spin network states based on graph @xmath16 with all vertices of valence less that @xmath55 are in the kernel of the volume operator .
consequently , all the previously mentioned states , belonging to the kernel of the super - hamiltonian constraint , correspond to degenerate metrics and , therefore , can not correspond to genuine quantum 3 geometries .
this urged the definition of the super - hamiltonian constraint @xmath56 on the whole diff - invariant hilbert space @xmath53 .
the situation improved in 1996 when thiemann @xcite was able to construct an operator with the _ naiv _ classical limit . using this definition , it is possibile investigate the true * physical * hilbert space @xmath54 . for the purpose of the present discussion
we do not need a complete account of its definition and its properties . what it is interesting for us it is that : 1 ) it is possible to define such operator ; 2 ) a rough idea of its action @xmath53 .
its main property is that it acts locally around each vertex of the spin network .
it adds or deletes an edge of color 1 @xmath57 \left| \begin{array}{c}\setlength{\unitlength}{.8 pt } \begin{picture}(65,50 ) \put ( 0 , 0){\line ( 1 , 1){20 } } \put ( 0,20){\line ( 1 , 0){20 } } \put ( 0,40){\line ( 1 , -1){20 } } \put(20,20){\circle*{3 } } \put(20,20){\line ( 3 , 2){30 } } \put(52,38){$\scriptstyle k$ } \put(20,20){\line ( 3 , -2){30 } } \put(52 , 2){$\scriptstyle j$ } \put(45 , 8){$\scriptstyle p$ } \put(45,30){$\scriptstyle q$ } \end{picture}\end{array } \right\rangle\ ] ] with well definite weight factors [ ( @xmath58 ) matrix elements of the operator ] that can be explicitly computed @xcite .
summarizing , the mathematical definition of the loop quantization of gravity , and the determination of the * physical * hilbert space @xmath54 , involves the analysis of the following hierarchy of hilbert spaces : @xmath59 where an arrow means imposition of a constraint and the questions mark emphasis the fact that , the last step , it is still not completely understood .
in fact : 1 ) the physical correctness of the thiemann s proposal for the regularization of the super - hamiltonian constraint has been questioned ; 2 ) more of one variant of this operator can be construct and indeed the problem of determine the correct one ; 3 ) despite @xcite there is not a clear understanding of the kernel of the super - hamiltonian constraint nor of the class of operator that are well defined on this hilbert space .
in the previous equation , @xmath18 is a diff - invariant definition of the hamiltonian constraints and @xmath75 is a 2-dimensional branched surface embedded over @xmath76 $ ] . in figure
[ fig : spine ] we reported an example of branched surface , and of the corresponding spin - network transition .
the kind of surfaces that appear on the sum is explicitly determined by the kind of non null matrix elements of the super - hamiltonian constraints .
we want to emphasize , once more , that this derivation is based on formal argument .
the use of formal argument is of great help to understand the general picture behind but one should stop here .
the result of this formal derivation is that the functional integral associate to quantum gravity is defined ( or , better , should be ) as a sum over branched surface interpolating between two spin networks .
moreover , it shows that one has to consider the sum over all the possible ( not just one ) such branched surfaces .
this in a deep breakthrough toward the understanding of loop quantum gravity and it shows a strict relationship between loop quantum gravity and topological field theories of bf type opening the possibility of use technics developed for such theories . the first occurrence , ( to our knowledge ) of the idea of formulating the transition amplitude between two hilbert space ( spanned by embedded colored graph ) associated to two boundary component of a manifolds , as a weight associated to the colored branched surface interpolending between the two graphs , first appeared in turaev and viro @xcite .
these authors used the terms : two dimensional polyhedral quantum field theory ( _ spin foam _ models are model of this kind ) .
the next section will be dedicated to the exposition of this idea in the case of 2 + 1 dimensional gravity or , more precisely , of the ponzano - regge - turaev - viro model .
the receipt for the construction of this model ( to which we will refer to as the prtv - model ) is the following .
color all the edge ( 1-dimensional object of the triangulation ) with irreducible representations of the @xmath11 group and assign the following factor to the various object of the triangulation : 1 ) to each edge the dimension of the representation @xmath81 ; 2 ) to each triangle the inverse of the 3j symbol made out by the three colors associated to each of its edges @xmath82 ; 3 ) to each tetrahedron the associated 6j symbol ( a tetrahedron has 6 edges ) @xmath83 $ ] . in the case of the prtv - model , it is useful to consider its dual formulation , i.e. to take advantage of the duality between triangulation and standard polyhedron suggested by figure [ duality ] . in the dual cellular complex @xmath84 of the triangulation @xmath77
we have that to each @xmath23-dimensional cell of @xmath77 is uniquely assigned a 3-@xmath23-dimensional cell of @xmath84 . in three dimension , the objects dual to an edge ( 1 dimensional object ) is a surface ( 2=3 - 1 dimensional object ) .
this implies that to each coloration of the edges of the triangulation @xmath77 is associated a unique coloration of the face ( two dimensional objects ) of its dual standard polyhedron @xmath84 . in turns , it is possible to consider the prtv - model as a model defined over a two dimensional branched surface .
i.e. , using this duality , we can interpret the prtv model both as a state sum model associated to a triangulation but also has a state sum model associated to a branched surface : @xmath85 = tv_g[{\cal s}_2^{\delta^ * } ] \label{shadowsum}\ ] ] where by @xmath86 we denote the collection of all the faces , edges and vertices ( the 2-scheleton ) of the cellular decomposition of @xmath78 dual to triangulation @xmath77 .
we will also use the notations @xmath87 , @xmath88 and @xmath89 to denote the collection of its faces , edges and vertices , respectively .
this express the fact that the definition of the prtv - model can be given as a series of rules with respect to the objects of the triangulation or as an equivalent set of rules with respect to the objects to its dual standard polyhedron : [ cols="<,<,<,^,<",options="header " , ] one of the nice characteristic of the dual formulation is that it involves only the coloring of a 2-dimensional standard polyhedron embedded into the manifold . completing the transition to the dual world it is possible to to forget the original triangulation and to directly associate a state - sum to any standard polyhedron of the ( @xmath20 ) manifold .
@xmath90 = \lambda^{\scriptscriptstyle -\#(3-cells ) } \!\!\!\!\ !
\sum_{\vec{c}\in d({\cal s}_2,g ) } \left[\!\prod_{f\in { \cal f}({\cal s}_2)}\ !
\left(\delta{c_f}\right)^{\chi(f ) } \right ] \left[\!\prod_{e\in { \cal e}({\cal s}_2)}\ !
\left({\theta{a_e}{b_e}{c_e}}\right)^{-\chi(e ) } \right ] \left[\!\prod_{v\in { \cal v}({\cal s}_2)}\ !
\tet{a_v}{b_v}{f_v}{d_v}{e_v}{c_v } \right ] \label{def : tv1}\ ] ] where @xmath91 and @xmath92 are the euler characteristics of the face @xmath93 and of the edge @xmath94 , respectively . for the purpose of the present discussion the introduction of the euler characteristics is completely irrelevant .
all the edges and faces of a standard 2-polyhedron dual to a triangulation of @xmath20 have euler - characteristic equal to 1 !
we explicitly wrote the euler characteristic because one of the interesting characteristic of the prtv partition function is that is is defined for any stratification of the manifolds @xmath78 whose two skeleton is a standard polyhedron .
that means that it is defined on the larger class of singular triangulations @xcite . for the rest of the discussion we will always deal with regular polyhedron dual to triangulation and
we will always replace the euler characteristic with the appropriate value for polyhedrons belonging to the dual of a triangulation : 1 .
this model manifest its topological nature from the fact that it is possible to prove that @xmath95 $ ] it is independ by the chosen standard polyhedron dual to a triangulation ( pseudo - triangulation ) of @xmath78 .
infact , it is possible to prove that , if @xmath96 and @xmath97 are two different standard polyhedron dual to triangulations of the same manifold @xmath78 , then : @xmath98=tv_{g}[{\cal s}_2]=tv_{g}[\tilde{\cal s}_2 ] \label{tnature}\ ] ] in the previous section we used the notation @xmath99 to denote all the possible decorations ( coloring ) of the 2-surfaces of a standard polyhedron with inequivalent irreducible representations of a group @xmath33 . the admissibility condition amount to the existence of an intertwining matrix ( in the case of the @xmath11 group , if it exists , it is unique up to normalization ) between the three representation associated to the three faces that meet at an edge ( a type ii object of figure [ duality ] ) .
it is natural to extend this concept of decoration to the more general situations where there are
n - faces that meet at a given edge . in this case
, we have to assign to each edge an invariant contractor on the tensor product of the spaces of all the representations associated to the faces that meet at the edge . in this way
, we have a decoration of the two- and one - imensional objects of the 2-polyhedron completly analogous to the decoration ( coloring ) of the edges and vertices of a graph @xmath16 that we used to label the spin networks basis of equation ( [ eq : defspinnet ] ) that , using the notation @xmath99 , can be rewrite as : @xmath100 .
\label{eq : decspinnet}\ ] ] this completes the definition of the concept of a group decorated 2 dimensional oriented polyhedron .
this extension to non simple polyhedra of this concept ( implicetly contained in @xcite ) of group decoratation of 2-polyhedron with boundary is due to baez .
he used , the now generaly accepted , term _ spin foam _ to denote a single group decorated 2-polyhedron and the term _ spin foam model _ to denote a theory based on the sum on all the possible decoration ( coloration ) .
we refer the reader to the baez work @xcite for more details and the discussion of the additional care needed on the case of a general group .
following turaev and viro @xcite , we note that it is possibile to generalize the prtv - model to include the case of a manifold @xmath78 with boundary with a decorated ( colored ) graph sitting on its boundary @xmath101 . the prtv partition function ( [ def : tv1 ] ) in the case of a decorated polyhedron @xmath96 with a graph @xmath16 on the boundary ( i.e. , @xmath102 ) is defined as @xmath103 \!\ !
& = & \frac{1}{n(\gamma,\vec{c } ) } \lambda^{\frac{\#(2-cells~of~\gamma)}{2 } } \lambda^{\scriptscriptstyle -\#(3-cells ) } \times \label{def : tv2}\\ & & ~~\times \sum_{\vec{c}\in d_{in}{({\cal s}_2,g ) } } \bigg[\prod_{f\in { \cal f}_{in}({\cal s}_2 ) } \!\!\ ! \left(\delta{c_f}\right ) \bigg ] \bigg[\prod_{e\in { \cal e}_{in}({\cal s}_2 ) } \!\!\ !
\left({\theta{a_e}{b_e}{c_e}}\right)^{-1 } \bigg ] \bigg[\prod_{v\in { \cal v}({\cal s}_2 ) } \!\!\ !
\tet{a_v}{b_v}{f_v}{d_v}{e_v}{c_v } \bigg ] \nonumber \\ \frac{1}{n(\gamma,\vec{c } ) } & = & \sqrt { \prod_{e\in { \cal e}(\gamma)}\delta{c_e } \prod_{v\in { \cal v}(\gamma ) } \left ( \theta{a_v}{b_v}{c_v}\right)^{-1 } } \label{snnorm}\end{aligned}\ ] ] where with @xmath104 and @xmath105 we have denoted the faces and edges of @xmath96 that do not end or belong to the boundary of @xmath96 .
moreover @xmath106 denotes a decoration of the internal faces @xmath104 compatible with the decoration of the graph @xmath16 on the boundary ( @xmath107 if @xmath108 , @xmath109 and @xmath110 ) . .
using the previous generalization of the prtv - model we can reinterpred the state sum model of equation ( [ shadowsum ] ) in the following way . as suggest by figure [ 3pieces ]
we can imagine to split the manifold @xmath78 in three disjoint components glued together along the common boundary and rewrite the partition function associate to any given triangulation as : @xmath111 & = & \sum_{\vec{c}'\in d(\gamma , g ) } ~ \sum_{\vec{c}\in d(\gamma , g ) } tv_g[{\cal s}_2^{\delta_1^*};(\gamma',c ' ) ] \frac{1}{n(\gamma',c')^2 } \nonumber \\ & & \qquad tv_g[{\cal s}_2^{\delta_2^*};(\gamma'\cup\gamma , c+c ' ) ] \frac{1}{n(\gamma , c)^2 } tv_g[{\cal s}_2^{\delta_3^*};(\gamma , c ) ] \label{split } \end{aligned}\ ] ] where @xmath112 and @xmath113 are the graphs determined by the intersection of the 2 bone of the dual cellular decomposition @xmath84 with @xmath114 and @xmath115 , respectively .
now , getting the intersection of @xmath84 with a 2 surface @xmath116 we will have that the 3 cells are mapped into 2 cells , the 2 cells are mapped into 1-cells ( the edge of @xmath16 ) and the 1 cells into vertices of @xmath16 . in this operation
we are indeed inducing a coloring of @xmath16 from the coloring of the branched polyhedron .
moreover , if @xmath77 is a triangulation of @xmath78 then the resulting graph is trivalent .
equation ( [ split ] ) has a nice interpretation in terms of an operator equation .
consider the hilbert spaces @xmath117 of section 2.1 associated to graphs embedded on the two cauchy surfaces @xmath114 ( @xmath118 ) and @xmath115 ( @xmath119 ) . in the normalization of @xcite and
@xcite the completeness of @xmath117 is given by @xmath120 moreover , we can define an operator @xmath121 @xmath122 \langle \gamma',c ' |\hat{p}_{\gamma',\gamma}^{m_2 } |\gamma , c\rangle & = & tv_g[{\cal s}_2^{\delta_2^*};(\gamma'\cup\gamma , c+c ' ) ] .
\label{def1}\end{aligned}\ ] ] this operator @xmath123 does not depends on the interior of the simple polyhedron @xmath124 used to compute the transition amplitude @xcite . using this notation we can rewrite write equation ( [ split ] ) as @xmath125 = \langle \emptyset | \hat{p}_{\emptyset,\gamma_2 } ^{m_3 } \hat{p}_{\gamma_2 , \gamma_1 } ^{m_2 } \hat{p}_{\gamma_1 , \emptyset}^{m_1 } |\emptyset\rangle ~~.\ ] ] in the previous notation we have implicitly assumed that the boundary of each of the manifold @xmath126 can be naturally seen as made of two disjoint components ( @xmath127 ) one associated to the _ future _ @xmath128 and the other to the _ past _ @xmath129 .
@xmath130 is the operator of equation ( [ def1 ] ) between the hilbert space associate to the graph @xmath131 and the hilbert space associated to the graph @xmath132 . with this convention
it is natural to define the composition of two manifolds with boundary @xmath133 and @xmath134 when @xmath135 as the manifold @xmath136 that one obtains gluing @xmath133 and @xmath134 along the isomorphic boundary @xmath137 .
at this point we are ready to discus the functorial nature of the invariant . from the independence of the prtv state sum from the chosen polyhedron we have
that the @xmath138 define a semifunctor on the space of manifolds with a fixed graph on their boundary : @xmath139 this @xmath138 does not properly define a topological field theory since it is not a functor ( @xmath140 ) .
there is however a standard procedure that allows to define a functor from a semifunctor .
in fact , if we consider the hilbert space @xmath141 from @xmath142 follows that its restriction to the family of hilbert spaces @xmath143 become a functor .
i.e , we used the operator @xmath144 as the projector on the physical hilbert space of the theory . in this way
the prtv - model defines a 2 + 1 dimensional quantum field theory @xcite in the atiyah terminology @xcite . from the physical point of view this discussion
can be seen as a computation of hartle - hawking wave function and of the topology change amplitude of the 3-dimensional lattice gravity of ponzano and regge @xcite . in consideration of a hamiltonian ( initial values ) formulation in terms of loop - variable
this is particularly nice .
in fact , in dimension @xmath23 the hamiltonian formulations is defined on an @xmath145 dimensional cauchy surface and in any dimension the intersection of an @xmath145 surface and any 2-dimensional shadow is an 1-dimensional graph embedded on the cauchy surface .
even more interesting is the fact the hilbert space structure associated to the bf theory is very similar to the one of loop quantum gravity .
in previous section we showed how the ponzano - regge model correspond to the bf theory in 3 dimension . in this section
we discuss an example of 4 dimensional theory that can be expressed within the spin foam formalism : the crane - yetter - ooguri model of bf theory and the spin foam model of barret and crane .
we do not discus here the model of reisenberger @xcite .
all this models share the characteristic to be based on a theory of the bf type .
this lead freidel and krasnov @xcite to propose a way to systematicaly derive the _
spin foam model _ associated to any theory that admits a formulation in terms of an action of bf type plus lagrange s multipliers imposed constraints . the state sum formulation of the topological @xmath146 theory in 4-dimensions @xmath147[{\cal d}b(x ) ] e^{{\rm i}\int { \rm tr}[b \wedge f[a ] ] + \lambda~ { \rm tr}[b\wedge b ] } \label{eq : actionbf}\\ & = & \int [ { \cal d}a(x ) ] e^{\frac{{\rm i}}{2\lambda } \int{\rm tr}[f[a ] \wedge f[a ] ] } % % % \nonumber\\ & = & = \int [ { \cal d}a(x ) ] e^{\frac{{\rm i}}{2\lambda } \int d\big ( { \rm tr}[a \wedge f[a]]+\frac{2}{3}{\rm tr}[a\wedge a\wedge a ] \big ) } \nonumber\end{aligned}\ ] ] with cosmological constant is given by the crane - yetter @xcite or by the ooguri ( without cosmological constant ) @xcite models . the partition function of the model associated to a triangulation @xmath77 of a manifold @xmath78 can be written as : @xmath148 where @xmath149
, @xmath150 denotes a coloring of the faces of @xmath77 by irreducible representation of @xmath151 , @xmath152 denotes a coloring of the tetrahedra of @xmath77 by intertwiners and the sum is over all such colorings with @xmath153 .
moreover @xmath154 denotes the quantum dimension of the representation of spin @xmath155 and @xmath156 denote the quantum @xmath157-j symbol associated with the 4-simplex @xmath158 .
more precisely , associated to a 4-simplex @xmath158 we consider a graph @xmath159 given by the intersection between the 2-skeleton of the complex which is dual to the triangulation and the boundary of the 4-simplex @xmath158 ( see figure [ fig : vertexbf ] ) .
@xmath159 corresponds to the pentagram graph and we color its 10 edges by @xmath160 and its 5 vertices by @xmath161 .
@xmath162 corresponds to the reshetikhin - turaev evaluation of the colored graph @xmath159 .
this state sum does not depend on the triangulation when correctly normalized by a factor @xmath163 , where @xmath164 is the number of @xmath165 simplices , and it corresponds to the evaluation of the bf partition function of equation ( [ eq : actionbf ] ) .
the relativistic - spin - foam model is a modification of the state sum ( [ eq : actionbf ] ) in the case of the @xmath166 gauge group in which the conditions corresponding to the quantum 4-simplex conditions ( when @xmath167 ) are imposed by hand .
yetter generalized this construction to the case of non vanishing @xmath168 @xcite .
yetter emphasized that the quantum group that should be considered is @xmath169 .
this state sum model can be written : @xmath170 so this model corresponds to two copies of @xmath171 @xmath146 theories ( or an @xmath166 @xmath146 theory ) together with @xmath157 constraints imposed for each 4-simplex .
the presence of @xmath172 and @xmath173 corresponds to the geometricity constraint . in @xcite
it is shown that the above spin foam model can be associated to a well defined field action : the @xmath166 plebanski action @xcite .
this is a deformed bf theory depending on an @xmath166 connection @xmath174 , a two form valued into @xmath166 @xmath175 , and a scalar symmetric traceless matrix @xmath176[kl]}$ ] ( @xmath177 ) that acts as lagrange multiplier fields .
this action reads : @xmath178 & = & \int_m\!\!\ !
\bigg [ b^{ij}\wedge f_{ij}(\omega ) - \frac{\lambda}{4}\epsilon_{ijkl } b^{ij } \wedge b^{kl } % % % # # % % % \nonumber \\ & & ~~ - \frac{1}{2 } \phi(b)_{ij}\wedge b^{ij}\bigg ] ~ , \label{ap}\end{aligned}\ ] ] where @xmath179 . in this context ,
the geometry constraints are imposed , at the quantum level , by the integration over lagrange multiplier fields @xmath176[kl]}$ ] .
this action has two nice characteristics : ( 1 ) one of its sectors of solution is closely connected to gravity ; ( 2 ) it is a deformation of a quantizable topological theory ( the bf theory ) that can play a role analogous to the one of free theory in standard quantum field theory .
moreover , it is intriguing that this description of gravity as a constraint bf theory can be generalized to higher dimension @xcite .
in this work we showed that the spin foam formalism gives an unified framework for the discussion of an important class of diff - invariant quantum theories . between them : topological bf theory in 3 and 4 dimension , the four dimensional formulation of loop quantization of gravity , the barret - crane model .
the main unsolved problems of the general picture we described here are : 1 .
the state sum model we described are properly defined only on a single given triangulation ( or its dual spin - foam ) .
if this is fine in the case of a topological field theory without local degree of freedom this is questionable in the case of gravity .
work is now in progress @xcite on the direction of dealing with all the possible triangulation .
this is somehow analogue to consider generalization ( where the triangulation is not the only dynamical variable ) of the model of dynamical triangulations @xcite .
2 . the connection between the spin foam models and the exponential of the constraints need to be improved with a rigorous mathematical analysis of the derivation .
a systematical study of the physical implication of taking different weight factor is missing .
acknowledgment : + this work was greatly influenced by numerous discussion with m. carfora , l. freidel , j. lewandowski , l. lusanna , a. marzuoli , m. pauri , c. rovelli and j. zapata , to whom all i m extremely grateful .
the work of r.d.p . at cpt marseille is supported by a dalla riccia fellowship .
abhay ashtekar , jerzy lewandowski , donald marolf , jose mourao , and thomas thiemann , _ quantization of diffeomorphism invariant theories of connections with local degrees of freedom _ , j. math . phys . * 36 * ( 1995 ) , 64566493 . carlo rovelli , _ strings , loops and others : a critical survey of the present approaches to quantum gravity _ ,
gravitation and relativity : at the turn of the millennium ( pune , india ) ( n. dadhich and j. narlikar , eds . ) , proceeding of the gr15 conference , i.u.c.a.a . , december , 16 -21 1998 , p. 281 | the canonical `` loop '' formulation of quantum gravity is a mathematically well defined , background independent , non perturbative standard quantization of einstein s theory of general relativity .
some among the most meaningful results of the theory are : 1 ) the complete calculation of the spectrum of geometric quantities like the area and the volume and the consequent physical predictions about the structure of the space - time at the planck scale ; 2 ) a microscopical derivation of the bekenstein - hawking black - hole entropy formula . unfortunately , despite recent results , the dynamical aspect of the theory ( imposition of the wheller - de witt constraint ) remains elusive . after a short description of the basic ideas and the main results of _ loop _ quantum gravity
we show in which sence the exponential of the super hamiltonian constraint leads to the concept of _ spin foam _ and to a four dimensional formulation of the theory .
moreover , we show that some topological field theories as the bf theory in 3 and 4 dimension admits a spin foam formulation .
we argue that the spin - foams / spin - networks formalism it is the natural framework to discuss loop quantum gravity and topological field theory .
-0.8 cm -0.8 cm -0.8 cm -1.4 cm -0.8 cm -0.8 cm # 1 ( # 1 ) # 1 c ( 35,25 ) ( 15 , 0)(0,1)10 ( 20 , 0)(0,1)10 ( 15 , 0)(1,0 ) 5 ( 15,10)(1,0 ) 5 ( 15,25)(1,0)5 ( 15,15)@xmath0 ( 15,15)(20,20)[l](20,15)(30,20)[r ] # 1#2#3 c ( 40,40 ) ( 18,32)@xmath0 ( 0,15)(1,0)40 ( 18,17)@xmath1 ( 20,15)(40,30 ) ( 18 , 2)@xmath2 ( 0,15 ) ( 40,15 ) # 1#2#3#4#5#6 c ( 50,30 ) ( 0,15)(1,-1)15 ( 0,22)@xmath3 ( 0,15)(1 , 1)15 ( 0 , 0)@xmath4 ( 0,15 ) ( 30,15)(-1 , 1)15 ( 28,20)@xmath5 ( 30,15)(-1,-1)15 ( 28 , 2)@xmath6 ( 30,15 ) ( 0,15)(1,0)30 ( 12,16)@xmath7 ( 15,30)(1,0)25 ( 15,30 ) ( 15 , 0)(1,0)25 ( 15 , 0 ) ( 40 , 0)(0,1)30 ( 42,12)@xmath8 cpt-99/p.3796 + university of parma preprint uprf-99 - 01 + _ xxx - archive : _ gr - qc/9903076 .
+ to appear in the proceeding of the xxiii congress of the italian society for general relativity and gravitational physics ( sigrav ) , monopoli ( italy ) , september @xmath9-@xmath10 , 1998 . |
this work is supported by nserc and trent university .
d. belitz and t. r. kirkpatrick , rev .
mod . phys .
* 66 * ( 1994 ) 261 .
a. georges and g. kotliar , rev .
* 68 * ( 1996 ) 13 .
v. dobrosavljevic , a. a. pastor and b. k. nikolic , europhys .
* 62 * ( 2003 ) 76 .
k. byczuk , w. hofstetter and d. vollhardt , phys .
* 94 * ( 2005 ) 056404 .
yun song , r. wortis and w. a. atkinson , cond - mat/0609590 . | with the goal of measuring localization in disordered interacting systems , we examine the finite - size scaling of the geometrically - averaged density of states calculated from the local green s function with finite energy resolution .
our results show that , unlike in a simple energy binning procedure , there is no limit in which the finite energy resolution is irrelevant . , , disorder ; interactions ; localization ;
dmft 71.30.+h,71.23.-k , 71.23.an , 71.27.+a significant disorder , for example due to non - stoichiometric doping , is a feature of many materials , and may in fact play a key role in many interesting behaviors , especially when combined with electronic correlations .
the study of systems with both disorder and interactions has a long history,@xcite but it has been difficult to access the regime in which both are strong , where the physics of both anderson localization and the mott gap are significant .
the technique of dynamical mean field theory@xcite ( dmft ) has been very successful in the study of systems with strong interactions , and there has recently been interest in applying it to disordered systems.@xcite dmft calculates the local green s function , @xmath0
. how can information on localization be drawn from this ?
the geometrically - averaged density of states ( gados ) appears to be ideally suited to the task .
this is the geometric average over sites of the local density of states ( ldos ) which is obtained directly from the local green s function : @xmath1^{1/n } , \\
\rho ( { \bf r}_i,\omega ) & = & -{1 \over \pi } { \rm i m } g({\bf r}_i,{\bf r}_i , \omega+i\gamma)\end{aligned}\ ] ] which is exact in the limit @xmath2 . in an infinite system , @xmath3 at the fermi level is an order parameter for the anderson transition , being nonzero when states at the fermi level are extended and zero when they become localized .
this is because when wavefunctions at a given energy become localized in real space , the ldos at that energy becomes discrete : localized wavefunctions at the same energy are well separated in position so the ldos at that energy goes to zero far from the wavefunction centers . to consider specific disorder configurations , it is necessary to work with finite size systems , and in this case finite - size scaling analysis of the gados
must be performed .
this procedure is well studied in non - interacting systems for which single - particle wavefunctions can be calculated , providing perfect energy resolution . in a numerical calculation on an interacting system , however , the green s function is obtained for finite @xmath4 .
we have explored the impact this has on the scaling of the gados in the context of a noninteracting model @xmath5 with site energies , @xmath6 , distributed with probability @xmath7 .
the width of the potential distribution , @xmath8 , parameterizes the strength of the disorder .
the model is solved exactly by numerical diagonalization using the lapack libraries . in earlier
work,@xcite we calculated the local density of states within a finite energy resolution @xmath4 using a simple binning procedure : @xmath9 each state which falls in the energy window is assigned equal weight . in this case , we demonstrated that the gados scales as a function of an effective length scale : @xmath10 where @xmath11 acts as an effective inelastic mean free path , limiting the relevant system size .
@xmath11 can also be seen as the system size divided by the average number of states in a bin : @xmath12 where in one dimension @xmath13 is the average energy spacing between states and @xmath14 is the density of states .
here we model more accurately the way the green s function introduces finite energy resolution : @xmath15}. \label{lor}\end{aligned}\ ] ] all results shown are at the band center on one - dimensional systems with disorder strength @xmath16 , for which all states are localized . .
all points represent averages over at least 1000 disorder configurations .
@xmath17 is the system size , and @xmath18 where @xmath19 is the density of states .
( a ) square energy bin ( eqn .
( [ box ] ) ) with constant width @xmath4 .
( b ) lorentzian energy bin ( eqn .
( [ lor ] ) ) with constant @xmath4 .
( c ) and ( d ) lorentzian energy bin with @xmath20.,scaledwidth=45.0% ] fig .
1(a ) shows the gados as a function of system size for three non - zero values of @xmath4 using the square binning in eqn .
( [ box ] ) reported earlier.@xcite fig .
1(b ) shows the gados as a function of system size for the same @xmath4 values using the lorentzian binning in eqn .
( [ lor ] ) .
two features are immediately apparent : the lack of exponential decay and the overall inflation of the gados . with the lorentzian bin , _ all _ states contribute to the sum in eqn .
( [ lor ] ) with rapidly decreasing weight away from @xmath21 . despite the small weight assigned to them ,
the many states outside the peak of the lorentzian have a dramatic impact on the gados .
this is because when these states are centered , in real space , at positions with otherwise small ldos they can significantly increase the ldos . even a modest increase in the _ minimum _ values of the ldos causes a dramatic increase in the gados .
a second important distinction between the two binning schemes arises in the limit @xmath22 . in this limit , for each @xmath21 there will be disorder configurations for which ( i ) in the square bin case there are no states in a bin and ( ii ) for the lorentzian bin case no state has significant weight . in the case of the square bin
, we removed these bins from the average . however , there is no systematic way to do this in the lorentzian case .
the inclusion of these states dramatically lowers the configuration averaged ldos and hence the gados even below the @xmath23 value .
it appears to be possible to account for the behavior of @xmath24 with lorentzian binning within the framework of an effective length ; however , a different form of @xmath25 is required . in the square bin case , in the limit of small @xmath4 , @xmath26 , because by construction there is always at least one state in each bin . in the lorentzian bin case ,
the average number of states in a bin may be less than one .
@xmath25 is now simply equal to @xmath27 .
this is consistent with the nearly size independent results when a fixed value of gamma is used .
furthermore , fig .
1(c ) shows the gados as a function of system size using bin widths @xmath28 .
fig .
1(d ) presents the same data plotted versus @xmath29 showing scaling .
the key point is that there is no limit in which the finite energy resolution is irrelevant . in conclusion
, there appears to be no finite - size scaling of the gados calculated from the local green s function with fixed energy resolution , and the scaling which occurs when @xmath30 arises at least in part from physics unrelated to localization .
we suggest that other measures of localization , such as the inverse participation ratio , which do not share the extreme sensitivity to small values of the ldos may be more effective than the gados in interacting systems , although changes to the scaling due to finite energy resolution will still need to be addressed . |
our objective in this paper is to study the following : [ quest ] let @xmath7 be a projective family of surfaces of degree @xmath8 in @xmath9 , with @xmath10 a smooth threefold , and @xmath11 the complex unit disc ( usually called a _ degeneration _ of the general @xmath12 , for @xmath13 , which is a smooth surface , to the _ central fibre _ @xmath14 , which is in general supposed to be singular ) .
what are the limits of tangent , bitangent , and tritangent planes to @xmath6 , for @xmath13 , as @xmath15 tends to @xmath3 ?
similar questions make sense also for degenerations of plane curves , and we refer to @xcite for a glimpse on this subject . for surfaces , our contribution is based on foundational investigations by caporaso and harris @xcite , and independently by ran @xcite , which were both aimed at the study of the so - called _ severi varieties _ ,
i.e. the families of irreducible plane nodal curves of a given degree .
we have the same kind of motivation for our study ; the link with question [ quest ] resides in the fact that nodal plane sections of a surface @xmath6 in @xmath9 are cut out by those planes that are tangent to @xmath6 .
ultimately , our interest resides in the study of severi varieties of nodal curves on @xmath1 surfaces .
the first interesting instance of this is the one of plane sections of smooth quartics in @xmath9 , the latter being primitive @xmath1 surfaces of genus @xmath16 .
for this reason , we concentrate here on the case @xmath17 .
we consider a couple of interesting degenerations of such surfaces to quite singular degree 4 surfaces , and we answer question [ quest ] in these cases .
the present paper is of an explorative nature , and hopefully shows , in a way we believe to be useful and instructive , how to apply some general techniques for answering some specific questions . on the way
, a few related problems will be raised , which we feel can be attacked with the same techniques . some of them we solve ( see below ) , and the other ones we plan to make the object of future research
. coming to the technical core of the paper , we start from the following key observation due to caporaso and harris , and ran ( see [ s : zrr ] for a complete statement )
. assume the central fibre @xmath14 is the transverse union of two smooth surfaces , intersecting along a smooth curve @xmath18 .
then the limiting plane of a family of tangent planes to the general fibre @xmath6 , for @xmath13 , is : [ i : node ] either a plane that is tangent to @xmath14 at a smooth point , or [ i : tangency ] a tangent plane to @xmath18 .
furthermore , the limit has to be counted with multiplicity 2 in case .
obviously , this is not enough to deal directly with all possible degenerations of surfaces .
typically , one overcomes this by applying a series of base changes and blow ups to @xmath0 , thus producing a semistable model @xmath19 of the initial family , such that it is possible to provide a complete answer to question [ quest ] for @xmath0 by applying a suitable extended version of the above observation to @xmath20 .
we say that @xmath0 is _ well behaved _ when it is possible to do so , and
@xmath20 is then said to be a _ good model _ of @xmath0 .
we give in
[ s : zrr ] a rather restrictive criterion to ensure that a given semistable model is a good model , which nevertheless provides the inspiration for constructing a good model for a given family .
we conjecture that there are suitable assumptions , under which a family is well behaved .
we do not seek such a general statement here , but rather prove various incarnations of this principle , thus providing a complete answer to question [ quest ] for the degenerations we consider .
specifically , we obtain : [ t : tetrahedron ] let @xmath7 be a family of general quartic surfaces in @xmath21 degenerating to a _ tetrahedron _
@xmath14 , i.e. the union of four independent planes .
the singularities of @xmath10 consist in four ordinary double points on each _ edge _ of @xmath14 .
the limits in @xmath22 of @xmath5-tangent planes to @xmath6 , for @xmath13 , are : + the @xmath23 webs of planes passing through a singular point of @xmath10 , plus the @xmath24 webs of planes passing through a _ vertex _ of @xmath14 , the latter counted with multiplicity @xmath16 ; + the @xmath25 pencils of planes passing through two double points of the total space @xmath10 that do not belong to an edge of @xmath14 , plus the @xmath26 pencils of planes passing through a vertex of @xmath14 and a double point of @xmath10 that do not belong to a common edge of @xmath14 ( with multiplicity @xmath16 ) , plus the @xmath27 pencils of planes containing an edge of @xmath14 ( with multiplicity @xmath28 ) ; + the @xmath29 planes containing three double points of @xmath10 but no edge of @xmath14 , plus the @xmath30 planes containing a vertex of @xmath14 and two double points of @xmath10 , but no edge of @xmath14 ( with multiplicity @xmath16 ) , plus the @xmath23 planes containing an edge of @xmath14 and a double point of @xmath10 not on this edge ( with multiplicity @xmath28 ) , plus the @xmath24 _ faces _ of @xmath14 ( with multiplicity @xmath31 ) .
[ t : kummer ] let @xmath7 be a family of general quartic surfaces degenerating to a general _
kummer surface _ @xmath14 .
the limits in @xmath22 of @xmath5-tangent planes to @xmath6 , for @xmath13 , are : + the dual surface @xmath32 to the kummer ( which is itself a kummer surface ) , plus the @xmath28 webs of planes containing a node of @xmath14 ( with multiplicity @xmath33 ) ; + the @xmath34 pencils of planes containing two nodes of @xmath14 , each counted with multiplicity @xmath24 ; + the @xmath28 planes tangent to @xmath14 along a _ contact conic _ ( with multiplicity @xmath35 ) , plus the @xmath25 planes containing exactly three nodes of @xmath14 ( with multiplicity @xmath36 ) .
we could also answer question [ quest ] for degenerations to a general union of two smooth quadrics , as well as to a general union of a smooth cubic and a plane ; once the much more involved degeneration to a tetrahedron is understood , this is an exercise .
we do not dwell on this here , and we encourage the interested reader to treat these cases on his own , and to look for the relations between these various degenerations .
however , a mention to the degeneration to a _
double quadric _ is needed , and we treat this in [ sec : pair of quadrics ] .
apparent in the statements of theorems [ t : tetrahedron ] and [ t : kummer ] is the strong enumerative flavour of question [ quest ] , and actually we need information of this kind ( see proposition [ p : deg - dual ] ) to prove that the two families under consideration are well behaved .
still , we hope to find a direct proof in the future . as a matter of fact , caporaso and
harris main goal in @xcite is the computation of the degrees of severi varieties of irreducible nodal plane curves of a given degree , which they achieve by providing a recursive formula . applying the same strategy ,
we are able to derive the following statement ( see [ s : triangle ] ) : [ t : triangle ] let @xmath37 be three independent lines in the projective plane , and consider a degree @xmath38 divisor @xmath39 cut out on @xmath40 by a general quartic curve . the sub linear system @xmath41 of @xmath42 parametrizing curves containing @xmath39 has dimension @xmath16 . for @xmath43 , we let @xmath44 be the zariski closure in @xmath41 of the locally closed subset parametrizing irreducible @xmath5-nodal curves .
then @xmath44 has codimension @xmath5 in @xmath45 , and degree @xmath46 for @xmath47 , degree @xmath48 for @xmath49 , degree @xmath31 for @xmath50 .
remarkably , one first proves a weaker version of this ( in [ s : triangle ] ) , which is required for the proof of theorem [ t : tetrahedron ] , given in [ s:4planes ] .
then , theorem [ t : triangle ] is a corollary of theorem [ t : tetrahedron ] .
it has to be noted that theorems [ t : tetrahedron ] and [ t : kummer ] display a rather coarse picture of the situation .
indeed , in describing the good models of the degenerations , we interpret all limits of nodal curves as elements of the limit @xmath51 of @xmath52 , for @xmath13 , inside the relative hilbert scheme of curves in @xmath10 .
we call @xmath51 the _ limit linear system _ of @xmath52 , for @xmath13 ( see [ s : resolving ] ) , which in general is no longer a @xmath21 , but rather a degeneration of it . while in @xmath22 , which is also a limit of @xmath52 , for @xmath13
, there are in general elements which do not correspond to curves ( think of the plane section of the tetrahedron with one of its faces ) , all elements in @xmath51 do correspond to curves , and this is the right ambient to locate the limits of nodal curves .
so , for instance , each face appearing with multiplicity @xmath31 in theorem [ t : tetrahedron ] is much better understood once interpreted as the contribution given by the @xmath31 curves in @xmath53 appearing in theorem [ t : triangle ] .
it should also be stressed that the analysis of a semistable model of @xmath0 encodes information about several flat limits of the @xmath6 s in @xmath9 , as @xmath2 tends to @xmath3 ( each flat limit corresponds to an irreducible component of the limit linear system @xmath51 ) , and an answer to question [ quest ] for such a semistable model would provide answers for all these flat limits at the same time .
thus , in studying question [ quest ] for degenerations of quartic surfaces to a tetrahedron , we study simultaneously degenerations to certain rational quartic surfaces , e.g. , to certain _ monoid _ quartic surfaces that are projective models of the faces of the tetrahedron , and to sums of a _ self dual _ cubic surface plus a suitable plane . for degenerations to a kummer , we see simultaneously degenerations to double quadratic cones , to sums of a smooth quadric and a double plane ( the latter corresponding to the projection of the kummer from one of its nodes ) , etc .
though we apply the general theory ( introduced in [ s : gnl - rules ] ) to the specific case of degenerations of singular plane sections of general quartics , it is clear that , with some more work , the same ideas can be applied to attack similar problems for different situations , e.g. , degenerations of singular plane sections of general surfaces of degree @xmath54 , or even singular higher degree sections of ( general or not ) surfaces of higher degree . for example , we obtain theorem [ t : triangle ] thinking of the curves in @xmath41 as cut out by quartic surfaces on a plane embedded in @xmath9 , and letting this plane degenerate . by the way , this is the first of a series of results regarding no longer triangles , but general configurations of lines , which can be proved , we think , by using the ideas in this paper . on the other hand , for general primitive @xmath1 surfaces of any genus @xmath55 , there is a whole series of known enumerative results @xcite , yet leaving some open space for further questions , which also can be attacked in the same way .
another application of our analysis of question [ quest ] is to the irreducibility of families of singular curves on a given surface .
this was indeed ran s main motivation in @xcite , since he applied these ideas to give an alternative proof to harris one @xcite of the irreducibility of severi varieties of plane curves .
the analogous question for the family of irreducible @xmath5nodal curves in @xmath56 , for @xmath10 a general primitive @xmath1 surface of genus @xmath57 is widely open .
in @xcite one proves that for any non negative @xmath58 , with @xmath59 and @xmath60 , the _
universal severi variety _
@xmath61 , parametrizing @xmath5nodal members of @xmath62 , with @xmath10 varying in the moduli space @xmath63 of primitive @xmath1 surfaces of genus @xmath64 in @xmath65 , is irreducible for @xmath66 .
one may conjecture that all universal severi varieties @xmath61 are irreducible ( see @xcite ) , and we believe it is possible to obtain further results in this direction using the general techniques presented in this paper . for instance , the irreducibility of @xmath67 , @xmath68 , which is well known and easy to prove ( see proposition [ prop : irr ] ) , could also be deduced with the degeneration arguments developed here .
note the obvious surjective morphism @xmath69 . for @xmath70 general
, one can consider @xmath71 the _ severi variety _ of @xmath5nodal curves in @xmath56 ( i.e. the fibre of @xmath72 over @xmath70 ) , which has dimension @xmath73 ( see @xcite ) .
note that the irreducibility of @xmath74 does not imply the one of the severi varieties @xmath75 for a general @xmath76 ; by the way , this is certainly not true for @xmath77 , since @xmath78 has dimension 0 and degree bigger than 1 , see @xcite .
of course , @xmath79 is isomorphic to the _ dual variety _ @xmath80 , hence it is irreducible . generally speaking ,
the smaller @xmath5 is with respect to @xmath64 , the easier it is to prove the irreducibility of @xmath81 : partial results along this line can be found in @xcite and ( * ? ? ?
* appendix a ) . to the other extreme
, the curve @xmath82 is not known to be irreducible for @xmath76 general . in the simplest case @xmath83 , this amounts to proving the irreducibility of @xmath84 for a general quartic @xmath10 in @xmath21 , which is the nodal locus of @xmath85 .
this has been commonly accepted as a known fact , but we have not been able to find any proof of this in the current literature .
we give one with our methods ( see theorem [ thm : v2 ] ) . finally , in
[ ssec : monod ] , we give some information about the monodromy group of the finite covering @xmath86 , by showing that it contains some _ geometrically interesting _ sugbroups .
note that a remarkable open question is whether the monodromy group of @xmath87 is the full symmetric group for all @xmath55 .
the paper is organized as follows . in
[ s : gnl - rules ] , we set up the machinery : we give general definitions , introduce limit linear systems , state our refined versions of caporaso and harris and ran s results , introduce limit severi varieties . in
[ s : enumerate ] , we state some known results for proper reference , mostly about the degrees of the singular loci of the dual to a projective variety . in [ s:4planes ] and [ s : kummer - degen ] , we give a complete description of limit severi varieties relative to general degenerations of quartic surfaces to tetrahedra and kummer surfaces respectively ; theorems [ t : tetrahedron ] and [ t : kummer ] are proved in
[ s : tetra - concl ] and [ s : kummer - concl ] respectively . in
[ sec : pair of quadrics ] we briefly treat other degenerations of quartics . section [ s : kummer - descr ] contains some classical material concerning kummer quartic surfaces , as well as a few results on the monodromy action on their nodes ( probably known to the experts but for which we could not find any proper reference ) : they are required for our proof of theorem [ thm : v2 ] and of the results in [ ssec : monod ] .
section [ s : triangle ] contains the proof of a preliminary version of theorem [ t : triangle ] ; it is useful for [ s:4planes ] , and required for [ s : irreducibility ] .
section [ s : irreducibility ] contains theorem [ thm : v2 ] and the aforementioned results on the monodromy .
we thank erwan brugall ( who , together with g. mikhalkin , was able to provide the enumerative numbers of theorems [ t : tetrahedron ] and [ t : triangle ] using tropical methods ) , and concettina galati for numerous discussions , both enlightening and motivating .
this project profited of various visits of the second author to the first , which have been made possible by the research group grifga , in collaboration between cnrs and indam .
the first author is a member of gnsaga of indam and was partly supported by the project `` geometry of algebraic varieties '' funded by the italian miur .
the second author is a member of projects class and mack of the french anr .
we will work over the field @xmath88 of complex numbers .
we denote the linear equivalence on a variety @xmath89 by @xmath90 , or simply by @xmath91 when no confusion is likely .
let @xmath92 be a group ; we write @xmath93 when @xmath94 is a subgroup of @xmath92 .
we use the classical notation for projective spaces : if @xmath95 is a vector space , then @xmath96 is the space of lines in @xmath95 , and if @xmath97 is a locally free sheaf on some variety @xmath89 , we let @xmath98 be @xmath99 . we denote by @xmath100 the projective space dual to @xmath101 , and if @xmath89 is a closed subvariety of @xmath101 , we let @xmath102 be its dual variety , i.e. the zariski closure in @xmath103 of the set of those hyperplanes in @xmath101 that are tangent to the smooth locus of @xmath89 . by a _ node _
, we always mean an ordinary double point .
let @xmath104 be an integer .
a _ nodal _ ( resp .
_ @xmath5-nodal _ ) variety is a variety having nodes as its only possible singularities ( resp .
precisely @xmath5 nodes and otherwise smooth ) . given a smooth surface @xmath10 together with an effective line bundle @xmath105 on it
, we define the _ severi variety _ @xmath106 as the zariski closure in the linear system @xmath107 of the locally closed subscheme parametrizing irreducible @xmath5-nodal curves .
we usually let @xmath94 be the line divisor class on @xmath108 ; when @xmath109 is a _ hirzebruch surface _ , we let @xmath110 be the divisor class of its ruling over @xmath111 , we let @xmath112 be an irreducible effective divisor with self - intersection @xmath113 ( which is unique if @xmath114 ) , and we let @xmath94 be the divisor class of @xmath115 . when convenient ( and if there is no danger of confusion ) , we will adopt the following abuse of notation : let @xmath116 be a birational morphism , and @xmath117 ( resp .
@xmath118 ) a divisor ( resp . a divisor class ) on @xmath89 ; we use the same symbol @xmath117 ( resp .
@xmath118 ) to denote the proper transform @xmath119 ( resp .
the pull - back @xmath120 ) on @xmath121 .
for example , let @xmath105 be a line in @xmath108 , and @xmath94 the divisor class of @xmath105 .
we consider the blow - up @xmath122 at a point on @xmath105 , and call @xmath123 the exceptional divisor .
the divisor class @xmath94 on @xmath124 is @xmath125 , and @xmath105 on @xmath124 is linearly equivalent to @xmath126 .
let then @xmath127 be the blow - up of @xmath124 at the point @xmath128 , and @xmath129 be the exceptional divisor . the divisor @xmath123 ( resp .
@xmath105 ) on @xmath130 is linearly equivalent to @xmath131 ( resp . to @xmath132 ) . in figures depicting series of blow
ups , we indicate with a big black dot those points that have been blown up .
in this section we explain the general theory upon which this paper relies .
we build on foundational work by caporaso and harris @xcite and ran @xcite , as reinvestigated by galati @xcite ( see also the detailed discussion in @xcite ) . in this paper we will consider flat , proper families of surfaces @xmath133 , where @xmath134 is a disc centered at the origin .
we will denote by @xmath6 the ( schematic ) fibre of @xmath135 over @xmath136 .
we will usually consider the case in which the _ total space _ @xmath10 is a smooth threefold , @xmath135 is smooth over @xmath137 , and @xmath6 is irreducible for @xmath138 .
the _ central fibre _
@xmath14 may be singular , but we will usually consider the case in which @xmath14 is reduced and with _ local normal crossing _ singularities . in this case
the family is called _
semistable_. another family of surfaces @xmath139 as above is said to be a _ model of _
@xmath140 if there is a commutative diagram @xmath141_{f ' } \ar@{}[dr]|{\box } & \bar s ' \ar[d ] \ar[l ] \ar@{-->}[r]^p & \bar s \ar[r ] \ar[d ] & s \ar[d]^f \ar@{}[dl]|{\box } \\
\delta & \delta \ar[l]^{t^{d ' } \mapsfrom t } \ar@{=}[r ] & \delta \ar[r]_{t\mapsto t^d } & \delta } \ ] ] where the two squares marked with a @xmath142 are cartesian , and @xmath72 is a birational map , which is an isomorphism over @xmath143 .
the family @xmath139 , if semistable , is a _ semistable model of _
@xmath140 if in addition @xmath144 and @xmath72 is a morphism . the _ semistable reduction theorem _ of @xcite asserts that @xmath7 always has a semistable model .
[ families of surfaces in @xmath9 ] [ ex : pi3 ] consider a _ linear pencil _ of degree @xmath146 surfaces in @xmath9 , generated by a _
surface @xmath147 and a _ special _ one @xmath14 .
this pencil gives rise to a flat , proper family @xmath148 @xmath149 , with @xmath150 a hypersurface of type @xmath151 in @xmath152 , isomorphic to the blow up of @xmath9 along the _ base locus _
@xmath153 of the pencil , and @xmath154 as fibres over @xmath155 , respectively .
we will usually consider the case in which @xmath14 is reduced , its various components may have isolated singularities , but meet transversely along smooth curves contained in their respective smooth loci
. thus @xmath14 has local normal crossing singularities , except for finitely many isolated _ extra singularities _ belonging to one , and only one , component of @xmath14 .
we shall study the family @xmath133 obtained by restricting @xmath150 to a disk @xmath156 centered at @xmath3 , such that @xmath157 is smooth for all @xmath158 , and we will consider a semistable model of @xmath133 . to do so , we resolve the singularities of @xmath10 which occur in the central fibre of @xmath135 , at the points mapped by @xmath159 to the intersection points of @xmath147 with the double curves of @xmath14 ( they are the singular points of the curve @xmath153 ) .
these are _ ordinary double points _ of @xmath10 , i.e. singularities analytically equivalent to the one at the origin of the hypersurface @xmath160 in @xmath161 .
such a singularity is resolved by a single blow up , which produces an exceptional divisor @xmath162 , and then it is possible to contract @xmath110 in the direction of either one of its rulings without introducing any singularity : the result is called a _ small resolution _ of the ordinary double point .
if @xmath14 has no extra singularities , the small resolution process provides a semistable model .
otherwise we will have to deal with the extra singularities , which are in any case smooth points of the total space .
we will do this when needed .
let @xmath163 be the semistable model thus obtained .
one has @xmath164 for @xmath138 .
if @xmath14 has irreducible components @xmath165 , then @xmath166 consists of irreducible components @xmath167 which are suitable blow
ups of @xmath165 , respectively .
if @xmath168 is the number of ordinary double points of the original total space @xmath10 , we will denote by @xmath169 the exceptional curves on @xmath167 arising from the small resolution process .
going back to the general case , we will say that @xmath133 is _ quasi semistable _ if @xmath14 is reduced , with _
local normal crossing _ singularities , except for finitely many isolated _ extra singularities _ belonging to one , and only one , component of @xmath14 , as in example [ ex : pi3 ] .
assume then that @xmath14 has irreducible components @xmath165 , intersecting transversally along the double curves @xmath170 , which are cartier divisors on the corresponding components .
[ l : triple - point ] assume @xmath133 is quasi
let @xmath171 be irreducible components of @xmath14 intersecting along the _
double curve _ @xmath18 .
then @xmath172 where a _
triple point _ is the intersection @xmath173 with a component @xmath174 of @xmath14 different from @xmath171 .
[ see @xcite ] [ rem : tpf ] there is a version of the triple point formula for the case in which the central fibre is not reduced , but its support has local normal crossings . then ,
if the multiplicities of @xmath171 are @xmath175 respectively , one has @xmath176 where each triple point @xmath173 has to be counted with the multiplicity @xmath177 of @xmath174 in @xmath14 .
let us consider a quasi
semistable family @xmath140 as in
[ s : setting ] .
suppose there is a fixed component free line bundle @xmath178 on the total space @xmath10 , restricting to a line bundle @xmath179 on each fibre @xmath6 , @xmath180 .
we assume @xmath178 to be ample , with @xmath181 constant for @xmath136 .
if @xmath182 is an effective divisor supported on the central fibre @xmath14 , we may consider the line bundle @xmath183 , which is said to be obtained from @xmath178 by _ twisting _ by @xmath182 .
for @xmath138 , its restriction to @xmath6 is the same as @xmath179 , but in general this is not the case for @xmath14 ; any such a line bundle @xmath184 is called a _
limit line bundle _ of @xmath179 for @xmath138 .
[ r : twist ] since @xmath185 is trivial , the divisor @xmath186 is linearly equivalent to @xmath3 .
so if @xmath182 is a divisor supported on @xmath14 , one has @xmath187 for all integers @xmath188 .
in particular if @xmath189 then @xmath190 . consider the subscheme @xmath191 of the _ relative hilbert scheme _ of curves of @xmath10 over @xmath11 , which is the zariski closure of the set of all curves @xmath192 , for @xmath138 .
we assume that @xmath191 is a component of the relative hilbert scheme , a condition satisfied if @xmath193 has no torsion , which will always be the case in our applications .
one has a natural projection morphism @xmath194 , which is a projective bundle over @xmath195 ; actually @xmath196 is isomorphic to @xmath197 over @xmath195 .
we call the fibre of @xmath198 over @xmath3 the _ limit linear system _ of @xmath199 as @xmath2 tends to @xmath3 , and we denote it by @xmath200 .
[ r : twist2 ] in general , _ the limit linear system is not a linear system_. one would be tempted to say that @xmath200 is nothing but @xmath201 ; this is the case if @xmath14 is irreducible , but it is in general no longer true when @xmath14 is reducible . in the latter case , there may be non zero sections of @xmath202 whose zero locus contains some irreducible component of @xmath14 , and accordingly points of @xmath203 which do not correspond to points in the hilbert scheme of curves ( see , e.g. , example [ ex : gh ] below ) . in any event , @xmath191 is a birational modification of @xmath204 , and @xmath200 is a suitable degeneration of the projective space @xmath199 , @xmath138 .
one has : [ l : lim - lin ] let @xmath205 be a flat and proper morphism , isomorphic to @xmath206 over @xmath143 , and such that @xmath207 is a zariski closed subset of the relative hilbert scheme of curves of @xmath10 over @xmath11
. then @xmath208 .
the two zariski closed subsets @xmath207 and @xmath209 are irreducible , and coincide over @xmath143 . in passing from @xmath206 to @xmath191 ,
one has to perform a series of blow
ups along smooth centres contained in the central fibre , which correspond to spaces of non trivial sections of some ( twisted ) line bundles which vanish on divisors contained in the central fibre .
the exceptional divisors one gets in this way give rise to components of @xmath200 , and may be identified with birational modifications of sublinear systems of twisted linear systems restricted to @xmath14 , as follows from lemma [ l : tg - id ] below .
we will see examples of this later ( the first one in example [ ex : gh ] ) .
[ l : tg - id ] [ pt : gnl - case ] let @xmath89 be a connected variety , @xmath178 a line bundle on @xmath89 , and @xmath210 a non zero global section of @xmath178 defining a subscheme @xmath39 of @xmath89 . then the projectivized tangent space to @xmath211 at @xmath212 canonically identifies with the _ restricted linear system _ @xmath213 also called the _ trace _ of @xmath214 on @xmath39 ( which in general is not the complete linear system @xmath215 ) .
[ subspace ] more generally , let @xmath216 be a linear subspace of @xmath211 with fixed locus scheme @xmath110 defined by the system of equations @xmath217 .
then the projectivized normal bundle of @xmath216 in @xmath211 canonically identifies with @xmath218 assertion comes from the identification of the tangent space of @xmath211 at @xmath219 with the cokernel of the injection @xmath220 , given by the multiplication by @xmath210 . as for , note that the normal bundle of @xmath216 in @xmath211 splits as a direct sum of copies of @xmath221 , hence the associated projective bundle is trivial .
then the proof is similar to that of .
[ see @xcite ] [ ex : gh ] consider a family of degree @xmath146 surfaces @xmath140 arising , as in example [ ex : pi3 ] , from a pencil generated by a general surface @xmath222 and by @xmath223 , where @xmath224 is a plane and @xmath110 a general surface of degree @xmath225 .
one has a semistable model @xmath163 of this family , as described in example [ ex : pi3 ] , with @xmath226 , where @xmath227 is the blow up of @xmath224 at the @xmath228 intersection points of @xmath222 with the smooth degree @xmath225 plane curve @xmath229 ( with exceptional divisors @xmath230 , for @xmath231 ) .
we let @xmath232 be the pull back by @xmath233 of @xmath234 , obtained by pulling back @xmath235 via the map @xmath236 .
the component @xmath209 of the hilbert scheme is gotten from the projective bundle @xmath237 , by blowing up the point of the central fibre @xmath238 corresponding to the 1dimensional space of non
zero sections vanishing on the plane @xmath224 .
the limit linear system @xmath200 is the union of @xmath239 , the blown up @xmath238 , and of the exceptional divisor @xmath240 , identified as the twisted linear system @xmath241 .
the corresponding twisted line bundle restricts to the trivial linear system on @xmath110 , and to @xmath242 on @xmath243 .
the components @xmath239 and @xmath244 of @xmath200 meet along the exceptional divisor @xmath245 of the morphism @xmath246 .
lemma [ l : tg - id ] shows that the elements of @xmath247 identify as the points of @xmath248 , whereas the plane @xmath249 is the set of elements @xmath250 containing the proper transform @xmath251 of @xmath18 on @xmath243 .
the corresponding element of @xmath252 is cut out on @xmath253 by the further component of @xmath254 , which is the pull back to @xmath243 of a line in @xmath224 .
let @xmath140 be a semistable family as in
[ s : setting ] , and @xmath178 be a line bundle on @xmath10 as in [ s : resolving ] .
we fix a non negative integer @xmath5 , and consider the locally closed subset @xmath255 of @xmath256 formed by all curves @xmath257 , for @xmath138 , such that @xmath118 is irreducible , nodal , and has exactly @xmath5 nodes .
we define @xmath258 ( resp .
@xmath259 ) as the zariski closure of @xmath255 in @xmath256 ( resp . in @xmath260 ) .
this is the _ relative severi variety _
the _ crude relative severi variety _ ) .
we may write @xmath261 , @xmath262 , and @xmath263 , rather than @xmath255 , @xmath258 , and @xmath259 , respectively .
we have a natural map @xmath264 . if @xmath265 , the fibre @xmath266 of @xmath267 over @xmath15 is the _
severi variety _
@xmath268 of @xmath5nodal curves in the linear system @xmath199 on @xmath6 , whose degree , independent on @xmath269 , we denote by @xmath270 ( or simply by @xmath271 ) .
we let @xmath272 ( or simply @xmath273 ) be the central fibre of @xmath264 ; it is the _
limit severi variety _ of @xmath268 as @xmath2 tends to @xmath3 .
this is a subscheme of the limit linear system @xmath200 , which , as we said , has been studied by various authors . in particular , one can describe in a number of situations its various irreducible components , with their multiplicities ( see [ s : zrr ] below ) .
this is what we will do for several families of quartic surfaces in @xmath21 . in a similar way
, one defines the _ crude limit severi variety _ @xmath274 ( or @xmath275 ) , sitting in @xmath276 .
[ rem : exp ] for @xmath138 , the _ expected dimension _ of the severi variety @xmath268 is @xmath277 .
we will always assume that the dimension of ( all components of ) @xmath268 equals the expected one for all @xmath138 .
this is a strong assumption , which will be satisfied in all our applications .
let @xmath140 be a family of degree @xmath146 surfaces in @xmath21 as in example [ ex : pi3 ] , and let @xmath163 be a semistable model of @xmath140 .
we consider the line bundle @xmath234 , defined as the pull back of @xmath235 via the natural map @xmath236 , and let @xmath278 be its pull back on @xmath279 .
we denote by @xmath280 ( resp .
@xmath281 ) , or simply @xmath282 , the limit severi variety @xmath283 ( resp .
@xmath284 ) . similar notation @xmath285 ( resp .
@xmath286 ) , or @xmath287 , will be used for the crude limit .
let again @xmath140 be a semistable family as in
[ s : setting ] , and @xmath178 a line bundle on @xmath10 as in
[ s : resolving ] . the local machinery developed in @xcite
enables us to identify the components of the limit severi variety , with their multiplicities . as usual
, we will suppose that @xmath14 has irreducible components @xmath165 , intersecting transversally along the double curves @xmath170 .
we will also assume that there are @xmath168 exceptional curves @xmath288 on @xmath14 , arising from a small resolution of an original family with singular total space , as discussed in
[ s : setting ] .
let @xmath289 be the set of sequences @xmath290 of non negative integers with only finitely many non vanishing terms .
we define two maps @xmath291 as follows : @xmath292 given a @xmath72-tuple @xmath293 , we set @xmath294 thus defining two maps @xmath295 .
given @xmath296 , we set @xmath297 given a subset @xmath298 , @xmath299 will denote its cardinality . [
d : zrr ] consider a divisor @xmath182 on @xmath10 , supported on the central fibre @xmath14 , i.e. a linear combination of @xmath165 .
fix @xmath300 , @xmath301 , and @xmath302 .
we let @xmath303 be the zariski locally closed subset in @xmath304 parametrizing curves @xmath118 such that : + [ no - double - curve ] @xmath118 neither contains any curve @xmath305 , with @xmath306 , nor passes through any triple point of @xmath14 ; + @xmath118 contains the exceptional divisor @xmath230 , with multiplicity 1 , if and only if @xmath307 , and has a node on it ; + @xmath308 has @xmath309 nodes on @xmath310 , for @xmath311 , off the singular locus of @xmath14 , and is otherwise smooth ; + [ c : tacnodes ] for every @xmath306 and @xmath312 , there are exactly @xmath313 points on @xmath305 , off the intersections with @xmath314 , at which @xmath118 has an _ @xmath188-tacnode _
( see below for the definition ) , with reduced tangent cone equal to the tangent line of @xmath305 there .
we let @xmath315 be the zariski closure of @xmath316 in @xmath317 .
recall that an _ @xmath188-tacnode _ is an @xmath318-double point , i.e. a plane curve singularity locally analytically isomorphic to the hypersurface of @xmath319 defined by the equation @xmath320 at the origin .
condition above requires that @xmath118 is a divisor having @xmath313 @xmath188th order tangency points with the curve @xmath305 , at points of @xmath305 which are not triple points of @xmath14 .
[ not:1 ] in practice , we shall not use the notation @xmath315 , but rather a more expressive one like , e.g. , @xmath321 for the variety parametrizing curves in @xmath304 , with two nodes on @xmath322 , one simple tacnode along @xmath323 , and containing the exceptional curve @xmath123 .
[ @xcite ] [ p : zrr ] let @xmath324 be as above , and set @xmath325 .
let @xmath95 be an irreducible component of @xmath326 . if + [ c : hilb ] the linear system @xmath317 has the same dimension as @xmath327 for @xmath2 , and + [ c : versal ] @xmath95 has ( the expected ) codimension @xmath5 in @xmath317 , + then @xmath95 is an irreducible component of multiplicity @xmath328 of the limit severi variety @xmath329 . [
r : smoothness ] same assumptions as in proposition [ p : zrr ] .
if there is at most one tacnode ( i.e. all @xmath313 but possibly one vanish , and this is equal to @xmath330 ) , the relative severi variety @xmath331 is smooth at the general point of @xmath95 ( see @xcite ) , and thus @xmath95 belongs to only one irreducible component of @xmath331 .
there are other cases in which such a smoothness property holds ( see @xcite ) .
if @xmath331 is smooth at the general point @xmath332 , the multiplicity of @xmath95 in the limit severi variety @xmath333 is the minimal integer @xmath188 such that there are local analytic _
@xmath188multisections _ of @xmath334 , i.e. analytic smooth curves in @xmath331 , passing through @xmath118 and intersecting the general fibre @xmath335 , @xmath138 , at @xmath188 distinct points . proposition [ p : zrr ] still does not provide a complete picture of the limit severi variety .
for instance , curves passing through a triple point of @xmath14 could play a role in this limit .
it would be desirable to know that one can always obtain a semistable model of the original family , where _ every _ irreducible component of the limit severi variety is realized as a family of curves of the kind stated in definition [ d : zrr ] .
[ def : reg - comps ] let @xmath140 be a semistable family as in
[ s : setting ] , @xmath178 a line bundle on @xmath10 as in [ s : resolving ] , and @xmath5 a positive integer .
regular part of the limit severi variety _
@xmath336 is the cycle in the limit linear system @xmath337 @xmath338 ( sometimes simply denoted by @xmath339 ) , where : + [ def:2 ] @xmath182 varies among all effective divisors on @xmath10 supported on the central fibre @xmath14 , such that @xmath340 for @xmath2 ; + @xmath341 denotes the set of all codimension @xmath5 irreducible components of a scheme @xmath39 .
proposition [ p : zrr ] asserts that the cycle @xmath342 is effective , with support disjoint in codimension @xmath330 from that of @xmath339 ( here , @xmath343 is the cycle associated to @xmath273 ) .
we call the irreducible components of the support of @xmath339 the _ regular components _ of the limit severi variety .
let @xmath344 be a semistable model of @xmath140 , and @xmath345 the pull
back on @xmath279 of @xmath178 .
there is a natural map @xmath346 , which induces a morphism @xmath347 .
[ def : good - model ] the semistable model @xmath344 is a @xmath5_good model _ of @xmath140 ( or simply _ good model _
, if it is clear which @xmath5 we are referring at ) , if the following equality of cycles holds @xmath348 note that the cycle @xmath349 is effective .
the family @xmath7 is said to be @xmath5_well behaved _ ( or simply _
well behaved _ ) if it has a @xmath5-good model .
a semistable model @xmath163 of @xmath140 as above is said to be @xmath5_absolutely good _ if @xmath350 as cycles in the relative hilbert scheme .
it is then a @xmath5good model both of itself , and of @xmath140 .
theorems [ t : tetrahedron ] and [ t : kummer ] will be proved by showing that the corresponding families of quartic surfaces are well behaved .
[ r : other - lim ] suppose that @xmath140 is @xmath5well behaved , with @xmath5good model @xmath351 .
it is possible that some components in @xmath352 are contracted by @xmath353 to varieties of smaller dimension , and therefore that their push forwards are zero .
hence these components of @xmath354 are _ not visible _ in @xmath355 .
they are however usually visible in the crude limit severi variety of another model @xmath139 , obtained from @xmath279 via an appropriate twist of @xmath178 .
the central fibre @xmath356 is then a flat limit of @xmath6 , as @xmath357 tends to @xmath3 , different from @xmath14 .
[ conj : goodness ] let @xmath7 be a semistable family of surfaces , endowed with a line bundle @xmath178 as above , and @xmath5 a positive integer . then : + * ( weak version ) * under suitable assumptions ( to be discovered ) , @xmath140 is @xmath5well behaved .
+ * ( strong version ) * under suitable assumptions ( to be discovered ) , @xmath140 has a @xmath5absolutely good semistable model . the local computations in @xcite
provide a criterion for absolute goodness : [ prop : criterion ] assume there is a semistable model @xmath351 of @xmath358 , with a limit linear system @xmath359 free in codimension @xmath360 of curves of the following types : + [ cat : first ] curves containing double curves of @xmath166 ; + [ cat : trip ] curves passing through a triple point of @xmath166 ; + [ cat : last ] non reduced curves . + if in addition , for @xmath361 as in definition [ d : zrr ] , every irreducible component of @xmath315 has the expected codimension in @xmath362 , then @xmath351 is @xmath5absolutely good , which implies that @xmath358 is @xmath5well behaved . unfortunately , in the cases we shall consider conditions in proposition [ prop : criterion ] are violated ( see propositions [ prop : limlin - tetra ] and [ prop : limlin - kummer ] ) , which indicates that further investigation is needed to prove the above conjectures .
the components of the various @xmath315 have nevertheless the expected codimension , and we are able to prove that our examples are well behaved , using additional enumerative information .
absolute goodness seems to be a property hard to prove , except when the dimension of the severi varieties under consideration is @xmath3 , equal to the expected one ( and even in this case , we will need extra enumerative information for the proof ) .
we note in particular that the @xmath5absolute goodness of @xmath351 implies that it is a @xmath5good model of every model @xmath139 , obtained from @xmath279 via a twist of @xmath345 corresponding to an irreducible component of the limit linear system @xmath363 . among the applications of the theory described above , there are the ones to enumerative problems , in particular to the computation of the degree @xmath271 of severi varieties @xmath364 , for the general member @xmath6 of a family @xmath140 as in [ s : setting ] , with @xmath178 a line bundle on @xmath10 as in [ s : resolving ] .
let @xmath138 be general , and let @xmath365 be the dimension of @xmath364 , which we assume to be @xmath366 . then @xmath271 is the number of points in common of @xmath364 with @xmath365 _ sufficiently general _ hyperplanes of @xmath199 .
given @xmath367 , @xmath368 \in |{\mathcal l}_t|\ \text{s.t.}\ x\in d \right\}\ ] ] is a plane in @xmath327 .
it is well known , and easy to check ( we leave this to the reader ) , that if @xmath369 are _ general points _ of @xmath6 , then @xmath370 are _ sufficiently general _ planes of @xmath199 with respect to @xmath364 . thus _
@xmath271 is the number of @xmath5nodal curves in @xmath199 passing through @xmath365 general points of @xmath6 . _ [ def : enne ] in the above setting
, let @xmath95 be an irreducible component of the limit severi variety @xmath336 , endowed with its reduced structure .
we let @xmath165 be the irreducible components of @xmath14 , and @xmath371 be such that @xmath372 .
fix a collection @xmath39 of @xmath373 general points on @xmath165 respectively .
the _ @xmath374degree _ of @xmath95 is the number @xmath375 of points in @xmath95 corresponding to curves passing through the points in @xmath39 .
note that in case @xmath376 , the above definition is somehow pointless : in this case , @xmath375 is simply the number of points in @xmath95 .
by contrast , when @xmath95 has positive dimension , it is possible that @xmath377 be zero for various @xmath374 s .
this is related to the phenomenon described in remark [ r : other - lim ] above .
we will see examples of this below . by flatness , the following result is clear : [ prop : deg ] let @xmath378 be a semistable model , and name @xmath379 the irreducible components of @xmath380 , in such a way that @xmath381 are the proper transforms of @xmath165 respectively . + for every @xmath382 such that @xmath383 , one has @xmath384 ( recall the definition of @xmath385 in proposition [ p : zrr ] ) .
+ if equality holds in for every @xmath386 as above , then @xmath378 is a @xmath5good model of @xmath140 endowed with @xmath178 .
in this section we collect a few results which we will use later .
first of all , for a general surface @xmath10 of degree @xmath146 in @xmath21 , we know from classical projective geometry the degrees @xmath387 of the severi varieties @xmath388 , for @xmath389 . for @xmath1 surfaces , this fits in a more general framework of known numbers ( see @xcite ) .
one has : [ p : deg - dual ] let @xmath10 be a general degree @xmath146 hypersurface in @xmath9 .
then @xmath390 for @xmath391 , these numbers are @xmath392 , @xmath393 , @xmath394 respectively . note that @xmath395 identifies with the _ dual surface _ @xmath396 .
the following is an extension of the computation of @xmath397 for surfaces with certain singularities .
this is well known and the details can be left to the reader . [
p : deg - dual - sing ] let @xmath10 be a degree @xmath146 hypersurface in @xmath9 , having @xmath398 and @xmath399 double points of type @xmath400 and @xmath401 respectively as its only singularities . then @xmath402 the following topological formula is well - known ( see , e.g. , ( * ? ? ? * lemme vi.4 ) ) .
[ l : pencil ] let @xmath403 be a surjective morphism of a smooth projective surface onto a smooth curve .
one has @xmath404 where @xmath405 and @xmath406 respectively denote the fibres of @xmath72 over the generic point of @xmath407 and a closed point @xmath408 , and @xmath409 is the set of points above which @xmath72 is not smooth .
as a side remark , note that it is possible to give a proof of the proposition [ p : deg - dual - sing ] based on lemma [ l : pencil ] .
this can be left to the reader .
propositions [ p : deg - dual ] and [ p : deg - dual - sing ] are sort of plcker formulae for surfaces in @xmath9 .
the next proposition provides analogous formulae for curves in a projective space of any dimension .
[ p : dj ] let @xmath410 be an irreducible , non degenerate curve of degree @xmath8 and of genus @xmath64 , the normalization morphism of which is unramified . let @xmath411 be a non - negative integer , and assume @xmath412 .
then the zariski closure of the locally closed subset of @xmath413 parametrizing _
@xmath414tangent _ hyperplanes to @xmath117 ( i.e. planes tangent to @xmath117 at @xmath414 distinct points ) has degree equal to the coefficient of @xmath415 in @xmath416 let @xmath417 be the normalization of @xmath117 , and let @xmath418 be the @xmath419 on @xmath420 defined as the pull back on @xmath420 of the hyperplane linear series on @xmath117 .
since @xmath398 is unramified , the degree of the subvariety of @xmath413 parametrizing @xmath414-tangent hyperplanes to @xmath117 is equal to the number of divisors having @xmath414 double points in a general sublinear series @xmath421 of @xmath418 .
this number is computed by a particular instance of de jonquires formula , see @xcite .
the last result we shall need is : [ l : tangency ] consider a smooth , irreducible curve @xmath18 , contained in a smooth surface @xmath10 in @xmath9 .
let @xmath422 be the irreducible curve in @xmath423 parametrizing planes tangent to @xmath10 along @xmath18 .
then the dual varieties @xmath424 and @xmath425 both contain @xmath422 , and do not intersect transversely at its general point .
clearly @xmath422 is contained in @xmath426 .
if either @xmath424 or @xmath425 are singular at the general point of @xmath422 , there is nothing to prove .
assume that @xmath424 and @xmath425 are both smooth at the general point of @xmath422 .
we have to show that they are tangent there .
let @xmath427 be general .
let @xmath94 be the tangent plane to @xmath10 at @xmath428 .
then @xmath429 is the general point .
now , the biduality theorem ( see , e.g. , ( * ? ? ?
* example 16.20 ) ) says that the tangent plane to @xmath424 and of @xmath425 at @xmath94 both coincide with the set of planes in @xmath21 containing @xmath428 , hence the assertion .
we consider a family @xmath358 of surfaces in @xmath9 , induced ( as in example [ ex : pi3 ] and in [ s : resolving ] ) by a pencil generated by a general quartic surface @xmath147 and a _ tetrahedron _
@xmath14 ( i.e. @xmath14 is the union of four independent planes , called the _ faces _ of the tetrahedron ) , together with the pull - back @xmath234 of @xmath235 .
we will prove that it is @xmath5well behaved for @xmath43 by constructing a suitable good model .
the plan is as follows .
we construct the good model in [ ssec : model ] , and complete its description in
[ s : identifications ] .
we then construct the corresponding limit linear system : the core of this is
[ s:4planes - limlin ] ; the paragraphs [ s : planes ] , [ s : rat4ics ] , and [ s : cubics ] , are devoted to the study of the geometry of the exceptional components of the limit linear system ( alternatively , of the geometry of the corresponding flat limits of the smooth quartic surfaces @xmath6 , @xmath357 ) ; eventually , we complete the description in
[ s : good ] .
we then identify the limit severi varieties in
[ s : tetra - concl ] .
the outline of the construction is as follows : + _ _ [ tetra : sm - resol ] we first make a small resolution of the singularities of @xmath10 as in example [ ex : pi3 ] ; + [ tetra : bs - change ] then we perform a degree @xmath27 base change ; + [ tetra : granchio - resol ] next we resolve the singularities of the total space arisen with the base change , thus obtaining a new semistable family @xmath430 ; + [ tetra : flop ] finally we will _ flop _ certain double curves in the central fibre @xmath431 , thus obtaining a new semistable family @xmath432 .
+ the central fibre of the intermediate family @xmath433 is pictured in figure [ f : granchio ] ( p. ; we provide a cylindrical projection of a real picture of @xmath431 , the dual graph of which is topologically an @xmath434 sphere ) , and the flops are described in figure [ f : flop ] ( p. ) .
the reason why we need to make the degree 6 base change is , intuitively , the following : a degree @xmath16 base change is needed to understand the contribution to the limit severi variety of curves passing through a
_ vertex _ ( i.e. a triple point ) of the tetrahedron , while an additional degree @xmath33 base change enables one to understand the contributions due to the _ edges _
( i.e. the double lines ) of the tetrahedron .
the singularities of the initial total space @xmath10 consist of four ordinary double points on each edge of @xmath14 .
we consider ( cf . example [ ex : pi3 ] ) the small resolution @xmath435 obtained by arranging for every edge the four @xmath436curves two by two on the two adjacent faces .
we call @xmath437 the new family .
let @xmath438 be the triple points of @xmath439 .
for each @xmath440 , we let @xmath441 be the irreducible component of @xmath439 which is opposite to the vertex @xmath442 : it is a plane blown - up at six points . for distinct @xmath443 , we let @xmath444 and @xmath445 be the two @xmath436-curves contained in @xmath441 and meeting @xmath446 .
we call @xmath447 and @xmath448 the two points cut out on @xmath441 by @xmath449 and @xmath450 respectively .
let now @xmath451 be the family obtained from @xmath437 by the base change @xmath452 .
the central fibre @xmath453 is isomorphic to @xmath166 , so we will keep the above notation for it . as a first step in the desingularization of @xmath454 , we perform the following sequence of operations for all @xmath455 .
the total space @xmath454 around @xmath442 is locally analytically isomorphic to the hypersurface of @xmath456 defined by the equation @xmath457 at the origin .
we blow - up @xmath454 at @xmath442 .
the blown up total space locally sits in @xmath458 .
let @xmath459 be the homogeneous coordinates in @xmath460 .
then the new total space is locally defined in @xmath458 by the equations @xmath461 the equation of the exceptional divisor ( in the exceptional @xmath462 of the blow up of @xmath463 ) is @xmath464 , hence this is the union of three planes meeting transversely at a point @xmath465 in @xmath460 . for @xmath466
distinct in @xmath467 , we call @xmath468 the exceptional planes meeting the proper transform of @xmath446 ( which , according to our conventions , we still denote by @xmath446 , see [ conv ] )
. the equation of the new family around the point @xmath465 given by @xmath469 is @xmath470 ( which sits in the affine chart @xmath471 ) .
next we blow - up the points @xmath465 , for @xmath440 .
the new exceptional divisor @xmath472 at each point @xmath465 is isomorphic to the cubic surface with equation @xmath470 in the @xmath9 with coordinates @xmath473 . note that @xmath472 has three @xmath401double points , at the vertices of the triangle @xmath474 .
next we have to get rid of the singularities of the total space along the double curves of the central fibre .
first we take care of the curves @xmath475 , for @xmath476 distinct in @xmath467 .
the model we constructed so far is defined along such a curve by an equation of the type @xmath477 , ( as it follows , e.g. , from the third equation in by setting @xmath478 ) .
the curve @xmath479 is defined by @xmath480 . if @xmath481 , the intersection point @xmath482 is cut out on @xmath479 by the hyperplane with equation @xmath483 .
away from the @xmath484 s , with @xmath481 , the points of @xmath479 are double points of type @xmath485 for the total space .
we blow up along this curve : this introduces new homogeneous coordinates @xmath486 , with new equations for the blow up
@xmath487 the exceptional divisor is defined by @xmath488 , and is the transverse union of two ruled surfaces : we call @xmath489 the one that meets @xmath490 , and @xmath491 the other .
the affine chart we are interested in is @xmath492 , where the equation is @xmath493 .
we then blow up along the curve @xmath494 , which gives in a similar way the new equation @xmath495 with the new coordinates @xmath496 .
the exceptional divisor consists of two ruled surfaces , and we call @xmath497 ( resp .
@xmath498 ) the one that meets @xmath489 ( resp .
@xmath491 ) . finally , by blowing - up along the curve @xmath499 , we obtain a new equation @xmath500 , with new coordinates @xmath501 .
the exceptional divisor is a ruled surface , with two @xmath401double points at its intersection points with the curves @xmath502 , with @xmath481 .
we call it either @xmath503 or @xmath504 , with no ambiguity . the final step of our desingularization process consists in blowing up along the twelve curves @xmath505 , with pairwise distinct @xmath506 . the total space is given along each of these curves by an equation of the type @xmath507 in the variables @xmath508 , obtained from the last equation in by setting @xmath478 .
the curve @xmath505 is defined by the local equations @xmath509 , which shows that they consist of @xmath401double points for the total space .
they also contain an @xmath401double point of @xmath503 and @xmath510 respectively .
a computation similar to the above shows that the blow up along these curves resolves all singularities in a single move .
the exceptional divisor over @xmath505 is the union of two transverse ruled surfaces : we call @xmath511 the one that meets @xmath512 , and @xmath513 the other . at this point
, we have a semistable family @xmath514 , whose central fibre is depicted in figure [ f : granchio ] : for each double curve we indicate its self intersections in the two components of the central fibre it belongs to .
this is obtained by applying the triple point formula ( see lemma [ l : triple - point ] ) .
of the degeneration into four planes , width=604 ] for our purposes , we need to further blow - up the total space along the twelve curves @xmath515 .
this has the drawback of introducing components with multiplicity two in the central fibre , namely the corresponding exceptional divisors . to circumvent this
, we will flop these curves as follows .
, width=453 ] let @xmath516 be the family obtained by blowing - up @xmath89 along the @xmath517 s .
we call @xmath518 ( or , unambiguously , @xmath519 ) the corresponding exceptional divisors : they appear with multiplicity two in the central fibre @xmath520 . by applying the triple point formula as in remark [ rem : tpf ] , one checks that the surfaces @xmath519 are all isomorphic to @xmath521
. moreover , it is possible to contract @xmath518 in the direction of the ruling cut out by @xmath511 and @xmath513 , as indicated on figure [ f : flop ] .
we call @xmath522 the contraction of the twelve divisors @xmath518 in this way , and @xmath523 the corresponding semistable family of surfaces
. even though @xmath524 is only a birational map , we have a birational morphism @xmath525 over @xmath11 . summarizing
, the irreducible components of the central fibre @xmath526 are the following : \(i ) the 4 surfaces @xmath441 , with @xmath527 .
each @xmath441 is a plane blown up at 6 + 3 points , and @xmath94 ( i.e. the pull - back of a general line in the plane , recall our conventions in [ conv ] ) is the restriction class of @xmath528 on @xmath441 . for @xmath529
, we set @xmath530 as indicated in figure [ fig : p ] .
in addition to the three @xmath436curves @xmath531 , we have on @xmath441 the six exceptional curves @xmath532 , for all @xmath533 , with @xmath532 intersecting @xmath534 at one point .
moreover , for @xmath535 , we have on @xmath534 the two points @xmath536 defined as the strict transform of the intersection @xmath537 in @xmath279 .
we will denote by @xmath538 the @xmath3dimensional scheme of length 6 given by @xmath539 .
we let @xmath540 be its defining sheaf of ideals .
, width=225 ] , width=268 ] \(ii ) the 24 surfaces @xmath541 , with @xmath443 distinct .
each of them is isomorphic to @xmath542 .
we denote by @xmath543 the ruling . note that the divisor class @xmath110 corresponds to the restriction of @xmath528 .
\(iii ) the 6 surfaces @xmath544 , with @xmath545 distinct . for each @xmath546 , we set @xmath547 and define similarly @xmath548 , @xmath549 , @xmath550 ( @xmath551 may be called @xmath552 without ambiguity ) .
this is indicated in figure [ fig : w ] .
a good way of thinking to the surfaces @xmath544 is to consider them as ( non minimal ) rational ruled surfaces , for which the two curves @xmath548 and @xmath553 are sections which do not meet , and the two rational chains @xmath554 are two disjoint reducible fibres of the ruling @xmath555 .
one has furthermore @xmath556 .
the surface @xmath544 has the length 12 anticanonical cycle @xmath557 cut out by @xmath558 , where we fixed @xmath146 and @xmath559 such that @xmath560 .
it therefore identifies with a plane blown up as indicated in figure [ f : wt ] : consider a general triangle @xmath561 in @xmath562 , with vertices @xmath563 , where @xmath564 is opposite to @xmath565 , etc . ; then blow up the three vertices @xmath566 , and call @xmath567 the corresponding exceptional divisors ; eventually blow up the six points @xmath568 , @xmath569 , and call @xmath570 the corresponding exceptional divisors .
the obtained surface has the anticanonical cycle @xmath571 which we identify term - by - term and in this order with the anticanonical cycle of @xmath544 . and @xmath572 as blown up planes ] we let @xmath94 be , as usual , ( the transform of ) a general line in the plane @xmath573 the ruling @xmath555 is the strict transform of the pencil of lines through the point @xmath564 , hence @xmath574
\(iv ) the 4 surfaces @xmath575 , with @xmath576 . here
we set @xmath577 also recall that @xmath578 for @xmath579 distinct .
this is indicated in figure [ fig : t ] .
, height=207 ] each @xmath575 identifies with a plane blown up as indicated in figure [ f : wt ] , as in the case of the @xmath544 s : it has the length 12 anticanonical cycle @xmath580 ( where we fixed indices @xmath581 such that @xmath582 ) cut out by @xmath583 on @xmath575 , which we identify term - by - term and in this order with the anticanonical cycle .
this yields @xmath584 we have on @xmath572 the proper transform of a pencil of ( bitangent ) conics that meet the curves @xmath585 and @xmath551 in one point respectively , and do not meet any other curve in the anticanonical cycle : we call this pencil @xmath586 , and we have @xmath587 the restriction of @xmath528 on @xmath572 is trivial .
\(v ) the 12 surfaces @xmath588 , with @xmath589 distinct . each of them identifies with a blown up plane as indicated in figure [ f : a ] .
it is equipped with the ruling @xmath590 , the members of which meet the curves @xmath531 and @xmath591 at one point respectively , and do not meet any other curve in the length 8 anticanonical cycle cut out by @xmath592 on @xmath593 .
the restriction of @xmath528 on @xmath593 is trivial . as a blown - up plane , width=415 ] \(vi ) the 24 surfaces @xmath594 with @xmath595 distinct .
these are all copies of @xmath562 , on which the restriction of @xmath528 is trivial . according with the general principles stated in
[ s : resolving ] , we shall now describe the limit linear system of @xmath596 as @xmath2 tends to @xmath3 .
this will suffice for the proof , presented in
[ s : good ] , that @xmath597 is a @xmath5good model for @xmath43 .
we start with @xmath598 , which is a @xmath21bundle over @xmath11 , whose fibre at @xmath136 is @xmath599 .
we set @xmath600 , and @xmath601 ; note that @xmath602 .
we will often use the same notation to denote a divisor ( or a divisor class ) on the central fibre and its restriction to a component of the central fibre , if this does not cause any confusion .
we will proceed as follows : + _ _ [ tetra : faces ] we first blow up
@xmath204 at the points @xmath603 corresponding to the irreducible components @xmath441 of @xmath14 , for @xmath440 ( the new central fibre then consists of @xmath604 blown up at four independent points , plus the four exceptional @xmath21 s ) ; + [ tetra : edges ] next , we blow up the total space along the proper transforms @xmath605 of the six lines of @xmath606 joining two distinct points @xmath607 , with @xmath443 , corresponding to pencils of planes with base locus an edge of @xmath14 ( the new central fibre is the proper transform of the previous one , plus the six exceptional @xmath608 s ) ; + [ tetra : vertices ] finally , we further blow up along the proper transforms of the planes @xmath609 corresponding to the webs of planes passing through the vertices @xmath610 of @xmath14 , for @xmath611 ( this adds four more exceptional divisors to the central fibre , for a total of fifteen irreducible components ) .
+ in other words , we successively blow
up @xmath204 along all the cells of the tetrahedron dual to @xmath14 in @xmath612 , by increasing order of dimension .
each of these blow ups will be interpreted in terms of suitable twisted linear systems as indicated in remark [ r : twist2 ] .
it will then become apparent that every point in the central fibre of the obtained birational modification of @xmath204 corresponds to a curve in @xmath613 ( see [ s : good ] ) , and hence that this modification is indeed the limit linear system @xmath200 . in @xmath614
there is for each @xmath440 the 1dimensional subspace of sections vanishing on @xmath441 , which corresponds to the sections of @xmath615 vanishing on the plane @xmath441 .
as indicated in remark [ r : twist2 ] , in order to construct the limit linear system , we have to blow up the corresponding points @xmath616 .
let @xmath617 be this blow up , and call @xmath618 , @xmath619 , the exceptional divisors .
each @xmath618 is a @xmath9 , and can be interpreted as the trace of the linear system @xmath620 on @xmath431 ( see lemma [ l : tg - id ] and example [ ex : gh ] ) .
however , any section of @xmath621 still vanishes on components of @xmath526 different from @xmath441 . by subtracting all of them with the appropriate multiplicities ( this computation is tedious but not difficult and
can be left to the reader ) , one sees that @xmath618 can be identified as the linear system @xmath622 , where @xmath623 with the notation introduced in [ s : identifications ] , one has : [ lem : comput ] the restriction class of @xmath624 to the irreducible components of @xmath526 is as follows : + [ lem : comput : p_i ] on @xmath441 , we find @xmath625 ; + [ i ] on @xmath446 , @xmath626 , we find @xmath627 ; + [ lem : comput : w ] for each @xmath626 , we find @xmath628 on each of the surfaces @xmath629 , @xmath630 , @xmath544 , @xmath631 , @xmath632 .
+ on the remaining components the restriction is trivial .
this is a tedious but standard computation . as a typical sample we prove , and leave the remaining cases to the reader .
set @xmath633 .
then , recalling and , we see that the restriction of @xmath624 to @xmath544 is the line bundle determined by the divisor class @xmath634 from this , we deduce that @xmath635 identifies with its restriction to @xmath441 : [ prop : descrl_i ] there is a natural isomorphism @xmath636 for each @xmath626 , the restriction of @xmath635 to @xmath446 has @xmath627 as its only member .
this implies that its restriction to @xmath632 has only one member as well , which is the sum of the two curves in @xmath637 intersecting @xmath449 and @xmath450 respectively . on @xmath631
, we then only have the sum of the two curves in @xmath637 intersecting the two curves on @xmath632 respectively , and so on on @xmath544 , @xmath630 , and @xmath629 .
now the two curves on @xmath629 impose the two base points @xmath638 and @xmath639 to the restriction of @xmath635 to @xmath441 . the right hand side in being @xmath16dimensional
, this ends the proof with of lemma [ lem : comput ] .
next , we consider the blow up @xmath640 along the proper transforms @xmath605 of the six lines of @xmath203 joining two distinct points @xmath607 , with @xmath443 , corresponding to the pencils of planes in @xmath22 respectively containing the lines @xmath641 .
the exceptional divisors are isomorphic to @xmath642 ; we call them @xmath643 , @xmath644 . arguing as in step ( [ tetra : faces ] ) and leaving the details to the reader
, we see that @xmath643 is in a natural way a birational modification ( see [ s : rat4ics ] below ) of the complete linear system @xmath645 , where @xmath646 we will denote by @xmath647 the two indices in @xmath648 , and go on using the notations introduced in [ s : identifications ] .
[ lem : comput2 ] the restriction class of @xmath649 to the irreducible components of @xmath613 is as follows : + [ i0 ] on @xmath650 ( resp .
@xmath490 ) we find @xmath651 ( resp .
@xmath652 ) ; + [ i00 ] on each of the surfaces @xmath653 , @xmath654 , @xmath504 , @xmath655 , and @xmath656 , we find @xmath110 ; + [ i1 ] on @xmath657 ( resp .
@xmath658 ) we find @xmath659 ( resp .
@xmath660 ) ; + [ ii1 ] on @xmath572 ( resp .
@xmath661 ) , we find @xmath662 ( resp .
@xmath663 ) ; + [ iii1 ] on @xmath441 ( resp .
@xmath446 ) , we find @xmath664 ( resp .
@xmath627 ) ; + [ iv1 ] on @xmath665 , we find @xmath628 ; + [ v1 ] on @xmath544 , with @xmath94 as in , we find @xmath666 [ vi1 ] on the remaining components the restriction is trivial . as for lemma [ lem : comput ] , this is a tedious but not difficult computation .
again we make a sample verification , proving above .
the restriction class is @xmath667 which , by taking into account the identification of figure [ f : wt ] , i.e. with and , is easily seen to be equivalent to the required class .
let @xmath668 be the two points cut out on @xmath544 by the two connected chains of curves in @xmath669 meeting @xmath670 respectively .
we let @xmath671 be the two points defined in a similar fashion by starting with @xmath672 .
define the @xmath3-cycle @xmath673 on @xmath544 , and let @xmath674 be its defining sheaf of ideals .
[ prop : descrl_ij ] there is a natural isomorphism between @xmath675 and its restriction to @xmath544 , which is the @xmath16dimensional linear system @xmath676 where we set @xmath677 , and @xmath94 as in .
consider a triangle @xmath561 in @xmath562 , with vertices @xmath563 , where @xmath564 is opposite to @xmath565 , etc .
consider the linear system @xmath678 of quartics with a double point at @xmath564 , two simple base points infinitely near to @xmath564 not on @xmath679 and @xmath680 , two base points at @xmath681 and @xmath682 with two infinitely near base points along @xmath680 and @xmath679 respectively , two more base points along @xmath565 .
there is a birational transformation of @xmath544 to the plane ( see figure [ f : wt ] ) mapping to a linear system of type @xmath678 .
one sees that two independent conditions are needed to impose to the curves of @xmath678 to contain the three lines @xmath561 and the residual system consists of the pencil of lines through @xmath564 .
this proves the dimensionality assertion ( see [ s : rat4ics ] below for a more detailed discussion ) .
consider then the restriction of @xmath675 to the chain of surfaces @xmath683 by taking into account , , and , of lemma [ lem : comput2 ] , we see that each divisor @xmath117 of this system determines , and is determined , by its restriction @xmath684 on @xmath544 , since @xmath117 consists of @xmath684 plus four rational tails matching it .
the remaining components of @xmath526 on which @xmath675 is non trivial , all sit in the chain @xmath685 the restrictions of @xmath675 to each irreducible component of this chain is a base point free pencil of rational curves , hence @xmath675 restricts on to the @xmath330dimensional system of connected chains of rational curves in these pencils : we call it @xmath686 . given a curve in @xmath675 , it cuts @xmath572 and @xmath687 in one point each , and there is a unique chain of rational curves in @xmath686 matching these two points . finally , we consider the blow - up @xmath688 along the proper transforms of the three planes that are strict transforms of the webs of planes in @xmath22 containing a vertex @xmath610 , with @xmath689 . for each @xmath146 , the exceptional divisor
@xmath690 is a birational modification ( see [ s : cubics ] below ) of the complete linear system @xmath691 , where @xmath692 [ lem : comput3 ] the restriction class of @xmath693 to the irreducible components of @xmath613 is as follows : + [ i2 ] on @xmath441 , @xmath694 , we find @xmath695 ; + [ ii2 ] on @xmath696 , @xmath694 , we find @xmath697 ; + [ iii2 ] on @xmath650 , as well as on the chains @xmath698 , @xmath694 , we find the restriction class of @xmath202 ; + [ iii3 ] on @xmath572 , we find @xmath699 with @xmath582 , and @xmath94 as in ; + [ iii4 ] on the remaining components it is trivial .
we limit ourselves to a brief outline of how things work for @xmath572 .
the restriction class is @xmath700 which is seen to be equal to the required class with the identification of figures [ f : wt ] and [ fig : t ] , i.e. with @xmath94 as in .
[ prop : descr2 ] there is a natural isomorphism between @xmath701 and its restriction to @xmath572 , which is the @xmath16dimensional linear system @xmath702 where we set @xmath582 , and @xmath94 as in .
this is similar ( in fact , easier ) to the proof of proposition [ prop : descrl_ij ] , so we will be sketchy here .
the dimensionality assertion will be discussed in [ s : cubics ] below . for each @xmath694 , the restriction of @xmath701 to each irreducible component of the chain
@xmath703 is a base point free pencil of rational curves , and @xmath701 restricts on to the @xmath330dimensional system of connected chains of rational curves in these pencils , that we will call @xmath704 .
now the general member of @xmath701 consists of a curve in @xmath705 , which uniquely determines three chains of rational curves in @xmath704 , @xmath694 , which in turn determine a unique line in @xmath706 .
let @xmath37 be three independent lines in @xmath562 , and consider a @xmath3dimensional scheme @xmath39 cut out on @xmath40 by a general quartic curve .
consider the linear system @xmath707 of plane quartics containing @xmath39 .
this is a linear system of dimension 3 . indeed containing
the union of the three lines @xmath37 is one condition for the curves in @xmath707 and the residual system is the @xmath33dimensional complete linear system of all lines in the plane .
proposition [ prop : descrl_i ] shows that @xmath635 can be identified with a system of type @xmath707 .
we denote by @xmath708 ( or simply by @xmath210 ) the rational map determined by @xmath635 and by @xmath121 its image , which is the same as the image of the plane via the rational map determined by the linear system @xmath707 .
[ prop : monoid ] the map @xmath709 is birational , and @xmath121 is a _ monoid _ quartic surface , with a triple point @xmath72 with tangent cone consisting of a triple of independent planes through @xmath72 , and with no other singularity .
the triple point @xmath710 is the image of the curve @xmath711 ( alternatively , of the sides of the triangle @xmath37 ) . by subtracting @xmath117
to @xmath635 one gets a homaloidal net , mapping to the net of lines in the plane .
this proves the assertion .
[ rem : monoid ] the image of @xmath712 by the complete linear system @xmath713 provides a model @xmath714 of the initial family @xmath7 , such that the corresponding flat limit of @xmath715 with @xmath13 , is @xmath716 the quartic monoid image of the face @xmath441 of the tetrahedron via @xmath210 .
the map @xmath717 contracts all other irreducible components of @xmath613 to the triple point of the monoid .
theorem [ t : triangle ] says that the degree of the dual surface of the monoid @xmath121 is 21 .
the strict transform of @xmath718 in @xmath719 ( which we still denote by @xmath718 , see [ conv ] ) can be identified as a blow up of @xmath720 : first blow up the three points corresponding to the three non reduced curves @xmath721 , @xmath722 , @xmath723 . then blow up
the proper transforms of the three pencils of lines with centres at @xmath724 plus the fixed part @xmath40 .
we will interpret this geometrically in
[ s : good ] , using lemma [ l : tg - id ] .
next , we need to study some of the geometric properties of the linear systems @xmath675 as in proposition [ prop : descrl_ij ] .
consider the rational map @xmath725 ( or simply @xmath198 ) determined by @xmath726 .
alternatively , one may consider the rational map , with the same image @xmath182 ( up to projective transformations ) , determined by the planar linear system @xmath678 of quartics considered in the proof of proposition [ prop : descrl_ij ] .
[ p : w - quartic ] the map @xmath198 is birational onto its image , which is a quartic surface @xmath727 , with a double line @xmath118 , and two triple points on @xmath118 .
first we get rid of the four base points in @xmath728 by blowing them up and taking the proper transform @xmath729 of the system .
let @xmath730 be this blow up , and let @xmath731 ( resp .
@xmath732 ) be the two @xmath436curves that meet @xmath553 ( resp .
@xmath548 ) .
the strict transform @xmath733 , has self intersection @xmath24 .
set , as usual , @xmath677 and consider the curves @xmath734 one has @xmath735 by mapping @xmath736 to @xmath544 , and this to the plane as in figure [ f : wt ] with and identified , one sees that @xmath737 goes to the line @xmath565 and @xmath738 to the union of the two lines @xmath739 .
the considerations in the proof of proposition [ prop : descrl_ij ] show that @xmath740 has no base points on @xmath741 ( i.e. , @xmath678 has only the prescribed base points along the triangle @xmath742 ) .
on the other hand , the same considerations show that the base points of @xmath740 may only lie on @xmath741 .
this shows that @xmath740 is base points free , and the associated morphism @xmath743 contracts @xmath737 and @xmath738 to points @xmath744 and @xmath745 respectively .
the points @xmath744 and @xmath745 are distinct , since subtracting the line @xmath565 from the planar linear system @xmath678 does not force subtracting the whole triangle @xmath742 to the system . by subtracting @xmath737 from @xmath740 ,
the residual linear system is a linear system of rational curves with self intersection 1 , mapping @xmath544 birationally to the plane .
indeed , this residual linear system corresponds to the residual linear system of @xmath565 with respect to @xmath678 , which is the linear system of plane cubics , with a double point at @xmath564 , two simple base points infinitely near to @xmath564 not on @xmath679 and @xmath680 , two base points at @xmath681 and @xmath682 , and this is a homaloidal system .
this shows that @xmath744 is a triple point of @xmath182 and that @xmath746 is birational .
the same for @xmath745 .
finally @xmath746 maps ( the proper transforms of ) @xmath551 and @xmath747 both to the unique line @xmath118 containing @xmath744 and @xmath745 . the subpencil of @xmath675 corresponding to planes in @xmath9 that contain the line @xmath118 corresponds to the subpencil of curves in @xmath678 with the triangle @xmath742 as its fixed part , plus the pencil of lines through @xmath564 . in this subpencil
we have two special curves , namely @xmath748 and @xmath749 .
this shows that the tangent cone to @xmath182 at the general point of @xmath118 is fixed , formed by two planes . the image of @xmath712 via the complete linear system @xmath750 provides a model @xmath714 of the initial family @xmath7 , such that the corresponding flat limit of @xmath715 with @xmath13 , is @xmath751 the image of @xmath544 via @xmath198 .
the map @xmath717 contracts the chain to the double line of @xmath182 , and the two connected components of @xmath752 minus the chain ( cf . figure [ f : granchio ] ) to the two triple points of @xmath182 respectively .
[ cor : excep ] the exceptional divisor @xmath643 of @xmath640 is naturally isomorphic to the blow up of the complete linear system @xmath753 along its subpencil corresponding to planes in @xmath9 containing the line @xmath118 .
this is a reformulation of the description of @xmath754 ( cf . step in [ s:4planes - limlin ] above ) , taking into account propositions [ prop : descrl_ij ] and [ p : w - quartic ] .
the divisor @xmath755 is a @xmath756 , and its structure of @xmath108bundle over @xmath149 is the minimal resolution of indeterminacies of the rational map @xmath757 to its intersection point with @xmath118 .
the next proposition provides an identification of the general fibres of @xmath643 over @xmath758 as certain linear systems .
[ p : projectw ] the projection of @xmath182 from a general point of @xmath118 is a double cover of the plane , branched over a sextic @xmath407 which is the union @xmath759 of a quartic @xmath760 with a node @xmath72 , and of its tangent cone @xmath761 at @xmath72 , such that the two branches of @xmath760 at @xmath72 both have a flex there ( see figure [ f : diramazione ] ; the intersection @xmath762 is concentrated at the double point @xmath72 , for @xmath763 ) .
, width=529 ] let us consider a double cover of the plane as in the statement .
it is singular .
following @xcite , we may obtain a resolution of singularities as a double cover of a blown up plane with non singular branch curve .
we will then observe that it identifies with @xmath764 blown up at two general _ conjugate _
points on @xmath551 and @xmath747 respectively ( here _ conjugate _ means that the two points are mapped to the same point @xmath428 of @xmath118 by @xmath746 ) .
we will denote by @xmath765 the surface @xmath736 blown up at two such points , and by @xmath766 the two exceptional divisors .
first note that our double plane is rational , because it has a pencil of rational curves , namely the pull - back of the pencil of lines passing through @xmath72 ( eventually this will correspond to the pencil of conics cut out on @xmath182 by the planes through @xmath118 ) . in order to resolve the singularities of the branch curve ( see figure [ f : diramazione ] ) , we first blow
up @xmath72 , pull back the double cover and normalize it . since @xmath72 has multiplicity @xmath24 , which is even , the exceptional divisor @xmath112 of the blow up does not belong to the branch curve of the new double cover , which is the proper transform @xmath407 ( still denoted by @xmath407 according to our general convention ) .
next we blow up the two double points of @xmath407 which lie on @xmath112 , and repeat the process .
again , the two exceptional divisors @xmath767 do not belong to the branch curve .
finally we blow up the two double points of @xmath407 ( which lie one on @xmath123 one on @xmath129 , off @xmath112 ) , and repeat the process .
once more , the two exceptional divisors @xmath768 do not belong to the branch curve which is the union of @xmath760 , @xmath769 and @xmath770 ( which denote here the proper transforms of the curves with the same names on the plane ) .
this curve is smooth , so the corresponding double cover is smooth .
the final double cover has the following configuration of negative curves : @xmath769 ( resp .
@xmath770 ) is contained in the branch divisor , so over it we find a @xmath436-curve ; @xmath771 ( resp .
@xmath772 ) meets the branch divisor at two points , so its pull - back is a @xmath773-curve ; @xmath123 ( resp .
@xmath129 ) does not meet the branch divisor , so its pull - back is the sum of two disjoint @xmath773-curves ; similarly , the pull - back of @xmath112 is the sum of two disjoint @xmath774-curves .
in addition , there are four lines through @xmath72 tangent to @xmath760 and distinct from @xmath769 and @xmath770 . after the resolution
, they are curves with self - intersection @xmath3 and meet the branch divisor at exactly one point with multiplicity @xmath33 .
the pull - back of any such a curve is the transverse union of two @xmath436-curves , each of which meets transversely one component of the pull - back of @xmath112 .
this configuration is precisely the one we have on @xmath765 , after the contraction of the four @xmath436-curves @xmath549 , @xmath775 , @xmath776 , and @xmath777 .
moreover , the pull - back of the line class of @xmath108 is the pull back to @xmath778 of @xmath779 .
[ cor : projectw ] in the general fibre of the generic @xmath562 bundle structure of @xmath780 , the severi variety of @xmath330nodal ( resp .
@xmath33nodal ) irreducible curves is an irreducible curve of degree @xmath781 ( resp .
the union of @xmath28 distinct points ) .
this follows from the fact that the above mentioned severi varieties are respectively the dual curve @xmath782 of a plane quartic as in proposition [ p : projectw ] , and the set of ordinary double points of @xmath782 .
one computes the degrees using plcker formulae .
here we study some geometric properties of the linear systems @xmath783 appearing in the third step of [ s:4planes - limlin ] . consider a triangle @xmath561 in @xmath562 , with vertices @xmath563 , where @xmath564 is opposite to @xmath565 , etc .
consider the linear system @xmath784 of cubics through @xmath563 and tangent there to @xmath785 respectively . by proposition [ prop : descr2 ]
, there is a birational transformation of @xmath575 to the plane ( see figure [ f : wt ] ) mapping @xmath783 to @xmath784 .
we consider the rational map @xmath786 ( or simply @xmath787 ) determined by the linear system @xmath788 , or , alternatively , the rational map , with the same image @xmath789 ( up to projective transformations ) , determined by the planar linear system @xmath784 .
the usual notation is @xmath790 .
[ p : cubics ] the map @xmath791 is a birational morphism , and @xmath789 is a cubic surface with three double points of type @xmath401 as its only singularities .
the minimal resolution of @xmath789 is the blow down of @xmath575 contracting the @xmath436curves @xmath792 .
this cubic contains exactly three lines , each of them containing two of the double points .
the linear system @xmath784 is a system of plane cubics with six simple base points , whose general member is clearly irreducible .
this implies that @xmath791 is a birational morphism and @xmath789 is a cubic surface . the linear system @xmath783 contracts the three chains of rational curves @xmath793 which map in the plane to the sides of the triangle @xmath561 . by contracting the @xmath436curves
@xmath794 , @xmath795 , @xmath796 , the three curves @xmath797 are mapped to three @xmath773cycles contracted by @xmath787 to double points of type @xmath401 .
the rest follows from the classification of cubic hypersurfaces in @xmath9 ( see , e.g. , @xcite ) .
the three lines on @xmath789 are the images via @xmath787 of the three exceptional divisors @xmath798 .
we now see that the image of @xmath712 by the complete linear system @xmath799 provides a model @xmath714 of the initial family @xmath7 , such that the corresponding flat limit of @xmath715 with @xmath13 , is @xmath800 , where @xmath789 is the image of @xmath572 via @xmath787 , and @xmath224 is the plane in @xmath9 through the three lines contained in @xmath789 , image of @xmath650 by the map associated to @xmath701 .
the three other faces of the initial tetrahedron @xmath14 are contracted to the three lines in @xmath789 respectively .
[ p : cubics - gauss ] the dual surface @xmath801 to @xmath789 is itself a cubic hypersurface with three double points of type @xmath401 as its only singularities .
indeed , the gauss map @xmath802 fits into the commutative diagram @xmath803_{\phi } \ar[dr]^{\check{\phi } } & \\ t \ar@{-->}[rr]_{\gamma_t } & & \check{t } } \ ] ] where @xmath804 is the morphism associated to the linear system @xmath805 which is mapped to the linear system @xmath806 of cubics through @xmath563 and tangent there to @xmath807 respectively , by the birational map @xmath808 identifying @xmath701 with @xmath784 . the dual hypersurface @xmath809 has degree @xmath16 by proposition [ p : deg - dual - sing ] .
let @xmath72 be a double point of @xmath789 .
the tangent cone to @xmath789 at @xmath72 is a rank 2 quadric , with vertex a line @xmath810 .
a local computation shows that the limits of all tangent planes to @xmath789 at smooth points tending to @xmath72 are planes through @xmath810 .
this means that @xmath802 is not well defined on the minimal resolution of @xmath789 , which is the blow down of @xmath575 contracting the @xmath436curves @xmath792 , its indeterminacy points being exactly the three points images of these curves .
the same local computation also shows that @xmath802 is well defined on @xmath575 , hence @xmath802 fits in the diagram as stated .
in @xmath784 there are the three curves @xmath811 , which implies that for any given line @xmath812 there is a plane @xmath813 in @xmath9 tangent to @xmath789 at the general point of @xmath814 ( actually one has @xmath815 ) .
then @xmath802 contracts each of the three lines contained in @xmath789 to three different points , equivalently @xmath804 contracts to three different points the three curves @xmath816 . being @xmath817 a ( weak ) del pezzo surface , this implies that @xmath804 must contract the three chains of rational curves @xmath818 , @xmath819 , and @xmath820 , because they have @xmath3 intersection with the anticanonical system , and the rest of the assertion follows . recalling the description of @xmath821 ,
one can realize @xmath690 as a birational modification of @xmath822 : first blow up the point corresponding to the plane containing the three lines of @xmath789 , then blow up the strict transforms of the three lines in @xmath823 corresponding to the three pencils of planes respectively containing the three lines of @xmath789 .
notice that @xmath690 has a structure of @xmath149bundle on the blow up of @xmath562 at three non coplanar points , as required .
alternatively , we have in @xmath784 the four curves @xmath824 , corresponding to four independent points @xmath825 of @xmath784 .
then @xmath690 is the blow up of @xmath784 at @xmath826 , further blown
up along the proper transforms of the lines @xmath827 , @xmath828 , and @xmath829 . via the map
@xmath830 , the projection of the @xmath111bundle structure corresponds to the projection of @xmath784 from @xmath826 to the plane spanned by @xmath831 .
this will be interpreted using lemma [ l : tg - id ] in [ s : good ] below .
we are now ready to prove : [ prop : limlin - tetra ] the limit linear system of @xmath832 as @xmath833 tends to @xmath3 is @xmath834 .
the identification of @xmath835 as @xmath209 will follow from the fact that every point in @xmath834 corresponds to a curve in @xmath613 ( see lemma [ l : lim - lin ] ) . having the results of [ s:4planes - limlin][s : cubics ] at hand ,
we are thus left with the task of describing how the various components of the limit linear system intersect each other .
we carry this out by analyzing , with lemma [ l : tg - id ] , the birational modifications operated on the components @xmath612 , @xmath618 , @xmath836 , and @xmath837 , during the various steps of the construction of @xmath835 ( see [ s:4planes - limlin ] ) . in @xmath838 , the strict transform of @xmath612 ( which we shall go on calling @xmath612 , according to the conventions set in
[ conv ] ) is the blow up of @xmath839 at the four points corresponding to the faces of @xmath14 . for each @xmath840 ,
the corresponding exceptional plane is the intersection @xmath841 , and it identifies with the subsystem of @xmath635 consisting of curves @xmath842 together with six rational tails respectively joining @xmath672 to @xmath536 , @xmath626 . for each @xmath843 , the intersection @xmath844 identifies as the exceptional @xmath521 of both the blow up of @xmath845 along the line @xmath605 , and the blow up
@xmath846 described in corollary [ cor : excep ] . as a consequence ,
it parametrizes the curves @xmath847 where @xmath117 is a chain in @xmath686 , and @xmath848 is the proper transform by @xmath849 of a conic through the two triple points of @xmath182 ( cf . proposition [ p : w - quartic ] ) , together with four rational tails respectively joining @xmath670 and @xmath672 to @xmath850 and @xmath851 .
the two components @xmath117 and @xmath852 are independent one from another , and respectively move in a @xmath330dimensional linear system .
the intersection @xmath853 is a @xmath108 . in @xmath836
, it identifies as the proper transform via @xmath854 of the linear system of curves @xmath855 while in @xmath618 it is the exceptional divisor of the blow up of @xmath856 at the point corresponding to the curve @xmath857 it follows that it parametrizes sums of a curve as in , plus the special member of @xmath686 consisting of double curves of @xmath613 and joining the two points @xmath858 and @xmath859 . for each @xmath860 , the intersection
@xmath861 is a @xmath108 blown up at three non colinear points . seen in @xmath612 , it identifies as the blow up of the web of planes in @xmath862 passing through the vertex @xmath146 of @xmath14 , at the three points corresponding to the faces of @xmath14 containing this very vertex . in @xmath837 on the other hand , it is the strict transform of the exceptional @xmath108 of the blow up
@xmath863 at the point @xmath864 $ ] .
it therefore parametrizes the curves @xmath865 where @xmath105 is a line in @xmath650 , together with three rational tails joining respectively @xmath866 to @xmath591 , @xmath694 . for @xmath694
, @xmath867 is a @xmath868 , identified as the exceptional divisor of both the blow up @xmath869 along the strict transform of the line parametrizing planes in @xmath870 containing the line @xmath871 , and the blow up of @xmath872 along the strict transform of the line parametrizing curves @xmath873 it therefore parametrizes sums of @xmath874 ( where @xmath875 , and the second summand is a member of @xmath705 ) , plus the fixed part @xmath876 , where @xmath877 is the special member of @xmath878 consisting of double curves of @xmath613 and joining the two points @xmath879 on @xmath650 and @xmath880 on @xmath572 , for each @xmath881 , with @xmath882 such that @xmath883 .
the two curves @xmath852 and @xmath117 are independent one from another , and respectively move in a @xmath330dimensional linear system . for each @xmath884
, @xmath885 is an @xmath886 , and identifies as the blow up of the plane in @xmath701 corresponding to divisors in @xmath823 passing through the double point @xmath887 , at the point @xmath888 $ ] ; it also identifies as the exceptional divisor of the blow up of @xmath889 along the @xmath149 corresponding to the curves as in , with @xmath852 the only member of @xmath890 containing @xmath551 .
we only need to identify the curves parametrized by the exceptional curve of this @xmath886 ; they are as in , with @xmath105 corresponding to a line in the pencil @xmath891 , @xmath892 . in conclusion
, @xmath835 is an irreducible zariski closed subset of the relative hilbert scheme of @xmath893 over @xmath11 , and this proves the assertion .
we shall now identify the regular parts of the limit severi varieties @xmath894 for @xmath43 ( see definition [ def : reg - comps ] ) . to formulate the subsequent statements , we use notation [ not:1 ] and the notion of @xmath374degree introduced in [ s : enumeration ] .
we will be interested in those @xmath374 that correspond to a choice of @xmath895 general base points on the faces @xmath441 of @xmath14 , with @xmath896 .
these choices can be identified with @xmath24tuples @xmath897 with @xmath898 ( by choosing @xmath899 general points on @xmath441 ) .
the vector @xmath374 is non zero only if @xmath900 . for @xmath47 ( resp . for @xmath49 ) , to give @xmath374 is equivalent to give two indices @xmath901 ( resp .
an @xmath902 ) : we let @xmath903 ( resp .
@xmath904 ) be the @xmath24tuple corresponding to the choice of general base points on @xmath441 and @xmath905 respectively if @xmath906 , and of two general base points on @xmath441 if @xmath907 ( resp .
a general base point on @xmath441 ) .
[ limits of @xmath330-nodal curves ] [ p:1 - 4planes ] the regular components of the limit severi variety @xmath908 are the following ( they all appear with multiplicity @xmath330 ) : + [ 4planes - sing1 ] the proper transforms of the 24 planes @xmath909 , where @xmath112 is any one of the @xmath436curves @xmath670 , for @xmath910 and @xmath911 .
the @xmath912degree is 1 if @xmath913 ; when @xmath914 , it is @xmath330 if @xmath915 , and @xmath3 otherwise ; + [ 4planes - triple1 ] the proper transforms of the four degree @xmath16 surfaces @xmath916 , @xmath689 .
the @xmath917degree is @xmath16 if @xmath906 ; when @xmath907 , it is @xmath16 if @xmath918 , and @xmath3 otherwise ; + [ 4planes - planes ] the proper transforms of the four degree @xmath46 surfaces @xmath919 , @xmath920 .
the @xmath912degree is @xmath46 if @xmath921 , and 0 otherwise ; + [ 4planesij ] the proper transforms of the six surfaces in @xmath922 , @xmath923 .
they have @xmath912degree @xmath3 for every @xmath924 .
this follows from , and from propositions [ p : cubics - gauss ] and [ p:1-triangle ] .
proposition [ p:1-triangle ] tells us that @xmath925 has degree at least @xmath46 in @xmath926 for @xmath920 ; the computations in remark [ rem : dual ] below yield that it can not be strictly larger than @xmath46 ( see also the proof of corollary [ cor : thmdi ] ) , which proves theorem [ t : triangle ] for @xmath47 .
the @xmath912degree computation is straightforward .
[ rem : dual ] [ tetra1-count ] the degree of the dual of a smooth surface of degree 4 in @xmath21 is 36 . it is instructive to identify , in the above setting , the @xmath392 limiting curves passing through two general points on the proper transform of @xmath14 in @xmath712 .
this requires the @xmath912degree information in proposition [ p:1 - 4planes ] .
if we choose the two points on different planes , 24 of the 36 limiting curves through them come from , and 4 more , each with multiplicity 3 , come from .
if the two points are chosen in the same plane , then we have 12 contributions from , only one contribution , with multiplicity 3 , from , and 21 more contributions form .
no contribution ever comes from if we choose points on the faces of the tetrahedron .
[ r:2-tetra ] we have here an illustration of remark [ r : other - lim ] : the components @xmath925 are mapped to points in @xmath927 , hence they do not appear in the crude limit @xmath928 ( see corollary [ cor : thmbi ] below ) ; they are however visible in the crude limit severi variety of the degeneration to the quartic monoid corresponding to the face @xmath441 . in a similar fashion , to see the component @xmath929 one should consider the flat limit of the @xmath6 , @xmath2 , given by the surface @xmath182 described in proposition [ p : w - quartic ] . [ theorem [ t : tetrahedron ] for @xmath47 ] [ cor : thmbi ] consider a family @xmath7 of general quartic surfaces in @xmath21 degenerating to a tetrahedron @xmath14 .
the singularities of the total space @xmath10 consist in 24 ordinary double points , four on each edge of @xmath14 ( see [ s : setting ] ) .
it is @xmath330well behaved , with good model @xmath931 .
the limit in @xmath22 of the dual surfaces @xmath932 , @xmath2 ( which is the crude limit severi variety @xmath928 ) , consists in the union of the @xmath23 webs of planes passing through a singular point of @xmath10 , and of the @xmath24 webs of planes passing through a vertex of @xmath14 , each counted with multiplicity @xmath16 .
the only components of @xmath933 which are not contracted to lower dimensional varieties by the morphism @xmath821 are the ones in and in of proposition [ p:1 - 4planes ] .
their push forward in @xmath934 has total degree @xmath392 .
the assertion follows .
[ cor : thmdi ] consider a family @xmath714 of general quartic surfaces in @xmath21 , degenerating to a monoid quartic surface @xmath121 with tangent cone at its triple point @xmath72 consisting of a triple of independent planes ( see remark [ rem : monoid ] ) .
this family is @xmath330well behaved , with good model @xmath935 .
the crude limit severi variety @xmath936 consists in the surface @xmath937 ( which has degree @xmath46 ) , plus the plane @xmath938 counted with multiplicity @xmath939 .
we have a morphism @xmath940 . the push
forward by this map of the regular components of @xmath908 are @xmath937 for @xmath925 , @xmath941 for @xmath942 , @xmath938 for each of the twelve @xmath943 corresponding to a @xmath436curve @xmath944 with @xmath945 , and @xmath3 otherwise .
the degree of @xmath925 in @xmath635 is at least @xmath46 by proposition [ p:1-triangle ] , so the total degree of the push forward in @xmath946 of the regular components of @xmath908 is at least @xmath392 .
the assertion follows .
[ limits of @xmath33-nodal curves ] [ p:2 - 4planes ] the regular components of the limit severi variety @xmath947 are the following ( they all appear with multiplicity @xmath330 ) : + [ 4planes - sing2 ] @xmath948 for each set of two curves @xmath949 that do not meet the same edge of the tetrahedron @xmath14 .
the @xmath950degree is @xmath330 if @xmath951 does not contain the two edges met by @xmath952 , and @xmath3 otherwise ; + [ 4planes - vertex2 ] @xmath953 for @xmath954 and @xmath955 , which is a degree @xmath16 curve in @xmath701 .
the @xmath950degree is @xmath16 if @xmath490 does not contain both the edge met by @xmath112 and the vertex corresponding to @xmath575 , it is @xmath3 otherwise ; + [ 4planes - edge2 ] @xmath956 for @xmath923 , which has @xmath957-degree @xmath28 for @xmath915 , and @xmath3 otherwise ; + [ 4planes-0 ] @xmath958 for @xmath959 , which which has @xmath957-degree @xmath48 for @xmath960 , and @xmath3 otherwise ; + [ 4planes-1 ] @xmath961 for @xmath923 , and @xmath962 , which is a curve of @xmath957-degree @xmath3 for @xmath963 .
it goes as the proof of proposition [ p:1 - 4planes ] .
again , proposition [ p:2-triangle ] asserts that @xmath958 has degree at least @xmath48 in @xmath635 , but it follows from the computations in remark [ rem:480 ] below that it is exactly @xmath48 , which proves theorem [ t : triangle ] for @xmath49 . [ rem:480 ] [ tetra2-count ] the degree of the severi variety @xmath964 for a general quartic surface @xmath965 is @xmath393 ( see proposition [ p : deg - dual ] ) .
hence if we fix a general point @xmath428 on one of the components @xmath490 of @xmath14 we should be able to see the @xmath393 points of the limit severi variety @xmath966 through @xmath428 .
the @xmath950degree information in proposition [ p:2 - 4planes ] tells us this . for each choice of two distinct edges of @xmath14 spanning a plane distinct from @xmath490 , and of two @xmath436-curves @xmath112 and @xmath967 meeting these edges
, we have a curve containing @xmath428 in each of the items of type .
this amounts to a total of @xmath30 such curves . for each choice of a vertex and an edge of @xmath14 , such that they span a plane distinct from @xmath968 , there are @xmath16 curves containing @xmath428 in each of the four corresponding items
this amounts to a total of @xmath969 such curves . for each choice of an edge of @xmath14
not contained in @xmath968 , there are @xmath28 curves containing @xmath428 in the corresponding item .
this gives a contribution of @xmath26 curves .
finally , there are @xmath48 plane quartics containing @xmath428 in the item for @xmath970 . adding up , one finds the right number @xmath393 .
considerations similar to the ones in remark [ rem : dual ] could be made here , but we do not dwell on this .
[ theorem[t : tetrahedron ] for @xmath49 ] [ cor : thmbii ] same setting as in corollary [ cor : thmbi ] .
the family @xmath972 is @xmath33well behaved , with good model @xmath973 .
the crude limit severi variety @xmath974 consists of the image in @xmath975 of : + the @xmath25 components in of proposition [ p:2 - 4planes ] , which map to as many lines in @xmath975 ; + the @xmath26 components in of proposition [ p:2 - 4planes ] , each mapping @xmath976 to as many lines in @xmath975 ; + the @xmath27 components in of proposition [ p:2 - 4planes ] , respectively mapping @xmath977 to the dual lines of the edges of @xmath14 .
the components in question are the only ones not contracted to points by the morphism @xmath978 , and their push forward sum up to a degree @xmath393 curve .
[ cor : thmdii ] same setting as in corollary [ cor : thmdi ] ; the family @xmath979 is @xmath33well behaved , with good model @xmath935 .
the crude limit severi variety @xmath980 consists of the ordinary double curve of the surface @xmath937 , which has degree @xmath48 , plus a sum ( with multiplicities ) of lines contained in the dual plane @xmath938 of the vertex of @xmath121 .
it is similar to that of corollary [ cor : thmdi ] .
the lines of @xmath980 contained in @xmath938 are the push
forward by @xmath981 of the regular components of @xmath947 listed in remark [ rem:480 ] , with the exception of @xmath958 .
they sum up ( with their respective multiplicities ) to a degree @xmath982 curve , while @xmath958 has degree at least @xmath48 in @xmath635 by proposition [ p:2-triangle ] .
[ limits of @xmath16-nodal curves ] [ p:3 - 4planes ] the family @xmath931 is absolutely @xmath16good , and the limit severi variety @xmath983 is reduced , consisting of : + the @xmath29 points @xmath984 , for @xmath985 such that the span of the three corresponding double points of @xmath10 is not contained in a face of @xmath14 ; + the @xmath30 schemes @xmath986 , for @xmath689 and @xmath987 , such that the two double points of @xmath10 corresponding to @xmath112 and @xmath967 and the vertex with index @xmath146 span a plane which is not a face of @xmath14 .
they each consist of @xmath16 points ; + the @xmath23 schemes @xmath988 , for @xmath989 , and @xmath990 , such that the double point of @xmath10 corresponding to @xmath112 does not lie on the edge @xmath991 of @xmath14 , and that these two together do not span a face of @xmath14 .
they each consist of @xmath28 points ; + [ 3nodal - plane4ics ] the @xmath24 schemes @xmath992 , each consisting of @xmath31 points .
the list in the statement enumerates all regular components of the limit severi variety @xmath983 with their degrees ( as before , corollary [ coro : d ] only gives @xmath31 as a lower bound for the degree of ) .
they therefore add up to a total of at least @xmath394 points , which implies , by proposition [ p:3 - 4planes ] , that @xmath983 has no component besides the regular ones , and that those in have degree exactly @xmath31 .
reducedness then follows from remark [ rem : d ] , . in conclusion , all the above degenerations of quartic surfaces constructed from @xmath993 with a twist of @xmath178 are @xmath16well behaved , with @xmath893 as a good model . in particular : [ theorem [ t : tetrahedron ] for @xmath50 ] [ cor : thmbiii ] same setting as in corollary [ cor : thmbi ] . the limits in @xmath22 of @xmath16-tangent planes to @xmath6 , for @xmath138 , consist of : + the @xmath29 planes ( each with multiplicity 1 ) containing three double points of @xmath10 but no edge of @xmath14 ; + the @xmath30 planes ( each with multiplicity @xmath16 ) containing a vertex of @xmath14 and two double points of @xmath10 , but no edge of @xmath14 ; + the @xmath23 planes ( each with multiplicity @xmath28 ) containing an edge of @xmath14 and a double point of @xmath10 on the opposite edge ; + the @xmath24 faces of @xmath14 ( each with multiplicity @xmath31 ) .
the degeneration of a general quartic we considered in [ s:4planes ] is , in a sense , one of the most intricate .
there are _ milder _ ones , e.g. to : + [ cp]a general union of a cubic and a plane ; + [ qq ] a general union of two quadrics ( this is an incarnation of a well known degeneration of @xmath1 surfaces described in @xcite ) .
+ though we encourage the reader to study in detail the instructive cases of degenerations and , we will not dwell on this here , and only make the following observation about degeneration .
let @xmath140 be such a degeneration , with central fibre @xmath995 , where @xmath996 are two general quadrics meeting along a smooth quartic elliptic curve @xmath18 .
then the limit linear system of @xmath52 as @xmath2 tends to @xmath3 is just @xmath22 , so that @xmath7 endowed with @xmath234 is absolutely good . on the other hand ,
there are also degenerations to special singular irreducible surfaces , as the one we will consider in [ s : kummer - descr ] below . in the subsequent sub section , we will consider for further purposes another degeneration , the central fibre of which is still a ( smooth ) @xmath1 surface .
let @xmath997 be a smooth quadric and let @xmath407 be a general curve of type @xmath998 on @xmath999 .
we consider the double cover @xmath1000 branched along @xmath407 .
this is a @xmath1 surface and there is a smooth family @xmath140 with general fibre a general quartic surface and central fibre @xmath14 .
the pull back to @xmath14 of plane sections of @xmath999 which are bitangent to @xmath407 fill up a component @xmath1001 of multiplicity 1 of the crude limit severi variety @xmath1002 .
note that @xmath1002 naturally sits in @xmath1003 in this case , hence one can unambiguously talk about its degree .
although it makes sense to conjecture that @xmath1001 is irreducible , we will only prove the following weaker statement : [ prop : irrcomp ] the curve @xmath1001 contains an irreducible component of degree at least 36 .
we point out the following immediate consequence , which will be needed in [ ssec : irred ] below : [ cor : irrcomp ] if @xmath89 is a general quartic surface in @xmath21 , then the severi variety @xmath1004 ( which naturally sits in @xmath1005 ) has an irreducible component of degree at least 36 . to prove proposition [ prop : irrcomp ] we make a further degeneration to the case in which @xmath407 splits as @xmath1006 , where @xmath118 is a general curve of type @xmath1007 on @xmath999 , and @xmath94 is a general curve of type @xmath1008 , i.e. a general plane section of @xmath1009 .
then the limit of @xmath1001 contains the curve @xmath1010 in @xmath1011 parametrizing those planes in @xmath9 tangent to both @xmath94 and @xmath118 ( i.e. , @xmath1012 is the intersection curve of the dual surfaces @xmath1013 and @xmath1014 ) .
note that @xmath1013 is the quadric cone circumscribed to the quadric @xmath1015 and with vertex the point @xmath1016 orthogonal to the plane @xmath224 cutting out @xmath94 on @xmath999 , while @xmath1014 is a surface scroll , the degree of which is @xmath1017 by proposition [ p : dj ] , hence @xmath1018 . to prove proposition [ prop : irrcomp ]
, it suffices to prove that : [ lem : irrcomp ] the curve @xmath1019 is irreducible . to show this , we need a preliminary information .
let us consider the irreducible , locally closed subvariety @xmath1020 of dimension @xmath1017 , consisting of all curves @xmath1006 , where @xmath118 is a smooth , irreducible curve of type @xmath1007 , and @xmath94 is a plane section of @xmath999 which is not tangent to @xmath118 .
consider @xmath1021 the zariski closure of the set of all pairs @xmath1022 such that the plane @xmath1023 is tangent to both @xmath118 and @xmath94 , i.e. @xmath1024 .
we have the projections @xmath1025 and @xmath1026 .
the curve @xmath1019 is a general fibre of @xmath1027 .
[ lem : irrcomp2 ] the variety @xmath1028 contains a unique irreducible component @xmath1029 of dimension 19 which dominates @xmath1011 via the map @xmath1030 .
let @xmath1023 be a general plane of @xmath21 .
consider the conic @xmath1031 , and fix distinct points @xmath1032 on @xmath254 . there is a plane @xmath224 tangent to @xmath254 at @xmath1033 , and a cubic surface @xmath110 passing through @xmath1034 and tangent to @xmath254 at @xmath1035 ; moreover @xmath224 and @xmath110 can be chosen general enough for @xmath1036 to belong to @xmath1037 , where @xmath1038 and @xmath1039 .
then @xmath1040 , which proves that @xmath1030 is dominant .
let @xmath1041 be the fibre of @xmath1030 over @xmath1023 .
the above argument shows that there is a dominant map @xmath1042 whose general fibre is an open subset of @xmath1043 : precisely , if @xmath1044 is a general point , the @xmath111 is the linear system of plane sections of @xmath999 tangent to @xmath254 at @xmath1033 , and the @xmath1045 is the linear subsystem of @xmath1046 consisting of curves passing through @xmath1034 and tangent to @xmath254 at @xmath1035 . the existence and unicity of @xmath1029 follow .
now we consider the commutative diagram @xmath1047^\nu \ar[dr ] \ar[d]_{p ' } & { \mathcal j } \ar[d]^{p_1 } \\ \mathcal u ' \ar[r]_{f } & { \mathcal}u } \ ] ] where @xmath398 is the normalization of @xmath1029 , and @xmath1048 is the stein factorization of @xmath1049 .
the morphism @xmath1050 is finite of degree @xmath559 , equal to the number of irreducible components of the general fibre of @xmath1027 , which is @xmath1019 .
the irreduciblity of @xmath1029 implies that the monodromy group of @xmath1050 acts transitively on the set of components of @xmath1019 .
we need to prove that @xmath1051 . to do this , fix a general @xmath1052 , and consider the curve @xmath1053 , with @xmath94 general , which consists of @xmath559 components .
we can move @xmath94 to be a section of @xmath999 by a general tangent plane @xmath39 .
then the quadric cone @xmath1013 degenerates to the tangent plane @xmath1054 to @xmath1015 at @xmath1055 , counted with multiplicity @xmath33 .
we claim that , for @xmath1056 general , the intersection of @xmath1054 with @xmath1014 is irreducible .
indeed , since @xmath1014 is a scroll , a plane section of @xmath1014 is reducible if and only if it contains a ruling , i.e. if and only if it is a tangent plane section of @xmath1014 .
since @xmath1057 , the biduality theorem implies the claim .
the above assertion implies @xmath1058 .
if equality holds , the general curve @xmath1019 consists of two curves which , by transitivity of the monodromy action of @xmath135 , are both unisecant to the lines of the ruling of @xmath1014 . to see that this is impossible , let us degenerate @xmath118 as @xmath1059 , where @xmath1060 is a general curve of type @xmath1061 and @xmath1062 is a general curve of type @xmath1063 on @xmath999 .
then @xmath1014 accordingly degenerates and its limit contains as irreducible components @xmath1064 and @xmath1065 , which are both scrolls of degree @xmath24 ( though we will not use it , we note that @xmath1066 and the ( crude ) limit of @xmath1014 in the above degeneration consists of the union of @xmath1064 , @xmath1065 , and of the five planes dual to the points of @xmath1067 , each of the latter counted with multiplicity @xmath33 ) .
we denote by @xmath1068 either one of the curves @xmath1069 .
let again @xmath94 be a general plane section of @xmath999 .
we claim that the intersection of @xmath1070 with @xmath1013 does not contain any unisecant curve to the lines of the ruling of @xmath1070 .
this clearly implies that the general curve @xmath1019 can not split into two unisecant curves to the lines of the ruling of @xmath1014 , thus proving that @xmath1051 . to prove the claim
, it suffices to do it for specific @xmath1071 , @xmath999 and @xmath94 . for @xmath1072
we take the rational normal cubic with affine parametric equations @xmath1073 , with @xmath1074 . for @xmath999
we take the quadric with affine equation @xmath1075 , and for @xmath94 the intersection of @xmath999 with the plane @xmath483 .
let @xmath1076 be affine coordinates in the dual space , so that @xmath1076 corresponds to the plane with equation @xmath1077 ( i.e. , we take as plane at infinity in the dual space the orthogonal to the origin ) .
then the affine equation of @xmath1070 is gotten by eliminating @xmath15 in the system @xmath1078 which defines the ruling @xmath1079 of @xmath1080 orthogonal to the tangent line to @xmath1072 at the point with coordinates @xmath1081 , @xmath1074 .
the affine equation of @xmath1013 is gotten by imposing that the system @xmath1082 has one solution with multiplicity @xmath33 ; the resulting equation is @xmath1083 . adding this to means intersecting @xmath1013 with @xmath1079 ;
for @xmath13 , the resulting system can be written as @xmath1084 for a general @xmath1074 , the first equation gives two values of @xmath72 and the remaining equations the corresponding values of @xmath168 and @xmath1085 , i.e. , we get the coordinates @xmath1076 of the two intersection points of @xmath1013 and @xmath1079 .
now we note that the discriminant of @xmath1086 as a polynomial in @xmath72 is @xmath1087 , which has the two simple solutions @xmath1088 .
this implies that the projection on @xmath1089 of the curve cut out by @xmath1013 on @xmath1080 has two simple ramification points .
in particular @xmath1090 is locally irreducible at these points , and it can not split as two unisecant curves to the lines of the ruling .
this proves the claim and ends the proof of the lemma .
this section is devoted to the description of some properties of quartic _
kummer surfaces _ in @xmath9 .
they are quartic surfaces with 16 ordinary double points @xmath1091 as their only singularities .
alternatively a kummer surface @xmath89 is the image of the jacobian @xmath1092 of a smooth genus @xmath33 curve @xmath117 , via the degree @xmath33 morphism @xmath1093 determined by the complete linear system @xmath1094 , where @xmath1095 is the abel
jacobi embedding , so that @xmath1096 is a principally polarised abelian surface ( see , e.g. , ( * ? ? ?
* chap.10 ) ) . since @xmath1097 is composed with the @xmath1098 involution on @xmath1092 , the 16 nodes of @xmath89 are the images of the 16 points of order @xmath33 of @xmath1092 . by projecting from a node
, kummer surfaces can be realised as double covers of the plane , branched along the union of six distinct lines tangent to one single conic ( see , e.g. , ( * ? ? ?
* chap.viii , exercises ) ) .
we refer to the classical book @xcite for a thorough description of these surfaces ( see also ( * ? ? ?
* chap.10 ) ) . an important feature of kummer surfaces is that they carry a so - called _ @xmath1099configuration _ ( see @xcite , as well as the above listed references ) .
let @xmath89 be such a surface .
there are exactly @xmath28 distinct planes @xmath1100 tangent to @xmath89 along a _ contact conic _
@xmath1101 , for @xmath1102 .
the contact conics are the images of the 16 symmetric theta divisors @xmath1103 on @xmath1092 .
each of them contains exactly @xmath27 nodes of @xmath89 , coinciding with the branch points of the map @xmath1104 determined by the canonical @xmath1105 on @xmath117 .
two conics @xmath1106 , @xmath906 , intersect at exactly two points , which are double points of @xmath89 : they are the nodes corresponding to the two order @xmath33 points of @xmath1092 where @xmath1107 and @xmath1108 meet .
since the restriction map @xmath1109 is an isomorphism , there is no pair of points of @xmath1092 contained in three different theta divisors .
this implies that , given a pair of nodes of @xmath89 , there are exactly two contact conics containing both of them . in other words , if we fix an @xmath1110 , the map from @xmath1111 to the set of pairs of distinct nodes of @xmath89 on @xmath1101 , which maps @xmath1112 to @xmath1113 , is bijective .
this yields that each node of @xmath89 is contained in exactly @xmath27 conics @xmath1101 .
the configuration of 16 nodes and 16 conics with the above described incidence property is called a @xmath1099configuration .
let @xmath1114 be the minimal smooth model of @xmath89 , @xmath1115 the @xmath773-curves over the nodes @xmath1116 of @xmath89 respectively , and @xmath1117 the proper transform of the conic @xmath1101 , for @xmath1102 .
since @xmath1118 is a @xmath1 surface and the @xmath1117 s are rational curves , the latter are @xmath773-curves . the @xmath1099configuration can be described in terms of the existence of the two sets @xmath1119 of @xmath28 pairwise disjoint @xmath773curves , enjoying the further property that each curve of a given set meets exactly six curves of the other set , transversely at a single point .
[ p : dual - kummer ] let @xmath89 be a kummer surface . then its dual @xmath1120 is also a kummer surface . by proposition [ p : deg - dual - sing ] , we have @xmath1121 . because of the singularities on @xmath89 , the gauss map @xmath1122 is not a morphism .
however we get an elimination of indeterminacies @xmath1123_f \ar[dr]^g & \\ x \ar@{-->}[rr]_{\gamma_x } & & \check{x } } \ ] ] by considering the minimal smooth model @xmath1114 of @xmath89 .
the morphism @xmath135 is the contraction of the sixteen curves in @xmath1124 , and @xmath64 maps each @xmath230 to a conic which is the dual of the tangent cone to @xmath89 at the node corresponding to @xmath230 . on the other hand , since @xmath1125 contracts each of the curves @xmath1126 to a point , then @xmath64 contracts the curves in @xmath1127 to as many ordinary double points of @xmath1128 .
the assertion follows .
let @xmath1129 be the locally closed subset of @xmath1130 whose points correspond to kummer surfaces and let @xmath1131 be the universal family : over @xmath1132 , we have the corresponding kummer surface @xmath1133 .
we have a subscheme @xmath1134 such that @xmath1135 is a finite morphism of degree 16 : the fibre @xmath1136 over @xmath1132 consists of the nodes of @xmath89 .
we denote by @xmath1137 the monodromy group of @xmath1138 there is in addition another degree 16 finite covering @xmath1139 : for @xmath1132 ,
the fibre @xmath1140 consists of the set of the contact conics on @xmath89 .
proposition [ p : dual - kummer ] implies that the monodromy group of this covering is isomorphic to @xmath1141 .
then we can consider the commutative square @xmath1142_{p ' } \ar[rr]^{q ' } & & \mathcal n \ar[d]^{p } \\
\mathcal g \ar[rr]_{q } & & { { \mathcal}k^\circ } } \ ] ] where @xmath1029 is the incidence correspondence between nodes and conics .
note that @xmath1143 are both finite of degree 6 , with isomorphic monodromy groups ( see again proposition [ p : dual - kummer ] ) . here
, we collect some results on the monodromy groups of the coverings appearing in .
they are probably well known to the experts , but we could not find any reference for them .
[ l : s_6 ] the monodromy group of @xmath1144 and of @xmath1145 is the full symmetric group @xmath1146 .
it suffices to prove only one of the two assertions , e.g. the one about @xmath1147 .
let @xmath89 be a general kummer surface and let @xmath1148 be a node of @xmath89 . as we noticed , by projecting from @xmath1148 , we realise @xmath89 as a double cover of @xmath562 branched along @xmath27 lines tangent to a conic @xmath112 , which is the image of the @xmath773curve over @xmath1148 .
these 6 lines are the images of the six contact conics through @xmath1148 , i.e. the fibre over @xmath1149 .
since @xmath89 is general , these 6 tangent lines are general .
the assertion follows .
[ cor : geer ] the group @xmath1141 acts transitively , so @xmath1150 and @xmath1151 are irreducible . it suffices to prove that the monodromy of @xmath1138 is transitive .
this follows from lemma [ l : s_6 ] and from the fact that any two nodes of a kummer surface lie on some contact conic .
it is also possible to deduce the transitivity of the monodromy action of @xmath72 and @xmath168 from the irreducibility of the igusa quartic solid , which parametrizes quartic kummer surfaces with one marked node ( see , e.g. , ( * ? ? ? * chap.10 ) ) .
the following is stronger : [ p:2-transitivity ] the group @xmath1141 acts @xmath33transitively .
again , it suffices to prove the assertion for @xmath1138 . by corollary [ cor : geer ] , proving that the monodromy is 2transitive is equivalent to showing that the stabilizer of a point in the general fibre of @xmath72 acts transitively on the remaining points of the fibre .
let @xmath89 be a general kummer surface and @xmath1152 a node .
consider the projection from @xmath1148 , which realizes @xmath89 as a double cover of @xmath562 branched along 6 lines tangent to a conic @xmath112 .
the 15 nodes on @xmath89 different from @xmath1148 correspond to the pairwise intersections of the 6 lines . moving the tangent lines
to @xmath112 one leaves the node @xmath1148 fixed , while acting transitively on the others .
look now at the pull back @xmath1153 .
of course @xmath1029 is a component of @xmath1153 .
we set @xmath1154 , and the morphism @xmath1155 which is finite of degree @xmath781 .
we let @xmath1156 be the monodromy of this covering .
[ l : preparation ] the group @xmath1157 acts transitively , i.e. @xmath678 is irreducible . let @xmath1158 be two nodes not lying on the contact conic @xmath254 .
there is a contact conic @xmath1159 that contains both @xmath1160 and @xmath1161 ; it meets @xmath254 transversely in two points , distinct from @xmath1160 and @xmath1161 , that we shall call @xmath1162 and @xmath8 . now a monodromy transformation that fixes @xmath1159 and fixes @xmath1162 and @xmath8 necessarily fixes @xmath254 .
it therefore suffices to find a monodromy transformation fixing @xmath1159 which fixes @xmath1162 and @xmath8 , and sends @xmath1160 to @xmath1161 .
such a transformation exists by lemma [ l : s_6 ] .
[ p:3-transitivity ] let @xmath89 be a general kummer surface . then : + [ it : alpha ] @xmath1141 acts transitively the set of unordered triples of distinct nodes belonging to a contact conic ; + [ it : beta ] the action of @xmath1141 on the set of unordered triples of distinct nodes not belonging to a contact conic has at most two orbits . to prove this , we need to consider degenerations of kummer surfaces when the principally polarised abelian surface @xmath1096 becomes non simple , e.g. , when @xmath117 degenerates to the union of two elliptic curves @xmath767 transversally meeting at a point . in this case
the linear system @xmath1163 on the abelian surface @xmath1164 , is still base point free , but it determines a degree 4 morphism @xmath1165 ( where @xmath1166 is a smooth quadric ) , factoring through the _ product kummer surface _
@xmath1167 , and a double cover @xmath1168 branched along a curve of bidegree @xmath998 which is a union of @xmath36 lines ; the lines in question are @xmath1169 ( resp .
@xmath1170 ) where @xmath1160 ( resp .
@xmath1161 ) ranges among the four branch points of the morphism @xmath1171 ( resp .
@xmath1172 ) .
we call the former _ horizontal lines _ , and the latter
_ vertical lines_. each of them has four marked points : on a line @xmath1173 ( resp .
@xmath1174 ) , these are the four points @xmath1175 where @xmath1161 ( resp .
@xmath1160 ) varies as above .
one thus gets 16 points , which are the limits on @xmath1176 of the 16 nodes of a general kummer surface @xmath89 .
the limits on @xmath1176 of the sixteen contact conics on a general kummer surface @xmath89 are the sixteen curves @xmath1177 .
on such a curve , the limits of the six double points on a contact conic on a general kummer surface are the six marked points on @xmath1173 and @xmath1174 that are distinct from @xmath1178 .
part ( i ) follows from lemma [ l : s_6 ] . as for part ( ii ) , consider three distinct nodes @xmath1160 , @xmath1179 and @xmath1180 ( resp .
@xmath1161 , @xmath1181 and @xmath1182 ) of @xmath89 that do not lie on a common conic of the @xmath1099 configuration on @xmath89 .
we look at their limits @xmath1183 , @xmath1184 and @xmath1185 ( resp .
@xmath1186 , @xmath1187 and @xmath1188 ) on the product kummer surface @xmath1176 ; they are in one of the two configurations ( a ) and ( b ) described in figure [ f : triples ] .
the result follows from the fact that the monodromy of the family of product kummer surfaces acts as the full symmetric group @xmath1189 on the two sets of vertical and horizontal lines respectively .
hence the triples in configuration ( a ) [ resp . in configuration ( b ) ] are certainly in one and the same orbit .
we consider a family @xmath358 of surfaces in @xmath9 induced ( as explained in [ s : setting ] ) by a pencil generated by a general quartic surface @xmath147 and a general kummer surface @xmath14 .
we will describe a related @xmath5-good model @xmath597 for @xmath1190 .
our construction is as follows : + _ _ [ kum : bs - change ] we first perform a degree 2 base change on @xmath358 ; + [ kum : resolution ] then we resolve the singularities of the new family ; + [ kum : blow - up ] we blow up the proper transforms of the sixteen contact conics on @xmath14 .
+ the base change is useful to analyze the contribution of curves passing through a node of @xmath14 .
the total space @xmath10 is smooth , analytically locally given by the equation @xmath1191 around each of the double points of @xmath14 .
we perform a degree 2 base change on @xmath135 , and call @xmath1192 the resulting family .
the total space @xmath454 has @xmath28 ordinary double points at the preimages of the nodes of @xmath14 .
we let @xmath1193 be the resolution of these @xmath28 points , gotten by a simple blow up at each point .
we have the new family @xmath514 , with @xmath1194 .
the new central fibre @xmath431 consists of the minimal smooth model @xmath439 of @xmath14 , plus the exceptional divisors @xmath1195 .
these are all isomorphic to a smooth quadric @xmath1196 .
we let @xmath1115 be the exceptional divisors of @xmath1197 .
each @xmath1198 meets @xmath439 transversely along the curve @xmath230 , and two distinct @xmath1199 do not meet . as in
[ s:16_6 ] , we let @xmath1200 be the proper transforms of the @xmath28 contact conics @xmath1126 on @xmath14 : they are pairwise disjoint @xmath773-curves in @xmath431 .
we consider the blow - up @xmath1201 of @xmath89 along them .
the surface @xmath439 is isomorphic to its strict transform on @xmath613 .
let @xmath1202 be the exceptional divisors of @xmath1203 .
each @xmath1204 meets @xmath439 transversely along the ( strict transform of the ) curve @xmath1117 .
note that , by the triple point formula [ l : triple - point ] , one has @xmath1205 , so that @xmath1204 is an @xmath1206hirzebruch surface , and @xmath1117 is the negative section on it .
we call @xmath1207 the strict transforms of @xmath1195 respectively .
they respectively meet @xmath439 transversely along ( the strict transforms of ) @xmath1115 . for @xmath1102 , there are exactly six curves among the @xmath1208 s that meet @xmath230 : we call them @xmath1209 . the surface
@xmath1210 is the blow - up of @xmath1198 at the six intersection points of @xmath230 with @xmath1209 : we call @xmath1211 respectively the six corresponding @xmath436-curves on @xmath1210 .
accordingly , @xmath1210 meets transversely six @xmath1212 s , that we denote by @xmath1213 , along @xmath1211 respectively .
the surface @xmath1210 is disjoint from the remaining @xmath1212 s . for @xmath1214 ,
we denote by @xmath1215 the six @xmath230 s that meet @xmath1208 .
there are correspondingly six @xmath1210 s that meet @xmath1212 : we denote them by @xmath1216 , and let @xmath1217 be their respective intersection curves with @xmath1212 .
note the equality of sets @xmath1218 we shall furthermore use the following notation ( cf .
[ conv ] ) . for @xmath1214
, we let @xmath1219 ( or simply @xmath110 ) be the divisor class of the ruling on @xmath1212 , and @xmath1220 ( or simply @xmath94 ) the divisor class @xmath1221 .
note that @xmath1222 and @xmath1223 , for @xmath1224 and @xmath1225 .
we write @xmath1226 for the pull - back to @xmath439 of the plane section class of @xmath1227 . for @xmath1228
, we let @xmath1229 and @xmath1230 be the two rulings of @xmath1198 , and @xmath1231 ( or simply @xmath94 ) be the divisor class @xmath1232 ; we use the same symbols for their respective pull - backs in @xmath1210 .
when designing one of these surfaces by @xmath1233 , we use the obvious notation @xmath1234 and @xmath1235 . we shall now describe the limit linear system of @xmath596 as @xmath2 tends to @xmath3 , and from this we will see that @xmath893 is a good model of @xmath10 over @xmath11 .
we start with @xmath1236 , which is a @xmath21bundle over @xmath11 , whose fibre at @xmath136 is @xmath599 ; we set @xmath600 , and @xmath601 for @xmath1237 .
note that @xmath862 .
we will proceed as follows : + the description of these steps parallels the one in
[ s:4planes - limlin ] , so we will be sketchy here .
the @xmath1238 planes in @xmath9 containing exactly three distinct nodes of @xmath14 correspond to the @xmath3-dimensional subsystems @xmath1239 of @xmath1240 , where @xmath1241 ranges through all subsets of cardinality 3 of @xmath1242 such that the nodes @xmath1243 corresponding to the @xmath773curves @xmath1244 do not lie in a plane of the @xmath1099 configuration of @xmath14 .
we denote by @xmath1245 the unique curve in the system and we set @xmath1246 , which lies in @xmath1247 . the exceptional component @xmath1248 of the blow - up of @xmath1249 at the point corresponding to @xmath1250 can be identified with the 3dimensional complete linear system @xmath1251 which is isomorphic to the projectivization of the kernel of the surjective map @xmath1252 the typical element of @xmath1253 consists of + the curve @xmath1245 on @xmath166 , plus + one curve in @xmath1254 for each @xmath1255 , matching @xmath1245 along @xmath1256 , plus + two rulings in each @xmath1212 ( i.e. a member of @xmath1257 ) , @xmath1214 , matching along the divisor @xmath1208 , while + the restriction to @xmath1258 is trivial for every @xmath1259 .
the strict transform of @xmath612 is isomorphic to the blow up of @xmath1260 at the point corresponding to @xmath1250 . by lemma [ l : tg - id ]
, the exceptional divisor @xmath1261 of this blow - up identifies with the pull
back linear series on @xmath1250 of the @xmath33-dimensional linear system of lines in the plane spanned by @xmath1243 ( note that in this linear series there are three linear subseries corresponding to sections vanishing on the curves @xmath1262 which are components of @xmath1263 ) .
the divisor @xmath1264 is cut out on the strict transform of @xmath1260 by @xmath1248 , along the plane @xmath1265 given by the equation @xmath1266 in the above notation .
the identification of @xmath1264 with @xmath1023 is not immediate .
it would become more apparent by blowing up the curves @xmath1245 in the central fibre ; we will not do this here , because we do not need it , and we leave it to the reader ( see step ( [ kum : coplanar]b ) for a similar argument ) .
however , we note that @xmath1267 coincides with the @xmath1268 spanned by three non zero sections @xmath1269 where @xmath1270 vanishes exactly on @xmath567 for each @xmath1271 .
these three sections correspond to three points @xmath1272 in @xmath1023 . in the identification of @xmath1023 with @xmath1273 the points
@xmath1274 are mapped to the respective pull
backs on @xmath1250 of the three lines @xmath1275 .
the @xmath28 planes of the @xmath1099 configuration correspond to the @xmath3-dimensional subsystems @xmath1276 consisting of the only curve @xmath1277 .
the blow - up of @xmath204 at these points introduces @xmath28 new components @xmath1278 , @xmath1214 , in the central fibre , respectively isomorphic to the linear systems @xmath1279 the corresponding line bundles restrict to the trivial bundle on all components of @xmath613 but @xmath1212 and @xmath1233 , for @xmath1225 , where the restriction is to @xmath1280 and to @xmath1281 , respectively . for each @xmath1282 ,
the complete linear system @xmath1283 is @xmath3-dimensional , its only divisor is the strict transform in @xmath1233 of the unique curve in @xmath1284 that is singular at the point @xmath1285 .
this is the union of the proper transforms of the two curves in @xmath1286 and @xmath1287 through @xmath1288 , and it cuts out a @xmath3-cycle @xmath1289 of degree 2 on @xmath1290 .
we conclude that @xmath1291 where @xmath1292 is the defining sheaf of ideals of the @xmath3-cycle @xmath1293 supported on the six fibres @xmath1217 of the ruling of @xmath1212 .
we shall later study the rational map determined by this linear system on @xmath1212 ( see proposition [ p:2:1 ] ) . for each @xmath1112 ,
the glueing of @xmath1294 with the strict transform of @xmath1260 is as follows : the exceptional plane @xmath1295 on the strict transform of @xmath1260 identifies with @xmath1296 by lemma [ l : tg - id ] , and the latter naturally identifies as the @xmath33-dimensional linear subsystem of @xmath1297 consisting of divisors of the form @xmath1298 let @xmath207 be the blow - up of @xmath1249 at the @xmath1299 distinct points described in the preceding step .
the next operation is the blow - up @xmath754 along the @xmath1300 pairwise disjoint respective strict transforms of the pencils @xmath1301 to describe the exceptional divisor @xmath1302 of @xmath1303 on the proper transform of , consider the 3dimensional linear system @xmath1304 , isomorphic to the projectivization of the kernel of the surjective map @xmath1305 then @xmath1302 identifies as the blow up of @xmath1306 along the line defined by @xmath1266 in the above notation ; in particular it is isomorphic to @xmath1307 , with @xmath108bundle structure @xmath1308 induced by the projection of the left - hand side of on its last summand , as follows from lemma [
l : tg - id ] .
the typical element of @xmath1302 consists of + [ comp(i ) ] a member @xmath117 of @xmath1309 , plus + two curves in @xmath1310 and @xmath1310 respectively , matching @xmath117 along @xmath1311 and @xmath1312 , together with + rational tails on the @xmath1212 s ( two on those @xmath1212 meeting neither @xmath1313 nor @xmath1314 , one on those @xmath1212 meeting exactly one component among @xmath1313 and @xmath1314 , and none on the two @xmath1212 s meeting both @xmath1313 and @xmath1315 ) matching @xmath117 along @xmath1208 . + the image by @xmath1316 of such a curve is the point corresponding to its component .
[ rem : two - nodes ] the image of @xmath712 via the complete linear system @xmath1317 provides a model @xmath714 of the initial family @xmath7 , with central fibre the transverse union of two double planes @xmath1318 and @xmath1319 . for @xmath1320 ,
the plane @xmath1321 is the projection of @xmath1322 from the point @xmath1323 corresponding to the direction of the line @xmath1324 in @xmath1325 , where @xmath1326 ;
there is a _ marked conic _ on @xmath1327 , corresponding to the branch locus of this projection .
the restriction to @xmath567 of the morphism @xmath1328 is a degree @xmath33 covering @xmath1329 .
the two marked conics on @xmath1318 and @xmath1319 intersect at two points on the line @xmath1330 , which are the two branch points of both the double coverings @xmath1331 and @xmath1331 .
these points correspond to the two points cut out on @xmath1311 ( resp .
@xmath1312 ) by the two curves @xmath1208 that correspond to the two double conics of @xmath14 passing through @xmath1332 and @xmath1333 .
there are in addition six distinguished points on @xmath1330 , corresponding to the six pairs of points cut out on @xmath1311 ( resp .
@xmath1312 ) by the six curves @xmath1245 on @xmath166 .
the last operation is the blow - up @xmath1334 along the 16 disjoint surfaces that are the strict transforms of the @xmath33-dimensional linear systems @xmath1335 we want to understand the exceptional divisor @xmath1336 . consider the linear system @xmath1337 which identifies with the projectivization of the kernel of the surjective map @xmath1338 ( itself isomorphic to @xmath1339 , by the way ) .
up @xmath1340 at the point @xmath1341 corresponding to @xmath1266 ; one thus gets a @xmath149bundle over the plane @xmath1342 .
then @xmath1336 is obtained by further blowing up along the proper transforms of the lines joining @xmath1341 with the @xmath1343 = 51 $ ] points of @xmath1344 we blew up in step ( [ kum : coplanar ] ) .
the typical member of @xmath1345 consists of two members of @xmath1342 and @xmath1346 respectively , matching along @xmath567 , together with rational tails on the surfaces @xmath1212 .
[ rem : one - node ] the image of @xmath712 by the complete linear system @xmath1347 provides a model @xmath714 of the initial family @xmath7 , with central fibre the transverse union of a smooth quadric @xmath999 , and a double plane @xmath1023 branched along six lines tangent to the conic @xmath1031 ( i.e. the projection of @xmath14 from the node @xmath1348 ) .
there are fifteen _ marked points _ on @xmath1023 , namely the intersection points of the six branch lines of the double covering @xmath1349 .
we shall now describe the curves parametrized by the intersections of the various components of @xmath834 , thus proving : [ prop : limlin - kummer ] the central fibre @xmath834 is the limit linear system of @xmath1350 as @xmath2 tends to @xmath3 .
we analyze step by step the effect on the central fibre of the birational modifications operated on @xmath204 in the above construction , each time using lemma [ l : tg - id ] without further notification . at this step ,
recall ( cf . [ conv ] ) that @xmath1351 denotes the proper transform of @xmath1352 in the blow up @xmath1353 . for each @xmath1354 such that @xmath1355 is a plane that does not belong to the @xmath1099 configuration , the intersection
@xmath1356 is the exceptional @xmath108 of the blow up of @xmath862 at the point corresponding to @xmath1250 .
its points , but those lying on one of the three lines joining two points among @xmath1357 which also have been blown up ( the notation is that of step ( [ kum : coplanar]a ) ) , correspond to the trace of the pull back of @xmath22 on @xmath1358 . for each @xmath1359 , the intersection @xmath1360 is a plane , the points of which correspond to curves @xmath1361 of @xmath526 , latexmath:[$\phi
\in corresponding to the cases when @xmath852 contains one of the six curves @xmath1217 . let @xmath1363 .
the intersection @xmath1364 is a @xmath521 ; the first factor is isomorphic to the proper transform of the line @xmath1309 in @xmath612 , while the second is isomorphic to the line @xmath1365 in the notation of step above .
then the points in @xmath1366 correspond to curves @xmath1367 in @xmath526 , with @xmath1368 , exception made for the points with second coordinate @xmath1369 $ ] or @xmath1370 $ ] in @xmath1306 , where @xmath1371 vanishes on @xmath567 for each @xmath1372 .
let @xmath1373 be such that @xmath1355 is a plane outside the @xmath1099 configuration .
the intersection @xmath1374 is the @xmath108 preimage of the point corresponding to @xmath1245 in @xmath1309 via @xmath1316 , and parametrizes curves @xmath1375 , with @xmath1376 and @xmath1377 matching @xmath1245 along @xmath1311 and @xmath1312 respectively . on the other hand , for @xmath1373 such that @xmath1355 belongs to the @xmath1099 configuration ,
let @xmath1359 be such that @xmath1277 is cut out on @xmath14 by @xmath1355 , and set @xmath1378 and @xmath1379 ; then @xmath1380 is the preimage by @xmath1316 of the point corresponding to @xmath1208 in @xmath1309 , and parametrizes the curves @xmath1381 where @xmath1382 is the proper transform by @xmath1383 of a member of @xmath1384 tangent to @xmath1311 at @xmath1385 , and similarly for @xmath1386 .
let @xmath1282 .
the intersection @xmath1387 is isomorphic to the plane @xmath1342 blown up at the @xmath1388 points corresponding to the intersection of at least two lines among the fifteen @xmath1389 , @xmath1390 .
each point of the non exceptional locus of this surface corresponds to a curve @xmath1391 , with @xmath1392 .
let @xmath1393 .
the intersection @xmath1394 is an @xmath886 , isomorphic to the blow
up at @xmath1341 of the plane in @xmath1340 projectivization of the kernel of the restriction of @xmath1395 to @xmath1396 .
it has the structure of a @xmath149bundle over @xmath1397 , and its points correspond to curves @xmath1398 , with @xmath1399 matching with @xmath1400 .
let @xmath1401 , and assume the plane @xmath1355 is outside the @xmath1099 configuration .
then @xmath1402 is a @xmath521 , the two factors of which are respectively isomorphic to the projectivization of the kernel of the restriction of @xmath1395 to @xmath1403 , and to the line @xmath1404 in @xmath1405 ( with the notations of step ( [ kum : coplanar]b ) ) .
it therefore parametrizes the curves @xmath1406 where @xmath1407 matches @xmath1408 along @xmath567 .
let @xmath1359 be such that @xmath1212 intersects @xmath1258 , and set @xmath1409 .
then @xmath1410 is a @xmath868 , the two factors of which are respectively isomorphic to the pencil of pull backs to @xmath1258 of members of @xmath1411 tangent to @xmath567 at the point @xmath1412 , and to the subpencil @xmath1413 of @xmath1414 .
it parametrizes curves @xmath1415 where @xmath1416 is the proper transform of a curve on @xmath310 tangent to @xmath1256 at @xmath1417 .
it follows from the above analysis that the points of @xmath834 all correspond in a canonical way to curves on @xmath526 , which implies our assertion by lemma [ l : lim - lin ] . in this section ,
we study the rational map @xmath1419 ( or simply @xmath198 ) determined by the linear system @xmath1420 on @xmath1212 , for @xmath1421 .
let @xmath1422 be the blow
up at the twelve points in the support of @xmath1423 . for @xmath1225 ,
we denote by @xmath1424 the strict transform of the ruling @xmath1290 , and by @xmath1425 the two exceptional curves of @xmath1426 meeting @xmath1424 .
then the pull back via @xmath1426 induces a natural isomorphism @xmath1427 we denote by @xmath1428 the right hand side linear system .
[ p:2:1 ] the linear system @xmath1428 determines a @xmath1429 morphism @xmath1430 where @xmath965 is a quadric cone .
the divisor @xmath1431 is contracted by @xmath746 to the vertex of @xmath965 .
the branch curve @xmath407 of @xmath746 is irreducible , cut out on @xmath965 by a quartic surface ; it is rational , with an ordinary six fold point at the vertex of @xmath965 .
before the proof , let us point out the following corollary , which we will later need .
[ c:2:1 ] the severi variety of irreducible @xmath5nodal curves in @xmath1432 is isomorphic to the subvariety of @xmath1433 parametrizing @xmath5tangent planes to @xmath407 , for @xmath1434 .
they have degree @xmath1435 , @xmath1436 , and @xmath35 , respectively . for the proof of proposition [ p:2:1 ] we need two preliminary lemmas .
[ lem : dim-2sigma ] the linear system latexmath:[$|{\mathcal{o}}_{w_j}(2h)\otimes\mathcal{i}_{z_j}|\subset the 0-cycle @xmath1423 is cut out on @xmath1438 by a general curve in @xmath1439 .
let then @xmath1440 be a non sero section vanishing at @xmath1423 .
then @xmath1441 where @xmath1442 is the restriction map .
the assertion now follows from the restriction exact sequence , since @xmath1443 [ lem:2sigma - deg2 ] the rational map @xmath1419 has degree @xmath33 onto its image , and its restriction to any line of the ruling @xmath1444 but the six @xmath1445 , @xmath1225 , has degree @xmath33 as well . let @xmath1446 be a general point and let @xmath1447 be the line of the ruling containing @xmath428 .
one can find a divisor @xmath1448 containing @xmath428 but not containing @xmath1447 .
let @xmath1449 be the length two scheme cut out by @xmath118 on @xmath1447 . by an argument similar to the one in the proof of lemma [ lem : dim-2sigma ]
, one has @xmath1450 .
this shows that @xmath428 and @xmath1451 are mapped to the same point by @xmath198 .
then , considering the sublinear system @xmath1452 of @xmath1418 , with fixed divisor @xmath1453 , the assertion follows from the base point freeness of @xmath1454 .
first we prove that @xmath1414 has no fixed components , hence that the same holds for @xmath1455 .
suppose @xmath852 is such a fixed component .
by lemma [ lem:2sigma - deg2 ] , @xmath1456 , hence @xmath852 should consist of curves contained in rulings .
the argument of the proof of lemma [ lem : dim-2sigma ] shows that no such a curve may occur in @xmath852 , a contradiction .
let @xmath1457 be a general element . by lemmas [ lem : dim-2sigma ] and [ lem:2sigma - deg2 ]
, @xmath118 is irreducible and hyperelliptic , since @xmath1458 .
moreover @xmath1459 and @xmath1460 .
this implies that @xmath118 is smooth and that @xmath1455 is base point free .
moreover the image @xmath965 of @xmath198 has degree 2 . since
@xmath1461 and @xmath1462 , the connected divisor @xmath1463 is contracted to a double point @xmath1464 of @xmath965 , which is therefore a cone .
since @xmath118 is mapped @xmath1429 to a general plane section of @xmath965 , which is a conic , we see that @xmath1465 .
let @xmath1466 be general , and @xmath814 its image via @xmath198 , which is a ruling of @xmath965 .
the restriction @xmath1467 is a degree 2 morphism , which is ramified at the intersection point of @xmath852 with @xmath1208 .
this implies that @xmath814 meets @xmath407 at one single point off the vertex @xmath1464 of @xmath965 .
hence @xmath407 has a unique irreducible component @xmath760 which meets the general ruling @xmath814 in one point off @xmath1464 .
we claim that @xmath1468 . if not
, @xmath1469 consists of rulings @xmath1470 , corresponding to rulings @xmath1471 , clearly all different from the @xmath1472 , with @xmath1225 .
then the restrictions @xmath1473 would be isomorphisms , for @xmath1474 , which is clearly impossible .
hence @xmath407 is irreducible , rational , sits in @xmath1475 .
finally , taking a plane section of @xmath965 consisting of two general rulings , we see that it has only two intersection points with @xmath407 off @xmath1464 .
hence @xmath407 has a point of multipilicity 6 at @xmath1464 and the assertion follows .
[ rem : singular ] each of the curves @xmath1476 , for @xmath1477 , is mapped by @xmath746 to a ruling @xmath1478 of @xmath965 , and this ruling has no intersection point with @xmath407 off @xmath1464 .
this implies that @xmath1464 is an ordinary 6tuple point for @xmath407 and that the tangent cone to @xmath407 at @xmath1464 consists of the rulings @xmath1479 of @xmath965 .
[ rem:2 - 1cone ] let @xmath1480 be the image of @xmath1481 via the map defined by the linear system @xmath1482 .
one has @xmath1483 for @xmath13 , and the new central fibre @xmath356 is a double quadratic cone @xmath965 in @xmath9 . in this section
we describe the regular components of the limit severi varieties @xmath1484 for @xmath43 .
the discussion here parallels the one in [ s : tetra - concl ] , therefore we will be sketchy , leaving to the reader most of the straightforward verifications , based on the description of the limit linear system in [ s : kummer - ls ] .
[ limits of @xmath330-nodal curves ] [ p:1-kummer ] the regular components of the limit severi variety @xmath908 are the following ( they all appear with multiplicity @xmath330 , but the ones in which appear with multiplicity @xmath33 ) : + [ it : kummer ] @xmath1485 , which is isomorphic to the kummer quartic surface @xmath1486 @xmath1487 , which is isomorphic to the smooth quadric @xmath1488 , for @xmath1489 ; + [ it : double ] @xmath1490 , which is isomorphic to a quadric cone in @xmath1491 , for @xmath1489 ; + @xmath1492 , which is isomorphic to @xmath1488 , for @xmath1493 ; + @xmath1494 , for @xmath1495 such that @xmath1496 are pairwise distinct and do not meet a common @xmath1212 : it is again isomorphic to @xmath1488 ; + [ it : doubleconics ] @xmath1497 , which is isomorphic to the degree @xmath1435 surface @xmath1498 , for @xmath1214 .
+ [ theorem [ t : kummer ] for @xmath47 ] [ cor : thmb2i ] the family @xmath7 of general quartic surfaces degenerating to a kummer surface @xmath14 we started with , with smooth total space @xmath10 , and endowed with the line bundle @xmath234 , is @xmath330well behaved , with good model @xmath1499 .
the limit in @xmath22 of the dual surfaces @xmath932 , @xmath357 , consists in the union of the dual @xmath1500 of @xmath14 ( which is again a kummer surface ) , plus the @xmath28 planes of the @xmath1099 configuration of @xmath1500 , each counted with multiplicity @xmath33 .
forward by the morphism @xmath1501 with their respective multiplicities in @xmath1502 is @xmath1500 in case , @xmath1503 in case , and @xmath3 otherwise .
the push - forward of @xmath933 has thus total degree @xmath392 , and is therefore the crude limit severi variety @xmath928 by proposition [ p : deg - dual ] .
[ rem : dual2 ] similar arguments show that @xmath931 is a @xmath330good model for the degenerations of general quartic surfaces obtained from @xmath993 via the line bundles @xmath1504 and @xmath1505 respectively ( see remarks [ rem:2 - 1cone ] and [ rem : one - node ] for a description of these degenerations ) . to see this in the former case ,
let us consider two general points on a given @xmath1212 , and enumerate the regular members of @xmath1506 that contain them .
there are @xmath33 curves in ( indeed , the two points on @xmath1212 project to two general points on @xmath1507 , which span a line @xmath1508 ; the limiting curves in @xmath1509 passing through the two original points on @xmath1212 correspond to the intersection points of @xmath1510 with @xmath32 ; now @xmath1510 meets @xmath32 with multiplicity @xmath33 at the double point which is the image of @xmath1511 via the gauss map , and only the two remaining intersection points are relevant ) .
there are in addition @xmath33 limiting curves in each of the @xmath781 components of type corresponding to the @xmath1512 s that do not meet @xmath1212 , and @xmath1435 in the relevant component of type . in this case , the crude limit severi variety therefore consists , in the notation of remark [ rem:2 - 1cone ] , of the degree @xmath1435 surface @xmath1513 , plus the plane @xmath1514 with multiplicity @xmath1515 ( this has degree @xmath392 as required ) . for the degeneration given by @xmath1516 , the crude limit severi variety consists , in the notation of remark [ rem : one - node ] , of the dual to the smooth quadric @xmath999 , plus the dual to the conic @xmath254 with multiplicity @xmath33 , plus the fifteen planes @xmath938 with multiplicity @xmath33 , where @xmath72 ranges among the fifteen marked points on the double plane @xmath1023 .
[ rem : not - good ] one can see that @xmath931 is not a @xmath330good model for the degeneration to a union of two double planes obtained via the line bundle @xmath1517 described in remark [ rem : two - nodes ] . in addition ( see step ( [ kum : coplanar]a ) ) the line bundles @xmath1518 , though corresponding to @xmath16dimensional components of the limit linear system , do not provide suitable degenerations of surfaces . despite all this , it seems plausible that one can obtain a good model by making further modifications of @xmath993 . the first thing to do would be to blow up the curves @xmath1245 .
[ limits of @xmath33-nodal curves ] [ p:2-kummer ] the regular components of the limit severi variety @xmath947 are the following ( they all appear with multiplicity @xmath330 , except the ones in appearing with multiplicity @xmath33 ) : + [ it:2quadrics ] @xmath1519 for @xmath1520 , proper transform of the intersection of two smooth quadrics in @xmath1306 ; + [ it:2quadrics2 ] @xmath1521 for @xmath1520 , proper transform of the intersection of a smooth quadric and a quadric cone in @xmath1306 ; + [ it:2quadrics - spe ] @xmath1522 for @xmath1495 such that @xmath1496 are pairwise distinct and do not meet a common @xmath1212 , proper transform of the intersection of two smooth quadrics in @xmath1253 ; + [ it:2conic-2 ] @xmath1523 for each @xmath1524 , proper transform of a degree @xmath1436 curve in @xmath1414 .
+ again , one checks that the components listed in the above statement are the only ones provided by proposition [ p : zrr ] , taking the following points into account : + the condition @xmath1525 is impossible to fulfil , because there is no plane of @xmath9 tangent to @xmath14 at exactly two points ( see proposition [ p : dual - kummer ] ) ; + the condition @xmath1526 is also impossible to fulfil , because there is no plane in @xmath9 tangent to @xmath14 at exactly one point , and passing through one of its double points .
indeeed , let @xmath442 be a double point of @xmath14 , the dual plane @xmath1527 is everywhere tangent to @xmath32 along the contact conic gauss image of @xmath230 ; + the condition @xmath1528 imposes to a member of @xmath1346 to be the sum of two rulings intersecting at a point on @xmath567 , and such a curve does not belong to the limit severi variety : + the condition @xmath1529 imposes to contain one of the two curves @xmath1208 intersecting both @xmath1311 and @xmath1312 , which violates condition of definition [ d : zrr ] .
[ rem:480_2 ] as in remark [ rem:480 ] , we can enumerate the @xmath393 limits of @xmath33nodal curves passing through a general point in certain irreducible components of @xmath613 : + [ case : kummer-2 ] for a general point on @xmath439 , we find @xmath24 limit curves in each of the @xmath1530 components in of proposition [ p:2-kummer ] ; + [ case:2conic-2 ] for a general point on a given @xmath1212 , we find @xmath1436 limit curves in the appropriate component in , and @xmath24 in each of the @xmath1531 different components of type such that @xmath1313 and @xmath1314 do not both meet @xmath1212 .
this shows that @xmath993 is a @xmath33good model for the degenerations of quartics corresponding to the line bundles @xmath178 and @xmath1504 .
in particular , it implies corollary [ cor : thmb2ii ] below .
[ theorem [ t : kummer ] for @xmath49 ] [ cor : thmb2ii ] same setting as in corollary [ cor : thmb2i ] .
the crude limit severi variety @xmath974 consists of the images in @xmath975 of the @xmath34 irreducible curves listed in case of remark [ rem:480_2 ] .
each of them projects @xmath1532 onto a pencil of planes containing two double points of @xmath14 .
[ limits of @xmath16-nodal curves ] [ p:3 - 4planes_kummer ] the family @xmath993 is absolutely @xmath16good , and the limit severi variety @xmath1533 is reduced , consisting of : + [ item : aa ] @xmath36 distinct points in each @xmath1534 , where @xmath1495 are such that @xmath1535 is a plane that does not belong to the @xmath1099 configuration of @xmath14 ; + the @xmath35 distinct points in each @xmath1536 , corresponding to the triple points of the double curve of @xmath1537 that are also triple points of @xmath1513 . there are @xmath25 unordered triples @xmath1241 such that the corresponding double points of @xmath14 do not lie on a common @xmath1208 , so @xmath1538 has degree @xmath394 , which fits with proposition [ p : deg - dual ] .
[ theorem [ t : kummer ] for @xmath50 ] [ cor : thmb2iii ] same setting as in corollary [ cor : thmb2i ] .
the crude limit severi variety @xmath1539 consists of : + the @xmath25 points corresponding to a plane through three nodes of @xmath14 , but not member of the @xmath1099 configuration , each counted with multiplicity @xmath36 ; + the @xmath28 points corresponding to a member of the @xmath1099 configuration , each counted with multiplicity @xmath35 .
in this section we prove the key result needed for the proof of theorem [ t : triangle ] , itself given in [ s : tetra - concl ] .
we believe this result , independently predicted with tropical methods by e. brugall and g. mikhalkin ( private communication ) , is interesting on its own .
its proof shows once again the usefulness of constructing ( relative ) good models .
the general framework is the same as that of [ s:4planes ] and [ s : kummer - degen ] , and we are going to be sketchy here .
we start with the trivial family @xmath1540 , together with flatly varying data for @xmath1237 of three independent lines @xmath1541 lying in @xmath6 , and of a @xmath3-dimensional scheme @xmath1542 of degree @xmath38 cut out on @xmath1543 by a quartic curve @xmath1544 in @xmath6 , which is general for @xmath2 .
we denote by @xmath234 the pull back line bundle of @xmath1545 via the projection @xmath1546 .
up @xmath10 along the line @xmath826 .
this produces a new family @xmath1547 , the central fibre @xmath1548 of which is the transverse union of a plane @xmath224 ( the proper transform of @xmath14 , which we may identify with @xmath224 ) and of an @xmath886 surface @xmath182 ( the exceptional divisor ) .
the curve @xmath1549 is the line @xmath826 in @xmath224 , and the @xmath436 section in @xmath182 .
the limit on @xmath1548 of the three lines @xmath1550 on the fibre @xmath1551 , for @xmath138 , consists of : + two general lines @xmath1552 in @xmath224 plus the curves @xmath1553 matching them on @xmath112 ; + a curve @xmath1554 on @xmath182 .
we denote by @xmath1555 the pull back of @xmath234 and we set @xmath1556 .
one has @xmath1557 for @xmath138 , whereas @xmath1558 restrict to @xmath1559 and @xmath1560 respectively .
we may assume that the quartic curve @xmath1561 cutting @xmath1542 on @xmath1543 for @xmath138 tends , for @xmath1562 , to a general curve @xmath1563 .
then @xmath1564 , where @xmath1565 is a general cubic in @xmath224 and @xmath1566 , with @xmath1565 and @xmath1567 matching along @xmath112 .
accordigly @xmath1568 , where @xmath1569 has length 6 consisting of @xmath16 points on @xmath1160 and 3 on @xmath1161 , and @xmath1570 consists of @xmath330 point on both @xmath1179 and @xmath1181 , and @xmath24 points on @xmath1162 ( see figure [ f : triangle ] ) .
next we consider the blow up @xmath1571 along the curve @xmath39 in @xmath121 described by @xmath1542 , for @xmath136 , and thus obtain a new family @xmath514 , where each @xmath1572 is the blow up of @xmath1573 along @xmath1542 .
we call @xmath1574 the exceptional divisor of @xmath1575 .
the fibre of @xmath1576 at @xmath136 consists of the twelve @xmath436curves of the blow up of @xmath1573 at @xmath1542 .
the central fibre @xmath431 is the transverse union of @xmath243 and @xmath778 , respectively the blow
ups of @xmath224 and @xmath182 along @xmath1569 and @xmath1570 ; we denote by @xmath1577 and @xmath1578 the corresponding exceptional divisors .
we let @xmath1579 .
recall from
[ s : planes ] that the fibre of @xmath1580 over @xmath833 has dimension @xmath16 .
we will see that @xmath1581 , endowed with @xmath178 , is well behaved and we will describe the crude limit severi variety @xmath275 for @xmath43
. this analysis will prove theorem [ t : triangle ] .
[ rem : central ] we shall need a detailed description of the linear system @xmath1582 . the vector space @xmath1583 is the subspace of @xmath1584 which is the fibred product corresponding to the cartesian diagram @xmath1585 \ar[rr ] & & { \mathrm{h}}^ 0(\tilde p , { \mathcal{o}}_{\tilde p}(3h - e_p ) ) \ar[d]^ { r_p } \\ { \mathrm{h}}^ 0(\tilde w , { \mathcal{o}}_{\tilde w } ( 4h-3e - e_w ) )
\ar[rr]_{r_w } & & { \mathrm{h}}^ 0(e,\left .{\mathcal l}\right|_{e})\cong { \mathrm{h}}^ 0({\mathbf{p}}^ 1 , { \mathcal{o}}_{{\mathbf{p}}^ 1}(3 ) ) } \ ] ] where @xmath1586 are restriction maps .
the map @xmath1587 is injective , whereas @xmath1588 has a 1-dimensional kernel generated by a section @xmath1589 vanishing on the proper transforms of @xmath40 . since @xmath1590 by semicontinuity
, one has @xmath1591 , and therefore @xmath1592 has also dimension 4 .
geometrically , for a general curve @xmath1593 , there is a unique curve @xmath1594 matching it along @xmath112 and @xmath1595 . on the other hand
@xmath1596 is the only non trivial section ( up to a constant ) identically vanishing on a component of the central fibre ( namely @xmath778 ) , and @xmath1597 .
therefore , if we denote by @xmath118 the point corresponding to @xmath1598 in @xmath203 , a line through @xmath118 parametrizes the pencil consisting of a fixed divisor in @xmath1599 on @xmath778 plus all divisors in @xmath1600 matching it on @xmath112 .
we will denote by @xmath1601 the @xmath1602 on @xmath112 given by @xmath1603 . to get a good model
, we first blow up the proper transform of @xmath1160 in @xmath243 , and then we blow up the proper transform of @xmath1161 on the strict transform of @xmath243 .
we thus obtain a new family @xmath935 .
the general fibre @xmath1604 , @xmath138 , is isomorphic to @xmath1572 .
the central fibre @xmath526 has four components ( see figure [ f : triangle2 ] ) : + the proper transform of @xmath243 , which is isomorphic to @xmath243 ; + the proper transform @xmath736 of @xmath778 , which is isomorphic to the blow up of @xmath778 at the two points @xmath1605 , with exceptional divisors @xmath1606 and @xmath1607 ; + the exceptional divisor @xmath1608 of the last blow up , which is isomorphic to @xmath1609 ; + the proper transform @xmath1610 of the exceptional divisor over @xmath1160 , which is the blow up of an @xmath1609surface , at the point corresponding to @xmath1611 ( which is a general point of @xmath1609 ) with exceptional divisor @xmath1612 . as usual , we go on calling @xmath178 the pull back to @xmath893 of the line bundle @xmath178 on @xmath89
. we shall now describe the limit linear system @xmath200 associated to @xmath178 . as usual , we start with @xmath1613 , and we consider the blow up
@xmath1614 at the point @xmath1615 is , as we will see , the limit linear system @xmath200 .
it consists of only two components : the proper transform @xmath239 of @xmath203 and the exceptional divisor @xmath1616 .
let us describe these two components in terms of twisted linear systems on the central fibre .
since the map @xmath1587 in is injective , it is clear that @xmath1617 .
the line bundle @xmath1618 is trivial on @xmath243 and restricts to @xmath1619 on @xmath1620 respectively .
once chosen @xmath1621 , there is only one possible choice of two curves @xmath1622 and @xmath1623 in @xmath1624 and @xmath1625 respectively , that match with @xmath1626 along @xmath1627 respectively .
they automatically match along @xmath1612 . in conclusion , by mapping @xmath736 to @xmath562 ( via @xmath1628 ) , we have : [ propo : descrlim ] the component @xmath1629 of @xmath838 is isomorphic to a @xmath16-dimensional linear system of plane quartics with an imposed double point @xmath428 , prescribed tangent lines @xmath1630 at @xmath428 , and six further base points , two of which general on @xmath1630 and the remaining four on a general line . to identify @xmath239 as the blow up of @xmath203 at @xmath118 , we take into account lemma [ l : tg - id ] , which tells us that the exceptional divisor @xmath1631 identifies with @xmath1601 . since @xmath1632 , the linear system @xmath1633 identifies with a sublinear system of codimension @xmath330 in @xmath244 , namely that of curves @xmath1634 ( in the setting of proposition [ propo : descrlim ] , @xmath117 corresponds to a quartic plane curve with a triple point at @xmath428 passing through the six simple base points ) .
it follows from this analysis that @xmath200 is the limit as @xmath1562 of the linear systems @xmath327 , @xmath2 , in the sense of [ s : resolving ] .
we will use the notion of @xmath1635degree introduced in definition [ def : enne ] .
however we will restrict our attention to the case in which we fix 1 or 2 points only on @xmath243 .
hence , if we agree to set @xmath1636 , then we call @xmath224degree of a component @xmath95 of @xmath273 its @xmath1635degree with @xmath1637 ; we denote it by @xmath1638 .
[ limits of @xmath330-nodal curves ] [ p:1-triangle ] the regular components of the limit severi variety @xmath1639 are the following , all appearing with multiplicity @xmath330 , except , which has multiplicity @xmath33 : + [ item:1 ] @xmath1640 , with @xmath224degree @xmath1641 ; + [ item:2 ] @xmath1642 , with @xmath224degree @xmath24 ; + [ item:3 ] @xmath1643 , with @xmath224degree @xmath24 ; + [ item:4 ] @xmath1644 , with @xmath224degree @xmath3 .
the list is an application of proposition [ p : zrr ] .
the only things to prove are the degree assertions . since @xmath244 is trivial on @xmath243 , case is trivial .
case follows from proposition [ p : deg - dual - sing ] , because the @xmath224degree of @xmath1640 is the degree @xmath1641 of the dual surface of the image @xmath1645 of @xmath243 via the linear system @xmath1600 , which is a cubic surface with an @xmath401 double point ( see proposition [ p : deg - dual - sing ] ) . as for , note that nodal curves in @xmath1646 on @xmath736 consist of a ruling in @xmath555 plus a curve @xmath117 in @xmath1647 .
if @xmath110 does not intersect one of the 4 exceptional curves in @xmath1578 meeting @xmath826 , then @xmath1648 and the matching curve on @xmath243 contains the proper transform of @xmath1160 and @xmath1161 , which is not allowed .
so @xmath110 has to contain one of the 4 exceptional curves in @xmath1578 meeting @xmath826 .
this gives rise to four pencils of singular curves in @xmath1599 , which produce ( see remark [ rem : central ] ) four @xmath33dimensional linear subsystems in @xmath203 , and this implies the degree assertion .
the degree assertion in follows from the fact that a @xmath1649 on @xmath112 has 4 ramification points .
[ limits of @xmath33-nodal curves ] [ p:2-triangle ] the regular components of the limit severi variety @xmath1650 are the following , all appearing with multiplicity @xmath330 , except and , which have multiplicity @xmath33 , and , which has multiplicity @xmath16 : + [ item:1a ] @xmath1651 , with @xmath224-degree @xmath1641 ; + [ item:2a ] @xmath1652 , with @xmath224-degree @xmath27 ; + [ item:3a ] @xmath1653 , with @xmath224-degree @xmath392 ; + [ item:4a ] @xmath1654 , with @xmath224-degree @xmath1655 ; + [ item:5a ] @xmath1656 , with @xmath224-degree @xmath36 ; + [ item:6a ] @xmath1657 , with @xmath224-degree @xmath16 ; + [ item:7a ] @xmath1658 , with @xmath224degree @xmath3 . again , the list is an immediate application of proposition [ p : zrr ] , and the only things to prove are the degree assertions .
once more case is clear . in case
the degree equals the number of lines on @xmath1645 ( the cubic surface image of @xmath243 ) , that do not contain the double point ; this is @xmath1641 . in case
, we have to consider the binodal curves in @xmath1599 not containing @xmath112 .
such curves split into a sum @xmath1659 , where @xmath1660 and @xmath1661 are the strict transforms of two curves in @xmath1662 .
they are uniquely determined by the choice of two curves in @xmath1578 meeting @xmath826 : these fix the two rulings in @xmath555 containing them , and there is a unique curve in @xmath1663 containing the remaining curves in @xmath1578 .
this shows that the degree is @xmath27 .
next , the limit curves of type consist of a nodal cubic in @xmath1664 and a nodal curve in @xmath1665 ; a ruling necessarily splits from the latter curve .
again , the splitting rulings @xmath110 are the ones containing one of the four curves in @xmath1578 meeting @xmath826 .
the curves in @xmath1666 containing the remaining curves in @xmath1578 , fill up a pencil .
let @xmath1667 be one of these four rulings .
the number of nodal curves in @xmath1664 passing through the base point @xmath1668 and through a fixed general point on @xmath243 equals the degree of the dual surface of @xmath1645 , which is 9 . for each such curve
, there is a unique curve in the aforementioned pencil on @xmath778 matching it .
this shows that the degree is 36 .
the general limit curve of type can be identified with the general plane of @xmath1669 which is tangent to both @xmath1645 and the curve @xmath1670 ( image of @xmath112 in @xmath1645 ) , at different points .
the required degree is the number of such planes passing through a general point @xmath72 of @xmath1645 .
the planes in question are parametrized in @xmath1671 by a component @xmath1672 of @xmath1673 : one needs to remove from @xmath1673 the component @xmath1674 , the general point of which corresponds to a plane which is tangent to @xmath1645 at a general point of @xmath1670 .
the latter appears with multiplicity @xmath33 in @xmath1675 by lemma [ l : tangency ] .
moreover , @xmath1676 and @xmath1677 have respective degrees @xmath1641 and @xmath24 by proposition [ p : deg - dual - sing ] .
thus we have @xmath1678 to compute @xmath1679 , take a general point @xmath1680 , and let @xmath1681 be the _ first polar _ of @xmath1645 with respect to @xmath168 , i.e. the surface of homogeneous equation @xmath1682 where @xmath1683 is the homogeneous equation of @xmath1645 .
the number of planes containing @xmath168 and tangent to @xmath1645 at a point of @xmath1670 is then equal to the number of points of @xmath1684 , distinct from the singular point @xmath1464 of @xmath1645 .
a local computation , which can be left to the reader , shows that @xmath1464 appears with multiplicity @xmath33 in @xmath1685 , which shows that @xmath1686 , whence @xmath1687 . in case
, we have to determine the curves in @xmath1599 with one node ( so that some ruling splits ) that are also tangent to @xmath112 . as usual , the splitting rulings are the one containing one of the four curves in @xmath1578 meeting @xmath826 . inside the residual pencil
there are 2 tangent curves at @xmath112 .
this yields the degree 8 assertion .
finally , in case , the degree equals the number of flexes of @xmath1670 , which is a nodal plane cubic : this is @xmath16 .
[ limits of @xmath16-nodal curves ] [ p:3-triangle ] the regular components of the limit severi variety @xmath1688 are the following @xmath3dimensional varieties , all appearing with multiplicity @xmath330 , except the ones in and appearing with multiplicity @xmath33 , and with multiplicity @xmath16 : + [ bitem1 ] @xmath1689 , which consists of @xmath27 points ; + [ bitem2 ] @xmath1690 , which consists of @xmath392 points ; + [ bitem3 ] @xmath1691 , which consists of @xmath1692 points ; + [ bitem4 ] @xmath1693 , which consists of @xmath1017 points ; + [ bitem5 ] @xmath1694 , which consists of @xmath1695 points ; + [ bitem6 ] @xmath1696 , which consists of @xmath1017 points ; + [ bitem7 ] @xmath1697 , which consists of @xmath27 points . in the course of the proof , we will need the following lemma . [
l : spec - pencil ] let @xmath1698 be general points on @xmath112 .
+ [ uno ] the pencil @xmath1699 of curves containing @xmath168 , and tangent to @xmath112 at @xmath72 , contains exactly @xmath1700 irreducible nodal curves not singular at @xmath72 .
+ [ due ] the pencil @xmath1701 of curves with a contact of order @xmath16 with @xmath112 at @xmath72 contains exactly @xmath27 irreducible nodal curves not singular at @xmath72 . first note that @xmath216 and @xmath1702 are indeed pencils by remark [ rem : central ] .
let @xmath1703 be the blow - up at @xmath72 and @xmath168 , with exceptional curves @xmath1704 and @xmath1705 above @xmath72 and @xmath168 respectively .
let @xmath1706 be the blow - up at the point @xmath1707 , with exceptional divisor @xmath1708 .
then @xmath216 pulls back to the linear system @xmath1709 , which induces an elliptic fibration @xmath1710 , with singular fibres in number of @xmath38 ( each counted with its multiplicity ) by lemma [ l : pencil ] . among them
are : the proper transform of @xmath1711 , which has @xmath16 nodes , hence multiplicity @xmath16 as a singular fibre ; the unique curve of @xmath216 containing the @xmath773-curve @xmath1704 , which has @xmath33 nodes along @xmath1704 , hence multiplicity @xmath33 as a singular fibre .
the remaining 7 singular fibres are the ones we want to count .
the proof of is similar and can be left to the reader .
there is no member of @xmath239 with 3 nodes on @xmath736 , because every such curve contains one of the curves @xmath1712 .
there is no member of @xmath239 with two nodes on @xmath736 and a tacnode on @xmath112 either .
indeed , the component on @xmath736 of such a curve would be the proper transform of a curve of @xmath182 consisting of two rulings plus a curve in @xmath1663 , altogether containing @xmath1570 .
each of the two lines passes through one of the points of @xmath1570 on @xmath826 .
the curve in @xmath1663 must contain the remaining points of @xmath1570 , hence it is uniquely determined and can not be tangent to @xmath112 .
then the list covers all remaining possible cases , and we only have to prove the assertion about the cardinality of the various sets .
the limiting curves of type are in one - to - one correspondence with the unordered triples of lines distinct from @xmath1160 and @xmath1161 in @xmath224 , the union of which contains the six points of @xmath1569 .
there are @xmath27 such triples .
the limiting curves of type consist of the proper transform @xmath1713 in @xmath243 of the union of a conic and a line on @xmath224 containing @xmath1569 , plus the union @xmath1626 of the proper transforms in @xmath736 of a curve in @xmath637 and one in @xmath1714 altogether containing @xmath1570 , with @xmath1713 and @xmath1626 matching along @xmath112 .
we have @xmath1641 possible pencils for @xmath1713 , corresponding to the choice of two points on @xmath1569 , one on @xmath1160 and one on @xmath1161 ; each such pencil determines by restriction on @xmath112 a line @xmath1715 .
there are @xmath24 possible pencils for @xmath1626 , corresponding to the choice of one of the points of @xmath1570 on @xmath826 : there is a unique ruling containing this point , and a pencil of curves in @xmath1714 containing the five remaining points in @xmath1570 ; each such pencil defines a line @xmath1716 . for each of the above choices , the lines @xmath216 and @xmath1702 intersect at one point , whence the order @xmath392 . we know from the proof of proposition [ p:2-triangle ] that there are six @xmath33nodal curves in @xmath1599 .
for each such curve , there is a pencil of matching curves in @xmath1600 .
this pencil contains @xmath1717 nodal curves , whence the number @xmath1692 of limiting curves of type .
the component on @xmath243 of a limiting curve of type is the proper transform of the union of a conic and a line on @xmath224 , containing @xmath1577 .
as above , there are @xmath1641 possible choices for the line . for each such choice
, there is a pencil of conics containing the @xmath24 points of @xmath1569 not on the line .
this pencil cuts out a @xmath1718 on @xmath112 , and therefore contains @xmath33 curves tangent to @xmath112 .
it follows that there are @xmath1017 limiting curves of type .
the component on @xmath736 of a limiting curve of type is the proper transform of a ruling of @xmath182 plus a curve in @xmath1714 tangent to @xmath112 , altogether passing through @xmath1570 .
the line necessarily contains one of the four points of @xmath1570 on @xmath826 .
there is then a pencil of curves in @xmath1714 containing the five remaining points of @xmath1570 .
it cuts out a @xmath1718 on @xmath112 , hence contains @xmath33 curves tangent to @xmath112 . for any such curve @xmath1626 on @xmath736 ,
there is a pencil of curves on the @xmath243side matching it . by lemma [ l : spec - pencil ] ,
this pencil contains @xmath1700 curves , the union of which with @xmath1626 is a limiting curve of type .
this proves that there are @xmath1695 such limiting curves . as for
, there are @xmath16 members of @xmath1601 that are triple points ( see the proof of proposition [ p:2-triangle ] ) .
each of them determines a pencil of curves on the @xmath243side , which contains six @xmath330-nodal curves by lemma [ l : spec - pencil ] .
this implies that there are @xmath1017 limiting curves of type
. finally we have to count the members of @xmath1697 .
they are in one to
one correspondence with their components on @xmath1719 , which decompose into the proper transform of unions @xmath1720 of two curves @xmath1721 and @xmath1722 , altogether containing @xmath1570 .
the curves @xmath1723 must contain the two base points on @xmath1724 respectively . we conclude that each limiting curve of type corresponds to a partition of the 4 points of @xmath1570 on @xmath826 in two disjoint sets of two points , and the assertion follows . in conclusion ,
the following is an immediate consequence of propositions [ p:1-triangle ] , [ p:2-triangle ] , and [ p:3-triangle ] , together with the formula .
[ preliminary version of theorem [ t : triangle ] ] [ coro : d ] let @xmath37 be three independent lines in the projective plane , and @xmath39 be a degree @xmath38 divisor on @xmath40 cut out by a general quartic curve .
we consider the @xmath16dimensional sub linear system @xmath41 of @xmath42 parametrizing curves containing @xmath39 , and we let , for @xmath43 , @xmath44 be the zariski closure in @xmath41 of the codimension @xmath5 locally closed subset parametrizing irreducible @xmath5nodal curves .
one has @xmath1725 [ rem : d ] * ( theorem [ t : triangle ] ) * the three inequalities in above are actually equalities .
this is proved in
[ s : tetra - concl ] , by using both and the degrees of the severi varieties of a general quartic surface , given by proposition [ p : deg - dual ] .
incidentally , this proves that @xmath935 is a good model for the family @xmath1726 obtained by blowing
up @xmath1727 along @xmath39 , and endowed with the appropriate subline bundle of @xmath1728 .
[ rem : d - red ] in particular , we have @xmath1729 .
it then follows from remark [ r : smoothness ] that the relative severi variety @xmath1730 is smooth at the points of @xmath1731 .
this implies that the general fibre of @xmath1730 is reduced .
therefore , in the setting of corollary [ coro : d ] , if @xmath40 and @xmath39 are sufficiently general , then @xmath1732 consists of @xmath31 distinct points .
set @xmath1733 .
we have the _ universal family _
@xmath1734 , such that the fibre of @xmath72 over @xmath1735 is the linear system @xmath1736 .
the variety @xmath707 is a component of the _ flag hilbert scheme _ , namely the one parametrizing pairs @xmath1737 , where @xmath117 is a plane quartic curve in @xmath21 and @xmath1735 contains @xmath117 .
so @xmath1738 , where @xmath678 is the component of the hilbert scheme of curves in @xmath21 whose general point corresponds to a plane quartic .
the map @xmath72 is the projection to the first factor ; we let @xmath168 be the projection to the second factor . denote by @xmath1739 the open subset parametrizing smooth surfaces , and set @xmath1740 . inside @xmath1741
we have the _ universal severi varieties _
@xmath1742 , @xmath43 , such that for all @xmath1743 , the fibre of @xmath1742 over @xmath10 is the severi variety @xmath1744 . since @xmath10 is a @xmath1 surface , we know that for all irreducible components @xmath95 of @xmath1744 , we have @xmath1745 , so that all components of @xmath1742 have codimension @xmath5 in @xmath1741 .
we then let @xmath1746 be the zariski closure of @xmath1742 in @xmath1747 ; we will call it universal severi variety as well .
it suffices to consider the projection @xmath1748 , and notice that its image is the irreducible variety whose general point corresponds to a quartic curve with @xmath5 nodes ( cf .
@xcite ) , and that the fibres are all irreducible of the same dimension 20 .
note that the irreducibility of @xmath1749 also follows from the fact that for all @xmath1743 , we have @xmath1750 . to the other extreme , @xmath1751 is a finite cover of degree 3200
. we will denote by @xmath1752 the _ monodromy group of this covering _ , which acts transitively because @xmath53 is irreducible .
in the middle we have @xmath1753 . though @xmath1754 is irreducible
, we can not deduce from this that for the general @xmath1743 , the severi variety @xmath1755 ( i.e. , the curve of binodal plane sections of @xmath10 ) is irreducible . though commonly accepted as a known fact , we have not been able to find any proof of this in the current literature .
it is the purpose of this paragraph to provide a proof of this fact . in any event
, we have a commutative diagram similar to the one in @xmath1756^\nu \ar[dr ] \ar[d]_{p ' } & { \mathcal v^\circ}_2 \ar[d]^p \\ \mathcal u ' \ar[r]_{f } & { \mathcal}u } \ ] ] where @xmath398 is the normalization of @xmath1757 , and @xmath1048 is the stein factorization of @xmath1758 .
the morphism @xmath1050 is finite , of degree @xmath559 equal to the number of irreducible components of @xmath1755 for general @xmath1743 .
the monodromy group of this covering acts transitively .
this ensures that , for general @xmath1743 , all irreducible components of @xmath1759 have the same degree , which we denote by @xmath1760 . by proposition [ prop : irrcomp ]
, we have @xmath1761 .
let @xmath14 be a general quartic kummer surface , and @xmath1763 a family of surfaces induced as in example [ ex : pi3 ] by a pencil generated by @xmath14 and a general quartic @xmath222 .
given two distinct nodes @xmath72 and @xmath168 of @xmath14 , we denote by @xmath1764 the pencil of plane sections of @xmath14 passing through @xmath72 and @xmath168 . corollary [ cor : thmb2ii ] asserts that the union of these lines , each counted with multiplicity 4 , is the crude limit severi variety @xmath1765 .
let @xmath1544 be an irreducible component of @xmath1766 , for @xmath2 , and let @xmath1767 be its ( crude ) limit as @xmath15 tends to @xmath3 , which consists of a certain number of ( quadruple ) curves @xmath1768 .
note that , by proposition [ p:2-kummer ] , the pull back of the lines @xmath1764 to the good limit constructed in [ s : kummer - degen ] all appear with multiplicity 1 in the limit severi variety .
this yields that , if @xmath216 is an irreducible component of @xmath1767 , then it can not be in the limit of an irreducible component @xmath1769 of @xmath1766 other than @xmath1544 .
if @xmath1767 does not verify , then it contains at most 8 curves of type @xmath1768 , a contradiction to @xmath1775 . to prove
, we consider two curves @xmath1768 and @xmath1770 contained in @xmath1767 , and assume that @xmath1772 do not lie on a contact conic , otherwise there is nothing to prove .
consider a degeneration of @xmath14 to a product kummer surface @xmath1776 , and let @xmath1777 be the limits on @xmath1776 of @xmath1772 respectively : they are necessarily in one of the three configurations depicted in figure [ f : triples - exc ] . in all three cases ,
we can exchange two horizontal lines in @xmath1778 ( as indicated in figure [ f : triples - exc ] ) , thus moving @xmath1779 to @xmath1780 , in such a way that @xmath1781 and @xmath1782 remain fixed , and there is a limit in @xmath1778 of contact conics that contains the three points @xmath1781 , @xmath1779 , and @xmath1780 . accordingly , there is an element @xmath1783 mapping @xmath1772 to @xmath1784 respectively , such that @xmath1785 lie on a contact conic @xmath118 of @xmath14 .
then @xmath1786 contains @xmath1787 . by the remark preceding the statement of , we have @xmath1788 .
it follows that @xmath1767 contains @xmath1789 and @xmath1790 , and therefore satisfies .
claim follows from and the fact that the monodromy acts as @xmath1791 on the set of nodes lying on @xmath118 ( see lemma [ l : s_6 ] ) . as for
, let @xmath1792 be any other contact conic of @xmath14 .
there exists @xmath1783 interchanging @xmath118 and @xmath1792 ( again by lemma [ l : s_6 ] ) .
the action of @xmath1793 preserves @xmath1794 .
we know that @xmath1767 contains @xmath1795 with @xmath1796 different from @xmath1797
. then the same argument as above yields that @xmath1767 contains @xmath1798 , where @xmath1799 and @xmath1800 .
this implies that @xmath1767 satisfies for @xmath1792 , and therefore holds for @xmath1792 .
finally implies .
[ rem : restrict ] let @xmath1806 be a dominant , generically finite morphism of degree @xmath1760 between projective irreducible varieties , with monodromy group @xmath1807 .
let @xmath1808 be an irreducible codimension @xmath330 subvariety , the generic point of which is a smooth point of @xmath121 .
then @xmath1809 is still generically finite , with monodromy group @xmath1810 .
if @xmath95 is not contained in the branch locus of @xmath135 , then @xmath1811 .
suppose to the contrary that @xmath95 is contained in a component of the branch locus of @xmath135 .
then @xmath1812 , with @xmath1813 , and @xmath1810 is no longer a subgroup of @xmath92 .
we can however consider the _
local monodromy group @xmath1814 of @xmath135 around @xmath95 _ , i.e. the subgroup of @xmath1807 generated by permutations associated to non trivial loops _
turning around _
precisely : let @xmath1815 be a tubular neighbourhood of @xmath95 in @xmath121 ; then @xmath1814 is the image in @xmath92 of the subgroup @xmath1816 of @xmath1817 .
there is an epimorphism @xmath1818 , obtained by deforming loops in @xmath1819 to loops in @xmath95 .
we let @xmath1820 be the kernel of this epimorphism , so that one has the exact sequence of groups @xmath1821 we first apply this to the degeneration studied in [ s:4planes ] . to this end , we consider the 12dimensional subvariety @xmath1822 of @xmath1823 which is the zariski closure of the set of fourtuples of distinct planes .
let @xmath1824 be the blow up of @xmath1825 along @xmath1822 , with exceptional divisor @xmath1826 .
the proof of the following lemma ( similar to lemma [ l : tg - id ] ) can be left to the reader : thus , for general @xmath1829 , a general point of the fibre of @xmath135 over @xmath89 can be identified with a pair @xmath1830 , with @xmath1831 general , where @xmath1828 . consider a family @xmath7 of surfaces in @xmath9 , induced as in example [ ex : pi3 ] by a pencil @xmath1832 generated by @xmath89 and a general quartic ; then the singular locus of @xmath10 is a member of @xmath1827 , which corresponds to the tangent direction normal to @xmath1822 defined by @xmath216 in @xmath1823 .
now the universal family @xmath1833 can be pulled back to @xmath1834 , and the analysis of [ s:4planes ] tells us that we have a generically finite map @xmath1835 , which restricts to @xmath1836 over @xmath1837 , and such that @xmath1838 is in the branch locus of @xmath1839 .
we let @xmath1840 be the monodromy group of @xmath1835 on @xmath1841 , and @xmath1842 , resp .
@xmath1843 , be as in . [ it:666 ] @xmath1850 , where @xmath1851 is the monodromy group of bitangent lines to @xmath330nodal plane quartics as in proposition [ p : projectw ] , and @xmath1852 is the monodromy group of irreducible trinodal curves in the linear system of quartic curves with 12 base points at a general divisor of @xmath1853 , with @xmath37 three lines not in a pencil ( see [ s : triangle ] ) .
the proof follows from corollary [ cor : thmbiii ] .
recall that a group @xmath1807 is equal to @xmath1854 , if and only if it contains a transposition and it is doubly transitive . using this , it is easy to verify the assertions in
( see @xcite ) . as for
, the factor @xmath1855 comes from the fact that the monodromy acts as the full symmetric group on a general line section of the irreducible cubic surface @xmath789 as in proposition [ p : cubics ] .
analogous considerations can be made for the degeneration studied in
[ s : kummer - degen ] . in that case , we consider the 18dimensional subvariety @xmath1856 of @xmath1823 which is the zariski closure of the set @xmath1857 of kummer surfaces .
let @xmath1858 be the blow up along @xmath1856 , with exceptional divisor @xmath1859 .
in this case we have : the universal family @xmath1833 can be pulled back to @xmath1863 .
the analysis of [ s : kummer - degen ] tells us that we have a map @xmath1864 , generically finite over @xmath1865 , which is in the branch locus of @xmath1839 .
we let @xmath1866 be the monodromy group of @xmath1839 on @xmath1867 , and set @xmath1868 and @xmath1869 as in . [
it : sic ] @xmath1870 , where @xmath1871 is the monodromy group of unordered triples of distinct nodes of a general kummer surface which do not lie on a contact conic ( see [ s : monodromy ] for the definition of @xmath1141 ) ; + part follows right away from proposition [ p:3 - 4planes_kummer ] .
part also follows , since the monodromy on complete intersections of three general quadrics in @xmath21 ( which gives the multiplicity 8 in of proposition [ p:3 - 4planes_kummer ] ) is clearly the full symmetric group . concerning the group @xmath1871 appearing in proposition [ prop : monodrom2 ] ,
, remember that it acts with at most two orbits on the set of of unordered triples of distinct of nodes of a general kummer surface which do not lie on a contact conic ( see proposition [ p:3-transitivity ] , ) . | we describe , for various degenerations @xmath0 of quartic @xmath1 surfaces over the complex unit disk ( e.g. , to the union of four general planes , and to a general kummer surface ) , the limits as @xmath2 tends to @xmath3 of the severi varieties @xmath4 , parametrizing irreducible @xmath5-nodal plane sections of @xmath6 .
we give applications of this to the counting of plane nodal curves through base points in special position , the irreducibility of severi varieties of a general quartic surface , and the monodromy of the universal family of rational curves on quartic @xmath1 surfaces . |
the @xmath1ray spectra and yields @xcite@xcite , produced by the atomic transitions @xmath2 in kaonic hydrogen , where @xmath3 is the _ principle quantum number _ of the energy levels , are the main experimental tool for the measurement of the energy level displacement of the ground state of kaonic hydrogen , caused by strong low energy @xmath4 interactions
it is known @xcite@xcite that the @xmath1ray yields related to the @xmath5 , @xmath6 and @xmath7 lines of kaonic hydrogen are very sensitive to the value of @xmath8 , the rate of hadronic decays of kaonic hydrogen from the @xmath9 state .
usually @xmath8 is used as an input parameter in the theories of the atomic cascades @xcite@xcite . recently @xcite we have calculated the rate @xmath10 of hadronic decays of kaonic hydrogen from the @xmath11 state . for the @xmath9 state
we have obtained : @xmath12 .
this agrees well with the assumption by koike , harada , and akaishi @xcite . in order to reconcile the experimental data on the @xmath5 , @xmath6 and @xmath7 lines with the theoretical analysis they assumed that @xmath13 . in this paper
we continue the analysis of the influence of strong low
energy interactions on the transitions from the @xmath11 state in kaonic hydrogen , which we have started in @xcite .
we investigate the radiative transitions @xmath14 , induced by strong low energy interactions and enhanced by the coulomb interaction of the @xmath15 pair , in kaonic hydrogen and extend the results on kaonic deuterium .
the paper is organized as follows . in section 2
we calculate the rates of the radiative transitions @xmath14 in kaonic hydrogen , induced by strong low
energy interactions and enhanced by the coulomb interaction .
the coulomb interaction is taken into account in the form of the explicit non relativistic coulomb wave functions of the relative motion of the @xmath15 pairs and the explicit non relativistic coulomb green functions for the calculation of the amplitude of the kaon
proton _ bremsstrahlung _
@xmath16 , defining the rate of the radiative atomic transition @xmath17 in our approach . in section 3
we modify the rates of the transitions @xmath14 , calculated in section 2 for kaonic hydrogen , for kaonic deuterium . in the conclusion
we discuss the obtained results .
the rate of the radiative transition @xmath18 can be defined by @xcite @xmath19 where @xmath20 is the photon energy , @xmath21 is the fine structure constant , @xmath22 is the reduced mass of the @xmath15 pair calculated for @xmath23 and @xmath24 .
the amplitude @xmath25 is given by @xcite @xmath26 where @xmath27 and @xmath28 are the wave functions of kaonic hydrogen in the ground @xmath29 and the @xmath11 state in the momentum representation @xcite ; @xmath30 is a spherical harmonic and @xmath31 and @xmath32 are radial wave functions of kaonic hydrogen in the momentum representation @xcite ( see also @xcite ) .
then , @xmath33 is the amplitude of the kaon
proton _ bremsstrahlung
_ @xmath34 .
the amplitude @xmath35 is defined by the feynman diagrams depicted in fig.1 . for the calculation of this amplitude we use the following effective lagrangian , describing strong low energy and electromagnetic interactions of the @xmath15 pairs : @xmath36 where @xmath37 is the electric charge of the proton such as @xmath38 , @xmath39 is the s wave scattering length of @xmath15 scattering @xcite , @xmath40 is the vector potential of the quantized electromagnetic field .
the ellipses denote interactions of order of @xmath41 , which we omit . in the effective lagrangian eq.([label2.3 ] ) strong low
energy interactions of the @xmath15 pair are described by @xmath42 . caused by strong low
energy interactions ] in the non relativistic limit the amplitude of the kaon proton _ bremsstrahlung
_ @xmath34 in fig.1 reads @xmath43 for the calculation of the amplitude of the atomic transition @xmath44 the amplitude of the kaon
proton _ bremsstrahlung
_ eq.([label2.4 ] ) should be weighted with the wave functions of kaonic hydrogen in the ground and excited @xmath11 states eq.([label2.2 ] ) . since the wave function of the ground state is spherical symmetric , the integration over @xmath45 leads to the vanishing of the term proportional to @xmath46 .
therefore , below we omit it . the appearance of the photon energy @xmath47 in the denominator of the amplitude of the kaon proton
_ bremsstrahlung _
@xmath48 is due to the fact that the virtual @xmath49meson and the proton are practically on mass shell .
our approach to the calculation of the amplitude of the kaon
proton _ bremsstrahlung _
@xmath34 is similar to that which has been used in @xcite for the derivation of the ericson
weise formula for the s wave scattering amplitude of @xmath50 scattering @xcite and runs parallel the non relativistic effective field theory ( the eft ) approach , based on chiral perturbation theory ( chpt ) by gasser and leutwyler @xcite , which has been applied by meiner _
@xcite to the systematic calculation of qcd isospin breaking and electromagnetic corrections to the energy level displacement of the @xmath51 state of kaonic hydrogen .
in @xcite we have analysed the quantitative agreement of the results on the energy level displacement of the @xmath51 state of kaonic hydrogen obtained by meiner s group @xcite , our group @xcite and the experimental data by the dear collaboration @xcite . due to the coulomb interaction of the @xmath15 pair in initial and final states @xcite@xcite the amplitude of the kaon
proton _ bremsstrahlung
_ eq.([label2.4 ] ) changes as follows @xmath52 where @xmath53 is the bohr radius and the @xmath54functions are defined by @xcite @xmath55\big\},\nonumber\\ e^{\textstyle\,\pi z/2}\,\gamma(2 - i z ) & = & ( 1 - iz)\ , e^{\textstyle\,\pi z/2}\,\gamma(1 - i z)\end{aligned}\ ] ] with euler s constant @xmath56 @xcite . for the calculation of the amplitude eq.([label2.5 ] ) , taking into account the coulomb interaction of the @xmath15 pair in the initial and final state , we have used the potential model approach . in this approach
we describe strong low
energy interactions by the effective zero range potential @xmath57 which is equivalent to the effective local strong low energy @xmath58interaction @xmath59 in eq.([label2.3 ] ) . including the coulomb interaction the term @xmath60 in eq.([label2.4 ] )
becomes replaced by @xmath61 where @xmath62 is the exact non relativistic coulomb wave function of the relative motion of the @xmath15 pair in the incoming scattering state with relative momentum @xmath63 .
it is given by @xcite @xmath64 here @xmath65 is the confluent function @xcite .
the factor @xmath66 in eq.([label2.5 ] ) is the rest of the asymptotic of the non relativistic coulomb green function @xmath67 @xcite @xmath68 of the relative motion of the @xmath15 pair at @xmath69 , where @xmath70 is the whittaker function @xcite , describing the outgoing spherical wave distorted by the coulomb interaction @xcite . for the amplitude of the transition
@xmath18 we get @xmath71 the main contributions to the integrals over @xmath45 and @xmath63 come from the regions @xmath72 and @xmath73 @xcite .
these momenta are of order of @xmath74 .
therefore , the contribution of the photon momentum can be dropped , since it is of order of @xmath75 . for the integration over @xmath63
we define the vector @xmath63 in the spherical basis @xcite @xmath76 where @xmath77 and @xmath78 are spherical unit vectors expended into cartesian unit vectors @xmath79 , @xmath80 and @xmath81 . using eq.([label2.12 ] ) for the integral over @xmath63 we get @xcite @xmath82 for the integration over @xmath83 and @xmath84 we use the wave functions @xmath27 and @xmath32 given by @xcite @xmath85 where @xmath86 is the gegenbauer polynomial @xcite . using @xcite and taking into account the results obtained in @xcite we make the integration over momenta @xmath83 and @xmath84 : @xmath87 where @xmath88 and @xmath89 and @xmath90 and @xmath91 are real phases , which do not contribute to the rate of the transition @xmath44 .
the amplitude of the transition @xmath18 reads @xmath92 the rate of the transition @xmath44 is equal to @xmath93 where the rate @xmath94 is defined by @xmath95 for the subsequent analysis of the rate of the transition @xmath96 it is convenient to represent @xmath97 in terms of the energy level displacement of the ground state of kaonic hydrogen @xmath98 this is the model independent dgbtt ( deser , goldberger , baumann , thirring @xcite and trueman @xcite ) formula . substituting eq.([label2.19 ] ) into eq.([label2.18 ] ) we get @xmath99 using the numerical values of the parameters @xmath100 and @xmath101 we obtain @xmath102 where @xmath103 and @xmath104 are measured in @xmath105 .
the recent theoretical value of the energy level displacement of the ground state of kaonic hydrogen reads @xcite @xmath106 inserting eq.([label2.22 ] ) into eq.([label2.21 ] ) for the rate of the transition @xmath94 we get @xmath107 according to eq.([label2.17 ] ) , the rate of the transition @xmath108 is equal to @xmath109 these rates should be compared with the rates of the pure electric dipole transitions @xmath110 and @xmath111 at the neglect of strong interactions .
using the results obtained by bethe and salpeter @xcite and adjusting them to kaonic hydrogen we get @xmath112 and @xmath113 , respectively .
hence , the rates of the transitions @xmath114 and @xmath115 , given by eqs.([label2.23 ] ) and ( [ label2.24 ] ) and induced by strong low energy interactions , make up about @xmath116 and @xmath117 of the rates of the pure electric dipole transitions @xmath110 and @xmath111 , respectively . for the experimental values of the shift and width of the energy level of the ground state of kaonic hydrogen , measured by iwasaki _ et al . _ ( the kek collaboration ) @xcite : @xmath118 , the rates of the transitions @xmath114 and @xmath108 , given by eqs.([label2.23 ] ) and ( [ label2.24 ] ) , become increased by a factor of three .
the formula ( [ label2.20 ] ) can be easily extended to radiative transitions in kaonic deuterium @xmath119 , induced by strong low
energy interactions .
this reads @xmath120 where @xmath121 is the reduced mass of the @xmath50 pair and @xmath122 is the deuteron mass .
recently the energy level displacement of the ground state of kaonic deuterium has been estimated in @xcite : @xmath123 according to eq.([label3.2 ] ) , the rates of the radiative transitions @xmath124 and @xmath125 are equal to @xmath126 these rates should be compared with the rates @xmath110 and @xmath111 of the pure electric dipole transitions . using the results obtained by bethe and salpeter @xcite and adjusting them to kaonic deuterium we get @xmath127 and @xmath128 , respectively . thus , the rates of the radiative transitions @xmath129 and @xmath130 , induced by strong low energy interactions , make up about @xmath131 and @xmath132 of the rates of the pure electric dipole transitions @xmath133 and @xmath111 , respectively . for the value of the shift and width of the energy level of the ground state of kaonic deuterium , predicted by barrett and deloff @xcite :
@xmath134 , the rates of the transitions @xmath129 and @xmath130 , given by eq.([label3.3 ] ) , become increased by more than three times .
we have calculated the rates of the radiative transitions @xmath135 for kaonic hydrogen and kaonic deuterium , induced by strong low
energy interactions and enhanced by the attractive coulomb interaction of the @xmath15 and @xmath50 pairs in the kaon proton and kaon deuteron _ bremsstrahlung _ , @xmath34 and @xmath136 .
the neglect of the coulomb interaction of the @xmath15 and @xmath50 pairs in these reactions corresponds to @xmath137 in eq.([label2.16 ] ) . for the calculation of the amplitudes of the kaon proton and kaon deuteron _ bremsstrahlung
_ we have used an approach analogous to that of the eft based on chpt by gasser and leutwyler @xcite and used in @xcite for the derivation of the ericson
weise formula for the s wave scattering length of @xmath50 scattering @xcite .
it agrees also well with the potential model approach .
we have found that for the @xmath9 states of kaonic atoms the contributions of the rates of the transitions @xmath138 and @xmath124 , induced by strong low energy interactions , relative to the rates of the pure electric dipole transitions @xmath110 , are of order of one percent .
precisions of this order @xmath139 and @xmath140 have been reached in the experiments of the psi collaboration @xcite for the measurements of the energy level shift and width of the ground state of pionic hydrogen , respectively .
these precisions are defined by the accuracy of the measurements of the @xmath1ray spectra and yields in pionic hydrogen .
measurements of the energy level displacements of the ground states for kaonic atoms at the same level of precision , would demand to take into account the rates of the transitions @xmath141 and @xmath142 , induced by strong low energy interactions , for the theoretical description of the experimental data on the @xmath1ray spectra and yields in kaonic atoms .
we are grateful to torleif ericson for numerous and fruitful discussions .
9 t. b. day , g. a. snow , and j. sucher , phys .
lett . * 3 * , 61 ( 1959 ) ; r. k. adair , phys .
rev . lett . * 3 * , 438 ( 1959 ) ; m. leon and h. a. bethe , phys . rev . * 127 * , 636 ( 1962 ) .
t. e. o. ericson and f. scheck , nucl .
b * 19 * , 450 ( 1970 ) .
e. borie and m. leon , phys .
a * 21 * , 1460 ( 1980 ) .
t. koike , t. harada , and y. akaishi , phys .
c * 53 * , 79 ( 1996 ) .
t. p. terada and r. s. hayano , phys .
c * 55 * , 73 ( 1997 ) .
m. p. faifman _
et al . _ , frascati physics series vol .
637641 , _ physics and detectors for da@xmath143ne_frascati , nov.1619 , 1999 , _ physics of the atomic cascades in kaonic and hydrogen and deuterium _ ; m. p. faifman and l. i. menshikov , _ cascade processes in kaonic and muonic atoms _ , proceeding of international workshop on _ exotic atoms future perspectives _ at institute of medium energy physics of austrian academy of sciencies , november 2830 , 2002 , vienna , austia , pp.185196 . v. e. markushin and t. s. jensen , nucl . phys . a * 691 * , 318c ( 2001 ) ; t. s. jensen and v. e. markushin , nucl .
a * 689 * , 537 ( 2001 ) ; eur .
j. d * 19 * , 165 ( 2002 ) ; _ collisional de excitations of exotic hydrogen atoms in highly excited states.i . cross sections .
_ , physics/0205076 ; _ collisional de excitations of exotic hydrogen atoms in highly excited states.ii
. cascade calculations .
_ , physics/0205077 .
m. cargnelli _
( dear collaboration ) , kaonic nuclear clusters
miniworkshop , ( imep , wien ) , 9 february 2004 ; m. cargnelli _
( dear collaboration ) , in proceedings of hadatom03 workshop , 1317 october 2003 , ect@xmath144 ( trento italy ) , hep
a. n. ivanov , m. cargnelli , m. faber , h. fuhrmann , v. a. ivanova , j. marton , n. i. troitskaya , and j. zmeskal , _ energy level displacement of excited @xmath11 state of kaonic hydrogen _ , nucl th/0411026 .
a. n. ivanov , m. cargnelli , m. faber , j. marton , n. i. troitskaya , and j. zmeskal , eur .
j. a * 21 * , 11 ( 2004 ) ; nucl
h. a. bethe and e. e. salpeter , in _ quantum mechanics of one and two electron atoms _ , springer
verlag , berlin , 1957 .
a. n. ivanov , m. faber , a. hirtl , j. marton , and n. i. troitskaya , eur .
j. a * 19 * , 413 ( 2004 ) ; nucl th/0310027 .
a. n. ivanov , m. cargnelli , m. faber , h. fuhrmann , v. a. ivanova , j. marton , n. i. troitskaya , and j. zmeskal , eur .
j. a * 23 * , 79 ( 2005 ) , nucl th/0406053 .
t. e. o. ericson and w. weise , in _ pions and nuclei _ , clarendon press , oxford , 1988 .
j. gasser and h. leutwyler , phys .
b * 125 * , 321 , 325 ( 1983 ) ; j. gasser , nucl .
. suppl . * 86 * , 257 ( 2000 ) and references therein .
h. leutwyler , pin newslett .
* 15 * , 1 ( 1999 ) ; ulf - g .
meiner , pin newslett .
* 13 * , 7 ( 1997 ) ; h. leutwyler , ann . of phys .
* 235 * , 165 ( 1994 ) ; g. ecker , prog .
. phys . * 36 * , 71 ( 1996 ) ; prog .
. phys . * 35 * , 1 ( 1995 ) ; nucl .
* 16 * , 581 ( 1990 ) ; j. gasser , nucl .
b * 279 * , 65 ( 1987 ) ;
j. gasser , h. leutwyler , nucl .
b * 250 * , 465 ( 1985 ) ; ann . of phys .
* 158 * , 142 ( 1984 ) ; phys .
b * 125 * , 321 ( 1983 ) .
meiner , u. raha , and a. rusetsky , eur .
j. c * 35 * , 349 ( 2004 ) ; hep ph/0402261 .
t. e. o. ericson and l. hambro , ann.of phys .
( ny ) * 107 * , 44 ( 1977 ) .
j. meixner , math . z.
* 36 * , 677 ( 1933 ) ; l. hostler and r. h. pratt , phys .
* 10 * , 469 ( 1963 ) .
l. d. landau and e. m. lifshitz , in _ quantum mechanics , non relativistic theory _ ,
volume 3 of course of theoretical physics , pergamon press , oxford , 1965 .
, edited by m. abramowitz and i. a. stegun , national bureau of standards , applied mathematics series @xmath145 55 , 1972 .
s. wolfram , in _ mathematica , a system for doing mathematics by computer _
, addison wesley publishing co. , inc . , the advanced book program , new york , 1988 . s. deser , m. l. goldberger , k. baumann , and w. thirring , phys
. rev . * 96 * , 774 ( 1954 ) .
t. l. trueman , nucl .
* 26 * , 57 ( 1961 ) .
m. iwasaki _
et al . _ ( the kek collaboration ) , phys .
* 78 * , 3067 ( 1997 ) ; t. m. ito _
et al . _ ( the kek collaboration ) , phys .
c * 58 * , 2366 ( 1998 ) .
r. c. barrett and a. deloff , phys .
c * 60 * , 025201 ( 1999 ) .
d. got ta _
et al . _ , physica scripta , t * 104 * , 94 ( 2003 ) , hep - ex/0305012 . | we calculate the rates of the radiative transitions @xmath0 in kaonic hydrogen and kaonic deuterium , induced by strong low energy interactions and enhanced by coulomb interactions .
the obtained results should be taken into account for the theoretical analysis of the experimental data on the @xmath1ray spectra and yields in kaonic atoms .
+ pacs : 36.10.gv , 33.20.rm , 13.75.jz , 11.10.ef _ atominstitut der
sterreichischen universitten , arbeitsbereich kernphysik und nukleare astrophysik , technische universitt wien , + wiedner hauptstr .
8 - 10 , a-1040 wien , sterreich + and + stefan meyer institut fr subatomare physik der sterreichische akademie der wissenschaften , + boltzmanngasse 3 , a-1090 , wien , sterreich _ |
the search for improved materials for magnetic recording and permanent magnets has been the driving force for research work in the area of magnetic condensed matter physics.@xcite artificially exchange - biased systems , like ferromagnetic - antiferromagnetic ( fm - afm ) multilayers , have been widely studied , and they have been shown to provide large saturation magnetizations ( @xmath4 ) as well as high coercive fields ( @xmath5).@xcite however , the usually small exchange bias field ( @xmath6 ) is ultimately limited by the ability of the system to create an uncompensated magnetization at the antiferromagnet interface ( @xmath7).@xcite this interfacial magnetic moment induces a unidirectional anisotropy on the ferromagnet , and depends on a number of parameters including disorder , afm domain- and domain - wall - size , surface roughness , etc.@xcite a different kind of system , known as exchange - spring magnet was proposed by skomski and coey@xcite to provide a rather large exchange anisotropy energy . in two - phase nanocomposites ,
small soft - fm inclusions adopt the anisotropy energy of a hard - fm matrix .
such a case was observed in nd@xmath8fe@xmath9b/@xmath10-fe alloys.@xcite similarly , exchange anisotropy interactions have been studied in another large group of inhomogeneous materials presenting fm - afm interfaces , like fe / fe - oxide,@xcite co / coo,@xcite and ni / nio@xcite nanocomposites , spin glasses , etc.@xcite in the heavily studied manganites , inhomogeneous magnetic phases are known to be at the root of the colossal magnetoresistance effect.@xcite considering the presence of fm - afm interfaces , exchange anisotropy interactions can also be expected in these materials .
however , we are not aware of any study of interface exchange anisotropy in inhomogeneous manganites so far . in this paper
, we present a meticulous magnetic study of the pr@xmath0ca@xmath1mno@xmath2 manganite , and show clear signatures of interface exchange coupling of fm nanodomains immersed in the afm background . through an analysis of the shift of the magnetization loops at low temperatures
, we find that the surface exchange interaction is of afm nature , and its magnitude is similar to the bulk exchange interaction within the afm volume .
finally , we also observe a training effect of the exchange bias , which is successfully analyzed using an existing relaxation model .
the polycrystalline pr@xmath0ca@xmath1mno@xmath2 compound was prepared by the nitrate decomposition route , as described elsewhere,@xcite from high purity pr@xmath11o@xmath12 , caco@xmath2 , and mno .
the final sintering process was made at @xmath13 @xmath14c for 24 h , after which the sample was slowly cooled to room temperature .
magnetization ( @xmath15 ) measurements were performed in a superconducting quantum interference device ( squid ) magnetometer with applied magnetic fields @xmath16 up to @xmath17 koe in the temperature range @xmath18 k@xmath19 k. the temperature dependence of the magnetization was measured on warming with an applied field @xmath20 koe , in both field - cooling ( fc ) and zero - field - cooling ( zfc ) processes .
figure 1 presents data of @xmath21 as a function of temperature for the pr@xmath0ca@xmath1mno@xmath22 manganite .
the physical properties of this compound are mainly determined by the electrons on the mn ions .
there is a mn core , where three electrons are strongly coupled to a total spin @xmath23 .
the pr@xmath24/ca@xmath25 ratio introduces one doping electron for every three mn sites , responsible for the electrical transport .
the peak of @xmath21 at @xmath26 k indicates the charge ordering transition of the material,@xcite where the double exchange interaction is suppressed due to the localization of the charge carriers , producing a large drop of the susceptibility . at a lower temperature ,
the charge - ordered phase undergoes a paramagnetic - to - antiferromagnetic transition at @xmath27 k.@xcite the inset of fig .
1 shows the linear @xmath28 curve of the afm phase at @xmath29 k. however , even though the overall @xmath15-@xmath16 response is as expected , the sample exhibits some peculiar behavior when is cooled with an applied magnetic field . the curve in the inset of fig .
1 was measured after cooling the sample from @xmath30 k to @xmath18 k with an applied cooling field @xmath31 koe .
a more careful examination of these data reveals two notable features , namely a hysteresis ( coercivity ) and a shift of the @xmath28 curve from the origin , i.e. exchange bias .
it is worth mentioning that the shift of the loops has the same magnitude but opposite sign when the cooling field is negative , confirming the existence of the exchange bias phenomenon . for @xmath32
the exchange bias effect goes away , while the still - present coercivity is greatly reduced . ) of the pr@xmath0ca@xmath1mno@xmath2 compound as a function of temperature ( @xmath20 koe ) .
inset : @xmath15-@xmath16 loop at @xmath29 k , after field cooling the sample from @xmath30 k with @xmath31 koe.,height=238 ] the two features clearly indicate that , superposed on the predominant afm signal , there is a minor fm component showing the presence of small fm inclusions .
this fm component becomes evident in fig .
2 , where we present the resulting magnetization ( @xmath33 ) after the afm moment is subtracted .
this intrinsic coexistence of different magnetic phases is commonly found in many manganese perovskite compounds in the whole range of carrier doping , and is usually referred to as electronic phase separation.@xcite ) vs @xmath16 loop , after subtracting the contribution of the afm matrix ( @xmath29 k , @xmath31 koe ) .
inset : blow - up of the low - field region , showing the determination of the remanent moments @xmath34 and @xmath35.,height=234 ] as the magnetic field is reduced from the maximum value @xmath17 koe , the ferromagnetic moment becomes negative at @xmath36 koe , while on the increasing branch of the loop @xmath33 changes sign at @xmath37 koe .
these two fields define a coercive field @xmath38 koe and an exchange bias field @xmath39 koe . however , the values of @xmath40 and @xmath41 depend on the value of the susceptibility used to subtract the afm background , thus preventing a precise determination of @xmath6 .
alternatively , the exchange bias behavior can be seen as a vertical shift of the @xmath42 loop , as evidenced by the positive remanent magnetization obtained even after recoil from a very negative reverse field ( @xmath43 koe ) .
this effect is a manifestation of the presence of a unidirectional exchange anisotropy interaction , which drives the fm domains back to the original orientation when the magnetic field is removed .
therefore , since for @xmath44 there is no contribution from the afm volume , we choose to quantify the exchange bias effect using the remanent magnetization .
for this reason , we define the magnetization shift @xmath45 and the magnetic coercivity @xmath46 as the vertical axis " equivalents of @xmath6 and @xmath5 , respectively . naming as @xmath34 and @xmath35 the remanent magnetizations on the field decreasing and increasing branches of the loop , respectively ( as in the inset of fig .
2 ) , we define @xmath47 and @xmath48 ( for the data shown in fig .
2 , @xmath49 gauss and @xmath50 gauss ) . with these definitions ,
@xmath46 and @xmath45 are quantitative indications of coercivity and exchange bias , respectively . indeed , in the simplest model of single - domain fm particles ( that we assume in this work )
, a relationship between @xmath45 and @xmath6 can be easily derived .
considering the magnetic moments switching by thermal activation over the anisotropy barrier @xmath51 ( @xmath52 is the anisotropy constant and @xmath53 the volume of the particles ) , the exchange bias field introduces an asymmetry in the activation energy for the backward and forward switching such that @xmath54 where @xmath4 is the saturation magnetization , @xmath55 sec@xmath56 is the switching attempt frequency , @xmath57 sec is the typical measurement time , @xmath58 is the boltzmann constant , and @xmath59 is the magnetic moment of the fm particles . for a small enough magnetic energy ( @xmath60 ) ,
( [ mevshe ] ) can be rewritten as @xmath61 indicating a direct equivalence between @xmath45 and @xmath6 .
our analysis in the next subsection shows that @xmath62 is always lower than 0.9 , justifying _ a posteriori _ the use of this relationship .
the fact that reverse fields as high as @xmath43 koe are not able to produce a negative @xmath35 , indicates that the exchange interactions giving rise to the described effects must be of considerable strength .
the origin of such exchange anisotropy is revealed by the temperature dependence of the magnetization shift , as presented in fig .
3 . in this case
, the sample was cooled down from @xmath30 k to @xmath18 k with an applied field @xmath63 koe , and warmed back to the measuring temperature without removing the applied field . at the appropriate temperature , the magnetization loop was measured between @xmath64 and @xmath65 koe , from which @xmath45 and @xmath46 were calculated .
it can be seen that , as the temperature increases , the magnetization shift vanishes right at the nel temperature of the afm background , @xmath66 k , while the coercivity remains present up to the temperature marked as @xmath67 , which is @xmath68 k higher than @xmath69 and indicates the onset of the fm component . and
@xmath46 , showing the onset of exchange bias and coercivity at @xmath69 and @xmath67 , respectively .
inset : @xmath45 and the ratio @xmath70 at low temperatures . both quantities
sharply increase below the freezing temperature , @xmath71 k.,height=234 ] the temperature evolution shown in fig .
3 is typical of exchange bias systems.@xcite as the sample is cooled through @xmath69 with an applied magnetic field , the exchange interaction between the afm phase and the polarized fm phase induces an interface magnetic moment on the antiferromagnet ( @xmath7 ) , which remains frozen at lower temperatures.@xcite in turn , below the freezing temperature @xmath72 ( @xmath73 ) , @xmath7 provides the necessary pinning forces ( exchange anisotropy ) to drive the ferromagnet back to a positive remanence on recoil .
then , the exchange bias behavior , observed through the appearance of a distinctive magnetization shift , is induced by the interface exchange coupling between the fm domains and the charge - ordered afm background .
although the magnetization shift first develops at @xmath74 k , it is useful to note that the main freezing temperature of the system is @xmath71 k. the inset of fig .
3 shows the temperature dependence of the ratio @xmath70 , the magnetization of the material measured in fc and zfc processes , respectively .
the increase of this ratio below @xmath72 is due to the enhancement of the magnetic irreversibility induced by the freezing of the surface spins .
simultaneous with the upward turn of @xmath70 , @xmath45 also exhibits a steep increase of more than one order of magnitude below @xmath75 k. this indicates that a small number of interfaces start to freeze at @xmath69 , but the dominant contribution to the low temperature @xmath45 comes from domains with @xmath71 k. the existence of low-@xmath72 spins has been also observed in other systems,@xcite and have been attributed to glass - like phases occurring in spin - disordered surfaces around the ordered particles.@xcite indeed , the irreversibility shown by the ratio @xmath70 is a fingerprint of the glassy behavior in nanoparticles systems . in this context
, a partial alignment of the glassy spins leads to the interface moment @xmath7 , and the temperature @xmath72 is a measure of the energy barriers separating multiple spin configurations of the glass - like phase . in order to get some insight into the properties of this exchange biased system , we studied the cooling field dependence of @xmath45 . for each data point shown in fig .
4 , the sample was cooled from @xmath30 k to @xmath18 k with the applied cooling field , and a hysteresis loop was measured between @xmath17 and @xmath43 koe . as shown in this figure
, @xmath45 presents a smooth increasing dependence with @xmath76 , which saturates at high cooling fields .
the exchange bias field is usually thought as the balance between the zeeman energy of the fm particles and the surface energy due to the interface exchange interaction,@xcite i.e. , @xmath77 @xmath78 and @xmath79 are the number of spins inside the fm volume and on the disordered afm shell , respectively , thus @xmath6 depends , as expected , on the surface - to - volume ratio .
the surface exchange constant is represented by @xmath80 , @xmath81 is the gyromagnetic factor , and @xmath82 is the bohr magneton . from eq .
( [ hevsmi ] ) , the usual meiklejohn - bean expression@xcite for fm - afm bilayers is easily recovered by making the appropriate replacements , e.g. @xmath83 , where @xmath84 is the out - of - plane lattice parameter and @xmath85 the thickness of the fm layer . .
the solid line is a fit with an exchange interaction model between the fm domains and a disordered afm shell ( see text).,height=241 ] for small cooling fields , the coupling of the interface moment to the fm volume dominates over the coupling of @xmath7 to @xmath76 , implying that the surface spin configuration minimizes the surface energy and @xmath86 , producing a negative @xmath6 .
however , it has been shown that for afm couplings ( @xmath87 ) , a large enough cooling field could prevail and flip the interface moment , producing a positive @xmath7 at the freezing temperature and changing the sign of the exchange bias field.@xcite assuming 1-to-1 core - shell spin interactions , these competing energies can be described in an effective magnetic field acting on the interface moment during the freezing process as @xmath88 with @xmath89 the magnetic moment per fm spin at the freezing temperature .
for our small magnetic domains , we assume @xmath90 , where @xmath91 @xmath82 is the magnetic moment of the mn core spin , @xmath92 is the langevin function , @xmath93 , and @xmath94 . at lower temperatures , because of the glassy behavior of the surface spins the interface moment @xmath7 ( @xmath95 ) remains frozen , inducing the exchange bias response .
therefore , combining eqs .
( [ mevshelinear ] ) , ( [ hevsmi ] ) , and ( [ hef ] ) we obtain @xmath96 \label{hcooldependence}\ ] ] in this equation , the competition between the exchange interaction and the cooling field becomes evident . for small
@xmath76 the first term usually dominates , and @xmath6 ( @xmath97 ) depends on @xmath98 .
however , for large cooling fields the second term ( @xmath99 ) becomes important , and for @xmath87 the absolute value of @xmath6 could decrease@xcite or even more , it could change sign , as observed in previous works.@xcite in fig .
4 , the solid line corresponds to a fit to eq .
( [ hcooldependence ] ) with @xmath80 and @xmath78 adjustable parameters , which indicates a quite good agreement of the experimental data with the simple model described above . moreover , the agreement is not only qualitative but also quantitative .
the exchange constant obtained from the fit was @xmath100 mev .
this coupling constant indicates an afm coupling between the fm domains and the afm host , explaining the tendency of @xmath101 towards a reduction at high @xmath76 .
on top of this , a mean field estimation of the exchange interaction for an antiferromagnet with @xmath102 k gives @xmath103 mev , in excellent agreement with the value obtained .
this indicates that the exchange interaction within the afm volume extends across the boundary with the fm domains without noticeable changes .
further , the number of spins per fm droplet is @xmath104 ( @xmath105 @xmath82 ) , corresponding to a droplet diameter @xmath106 .
this kind of fm nanodomain was already observed in the la@xmath107ca@xmath108mno@xmath22 manganite@xcite and in la@xmath109sr@xmath110mno@xmath22,@xcite where small angle neutron scattering experiments show the presence of fm droplets with the same characteristic size , and a density of domains @xmath111 @xmath112 . in our pr@xmath0ca@xmath1mno@xmath22 sample ,
an estimation of the density of fm droplets can be made from the saturation magnetization , which can be written as @xmath113 with @xmath114 gauss obtained from fig .
2 , we estimate @xmath115 @xmath112 , which is very similar to the mentioned neutron scattering results .
this density of droplets implies that the fm phase occupies only around @xmath116 of the volume of the material , explaining the almost imperceptible fm signal that superposes on the afm magnetization . at this point
it is worth noting that , although in previous works the small interface moment @xmath7 has been detected by different techniques,@xcite in our case this moment is negligible as compared to @xmath45 . an estimation from eq .
( [ hevsmi ] ) indicates that the maximum contribution of the interface is @xmath117 @xmath82 per surface spin .
this represents only a @xmath118 of the maximum possible value ( @xmath119 @xmath82 ) , in coincidence with measurements in other different systems.@xcite this @xmath7 is indeed much smaller than the shift of the loops , which is @xmath120 @xmath82 per fm spin ( @xmath121 ) at the same cooling field .
moreover , while @xmath7 is negative the measured @xmath45 is always positive .
therefore , the magnetization shift measured in our system corresponds almost in its entirety to the remanence of the fm domains , as assumed in our analysis .
the exchange bias phenomenon observed in our material would be similar to that taking place in artificially fabricated nanocomposites,@xcite where fm nanoparticles are surrounded by the afm oxide .
this kind of core - shell interaction has been shown to provide quite strong magnetic anisotropies . in our case , using @xmath121 and @xmath122 koe ( for @xmath31 koe ) , from eq .
( [ mevshe ] ) we can estimate an anisotropy constant @xmath123 erg/@xmath124 .
this is a quite important magnetic anisotropy , two orders of magnitude larger than the usual magnetocrystalline anisotropy measured in bulk fm manganites.@xcite it is clear that this notable anisotropy constant is related to the particular magnetic configuration of the pr@xmath0ca@xmath1mno@xmath22 compound . in spring - magnet materials , it has been shown that the exchange anisotropy increases when the soft - fm domain size decreases , acquiring the anisotropy energy of the hard - fm matrix.@xcite similarly , the exchange bias phenomenon is related to the surface exchange interaction between the fm and afm phases .
therefore , it is widely accepted that , in nanocomposite materials , @xmath52 is proportional to the surface - to - volume ratio of the fm domains . in the case of our nanometric fm domains , where @xmath125 , the ratio @xmath126 is close to 1 , being compatible with the high anisotropy constant estimated .
elastic effects in the charge - ordered phase tend to produce not spherical but elongated secondary phases,@xcite therefore a shape anisotropy energy could be also contributing to the anisotropy constant . from the structural point of view , the strength of the exchange anisotropy is influenced by the existence of surface strain and atomic displacement .
this disorder modifies the interface exchange constant,@xcite and could be particularly relevant in nanocomposites where the constituents are structurally and chemically different .
however , our manganite is an _ intrinsic nanocomposite _
, where the fm droplets appear naturally embedded in the afm background and where structural and chemical differences should be less important . as a consequence ,
the obtained interface exchange constant is as large as in the bulk of the material , favoring a high exchange anisotropy as well .
koe.,height=249 ] in exchange bias systems , a gradual decrease of the anisotropy interaction is commonly found as the material is continuously field - cycled ; the so - called training effect.@xcite as a result , the exchange bias and the coercive fields decrease with increasing loop index number . in our manganite , the training effect is also present , as shown in fig .
5 . this figure presents the consecutive @xmath33-@xmath16 loops measured for fields between @xmath17 and @xmath43 koe , after the sample was field - cooled with @xmath31 koe ( only loops with index number @xmath127 , @xmath128 , @xmath129 , and @xmath130 are shown ) .
the relaxation of the magnetization shift is evident and , as shown by the evolution of @xmath45 with @xmath131 ( fig .
6 ) , this relaxation is particularly important between the first and second loops , where @xmath45 falls by @xmath132 . )
vs loop index number ( @xmath131 ) .
the solid triangles correspond to a fit with a recursive law , that results from a discretized relaxation model .
inset : @xmath45 vs @xmath133 , showing the linear behavior usually obtained .
the line is a least squares fit.,height=245 ] the training effect has been explained in terms of the demagnetization of the afm surface moment @xmath7.@xcite as the fm domain switches back and forth under the influence of the applied field , a relaxation of the surface spin configuration towards the equilibrium is induced due to the _ surface drag _ of the exchange interaction .
this is particularly relevant for glass - like phases , where multiple spin configurations are available .
indeed , it has been shown that magnetic frustration and the existence of multiple easy axes orientations on the afm counterpart of the fm / afm interface are essential for the occurrence of training effects.@xcite the usual experimentally observed relationship between @xmath45 and @xmath131 is given by @xmath134 where @xmath135 is the equilibrium magnetization shift .
as shown in the inset of fig .
6 , this equation holds for @xmath136 , with @xmath137 gauss .
however , as already stated in previous works@xcite this equation holds only for loop indices @xmath136 , and can not explain the steep relaxation between the first and second loops . in recent work binek,@xcite using a discretized relaxation model
deduced a recursive formula that relates the @xmath138-th loop shift with the @xmath131-th one as @xmath139 with @xmath140 a sample dependent constant determined by microscopic parameters like the spin damping constant . with this simple relationship , and using @xmath141 gauss@xmath142 and @xmath143 gauss , the whole set of data
is reproduced ( solid triangles in fig .
6 ) , including the large initial drop of @xmath45 . from eq .
( [ eq5 ] ) , the recursive law is recovered for @xmath144 , therefore explaining its applicability in this regime .
in summary , we provide the first evidence of intrinsic interface exchange coupling in phase separated manganites .
our magnetization measurements show the coexistence of a predominant charge - ordered afm phase and a minor fm component in the pr@xmath0ca@xmath1mno@xmath2 compound . furthermore ,
after cooling the sample through the nel temperature with an applied magnetic field this manganite presents exchange bias .
this is clearly revealed by a distinctive magnetization shift ( @xmath45 ) in the remanence of the @xmath15-@xmath16 loops .
the cooling field dependence of @xmath45 at low temperature was successfully analyzed in terms of a simple exchange interaction model between fm nanodroplets of @xmath3 @xmath145 in size and a disordered ( glassy - like ) afm shell .
the observed training effect of the exchange bias can be described with the same relaxation model used for other classical exchange bias systems , indicating a similar mechanism for exchange anisotropy . moreover ,
the anisotropy constant deduced , @xmath123 erg/@xmath124 , is two orders of magnitude larger than the magnetocrystalline anisotropy of other common fm manganites .
this large anisotropy energy opens a new possibility for the development of hard fm materials , where important exchange anisotropy interactions are introduced by intrinsic inhomogeneities that naturally occur in phase separated magnetic materials .
antel , jr . , f. perjeru , and g.r .
harp , phys .
lett . * 83 * , 1439 ( 1999 ) ; t.p.a .
hase , b.d .
fulthorpe , s.b .
wilkins , b.k .
tanner , c.h .
marrows , and b.j .
hickey , appl .
. lett . * 79 * , 985 ( 2001 ) .
j. lffler , h. van swygenhoven , w. wagner , j. meier , b. doudin , and j .- ph . ansermet , nanostruct . mater .
* 9 * , 523 ( 1997 ) ; y.d .
yao , y.y .
chen , m.f .
tai , d.h .
wang , and h.m .
lin , mater .
a * 217 * -*218 * , 281 ( 1996 ) .
f. nolting , a. scholl , j. sthr , j.w .
seo , j. fompeyrine , h. siegwart , j .-
locquet , s. anders , j. lning , e.e .
fullerton , m.f .
toney , m.r .
scheinfein , and h.a .
padmore , nature ( london ) * 405 * , 767 ( 2000 ) ; h. ohldag , a. scholl , f. nolting , s. anders , f.u .
hillebrecht , and j. sthr , phys .
lett . * 86 * , 2878 ( 2001 ) .
e. granado , c.d .
ling , j.j .
neumeier , j.w .
lynn , and d.n .
argyriou , phys . rev .
b * 68 * , 134440 ( 2003 ) ; m. hennion , f. moussa , g. biotteau , j. rodrguez - carvajal , l. pinsard , and a. revcolevschi , phys . rev . lett . *
81 * , 1957 ( 1998 ) ; m. hennion , f. moussa , j. rodrguez - carvajal , l. pinsard , and a. revcolevschi , phys . rev .
b * 56 * , r497 ( 1997 ) . s.e .
lofland , s.m .
bhagat , c. kwon , s.d .
tyagi , y.m .
mukovskii , s.g .
karabashev , and a.m. balbashov , j. appl .
phys . * 81 * , 5737 ( 1997 ) ; f. rivadulla , l.e .
hueso , c. jardn , c. vzquez - vzquez , m.a .
lpez - quintela , j. rivas , m.t .
causa , c.a .
ramos , and r.d .
snchez , j. magn .
* 196 * -*197 * , 470 ( 1999 ) . | we present a detailed magnetic study of the pr@xmath0ca@xmath1mno@xmath2 manganite , where we observe the presence of small ferromagnetic ( fm ) domains ( diameter @xmath3 ) immersed within the charge - ordered antiferromagnetic ( afm ) host .
due to the interaction of the fm nanodroplets with a disordered afm shell , the low - temperature magnetization loops present exchange bias ( eb ) under cooling in an applied magnetic field .
our analysis of the cooling field dependence of the eb yields an antiferromagnetic interface exchange coupling comparable to the bulk exchange constant of the afm phase .
we also observe training effect of the eb , which is successfully described in terms of a preexisting relaxation model developed for other classical eb systems .
this work provides the first evidence of intrinsic interface exchange coupling in phase separated manganites . |
quantum computation is a subject that is strongly attracting the interest of the physics community in recent times @xcite , mainly because of the vast potential it has for extremely high - performance calculations .
the main reason for that derives from the fact that instead of employing a binary system as the basic computing unit , it uses the infinitely many quantum states obtained by coherent linear combinations of certain base states .
loss of quantum coherence through interaction with the environment , however , is a fatal threat for a quantum computer , since it implies complete loss of information . for this reason it is important to find systems , capable of performing the required operations of quantum computation , being at the same time protected against the process of decoherence .
systems presenting non - abelian statistics can provide the required stability through the mechanism known as topological quantum computation @xcite .
for these systems , exchanging particles in a many - particle state produces the entanglement thereof , thus making them robust against decoherence . an intense search for many - particle systems presenting such peculiar behavior under the particle exchange or braiding operation
has then started . in this work ,
we consider the non - abelian chern - simons field in 2 + 1 d , minimally coupled to a higgs field in the adjoint representation of the su(2 ) group ( the csh - theory ) .
the higgs field potential is such that there are two phases according to whether the vacuum expectation value of this field vanishes or not .
we study in detail the quantum magnetic vortex excitations of this system in the ordered phase . in order to accomplish the full quantization of such excitations
, we apply to the csh - theory the method of quantization of topological excitations , which is based on the concept of order - disorder duality @xcite .
we obtain , in particular , the explicit form of the creation operator of quantum excitations carrying both magnetic flux and charge , as well as their euclidean correlation functions .
these electrically charged magnetic vortices may be boson , fermion or , more generally anyons .
we show that special self - adjoint combinations of vortices and anti - vortices possess non - abelian statistics whenever the vortices are anyonic .
furthermore , for a specific value of the anyon spin , namely @xmath0 , we show that we can construct the not and cnot logic gates , required in quantum computation , from the corresponding monodromy matrices .
our results , based on a fully quantized approach of topological excitations , therefore show that the csh model with an su(2 ) group is an excellent example of a system exhibiting the requisites needed for the operation of a quantum computer .
related results have been reported in the literature .
for instance , non - abelian statistics has been obtained for certain quasi - particle excitations of the quantum hall liquid @xcite and also for ising anyons @xcite .
models inspired in non - abelian anyons have been proposed in @xcite .
let us consider the su(2 ) non - abelian chern - simons theory to which we couple a higgs field in the adjoint representation : @xmath1 = \frac{\kappa}{4\pi}\int d^{3}z\,\epsilon^{\mu\nu\rho}\left(a^{a}_{\mu } \partial_{\nu}a^{a}_{\rho } + \frac{2}{3}\,\epsilon^{abc}a^{a}_{\mu } a^{b}_{\nu}a^{c}_{\rho}\right ) + \mbox{tr}d_{\mu}\phi d^{\mu}\phi - v(\vert \phi\vert , \eta)\end{aligned}\ ] ] where the higgs self - interaction potential is given by @xcite @xmath2 the euler - lagrange equation is @xmath3 where @xmath4^{a}$ ] .
the theory presents ordered or disordered phases , according to whether @xmath5 or @xmath6 , where , respectively , we have @xmath7 and @xmath8 . we are now going to obtain the operators @xmath9 and @xmath10 , which create states carrying , respectively , charge and magnetic flux .
these two quantities are given respectively by @xmath11 where @xmath12 is an unit vector , subject to the action of the group ( for instance @xmath13 where @xmath14 ) and @xmath15 in order to construct the local operators @xmath9 and @xmath10 , we will follow the method for quantization of topological excitations that was developed with basis on the order - disorder duality @xcite . according to this ,
the @xmath9 and @xmath10 operators act , respectively , as order and disorder operators and therefore satisfy the corresponding dual algebra .
then , correlation functions determined by this algebra are obtained by coupling certain special external fields to the dynamical fields of the system . in the case of the @xmath10-operator these external fields
are given by @xmath16 , whereas , for the @xmath9-operator , by @xmath17 where @xmath18 and @xmath19 in the above expressions @xmath20 is the covariantized vector surface integration element , perpendicular to the integration surface @xmath21 .
this consists of the complex plane , excluding the singularities at @xmath22 and along the cut of the function @xmath23 .
it turns out that the mixed multicorrelation function is given by the vacuum functional in the presence of these external fields : @xmath24 \partial_{\nu } [ \nonumber\\ & & a^{d}_{\rho } + \bar{a}^{d}_{\rho}(x^{b}_{1 } , \ldots , y^{b}_{m } ) + \bar{c}^{d}_{\rho}(x^{a}_{1 } , \ldots , y^{a}_{m } ) ] + \frac{2}{3}\,\epsilon^{def}[a^{d}_{\mu } + \bar{a}^{d}_{\mu}(x^{b}_{1 } , \ldots , y^{b}_{m } ) + \bar{c}^{d}_{\mu}(x^{a}_{1 } , \ldots , y^{a}_{m } ) ] \nonumber\\ & & a^{e}_{\nu } + \bar{a}^{e}_{\nu}(x^{b}_{1 } , \ldots , y^{b}_{m } ) + \bar{c}^{e}_{\nu}(x^{b}_{1 } , \ldots , y^{b}_{m } ) ] [ a^{f}_{\rho } + \bar{a}^{f}_{\rho}(x^{b}_{1 } , \ldots , y^{b}_{m } ) + \bar{c}^{f}_{\rho}(x^{a}_{1 } , \ldots , y^{a}_{m})]\big ] + \mbox{tr}d_{\mu}\phi d^{\mu}\phi \nonumber\\ & & - ( 4\lambda)^{2}\mbox{tr}\phi^{2}(\eta^{2 } + \phi^{2})^{2 } + { \cal l}_{gf}[a ] + { \cal l}_{gh}[a]\bigg]\bigg\ } , \nonumber\end{aligned}\ ] ] where @xmath25 and @xmath26 are respectively the gauge - fixing and ghost lagrangians . we may obtain an equivalent expression for the @xmath27-correlation function , by shifting the @xmath28 functional integration variable in the above equation as @xmath29 this produces the equivalent expression @xmath30 + { \cal l}_{gh}[a \rightarrow a^{\mu } - \bar{a}^{\mu } - \bar{c}^{\mu}]\bigg\},\end{aligned}\ ] ] where @xmath31.\end{aligned}\ ] ] from this form of the correlation function we can extract the operators carrying , respectively , charge and magnetic flux for the non - abelian chern - simons - higgs theory .
this is easily done , because , according to ( [ fff ] ) , a 2-point correlator of the @xmath10-operator , for instance , is expressed as a functional integral having an integrand of the form : @xmath32 \right\}\end{aligned}\ ] ] where @xmath33 is given in terms of the dynamical fields and @xmath34 is given by ( [ campoexterno1 ] ) .
the last term clearly does not involve dynamical fields , being therefore a kind of renormalization factor .
the first term is the measure weight , which is used for computing averages , namely @xmath35 \right\ } \mu(x ) \mu^\dagger ( y)\end{aligned}\ ] ] it follows that , appart from the ( c - number ) renormalization factor , the @xmath36 operators must correspond to the second term in ( [ mu - op ] ) , which has an exponent linear in @xmath37 .
the first one will give @xmath38 , whereas the second , @xmath39 .
the @xmath9-operators , conversely , must be exponentials of those terms that are linear in the external fields @xmath40 , given by ( [ campoexterno2 ] ) .
following the previously described inspection procedure , we obtain @xmath41 and @xmath42 with @xmath43 given by ( [ el ] ) . in the above equations , @xmath44 is the covariantized vector line integration element along the integration line @xmath45 .
this is a line going from @xmath46 to @xmath47 along the cut of the @xmath23 function .
we investigate now the commutation rules of @xmath9 and @xmath10 operators obtained above with the charge and magnetic flux operators .
let us evaluate firstly the equal time commutator of @xmath10 with the magnetic flux operator . using the field equation
, we may cast the @xmath10-operator in the form @xcite @xmath48 from this , we get @xmath49 & = & \frac{1}{2 } \mu(x_{i})\kappa a\,n^{b}n^{c}\int d^{2}y\int_{x_{i } , l}^{+\infty } d\xi^k [ a^{b}_{k}(\xi ) , \epsilon^{jl}f_{jl}^{c } ] \nonumber\\ & = & - 2\pi a \mu(x_{i})\,n^{b}n^{c}\int d^{2}y\int_{x_{i } , l}^{+\infty } d\xi^k \partial_k^{(\xi ) } \delta^{bc } \delta(\xi - y ) \nonumber\\ & & + \,\mu(x_{i})\kappa a\,n^{b}n^{c}\int d^{2}y\int_{x_{i } , l}^{+\infty } d\xi^k \epsilon^{jl } \epsilon^{ced}[a^{b}_{k}(\xi ) , a^{e}_{j}(y)a^{d}_{l}(y ) ] \nonumber\\ & = & 2\pi a\mu(x_{i } ) + \mu(x_{i } ) \kappa a n^{b}n^{c}\int d^{2}y\int_{x_{i } , l}^{+\infty}d\xi^k \epsilon^{jl } \epsilon^{ced}\left(a^{e}_{j}(y)[a^{b}_{k}(\xi ) , a^{d}_{l}(y ) ] + [ a^{b}_{k}(\xi ) , a^{e}_{j}(y)]a^{d}_{l}(y)\right ) \nonumber\\ & = & 2\pi a\mu(x_{i } ) \nonumber\\ & & + \mu(x_{i } ) \kappa a n^{b}n^{c } \int d^{2}y\int_{x_{i } , l}^{+\infty } d\xi^k \epsilon^{jl } \epsilon^{ced}\big(a^{e}_{j}(y)\frac{2\pi}{\kappa}\epsilon_{kl}\delta^{bd}\delta^{2}(\xi - y ) + \frac{2\pi}{\kappa}\epsilon_{kj}\delta^{be}\delta^{2}(\xi - y)a^{d}_{l}(y)\big ) \nonumber\\ & = & 2\pi a\mu(x_{i})\end{aligned}\ ] ] where we used the equal - time commutator @xmath50 = 2\pi/\kappa\ , \epsilon_{ij}\delta^{ab}\delta^{2}(\vec{x } - \vec{y})$ ] and the fact that @xmath51 .
the second term in the rhs above vanishes because it is proportional to @xmath52 .
this result shows that @xmath53 creates states bearing a magnetic flux @xmath54 , being therefore , a magnetic vortex creation operator .
notice that @xmath55 is the quantum of magnetic flux for @xmath56 , hence the free parameter @xmath57 determines the number of flux units created by @xmath10 . a natural choice , therefore would be @xmath58 . in order to evaluate the commutator of @xmath9 with the matter charge operator @xmath59
, we must consider the current - current commutator .
this is given in general by the current algebra relation @xcite @xmath60 = { \cal m}\delta^{ab}\partial^{i}\delta^{(2 ) } ( \vec{x } - \vec{y})\end{aligned}\ ] ] where @xmath61 is a functional of the spectral density of the theory .
using this , we find @xmath62 = b { \cal m}\sigma(\vec{x } , t ) \delta^{(2)}(\vec{x } - \vec{y})\end{aligned}\ ] ] or @xmath63 = b { \cal m}\sigma(\vec{x } , t),\end{aligned}\ ] ] indicating that @xmath9 bears a charge @xmath64 from this , we can see that the choice @xmath65 would imply that the operator @xmath9 carried one unit of electric charge .
we see that the @xmath10 and @xmath9 operators obtained above carry , respectively , magnetic flux and charge .
we therefore expect their product will be , in general anyon fields @xcite .
it is precisely out of combinations of these that we will construct the fields with non - abelian statistics . in the symmetric phase , @xmath66 in ( [ csh ] )
, we have to add to ( [ csh ] ) a gauge - fixing term @xmath25 along with the corresponding ghost term @xmath26 . for the gauge - fixing term
, we are going to choose a lorentz - type gauge .
then , we add to ( [ csh ] ) , in the symmetric phase , the terms @xmath67[d_{\mu}^{\tiny{adj}}\eta]^{a}\end{aligned}\ ] ] where @xmath68 are ghosts fields and @xmath69 is the gauge parameter . in the broken phase , @xmath70 in ( [ csh ] ) , the potential has a minimum at @xmath71 , with @xmath72 . taking the vacuum pointing along the third direction , that is , @xmath73
, we can see that the fields will be given by ( @xmath74 ) , with @xmath75 .
the fields @xmath76 , @xmath77 , and @xmath78 have a zero vacuum expection value .
then , in the broken phase we choose an t hooft gauge , where the quadratic mixed terms involving ( @xmath79 ) in the expression @xmath80 disappear .
to be more general , however , this unwanted term disappears if we add a gauge - fixing term of the form @xmath81^{2}.\end{aligned}\ ] ] where @xmath82 is the vacuum expectation value of the higgs field . from this gauge - fixing
we have @xmath83[d_{\mu}^{\tiny{adj}}\eta]^{a } - \bar{\eta}^{a}\left[\frac{2m}{\xi}\phi_{3 } + \frac{4m^{2}}{\xi}\right ] ( \delta^{ab } - \delta^{a3}\delta^{b3})\eta^{b}.\end{aligned}\ ] ] from eqs .
( [ csh ] ) and ( [ gauge - symmetric ] ) , we have the langrangian density in the symmetric phase , @xmath84 , while in the broken phase , the lagrangian density is @xmath85 . from the quadratic terms in @xmath86 and @xmath87 we obtain the propagators for the fields . in euclidean space
these are @xmath88 , \nonumber\\ d^{\mu\nu}_{(3)}(x ) & = & \int \frac{d^{3}k}{(2\pi)^{3 } } e^{ik\cdot x}\left[\frac{1}{2\alpha}\epsilon^{\mu\lambda\nu}\frac{k_{\lambda}}{k^{2 } } + \frac{1}{\xi}\frac{k^{\mu}k^{\nu}}{k^{4}}\right ] , \nonumber\\ \delta_{gh}^{(i)}(x ) & = & \int\frac{d^{3}k}{(2\pi)^{3}}\frac{e^{ik\cdot x}}{k^{2 } + m_{gh_{(i)}}^{2}},\quad i = 1 , 2 , 3 , \label{propagators}\end{aligned}\ ] ] where @xmath89 and @xmath90 are the propagators for the higgs - field components , @xmath76 , @xmath77 , and @xmath91 ( @xmath78 in the broken phase ) , @xmath92 are the propagators for the gauge fields @xmath93 and @xmath94 are the propagators for ghosts - field components . in the symmetric phase we have @xmath95 and @xmath96 and @xmath97 ( @xmath98 ) . in the broken phase we have @xmath99 , @xmath100 , and @xmath101 and @xmath102 .
let us write the exponent in ( [ fff ] ) as @xmath104 $ ] , where @xmath105 contains all the dependence on the external field @xmath106 and @xmath107 . in the symmetric phase , from ( [ fff ] ) and ( [ gauge - symmetric ] )
, we obtain that @xmath108^{2 } \nonumber\\ & & - 2\partial_{\mu}(\bar{a}_{\mu}^{a } + \bar{c}_{\mu}^{a})\partial_{\nu}(\bar{a}_{\nu}^{a } + \bar{c}_{\nu}^{a } ) ] -\bar{\eta}^{a}[\epsilon^{abc}\partial^{\mu}(\bar{a}_{\mu}^{c } + \bar{c}_{\mu}^{c } ) + \epsilon^{abc}(\bar{a}_{\mu}^{c } + \bar{c}_{\mu}^{c})\partial^{\mu}]\eta^{b}\end{aligned}\ ] ] while in the broken phase , we obtain @xmath109 \nonumber\\ & & - [ ( \bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1})(\phi_{2}\partial_{\mu}\chi - \chi\partial_{\mu}\phi_{2 } ) + ( \bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2})(\chi\partial_{\mu}\phi^{1 } - \phi^{1}\partial_{\mu}\chi ) \nonumber\\ & & + ( \bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3})(\phi^{1}\partial_{\mu}\phi^{2 } - \phi^{2}\partial_{\mu}\phi^{1 } ) ] + [ ( \bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1})^{2 } - 2a^{\mu}_{1}(\bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1 } ) ] ( \phi_{2}^{2 } \nonumber\\ & & + \chi^{2 } + 2\phi_{0}\chi ) + [ ( \bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2})^{2 } - 2a^{\mu}_{2}(\bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2})](\phi_{1}^{2 } + \chi^{2 } + 2\phi_{0}\chi ) \nonumber\\ & & + [ ( \bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3})^{2 } - 2a^{\mu}_{3}(\bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3 } ) ] ( \phi^{2}_{1 } + \phi^{2}_{2 } ) - 2[(\bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1})(\bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2 } ) \nonumber\\ & & a^{\mu}_{1}(\bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2 } ) - a^{\mu}_{2}(\bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1 } ) ] \phi_{1}\phi_{2 } - 2[(\bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1})(\bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3 } ) \nonumber\\ & & a^{\mu}_{1}(\bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3 } ) - a^{\mu}_{3}(\bar{a}^{\mu}_{1 } + \bar{c}^{\mu}_{1 } ) ] ( \phi_{1}\chi + \phi_{0}\phi_{1 } ) - 2[(\bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2})(\bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3 } ) \nonumber\\ & & a^{\mu}_{2}(\bar{a}^{\mu}_{3 } + \bar{c}^{\mu}_{3 } ) - a^{\mu}_{3}(\bar{a}^{\mu}_{2 } + \bar{c}^{\mu}_{2 } ) ] ( \phi_{2}\chi + \phi_{0}\phi_{2 } ) ] - \frac{\xi}{2 } [ [ \partial_{\mu}(\bar{a}_{\mu}^{a } + \bar{c}_{\mu}^{a})]^{2 } \nonumber\\ & & - 2\partial_{\mu}(\bar{a}_{\mu}^{a } + \bar{c}_{\mu}^{a})\partial_{\nu}(\bar{a}_{\nu}^{a } + \bar{c}_{\nu}^{a } ) ] -\bar{\eta}^{a}[\epsilon^{abc}\partial^{\mu}(\bar{a}_{\mu}^{c } + \bar{c}_{\mu}^{c } ) + \epsilon^{abc}(\bar{a}_{\mu}^{c } + \bar{c}_{\mu}^{c})\partial^{\mu}]\eta^{b}.\end{aligned}\ ] ] from @xmath110 and @xmath111 , we can extract the feynman rules involving the external field .
the relevant vertices are shown in fig .
[ figura-1 ] .
( 440,130)(0,0 ) ( 50,20)(20,50)45 ( 50,20)4 ( 80,50)(50,20)45 ( 85,45)@xmath57 ( 10,45)@xmath57 ( 100,30)@xmath112^{2}$ ] ( 241,105)@xmath57 ( 280,80)(250,110)45 ( 280,80)4 ( 310,110)(280,80)45 ( 315,105)@xmath57 ( 330,90)@xmath113 ( 50,80)(20,110)45 ( 50,80)4 ( 50,80)(90,80)45 ( 10,105)@xmath57 ( 81,68)@xmath57 ( 100,90)@xmath114 ( 280,20)(250,50)45 ( 280,20)4 ( 280,20)(320,20)45 ( 241,45)@xmath57 ( 311,8)@xmath57 ( 330,30)@xmath115 the mixed correlation function can be expressed as @xmath116 it has been shown @xcite that only the two legs graphs , containing the external field @xmath117 will contribute to the large distance behavior of @xmath118 . at tree level the relevant graphs are depicted in fig .
[ figura-5 ] in the symmetric phase only the first two graphs in fig [ figura-5 ] would contribute .
their sum , however , actually vanishes , as we can see by using the gauge field propagators given in eq .
( [ propagators ] ) .
this result leads us to the conclusion that in the symmetric phase we have @xmath119 ( 400,130)(0,0 ) ( 10,105)@xmath57 ( 50,80)(20,110)45 ( 50,80)4 ( 80,110)(50,80)45 ( 85,105)@xmath57 ( 110,85)@xmath120 ( 135,105)@xmath57 ( 175,80)(145,110)45 ( 175,80)4 ( 175,80)(225,80)45 ( 255,110)(225,80)45 ( 225,80)4 ( 260,105)@xmath57 ( 285,85)@xmath120 ( 306,105)@xmath57 ( 345,80)(315,110)45 ( 345,80)4 ( 375,110)(348,80)45 ( 380,105)@xmath57 ( 35,30)@xmath120 ( 60,45)@xmath57 ( 100,20)(70,50)45 ( 100,20)4 ( 100,20)(150,20)45 ( 180,50)(150,20)45 ( 150,20)4 ( 185,45)@xmath57 ( 210,30)@xmath120 ( 232,45)@xmath57 ( 270,20)(240,50)45 ( 270,20)4 ( 274,20)(320,20)45 ( 350,50)(320,20)45 ( 320,20)4 ( 355,45)@xmath57 on the other hand , the last three graphs of fig . [ figura-5 ] only occur in the broken symmetry phase where the higgs field possesses a nonzero vacuum expectation value , @xmath82 . from these , using the gauge propagators @xmath121 given in ( [ propagators ] ) we can write explicitly @xmath122\big[i\alpha\epsilon^ { \mu\lambda\nu}\partial_{\lambda } + \frac{\alpha^{2}}{m^{2}}\left(-\box\delta^{\mu\nu } + \partial^{\mu}\partial^{\nu}\right)\big]f(z - z ' ) [ \bar{a}^{d}_{\mu}(z ; x^{b}_{1},\ldots , y^{b}_{m } ) + \bar{c}^{d}_{\mu}(z ; x^{a}_{1},\ldots , y^{a}_{m } ) ] , \nonumber\end{aligned}\ ] ] where @xmath123 expression ( [ 26 ] ) has been evaluated in @xcite giving @xmath124\bigg\ } \label{mcf}\end{aligned}\ ] ] now , we can introduce a composite operator @xmath125 bearing charge and magnetic flux , through @xmath126 from this and ( [ mcf ] ) we obtain the large distance behavior of the composite operator correlation function : @xmath127 \nonumber\\ & & \hspace{6.5 cm } + 4\pi iabm^{2 } \sum_{i \neq j = 1}^{n}[\mbox{arg}({\bf x}_{i } - { \bf x}_{j } ) + \mbox{arg}({\bf y}_{i } - { \bf y}_{j})]\bigg\}\end{aligned}\ ] ] where @xmath128 and we choose the cuts of the @xmath129 functions as @xmath130 and @xmath131 , in such a way that we may write @xmath132 .
the composite field @xmath133 carries magnetic flux and charge , which are both conserved quantities in the broken phase , hence the only non - vanishing functions are the ones with the same number of operators and their hermitean conjugates .
this selection rule appears naturally in the calculation leading to ( [ phiphi ] ) @xcite . the first term in ( [ phiphi ] )
produces the exponential decay of the vortex correlation function .
this implies the energy spectrum of the quantum vortices @xmath133 possesses a gap proportional to the vacuum expectation value of the higgs field squared : @xmath134 .
the fact that the above two - point function vanishes asymptotically at large - distances means the quantum vortex states @xmath135 are orthogonal to the vacuum .
they are also orthogonal to isolated charge and magnetic flux states @xmath136 and @xmath137 .
these properties indicate that the @xmath133-states are genuine and stable quantum excitations of the system .
the analytic structure of the eucllidean correlation functions is closely related to the spin value . except for the case of bosons ,
the euclidean correlators are multivalued , each sheet corresponding to a different ordering of operators in the vacuum expectation values that correspond to these euclidean functions @xcite .
observe that the composite field @xmath133 ( euclidean ) correlation function above is multivalued whenever @xmath138 is not an integer .
the composite field @xmath133 , indeed , has spin @xmath139 and , as expected , is in general an anyon .
notice that @xmath140 and @xmath141 are respectively the magnetic flux and the charge carried by the vortex operator .
the spin , consequently , can be written in a more physical way as @xmath142 .
the values of the euclidean functions on adjacent sheets differ by a phase @xmath143 , hence , we have the following property for the real time vacuum expectation values of fields @xcite .
@xmath144 notice first that whenever @xmath145 _ integer _ ( bosons or fermions ) , @xmath146 .
it follows that each of the vev s @xmath147 and @xmath148 is univalued in this case .
notice now that , conversely , in the case of anyons , @xmath149 _ integer _ and consequently the previous vev s are themselves multivalued , the particular sheet being indicated by the superscript .
observe that the values of the function in two adjacent sheets differ by a factor @xmath150 .
this means the vev s of field operators have a branch cut , the number of sheets being determined by the spin . for @xmath151 ( @xmath149 _
integer _ ) , for instance , there are n sheets . for irrational
spin , the vev s above would have an infinite number of sheets .
this analytic structure will be the basis for our costruction of states with non - abelian statistics .
in this section , we show how to construct states with non - abelian statistics out of the composite anyon vortex fields .
we start by considering the @xmath152point correlation function of this field , namely @xmath153\right\}\,e^{-d_{2 } } \nonumber\\ & = & e^{-2is\mbox{arg}(\bf{x } - \bf{y})}e^{-is\pi}\,e^{-d_{2}}\end{aligned}\ ] ] in which @xmath154 and @xmath155 .
we now introduce the fields that will present non - abelian statistics . for that purpose ,
let us consider the combined fields @xmath156 which are , respectively , self - adjoint and anti - self - adjoint .
using the fact that @xmath157 we conclude that their correlation functions satisfy @xmath158 and @xmath159 using eq .
( [ doispontos ] ) we can write @xmath160 \nonumber\\ \langle \psi_{-}(x)\psi_{+}(y)\rangle & = & \frac{1}{4}\left[e^{-2is\mbox{arg}(\bf{x } - \bf{y})}e^{-is\pi}e^{-d_{2 } } - e^{-2is\mbox{arg}(\bf{y } - \bf{x})}e^{-is\pi}e^{-d_{2}}\right]\end{aligned}\ ] ] let us see what are the braiding properties of the states associated to the fields @xmath161 . using the properties of @xmath162 in ( [ mm ] )
we obtain @xmath163 \stackrel{\longrightarrow}{x \leftrightarrow y } \big[e^{-2\pi is}e^{-2is\mbox{arg}(\bf{x } - \bf{y})}e^{-is\pi}e^{-d_{2 } } \nonumber\\ & & \hspace{10.0 cm } + \,\,e^{2\pi is}e^{-2is\mbox{arg}(\bf{y } - \bf{x})}e^{-is\pi}e^{-d_{2}}\big ] \nonumber\end{aligned}\ ] ] and @xmath164 \stackrel{\longrightarrow}{x \leftrightarrow y } \big[e^{-2\pi is}e^{-2is\mbox{arg}(\bf{x } - \bf{y})}e^{-is\pi}e^{-d_{2 } } \nonumber\\ & & \hspace{10.0 cm } -\,\,e^{2\pi is}e^{-2is\mbox{arg}(\bf{y } - \bf{x})}e^{-is\pi}e^{-d_{2}}\big ] .
\nonumber\end{aligned}\ ] ] observe that , whenever the operator @xmath133 is bosonic or fermionic the phases generated by braiding the @xmath165-particles are identical , i. e. , @xmath166 .
this implies @xmath167 in this case , the above expression shows that whenever the charged vortex operator @xmath133 is bosonic or fermionic , then the self - sdjoint operators @xmath165 are also bosonic or fermionic . on the other hand ,
when the vortex field is an anyon , namely , for @xmath168 _ integer _ , the @xmath169 fields have non - abelian braiding given by @xmath170 \nonumber\\ & = & \frac{1}{2}\big[(\alpha + \alpha^{*})\langle \psi_{+}(x)\psi_{+}(y)\rangle - ( \alpha - \alpha^{*})\langle \psi_{-}(x)\psi_{-}(y)\rangle\big ] \nonumber\\ & = & \cos \delta \langle \psi_{+}(x)\psi_{+}(y)\rangle - i\sin\delta\langle \psi_{-}(x)\psi_{-}(y)\rangle\end{aligned}\ ] ] and @xmath171 \nonumber\\ & = & \frac{1}{2}\big[-(\alpha - \alpha^{*})\langle \psi_{+}(x)\psi_{+}(y)\rangle + ( \alpha + \alpha^{*})\langle \psi_{-}(x)\psi_{-}(y)\rangle\big ] \nonumber\\ & = & -i\sin\delta\langle \psi_{+}(x)\psi_{+}(y)\rangle + \cos\delta\langle \psi_{-}(x)\psi_{-}(y)\rangle\end{aligned}\ ] ] where in the above expression @xmath172 and @xmath173 .
we conclude that when the composite vortex field @xmath133 is an anyon it follows that the @xmath165 fields will have non - abelian braiding given by @xmath174 the braiding matrix and its hermitean adjoint @xmath175 satisfy @xmath176 , being therefore unitary .
we now come to one of our most interesting results : observe that a not gate can be obtained out of the braiding matrix @xmath82 ( up to an i - factor ) by making @xmath177 or , equivalently , @xmath0 , namely , @xmath178 let us consider here the 4-point function of the vortex operator in the broken phase . from eq .
( [ phiphi ] ) we can extract the following expression @xmath180 - 2\pi is + c_{4a}\big\ } \nonumber\\ \langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle & = & \langle \psi(x_{1})\psi^{\dagger}(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle \nonumber\\ & = & \exp\big\{2is[-\mbox{arg}(\vec{x}_{1 } - \vec{x}_{2 } ) + \mbox{arg}(\vec{x}_{1 } - \vec{x}_{3 } ) - \mbox{arg}(\vec{x}_{1 } - \vec{x}_{4 } ) \nonumber\\ & & - \mbox{arg}(\vec{x}_{2 } - \vec{x}_{3 } ) + \mbox{arg}(\vec{x}_{2 } - \vec{x}_{4 } ) - \mbox{arg}(\vec{x}_{3 } - \vec{x}_{4 } ) ] - 2\pi is + c_{4b}\big\ } \nonumber\\ \langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle & = & \langle \psi(x_{1})\psi^{\dagger}(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle \nonumber\\ & = & \exp\big\{2is[-\mbox{arg}(\vec{x}_{1 } - \vec{x}_{2 } ) - \mbox{arg}(\vec{x}_{1 } - \vec{x}_{3 } ) + \mbox{arg}(\vec{x}_{1 } - \vec{x}_{4 } ) \nonumber\\ & & + \mbox{arg}(\vec{x}_{2 } - \vec{x}_{3 } ) - \mbox{arg}(\vec{x}_{2 } - \vec{x}_{4 } ) - \mbox{arg}(\vec{x}_{3 } - \vec{x}_{4 } ) ] - 2\pi is + c_{4c}\big\ } \nonumber\\ \label{1a}\end{aligned}\ ] ] where @xmath181 the correlation functions of the new fields given by ( [ 29 ] ) may be expressed in terms of the correlation functions above as @xmath182 \nonumber\\ \nonumber\\ & & \hspace{-0.5 cm } \langle \psi_{+}(x_{1})\psi_{+}(x_{2})\psi_{-}(x_{3})\psi_{-}(x_{4})\rangle = \langle
\psi_{-}(x_{1})\psi_{-}(x_{2})\psi_{+}(x_{3})\psi_{+}(x_{4})\rangle = \nonumber\\ & & \hspace{-0.5 cm } 2[\langle\psi(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi^{\dagger}(x_{4})\rangle - \langle\psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle - \langle\psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle ] \nonumber\\ \nonumber\\ & & \hspace{-0.5 cm } \langle \psi_{-}(x_{1})\psi_{+}(x_{2})\psi_{-}(x_{3})\psi_{+}(x_{4})\rangle = \langle \psi_{+}(x_{1})\psi_{-}(x_{2})\psi_{+}(x_{3})\psi_{-}(x_{4})\rangle = \nonumber\\ & & \hspace{-0.5 cm } 2[-\langle\psi(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi^{\dagger}(x_{4})\rangle + \langle\psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle - \langle\psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle ] \nonumber\\ \nonumber\\ & & \hspace{-0.5 cm } \langle \psi_{-}(x_{1})\psi_{+}(x_{2})\psi_{+}(x_{3})\psi_{-}(x_{4})\rangle = \langle \psi_{+}(x_{1})\psi_{-}(x_{2})\psi_{-}(x_{3})\psi_{+}(x_{4})\rangle = \nonumber\\ & & \hspace{-0.5 cm } 2[-\langle\psi(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi^{\dagger}(x_{4})\rangle - \langle\psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle + \langle\psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle ] \nonumber\\ \label{2a}\end{aligned}\ ] ] let us see what are the braiding properties of the above functions . using ( [ 1a ] ) and the expression above , we get @xmath183 \nonumber\\ \nonumber\\ % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % & & \langle \psi_{+}(x_{1})\psi_{+}(x_{2})\psi_{-}(x_{3})\psi_{-}(x_{4})\rangle \stackrel{\longrightarrow}{x_{1 } \leftrightarrow x_{2 } } \quad 2\big [ e^{2\pi is}\langle \psi(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi^{\dagger}(x_{4})\rangle \nonumber\\ & & \hspace{1.0 cm } - e^{-2\pi is}\langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle - e^{-2\pi is}\langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle\big ] \nonumber\\ \nonumber\\ % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % & & \langle \psi_{-}(x_{1})\psi_{+}(x_{2})\psi_{-}(x_{3})\psi_{+}(x_{4})\rangle \stackrel{\longrightarrow}{x_{1 } \leftrightarrow x_{2 } } \quad 2\big[- e^{2\pi is}\langle \psi(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi^{\dagger}(x_{4})\rangle \nonumber\\ & & \hspace{1.0 cm } + \ , e^{-2\pi is}\langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle - e^{-2\pi is}\langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle\big ] \nonumber\\ \nonumber\\ % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % & & \langle \psi_{-}(x_{1})\psi_{+}(x_{2})\psi_{+}(x_{3})\psi_{-}(x_{4})\rangle \stackrel{\longrightarrow}{x_{1 } \leftrightarrow x_{2 } } \quad 2\big[- e^{2\pi is}\langle \psi(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi^{\dagger}(x_{4})\rangle \nonumber\\ & & \hspace{1.0 cm } - e^{-2\pi is}\langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi(x_{3})\psi^{\dagger}(x_{4})\rangle + e^{-2\pi is}\langle \psi^{\dagger}(x_{1})\psi(x_{2})\psi^{\dagger}(x_{3})\psi(x_{4})\rangle\big ] .
% % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % \label{e1 } % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % % \end{aligned}\ ] ] now with the help of the eq .
( [ 2a ] ) we can write the right - hand side of eq .
( [ e1 ] ) in terms of correlators of the new fields @xmath184 and @xmath185 , namely @xmath186 from ( [ 85 ] ) we can determine the unitary matrix corresponding to the braiding operation ( monodromy matrix ) @xmath187 .
indeed , we may write the above equation as @xmath188 where @xmath189 and @xmath190 .
we see that it satisfies @xmath191 , being therefore unitary . using the same procedure
we get the monodromy matrices that correspond to the braiding operations @xmath192 , @xmath193 , @xmath194 , @xmath195 and @xmath196 .
these are given by @xmath197 @xmath198 where @xmath199 and @xmath200 .
the only commuting braiding matrices are @xmath201 and @xmath202 , i. e. @xmath203 = 0 \nonumber\end{aligned}\ ] ] it can be easily verified that the unitary monodromy braiding matrices satisfy the yang - baxter relations , @xmath204 or , equivalently @xmath205 we now come to another of our most interesting results .
again we will see that for a particular choice of the spin , the monodromy matrices become logic gates , which are essential for quantum computation algorithms .
indeed , a simply controlled - not operation ( cnot gate ) , then can be obtained by choosing @xmath206 or @xmath177 in @xmath207 ( or @xmath208 ) , namely @xmath209 more logic keys may be obtained accordingly by a straightforward generalization . at this point , one should inquire more precisely about what ultimately determines the value of the spin of the vortices .
we have seen that @xmath210 . assuming the vortices carry one unit of magnetic flux ( @xmath58 ) , we have @xmath211 , hence the spin is @xmath212 where we retrieved the physical units of magnetic flux .
we conclude that the spin is determined by the ratio of the squared higgs vacuum expectation value to the current algebra scalar functional .
the latter is a fixed number , determined by the spectral density of the theory .
the former , is determined in principle by the higgs potential parameters , however , in any concrete associate condensed matter system , the higgs expectation value is a physically adjustable parameter , which will depend on the temperature of the system .
this means , therefore that the value of the spin s is ultimately determined by the temperature and therefore could be adjusted to the value @xmath213 or to any other value by tuning the temperature appropriately .
we have shown that the 2 + 1 d chern - simons - higgs theory in the broken phase contains quantum states with non - abelian statistics .
these states are created by operators that are combinations of electrically charged magnetic vortex fields and their hermitean adjoints .
the euclidean correlation functions of these composite vortex operators have been obtained by the method of quantization of topological excitations , which is based on the idea of order - disorder duality .
all properties of such states may be derived from these correlation functions .
for instance , we may infer from their large distance behavior , that in the ordered phase the quantum vortices are in excited quantum states carrying both nonzero magnetic flux and charge and having a gap proportional to the vacuum expectation value of the higgs field .
we may also show , out of these correlation functions behavior , that the quantum vortices are in general anyons , as one should expect from a field carrying both magnetic flux and charge .
in special cases , however , they can be fermions or bosons , depending on the value of some input parameters . the analytic properties of the euclidean vortex correlation functions , by its turn ,
have been used in order to show that proper ( self - adjoint or anti - self - adjoint ) combinations of vortices and anti - vortices possess non - abelian statistics , whenever these electrically charged vortices are anyons .
the unitary matrices corresponding to each of the non - commuting braiding operations have been explicitly determined as a function of the spin @xmath214 for the case of the two and four - point correlation functions . for a special value of the spin , namely @xmath0 ,
we have shown that the monodromy matrices , which result from the exchange of the ( anti ) self - adjoint vortex states , become the basic logic gates ( not , cnot , and so on ) required by the algorithm of a quantum computer .
the spin value @xmath214 , being proportional to the vacuum expectation of the higgs field can be tuned by the temperature in a in any associated condensed matter system .
it would be nice to find a concrete material realization for the system studied here .
pure non - abelian chern - simons theory has been claimed to describe the state corresponding to the plateau @xmath215 of quantum hall systems @xcite .
it should be investigated , for that matter , whether the coupling of a higgs field could have any physical meaning in this or in any related system .
e. c. marino , b. schroer and j. a. swieca , nucl .
b @xmath225 @xmath226 $ ] , @xmath227 ( @xmath228 ) .
e. c. marino , phys .
d @xmath229 , @xmath230 ( @xmath231 ) ; ; e. c. marino , phys .
b @xmath232 , @xmath233 ( @xmath234 ) .
e.c.marino , _ dual quantization of solitons _ in proceedings of the nato advanced study institute , applications of statistical and field theory methods in condensed matter , ( d.baeriswyl , a.bishop and j.carmelo , eds . ) plenum , ny , @xmath238 . | we naturally obtain the not and cnot logic gates , which are key pieces of quantum computing algorithms , in the framework of the non - abelian chern - simons - higgs theory in two spatial dimensions .
for that , we consider the anyonic quantum vortex topological excitations occurring in this system and show that self - adjoint ( majorana - like ) combinations of these vortices and anti - vortices have in general non - abelian statistics .
the associated unitary monodromy braiding matrices become the required logic gates in the special case when the vortex spin is @xmath0 .
we explicitly construct the vortex field operators , show that they carry both magnetic flux and charge and obtain their euclidean correlation functions by using the method of quantization of topological excitations , which is based on the order - disorder duality .
these correlators are in general multivalued , the number of sheets being determined by the vortex spin .
this , by its turn , is proportional to the vacuum expectation value of the higgs field and therefore can be tuned both by the free parameters of the higgs potential and the temperature . |
knowledge of the @xmath12 coherence factor and associated hadronic parameters is necessary for a measurement of the unitarity triangle angle @xmath4 ( also denoted @xmath13 ) in @xmath14 decays .
( the symbol @xmath6 is used to denote a neutral charm - meson that is not in a flavour eigenstate , or where the flavour eigenvalue is not relevant for the discussion . ) the coherence factor @xmath0 and average strong - phase difference @xmath1 for the decay @xmath15 are defined as follows @xcite : @xmath16 where @xmath17 is the decay amplitude of @xmath18 at a point in multi - body phase space described by parameters @xmath19 , and @xmath20 therefore @xmath21 is the cabibbo - favoured ( cf ) amplitude , averaged over phase space , and @xmath22 is the corresponding doubly cabibbo - suppressed ( dcs ) quantity .
the coherence factor takes a value between 0 and 1 .
it is also useful to define the parameter @xmath23 . in this letter
the term ` hadronic parameters ' of the decay @xmath24 is employed to refer collectively to @xmath0 , @xmath25 and @xmath2 . throughout the discussion charge conjugation
is implicit , the good approximation is made that @xmath11-violation can be neglected in @xmath7 mixing and decay @xcite , and expressions are given in the convention @xmath26 .
the role of the hadronic parameters in measurements sensitive to @xmath4 can be appreciated by considering the decay of @xmath27 mesons to a neutral charm - meson , reconstructed in the inclusive @xmath28 final state , and a charged kaon . neglecting small corrections from @xmath10 mixing , which can be included in a straightforward manner @xcite ,
the decay rates are given by : @xmath29 here @xmath30 is the absolute ratio of @xmath31 to @xmath32 amplitudes .
the phase difference between these two amplitudes is @xmath33 , where @xmath34 is a @xmath11-conserving strong phase .
the coherence factor , which is replaced by unity in the equivalent expression for a single point in phase space or two - body @xmath6-meson decays , modulates the size of the interference term that carries the dependence on @xmath4 .
the hadronic parameters can be measured in the decays of coherently produced @xmath35 pairs at the @xmath9 resonance @xcite , such as are available in the data sets collected by the cleo - c and besiii experiments .
a double - tag technique is employed where one meson is reconstructed in the signal decay @xmath12 , and the other in a tagging mode , for example , a @xmath11 eigenstate .
this method was applied by the cleo collaboration to obtain first constraints on @xmath0 and @xmath1 , together with @xmath36 and @xmath37 , the analogous parameters for @xmath38 decays @xcite .
a later analysis augmented the set of tags with the decay @xmath39 in order to improve the sensitivity to the hadronic parameters of both signal modes @xcite .
these results have been exploited by the lhcb collaboration to interpret a set of measurements with the decays @xmath5 , @xmath24 @xcite and @xmath38 @xcite , in combination with observables from other modes , to yield @xmath40 @xcite .
observables sensitive to @xmath10 mixing in multibody @xmath6-meson decays are also affected by the same hadronic parameters @xcite .
the time - dependent ratio between @xmath41 and @xmath42 decay rates is @xmath43 where @xmath44 is the proper decay time , @xmath45 is the mean @xmath7 lifetime , @xmath46 @xmath47 , @xmath48 , and @xmath49 and @xmath50 are the commonly used mixing parameters , defined for example in ref .
@xcite . as noted in refs .
@xcite , a measurement of @xmath2 , @xmath51 and @xmath52 can either be used to determine @xmath49 and @xmath50 , given external information on @xmath0 and @xmath25 , or employed to set constraints on the latter parameters , provided @xmath49 and @xmath50 are known .
both interpretations are presented in ref .
@xcite , which reports a first observation of @xmath10 mixing with this multibody final state .
improved knowledge of @xmath0 , @xmath1 and @xmath2 is also needed to help understand the contribution of @xmath12 decays to the width difference in the @xmath10 system @xcite .
this letter reports a new analysis of @xmath24 decays in the cleo - c data set , and benefits from an updated monte carlo simulation sample to correct a biased estimate of the background contamination in several sets of double tags that afflicted the results reported in refs
. @xcite .
in addition , the quasi-@xmath11-eigenstate @xmath53 @xcite is included as a new tagging mode to augment the sensitivity of the analysis .
furthermore , an error is corrected in the formalism used to interpret the yields found with the @xmath39 tags .
the observables from this new analysis are then used to determine updated constraints on the hadronic parameters .
finally , a global fit is made to the observables from this study and those measured in @xmath10 mixing by lhcb , to obtain more precise results for the @xmath24 hadronic parameters . since one of the re - measured @xmath9 observables
couples the and systems , it is also possible to determine updated values of @xmath36 , @xmath54 and @xmath55 .
in this section the observables sensitive to the hadronic parameters are reviewed , and a new analysis of the cleo - c data set is presented that updates the measured values with respect to those reported in ref .
information on the hadronic parameters is then obtained from a fit to the updated measurements .
two categories of observables exist , both based on double - tagged measurements in which one @xmath6 meson decays to the signal mode , and the other decays to a tagging mode . in the first category the tagging modes are two - body or higher multiplicity final states , which are treated in an inclusive fashion . in the second case ,
the tagging mode is the self - conjugate final - state and yields are measured in different phase - space bins of this tagging decay , defined in the plane of the dalitz plot . in the category involving inclusive tags ,
so - called @xmath56 observables are constructed , which are the ratio of the measured double - tag yields , after background subtraction and efficiency correction , to the yields expected in the absence of quantum - correlations .
@xmath57 : : these are the ratios where the tagging mode is a @xmath11-even eigenstate ( @xmath58 ) or a @xmath11-odd eigenstate ( @xmath59 ) . neglecting corrections from @xmath10 mixing , which enter through the definitions of the branching fractions used in the normalisation factors , @xmath60 @xmath61.$ ] precise definitions and mixing - corrected expressions for this , and subsequent relations , can be found in ref .
@xmath62 : : this is a @xmath11-invariant observable defined @xmath63 it allows the results for the @xmath11-even and @xmath11-odd tags to be combined together . neglecting mixing , @xmath64 $ ] .
@xmath65 : : this is the ratio involving vs. events , _
i.e. _ those where the kaons are of like sign ( ls ) . neglecting mixing , @xmath66 @xmath67
: : this is the ratio involving vs. events . neglecting mixing ,
@xmath68 $ ] , where @xmath69 is the ratio between the dcs and cf @xmath70 amplitudes and @xmath71 the accompanying strong - phase difference .
@xmath72 : : this is the ratio involving vs. events .
neglecting mixing , @xmath73 @xmath74.$ ] the deviation of any of the @xmath56 observables from unity or @xmath75 from zero is indicative of a non - zero coherence factor . recalling that @xmath2 , @xmath69 and @xmath76 , it is expected that @xmath77 , whereas larger effects can occur for the like - sign observables . the observables involving decays comprise the yields of double tags , after background subtraction and bin - to - bin relative efficiency correction , in eight pairs of bins in the plane of the dalitz plot symmetric about the line @xmath78 ^ 2 = m[k^0_{\rm s } \pi^-]^2 $ ] .
the binning scheme follows the ` equal @xmath79 ' definition of ref .
@xcite , where the partitioning is guided by an amplitude model developed by the babar collaboration @xcite .
the observables are denoted @xmath80 , where the subscript gives the bin number ( @xmath81 , excluding 0 ) .
the values of @xmath80 differ from those expected in the incoherent case in a manner that is dependent on the values of the coherence factor and average strong - phase difference of the signal mode @xcite : @xmath82 { \big ) } .
\label{eq : kspipiyield}\end{aligned}\ ] ] ( in ref .
@xcite this expression is written with an incorrect sign before the term @xmath83 ) . here
@xmath84 is a bin - independent normalisation factor and @xmath85 is the fractional yield of @xmath7 decays that fall into bin @xmath86 .
the parameters @xmath87 and @xmath88 are the amplitude - weighted averages over bin @xmath86 of @xmath89 and @xmath90 , respectively , where @xmath79 is the strong - phase difference between the @xmath91 and @xmath92 amplitudes at a single point in the dalitz plot .
all these @xmath93 quantities are defined ignoring @xmath94 mixing effects , which is appropriate for double - tag yields arising from @xmath9 mesons produced at rest in the laboratory , as is the case for the cleo - c experiment @xcite .
an 818 @xmath95 data set of @xmath96 collisions produced by the cornell electron storage ring ( cesr ) at @xmath97 gev and collected with the cleo - c detector is analysed .
the cleo - c detector is described in detail elsewhere @xcite .
in addition , simulated monte carlo samples are studied to assess possible background contributions and to determine efficiencies .
the evtgen package @xcite is used to generate the decays and geant @xcite is used to simulate the cleo - c detector response . to ensure full understanding of all the inputs to the analysis ,
the selection of all double tags involving the decay @xmath98 is re - performed , although the selection criteria are intended to be identical to those reported in refs .
the full list of final states reconstructed is given in table [ tab : tags ] , with @xmath99 , @xmath100 , @xmath101 , @xmath102 , @xmath103 and @xmath104 .
a single new addition to the list of tag modes is the abundant decay @xmath105 , which has recently been measured to be very close to a @xmath11 eigenstate , with a @xmath11-even fraction of @xmath106 @xcite .
the selection requirements for this new mode are similar to those reported in ref .
in particular , a @xmath107 veto is applied to the @xmath108 combination , as described in ref .
@xcite , in order to suppress contamination from @xmath109 decays .
this veto rejects candidates where the two - track vertex is significantly displaced from the beamspot , or either track has a significant impact parameter . 00
d. atwood and a. soni , _ role of charm factory in extracting ckm phase information via @xmath110 _ , phys . rev .
* d 68 * ( 2003 ) 033003 , arxiv : hep - ph/0304085
. y. amhis _ et al . _
( hfag ) , _ averages of b - hadron , c - hadron , and @xmath45-lepton properties as of summer 2014 _ , arxiv:1412.7515 , online updates at http://www.slac.stanford.edu/xorg/hfag . numbers taken from ` charm 2015 ' update .
m. rama , _ effect of @xmath111 mixing in the extraction of @xmath4 with @xmath32 and @xmath112 decays _ , phys .
d 89 * ( 2014 ) 014021 , arxiv:1307.4384 [ hep - ex ] .
n. lowrey _ et al . _
( cleo collaboration ) , _ determination of the @xmath113 and @xmath42 coherence factors and average strong - phase differences using quantum - correlated measurements _ , phys
* d 80 * ( 2009 ) 031105(r ) , arxiv:0903.4853 [ hep - ex ] . j. libby _ et al .
_ , _ new determination of the @xmath113 and @xmath42 coherence factors and average strong - phase differences using quantum - correlated measurements _ , phys
* b 731 * ( 2014 ) 197 , arxiv:1401.1904 [ hep - ex ] .
r. aaij _ et al . _
( lhcb collaboration ) , _ observation of the suppressed ads modes @xmath114_d k^\pm$ ] and @xmath114_d \pi^\pm$ ] _ , phys .
* b 723 * ( 2013 ) 44 , arxiv:1304.4646 [ hep - ex ] .
r. aaij _ et al . _
( lhcb collaboration ) , _ a study of @xmath11 violation in @xmath115 ( @xmath116 ) with the modes @xmath117 , @xmath118 and @xmath119 _ , phys . rev . * d 91 * ( 2015 ) 112014 , arxiv:1504.05442 [ hep - ex ] .
lhcb collaboration , _ improved constraints on @xmath4 : ckm2014 update _ , lhcb - conf-2014 - 004 ; r. aaij _ et al .
_ ( lhcb collaboration ) , _ measurement of the ckm angle @xmath4 from a combination of @xmath120 analyses _ , phys . lett . * b 726 * ( 2013 ) 151 , arxiv:1305.2050 [ hep - ex ] .
s. malde and g. wilkinson , _ @xmath121 mixing studies with the decays @xmath122 _ , phys
* b 701 * ( 2011 ) 353 , arxiv:1104.2731 [ hep - ph ] . s. harnew and j. rademacker , _ charm mixing as input for model - independent determination of the ckm phase @xmath4 _ , phys . lett .
* b 728 * ( 2014 ) 296 , arxiv:1309.0134 [ hep - ph ] . s. harnew and j. rademacker , _ model independent determination of the ckm phase @xmath4 using input from @xmath123 mixing _ , jhep * 03 * ( 2015 ) 169 , arxiv:1412.7254 [ hep - ph ] .
r. aaij _ et al . _
( lhcb collaboration ) , _ first observation of @xmath124 oscillations in @xmath41 decays and a measurement of the associated coherence parameters _ , phys . rev .
* 116 * ( 2016 ) 241801 , arxiv:1602.07224 [ hep - ex ] .
t. gershon , j. libby and g. wilkinson , _ contributions to the width difference in the neutral @xmath6 system from hadronic decays _ , phys .
* b 750 * ( 2015 ) 338 ; arxiv:1506.08594 [ hep - ph ] . s. malde _ et al .
_ , _ first determination of the @xmath11 content of @xmath125 and updated determination of the @xmath11 contents of @xmath105 and @xmath119 _ , phys . lett .
* b 747 * ( 2015 ) 9 , arxiv:1504.05878 [ hep - ex ] ; m. nayak _ et al .
_ , _ first determination of the @xmath11 content of @xmath105 and @xmath119 _ , phys
* b 740 * ( 2015 ) 1 , arxiv:1410.3964 [ hep - ex ] .
j. libby _ et al .
_ ( cleo collaboration ) , _ model - independent determination of the strong - phase difference between the decays @xmath7 and @xmath126 ( @xmath127 and its impact on the measurement of the ckm angle @xmath4 _ , phys . rev . * d 82 * ( 2010 ) 112006 , arxiv:1010.2817 [ hep - ex ] .
a. bondar and a. poluektov , _ feasibility study of model - independent approach to @xmath13 measurement using dalitz plot analysis _ ,
j. * c 47 * ( 2006 ) 347 , arxiv : hep - ph/0510246 ; a. bondar and a. poluektov , _ the use of quantum - correlated @xmath7 decays for @xmath13 measurement _ , eur .
j. * c 55 * ( 2008 ) 51 , arxiv:0801.0840 [ hep - ex ] .
b. aubert _ et al .
_ ( babar collaboration ) , _ improved measurement of the ckm angle @xmath4 in @xmath128 decays with a dalitz plot analysis of @xmath6 decays to @xmath129 and @xmath130 _ , phys . rev . *
d 78 * ( 2008 ) 034023 .
, arxiv:0804.2089 [ hep - ex ] .
a. bondar , a. poleuktov and v. vorobiev , _ charm mixing in the model - independent analysis of correlated @xmath131 decays _ , phys .
d 82 * ( 2010 ) 034033 , arxiv:1004.2350 [ hep - ex ] .
y. kubota _ et al .
_ ( cleo collaboration ) , _ the cleo ii detector _ , nucl .
. meth . * a 320 * ( 1992 ) 66
; d. peterson _
_ , _ the cleo iii drift chamber _ ,
. meth . * a 478 * ( 2002 ) 142 ; m. artuso _
et al . _ , _ construction , pattern recognition and performance of the cleo iii lif - tea rich detector _ , nucl .
. meth . * a 502 * ( 2003 ) 91 ; r.a .
briere _ et al . _
( cleo - c / cesr - c taskforces and cleo - c collaboration ) , _
cleo - c and cesr - c : a new frontier of weak and strong interactions _
, cornell lepp report clns report no .
01/1742 ( 2001 ) .
lange , _ the evtgen particle decay simulation package _ , nucl .
meth . * a 462 * ( 2001 ) 152 .
r. brun _ et al .
_ , geant 3.21 , cern program library long writeup w5013 , unpublished .
particle data group , k.a .
olive _ et al .
_ , _ review of particle physics _ ,
* c 88 * ( 2014 ) 090001 .
j. insler _ et al .
_ , _ studies of the decays @xmath132 and @xmath133 _ , phys
* d 85 * ( 2012 ) 092016 , arxiv:1203.3804 [ hep - ex ] .
r. aaij _ et al . _
( lhcb collaboration ) , _ studies of the resonance structure in @xmath134 decays _ , phys . rev .
* d 93 * ( 2016 ) 052018 , arxiv:1509.06628 [ hep - ex ] .
x. shi _ et al .
_ ( cleo collaboration ) , _ updated measurements of absolute @xmath135 and @xmath136 hadronic branching fractions and @xmath137 _ , phys . rev . *
d 85 * ( 2012 ) 092016 , arxiv:1312.6775 [ hep - ex ] .
a. poluektov _ et al .
_ ( belle collaboration ) , _ evidence for direct @xmath11 violation in the decay @xmath138 , @xmath139 and measurement of the ckm phase @xmath13 _ , phys . rev . * d 81 * ( 2010 ) 112002 , arxiv:1003.3360 [ hep - ex ] .
del amo sanchez _ et al . _
( babar collaboration ) , _ evidence for direct @xmath11 violation in the measurement of the ckm angle @xmath4 with @xmath140 decays _ , phys .
* 105 * ( 2010 ) 121801 , arxiv:1005.1096 [ hep - ex ] .
b. aubert _ et al .
_ ( babar collaboration ) , _ measurement of the cabibbo - kobayashi - maskawa angle @xmath4 in @xmath141 decays with a dalitz analysis of @xmath142 _ , phys
lett . * 95 * ( 2005 ) 121802 , arxiv : hep - ex/0504039 . | measurements of the coherence factor @xmath0 , the average strong - phase difference @xmath1 and mean amplitude ratio @xmath2 for the decay @xmath3 are presented .
these parameters are important inputs to the determination of the unitarity triangle angle @xmath4 in @xmath5 decays , where @xmath6 designates a superposition of @xmath7 and @xmath8 mesons decaying to a common final state .
the results are based on a combined fit to observables obtained from a re - analysis of the cleo - c @xmath9 data set and those measured in a @xmath10 mixing study performed by the lhcb collaboration .
charm decay , quantum correlations , @xmath11 violation |
the mechanisms that induce magnetic activity in field stars have been intensively studied and discussed over the last decades and this effort is now converging in a fairly consistent scenario that explains both magnetic field generation and angular momentum evolution in both partially and fully convective stars .
magnetic fields in partially convective ( f , g , k and early m ) stars are believed to be generated at the transition region between the radiative interior and the differentially rotating outer convective zone ( tachocline ) through a solar - type @xmath3 dynamo . in this scenario
magnetic fields are generated as a result of a combination of differential rotation ( @xmath4 effect ) and cyclonic convection ( @xmath5 effect ) @xcite .
lower mass m dwarfs of spectral type later than @xmath6m3 are fully convective and therefore do not posses a tachocline , however display clear signs of magnetic activity ( e.g. * ? ? ?
such stars are suggested to rotate as rigid bodies @xcite , implying that the @xmath3 dynamo can not be responsible for producing magnetic fields .
instead , an @xmath7 dynamo has been proposed for magnetic field generation in fully convective stars @xcite .
the efficiency of both the @xmath3 dynamo and the @xmath7 dynamo are expected to be strongly correlated with the rossby number ( * ? ? ?
* @xmath8 , with @xmath9 the rotational period and @xmath10 the convective overturn timescale ) . in partially convective stars magnetic
activity rises for decreasing rossby numbers , implying that larger rotational velocities lead to higher levels of magnetic activity ( for a given @xmath10 , * ? ? ?
similarly , a correlation between activity and rotation is also expected in fully convective stars in which magnetic fields may be generated by the @xmath7 dynamo @xcite .
several early observational studies have explored the connection between magnetic activity and rotation @xcite .
more recently , @xcite showed that the x - ray emission increases with rotation for slowly rotating stars but saturates below a threshold of @xmath11 .
the same relation seems to apply for the magnetic field strength @xcite .
unfortunately , the phenomenon of saturation is not fully understood yet , one possible explanation is that saturation occurs due to a change of dynamo configuration @xcite .
an additional explanation is that saturation is reached when a certain maximum fraction of the available energy flux ( from convection ) is converted into magnetic energy @xcite . for ultracool dwarfs of spectral type m7 - 9.5 this clear rotation - activity relation , however , seems to be no longer valid @xcite .
moreover , it has to be stressed that rotation and magnetic activity in single ( late - type ) m dwarfs might not always be linked @xcite .
rotation and magnetic activity decrease in time @xcite as a consequence of magnetic braking , i.e. angular momentum is extracted from the convective envelope and lost through a magnetized wind .
the braking timescales strongly depend on the spectral type ( or mass ) of the star @xcite , which results in lower - mass partially convective stars spinning - down slower than their higher mass counterparts .
fully convective stars have even longer rotational braking timescales @xcite .
unfortunately the reason for magnetic braking being less efficient in fully convective stars is not completely understood .
this might be related to a change in the magnetic field topology , which has been observed to switch from less ordered field structures to dipoles at the fully convective boundary going from partially to fully convective stars @xcite .
however such a switch should lead to increased magnetic braking for fully convective stars , exactly opposite to what is indicated by the observed rotation rates .
a new scenario for magnetic braking has been recently developed by @xcite who suggest that the steep transition of the braking rates at the fully convective boundary arises due to the significant drop in radius , without invoking any change in the dynamo theory .
m dwarf companions to white dwarfs that form part of detached binaries ( hereafter wd / dm binaries ) offer an ideal test bed for studying magnetic activity , rotation and magnetic braking across the fully convective boundary for three reasons .
( 1 ) wd / dm binaries can easily be separated into wide binaries with orbital periods typically exceeding @xmath12days , and close binaries that evolved through common envelope evolution @xcite with final orbital periods generally below one day @xcite . while the m dwarfs in wide binaries should be unaffected by the presence of the white dwarf , those in close post - common envelope binaries ( pcebs ) are tidally locked , i.e. they are rapidly rotating .
comparing the two samples can provide clear constraints on the impact of fast rotation on activity .
( 2 ) using pcebs with measured orbital periods one can furthermore directly relate the strength of activity to the rotational velocity of the m dwarf ( the orbital period of the binary equals to the rotational period of the m dwarf which then gives the rotational velocity by adopting a spectral type - radius relation ) .
( 3 ) the age of the m dwarfs in wide wd / dm systems can be estimated relatively accurately using cooling tracks plus an initial - to - final mass relation for their white dwarf companions .
this allows to test activity lifetimes that are presumably related to magnetic braking models .
the first studies of magnetic activity in m dwarfs that form part of wd / dm binaries from the sloan digital sky survey ( sdss , * ? ? ?
* ; * ? ? ?
* ) were performed by @xcite and more recently by @xcite .
however , these two studies consider all wd / dm binaries in sdss to be part of close binaries , which is clearly not the case @xcite , and their results should be interpreted with extreme caution .
only our dedicated radial velocity survey of wd / dm binaries from sdss allows to accurately separate wide binaries ( in which the stellar components have evolved as if they were single ) from close binaries that underwent a common envelope evolution and therefore contain tidally locked fast rotating m dwarfs @xcite .
this is fundamental to explore the relation between activity and rotation .
the largest and most homogeneous sample of wd / dm binaries from sdss including radial velocity information has been presented in @xcite .
this sample forms the input for the present paper where we analyse the magnetic activity of m dwarf components in wd / dm binaries , thereby testing possible relations between age , activity and rotation .
the catalogue of @xcite consists of 2248 white dwarf - main sequence binaries from the data release 7 ( dr7 ) of sdss , of which 2019 contain an m dwarf ( wd / dm binaries ) .
the sdss wd / dm binary sample is described in more detail in this section .
we begin introducing our radial velocity survey that allows to separate between close and wide binaries .
we then explain how sdss wd / dm binary spectra can be used for obtaining reliable stellar parameters of both components and finally give indications of how we estimate the ages of our systems .
the population of wd / dm binaries contains close binaries ( post - common envelope binaries or pcebs ) that underwent common envelope evolution @xcite , and wide binaries that never interacted .
the m dwarfs that form part of pcebs are tidally locked and are therefore fast rotators .
the properties of the m dwarfs that are part of wide binaries should resemble those of single m dwarfs as the stellar components in wide wd / dm binaries have evolved as if they were single stars @xcite . to characterize large samples of both close pcebs and wide wd / dm binaries we are performing a large radial velocity follow - up program of sdss wd / dm binaries .
our observing strategy follows a two - step procedure . in a first step
we obtain at least two spectra per target , separated by at least one night , and identify systems as pcebs if the radial velocities we measure from these spectra show more than 3@xmath13 radial velocity variation .
if we do not detect significant radial velocity variation , we consider the system as a wide wd / dm binary candidate .
this technique is obviously less sensitive towards the detection of long orbital period pcebs and/or low - inclination systems and we predict that , on average , @xmath616 per cent of the wide wd / dm binary candidates are in fact unrecognised pcebs @xcite . in a second step , the orbital periods of the pcebs identified in this way are measured from higher cadence radial velocity follow - up . so far we have identified 191 pcebs and 1055 wide wd / dm binary candidates ( e.g. * ? ? ? * ) and have measured the orbital periods of 81 pcebs ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these samples of wide and close wd / dm binaries offer the potential to e.g. directly compare activity fractions of slowly and fast rotating m dwarfs as a function of spectral type .
furthermore , the rotational velocities of the m dwarfs in pcebs with known orbital periods can also be known ( assuming a spectral type - radius relation ) .
these m dwarfs are very fast rotators and should populate the saturated regime , therefore allowing us to test the saturation - type rotation - activity relation .
reliable stellar parameters of the wd / dm binaries are obtained from the sdss spectra using our decomposition / fitting routine @xcite .
the sdss spectrum is initially fitted with a two - component model using a grid of observed m - dwarf templates and a grid of observed white dwarf templates ( see figure[f - norm1 ] ) , and the spectral type of the m dwarf is determined . then , the best - fit m dwarf template , scaled by the appropriate flux scaling factor , is subtracted and the residual white dwarf spectrum is fitted with a model grid of da white dwarfs @xcite to determine the white dwarf effective temperature and surface gravity . from an empirical spectral type - radius - mass relation for
m - dwarfs @xcite and a mass - radius relation for white dwarfs @xcite the masses and radii of the m dwarf and the white dwarf components are calculated . in the context of the present paper the spectral types of the m - dwarfs are of fundamental importance .
we therefore compared the spectral types of our template fitting routine with those obtained by @xcite .
fitting several hundred m dwarfs from the list of @xcite , equally distributed in spectral type , with our m dwarf templates and adopting an uncertainty of half a spectral type sub - class , we find that 98 per cent of our spectral classifications agree with those of @xcite within one spectral subclass .
for a given mass , the magnetic activity of m dwarfs is closely related to their rotational velocity ( see for example @xcite ) .
if other effects , such as a close companion , do not affect the star , magnetic braking is slowing down the rotation with time .
the magnetic activity in our sample of m dwarfs that form part of wide sdss wd / dm binaries should therefore primarily depend on the age of the system .
fortunately , for a given wide wd / dm binary one can estimate the age of the system , and consequently the age of the m dwarf companion , by summing the white dwarf cooling age to its main sequence progenitor lifetime .
sdss wd / dm binary ages were previously calculated by @xcite for a sub - sample of sdss wd / dm binaries .
we estimate the ages for the wide sdss wd / dm binary sample studied in this work ( section[s - ident ] ) following the same strategy . the white dwarf cooling ages were calculated using the evolutionary tracks of of @xcitebergeron / coolingmodels/ ] .
the initial main sequence progenitor masses were calculated using the initial - to - final mass relation by @xcite and we used the equations of ( * ? ? ?
* see their appendix ) to calculate their main sequence lifetime . in this exercise we excluded wide wd / dm binaries containing white dwarfs cooler than 12000k , as the white dwarf spectral fitting overestimates the mass in such systems ( e.g. * ? ? ?
* ; * ? ? ?
* ) , which affects the resulting cooling age .
for the same reason we only determined the age of wide wd / dm binaries if the error of the mass of the white dwarf is smaller than @xmath14 , and given that the initial - to - final mass relation is only well defined for systems with white dwarfs masses @xmath150.55@xmath16 , we only considered wide wd / dm binaries containing white dwarfs with masses above this value .
in the context of this paper it is necessary to unambiguously identify both active and inactive m dwarfs among the sdss wd / dm binary catalogue described in section[s - cat ] .
two features have been predominantly used as activity indicators in optical spectroscopy , namely the caii h and k resonance lines and h@xmath5 emission .
h@xmath5 emission is more accessible in wd / dm binaries because the caii h and k lines fall at wavelengths generally dominated by the flux of the white dwarf .
we therefore consider a given m dwarf component as active if h@xmath5 emission is detected in its sdss spectrum .
we developed an automatic routine for this purpose that is described below . in a first step
the best white dwarf model fit ( see section[s - param ] , figure[f - norm1 ] ) is subtracted from the sdss wd / dm binary spectrum . in order to avoid our results being dominated by low quality data , only residual m dwarf spectra with a signal - to - noise ratio ( s / n ) of at least @xmath15 10 are used3 was used , however the activity fractions ( see details in section[s - frac ] ) we obtained for late - type m dwarfs where very much affected by low - quality data , and we therefore decided to increase the s / n threshold to a least @xmath1510 . ] , which results in a sample of 739 objects .
for these 739 systems the continuum flux of the residual m dwarf around h@xmath5 ( 6420 @xmath17 6700 ) is fitted with a parabola ( excluding the range ha-7@xmath18 ha@xmath19 , where ha@xmath20 ) .
the parabolic fit is then used to normalize the spectrum .
examples are shown in figure[f - norm2 ] . once the spectrum is normalized , the equivalent width of h@xmath5 ( ew@xmath1 ) is calculated within the range ha-7@xmath21 ha@xmath19 ( emission in our case is indicated by a negative ew ) .
active m dwarfs are then selected applying the following criteria : @xmath22 where eew@xmath1 is the ew@xmath1 error , h is the peak of the h@xmath5 emission above the continuum level , and n@xmath23 is the noise at continuum level .
an m dwarf is considered inactive if , conversely , any of the following conditions apply : @xmath24 the above procedure results in 373 active and 366 inactive m dwarfs among the 739 sdss wd / dm binaries that form our final sample ( note that for objects with multiple sdss exposures we consider as active those systems for which h@xmath5 emission is detected in at least one of the sdss available m dwarf residual spectra ) .
visual inspection of the active and inactive samples confirms the efficiency of our method , with only three per cent of miss - classifications .
note that these miss - classifications are due to the fact that we are subtracting the white dwarf contribution from the sdss spectra , which may introduce uncertainties in the residual flux of the m dwarf if the white dwarf dominates the spectral energy distribution of the sdss wdms binary .
examples of active and inactive m dwarf residual spectra are shown in the bottom panels of figure[f - norm2 ] . from our radial velocity survey ( section[s - sample ] ) we identify 63 pcebs and 599 wide wd / dm binaries among the 739 systems .
ages are available for 118 of the 599 wide binaries ( section[s - age ] ) .
sdss single m dwarf activity fractions have been published by @xcite for different sdss data releases ( dr4 , dr5 and dr7 respectively ) , as well for independent ( non - sdss ) m dwarf samples ( e.g. * ? ? ?
* ; * ? ? ?
all show the same general trend , i.e. a significant increase of the fraction of active systems going from partially to fully convective stars .
the m dwarf activity fractions of our 739 sdss wd / dm binaries as a function of spectral type are shown on the left panel of figure[f - frac ] ( black dots ) together with the activity fractions obtained by @xcite for single m dwarfs from the sdss .
the two distributions differ significantly at early - to - mid ( m0 - 4 ) spectral types . at first glance
this difference may not seem entirely surprising , as a significant fraction of wd / dm binaries are in fact close pcebs containing rapidly rotating m dwarfs and , as outlined in the introduction , fast rotation induces magnetic activity .
one might therefore speculate , as e.g. @xcite and @xcite , that the increased activity fraction of wd / dm binaries containing early - to - mid spectral type m dwarfs is caused by a fraction of short orbital period pcebs .
however , we have shown that very few wd / dm binaries containing early type m dwarfs ( @xmath25m3 ) are pcebs @xcite , i.e. the majority of these pcebs have evolved towards shorter orbital periods due to angular momentum loss via magnetic braking and entered a semi - detached configuration becoming cataclysmic variables @xcite .
therefore the fraction of pcebs misclassified as wide systems in our radial velocity survey should be low in this spectral type range too .
clearly , the observed activity fraction does not only depend on the rotational velocities but also on the intrinsic ages of the m dwarfs .
therefore , to understand the measured activity fractions requires to separately investigate close pcebs and wide wd / dm binaries , and to take into account the ages of the latter population .
the sdss pceb orbital period distribution peaks at @xmath68 hours @xcite , and the longest sdss pceb orbital period is slightly below 10 days @xcite .
consequently m dwarfs in sdss pcebs are tidally locked @xcite and are expected to be magnetically active .
the middle panel of figure[f - frac ] shows that all m dwarfs within sdss pcebs are active independently of the spectral type .
one might be inclined to interpret this result as an obvious consequence of rotation causing activity , nevertheless it is important to keep in mind that h@xmath5 emission in pcebs may arise from irradiation effects if the white dwarf is relatively hot and/or the pceb orbital period is relatively short ( e.g. * ? ? ?
* ; * ? ? ?
therefore , we investigate below the effect of irradiation present among the 63 pcebs in our sample . for 47 of these pcebs
the orbital periods are well known and the stellar parameters relatively well constrained @xcite . with this information at hand
we can easily estimate the influence of irradiation by calculating the ratio between the intrinsic flux of the m dwarf ( f@xmath26 ) and the irradiating flux from the white dwarf on the m dwarf s surface ( f@xmath27 ) , @xmath28 where @xmath29 and @xmath30 are the m dwarf and white dwarf effective temperatures respectively , @xmath31 is the white dwarf radius , and @xmath32 is the orbital separation .
the resulting f@xmath26/f@xmath27 ratios are shown in figure[f - ratio ] as a function of white dwarf effective temperature ( top panel ) and orbital period ( bottom panel ) . among the 47 pcebs , eight were followed - up photometrically as part of our survey ( section[s - sample ] ) and
inspection of their light curves reveals clear irradiation effects in only two objects ( red solid triangles in figure[f - ratio ] , cyan dots represent the remaining six systems ) . based on the evidence drawn from these eight pcebs
we define f@xmath33 f@xmath27 as the threshold separating systems dominated by irradiation from those that are dominated by the intrinsic emission of the m dwarf ( horizontal gray dashed line in figure[f - ratio ] ) .
93 per cent of our systems lie above this limit and are presumably unaffected by irradiation . although the limit defined in that way just represents an empirically motivated estimate , we conclude that irradiation induced h@xmath5 plays a minor role for our sample of pcebs and that _ all m dwarfs in pcebs are magnetically active_. the m dwarf intrinsic flux / white dwarf irradiated flux ratio ( f@xmath26/f@xmath27 ) as a function of white dwarf effective temperature ( top ) and orbital period ( bottom ) .
the horizontal gray dashed line represents f@xmath34 f@xmath27 and is adopted to be the maximum limit for irradiation effects to be strong enough for h@xmath5 to be in emission .
red solid triangles represent confirmed pcebs showing signs of irradiation effects in their light curves , cyan solid dots pcebs displaying no signs of irradiation in their light curves .
the large errors obtained for several of our systems are a simple consequence of propagating relatively large uncertainties on the estimated effective temperatures @xcite .
] it is generally accepted that the stellar components in wide wd / dm binaries have evolved in the same way as if they were single stars @xcite .
consequently the properties of the m dwarf companions in such binaries should be fairly similar to those of single m dwarfs , and naively one would expect similar activity fractions in both populations .
this is clearly not the case ( see the right panel of figure[f - frac ] ) .
the larger activity fractions detected in our sample must be caused by the fact that these m dwarfs are part of binary systems , but as these binaries are fairly wide ( @xmath35days , * ? ? ?
* ) , interactions between the two stars can be excluded to to be the cause of the increased activity fractions . as mentioned in section[s - sample ]
, our radial velocity survey suffers from observational biases and we expect a small percentage of the wide wd / dm binary sample to be formed by unrecognised pcebs . it is therefore important to evaluate whether this effect can explain the increase of the activity fractions seen in m dwarfs that are part of wide wd / dm binaries as compared to those of single m dwarfs .
for this purpose we calculate in what follows the pceb contamination in our wide wd / dm binary sample at a given spectral type . the exact number of unrecognised pcebs depends on the spectral type of the m dwarf components .
@xcite demonstrated that the observed fraction of pcebs containing early - type m dwarfs is @xmath610 per cent at m0 - 1 spectral types , and increases to only @xmath620 per cent at m2 . combining these fractions with the pceb detection probability of our radial velocity survey of @xmath60.8 @xcite
we calculate a pceb contamination of less than five per cent in the m0 - 2 range ( where the activity fractions most differ ) .
therefore this effect can certainly not explain the difference between the activity fractions in the single and wide wd / dm binary samples ( right panel of figure[f - frac ] ) .
identifying both components in a sdss wd / dm binary spectrum requires that neither the white dwarf nor the m dwarf completely outshines the other component .
cold white dwarfs are too faint to be detected in binaries with early - type m dwarfs .
this implies the white dwarf to have an effective temperature @xmath36k thus excluding the coolest ( i.e. oldest ) white dwarfs @xcite . ) .
these two effects result in no systems with m01 m dwarf components being included in the pceb sample studied in this work ( figure[f - frac ] , middle panel ) . ]
the mass distribution of white dwarfs in wide wd / dm binaries is similar to the mass distribution of single white dwarfs @xcite implying that the progenitors of such white dwarfs were mainly f stars that spent a rather short time on the main sequence .
in other words , the simple requirement that we need to be able to detect the white dwarf in the sdss spectrum to classify a system as wd / dm binary introduces a significant bias towards young systems that is not present in the samples of single m dwarfs . in order to investigate this hypothesis
further we show in figure[f - age ] the activity fractions as a function of age for 118 wide wd / dm binaries with available estimated ages ( section[s - age ] ) .
two important results are extracted from inspecting this figure .
( 1 ) as expected , m dwarfs in wide wd / dm binaries are in general significantly younger than the average age of field m dwarfs of @xmath37gyrs @xcite . on average ,
younger stars rotate faster and therefore a considerably higher fraction of m dwarfs in wd / dm binaries are active .
( 2 ) there is a clear decrease of the activity fractions for older systems . even though the errors in the activity fractions are fairly broad , this is in very good agreement with magnetic braking slowing down the rotation of the m dwarfs that are part of wide wd / dm binaries with time .
it is worth mentioning that the activity fractions measured for the systems with available ages seem to be higher than those obtained for the complete wide wd / dm binary sample ( see figure[f - age ] and right panel of figure[f - frac ] ) .
this is simply because we can only obtain reliable ages if the white dwarfs are hot enough to dominate the sdss wd / dm binary spectra , which reflects in an effective temperature of @xmath38k , depending on the spectral type of the m dwarf .
thus an older population containing inactive systems exists in our wide wd / dm binary sample for which we are not able to estimate the ages , making the activity fractions lower for the complete wide wd / dm binary sample but still considerably higher than those measured for field m dwarfs .
we conclude that the age effect described above can explain the observed difference between the activity fractions measured by @xcite for single m dwarfs and those obtained here for the m dwarfs in wide wd / dm binaries .
as outlined in the introduction , rotation and magnetic activity are strongly correlated .
the strength of activity increases with rotation and saturates around @xmath39 , the exact value of the saturation threshold velocity depending on the mass of the star @xcite .
such a saturation - type rotation - activity relation has been observed for the entire spectral type range of m dwarfs @xcite . in this section
we investigate the saturation - type relation in our sample of 47 pcebs for which we know the orbital periods ( section[s - pcebfrac ] ) . for this purpose , the strength of magnetic activity of the 47 systems is quantified by the ratio between the h@xmath5 and bolometric luminosities , l@xmath1/l@xmath40 .
this luminosity ratio is obtained from the measured ew@xmath1 values ( see section[s - ident ] ) following the equations of @xcite , who obtained conversion factors between the two considered quantities .
compared to single field stars , m dwarf companions in pcebs are tidally locked and are therefore expected to be very fast rotators that should populate the saturated regime .
top panel : the strength of magnetic activity ( evaluated through the l@xmath1/l@xmath40 luminosity ratio ) as a function of rotational velocity ( v@xmath41 ) of 47 m dwarfs in close pcebs .
our data ( black solid dots ) extend the work by ( * ? ? ?
* red triangles ) to faster rotators and unambiguously confirm the saturation limit of the strength of magnetic activity at @xmath0l@xmath1/l@xmath42 .
bottom panel : the strength of magnetic activity as a function of the rossby number
. the increased scatter of @xmath0l@xmath1/l@xmath40 at @xmath43 may indicate chromospheric supersaturation ] figure[f - vrot ] ( top panel ) shows the activity strength as a function of rotational velocity v@xmath41 for the 47 m dwarf companions in pcebs with well known orbital periods ( black solid dots ) and the single m dwarfs from the catalogue of ( * ? ? ?
* red triangles ) .
note that the latter sample is divided into systems with accurately measured rotational velocities ( solid red triangles ) and systems with upper limits to the rotational velocities ( open red triangles ) , and that these values , contrary to ours , represent projected rotational velocities ( v@xmath44 ) .
inspection of the top panel of figure[f - vrot ] clearly reveals that the m dwarfs in pcebs populate the regime of the fastest rotators in which the strength of magnetic activity saturates at @xmath0l@xmath1/l@xmath42 .
this result represents a crucial and very robust confirmation of the proposed rotation - activity relations in m dwarfs . to further illustrate this point
, we show in the bottom panel of figure[f - vrot ] the activity strength as a function of the rossby number @xmath45 .
the latter has been calculated as described in @xcite .
all our systems fall below the threshold of @xmath11 , well within the saturated regime .
a closer inspection of the bottom panel of figure[f - vrot ] reveals that whilst practically all systems with @xmath46 have -3.5@xmath47l@xmath1/l@xmath48 , there is substantial system - to - system scatter for smaller rossby numbers .
interestingly , @xcite found a decline in x - ray saturation for stars with @xmath49 in the coronal emission of young g and k dwarfs , an effect termed `` supersaturation '' . whilst evidence for coronal supersaturation in f , g ,
k dwarfs is growing @xcite , coronal supersaturation has not been identified in m dwarfs @xcite and evidence for chromospheric supersaturation has not been identified neither in f , g , k nor m dwarfs @xcite . while our data do not show a statistically significant decline of the strength of activity for very fast rotators ( top panel of figure[f - vrot ] ) , the increased scatter of the chromospheric activity level for systems with very small rossby numbers in our pcebs sample is intriguing in the context of supersaturation .
one could speculate that we see first indications of chromospheric supersaturation in m dwarfs .
however , the current data are clearly not conclusive and additional measurements of chromospheric activity of m dwarfs in fast rotating pcebs are required to test this hypothesis .
angular momentum evolution , rotation and age are closely related in low - mass stars .
magnetic field generation strongly depends on the rossby number , therefore the rotation , of a given star @xcite .
thus , magnetic activity increases with rotation until its strength saturates for fast rotation rates @xcite .
magnetic braking slows down the rotation of the star and consequently magnetic activity and rotation decrease in time @xcite .
the timescale of magnetic braking depends on the mass of the star , i.e. in fully convective stars magnetic braking becomes inefficient @xcite and the activity lifetimes increase @xcite .
wd / dm binaries offer an excellent opportunity to independently investigate the relations among magnetic activity , rotation and age in m dwarfs .
this is because wd / dm binaries can be separated into wide binaries that never interacted , and in which the properties of the m dwarf components should be fairly similar to those of single stars , as well as close binaries that underwent common envelope evolution ( pcebs ) in which the m dwarf is tidally locked and rotates very rapidly @xcite . to test the activity - rotation - age connection with the largest and most homogeneous sample of sdss wd / dm dwarf binaries
currently known @xcite has been the main goal of this paper .
our main results can be summarised as follows : * we have demonstrated that for m dwarfs in pcebs fast rotation implies activity independently of the spectral type .
this supports the connection between rotation and activity as a necessary ingredient for dynamo models responsible for magnetic field generation in both partially and fully convective stars @xcite .
* we have shown that m dwarfs in close wd / dm binaries populate the saturated regime ( @xmath50 ) and that the strength of magnetic activity saturates at @xmath0l@xmath1/l@xmath51 .
this represents a crucial confirmation of the saturation - type rotation - activity relation for m dwarfs that are very fast rotators , in perfect agreement with the results by @xcite for single m dwarfs .
* we have provided additional observational evidence for magnetic braking becoming inefficient in fully convective stars i.e. the activity fractions of m dwarfs in wide wd / dm binaries show a significant increase at the fully convective boundary .
we have also shown that the activity fractions of m dwarfs in wide wd / dm binaries decrease in time , which proves the spin - down of these m dwarfs due to magnetic braking .
we conclude the observational findings from the sdss wd / dm binary sample are consistent with the current picture of angular momentum evolution and dynamo generation of magnetic fields in low - mass stars , and add important information to our understanding of the relations between magnetic activity , rotation , age and magnetic braking in m stars .
arm acknowledges financial support from fondecyt , grant number 3110049 .
mrs acknowledges support from milenium science initiative , chilean ministry of economy , nucleus p10 - 022-f , and fondecyt ( 1100782 ) .
we thank ada nebot gomez - moran for helpful discussions and the anonymous referee for his / her suggestions that helped improving the quality of the paper . | we make use of the largest and most homogeneous sample of white dwarf / m dwarf ( wd / dm ) binaries from the sloan digital sky survey ( sdss dr7 ) to investigate relations between magnetic activity , rotation , magnetic braking and age in m stars .
these relations are studied separately for close wd / dm binaries that underwent a common envelope phase and thus contain tidally locked and hence rapidly rotating m dwarfs , and for wide wd / dm binaries that never interacted . for the wide wd / dm binary sample
we find that the m dwarf activity fractions are significantly higher than those measured in single m stars of spectral type m0 to m5 .
we attribute this effect as a consequence of m dwarfs in wide sdss wd / dm binaries being , on average , significantly younger and hence more active than the field m dwarf population .
the measured m dwarf activity fractions in wide wd / dm binaries show as well a significant increase from spectral types m3 to m5 , where these low - mass stars become fully convective .
this provides additional observational evidence for magnetic braking being less efficient below the fully convective boundary , in agreement with the hypothesis of fully convective stars having considerably longer activity lifetimes than partially convective stars .
the m dwarfs in all our close binaries are active , independently of the spectral type , giving robust observational evidence for magnetic activity being enhanced due to fast rotation .
the rotational velocities of the m dwarfs in our close binary sample are significantly higher than seen among field m dwarfs , however the strength of magnetic activity remains saturated at @xmath0l@xmath1/l@xmath2 .
this unambiguously confirms the m dwarf saturation - type rotation - activity relation .
binaries : spectroscopic stars : low - mass stars : white dwarfs binaries : close stars : post - agb [ firstpage ] |
the thirty seven years ago we discovered theoretically a threshold phenomenon known as the mixing of @xmath0 and @xmath1 resonances that breaks the isotopic invariance considerably , since the effect @xmath2 in the module of the amplitude @xcite ; see also ref .
this effect appears as the narrow , @xmath3 mev , resonant structure between the @xmath4 and @xmath5 thresholds , @xmath6 and vice versa . since that time many new proposals were appeared , concerning both the searching it and estimating the effects related with this phenomenon [ 3 - 29 ] .
nowadays , this phenomenon has been discovered experimentally and studied with the help of detectors ves in protvino @xcite and besiii in beijing @xcite in the processes + ( a ) @xmath7 @xmath8@xmath9 @xcite , + ( b ) @xmath10 @xcite , + ( c ) @xmath11@xmath12@xmath13@xmath12@xmath14@xmath12@xmath15 @xcite , + ( d ) @xmath16 @xcite , + ( e ) @xmath17 @xcite , + ( f ) @xmath18 @xcite + it has become clear @xcite that the similar isospin breaking effect can appear not only due to the @xmath19 mixing , but also for any mechanism of the production of the @xmath20 pairs in the @xmath21 wave , @xmath22 . and @xmath5 thresholds . ] thus a new tool to study the production mechanism and nature of light scalars is emerged .
the main contribution to the @xmath23 mixing amplitude , caused by the diagrams shown in fig . 1 , has the form the @xmath20 loop mechanism of the @xmath23 mixing . ]
@xmath24\nonumber\\ \hspace*{-35pt } & & \approx\frac{g_{a^0_0k^+k^-}g_{f_0k^+k^- } } { 16\pi}\,i\,\bigl(\rho_{k^+k^-}(m)-\rho_{k^0\bar k^0}(m)\bigr),\nonumber\end{aligned}\ ] ] where @xmath25 ( invariant virtual mass of scalar resonances ) @xmath26 and @xmath27 ; in the region @xmath28 , @xmath29 should be replaced by @xmath30 .
the modulus and the phase of @xmath31 are shown in fig .
2 . ( a )
an example of the modulus of the @xmath23 mixing amplitude .
( b ) the phase of the @xmath23 mixing amplitude . ] in the region between the @xmath4 and @xmath5 thresholds , which is the 8mev wide , @xmath32 note that @xmath33 the branching ratios of the isospin - breaking decays @xmath34 and @xmath35 , caused by the @xmath23 mixing , are @xcite @xmath36 @xmath37 where @xmath38 and @xmath39 are the propagators of the @xmath0 and @xmath1 resonances , respectively .
mass spectra in the isospin - violating decays @xmath34 and @xmath35 , caused by the @xmath23 mixing .
the solid and dashed lines are generally similar each other .
the dotted vertical lines show the locations of the @xmath4 and @xmath5 thresholds . ]
figure 3 shows the mass spectra correspond to the integrands in the above equations .
and @xmath0 resonances with the @xmath40 , @xmath20 , and @xmath41 channels obtained in ref .
@xcite from the besiii data on the intensities of the @xmath42 and @xmath43 transitions measured in the reactions ( b ) and ( c ) @xcite . ]
the phase jump ( see fig .
2(b ) ) suggests the idea to study the @xmath44 mixing in polarization phenomena @xcite . if the process amplitude with the spin configuration is dominated by the @xmath44 mixing then the spin asymmetry of the cross section jumps near the @xmath20 thresholds .
an example is the reaction @xmath45 on a polarized proton target .
the corresponding differential cross section has the form @xmath46,\nonumber\end{aligned}\ ] ] and the dimensionless normalized spin asymmetry is @xmath47 , @xmath48 . and @xmath49 are the @xmath50-channel helicity amplitude with ana without nucleon helicity flip , @xmath51 is the angle between the normal to the reaction plain formed by the momenta of the @xmath52 and @xmath53 system , and the transverse ( to the @xmath52 beam axis ) polarization of the the proton target , and @xmath54 is a degree of this polarization . ]
figure 4 illustrates the strong asymmetry jump which is straightforward manifestation of the @xmath55 mixing amplitude interfering with the isospin preserving one in the @xmath56 and @xmath57 regge exchange model .
details and various variants may be found in refs .
@xcite .
these effects are still in waiting for their studies .
manifestation of the @xmath55 mixing effect in the reaction @xmath58 on a polarized target at @xmath59 gev in the @xmath56 and @xmath57 exchange model .
the solid ( dotted ) curves show the spin asymmetry @xmath60 as a function of @xmath53 invariant mass , @xmath25 ( smoothed with 10 mev mass resolution ) . ]
estimated are the contributions of the following mechanisms responsible for the decay @xmath62 @xcite : \(1 ) the contribution of the @xmath23 mixing , @xmath63 , \(2 ) the contribution of the transition @xmath64 , arising due to the pointlike decay @xmath65 , \(3 ) the contribution of the transition @xmath66 , where @xmath67 , and \(4 ) the contribution of the transition @xmath68 , where @xmath69 ( or @xmath70 ) and @xmath71 .
these mechanisms break the conservation of the isospin due to the nonzero mass difference of the @xmath72 and @xmath73 mesons .
they result in the appearance of the narrow resonance structure in the @xmath74 mass spectrum in the region of the @xmath20 thresholds , with the width @xmath75 mev .
the observation of such a structure in experiment is the direct indication on the @xmath20 loop mechanism of the breaking of the isotopic invariance .
we point out that existing data should be more precise , and it is difficult to explain them using the single specific mechanism from those listed above . taking the decay @xmath76 as the example , we discuss the general approach to the description of the @xmath20 loop mechanism of the breaking of isotopic invariance .
\(1 ) the matter is that the @xmath77@xmath12@xmath78 @xmath12@xmath79@xmath12@xmath80 @xcite and @xmath81@xmath12@xmath13@xmath12 @xmath82@xmath83@xmath84 @xcite decays are described by the @xmath23 mixing well enough : @xmath85 @xmath86 as for the @xmath61 decay @xcite , its description requires the `` terrible '' @xmath23 mixing : @xmath87 and , as a result , the inconvenient coupling constants of the scalar mesons with the pseudo - scalar mesons in the many cases @xmath88 for example , due to the very strong coupling of @xmath0 with the @xmath20 channel , the width of the @xmath0 resonance in the @xmath53 mass spectrum turns out to be near 15 mev .
\(2 ) the the pointlike decay @xmath89 gives @xmath90 instead of the experimental value @xmath91 the @xmath74 mass spectrum in the decay @xmath92 looks similar to the curves in fig .
3 for the @xmath93-@xmath1 mixing case .
however , it is clear that the pointlike mechanism of the decay @xmath94 can not by itself provide the considerable probability of the @xmath95 transition .
\(3 ) the isospin - breaking decay @xmath96@xmath12@xmath14 @xmath83@xmath97 is induced by the diagram shown in fig . 5 , [ !
ht ] the diagram of the decay @xmath62 via the @xmath98 intermediate states.,title="fig:",width=287 ] because the contributions from the @xmath4 and @xmath5 pair production are not compensated entirely .
the transition @xmath99 gives the shape of the @xmath74 spectrum practically coincides with the corresponding spectrum caused by the @xmath23 mixing , but its @xmath100 is much less then the experimental value @xmath101 so , the @xmath102 transition mechanism alone is also insufficient to understand the experimental data .
\(4 ) the variant @xmath103 is rejected by the shapes of the @xmath104 and @xmath20 mass spectra in the decay @xmath105 . as for @xmath106
, it provides the results similar to @xmath107 and consequently can not describe the data alone .
the isospin breaking amplitude @xmath108 can be expanded near the @xmath20 threshold into the series in @xmath109 : @xmath110 + b(m)[\rho^2_{k^+k^-}(m)\ \nonumber \\ \hspace*{-22.5pt } & -\rho^2_{k^0 \bar k^0}(m ) ] + o[\rho^3_{k^+k^-}(m)-\rho^3_{k^0\bar k^0}(m)]+\cdot\cdot\cdot \left.\right\}.\nonumber\end{aligned}\ ] ] with a good accuracy @xmath111.\ \ \qquad\end{aligned}\ ] ] the amplitude @xmath112 contains the information about all possible mechanisms of production of the @xmath20 system with isospin @xmath113 in @xmath21 wave in the process @xmath105 . from the data on the decay
@xmath114 one can extract the information about @xmath115 in the region of the @xmath4 and @xmath5 thresholds , @xmath116 where @xmath117=@xmath118^{1/2}\\ /(2m_{f_1})$ ] .
moreover , the information about @xmath115 at @xmath119 can be obtained from the data on the @xmath20 mass spectra measured in the decays @xmath105 .
for instance , the @xmath4 spectrum in the decay @xmath120 can be represented in the form @xmath121 fitting the data on @xmath122 , one can find the value @xmath123 and obtain the following approximate estimate @xcite @xmath124 thus its comparison with the data on the decay @xmath125 permits one to verify their consistence with the data on the decay @xmath94 and with the idea of the breaking of isotopic invariance caused by the mass difference of @xmath72 and @xmath73 mesons .
according to besiii @xcite , the mass and the width of the @xmath128 peak in the @xmath15 channel are @xmath129 mev and @xmath130 mev , respectively , while the branching ratio is @xmath131 in addition , the besiii gives the ratio @xmath132 that rules out practically the explanation of the discovered effect by means of the @xmath55 mixing .
the diagram of the decay @xmath133 . in the region of the @xmath128 resonance
all intermediate particles in the loop of this triangle diagram can lie on their mass shells .
that is , in the hypothetical case of the stable @xmath134 meson the logarithmic singularity appears in the imaginary part of the triangle diagram.,width=283 ] we discuss the possibility of the theoretical explanation of the large breaking of isotopic invariance in the decay @xmath128@xmath12@xmath14@xmath12@xmath84 by means of the anomalous landau thresholds ( the logarithmic triangle singularities ) , which are in the transition @xmath135 ( see fig .
6 ) , and show that the account of the finite width of the @xmath136 ( @xmath137 mev ) smoothes the logarithmic singularities in the amplitude and results in the suppression of the calculated width of the decay @xmath138 by the factor of @xmath139 in comparison with the case of @xmath140 @xcite . the accounting of the finite width of the @xmath134 resonance , i.e. , the averaging of the amplitude over the resonance breit wigner distribution in accord with the spectral klln
lehmann representation for the propagator of the unstable @xmath134 meson , smoothes the logarithmic singularities of the amplitude and hence makes the compensation of the contributions of the @xmath141 and @xmath142 intermediate states more strong .
this results in both the diminishing of the calculated width of the decay @xmath143 by a number of times in comparison with the case of @xmath140 , and in the concentration of the main effect of the isospin breaking in the domain of the @xmath74 invariant mass between the @xmath20 thresholds .
we also analyze the difficulties related with the assumption of the dominance of the @xmath147 decay mechanism and discuss the possible dynamics of the decay @xmath148 @xcite .
the decisive improvement of the experimental data on the @xmath20 , @xmath149 , @xmath41 , and @xmath40 mass spectra in the decay of the resonance structure @xmath150 to @xmath151 and @xmath152 , and on the shape of the resonance peaks themselves in the @xmath151 and @xmath153 decay channels is necessary for the further establishing the @xmath154 decay mechanism .
99 n. n. achasov , s. a. devyanin , and g. n. shestakov , phys .
b * 88 * , 367 ( 1979 )
. n. n. achasov , s. a. devyanin , and g. n. shestakov , yad .
fiz . * 33 * , 1337 ( 1981 ) [ sov
. phys . * 33 * , 715 ( 1981 ) ] ; usp .
nauk * 142 * , 361 ( 1984 ) [ sov .
* 27 * , 161 ( 1984 ) ] .
a. r. dzierba , in _ proceedings of the second workshop on physics and detectors for da@xmath155ne95 , frascati , 1995 _ , edited by r. baldini , f. bossi , g. capon , and g. pancheri , frascati physics series vol . 4 ( infn , frascati , 1996 ) , p. 99 . n. n. achasov and g. n. shestakov , phys .
d * 56 * , 212 ( 1997 ) ; yad . fiz . * 60 * , 1669 ( 1997 ) [ phys . at
. nucl . * 60 * , 1522 ( 1997 ) ] .
b. kerbikov and f. tabakin , phys .
c * 62 * , 064601 ( 2000 ) .
f. e. close and a. kirk , phys .
b * 489 * , 24 ( 2000 ) .
v. yu . grishina , l. a. kondratyuk , m. bscher , w. cassing , and h. strher , phys . lett .
b * 521 * , 217 ( 2001 ) .
n. n. achasov and a. v. kiselev , phys .
b * 534 * , 83 ( 2002 ) .
d. black , m. harada , and j. schechter , phys .
lett . * 88 * , 181603 ( 2002 ) .
a. e. kudryavtsev , v. e. tarasov , j. haidenbauer , c. hanhart , and j. speth , phys .
c * 66 * , 015207 ( 2002 ) ; yad .
fiz . * 66 * , 1994 ( 2003 ) [ phys . at
. nucl . * 66 * , 1946 ( 2003 ) ] .
m. buescher , f. p. sassen , n. n. achasov , and l. kondratyuk , arxiv : hep - ph/0301126 . l. a. kondratyuk , e. l. bratkovskaya , v. yu . grishina , m. bscher , w. cassing , and h. strher , yad . fiz . * 66 * , 155 ( 2003 ) [ phys . at . nucl . * 66 * , 152 ( 2003 ) ] . c. hanhart , in _
scalar mesons : an interesting puzzle for qcd _ , edited by a. h. fariborz , aip conf .
688 ( aip , new york , 2003 ) , p. 61
rep . * 397 * , 155 ( 2004 ) . c. amsler and n. a. trnqvist , phys .
rep . * 389 * , 61 ( 2004 ) .
m. buescher , acta phys . pol .
b * 35 * , 1055 ( 2004 ) .
z. g. wang , w. m. yang , and s. l. wan , eur .
j. c * 37 * , 223 ( 2004 ) .
n. n. achasov and g. n. shestakov , phys .
lett . * 92 * , 182001 ( 2004 ) .
n. n. achasov and g. n. shestakov , phys .
d * 70 * , 074015 ( 2004 ) .
j. j. wu , q. zhao , and b. s. zou , phys .
d * 75 * , 114012 ( 2007 ) .
j. j. wu and b. s. zou , phys .
d * 78 * , 074017 ( 2008 ) .
j. j. wu , x. h. liu , q. zhao , and b. s. zou , phys .
* 108 * , 081803 ( 2012 ) .
f. aceti , w. h. liang , e. oset , j. j. wu , and b. s. zou , phys .
d * 86 * , 114007 ( 2012 ) .
x. g. wu , j. j. wu , q. zhao , and b. s. zou , phys .
d * 87 * , 014023 ( 2013 ) .
l. roca , phys .
d * 88 * , 014045 ( 2013 ) .
f. e. close and a. kirk , phys .
d * 91 * , 114015 ( 2015 ) .
t. sekihara and s. kumano , phys .
d * 92 * , 034010 ( 2015 ) .
t. sekihara and s. kumano , j. phys .
soc . jpn .
* 8 * , 022006 ( 2015 ) .
f. aceti , j. m. dias , and e. oset , eur .
j. a * 51 * , 48 ( 2015 ) .
w. wang , phys .
b * 759 * , 501 ( 2016 ) .
v. dorofeev _ et al .
j. a * 38 * , 149 ( 2008 ) .
v. dorofeev _ et al .
j. a * 47 * , 68 ( 2011 ) .
m. ablikim _ et al .
d * 83 * , 032003 ( 2011 ) .
m. ablikim _ et al .
_ , phys . rev
. lett . * 108 * , 182001 ( 2012 ) .
m. ablikim _ et al .
d * 92 * , 012007 ( 2015 ) .
n. n. achasov , a. a. kozhevnikov , and g. n. shestakov , phys .
d * 92 * , 036003 ( 2015 ) .
n. n. achasov , a. a. kozhevnikov , and g. n. shestakov , phys .
d * 93 * , 114027 ( 2016 ) . | it is discussed breaking the isotopic symmetry as the tool of studying the production and nature of light scalar mesons . |
the pseudospin symmetry ( pss ) , manifested by the quasidegeneracy between single - nucleon states with quantum numbers @xmath2 and @xmath3 , was found more than 40 years ago @xcite .
subsequently , substantial efforts had been devoted to understanding the dynamic origin of the pss until it was recognized that the pss is a symmetry hidden in the equations for the small component of dirac spinors .
this might be natural and became clear later on since the relativistic mean - field ( rmf ) models are characteristic of the dynamical description of the spin degree of freedom and spin - orbit interactions @xcite .
the exact pss is a consequence of the fact that the scalar and vector potentials are equal in size but opposite in sign , i.e. , @xmath4 , while in practice the pss is approximate since the condition @xmath5 gives no bound state as in the rmf theory .
however , the exception was found in ref .
@xcite when the confining potential exists . in the schrdinger - like equation for the small component of the dirac spinor
, the term proportional to @xmath6 is related to the breaking of the pss , and the small magnitude of that term gives the approximate pss . for further understanding the properties of the pss
, people have studied the pss in a number of physical processes and phenomena including superdeformation @xcite , nucleon - nucleon and nucleon - nucleus scatterings @xcite , single particle resonant states @xcite , superheavy magic structures @xcite , identical bands @xcite , and pseudospin partner bands @xcite , and so on .
the pss breaking , proportional to the radial gradient of the potential @xmath7 , can vary with the isospin and charge of nuclei through the vector potential entry @xcite . in the past ,
the pss breaking had been studied extensively , and some very useful tools were also developed to analyze the breaking @xcite .
in addition to the isospin effects on the pss breaking , the variation of the isoscalar potentials can also become one pss breaking source , although this variation in different normal nuclei in the nuclear chart is not significant . in this work ,
our attention to the pss breaking is focused on the exotic systems , the kaonic nuclei that may feature a characteristic enhancement of the isoscalar density in the core of nuclei . by analyzing experimental data of kaonic atoms and @xmath0 scatterings ,
people have found that @xmath0 interaction is strongly attractive at saturation density but with an optical potential depth roughly ranging from 40 to 200 mev @xcite .
the great interest in studying kaonic nuclei has been attracted by the fact that the same sign of the vector and scalar potentials of the @xmath8 creates a strong attraction that may lead to high - density @xmath1-nuclear bound states . in the past
, lots of theoretical works have flourished .
for instance , few - body calculations @xcite , the rmf and non - relativistic skyrme - hartree - fock researches @xcite , were performed to obtain the @xmath8 binding energy , the width and ground - state properties of @xmath1-nuclear bound states . along with various predictions ,
experiments for the @xmath8 bound states have also been progressive with the construction of new facilities@xcite .
though the pss is one of fundamental nuclear properties , it has not received due attention in kaonic nuclei .
indeed , the pss has seldom been examined in kaonic nuclei .
thus , it is the aim of this work to investigate how the pss is affected by the strong @xmath8-nucleon attraction in the rmf theory . for completeness and comparison
, we will also examine the corresponding effect on spin - orbit splittings that are associated with the spin symmetry .
the paper is organized as follows . the rmf formalism for kaonic ( @xmath8 )
nuclei is given in section [ rmf ] , and the pseudospin and spin symmetries in rmf are briefly manifested in section [ pss ] .
the results and discussions are given in section [ result ] , followed by a brief summary in section [ summary ] .
the relativistic lagrangian containing @xmath1-nucleon interaction can be written as @xmath9\psi_b \nonumber\\ & & - \frac{1}{4 } f_{\mu \nu } f^{\mu \nu}+\frac{1}{2}m^2_{\omega } \omega_{\mu } \omega^{\mu } - \frac{1}{4 } b_{\mu \nu } b^{\mu \nu}+\frac{1}{2}m^2_{\rho } b_{0\mu } b_0^{\mu}\nonumber \\ & & - \frac{1}{4 } a_{\mu \nu}a^{\mu \nu}+\frac{1}{2}(\partial _ { \mu}\sigma \partial ^{\mu }
\sigma - m^2_{\sigma}\sigma^2 ) - \frac{1}{3}g_2\sigma^3 - \frac{1}{4}g_3\sigma^4 + \mathcal{l}_k,\end{aligned}\ ] ] where @xmath10 , @xmath11 , @xmath12 , and @xmath13 are the fields of the baryon , scalar , vector , and charge - neutral isovector - vector mesons , with their masses @xmath14 , @xmath15 , @xmath16 , @xmath17 , respectively . the @xmath18 is the field of photon .
the @xmath19 are the corresponding meson - baryon coupling constants .
the @xmath20 is the third component of isospin pauli matrix for nucleons .
@xmath21 , @xmath22 , and @xmath23 are the strength tensors of the @xmath24 , @xmath25 meson and the photon , respectively @xmath26 @xmath27 , the lagrangian of the kaonic sector @xcite , is written as @xmath28 where the covariant derivative is given by @xmath29 with the @xmath30 being the corresponding @xmath8-meson coupling constants . here ,
@xmath31 and @xmath32 denote the kaon and antikaon doublet , i.e. @xmath33 and @xmath34 , respectively .
since our investigation is limited to @xmath8 nuclei , the @xmath35 and @xmath36 are left out in the following .
the equations of motion are derived from the lagrangian ( [ lag ] ) . in the rmf approximation ,
the nucleons obey the following equation @xmath37\psi_b = e_b\psi_b . \label{eomd}\ ] ] in spherical systems we consider in this work , the equations of motion for non - strange mesons are given by @xmath38 where @xmath39 , @xmath40 , @xmath41 and @xmath42 are the scalar , vector , proton , and isovector densities , respectively , and we refer readers to ref .
@xcite for detailed expressions .
the @xmath43 denotes the @xmath8 density @xmath44 where the integration of @xmath43 over the whole volume is normalized to the @xmath8 number which is one in this work .
the klein - gordon equation for @xmath8 reads @xmath45 where @xmath46 is the single - particle energy of the @xmath8 .
the real part of the @xmath8 self - energy @xmath47 is written as @xmath48 the imaginary part @xmath49 is considered as the absorptive contribution to the @xmath8 self - energy which can be taken from some optical models phenomenologically . following works of mare
@xcite , we adopt the simple @xmath50 form @xmath51 where the depth @xmath52 is obtained by fitting @xmath8 atomic data @xcite .
the concept of pss is introduced to describe the quasidegeneracy in some nuclei between single - nucleon states with quantum numbers @xmath53 and @xmath3 , e.g. , @xmath54 and @xmath55 .
the pss can be understood as a relativistic symmetry of the dirac hamiltonian originating from the near equality in magnitude of the scalar potential s(r ) and vector potential v(r ) but different in sign , i.e. @xmath56 @xcite .
the quality of the approximate symmetry was found to be associated with the competition between pseudocentrifugal barrier and the pseudospin - orbit potential @xcite . in the following ,
we demonstrate in the rmf the conditions for the pss .
since the meson fields are classical in the rmf , one can write the dirac hamiltonian for nucleons in spherical nuclei as @xmath57 it is well - known that the total angular momentum * j * commutates with the hamiltonian and is a conserved operator .
while the orbital angular momentum @xmath58 does not commute with the _ _ , to characterize the full set of quantum numbers of single - particle states , one invokes another conserved operator * k * @xmath59 with its eigenvalues @xmath60 then we have @xmath61 and @xmath62 . for a spherical system , the quantum number set is @xmath63 where _ n _ is the principle quantum number , _ m _ is the magnetic quantum number , and @xmath64 denotes the isospin . with these quantum numbers ,
the single - particle wave functions can be written as @xmath65 where @xmath66 and @xmath67 are the big and small components of the ( radial ) spinor , respectively , @xmath68 , @xmath69 is the spinor spherical harmonic , and @xmath70 is the isospinor with @xmath71 for protons and neutrons , respectively .
substituting eq .
( [ gf ] ) into the dirac equation ( [ eomd ] ) , one can immediately obtain the radial equations for nucleons @xmath72 where @xmath73 with @xmath74 and @xmath75 . by performing radial derivative on the both sides of eq .
( [ up ] ) and ( [ low ] ) , one arrives at the two second - order differential equations for the big and small components , respectively @xmath76g_a(r ) & = & 0,\\ \label{eqf } [ \frac{d^2}{dr^2}-\frac{1}{u_f}\frac{d\sigma}{dr}\frac{d}{dr } + \frac{1}{u_f}\frac{d\sigma}{dr}\frac{\kappa}{r}-\frac{\kappa(\kappa-1)}{r^2}-u_gu_f ] f_a(r ) & = & 0,\end{aligned}\ ] ] with @xmath77 and @xmath78 . these two equations are equivalent for obtaining the eigenvalues @xmath79 .
similar to the description in ref .
@xcite , the centrifugal barrier ( cb ) and pseudocentrifugal barrier ( pcb ) are here defined as @xmath80 and @xmath81 , respectively . the spin - orbit potential ( sop ) and pseudospin - orbit potential ( psop ) terms are those in eqs.([eqg ] ) and ( [ eqf ] ) proportional to @xmath82 and @xmath83 , respectively .
if the sop term equals exactly to zero , say , @xmath84 or @xmath85 , it can be obviously seen from eq .
( [ eqg ] ) that the eigenvalue @xmath79 only depends on @xmath86 .
one can easily find out that those states with @xmath87 and @xmath88 share the same eigenvalues .
this is nothing but the spin symmetry leading to the spin degeneracy .
similarly , if the psop term equals exactly to zero , say , @xmath89 or @xmath90 , then the eigenvalue @xmath79 only depends on @xmath91 as seen from eq .
( [ eqf ] ) . introducing @xmath92
, we can see that those states with @xmath93 and @xmath94 have the same eigenvalues .
this type of degeneracy was named the pss in resemblance to the spin symmetry . however , the pss is nonexistent once the psop term is far away from zero . generally speaking , as already pointed out in ref .
@xcite , the quality of the pss is tightly associated with the relative magnitude of pcb to the psop in the rmf , while the correlation between psop term and the pss is not simply linear @xcite . in a normal nuclear system ,
the attractive potential is usually around 380mev , while the repulsion is around 320mev .
it gives rise to @xmath95mev , corresponding to relative large sop term , while the @xmath96 is around @xmath97 mev , leading to a relative small psop term .
this estimation indicates that the pss is developed much better than the spin symmetry in normal nuclear single - particle spectra .
in plenty of nuclei , the pss exists approximately and the splitting of pseudospin doublets @xmath53 and @xmath3 is relatively small , compared to the separation between two levels nearby .
but the appearance of @xmath8 could change the proportion of the attraction to the repulsion , leading to the alteration of @xmath98 and @xmath96 .
thus , the general situation for the pss in kaonic nuclei can be rather different and it is worthy of careful investigations .
as one source term of the mean field , the @xmath8 diminishes its role in the mean field with the increase of the nuclear mass . thus , our investigation is limited to the medium and light nuclei in which the @xmath8 has more distinct effects on nuclear properties , especially the bulk density of the nuclear system and the single - particle energy for the pseudospin and spin doublets .
we perform calculations with the nl3 parameter set @xcite .
the @xmath99 and @xmath100 are chosen from su(3 ) relations : @xmath101 , and the @xmath102 is adjusted to yield a @xmath8 binding energy @xmath103=100mev for @xmath104ca where the @xmath103 is defined as the difference of the total binding energy between the kaonic nucleus and its normal counterpart without @xmath8 . though the moderately deep optical potential is adopted here ( @xmath105 mev ) , as in some pioneer works @xcite , we will examine the cases for various binding energies at last .
the coupled equations ( [ eomr]),([kg - k ] ) , ( [ up ] ) and ( [ low ] ) are solved self - consistently by an iterative procedure .
some details of solving the @xmath8 equation are given in the appendix .
c | c c c || c c |c c c & @xmath54 & @xmath55 & @xmath98 ( 2s-1d ) & @xmath106 & @xmath98 ( 1d ) & @xmath107 & @xmath108 & @xmath98 ( 1p ) + @xmath109o & - & - & - & - & - & 21.73 & 15.25 & 6.48 + @xmath110o & - & - & - & - & - & 27.43 & 6.75 & 20.68 + + @xmath111s & 13.95 & 10.45 & 3.50 & 18.59 & 8.14 & 35.85 & 28.82 & 7.03 + @xmath112s & 20.81 & 8.45 & 12.36 & 18.85 & 10.40 & 40.18 & 24.52 & 15.66 + + @xmath113ca & 16.96 & 16.17 & 0.79 & 22.88 & 6.71 & 37.98 & 33.50 & 4.48 + @xmath114ca & 25.86&14.49&11.37&23.19 & 8.70&41.57 & 30.94&10.63 + + @xmath115ca & 17.56 & 17.73&-0.17&23.88&6.15 & 38.94 & 35.63 & 3.31 + @xmath116ca & 21.36 & 16.77&4.59 & 25.00 & 8.23 & 42.61 & 36.48&6.13 + + @xmath117cr & 20.21 & 21.89 & -1.68 & 27.95 & 6.06 & 42.95 & 40.11 & 2.84 + @xmath118cr & 22.95 & 21.44 & 1.51 & 29.52 & 8.08 & 46.80 & 42.12&4.68 + + @xmath119ni & 22.85 & 26.08 & -3.23 & 31.38 & 5.30 & 45.71 & 43.67&2.04 + @xmath120ni & 25.10 & 26.22 & -1.12 & 33.23&7.01&49.59&46.41&3.18 + + @xmath121se & 25.63 & 29.42 & -3.79 & 33.55 & 4.13 & 45.21 & 43.41 & 1.80 + @xmath122se & 28.05&29.78&-1.73 & 34.90 & 5.12 & 47.93 & 45.40 & 2.53 + + @xmath123zr & 30.29 & 32.43 & -2.14 & 36.32 & 3.89 & 48.38 & 46.56 & 1.82 + @xmath124zr & 33.35 & 32.62 & 0.73 & 37.46 & 4.84 & 51.02 & 48.14 & 2.88 + tabulated in table [ spectra ] are the single - neutron binding energies and the splittings of pseudospin and spin doublets for a few states in some spherical nuclei and in the corresponding kaonic nuclei .
those results for protons are not presented here because of no qualitative difference therein . seen from this table , the most noticeable change of the single - particle binding energy in these kaonic nuclei is the considerable increase in the binding energy of the @xmath54 and @xmath107 states while just a little effect on @xmath55 , especially for those light nuclei . for instance , in @xmath113ca , the pseudospin doublets @xmath54 and @xmath55 are very close , showing that the pss is satisfied approximately , whereas the situation is quite different when the @xmath8 is implanted into @xmath113ca .
the big separation between the corresponding doublets arises in @xmath114ca due to a dramatic increase of the binding energy of the @xmath54 state while a small reduction in @xmath55 , leading to the manifest breaking of the pss . besides in @xmath114ca , similar orbital shifts
take place in other kaonic nuclei as seen in table [ spectra ] .
interestingly , with the increase of the nucleon number , the orbital shifts in heavier nuclei @xmath120ni , @xmath122se , and @xmath124zr even favor the approximate pss , being more satisfactory than those in corresponding normal nuclei .
this takes place because the shallower binding of the @xmath54 state in the pseudospin doublet of heavier nuclei accompanies with the smaller binding enhancement of such a state in corresponding kaonic nuclei . on the other hand ,
the spin - orbit splittings between @xmath107 and @xmath108 all get a rise in kaonic nuclei listed in table [ spectra ] .
for instance , the energy interval between @xmath107 and @xmath108 is 4.48 mev for @xmath113ca , while it is 10.63 mev for @xmath114ca .
needless to say , the breaking of the spin symmetry becomes more distinctive in these kaonic nuclei , in comparison to that in normal nuclei .
we mention that the effect of the imaginary part of @xmath8 on the nuclear density and single - particle energies is insignificant for the given @xmath8 binding energy 100 mev , and it just becomes moderate for much weaker binding energies .
thus , we will not regard it specifically in the following discussion .
ca and @xmath114ca as a function of radius .
those labelled `` w / o @xmath8 '' and `` with @xmath8 '' represent for normal nuclei and corresponding kaonic nuclei , respectively , and we will keep these abbreviations throughout.,width=325,height=264 ] now , we take @xmath113ca and @xmath114ca as a typical example to understand the pss features in kaonic nuclei . in fig . [ wf ] , we plot the small components of the dirac spinor , i.e. @xmath67 . as discussed in sec.[pss ] , the similarity in the small components of the dirac spinor is associated with the approximate pss @xcite . indeed , this is the case in @xmath113ca : the @xmath125s for the pseudospin doublets @xmath54 ( dotted curve ) and @xmath55 ( dashed curve ) are rather close , though one has a node and the other without .
but the addition of the @xmath8 in @xmath113ca forces the small component of the @xmath54 state to move inwards considerably . as a result , the breaking of the pss appears with the similarly explicit shift in the small component of the dirac spinor .
while with the increase of the nuclear number , the relative shift of @xmath125 in kaonic nuclei becomes trivial , consistent with the much smaller shift in single - particle binding energy . ) for pseudospin doublets @xmath54 and @xmath55 in @xmath113ca and @xmath121se systems .
also see text . , width=325,height=377 ] and sops @xmath126 ( labelled by orbital quanta ) , see eq .
( [ eqg ] ) , for spin doublets @xmath108 and @xmath107 in @xmath113ca and @xmath121se systems.,width=325,height=377 ] for a deeper understanding of these phenomena concerning the addition of the @xmath8 , we illustrate in fig .
[ potps ] the radial distributions of some relevant potential terms for pseudospin doublets ( @xmath54 and @xmath55 ) in @xmath113ca , @xmath121se and in the corresponding kaonic nuclei .
they are the pcb @xmath127 and the psop @xmath126 , see eq .
( [ eqf ] ) .
because the @xmath8 is trapped in the interior nutshell by the strong attraction and its effect becomes insignificant beyond @xmath128 , only shown in the figure are the results in the region @xmath129 for clarity .
the relative magnitude of the pcb to the psop , closely associated with the quality of the pss , can be used to evaluate the extent of the corresponding symmetry breaking .
it is seen from the upper panel of fig .
[ potps ] that psop is largely enhanced by the @xmath8 in @xmath114ca , especially in the region @xmath130 . due to the large enhancement , the psop and pcb in @xmath114ca are comparable .
it is now not difficult for us to understand why the pss is destroyed in @xmath114ca : the clear enhancement of the psop brings out the importance of the @xmath131 dependence of eigen energies and vectors ( wave functions ) that is the exact factor for the pss breaking , see eq .
( [ eqf ] ) .
we have already seen in table [ spectra ] that the pss in @xmath122se is a little better than that in @xmath121se .
it is again illustrated in terms of the corresponding potentials for @xmath121se and @xmath122se in the lower panel of fig .
[ potps ] that the psop is small compared to the pcb and is much less affected by the @xmath8 , which justifies the approximate pss both in @xmath121se and @xmath122se .
we may further examine the potentials in @xmath121se and @xmath122se in more details .
for @xmath132 , the psop difference between @xmath121se and @xmath122se is not significant , while in the range of @xmath133 , the psop of @xmath121se is clearly larger than that of @xmath122se .
the latter may be responsible for the improvement of the pss in @xmath122se , as this is consistent with the situation in heavier nuclei that more nucleons in exterior shells produce the more radial extension of the @xmath8 by the attraction .
in general , the difference in the psop caused by the @xmath8 embedment becomes small in heavy nuclei , and distinctive features with the embedment of the @xmath8 tend to disappear .
the situation for the shift of spin - orbit splitting is quite analogous .
the attraction provided by the @xmath8 leads to the deepening of the sop , and the @xmath131 dependence of eigenvalues of the spin doublets is magnified . for heavier and heavier nuclei , the @xmath8 effect on the spin - orbit splitting drops off , just like the case of the pss breaking .
these phenomena are clearly shown in fig .
[ pots ] , in resemblance to the pseudospin case in fig .
[ potps ] , and are consistent with those for the spin - orbit splittings listed in table [ spectra ] .
ca ( left panels ) and @xmath121se systems ( right panels ) .
, width=566,height=377 ] the most striking phenomenon of the kaonic nuclei should be the enhancement of the central density @xcite .
the enhancement in light nuclei could be very prominent . here
, we take @xmath113ca and comparably @xmath121se systems as examples . in fig .
[ dens ] , it displays the radial distributions of the densities and potentials . for @xmath113ca systems
, we see that the @xmath8 is deeply bound inside the interior of the kaonic nuclei(@xmath130 ) due to the strong attraction .
this deeply trapped @xmath8 distribution is consistent with the sizeable rise of the sop and psop in the core region , as shown in fig .
[ potps ] and [ pots ] , respectively . as a result ,
the strong attraction provided by the @xmath8 in @xmath113ca pulls nucleons inwards to form a dense core with a density up to twice the saturation density , as shown in the upper left panel of fig .
we recall that the strong attraction responsible for this is produced by the coherent coupling of @xmath11 and @xmath24 mesons to the @xmath8 .
feeding back to the mean field in kaonic nuclei , the attractive potential ( @xmath134 ) in the shrunk core acquires an enhancement greater than the one for the repulsive potential ( @xmath135 ) , as clearly shown in the lower left panel in fig .
the enhancement of the core density also exists in all the kaonic nuclei , while it fades away gradually in heavy nuclei . shown in the right panels of fig .
[ dens ] is the case for @xmath121se systems . as can be seen , the @xmath8 is pulled outwards by the attraction provided by more out - layer nucleons . as a consequence ,
just moderate increase of the core density and mean - field potentials in @xmath122se , instead of the dramatic increase in light nuclei , is observed .
consistently , the shifts of single - particle energy are much smaller than those in light nuclei .
we note that the calculations with the non - relativistic skyrme - hartree - fock approach also found the similar shrinkage @xcite .
these results are inspiring as kaonic nuclei could probably provide a natural cold dense nuclear system rather than a hot dense one that should be created by heavy - ion collisions .
besides the above general analysis for the pss breaking , we have not paid much attention to a specific phenomenon , the almost unilateral shift of the pseudospin doublets .
looking back to fig .
[ wf ] , we see that it is the shift of the 2@xmath136 , rather than that of the 1@xmath137 , which dominates the pss breaking .
this can be roughly understood by the spatial proximity between the @xmath138 @xmath8 and @xmath54 nucleons . to make it clear , we plot the simple product between the neutron wave function square of different states in @xmath113ca and @xmath8 density distribution in @xmath114ca in fig . [ overlap ] .
we see that unlike the @xmath136 orbitals , the @xmath139 orbitals are almost uncorrelated with the @xmath8 occupation .
this illustrates why the @xmath54 state of pseudospin doublets is affected much more dramatically by the @xmath8 .
further , this is associated with the change of the shell structure . shown in fig .
[ spe ] is the single - particle spectra for protons and neutrons in @xmath140ca and @xmath141ca .
it is clearly shown that the @xmath8 implantation leads the migration of @xmath138 and @xmath54 nucleons downwards to the deep fermi sea .
the orbital migration certainly changes the original magic numbers in normal nuclei that are 2 , 8 , 20 ... in @xmath141ca , the magic numbers now become 2 , 6 and 16 that are dictated by the migrated @xmath136 orbitals and enlarged spin - orbit splittings of @xmath142 and @xmath139 orbitals , as shown in fig .
though it is perhaps premature to speak of the new magic numbers because the magic gaps should rely on the depth of the potential well for the @xmath8 , it is definite that the nuclear structure is changed by the @xmath8 implantation .
ca and @xmath8 density distribution in @xmath114ca .
, width=325,height=226 ] ca and @xmath141ca .
the left column in each panel represents energy levels for @xmath140ca , while the right ones are for @xmath141ca , width=325,height=264 ] .
[ spe ] by now , there is no conclusive value of the depth of the @xmath8-nuclear potential .
thus , it is valuable to investigate the dependence of energy splittings of the pseudospin and spin doublets on the @xmath8 binding energy .
specifically , we obtain various splittings of pseudospin and spin doublets by changing @xmath143 and @xmath144 , which is equivalent to varying the @xmath8 binding energy .
results are displayed in fig .
[ ls - b ] for the pseudospin doublets ( @xmath54 and @xmath55 ) , ( @xmath108 and @xmath107 ) and ( @xmath55 and @xmath106 ) . at the point @xmath103=0 , i.e. the normal @xmath113ca , the pseudospin doublet splitting between @xmath54 and @xmath55 is less than 1 mev .
the splitting becomes increasingly large with the enhancement of the @xmath8-nuclear attraction , while a clear increase appears at @xmath145 .
similarly , the splitting of the spin doublets @xmath108 and @xmath107 is apparently amplified by increasing the @xmath8-nuclear attraction .
we see that the effect on the doublet @xmath55 and @xmath106 is relatively smaller .
this is understandable since the out - layer states are less affected by the interior @xmath8 .
nevertheless , larger @xmath8 binding can generally result in more prominent phenomenon of the pseudospin and spin symmetry breakings , especially for interior states .
the above discussion on the association between the doublet splittings and the @xmath8 binding energy also leads our attention to more details of the model dependence .
we find that the splittings of the pseudospin and spin doublets are almost independent of various combinations of different @xmath143 and @xmath144 for a given @xmath8 binding energy . while we change to other models like the nl - sh @xcite , we find that the size of the doublet splittings is quantitatively different not only due to various incompressibility at saturation density but also because of the rather different equations of state at suprasaturation densities . nevertheless , the splittings are still very large for light kaonic nuclei
. moreover , we should point out that all results with the nl - sh are qualitatively similar to those obtained from the nl3 , without changing the conclusions drawn above .
ca as a function of @xmath8 binding energy .
@xmath103=0 represents normal @xmath113ca .
[ ls - b ] , width=325,height=226 ] at last , it is worthy to point out the dependence of the spin or pseudospin doublet splittings on the isospin . with the increase of the neutron number , the repulsion provided by the isovector meson increases .
this causes the moderate reduction of the relevant doublet splittings .
however , the splittings are still well above those for the normal nuclei .
we have investigated the ground - state properties of light kaonic nuclei and the relativistic symmetry breakings with the rmf model . it is found that the relativistic symmetry breakings , underlying potentials associated with the breakings , and the appreciable shrinkage effect , i.e. , the significantly enhanced core density take place consistently due to the strong attraction provided by the @xmath8 embedment in light nuclei . for normal nuclei entertaining the approximate pseudospin symmetry ,
the @xmath8 embedment can enhance @xmath131-dependence of the eigen energies and wave functions , break the original balance between the attractive and repulsive potentials consequently , and result in the pseudospin symmetry breaking and the spin symmetry deterioration . with the @xmath8 embedment , the shell structures are also changed dramatically in light nuclei . in particular
, the migration of @xmath136 orbitals towards the interior fermi sea and the enlarged spin - orbit splittings can reshape the order of energy levels and form new magic numbers .
moreover , we have examined the dependence on the @xmath8 binding energy and the model dependence for these phenomena .
the model dependence is rather weak , and we find that for a large parameter space of the @xmath8 binding energy all the phenomena are quite general .
we thank prof .
xian - hui zhong for useful discussions .
the work was supported in part by the national natural science foundation of china under grant no .
11275048 , the china jiangsu provincial natural science foundation under grant no.bk20131286 , and the china scholarship council .
we give some necessary details for solving the klein - gordon equation for @xmath8 ( eq . [ kg - k ] ) herein . eq .
( [ kg - k ] ) is a complex equation which should firstly be divided into two coupled real equations . in a spherical system , eq .
( [ kg - k ] ) can be written as @xmath146 it is decomposed as @xmath147 with @xmath148 , @xmath149 .
these are second - order differential equations , and can be solved by the runge - kutta method , or the numerov method . in this work ,
the runge - kutta method is adopted .
given the boundary values and a trial eigenvalue @xmath150 , we can integrate those equations from @xmath151 outwards and from @xmath152 ( far away enough ) inwards to a match point @xmath153 , and then connect the wave function at this point . by integrating the integrand that is @xmath154 times
( [ eqk ] ) in the range @xmath155 and @xmath156 ( @xmath157 is a positive infinitesimal ) , we can get @xmath158 however , we must keep in mind that @xmath159 is discontinuous at @xmath153 as long as @xmath150 is still not the eigenvalue .
this discontinuity is used to adjust the eigenvalue in an iterative way @xmath160 with @xmath161 being given as @xmath162 /\int_{0}^{\infty } k_{tr}^{-*}(r)k_{tr}^{-}(r)dr,\end{aligned}\ ] ] where we have disregarded all infinitesimal terms in @xmath157 by considering that the remaining term is from the derivative of the discontinuity which embodies the property of the dirac @xmath163 function .
a. ramos , and e. oset , nucl .
a * 671 * , 481 ( 2000 ) .
j. mare , e. friedman , and a. gal , phys .
b * 606 * , 295 ( 2005 ) .
j. mare , e. friedman , and a. gal , nucl .
a * 770 * , 84 ( 2006 ) .
e. friedman , a. gal , and c. j. batty , phys .
b * 308 * , 6 ( 1993 ) .
e. friedman , a. gal , and c. j. batty , nucl .
a * 579 * , 518 ( 1994 ) .
e. friedman , and a. gal , phys .
rep . * 452 * , 89 ( 2007 ) .
d. gazda , and j. mare , nucl .
a * 881 * , 159 ( 2012 ) .
y. akaishi , and t. yamazaki , phys .
c * 65 * , 044005 ( 2002 ) . a. gal , nucl . phys .
a * 914 * , 270 ( 2013 ) .
t. yamazaki , and y. akaishi , phys .
b * 353 * , 70 ( 2002 ) .
t. suziki et al , phys .
b * 597 * , 263 ( 2004 ) .
m. agnello et al . , phys .
lett . * 94 * , 212303 ( 2005 ) .
m. agnello et al . , phys .
b * 654 * , 80 ( 2007 ) . t. yamazaki et al , phys . rev .
lett . * 104 * , 132502 ( 2010 ) .
e. oset , and h. toki , phys .
c * 74 * , 015207 ( 2006 ) .
d. gazda et al , phys .
c * 77 * , 045206 ( 2008 ) .
b. d. serot , j. d. walecka , adv .
* 16 * , 1 ( 1986 ) .
w. j. jiang , et al , phys .
c * 72 * , 024313 ( 2005 ) .
p. alberto et al , phys .
c * 65 * , 034307 ( 2002 ) g. a. lalazissis , j. konig , and p. ring , phys . rev .
c * 55 * , 540 ( 1997 ) . | as the experimental data from kaonic atoms and @xmath0 scatterings imply that the @xmath1-nucleon interaction is strongly attractive at saturation density , there is a possibility to form @xmath1-nuclear bound states or kaonic nuclei . in this work ,
we investigate the ground - state properties of the light kaonic nuclei with the relativistic mean - field theory .
it is found that the strong attraction between @xmath1 and nucleons reshapes the scalar and vector meson fields , leading to the remarkable enhancement of the nuclear density in the interior of light kaonic nuclei and the manifest shift of the single - nucleon energy spectra and magic numbers therein . as a consequence
, the pseudospin symmetry is shown to be violated together with enlarged spin - orbit splittings in these kaonic nuclei . |
coronal magnetic flux ropes are believed to have close relationship with solar eruptive activities ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , including prominence / filament eruptions , flares , and coronal mass ejections ( cmes ) , which are widely considered as different manifestations of a single physical process ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , corresponding to a sudden destabilization of the coronal magnetic configuration @xcite .
flux ropes can be triggered to erupt by many different mechanisms such as magnetic reconnections and various instabilities ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
it was also suggested by many authors that catastrophe could be responsible for the solar eruptive activities @xcite .
catastrophe occurs by a catastrophic loss of equilibrium as a control parameter of the magnetic system exceeds a critical value @xcite . here
the control parameter characterizes the physical properties of the magnetic conguration .
any parameter can be selected as a control parameter provided that different values of this parameter will result in different configurations @xcite , and different kinds of control parameters correspond to different evolutionary scenarios @xcite . during catastrophe ,
magnetic free energy is quickly released and converted to kinetic and thermal energy @xcite .
catastrophe and instability are intimately related in the evolution of different magnetic systems @xcite .
many solar eruptive activities originate from active regions @xcite , and cartesian coordinates are widely used to investigate active region activities . in cartesian coordinates , both analytical and numerical analyses have been performed to explore catastrophic behaviors of flux ropes in bipolar background field .
if the bipolar field is completely closed , no catastrophe occurs for the flux rope of finite cross section @xcite ; only if the radius of the flux rope is small compared to the length scale of the photospheric magnetic field will there exist a catastrophe @xcite . during the catastrophe , the flux rope , originally attached to the photosphere , loses equilibrium at a critical value of the control parameter and
reaches a new equilibrium levitating in the corona . in a partially open bipolar field
, however , @xcite found that catastrophe also occurs for the flux rope of finite cross section .
these studies imply that there are two branches of equilibrium states of a flux rope with catastrophe : the lower branch and the upper branch .
the flux rope sticks to the solar surface for solutions at the lower branch and levitates in the corona for those at the upper branch , with a vertical current sheet below it .
the catastrophe mentioned above corresponds to a jump from the lower branch to the upper , and thus is called `` upward catastrophe '' hereinafter .
all the previous studies only analysed the upward catastrophe .
some questions are still standing in theory : is there a catastrophe during which the flux rope falls back from the upper branch to the lower ( called `` downward catastrophe '' hereinafter ) ? will it also release magnetic free energy ? to answer these questions , we follow the work by @xcite to study the equilibrium solutions , but change to force - free fields .
the motivations of using a force - free field structure rather than a magnetostatic structure are as follows .
first , the strong magnetic fields over active regions are usually considered to be force - free @xcite .
second , under force - free conditions , the system is dominated by magnetic fields , and its energy is limited to magnetic energy so as to substantially simplify the energy analysis ( see section [ sec : energy ] ) . by analysing the evolution of the equilibrium solutions versus the control parameters in cartesian coordinates ,
the properties of the catastrophes in partially open bipolar background field under the force - free approximation are investigated .
we mainly focus on the existence of the downward catastrophe , and the evolutions of the magnetic energy during the catastrophes .
the sections are arranged as follows : simulation methods are introduced in section [ sec : equation ] ; two kinds of catastrophes are demonstrated in section [ sec : catastrophe ] ; the variations of magnetic energy during upward and downward catastrophes are analysed in section [ sec : energy ] ; by summarizing the simulation results , the whole evolution of a flux rope in partially open bipolar field is illustrated in section [ sec : profile ] ; the mechanism by which magnetic energy is released is analysed in section [ sec : work ] ; the significance of catastrophe in both observational and theoretical analyses are discussed in section [ sec : discuss ] .
a cartesian coordinate system is taken and the magnetic flux function is used here to denote the magnetic fields as follows : @xmath0 and the 2.5-d mhd equations can be written in the non - dimensional form : @xmath1 where @xmath2 correspond to the density , velocity , temperature and magnetic flux function , respectively ; @xmath3 and @xmath4 denote the z - component of the magnetic field and the velocity , parallel to the axis of the flux rope ; @xmath5 is the normalized gravity , @xmath6 is the characteristic ratio of the gas pressure to the magnetic pressure , where @xmath7 is the vacuum magnetic permeability , and @xmath8 is the gas constant ; @xmath9 , @xmath10 , @xmath11 , and @xmath12 are the characteristic values of density , temperature , length and magnetic flux function , respectively .
the radiation and heat conduction in the energy equation have been neglected .
the initial corona is isothermal and static with @xmath13 the background field , assumed to be symmetrical relative to the @xmath14-axis , is a partially open bipolar magnetic field , which is calculated by complex variable method ( see * ? ? ?
. a positive and a negative surface magnetic charges are located at @xmath15 within @xmath16 and @xmath17 , respectively .
the computational domain is taken to be @xmath18 , @xmath19 , with symmetrical conditions used for the left side ( @xmath20 ) .
the lower boundary @xmath15 corresponds to the photosphere . during the simulation , top ( @xmath21 ) and right ( @xmath22 )
boundary conditions are fixed .
fixed boundary conditions greatly help to maintain stability of the simulation .
we have tried the simulations with computational domains of different size ( e.g. @xmath23 ) , and found that the size of the domain does not influence the existence of the catastrophes . although the critical values of the control parameters at which catastrophe takes place vary slightly with the domain size , the deviations are smaller than 10% .
this indicates that the computational domain used here is large enough so that the influence of the boundary condition on the evolution of the magnetic system is almost negligible .
starting from the background field , with the same procedures as in @xcite , a magnetic flux rope emerges from the central area of the base , resulting in an equilibrium state with a flux rope sticking to the photosphere , i.e. a lower - branch solution , which is the initial state of our simulation ( see fig .
1(a ) in @xcite ) . with the initial and boundary conditions ,
equations [ equ : cal - st ] to [ equ : cal - en ] are solved by the multi - step implicit scheme @xcite to let the system evolve to equilibrium states .
relaxation method was used by @xcite to obtain force - free equilibrium solutions in spherical coordinates .
this method resets the temperature and density in the computational domain to their initial values , so that the pressure gradient force is always balanced everywhere by the gravitational force .
using the same method in cartesian coordinates , we obtain force - free equilibrium solutions .
moreover , during the whole simulation , both numerical and physical magnetic reconnections are prohibited . in theory
, @xmath24 is constant along a current sheet .
any reconnection will reduce the value of @xmath24 at the reconnection site .
therefore , by reassigning @xmath24 with the initial value at each time step along the entire current sheet ( if it exists ) , we keep @xmath24 invariant , so that reconnections are completely prevented across the current sheet .
in this paper , we select the axial magnetic flux passing through the cross section of the flux rope , @xmath25 , and the annular magnetic flux of the rope per unit length along @xmath26-direction , @xmath27 , which is simply the difference of @xmath24 between the rope axis and the outer border of the rope , as the control parameters .
the varying @xmath25 and @xmath27 represent the evolutionary scenarios , e.g. flux emergence and twisting / untwisting motions .
if not changed manually , @xmath25 and @xmath27 of the rope should be maintained to be conserved , which is achieved by special numerical measures : following similar procedures proposed by @xcite , a slight adjustment of @xmath3 inside the flux rope and @xmath24 at the rope axis is taken at the end of each time step .
the evolution of the equilibrium solutions is described by the geometric parameters of the flux rope , including the height of the rope axis , @xmath28 , and the length of the current sheet below the rope , @xmath29 .
note that at the lower branch , @xmath30 , and at the upper branch , @xmath29 is finite .
starting from the initial state , equilibrium solutions with different control parameters are calculated in the following way : from 0 to @xmath31 , @xmath25 and @xmath27 are linearly changed from the same initial values to target values ( @xmath32 , @xmath33 ) , and from @xmath31 to @xmath34 , the two fluxes are kept invarant till the magnetic system reaches an equilibrium state . here
@xmath35 is the typical alfvn transit time .
the final state of this calculation at @xmath36 is regarded as the equilibrium solution with ( @xmath32 , @xmath33 ) .
similar calculations are repeated for different target values of @xmath25 and @xmath27 to obtain different equilibrium solutions .
first , in order to analyse the evolution of the equilibrium solutions versus @xmath25 , we focus on the solutions with different @xmath25 but the same @xmath37 .
the geometric parameters describing these equilibrium solutions are plotted by the red dots in fig .
[ fig : pf](a ) and [ fig : pf](c ) .
the flux rope keeps sticking to the photosphere for @xmath38 ( see fig .
[ fig : conf](a ) ) and the equilibrium solutions remain at the lower branch .
once the control parameter reaches @xmath39 , however , the flux rope jumps to the equilibrium state at the upper branch , levitating in the corona , as shown in fig .
[ fig : conf](b ) .
this indicates that a catastrophe with the flux rope jumping from the lower branch to the upper branch takes place , which is the upward catastrophe analysed in previous studies , and the critical value of the control parameter @xmath25 is called the upward catastrophic point : @xmath40 .
our simulation reveals that the upward catastrophe also occurs in partially open bipolar field under force - free approximations . as mentioned above
, all the previous studies only analysed the upward catastrophe . here
are the questions to be considered : what will happen to the flux rope at the upper branch with decreasing @xmath25 ?
will the flux rope fall back to the lower branch ?
if so , will the flux rope fall back by a catastrophic jump or a continuous transition ? to solve these problems ,
we start from the equilibrium solution at the upper branch obtained above ( fig .
[ fig : conf](b ) ) and decrease @xmath25 to calculate the equilibrium solutions at the upper branch with similar procedures , i.e. from 0 to @xmath31 decrease @xmath25 to a certain value and from @xmath31 to @xmath34 the system relaxes to equilibrium state with the given @xmath25 .
@xmath27 is still fixed to be @xmath41 . as shown by the blue dots in fig .
[ fig : pf](a ) and [ fig : pf](c ) , with decreasing @xmath25 , the flux rope does not fall back to the lower branch at @xmath42 , indicating that the upward catastrophe is an irreversible process .
the simulation result reveals that the flux rope also keeps suspended in the corona , and the associated equilibrium solutions remain at the upper branch , till @xmath43 ( fig .
[ fig : conf](c ) ) , at which the flux rope suddenly falls back to the photosphere ( fig .
[ fig : conf](d ) ) .
this indicates that there exists another kind of catastrophe , during which the flux rope falls down from the upper branch to the lower branch , i.e. a downward catastrophe .
the downward catastrophic point is @xmath44 .
furthermore , we increase @xmath25 again , but from the lower - branch solution right after the downward catastrophe ( fig . [
fig : conf](d ) ) , to calculate equilibrium solutions with varying @xmath25 , as shown by the black triangles in fig .
[ fig : pf](a ) and [ fig : pf](c ) .
similarly , the flux rope does not jump to the upper branch at @xmath45 as well , indicating that the downward catastrophe is also an irreversible process . from fig .
[ fig : pf](a ) and [ fig : pf](c ) , it can be seen that the equilibrium solutions calculated from the solution right after the downward catastrophe ( black triangles ) follow almost the same profile as the red dots do .
when @xmath25 reaches the catastrophic point , an upward catastrophe takes place as before .
consequently , a circulation is formed by these two different kinds of catastrophes ( see section [ sec : profile ] ) .
the values of the catastrophic points are close to the observed magnetic fluxes of cmes ( e.g. * ? ? ?
there are also upward and downward catastrophes as the annular magnetic flux @xmath27 is taken to be the control parameter with fixed @xmath46 . by similar procedures mentioned above , the evolution of equilibrium solutions versus @xmath27 is obtained , as plotted in fig .
[ fig : pf](b ) and [ fig : pf](d ) .
the meanings of the symbols are the same : red dots denote the process of searching for the upward catastrophe , blue for the downward catastrophe , and black triangles are the equilibrium states calculated from the state after the downward catastrophe has taken place .
both upward and downward catastrophes exist and belong to irreversible processes .
the upward catastrophic point is @xmath47 , and the downward catastrophic point is @xmath48 .
as pointed in section [ sec : catastrophe ] , there exist upward and downward catastrophes ; they are both irreversible processes , but the flux rope moves in opposite directions . whether magnetic energy is released or stored during the downward catastrophe will help to understand the nature of these two catastrophes .
since the magnetic energy per unit length along the @xmath26-axis is infinite for a partially open bipolar field in cartesian coordinates , the evolution of magnetic energy has never been touched upon before . as an expedient measure
, we evaluate the magnetic energy in the computational domain only , and think it approximately represents the evolutionary tendency of the magnetic energy of the whole system .
as long as the domain is sufficiently large in size , as we did and explained in section [ sec : equation ] , the conclusions obtained should be qualitatively correct .
the magnetic energy per unit length in @xmath26-direction in the computational domain is calculated by : @xmath49 the evolutions of the magnetic energy within the domain are plotted in fig .
[ fig : pf](e ) and fig .
[ fig : pf](f ) with @xmath25 and @xmath27 as the control parameter , respectively .
magnetic energy is released during both of the two catastrophes .
it is natural that magnetic energy is always released during catastrophes : any catastrophe belongs to a spontaneous process in the absence of external influences , and the system must evolve to a state of lower energy after catastrophe .
we may conclude that , although different kinds of catastrophes could have different kinematic characters ( e.g. in this paper the moving directions of the flux rope are opposite during the upward and downward catastrophes ) , catastrophe should always correspond to a sudden energy release . by assuming that spatial scale in @xmath26-direction is the same as @xmath50-direction and @xmath14-direction , i.e. @xmath51
, the released energy of the catastrophes is estimated to be : @xmath52 j for the upward catastrophe and @xmath53 j for the downward catastrophe with @xmath25 as the control parameter ; @xmath54 j for the upward and @xmath55 j for the downward with @xmath27 as the control parameter .
more magnetic energy is released for the upward catastrophe than the downward .
it is revealed that the released magnetic energy during the catastrophes is of the order of @xmath56 j. it should be noted that the amount of the released energy is inversely proportional to the value of @xmath57 , which is set to be 0.1 in this study .
larger amount of released magnetic energy can be obtained with smaller @xmath57 .
for example , if @xmath58 , which is more proper for coronal situation @xcite , the released energy should be as large as @xmath59 j , which is comparable to the released energy of a medium flare . the evolution of a flux rope in a partially open bipolar background field as discussed in section [ sec : catastrophe ] and section [ sec : energy ] is summarised in fig .
[ fig : cart ] , where @xmath60 stands for the geometric parameter and @xmath61 the control parameter .
there are two branches of equilibrium states for the flux rope : the lower branch ( @xmath62 in fig .
[ fig : cart ] ) and the upper branch ( @xmath63 in fig .
[ fig : cart ] ) .
the two branches are separated by the upward ( from lower to upper branch ) and downward ( from upper to lower branch ) catastrophes .
the upward ( downward ) catastrophe is an irreversible but reproducible process , and it takes place when the solution jumps from the lower ( upper ) branch to the upper ( lower ) branch with increasing ( decreasing ) control parameter .
therefore , the two branches and the two catastrophes form a circulation , as shown by @xmath64 in fig .
[ fig : cart](a ) . the evolution of magnetic energy ( fig .
[ fig : cart](b ) ) is also a circulation consisting of the two branches and two catastrophes , but it is quite different from that of the geometric parameter ( fig . [
fig : cart](a ) ) . at the lower branch ( @xmath65 ) ,
magnetic energy is stored with increasing control parameter , whereas it is reduced during both upward and downward catastrophes , and during the shift of the solution along the upper branch from @xmath66 to @xmath67 as well .
the constraint exerted by the background field on the flux rope is larger for solutions at the lower branch than that at the upper branch , so that the growth rate of the magnetic energy with @xmath61 , @xmath68 , is larger for lower - branch solutions than upper - branch solutions .
consequently , at the upward catastrophic point ( @xmath69 ) , the magnetic energy of the upper - branch solution is smaller than that of the lower branch solution ( @xmath70 ) , whereas at the downward catastrophic point ( @xmath71 ) , the opposite is true ( @xmath72 ) .
this provides the essential energy condition under which the downward catastrophe could occur .
it is widely accepted that magnetic energy is released by magnetic reconnection in solar eruptive activities @xcite .
however , magnetic reconnections are prohibited during our simulation ( see section [ sec : equation ] ) .
there should be other mechanisms of magnetic energy release in addition to magnetic reconnection . by simulations in
spherical coordinates , @xcite found that the magnetic force acting on the flux rope vanishes for equilibrium states , but becomes upward if the rope erupts after the upward catastrophe . as a result
, the flux rope is expected to be accelerated by the ampre s force . here
, we calculate the work done by ampre s force during catastrophe in the entire computational domain , so as to reveal the mechanism of magnetic energy release .
[ fig : up ] shows the temporal profiles of relevant geometric and physical parameters during a transition toward the equilibrium solutions right before and after the upward catastrophe , which correspond to @xmath73 and @xmath74 , respectively .
the first four panels show the geometric parameters , including @xmath28 and @xmath29 , whereas panels ( e ) and ( f ) show the magnetic energy .
finally , panels ( g ) and ( h ) depict the rate of doing work by ampre s force , which is determined by : @xmath75 \cdot\overrightarrow{v}dxdy,\end{aligned}\ ] ] where @xmath76 and @xmath77 denote the current density and electric field . from 0 to @xmath31
, @xmath25 and @xmath27 are smoothly adjusted .
note that at the beginning ( 0@xmath7840@xmath79 ) , @xmath80 is small but finite .
this should result from the deviation of the system from equilibrium at the beginning caused by adjusting the control parameters . from @xmath31 to @xmath34 , during which @xmath25 and @xmath27 is constant , the magnetic energy is almost constant if the upward catastrophe does not take place ( fig . [
fig : up](e ) ) , and @xmath80 is almost zero accordingly ( fig .
[ fig : up](g ) ) .
if the upward catastrophe takes place , however , not only magnetic energy is released ( fig .
[ fig : up](f ) ) , but also a significant peak value of @xmath80 appears in fig .
[ fig : up](h ) . the total work done by ampre s force during upward catastrophe
is roughly estimated to be about @xmath81 , in the same order of the released magnetic energy ( @xmath82 ) .
the deviation might result from the poynting energy flow and the influence of relaxation method .
therefore , we may conclude that the magnetic energy is released primarily by the work done by ampre s force during the upward catastrophe . for downward catastrophe
, although the deviation of the work done by ampre s force from the released energy is larger , basic conclusion remains the same as for the upward catastrophe : the work done by ampre s force appears as a positive peak during the downward catastrophe .
to investigate the catastrophic behavior of coronal flux ropes , we simulate the evolution of the equilibrium states associated with a flux rope in a force - free partially open bipolar field versus different control parameters .
it is found that , under force - free approximation , there is also upward catastrophe .
what s more , under the force - free approximation , there exists another possibility that a downward catastrophe takes place , during which a levitating flux rope may fall down to the photosphere . the evolutionary scenario represented by the control parameter might cause a catastrophe to take place .
for example , the `` flux - feeding '' procedure , during which chromospheric fibrils rise upward and merge with the prominence above @xcite , will result in varying @xmath25 , and varying @xmath27 could be caused by twisting or untwisting motions of the flux rope @xcite .
all these phenomena are the possible triggers of catastrophes .
upward and downward catastrophes connect the two branches of equilibrium states so as to form a cycle .
these two catastrophes are both irreversible but reproducible processes . thus there
might exist activities during which more than one catastrophe take place : e.g. a flux rope is suspended in the corona at first , then falls back to the photosphere , and at last jumps upward , resulting in an eruptive activity . by calculating the magnetic energy within the numerical domain ,
the evolution of magnetic energy is analysed semi - quantitatively .
although the moving directions of the rope are opposite for the upward and downward catastrophes , magnetic energy is always released .
the order of the released energy is rather large , comparable to a medium flare ( see section [ sec : energy ] ) .
since there are no magnetic reconnections , magnetic energy is mainly released by the work done by ampre s force .
our calculation demonstrates that the magnetic energy released by the work done by ampre s force during catastrophe is sufficient for solar eruptive activities , indicating that magnetic reconnection is not always necessary .
if magnetic reconnection is included in the simulation , the eruptive speed can be signicantly enhanced @xcite , and magnetic energy is released by both magnetic reconnection and the work done by ampre s force .
previous studies have proposed that an upward catastrophe can serve as an effective mechanism for solar eruptive activities .
since catastrophe occurs via a loss of equilibrium at a critical point , it can be triggered by very small disturbances .
in addition , upward catastrophes can not only account for cmes and flares , they also provide sites for fast reconnection @xcite .
apart from eruptive activities , there are also energetic but non - eruptive activities , such as confined flares @xcite .
the physical mechanism of confined flares has been discussed in many previous studies ( e.g. * ? ? ?
* ; * ? ? ?
during a downward catastrophe , although magnetic energy is still released , the flux rope falls back to the photosphere .
therefore , a downward catastrophe might be another possible cause of energetic but non - eruptive activities .
observational evidences are still needed to confirm these conjectures .
this research is supported by grants from nsfc 41131065 , 41574165 , 41421063 and 41222031 , moec 20113402110001 , cas key research program kzzd - ew-01 - 4 , and the fundamental research funds for the central universities wk2080000077 . | 2.5d time - dependent ideal magnetohydrodynamic ( mhd ) models in cartesian coordinates were used in previous studies to seek mhd equilibria involving a magnetic flux rope embedded in a bipolar , partially open background field . as demonstrated by these studies ,
the equilibrium solutions of the system are separated into two branches : the flux rope sticks to the photosphere for solutions at the lower branch but is suspended in the corona for those at the upper branch . moreover , a solution originally at the lower branch jumps to the upper , as the related control parameter increases and reaches a critical value , and the associated jump is here referred to as upward catastrophe .
the present paper advances these studies in three aspects .
first , the magnetic field is changed to be force - free .
the system still experiences an upward catastrophe with an increase in each control parameter .
secondly , under the force - free approximation , there also exists a downward catastrophe , characterized by a jump of a solution from the upper branch to the lower .
both catastrophes are irreversible processes connecting the two branches of equilibrium solutions so as to form a cycle . finally , the magnetic energy in the numerical domain is calculated .
it is found that there exists a magnetic energy release for both catastrophes .
the ampre s force , which vanishes everywhere for force - free fields , appears only during the catastrophes and does a positive work , which serves as a major mechanism for the energy release .
the implications of the downward catastrophe and its relevance to solar activities are briefly discussed . |
the dyakonov wave is an electromagnetic surface wave guided by the planar interface of two homogenous dielectric materials , at least one of which is anisotropic .
the existence of the dyakonov wave was predicted in the 1980s @xcite and experimentally verified about five years ago @xcite .
the tamm wave is guided by the planar interface of two isotropic dielectric materials , at least one of which is periodically nonhomogeneous in the direction normal to the interface . predicted in 1977 @xcite , tamm waves
were experimentally observed shortly thereafter @xcite and are being applied for optical sensing of analytes @xcite . whereas the anisotropy of a partnering material provides the dyakonov wave a sensitivity to the direction of propagation @xcite ,
the periodic nonhomogeneity of a partnering material is responsible for multiple tamm waves to propagate in any direction @xcite .
both of these attributes are combined in dyakonov
tamm waves , whose propagation is guided by the interface of two dielectric materials , one of which must be anisotropic and periodically nonhomogeneous normal to the interface @xcite .
recently , the existence of these dyakonov
tamm waves was confirmed experimentally using a prism - coupled configuration @xcite . in the foregoing papers on dyakonov tamm waves ,
one partnering material is an ordinary , isotropic , homogeneous , dielectric material with a purely real relative permittivity @xmath0 , whereas the real part of the relative permittivity dyadic @xmath1 of the anisotropic , periodically nonhomogeneous , dielectric partnernering material is positive definite , i.e. , all three eigenvalues of @xmath1 have positive real parts @xcite . what would happen if @xmath2 were indefinite , i.e. , if either one or two of its eigenvalues had negative real parts , the remaining having positive real parts ?
such materials are nowadays called hyperbolic materials .
homogeneous hyperbolic materials exist in nature @xcite and have also been manufactured @xcite .
periodically nonhomogeneous hyperbolic materials appear very likely to be manufacturable @xcite using physical - vapor - deposition techniques @xcite . in this letter , we present the results of our investigations on dyakonov
tamm waves guided by the planar interface of an ordinary , isotropic , homogenous , dielectric material and a hyperbolic , dielectric , structurally chiral material which is nonhomogeneous in the direction normal to the interface .
section [ theory ] briefly describes the relevant one - point boundary - value problem .
numerical results are provided and discussed in sec .
[ nrd ] . an @xmath3 dependence on time
@xmath4 is implicit , with @xmath5 and @xmath6 denoting the angular frequency .
the one - point boundary - value problem of dyakonov tamm - wave propagation is as follows : suppose that the plane @xmath7 is the interface between the two chosen partnering materials .
the half space @xmath8 is occupied by an ordinary , isotropic , homogenous , dielectric material with relative permittivity @xmath9 such that @xmath10 and @xmath11 .
the half space @xmath12 is occupied by a hyperbolic , dielectric , structurally chiral material with relative permittivity dyadic @xcite @xmath13 where the direction of nonhomogeneity is parallel to the @xmath14 axis ; the cartesian unit vectors are identified as @xmath15 , @xmath16 , and @xmath17 ; and @xmath18 is the permittivity of free space ; the periodic nonhomogeneity is expressed through the rotation dyadic @xmath19 with @xmath20 as the period and either @xmath21 for structural right - handedness or @xmath22 for structural left - handedness ; the dyadic @xmath23 contains the tilt angle @xmath24 $ ] with respect to the @xmath25 plane ; @xmath26 , @xmath27 , and @xmath28 are the three @xmath14-independent eigenvalues of @xmath29 ; and either one or two of these three eigenvalues have negative real parts but the remainder do not . the electromagnetic field phasors everywhere can be written as @xcite @xmath30 \#h(\#r)=\#h(z ) \exp\les{i}q(x\cos\psi+y\sin\psi)\ris \end{array}\ric\,,\ ] ] with the unknown surface wavenumber @xmath31 and unknown functions @xmath32 and @xmath33 .
the angle @xmath34 denotes the direction of propagation in the @xmath25 plane .
as the procedure to obtain and solve a dispersion equation for @xmath31 , and then determine @xmath32 and @xmath33 , for a specific @xmath35 has been described elsewhere in detail @xcite , it is not repeated here .
for illustrative results , we set @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath21 , @xmath40 nm , @xmath41 , and @xmath42 . without loss of generality , we took the isotropic partner to be free space ( i.e. , @xmath43 ) flp .
our search was mostly restricted to the regime @xmath44 $ ] , where @xmath45 is the free - space wavenumber . at every free - space wavelength @xmath46 $ ] nm
, we found three solutions @xmath31 of the dispersion equation .
the solutions were organized in three branches , as shown in fig .
thus , three distinct dyakonov
tamm waves differing in phase speed @xmath47 and propagation length @xmath48 can propagate along the @xmath49 axis in the interface plane for @xmath46 $ ] nm .
tamm waves on the highest-@xmath50 branch have @xmath51 , where @xmath52 is the speed of light in free space , and @xmath53 ranging from @xmath54 nm to @xmath55 nm .
tamm waves on the lowest-@xmath50 branch have @xmath50 ranging from @xmath56 to @xmath57 , and @xmath58 @xmath59 m .
clearly from the presented data , a lower phase speed is associated with higher attenuation along the direction of propagation .
the dyakonov
tamm waves on the three branches in fig .
[ fig1 ] are not only dissimilar in phase speed and propagation length , but the spatial profiles of their field along the @xmath14 axis also differ .
let us focus on the three solutions of the dispersion equation for @xmath60 nm .
figure [ fig2 ] presents plots of the magnitudes of the cartesian components of @xmath32 and @xmath33 as functions of @xmath14 for the dyakonov tamm wave with @xmath61 , fig .
[ fig3 ] for the dyakonov
tamm wave with @xmath62 , and fig .
[ fig4 ] for the dyakonov
tamm wave with @xmath63 . on the isotropic side , the degree of localization of the dyakonov
tamm wave to the interface is greater if @xmath64 is higher .
however , on the anisotropic side of the interface , a clear dependence of localization to the interface does not emerge from the three figures .
indeed , of the three dyakonov tamm waves , the one in fig .
[ fig4 ] is the most strongly localized to the interface , whereas the one in fig .
[ fig3 ] is the most weakly localized to the interface .
the dyakonov tamm - wave - propagation phenomenon depends strongly on @xmath35 .
this is exemplied by the plots in fig .
[ fig5 ] of @xmath50 and @xmath53 as functions of @xmath65 $ ] when @xmath60 nm and @xmath66 .
the solutions can be organized in six branches .
one branch , with the highest values of @xmath50 and @xmath53 spans the entire angular regime @xmath65 $ ] available for propagation .
each of the remaining five branches spans a finite range of @xmath35 .
although the five branches appear to be clustered together , the phase speeds range from @xmath67 to @xmath68 and the propagation lengths from @xmath69 @xmath59 m to @xmath70 @xmath59 m , which are quite wide ranges .
either one or two or three dyakonov
tamm waves can propagate in the interface plane at an angle @xmath35 with respect to the @xmath49 axis .
let us also note that , because of chiral symmetry about the @xmath14 axis , the solutions of the dispersion equation are the same for @xmath35 and @xmath71 .
the dyakonov tamm - wave - propagation phenomenon also depends strongly on @xmath9 .
suppose that free space , the isotropic partnering material for figs .
[ fig2][fig4 ] , were to be replaced by a dielectric material with @xmath72 .
then the dispersion equation has just only two solutions@xmath73 and @xmath74instead of the three solutions for @xmath43 .
the large changes between the solutions for the two values of @xmath9 chosen here strongly suggest the potential of dyakonov
tamm waves supported by a partnering material that is hyperbolic , dielectric , and structurally chiral for optical sensing of analytes in a prism - coupled configuration , just as both surface - plasmon - polariton waves @xcite and tamm waves @xcite are being used by technoscientists .
* acknowledgments . * a.l . is grateful to the charles godfrey binder endowment at the pennsylvania state university for the financial support of his ongoing research .
m.f . was supported by grant no .
dmr-1125591 from the u.s .
national science foundation . and propagation length @xmath53 as functions of @xmath75 , when @xmath43 , @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath21 , @xmath40 nm , @xmath41 , and @xmath42 .
the solutions of the dispersion equation are organized in three branches .
, width=377 ] of the magnitudes of the cartesian components of @xmath32 and @xmath33 of an spp wave when @xmath60 nm , @xmath43 , @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath21 , @xmath40 nm , @xmath41 , and @xmath42 .
the @xmath49 , @xmath76 , and @xmath14-components are represented by red solid , blue dashed , and black chain - dashed lines , respectively . for this dyakonov
tamm wave , @xmath61 .
, width=377 ] and propagation length @xmath53 as functions of @xmath35 , when @xmath60 nm , @xmath43 , @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath21 , @xmath40 nm , and @xmath66 . the solutions of the dispersion equation are organized in six branches.,width=377 ] | the planar interface of an ordinary , isotropic , homogeneous , dielectric material and a hyperbolic , dielectric , structurally chiral material can support the propagation of one or multiple dyakonov
tamm waves , at a specified frequency and along a specified direction in the interface plane . when multiple dyakonov
tamm waves can exist , they differ in phase speed , propagation length , degree of localization to the interface , and spatial profiles of the associated electromagnetic fields .
dependence on the relative permittivity scalar of the isotropic partnering material suggests exploitation for optical sensing of analytes .
* dyakonov tamm waves guided jointly by an ordinary , isotropic , homogeneous , dielectric material and a hyperbolic , dielectric , structurally chiral material * + akhlesh lakhtakia and muhammad faryad + pennsylvania state university , department of engineering science and mechanics , nanoengineered metamaterials group ( nanomm ) , university park , pa 168026812 , usa + |
the hypothesis that the characteristic energy scale of quantum gravity lies just beyond the electroweak scale finds its motivation in the hierarchy problem and its operative realization in the existence of extra spatial dimensions @xcite .
if this hypothesis is verified in nature , future experiments at the lhc have the ground - breaking opportunity of testing the dynamics of gravitation in the high - energy and quantum regimes .
unfortunately , since quantum - gravity dynamics are still unknown , systematic theoretical predictions can not be made .
it is then important to identify certain collider topologies and kinematical regions that allow a description based only upon general principles and not on model peculiarities , and which can be used in the experimental program as a handle to demonstrate the gravitational nature of the interaction .
single graviton emissions satisfy these requirements and allow for a model - independent study of the propagation of gravitational forces in extra spatial dimensions @xcite . on the other hand , model - dependent experimental signals , as those originating from contact interactions @xcite or ,
more likely , the observation of a specific spectrum of string regge excitations @xcite will be very important to discriminate among various models and to guide theoretical research .
the model - independent experimental signals discussed in refs .
@xcite are based upon interactions at energies below the quantum gravity scale @xmath5 ( the analogue of the planck mass for a @xmath6-dimensional theory ) , and their theoretical description fails as we approach @xmath5 .
however , there is another kinematical regime that can be tackled in a fairly model - independent fashion : the transplanckian region @xmath7 .
this is the subject of this paper .
a pecularity of hadron machines is that , because of the composite structure of the proton , one can study , in the same environment , parton collisions at different center - of - mass energies .
consequently , if @xmath8 experiments at the lhc could probe the cisplanckian ( graviton emission , contact interactions ) , the planckian ( the core of the quantum - gravity dynamics ) , and the transplanckian regions . to understand the nature of transplanckian collisions , we first identify the relevant scales ( see ref .
let us consider general relativity in @xmath9 dimensions with ( generalized ) newton s constant @xmath10 . for @xmath11 flat dimensions compactified in a volume @xmath12
, @xmath10 is related to the usual newton s constant by @xmath13 .
to better elucidate the relation between the transplanckian and the classical limit , we reinstate for a moment the correct powers of @xmath14 and @xmath15 .
the relation between @xmath10 and the @xmath6-dimensional planck scale is g_d= , where the proportionality constant follows from the conventions of ref .
notice that @xmath10 has dimensions @xmath16=\ell^{n+5}e^{-1}t^{-4}$ ] ( @xmath17 length , @xmath18 energy , @xmath19 time ) .
starting from @xmath10 , we can construct the planck length _
p = ( ) ^. [ lamp ] this is the distance below which quantum gravity effects become important . using the center - of - mass energy of the collision
, we can construct the length _
this is the de broglie wavelength of the colliding particles , which characterizes their quantum length scale . combining @xmath10 and @xmath20
, we can form the radius of a system with center - of - mass energy @xmath20
@xcite r_s = ^ ( ) ^. this is the length at which curvature effects become significant . in the limit
@xmath21 , with @xmath10 and @xmath22 fixed , @xmath23 vanishes , showing that classical physics correspond to transplanckian ( macroscopically large ) energies .
moreover , in the same limit , @xmath24 remains finite , while the two length scales @xmath25 and @xmath26 go to zero .
therefore , the transplanckian regime corresponds to a classical limit in which the length scale @xmath24 characterizes the dynamics , m_d r_s_p _ b. for instance , by simple dimensional analysis and analiticity in @xmath10 , we expect that the scattering angle for a collision with impact parameter @xmath27 is @xmath28 .
this behavior shows that by increasing the energy we can obtain a finite angle scattering by going to larger @xmath27 , and therefore further suppressing short distance quantum gravity effects .
this property of gravity should be contrasted to what happens in an ordinary gauge theory , where spin-1 particles mediate the force . in @xmath29dimensional space - time
the gauge coupling @xmath30 has dimension @xmath31=\ell^{n+1 } e$ ] .
the analogues of @xmath23 , @xmath25 and @xmath24 are given respectively by @xmath32 , @xmath33 and @xmath34 . in the @xmath21 limit , with @xmath30 and @xmath22 fixed ,
we find that @xmath35 ( we focus for simplicity on @xmath36 ) indicating that quantum fluctutations are not suppressed at any finite length . indeed
the same dimensional argument we used before gives a classical scattering angle @xmath37 , which remains finite as @xmath38 provided @xmath39 .
but @xmath40 is the regime where quantum effects are unsuppressed , so we conclude that there is no classical limit .
the different properties of spin-2 and spin-1 exchange that are elucidated by this simple dimensional analysis are the same that render the eikonal approximation consistent for the first and inconsistent for the second .
basically it is because energy itself plays the role of charge in gravity : when the energy is large , gravity gives sizeable effects also at large distance where quantum fluctuations of the geometry are irrelevant .
having established that scattering at transplanckian energies is described by classical physics , one also realizes that the corresponding amplitude should only be calculable by a non - perturbative approach .
this is always the case for semiclassical amplitudes ( like instantons for example ) , for which the classical action @xmath41 is much larger than @xmath14 , and we can not perturbatively expand @xmath42 . in order to tackle the non - perturbative problem
, we will further restrict the range of distance scales under investigation , and only consider collisions with impact parameter @xmath27 much larger than @xmath24 . in this regime ,
the curvature is small , the metric is nearly flat , and we can work in the limit of weak gravitational field , neglecting non - linear effects of the graviton couplings .
the non - perturbative interactions between the high - energy colliding partons and the weak gravitational field can be computed using the eikonal approximation , which can be trusted at small scattering angle . from a field - theoretical point of view
, this approximation corresponds @xcite to a resummation of an infinite set of feynman diagrams which , at each order in perturbation theory , give the leading contributions to the forward scattering .
the eikonal approximation has been employed for 4-dimensional gravity @xcite and for string theory @xcite ( see also ref .
the results for the scattering amplitude agree with those obtained @xcite by solving the klein - gordon equation for a particle propagating in the classical gravitational shock wave produced by the other particle @xcite , or by reformulating quantum gravity as a topological field theory @xcite .
all these different approaches give equivalent results @xcite . in sect .
[ sec2 ] we will derive and discuss the scattering amplitude in the eikonal approximation for 4-dimensional colliding particles ( living on a 3-brane ) and @xmath6-dimensional gravity @xcite ( see also ref .
@xcite ) . as explained above
, the process is essentially classical , like the motion of two ( relativistic ) stars coming within a close distance .
however , an important difference is that the gravitational potential in the non - relativistic analogue of our process decreases with the distance @xmath43 as @xmath44 . as discussed in sect .
[ sec2 ] this will have significant consequences , and it will lead to quite different results from the case of 4-dimensional newtonian potentials . a novel feature is the emergence of a new length scale b_c ( ) ^. notice that @xmath45 can not be defined in 4 dimensions ( @xmath46 ) . in the limit @xmath21 , we find that @xmath45 goes to infinity and therefore the classical region extends only up to length scales of order @xmath45 . for impact parameters larger than @xmath45 ,
the process is no longer determined by classical properties . however , since quantum gravity effects are always negligible , ordinary relativistic quantum mechanics is sufficient to compute scattering amplitudes within our approximation .
as previously discussed , we are limited by our computational ability to work at @xmath47 . as the impact parameter approaches the radius ( and
our expressions become unreliable ) , we expect to enter a ( classical ) regime in which the gravitational field is strong , non - linear effects are important and , plausibly , black holes are formed .
this regime , discussed in refs .
@xcite and @xcite ( see also ref .
@xcite ) , is very exciting from the experimental point of view , but more arduous from a theoretical point of view .
while in the eikonal region we are able to make quantitatively reliable predictions with controllable expansion parameters , the black - hole production cross section can only be estimated with dimensional arguments and more precise information should wait for numerical simulations in @xmath6-dimensional general relativity .
black - hole production results in a black disk of radius @xmath24 in the cross section , since any initial state with @xmath48 is absorbed by the black hole ( this point of view has been criticized in ref .
collisions with an impact parameter @xmath27 smaller than the radius do not directly probe physics at the energy scale @xmath49 .
ultraviolet dynamics and quantum effects are screened by the black hole .
the physical picture we have been describing holds if there is a clear separation of the various length scales and if no other new dynamics intervene at distances between @xmath24 and @xmath45 . as we will discuss in this paper ,
this is not necessarily the case for particle collisions at the lhc .
given the present lower limits on @xmath23 , the center - of - mass energy at the lhc can be only marginally in the transplanckian region , and a clear separation between the quantum - gravity scale @xmath25 and the classical scale @xmath24 is not completely achieved , endangering our approximations .
the problem is particularly apparent in string theory , which contains a new scale , the string length _
s = , where @xmath50 is the string tension . in weakly - coupled string theory ,
@xmath51 is larger than @xmath25 and string effects can set in and modify gravity in an intermediate regime between the eikonal elastic scattering and black - hole formation @xcite .
these are interesting effects , as the lhc can directly probe the nature of string theory @xcite . however , model - independent calculability of the elastic channel is lost . in this paper
we will consider only the pure transplanckian gravitational scattering , assuming that quantum - gravity effects are small .
the relevance of our considerations for experiments at the lhc and other future colliders is discussed in sect .
we will compute the elastic cross section in the transplanckian eikonal regime , m_d/1 , -t / s1 , [ pit ] where @xmath52 is the lorentz - invariant mandelstam variable , and @xmath53 is a measure of the scattering angle in the center - of - mass frame .
the experimental signal will be discussed and compared with the expected standard model ( sm ) rate and with black - hole production .
we are interested in the elastic scattering of two massless 4-dimensional particles ( living on a 3-brane ) due to a @xmath6-dimensional weak gravitational field in the kinematic limit of .
we will focus on the case in which the two colliding particles are different , but the case of identical particles is analogous . in the eikonal approximation
, we consider the infinite set of ladder and cross - ladder feynman diagrams which give the leading contribution to forward scattering .
since we are working in the approximation of small momentum transfer , we can compute the coupling between the colliding particle @xmath54 and the small fluctuation around the flat metric ( the graviton ) , taking @xmath54 to be on - shell : ( p)|t^|(p)= .
[ tmu ] here @xmath55 is the energy - momentum tensor and @xmath56 is the 4-momentum of the particle @xmath54 .
notice that is equally valid for colliding particles of any spin . in the @xmath54 propagators
we keep only the leading terms in the momentum transfer @xmath57 : & & for bosons , + & & for fermions .
[ ferm ] since the factor @xmath58 in is used to reconstruct the on - shell vertices of in the ladder diagrams , we observe that not only the interaction vertices but also the propagators of bosons and fermions are identical in the eikonal approximation .
this universality of bosons and fermions is not surprising since , as discussed in the introduction , we are performing a semi - classical calculation , and the spin of the @xmath54 particle should not matter .
the tree - level exchange of a @xmath6-dimensional graviton in the diagram of fig .
[ fig1]a gives the scattering amplitude _ born ( -t)= = ^ ( 1-n/2 ) ( ) ^ -1 ( ) ^2 , [ born ] where @xmath59 is the momentum transfer in the extra dimensions .
although is a tree - level amplitude , it is divergent due to the infinite number of ( extra - dimensional ) momentum configurations of the exchanged gravitons , allowed by the non - conservation of momenta transverse to the brane .
divergences have been subtracted using dimensional regularization , considering non - integer @xmath11 , but the regularization prescription is not important since the eikonalization will consistently eliminate any ultraviolet sensitivity .
basically this is because the eikonalization selects in only the partial waves with large angular momentum .
these do not depend on the local counter - terms as they are determined by the finite ( calculable ) non - analytic terms in @xmath60 , which depend only on infrared physics @xcite .
these terms are proportional to @xmath61 for even @xmath11 , and to @xmath62 for odd @xmath11 .
let us now turn to the one - loop amplitude .
the leading term in the limit of small momentum transfer is given by a feynman diagram with a cut arising from a physical - region singularity in which the intermediate @xmath54 particles are on their mass shells .
there is only one such diagram at one loop , and it is drawn in fig .
all other diagrams are subleading .
for instance , the one - loop @xmath52-channel exchange of two gravitons @xmath63 coupled via quartic @xmath64 vertices is suppressed by a power of @xmath65 . diagrams with one @xmath64 vertex and two @xmath66 vertices simply vanish ( as is evident by working in the de donder gauge ) . the diagram of fig .
[ fig1]b with on - shell intermediate states can be calculated using the cutkosky rule .
its imaginary part is 1/2 times the discontinuity across the cut , which is obtained by replacing the @xmath54 propagators in with @xmath67 . if @xmath68 and @xmath69 are the momenta of the incoming particles , we obtain _ 1-loop ( -q^2 ) & = & ( ) ^2 d^n k_t d^n k_t^ ( 2i)^2 ( 2p_1k)(2p_2k ) + & = & _ born(k_^2 ) a_born[(q_-k_)^2 ] .
[ 1loop ] the delta functions from the cutkosky rule had the effect of reducing the 4-dimensional integral into a 2-dimensional integral over momenta perpendicular to the beam ( and along the brane ) .
notice that , in the eikonal limit , the momentum transfer @xmath57 is mainly in the perpendicular direction and therefore , we can use the approximation t = q^2-q_^2 . equation ( [ 1loop ] ) is merely a convolution of the born amplitude .
it is then convenient to perform a fourier transform of the amplitudes with respect to the transverse momentum @xmath70 .
this amounts to trading @xmath70 with its conjugate variable @xmath71 , the impact parameter . in impact parameter space
becomes a simple product .
summing and , and transforming back to momentum space we can recast the loop expansion in the form _
born(q_^2)+a_1-loop ( q_^2)+ =-2is d^2b_e^iq_b _ ( i- ^2 + ) , [ insie ] ( b_)= e^-iq_b_a_born(q_^2 ) .
[ eikp ] the combinatorics of higher - order loop terms is such that one can resum all terms in to obtain _ eik = -2is d^2b_e^iq_b _ ( e^i -1 ) .
[ eicon ] we have recovered the known result @xcite that the eikonal scattering phase @xmath72 , function of the impact parameter plane coordinates @xmath71 , is the 2-dimensional fourier transform of the born amplitude in the direction perpendicular to the beam .
one crucial feature of the eikonal procedure is that @xmath73 for @xmath74 depends only on the calculable , ultraviolet finite , terms in @xmath75 .
the ultraviolet divergent terms correspond to @xmath76-function singularities localized at @xmath77 .
while these terms are obviously irrelevant for scattering processes where the initial states are prepared in such a way that they have no overlap at @xmath77 , we will also define our full amplitude in by neglecting these terms .
we believe this to be a consistent procedure . indeed , it would not make any sense to include @xmath78-functions in the exponent of . from a physical point of view
, we expect these singularities to be softened by the fundamental theory of gravity , at some finite , but small , impact parameter @xmath79 ( or , more likely , @xmath80 ) .
so the presence of these terms is just evidence that there should be corrections to our simple eikonal expression coming from the underlying theory of quantum gravity .
we will discuss these effects in sect .
[ seccor ] . for the moment
we limit ourselves to note that these short - distance contributions should plausibly give rise to @xmath81 corrections to the cross section , while , as we will see shortly , the long - distance eikonal amplitude gives a cross section that grows with a power of @xmath22 , thus dominating at large energies .
we can now calculate explicitly the eikonal phase @xmath72 in .
one finds ( see appendix ) = _ 0^dq
q^n-1 j_0(qb ) =( ) ^n , [ bbc ] b_c^1/n .
[ defbc ] here @xmath82 , @xmath83 and @xmath84 is the zero - th order bessel function .
notice that by inserting this result for @xmath72 in the integral in we obtain an ultraviolet finite result .
this is so , even though the contributions to from the individual terms in the expansion of @xmath85 are ultraviolet divergent , corresponding to the fact that each individual feynman diagram in the ladder expansion is ultraviolet divergent but the complete sum is finite . moreover , since @xmath86 , the integrand in oscillates very rapidly as @xmath39 , showing that the ultraviolet region gives but a small contribution to the amplitude . replacing into
, we obtain the final expression for the eikonal amplitude @xcite _ eik = 4s b_c^2 f_n ( b_c q ) , [ eikfin ] f_n ( y)=-i_0^dx x j_0 ( xy ) ( e^ix^-n-1 ) , [ funn ] where the integration variable is @xmath87 .
the functions @xmath88 , which can be written in terms of meijer s g - functions , are shown in fig .
[ fig2 ] . at small @xmath89 , the functions @xmath90 can be expanded as ( see appendix ) f_2(y)&&- + i + f_3(y ) & & ( ) e^i -y + f_4(y ) & & i e^i+y + f_n(y ) & & ( 1- ) e^-i -y^2 ( 1- ) e^-i for n>4 .
[ singg ] these functions develop non - analytic terms in @xmath91 of the form @xmath92 for even @xmath11 , and @xmath93 for odd @xmath11 , which reproduce the non - analytic structure of @xmath94 previously mentioned .
notice that @xmath95 is finite for @xmath96
. for @xmath97 ( @xmath98 ) , the stationary - phase approximation applies , yielding f_n(y ) .
[ statp ] the function @xmath88 oscillates around its asymptotic value given in , before converging to it at large @xmath89 , as illustrated in the bottom - right panel of fig .
[ fig2 ] .
let us now consider the differential cross section for elastic scattering .
from we obtain = b_c^4 | f_n ( b_c)|^2 .
[ sigpr ] we recall the definitions r_s^2s = a_n ( ) ^ , b_c^2s = b_n ( ) ^ , [ defab ] where the coefficients @xmath99 and @xmath100 are given in table 1 . in terms of the geometric black - hole cross section @xmath101 , can be written as & = & ( b_c^2 s ) ^
| f_n ( b_c)|^2 + & = & ( ) ^ | f_n ( b_c)|^2 .
[ tab1 ] .the coefficients @xmath99 and @xmath100 entering and their numerical values for @xmath11 between 2 and 6 . [ cols="^,^,^,^,^,^,^",options="header " , ] in the limit @xmath102 , or equivalently @xmath103 , the differential cross section becomes ( for @xmath96 ) & = & ( b_c^2 s ) ^ + & = & ( ) ^ .
[ diff1 ] in the opposite limit ( @xmath98 ) , we find = ( -)^. [ diff2 ] notice how the two formul valid in the two regimes match since , for @xmath104 , parametrically agrees with .
moreover , extrapolating beyond its range of validity towards small impact parameters ( _ i.e. _ large angle , or @xmath105 ) , we find a result that matches parametrically the black - hole production cross section derived from dimensional analysis .
the integrated elastic cross can be derived directly from : _
el&= & d^2q_| a_eik |^2 + & = & d^2 b _
( 1+e^-2im -2e^-im ) .
[ sigtot ] the total cross section , inclusive of elastic and inelastic channels , can be derived from the optical theorem _ tot = = 2 d^2 b _ ( 1-e^-im ) .
[ sigop ] in the absence of absorptive parts ( @xmath106 ) and coincide .
black - hole production may be modelled by a large absorptive part ( @xmath107 ) at impact parameters smaller than @xmath24 . before proceeding to the phenomenological analysis , we remark on the physical meaning of the expression of the eikonal amplitude given by and .
we can first consider a non - relativistic analogue of the problem under investigation @xcite by studying the scattering of a particle with mass @xmath108 , velocity @xmath109 , and impact parameter @xmath27 by a radial `` gravitational '' potential v(r)=. here @xmath108 and @xmath110 are the non - relativistic analogues of @xmath111 , @xmath109 is the analogue of @xmath15 , and @xmath11 corresponds to the number of extra dimensions . when @xmath27 is not too small , the quantum - mechanical scattering phase of the wave associated with angular momentum @xmath112 is _ b = -
_ b^dr v(r ) .
[ delsc ] the intergral in converges only for @xmath36 , leading to @xmath113 , where @xmath114^{1/n}$ ] exactly corresponds to the definition in .
therefore , the forward amplitude ( which is proportional to @xmath115 ) is finite only for @xmath96 . for @xmath46
one finds the coulomb singularity , and for @xmath116 a logarithmic divergence .
this is in agreement with and with the relativistic results presented in sect .
[ sec21 ] .
a classical description of the process is valid if the quantum - mechanical uncertainties in the impact parameter @xmath27 and in the scattering angle @xmath117 are small with respect to their classical values . using the heisenberg principle ,
the quantum uncertainties are estimated to be ~ ~ , where @xmath57 is the momentum in the direction orthogonal to the initial velocity @xmath118 .
the classical scattering angle is equal to the force ( @xmath119 ) at the minimum distance ( @xmath120 ) , times the collision time ( @xmath121 ) , divided by the momentum ( @xmath122 ) , ~ .
the condition for the validity of the classical approximation ( @xmath123 , @xmath124 ) implies @xmath125 . for a coulomb - like potential @xmath126 ,
this condition is satisfied in the non - perturbative region @xmath127 . in our case , it implies ( for @xmath36 ) @xmath128 , with @xmath129^{1/n}$ ] .
the length scale @xmath45 is a quantum scale ( @xmath130 as @xmath21 ) characteristic of the extra - dimensional scattering ( it is undefined at @xmath46 ) .
therefore , at @xmath131 , quantum mechanical effects can not be neglected . in the classical process ( for @xmath132 ) , a projectile with an arbitrarily large impact parameter @xmath27 is deflected by a non - vanishing ( although very small ) angle , and therefore the cross section diverges at @xmath133 ( @xmath134 ) . in our case , for @xmath131 , quantum mechanics sets in , and the quantum fluctuations of the scattering angle become larger than its classical value , rendering ill - defined the notion of a trajectory .
this is why the quantum scattering amplitude can remain finite at @xmath135 .
it is also interesting to compare the result of with the fully @xmath6-dimensional case , in which both gravity and the colliding particles propagate in the extra dimensions @xcite . in this case
becomes _ eik = -2is d^2+nb_e^iq_b _ .
[ eqsuper ] for @xmath133 , the integral is infrared dominated by large values of @xmath27 and it is quadratically divergent ( for any @xmath11 ) .
therefore we find @xmath136 , and we encounter the coulomb singularity characteristic of long - range forces . on the other hand ,
in the case of particles living on the 3-brane , we have found ( for @xmath96 ) that the forward scattering amplitude is finite and that , although the force does not have a finite range , the integral in at @xmath133 is dominated by distances of order @xmath45 . indeed , by working with localized particles we lose momentum conservation in the transverse direction , and moreover the born amplitude becomes dominated by contact terms of ultraviolet origin .
the fact that is not infrared dominated even for @xmath135 is a remnant of this property of the born amplitude .
let us now come back to the relativistic elastic scattering under consideration .
as we have seen in sect .
[ sec21 ] , there exist two different kinematic regions in which the eikonal amplitude has distinct behavior
. region _
( i ) _ : @xmath137 .
+ this is the region in which the integral in is dominated by the stationary - phase value of the impact parameter @xmath138 , where b_sb_c ( ) ^ = ( - ) ^- ( ) ^. [ defbs ] in this region @xmath128 , and therefore the eikonal phase @xmath139 is large ( in units of @xmath14 ) : we are in the classical domain . indeed , the classical scattering angle is given by the derivative of the scattering phase with respect to the angular momentum @xmath140 , _
cl=-= . [ einsa ] in the limit @xmath141 , we recover the einstein angle @xmath142 while , for @xmath36 , gives its higher - dimensional generalization particle at rest : @xmath143 .
it is easy to see that by boosting @xmath144 to the center - of - mass frame , in the limit @xmath145 , we obtain our expression of @xmath146 . ] .
notice that the relation between scattering angle and impact parameter in is equivalent to @xmath138 .
( ii ) _ : @xmath147 .
+ in this region the integral in is dominated by @xmath27 of the order of ( or slightly smaller than ) @xmath45 .
this means that the eikonal phase @xmath139 is of order one ( in units of @xmath14 ) and the quantum nature of the scattering particles is important ( although the exchanged graviton is treated as a classical field , and so quantum gravity effects are negligible ) , as we could have expected from the discussion of the non - relativistic analogue presented above .
there is no one - to - one relation between impact parameter and scattering angle in region _
( ii ) _ , since the classical concept of a trajectory is not applicable . moreover ,
notice that the relevant @xmath72 never becomes much smaller than 1 , and therefore we never enter the perturbative regime in which a loop expansion for the amplitude applies . even though the interaction vanishes at @xmath134 ( where @xmath148 ) , we never reach the born limit . even for @xmath133 , the scattering is dominated by @xmath149 and not by @xmath150 , as in the coulomb case .
as explained after , this result simply follows from the different dimensionalities of the spaces relevant for scattering particles ( living on a 3-brane ) and exchanged graviton ( propagating in the bulk ) , and it does not hold in the case of scattering of bulk particles . the conditions required by our approximations are expressed in . in order to assess the theoretical uncertainty inherent to our calculation , in this section we estimate the size of the effects that we have neglected .
we first discuss classical effects , for which general relativity suffices .
we also include the quantum - mechanical effects of region _
( ii ) _ , but we leave quantum - gravity contributions for the next section .
the neglected classical terms come from the approximations of small angle ( @xmath151 ) and of weak gravitational field ( @xmath152 ) . associated with the subleading classical effects
there is also the emission of gravitational radiation , leading to missing energy signals at high - energy colliders .
let us estimate these effects .
notice that in the classical region _
( i ) _ where @xmath153 , the two requirements coincide since @xmath138 , see . here
, scattering can be described in terms of classical trajectories .
the equation @xmath138 represents a relation between transferred momentum and impact parameter -~~ ( ) ^2n+2 , [ leading ] where we have neglected factors of order unity .
this result represents an approximation @xcite in which one considers only a linear superposition of the gravitational shockwave fields generated by the two colliding particles @xcite , thereby neglecting non - linear effects of their mutual interactions . in this computation
, each field represents an exact solution to einstein s equations in the absence of the other field .
inclusion of the non - linear effects will induce @xmath154 corrections to , where @xmath155 is an unknown power . by dimensional
analysis the relative size of these corrections must be @xmath156 . moreover since we expect that the evaluation of these effects is perfectly perturbative , analyticity in @xmath157 should hold , and therefore only even integers of the exponent @xmath155 will appear . we conclude that will be corrected by a factor of the form 1+o ( ) = 1+o ( ) . [ subleading ] this will induce a correction of the same size to the cross section , in the classical region _ ( i)_. notice that corresponds to a @xmath158 correction term to the relation between scattering angle and impact parameter , . in the scattering by a static potential
we would get corrections already at @xmath159 .
our different result follows from the relativistic nature of the process . in region _
( ii ) _ where @xmath102 , the leading correction to the cross section will then have the form = , [ fullcorr ] where @xmath160 is expected to be a constant of order unity in the classical region @xmath161 . in the region
@xmath162 , the quantum nature of the scattering particles becomes relevant , and @xmath160 will have non - trivial behavior .
however , admitting only finite corrections at @xmath163 implies that the size of the relative correction to @xmath164 is at most @xmath165 . in summary , we have inferred that the leading classical gravity corrections to the eikonal amplitude are ( - ) + o , [ correx ] and therefore suppressed and under control for the conditions we are applying , as expressed by .
the discussion outlined above is consistent with the analysis of ref .
@xcite where classical corrections to the eikonal are identified by resumming an improved ladder series including a class of two - loop graphs , the so - called h - diagrams . after improving the born term by the h - diagram
, the eikonal phase is modified to ( ) ^n .
[ quarantaq ] it is easy to check that altering according to the above equation produces corrections to the leading amplitude consistent with .
notice that the corrections to the eikonal become large for impact parameters comparable to the schwarzschild radius , for which the production of black holes presumably sets in . ] .
gravitational radiation is associated with subleading classical effects .
emission of gravitons is signaled by the presence of an imaginary part in the h - diagram contribution to the eikonal phase @xcite ( _ h ) & ~&()^3n+2 , [ imclass ] + b_r & & ( b_c^n r_s^2n+2)^~ ( g_d^3 s^2 ) ^. this absorptive term corresponds to a depletion of purely elastic scattering due to the emission of gravitational radiation .
when @xmath166 the probability of elastic scattering becomes exponentially small .
the calculation of ref .
@xcite shows that the emitted radiation has typical transverse momentum @xmath167 .
however the longitudinal momentum is distributed up to values of order @xmath22 .
therefore we expect that for @xmath168 a significant fraction of the energy is radiated in the form of forward gravitational radiation .
notice that @xmath168 is typically outside our chosen kinematical regime , so that we expect only a fraction @xmath169 of initial energy to be lost to invisible gravitational radiation . in the semiclassical region ,
we find @xmath170 , and therefore gravitational radiation for small - angle scatterings is not large
. one may worry that important emission of radiation , already at @xmath171 , may drastically reduce the naive geometric estimate of the black - hole production cross - section . indeed if one simply interpreted , _ la _ block - nordsiek , as the number @xmath172 of emitted gravitons , one would conclude that hard bremsstrahlung depletes all the energy , as soon as @xmath27 is smaller than @xmath173 .
however this naive interpretation is manifestly inconsistent as the total emitted energy would quickly exceed @xmath22 .
indeed it is likely that the exponentiation of the h - diagram imaginary part is inconsistent for hard radiation . on the other hand for soft radiation with @xmath174 ,
the block - nordsiek interpretation is likely to be reasonable .
thus if the number of soft gravitons is @xmath175 then the total energy lost to soft radiation at large angle is e_rad~~s ( ) ^3n+3 , which becomes important only for @xmath176 . on general grounds we expect that quantum - gravity contributions to the eikonal phase must be of the form @xmath177 , with @xmath178 and @xmath25 given in .
if we work in the true transplanckian region ( @xmath179 ) , these effects are subleading ( we take @xmath180 ) .
notice that , as opposed to gravity , there is no factor of @xmath157 enhancement @xcite .
it is easy to realize that quantum corrections will also go like a power of @xmath181 .
this is for instance the case for vacuum - polarization improvement of the vector propagator .
therefore , there is no impact parameter choice for which both @xmath182 and quantum corrections are small , where the eikonal approximation can be useful . ]
since , as we have discussed in sect . 1
, the classical scale @xmath24 is much larger than the quantum scale @xmath25 .
however , as it will be clear in the following , collisions at the lhc can only barely satisfy the transplanckianity condition .
therefore , quantum - gravity corrections are potentially large , and they are likely to be the limiting factor of our approximations .
this happens because , although quantum - gravity effects are characterized by a length scale @xmath25 smaller than the radius @xmath24 , classical corrections appear with a large exponent , see eqs .
( [ subleading ] ) and ( [ correx ] ) , remnant of the geometric nature of the interaction , while this is not expected to be the case for the quantum - gravity corrections @xmath183 .
however in order to be able to say more we need a quantum - gravity model or at least a framework .
string theory is the only possibility at hand , and the transplanckian regime has been discussed in this context by several authors , see _
e.g. _ refs .
indeed , we will show that string corrections appear to be large for energies relevant to lhc experiments . in ref .
@xcite string corrections to eikonalized graviton scattering for type ii superstrings were studied .
since we do not stick to a particular string realization of our brane - world scenario , we will limit ourselves to a qualitative discussion .
following ref .
@xcite we may imagine a realization of the brane - world in type i , where @xmath184 spatial dimensions out of the 10 space - time dimensions are compactified on a torus @xmath185 with radius @xmath186 , and @xmath187 is the string scale .
of the remaining dimensions , @xmath11 have a `` large '' compactification radius and 3 are the usual non - compact ones .
the brane - world is realized by a @xmath188-dimensional d - brane , in which @xmath108 dimensions span the small manifold @xmath185 . in terms of @xmath187 and of the string
coupling @xmath189 , the @xmath190 dimensional planck scale @xmath23 and the gauge coupling @xmath191 on the brane are given by m_d^2+n = m_s^2+n,= , g_s = e^ , [ ancora ] where @xmath192 is the dilaton vacuum expectation value .
of course these relations will receive corrections of order unity in realistic models with broken supersymmetry and should be viewed , here and in the following , only as indicative estimates
. equations ( [ ancora ] ) imply ( ) ^2+n=. if we identify @xmath191 with the gauge couplings of the sm , say for instance @xmath193 , we conclude that there should be a mild hierarchy between @xmath23 and @xmath187 .
for instance , we find @xmath194 equals 1.7 and 2.8 for @xmath195 and @xmath116 , respectively . _ a priori
_ one could also conceive string models where @xmath23 and @xmath187 coincide , with a self - dual dilaton vacuum expectation value @xmath196 , corresponding to a strongly coupled string . in this case
the weakness of the sm gauge couplings would have to be explained by some additional mechanism . as we will discuss in sect .
[ sec3 ] , in order to observe transplanckian scatterings at the lhc , we have to consider values of @xmath23 smaller than a few tev .
the existence of string theory at such a low scale is already limited by precision electroweak data and by the non - observation of new contact interactions at lep .
exchanges of gravitons and massive string states give rise to effective operators involving sm fields .
the leading effects come from 4-fermion operators of dimension 6 which , if present at tree level , give an approximate bound @xmath197 tev @xcite .
this bound is in conflict with the working assumptions necessary to have observable transplanckian signals at the lhc , and therefore we have to assume a mild suppression of the dimension-6 operators .
this assumption is probably not unreasonable , as there exist examples @xcite where these effects vanish at tree level . in this respect
the presence of supersymmetry at some stage in the construction may help suppress these contributions .
loop effects are within current experimental bounds even for @xmath23 of order a few tev , if the underlying quantum - gravity theory does not become strongly - interacting or , in a more operative sense , if the divergent loops are cut off at a scale slightly smaller than @xmath23 @xcite .
this allows for the weakly - interacting string scenario in which there is a small hierarchy between @xmath187 and @xmath23 , but it essentially rules out the case of strongly - coupled strings with @xmath187 of a few tev , where we would need the rather strong , and implausible , assumption that dimension-6 operators are suppressed to all orders in perturbation theory . with these limitations in mind
we now consider string corrections @xcite .
the basic result is that the eikonal elastic field - theory amplitude gets promoted to a unitary matrix acting non - trivially over a subspace of the string fock space & & k , [ diffractive ] + & & k = , b_d^n+2 b_c^n . here
@xmath198 is a hermitian operator involving the creation and annihilation operators for the string oscillator modes .
the above equation shows that , for @xmath199 , a diffractive production of excited string modes takes place without any suppression .
basically what happens is that the two colliding partons are excited into string modes through multiple soft graviton exchange .
the average mass of the string modes is roughly given by @xmath200 .
this is smaller than the available energy @xmath22 as long as @xmath201 .
this condition is clearly satisfied within our chosen kinematical regime , and therefore the produced particles are relativistic .
the opening of an ( exponentially ) large number of new channels drastically depletes the elastic one .
for instance in the example discussed in ref .
@xcite , where gravitons are scattered in type ii superstring , the elastic amplitude correction factor becomes k=^2n+2(1-i ) ^2[1-i(n+1 ) ] , [ elastdiff ] where @xmath202 and where by @xmath203 we indicate the expectation value of the operator @xmath204 on the initial state . for @xmath205 ,
the correction factor in becomes k1+i 2 n(n+1 ) ( ) ^n+2 , [ corrinf ] where @xmath206 is the euler number .
therefore , the effects of string excitations set in much before the string scale . already at @xmath207 , the terms modifying the gravitational eikonal phase implied by are as large as @xmath208 .
as we have previously anticipated , these effects can be much more sizable than the classical corrections in , even if the string length is smaller than the classical radius . for @xmath209
, we find |k| ( n+1 ) ( n)^n+2 ( ) ^(n+2)^2 , showing that the absolute value of the correction factor decreases exponentially at @xmath199 , because of the appearence of the string excitations .
this string effect is fairly general and we expect it in realistic models not to differ too much from .
however the sharp drop in the elastic cross section can be compensated by looking at the inclusive diffractive cross section . since @xmath210 we can use the saddle point approximation to evaluate the amplitude in transverse momentum space ( q)= a_eik(q ) , [ unitary ] where @xmath211 is the eikonal amplitude in the stationary phase regime . due to unitarity of the second factor ,
the total cross - section obtained by summing on the diffractive channels is just the same as in the absence of strings , _
i.e. _ there is no exponential suppression .
notice that this inclusive cross section is consistently defined in the limit in which the string modes are relativistic : at a fixed @xmath57 all modes come out at essentially the same angle @xmath117 .
the question of how well one can measure experimentally the inclusive cross section depends very much on the model , and in particular on the decay properties of the produced string modes .
these states are excited by gravitons , so that they have the same color and electric charge of the incoming partons .
if they are stable , the final state of the collision is given by two ( massive ) jet events .
if they are unstable and their dominant decay mode involves bulk states , like the graviton , then the presence of missing energy would make it impossible to deduce the original center - of - mass energy and to make sure that only events at a given center - of - mass energy are selected . on the other hand
if they dominantly decay on the brane this reconstruction might be possible .
notice that the coupling to bulk modes ( closed strings ) and to brane modes ( open strings ) are proportional to @xmath212 and @xmath189 , respectively .
therefore the decay to brane modes , when possible , dominates over bulk decays in weakly - coupled string models .
we also want to comment on the string effects arising in the regime @xmath213 @xcite . in this regime , although the energy is transplanckian , @xmath214 holds .
then the large angle , small impact parameter effects are dominated by string physics instead of classical gravitational effects . in this regime
one should use the string version of the eikonal phase @xcite .
the main effect is that @xmath215 , corresponding to inelastic production of very excited strings ( string balls ) dominating the scattering for @xmath216 .
the corresponding cross section is @xmath217 for @xmath218 . at @xmath219
, we have that @xmath220 , consistent with the string balls becoming indinguishable from ordinary black holes @xcite .
the onset of inelastic string production marks the end of the validity of our description , since any reminiscence of classical gravitational dynamics is lost .
notice that in the context of strings there exists another approach to calculate fixed - angle scattering in the regime @xmath221 @xcite .
this different approach leads to an elastic amplitude that vanishes at small fixed angle like @xmath222 @xcite .
recently this approach was applied @xcite to the tev string scenario .
the conclusions of ref .
@xcite would seem to disagree with those of ref .
@xcite , based on the string eikonal , according to which the scattering at @xmath223 is well described by ( semi)classical general relativity . however , as explained in ref .
@xcite , the borel resummation leading to the result @xmath224 applies without relevant corrections only in a limited range of energies @xmath225 .
indeed for energies such that @xmath226 the two approaches give results that are fairly consistent with each other @xcite .
so we believe that in the transplanckian regime our simple approach is valid .
notice , that also in the approach of ref .
@xcite the elastic amplitude has an exponential suppression , different than that in refs .
@xcite , and due to gravitational radiation and diffractive string production
. moreover the discussion of ref .
@xcite shows that , even if the elastic channel is suppressed , the total cross section at finite angle is large and grows with @xmath22 .
notice also that if @xmath22 is large enough the average mass of the diffractive states eventually exceeds both @xmath187 and @xmath23 , so it is possible that the diffractive channel becomes at extremely high energies the production of a pair of small black holes .
next , we would like to comment on the possible effects of the modes corresponding to the brane motion , the branons @xcite .
the first comment is that these modes do not necessarily exist .
if the sm fields are localized at some fixed point , as in orbifold constructions , then it is consistent to assume that the corresponding `` brane '' is not a dynamical object , and could also have vanishing tension .
so when the sm lives at a fixed point there is no effect ( apart from a `` trivial '' reduction of the phase space of bulk gravitons ) .
two types of effects can be associated with brane dynamics .
first , the presence of a brane tension @xmath227 gives rise to a gravitational field , which at a tranverse distance @xmath89 from the brane vanishes like @xmath228 , where the gravitational radius @xmath229 is @xcite r_^n-2=. the wave function of a bulk graviton with transverse momentum @xmath230 is expected to receive a correction of order @xmath231 from the presence of this background . for a given impact parameter @xmath27
, the tree - level scattering amplitude is precisely saturated by the exchange of graviton kk modes with mass @xmath232 .
therefore we expect the background metric to correct the eikonal phase by a relative amount @xmath233 . for @xmath96 , at large enough impact parameter we can neglect the brane radius as much as we can neglect the schwarzschild radius @xmath24 .
this gravitational brane effect becomes less important as the tension is decreased .
= 1.0 in the second class of effects is associated with the exchange of branons .
branon exchange becomes more important the smaller the tension .
notice first of all that at the classical level branons are not excited by the gravitational shockwave of a fast moving particle , _ i.e. _ the induced metric on the brane remains of the aichelburg - sexl kind .
at the quantum level , branons can be exchanged in pairs between the colliding particles .
the exchange of a pair of branons is very much equivalent to the exchange of a``composite '' graviton . in order to get an estimate of the size of the effect
, we may generalize the graviton eikonal including the ladder and cross - ladder diagrams with two - branon blobs ( see fig .
[ figb ] ) .
the eikonal phase gets modified into ( b)=()^n + .
[ braneik ] in the case of @xmath195 the effect is equivalent to a renormalization of @xmath234 , but in all other cases it can distort our result from pure graviton exchange . including both effects from brane excitations , we can write the relative correction to the elastic amplitude ( in the semiclassical region ) as = ( ) ^4 ( ) ^+ ( ) ^8 ( ) ^ , where we have dropped all numerical factors .
this equation shows that for the interesting case @xmath235 and for the natural choice @xmath236 , both branon effects are small in the region of small angle and large @xmath157 .
notice that for a 3-dimensional d - brane arising by wrapping a @xmath237brane over the @xmath238 dimensions of radius @xmath239 along the lines of the example above , the tension is given by = .
then at weak coupling we have @xmath240 and the effects of the first class are very small .
effects of the second class are also small as long as @xmath241 and @xmath242 . for @xmath195 ,
the contribution from branon exchange leads to the peculiar result @xmath243 , see . to summarize , quantum - gravity corrections are potentially very sizable because , although they involve smaller length scales , they do not appear with large exponents , as in the case of the semiclassical corrections
of course , they can not be computed in the absence of a complete quantum - gravity theory .
we have estimated their effects in the context of string theory , borrowing the results of ref .
@xcite ( which are however strictly valid only for graviton - graviton scattering ) . for @xmath199 , a diffractive production of string states occurs , depleting the elastic scattering cross section , although it may still be possible to recover the gravitational scattering properties by studying inclusive cross sections .
this gravitational excitation of string modes looks like a very promising way to investigate , in a rather clean channel , string effects at high - energy colliders . as the impact parameter
is further decreased and becomes of the order of @xmath244 , we are entering the regime of head - on collisions between string modes , with inelastic production of regge excitations eventually leading to multi - string states @xcite . in the following ,
we will focus on gravitational transplanckian collisions , neglecting quantum - gravity effects .
one should keep in mind that such ( at the moment incalculable ) effects may give significant modifications to the signal discussed below .
although potentially disrupting to the predictivity of the elastic channel , these new - physics effects are of course very interesting from the experimental point of view .
at the lhc , the observable of interest is jet - jet production at small angle ( close to beam ) with large center - of - mass collision energy .
the amplitudes derived in previous sections are applicable for the scattering of any two partons .
the total jet - jet cross - section is then obtained by summing over all possible permutations of initial state quarks and gluons , using the appropriate parton distribution weights and enforcing kinematic cuts applicable for the eikonal approximation . defining @xmath245 and @xmath246 as mandelstam variables of the parton - parton collision , we are interested in events that have @xmath247 and @xmath248
. we can extract @xmath249 from the jet - jet invariant mass @xmath250 , and @xmath246 from the rapidity separation of the two jets @xmath251 , where @xmath252 .
the variable @xmath253 is especially useful since it is invariant under boosts along the beam direction , and it is simply related to the @xmath254 scattering angle in the center - of - mass frame : = .
therefore , the kinematical region of interest is defined by the equivalent statements 0 0 . in computing the cross - sections we use the cteq5 @xcite parton - distribution functions .
we have also compared results with grv @xcite and find little difference .
we evaluate the parton - distribution functions at the scale @xmath255 , see , if @xmath256 and @xmath257 otherwise ( @xmath258 ) @xcite .
we require that both jets have @xmath259 and @xmath260 from conservative detector requirements .
the sm di - jet cross section is computed using pythia @xcite , ignoring higher - order qcd corrections , and the gravitational signal is calculated using our own vegas monte carlo integrater . for simplicity
we are defining the background as the jet - jet cross - section from qcd with gravity couplings turned off , and the signal as the jet - jet cross - section from the eikonal gravity computation with qcd turned off . in reality ,
sm and gravity contributions would be simultaneosly present .
nevertheless , there is no interference between the leading qcd contribution in the limit @xmath261 and the gravitational contribution , since graviton couplings are diagonal in color indices .
however , terms @xmath262 are generated by the interference between gravitational and @xmath263 contributions . at order @xmath264
, the di - jet cross section receives contributions from the square of the amplitude obtained by exchanging 1 gluon and any number of gravitons ( resummed in the eikonal approximation ) and from the interference between the gravitational amplitude and resummed diagrams with exchange of 2 colored particles and any number of gravitons . in the small regions of parameter space where the qcd and gravity contributions are comparable , these terms , which will not be computed in this paper ,
should be taken into account .
we will not make any detailed claims about that region here .
although we are working in the transplanckian regime , we are considering processes with low virtuality ( @xmath265 ) and , as long as no new physics appear at momenta smaller than @xmath266 , the sm dynamics give a reliable estimate of the background . in the context of string theory ,
the qcd background has to be interpreted as the leading contribution from exchange of open strings , in the limit in which the transfer momentum is smaller than the mass of the regge excitations .
the jet - jet process is best described by the two kinematic variables @xmath267 and @xmath253 .
although experimental searches for new physics in the jet - jet observable should investigate the full range of these two kinematic variables , our goal is to study the region in which the eikonal computation is valid , where we can make reliable theoretical predictions .
this leads us to the region of large @xmath253 and large @xmath267 .
let us begin by studying the parton differential cross section signal distribution as a function of the rapidity separation @xmath268 for a fixed @xmath269 , = | f_n ( y)|^2 , where @xmath270 .
this distribution has a series of peaks whose maxima and minima are determined by the values of @xmath253 that satisfy the equation 1-e^- = - .
|_y=. [ picco ] these peaks arise from the oscillations of the function @xmath271 around its asymptotic value given in , and characterize the transition region between the `` classical '' ( small @xmath253 ) and `` quantum '' ( large @xmath253 ) regimes discussed in sect . [ sec2 ] .
therefore , we can not solve using the asymptotic expressions given in or .
approximate solutions of can be found if the peaks are located at sufficiently large values of @xmath253 , since in this case we can neglect @xmath272 with respect to 1 in the left - hand side of .
then , becomes a function only of the variable @xmath89 . for our consideration , this approximation is adequate for determining at least the first peak .
we find that the first peak of the @xmath253 distribution , for a fixed value of the two - jet invariant mass @xmath267 , is given by ^(peak ) = ( ) .
[ piccolin ] the coefficients @xmath273 can be obtained by numerically solving and are given by @xmath274 .
these peaks are a characteristic feature of the higher - dimensional gravitational force , and correspond to the diffraction pattern of the scattered particles . as discussed in sect .
[ physical ] , higher - dimensional gravity , albeit leading to a force with infinite range , defines a length scale ( @xmath45 ) in the quantum theory . a diffraction pattern emerges when @xmath275 corresponding to the interference of the scattered waves in the region @xmath207 . in the coulomb case ( @xmath46 ) ,
such a scale does not exist and therefore no diffractive pattern is produced . the di - jet differential cross section @xmath276 is plotted in fig .
[ fig - dely ] for @xmath195 , @xmath277 and @xmath278 and @xmath279 .
since the parton - distribution functions decrease rapidly at higher @xmath267 , the value of @xmath267 in is well approximated by @xmath280 in this example .
the first peaks are then calculated from to be at @xmath281 ( @xmath278 ) and @xmath282 ( @xmath283 ) .
the monte carlo integrated distributions have peaks that agree well with these numbers .
the same differential cross section is shown in fig .
[ cycle - dely - n ] for different values of @xmath11 and for @xmath278 . for @xmath116 the first peak
is partly hidden by the logarithmic divergence for @xmath284 , and no structure after the first peak is visible . as @xmath11 increases ,
the peaks become more evident .
the study of the peak structure could be a feasible experimental technique to measure the number of extra spatial dimensions @xmath11 .
however , such a study can only be pursued with a full detector simulation , taking into account the rapidity and jet mass resolutions .
since the two jets are experimentally indistinguishable , we have used @xmath285 , instead of @xmath253 , as the appropriate kinematical variable to plot .
this means that the experimental signal considered here contains also contributions from scattering with large and negative @xmath253 , which corresponds to partons colliding with large momentum transfer and retracing their path backwards . for the background ,
these effects are calculable and taken into account .
however , the theoretical estimate of the signal at negative @xmath268 lies outside the range of validity of the eikonal approximation .
nevertheless , this is expected to be negligible and can be safely ignored .
indeed , by dimensional arguments , the differential cross section for @xmath286 is estimated to be @xmath287 , see .
when compared to the signal at small @xmath52 ( @xmath288 ) , this gives a correction @xmath289 \sim { \cal o } [ ( m_d^2/s)^{\frac{(n+2)^2}{n(n+1)}}]$ ] , which is smaller than the terms we have neglected in our calculation , see .
notice that this complication is unavoidable in the presence of identical colliding particles , since the amplitude is symmetric in @xmath52 and @xmath290 ( @xmath291 ) .
recalling our discussion of the corrections to the eikonal amplitude given in sect .
[ seccoreik ] , we notice that semiclassical corrections are indeed small . in the kinematical region of interest , the second term in amounts to less than 1% for @xmath292 and @xmath242 and less than 5% for @xmath293 and @xmath242 , while the first term amounts to about 5% at @xmath294 .
gravitational radiation at @xmath294 and @xmath293 gives corrections of about 7% for @xmath116 and 5% for @xmath195 , see .
quantum gravity can however severely affect our signal .
if the corresponding corrections had the form @xmath295 , then they would amount to 5% ( 6% ) for @xmath292 and @xmath116 ( @xmath195 ) or to 20% ( 9% ) for @xmath293 and @xmath116 ( @xmath195 ) , and they would be under control
. however , it is rather likely that the dynamics taming the ultraviolet behavior of gravity sets in at energies lower than @xmath23 , enhancing these corrections .
this is indeed the case of weakly - coupled strings , where in the previous estimate @xmath25 should be replaced by @xmath51 , giving an enhancement of @xmath296 .
for instance , if the diffractive string production discussed in sect .
[ seccor ] takes place , it will deplete the elastic channel at impact parameters smaller than @xmath297 , which is typically very close to @xmath45 , for values of @xmath23 relevant for the lhc and for weakly - coupled strings . in this case
, string production could be the discovery process at the lhc . as is apparent from fig .
[ fig - dely ] , the background increases with @xmath253 faster than the signal , and so the best signal to background ratio is found at the smallest @xmath268 . however , small @xmath298 are not in accord with the eikonal approximation . to stay within the acceptable kinematic range of the eikonal approximation we impose the condition @xmath299 , which is equivalent to @xmath300 and @xmath301 .
we then select a maximum value of @xmath298 in order to reject the background . in our study
we make a simple universal choice @xmath302 , but in practice one could choose the maximum @xmath298 using only the criterion of retaining enough events to be detected .
we want to stress that this universal range of @xmath253 is chosen just for illustration , since its optimal choice depends on @xmath23 .
in particular , if @xmath23 is below about 2 tev , visible signals certainly can be obtained with a more stringent lower limit on @xmath253 , therefore investigating a region of smaller scattering angles , where the eikonal approximation is even more accurate .
furthermore , with the current computation technology of qcd , the dijet rate in our kinematic configuration is not a precision observable to compare with theory .
this is especially true at large rapidity separation .
however , our gravity signal can swamp the qcd expectation most notably at smaller rapidity separation , where the hard - scattering qcd rate is smaller and more reliably known .
that is partly the reason why we choose for some plots to illustrate the gravity signal in the modest interval @xmath303 . at very large rapidity separation
, the qcd dijet observable may be dominated by bfkl dynamics @xcite , in which case a decorrelation of the azimuthal angles of the two jets ensues @xcite .
such a decorrelation is not expected in the color - singlet gravity signal .
it may even be possible to use this to distinguish qcd from gravity .
similarly , the gravity signal is not expected to fill the central region with soft gluons , and the lack of this central jet activity could be an additional discriminating tool .
after having selected the range of @xmath304 , we now show in fig .
[ fig - mjj ] the cross - section as a function of minimum jet - jet invariant mass cut for @xmath278 and @xmath279 .
we plot results for all @xmath305 , but we recall that the eikonal approximation is valid only for @xmath306 . this plot shows the important feature that the signal cross - section is flatter in @xmath267 than the background .
this enables better signal to background for larger @xmath267 cuts , which is the preferred direction to go for eikonal approximation validity .
therefore , one should make the largest possible @xmath267 cut that still has a countable signal rate for a given luminosity . finally , in fig .
[ fig - md ] we plot the total integrated cross - section as a function of @xmath23 for @xmath307 ( left panel ) and @xmath308 ( right panel ) .
we also have required @xmath309 .
the two solid lines correspond to @xmath195 ( upper line ) and @xmath116 ( lower line ) .
they are not far separated in this log plot .
figure [ fig - md ] demonstrates several universal features .
first , larger @xmath310 corresponds to a larger ratio of gravity signal to expected qcd rate , and to a more reliable applicability of the eikonal approximation .
larger @xmath310 also means less range of @xmath23 probed at the lhc since parton luminosity drops rapidly as @xmath267 approaches the @xmath311 collider limit . in this paper
, we will restrict our considerations to @xmath312 and so elastic scattering with @xmath23 as large as @xmath313 can be tested with more than 10 events in @xmath314 . if we adopted the more conservative constraint @xmath315 , the range of @xmath23 where large and calculable signals are expected reduces to @xmath316 . before concluding this section on the lhc signatures , we recall that the eikonal jet - jet observable analyzed here is not the only transplanckian non - sm process . black - hole production is also expected to be very large with a parton - parton production cross - section given approximately by @xmath317 .
as discussed earlier , this is just a dimensional - analysis estimate and reflects only the expectation that all collisions with impact parameter @xmath48 get absorbed into a black hole .
the description of the scattering process in terms of black - hole production is expected to be reliable only for black hole masses
@xmath318 well above the @xmath6-dimensional gravity scale @xmath23 . in the left panel of fig . [ fig - bh ]
we plot the production cross - section as a function of minimum black - hole mass for @xmath278 and @xmath283 .
although the lines are extended down to @xmath319 , the reliable region is only @xmath320 .
the signatures of black holes are spectacular high - multiplicity events with almost no sm background @xcite .
only a few events at very high invariant mass are needed to identify a signature .
the right panel of fig .
[ fig - bh ] shows the cross - section for black - hole production for @xmath321 and @xmath322 . with an integrated luminosity of @xmath314
, a minimum of several events can be achieved for @xmath323 , if we require @xmath321 , or for @xmath324 if @xmath322 .
these values of @xmath23 are nearly identical to what one can reach with the jet - jet observable discussed above . the combined search for black - hole production and small angle jet - jet events will be important for a full experimental characterization of the extra dimensions and for our ability to determine the underlying parameters of the theory .
in addition to tevatron upgrades and lhc , there are many colliders envisioned for the future .
these include variants on the @xmath325 linear collider theme , the muon collider , and the very large hadron collider ( vlhc ) .
observables associated with gravity are very sensitive to collider energies .
the search for phenomena of almost all beyond - the - sm schemes is interested in higher and higher energies , but gravity is especially dependent on this progress .
the reason is that either the rate increases at a high power of the collision energy if @xmath326 , ( , @xmath327 for jet plus graviton signatures ) ; or , well - chosen observables become calculable if @xmath328 , ( , in the case of eikonal gravity jet - jet production ) . such calculability is taken for granted in many other beyond - the - sm frameworks .
either way , the highest energies attainable are important for increased understanding if low - scale gravity is correct .
we would therefore like to briefly discuss transplanckian collisions in the context of two of the highest energy colliders being considered in the not - too - distant future .
the first is the vlhc @xcite , whose center - of - mass energy for proton - proton collisions is envisaged to be between 50 tev and 200 tev .
obviously the higher energy of this post - lhc machine will be much more probing .
much of the analyses that we have described for the lhc would be applicable for vlhc analyses . by going to higher center - of - mass energies , we can gain in the range of sensitivity to @xmath23 and , more importantly , we can study kinematical regions where quantum - gravity effects are expected to be smaller and the eikonal approximation becomes more trustworthy .
for instance , by analysing dijet final states at the vlhc with invariant masses larger than 150 tev , we can make quantum - gravity corrections ( proportional to @xmath329 ) , for a fixed value of the string scale , about a factor of 200 ( for @xmath116 ) or 6 ( for @xmath195 ) smaller than their expected size in lhc experiments .
the other collider discussed at the highest currently imagined energy is clic @xcite , which is a two - beam @xmath325 linear collider design .
energies as high as @xmath330 are being considered .
although such high - energy machines will have some spread in luminosity as a function of the center - of - mass energy , one still expects that approximately @xmath331 of the beam luminosity will be within 1% of designed center - of - mass energy . with large gravity - induced cross sections ( @xmath332pb or more ) for all @xmath333 , combined with relatively low background ( @xmath332fb ) and impressive luminosity design goals of @xmath334 , clic
could certainly improve , verify and even discover the gravitational origin of beyond the sm signatures .
the maximum values of @xmath23 that can be studied at clic in the gravitational deflection process ( bhabha scattering ) are uniquely determined by the condition for validity of the eikonal approximation . for comparison ,
the condition equivalent to the one chosen in our lhc analysis is @xmath335 .
therefore , a linear collider with @xmath336 tev could probe a parameter - space region , where theoretical calculations are reliable , which is very similar to the one that can be studied by lhc .
however , the cleaner @xmath325 environment offer several advantages for precision tests and parameter determinations . in fig .
[ clic - theta ] we give one example of the signal and background distributions in the scattering angle @xmath117 for @xmath337 tev and @xmath330 . here
@xmath117 is the scattering angle of the electron in the bhabha process , and @xmath338 indicates the electron going down the beam pipe undeflected by the collision .
notice that the sm background is completely insignificant as long as we exclude a small region around @xmath339 .
this allows experimental studies of the cross section in a much more forward region than what is possible at the lhc .
the signal angular distribution is characterized by the peak structure encountered before , with minima and maxima described by the equation ( 1 -^2 ) = - .
|_y = b_c .
the first peak is approximatly given by ^(peak ) ( ) ^ , where @xmath340 is a numerical coefficient with value @xmath341 .
the @xmath195 line ( solid line ) in fig .
[ clic - theta ] is again the most telling line , since a careful measurement of the ups and downs of @xmath342 would be hard to reproduce in another framework .
given the good energy and angular resolutions that can be achieved at a linear collider , clic could perform a much more precise study of the peak structure than what is feasible at the lhc .
another advantage of clic is that unlike a @xmath343 collider , the scattering particles are distinguishable .
this means that the calculable limit of @xmath163 ( small angle scattering ) is unambiguously identifiable .
recall that in the @xmath343 collider case , the case of partons glancing off each other at @xmath344 was not distinguishable from partons bouncing backwards at large momentum transfer @xmath345 .
we argued that these non - calculable large momentum transfer contributions are negligible , but clic can test that assertion .
let us conclude by comparing the various signals and experimental strategies for the discovery and the study of higher - dimensional gravitational interactions at the lhc .
the two relevant parameters are the fundamental planck mass @xmath23 and the number of extra spatial dimensions @xmath11 .
we are tacitly assuming that the new dynamics of quantum gravity does not introduce a new mass scale smaller than @xmath23 , or new phenomena in the relevant kinematical region .
the searches for gravity signals at the lhc are classified in terms of three different kinematical regions .
_ cisplanckian region .
_ this is the region in which the center - of - mass energy of the parton collision is smaller than the planck mass , @xmath346 .
the theory can be described by an effective field - theory lagrangian with non - renormalizable interactions , and the seach for new contact interactions is an adequate experimental tool . however ,
the relations between the coefficients of the contact interactions and the fundamental gravity parameters is model dependent . in this region ,
the search for jet plus missing energy events is particularly interesting , because graviton emission can be reliably calculated with a perturbative expansion .
the region @xcite in which we can trust the effective theory for graviton emission is @xmath347 tev ( for @xmath116 ) or @xmath348 tev ( for @xmath349 ) .
lhc is sensitive to the graviton signal up to @xmath350 tev ( for @xmath116 and an integrated luminosity @xmath351 fb@xmath352 ) or @xmath353 tev ( for @xmath349 and @xmath351 fb@xmath352 ) . in this window
, lhc can perform a quantitative test of high - dimensional gravity in the cisplanckian region .
_ planckian region .
_ this is the intermediate region @xmath354 , where no experimental signals can be predicted without knowledge of the quantum - gravity theory .
signals from graviton emissions or elastic gravitational scattering are also present , but their rates can not be reliably computed .
new and unexpected phenomena will be the key to understand the underlying dynamics .
of course this is the region which will eventually yield the crucial experimental information .
_ transplanckian region .
_ this is the region discussed in this paper , characterized by @xmath355 .
the two experimental signals of interest in the transplanckian region are di - jet events ( from elastic parton scattering ) and black - hole production . unfortunately , in the case of the lhc we are rather limited in energy for both processes
. the actual conditions for calculability we have taken are @xmath356 ( in which the transplanckian region extends up to about 1.8 tev ) or the more optimistic @xmath357 ( in which the transplanckian region extends up to about 3.5 tev ) . in both cases
we are only marginally inside the transplanckian region since a complete separation between the quantum - gravity scale @xmath25 and the classical radius @xmath24 has not been fully reached . therefore quantum - gravity contributions can potentially modify in a significant way both elastic scattering and black - hole formation . in our analysis , we have assumed that such contributions are small .
while the black - hole production cross section can only be estimated by dimensional analysis , the elastic cross section in the small - angle region can be computed as a perturbative expansion over controllable parameters .
the elastic cross section is larger than the one for black - hole production .
the observation of a cross section at finite angle growing with a power of @xmath157 would be a clean signal that the high - energy dynamics of gravity has been detected . given the highly characteristic events from black - hole evaporation , we expect negligible backgrounds for the black - hole events .
as we have shown in this paper , the qcd background for di - jet events can be overcome by studying the distributions in di - jet invariant mass and rapidity separation .
therefore , the di - jet signal from the gravitational deflection of partons is not limited by the background .
its characteristic distributions can be used to test the gravitational nature of the interaction and to determine the parameters of the underlying theory .
* acknowledgments * we wish to thank i. antoniadis , j. barbon , j. bartels , s. catani , p. creminelli , a. de roeck , r. emparan , a. ringwald , a. strumia and g. veneziano for useful discussions .
in this appendix we collect some mathematical formul which are useful to reproduce computations presented in this paper .
the bessel functions @xmath358 have the following series expansions j_n(x)=_k=0^ , j_n(x)=\ { -+o ( ) } .
some useful integrals are d^ny e^ix f(y)= _ 0^dy
y^j_-1(xy)f(y ) , for a generic function @xmath359 , and with @xmath360 , @xmath361 . _
0^dx x^a j_n ( x ) = 2^a ( ) / ( ) . _ 0^dx x^a-1 ( e^ix^-n-1 ) = - ( 1- ) e^- . _ 0^dx x^a e^ix^-n = e^i ( ^-n + ) for a>-n-1 .
s. cullen , m. perelstein and m. e. peskin , phys .
d * 62 * , 055012 ( 2000 ) ; e. dudas and j. mourad , nucl .
b * 575 * , 3 ( 2000 ) ; e. accomando , i. antoniadis and k. benakli , nucl .
b * 579 * , 3 ( 2000 ) ; h. l. lai _ et al .
_ , phys .
d * 51 * , 4763 ( 1995 ) . m. gluck , e. reya and a. vogt , z. phys .
c * 67 * , 433 ( 1995 ) .
t. sjostrand , l. lonnblad and s. mrenna , hep - ph/0108264 .
a. h. mueller and h. navelet , nucl .
b * 282 * , 727 ( 1987 ) ; l. h. orr and w. j. stirling , phys .
b * 436 * , 372 ( 1998 ) .
v. del duca and c. r. schmidt , phys .
d * 49 * , 4510 ( 1994 ) ; phys .
d * 51 * , 2150 ( 1995 ) ; w. j. stirling , nucl .
b * 423 * , 56 ( 1994 ) ; l. h. orr and w. j. stirling , phys .
d * 56 * , 5875 ( 1997 ) .
see http://www - ap.fnal.gov / vlhc/. | elastic collisions in the transplanckian region , where the center - of - mass energy is much larger than the fundamental gravity mass scale , can be described by linearized general relativity and known quantum - mechanical effects as long as the momentum transfer of the process is sufficiently small . for larger momentum transfer , non - linear gravitational effects become important and , although a computation is lacking , black - hole formation is expected to dominate the dynamics .
we discuss how elastic transplanckian collisions can be used at high - energy colliders to study , in a quantitative and model - independent way , theories in which gravity propagates in flat extra dimensions . at lhc energies , however , incalculable quantum - gravity contributions may significantly affect the experimental signal .
= 0.1 in # 1*[#1 ] * # 1*[#1 ] * # 1m_^_#1 # 1m_t_#1 # 1f^(#1)_,g # 1f^(#1 ) _ # 1f^(#1)_g dt ^0_1 _ r _ 1 m _ m_h^ .7ex .7ex
_ i.e. _ _ et al . _
_ e.g. _ # 1o(#1 ) m_t m_b m_z m_w _ s(m_z^2 ) ^2_w _ w _ w _ w
zb| ( zb|b ) ( z ) r_b^ r_b^ r_b # 1eq .
( [ # 1 ] ) /#1 # 1#2#3eur .
phys .
j. c * # 1 * , # 3 ( 19#2 ) # 1#2#3nucl .
phys .
b * # 1 * , # 3 ( 19#2 ) # 1#2#3phys .
lett .
b * # 1 * , # 3 ( 19#2 ) # 1#2#3phys . lett .
b * # 1 * ( 19#2 ) # 3 # 1#2#3phys . rev .
d * # 1 * , # 3 ( 19#2 ) # 1#2#3phys .
rev .
lett . *
# 1 * , # 3 ( 19#2 ) # 1#2#3phys . rep . *
# 1 * ( 19#2 ) # 3 # 1#2#3ann .
rev .
astron .
astrophys . *
# 1 * ( 19#2 ) # 3 # 1#2#3ann .
rev .
nucl . part .
sci .
* # 1 * ( 19#2 ) # 3 # 1#2#3mod .
phys .
lett . a * # 1 * ( 19#2 )
# 3 # 1#2#3zeit .
fr physik c * # 1 * , # 3 ( 19#2 ) # 1#2#3ap .
j. * # 1 * ( 19#2 ) # 3 # 1#2#3mod .
phys .
lett . a * # 1 * ( 19#2 )
# 3 # 10= 0=0 1= 1=1 0>1 # 1 /
@=11 caption#1[#2]#3 @=12 # 1#2 = # 1 # 1#2#3 = # 1 = # 1 # 1#2#3 = 2.8 in # 1#2#3 = 2.1 in cern - th/2001 - 306 + desy 01 - 206 + hep - ph/0112161 * * transplanckian collisions at the lhc and beyond * gian f. giudice@xmath0 , riccardo rattazzi@xmath0 , james d. wells@xmath1 * _ @xmath2cern , theory division , ch-1211 geneva 23 , switzerland _ + _ @xmath3physics department , university of california , davis , ca 95616 , usa _ + _
@xmath4deutsches elektronen - synchrotron desy , d-22603 hamburg , germany _
december 2001 |
large modern surveys of metal - deficient stars in the galaxy have revealed that the fraction of carbon - enhanced objects is significant at low metallicity @xcite .
the fraction appears to increase with decreasing metallicity ( beers & christlieb 2005 ; lucatello et al .
2006 ) , although the derived fraction depends somewhat on one s choice of the definition of the carbon - enhancement phenomenon .
many years ago , @xcite coined the term `` ch stars '' to refer to `` high - velocity '' carbon stars that exhibit very strong g - bands due to the ch molecule , and otherwise weak lines of other metals usually used to classify stellar spectra .
the small sample of stars he had available at that time had spectra that suggested high luminosities , i.e. , membership on the giant branch .
the nomenclature was extended to objects of lower luminosity by @xcite , who found subgiants with similar elemental compositions .
more recently , surveys for carbon stars at high galactic latitude have also been made @xcite .
the carbon - enhanced metal - poor ( cemp ) stars found by recent surveys of metal - poor stars may well be closely related to these objects .
an important question concerning the nature of cemp stars is that of the astrophysical origin of the carbon excess that is observed in these objects .
one established scenario is carbon production by nucleosynthesis in an asymptotic giant branch ( agb ) star .
since most cemp stars that are observed in the present galaxy are too old to be intrinsic agb stars , it is usually assumed that the agb star responsible for carbon production was once the primary of a binary system , and the elements produced during its lifetime and dredged to its outer atmosphere were transferred to its companion prior to the primary evolving to become a faint white dwarf .
observational support for this scenario is found from the observed excess of @xmath5-process elements , as well as carbon , in classical ch stars ( e.g. * ? ? ?
* ; * ? ? ?
* ) and some cemp stars ( e.g. , * ? ? ?
other evidence comes from the binarity of the objects found in some ch and cemp stars identified from radial - velocity monitoring @xcite .
the fraction of binary stars among cemp stars has been recently studied by @xcite , who concluded that all cemp stars with s - process element overabundances ( which beers & christlieb 2005 refer to as cemp - s stars ) could statistically belong to binary systems , given the limited observational constraints on the radial - velocity variations observed to date .
, [ eu / fe ] , and [ ba / eu ] . for instance
, the term s - process element - enhanced stars indicates objects that have high [ ba / fe](@xmath9 ) and high [ ba / eu ] ( @xmath10 ) . ] however , in some cemp stars , no excess of neutron - capture elements is found ( e.g. , * ? ? ?
* ; * ? ? ?
* ) ; these stars are referred to as cemp - no stars @xcite .
at least one cemp star ( cs 22892052 ; * ? ? ?
* ) has been reported with an excess of neutron - capture elements associated with the r - process .
lucatello et al .
( in preparation ) has identified a few other such stars , hence the class cemp - r is no longer based on a single object .
there are a growing number of cemp stars with apparent contributions from _ both _ the r - process and the s - process , the cemp - r / s stars ( see beers & christlieb 2005 for further discussion of these classes ) .
the existence of such a wide variety of elemental abundance patterns associated with cemp stars strongly suggests that other mechanisms may be responsible for their nucleosynthetic history beyond those involving agb stars .
@xcite compiled the elemental and isotopic abundances of 19 cemp stars measured by previous work where the @xmath11c/@xmath12c ratio was studied , and considered possible differences between ba - enhanced cemp stars ( cemp - s ) and ba - normal ones ( cemp - no ) .
this discussion was necessarily somewhat limited because of the small sample size .
our long - running investigation of these interesting objects has led us to undertake a high - resolution spectroscopic study for an expanded sample of candidate cemp stars .
we have thus far obtained spectra for 26 stars with the subaru telescope high dispersion spectrograph @xcite ; we report on their chemical compositions in this paper .
we also analyze redder spectra , in order to measure their na abundances , for another nine stars for which abundances of other elements have been reported by previous work .
based on the results of our present abundance study and on previous work , we discuss the distribution of the carbon - enhancement , the fraction of ba - enhanced stars , and the metallicity distribution among cemp stars .
the evolutionary status of our sample is taken into consideration in our discussion .
the sample selection , observations , and data - reduction procedures are reported in 2 .
details of the abundance analyses are presented in 3 .
our modified definition of carbon - enhanced objects is given in 4 .
we discuss the abundance distribution of barium and the differences between the ba - enhanced and ba - normal stars in 5 . in 6 we present our conclusions .
stars in our main program were selected on the basis of medium - resolution spectroscopic follow - up of metal - poor candidates from the hk survey @xcite and the hamburg / eso survey ( he s ; christlieb 2003 ) that exhibited apparently strong ch g bands and metallicities of [ fe / h ] @xmath13 .
= @xmath14 , and @xmath15 for elements a and b. ] the majority of the he s stars considered in the present paper were drawn from the list of carbon - enhanced stars of christlieb et al .
( 2001 ) , which sought to identify stars on the basis of their carbon enhancement , rather than low metallicity .
the list of objects is provided in table [ tab : obs ] , where coordinates and details of the observations are given .
we note that he 00392635 is identical to cs 29497034 , which was studied in detail by @xcite , @xcite , and @xcite .
this star was re - identified as a candidate cemp star by the he s .
effective temperatures for the stars in our sample are estimated from photometric data using the color - temperature relation provided by @xcite and @xcite ( see [ sec : param ] ) .
optical @xmath16 photometry ( johnson - kron - cousins ) for these stars are reported by beers et al .
the results are listed in table [ tab : photo ] .
errors in the photometric measurements are typically on the order of 0.010.02 mag . near infrared @xmath17 photometry was collected from the database of the two micron all sky survey ( 2mass ) point source catalog @xcite .
table [ tab : photo ] also lists our adopted estimates of interstellar reddening ( @xmath18 ) , which are total line - of - sight values derived from the dust map of @xcite .
the extinction in each band is obtained from the reddening relation provided in their table 6 .
high - resolution spectroscopic observations for the stars in our sample was obtained with subaru / hds , with a spectral resolving power @xmath19 ( a slit width of 0.72 arcsec ) , in october 2002 and may 2003 .
the atmospheric dispersion corrector ( adc ) was used in all observing runs .
two eev - ccds were used with two by two on - chip binning , providing approximately a three - pixel sampling of the resolution element .
the spectra obtained in october 2002 cover the wavelength range 41006850 , with a small gap in the coverage between 5440 and 5520 due to the physical separation between the two ccds .
the wavelength coverage was slightly ( @xmath20 ) shifted to shorter wavelengths in the may 2003 run . for cs 29528028 , which is the hottest star in our sample , a blue spectrum ( 35505250
) was also obtained during the october 2002 run .
spectra of several stars not listed in table [ tab : obs ] were also obtained .
these stars , however , turned out to exhibit severe molecular absorption ( c@xmath0 and cn ) , and special care is required for their abundance analysis .
these spectra will be reported on separately @xcite .
table [ tab : naobs ] lists the object names and observing log for stars for which we have also provided an analysis of na .
all stars other than cs 22183015 were studied by our previous work @xcite , based on blue spectra .
cs 22183015 was studied by @xcite and @xcite , but the na abundance was not measured by these authors . most of the red spectra were obtained with subaru / hds in 2005 during scheduled service observing .
the single exception is cs 22898027 , which was observed during the hds commissioning phase in 2000 .
photon counts per 0.18 km s@xmath21 pixel at 5900 are listed in table [ tab : naobs ] .
note that although some objects were observed for the purpose of radial - velocity monitoring , the counts are sufficient for our present analysis .
standard data reduction procedures ( bias subtraction , flat - fielding , background subtraction , extraction , and wavelength calibration ) were carried out with the iraf echelle package .
suspected cosmic - ray hits are removed using the technique described by @xcite .
the sixth column of table [ tab : obs ] lists the photon count ( per 1.8 km s@xmath21 pixel ) collected at 5200 . since the primary purpose of this work is to investigate the abundance characteristics for a large sample , we pursued only moderate signal - to - noise ( s / n ) ratios that could be obtained with relatively short integrations .
note also that the data quality is not uniform in our sample .
one reason for this is that complete optical photometry was not available at the time of the spectroscopic observing run .
moreover , our sample consists of stars having quite different colors .
these factors made the estimation of exposure times required to obtain uniform s / n ratios quite difficult .
equivalent widths are measured for isolated atomic lines by fitting gaussian profiles , while a spectrum - synthesis technique is applied to molecular bands .
the measured equivalent widths of a number of elements are listed in tables [ tab : ew1 ] , [ tab : ew2 ] , and [ tab : ew3 ] .
we selected 10 elements ( 12 species ) for the measurements of equivalent widths .
table [ tab : ew4 ] lists the equivalent widths of lines in the blue range for cs 29528028 . as shown below ,
the heavy neutron - capture elements in this star are significantly overabundant , and equivalent widths of the sr , zr , la , ce , and nd lines are measured .
the equivalent widths of na lines for the nine other stars are listed in table [ tab : ew5 ] .
the na d lines could not be measured for cs 31062050 because of severe contamination from interstellar na absorption .
the na abundance of this object is determined from the @xmath225688 line .
we measure heliocentric radial velocities ( @xmath23 ) using isolated lines .
the results are listed in table [ tab : obs ] . the standard error ( @xmath24 , where @xmath25 is the standard deviation of the values from individual lines and
@xmath26 is the number of lines used ) is adopted as the random error of the measurements reported in the table .
the wavelength calibration was made using th - ar comparison spectra that were obtained during the observing runs .
possible systematic errors of the wavelength calibration due to the stability of the spectrograph are no more than 0.4 km s@xmath21 ( see * ? ? ?
combining this possible systematic error with the random errors ( 3 @xmath25 ) , the uncertainties of the reported radial velocity measurements are 0.51.0 km s@xmath21 .
the @xmath23 of bs 16929005 measured in the present work agrees with that of @xcite ( @xmath27 km s@xmath21 ) within the stated errors .
no significant variation was found for this object between the two observations , taken roughly two years apart .
the radial velocities and the binarity of cs 22948027 and cs 29497034 were investigated by @xcite .
our results agree , within the measurement errors , with the predictions from the orbital parameters for these two stars derived by these authors .
different orbital parameters for cs 22948027 were determined by @xcite .
although the values of the orbital parameters are quite different between the two studies , both predict a similar velocity at the epoch of our observation .
hence , our measurement for a single epoch does not provide meaningful constraints on the determination of the orbital parameters for cs 22948027 at present .
further monitoring of the radial velocity of this star is strongly desired .
a standard lte analysis , employing the model atmospheres of @xcite , is performed for the measured equivalent widths for most of the detectable elements , while a spectrum - synthesis technique is applied for the ch , c@xmath0 , and cn molecular bands . since the calculations of model atmospheres used in the present analyses are made for scaled solar abundances , the effect of the enhancements of carbon and nitrogen on thermal structures is not included .
this may have some affect on cool stars with large excesses of these elements .
@xcite showed that the temperature at the layers where weak absorption lines form ( @xmath28 to 0 ) is a few hundred kelvin higher in the cn - enhanced ( [ c / fe ] = [ n / fe ] = + 2 for [ fe / h]@xmath29 ) model for @xmath30=4500 k than in the standard model ( that is , the model based on scaled solar abundances ) .
the effect of the enhancements of carbon and nitrogen on , for instance , iron abundances derived from fe i lines , might be 0.20.3 dex , which is estimated from the sensitivity of the derived abundances to the effective temperatures adopted for a few cool stars ( @xmath31 k ) in our sample ( e.g. , he 1447 + 0102 ) .
we note that the carbon enhancement of cs 30322023 , the coolest object in our sample , is relatively small ( [ c / fe]@xmath32 , see [ sec : cn ] ) , thus the effect of the carbon enhancement on the thermal structure of the model atmosphere for this star might be less significant .
effective temperatures are often determined in high - resolution spectroscopic studies by ensuring that the derived abundance of some species ( e.g. , fe ) is independent of the excitation potential of the spectral lines employed . however , since the measurement accuracy of the weak absorption lines is not so high in our moderate s / n spectra , the spectroscopic method to determine effective temperature is less than ideal .
instead , we estimate effective temperatures of our program stars from broad - band photometry , although we note that the determination of effective temperatures for cemp stars from photometric data can also present a challenge , due to the presence of severe molecular absorption in some photometric bands .
we first estimate the effective temperatures for the stars in our sample from their @xmath33 , @xmath34 , @xmath35 , and @xmath36 colors , using the empirical temperature scale of @xcite for cs 29503010 and cs 29528028 ( warm stars ) and @xcite for the others , after reddening corrections were applied ( table [ tab : photo ] ) .
corrections for the filter systems which we employed , in order to put them onto the alonso et al .
scale , were as reported by @xcite .
the metallicity applied to the effective temperature scale is first assumed to be [ fe / h ] @xmath37 , and is then corrected after the first - pass abundance analysis has been completed .
the determination of the effective temperature and abundance analysis are iterated until the derived iron abundance is no longer significantly different from the assumed value .
we give priority to the results from @xmath33 ( @xmath38 ) , as mentioned below , which are listed in table [ tab : photo ] .
the average of the effective temperatures from the other three colors is calculated , and the differences of this average with respect to @xmath38 are also listed in the table . a typical error of the @xmath33 color is 0.03 mag ( errors of the @xmath39-band photometry and reddening estimates are usually dominant ) .
the error of the @xmath40 of 0.03 mag results in an uncertainty in @xmath30 of 90 k. the effect of molecular absorption is relatively small in the @xmath41 and @xmath39 bands for most stars in our sample .
indeed , the @xmath33 colors calculated by models including enhancements of carbon and nitrogen by @xcite are in reasonable agreement with the values derived by the empirical color@xmath42 relation by @xcite . for this reason the results from the @xmath33 color is adopted for most of our program stars . however , we made corrections to the effective temperatures of the following five stars : \(1 ) * cs 22948027 and cs 29497034 * : these stars have been studied in previous work by @xcite and @xcite , who adopted @xmath43 k for both objects .
the effective temperatures we derived from the @xmath33 colors are 330 k and 230 k higher than their values , respectively .
these differences could be due to the differences in the reddening corrections adopted . moreover , other colors yield lower temperatures . including these results , we adopt 5000 k and 4900 k for cs 22948027 and cs 29497034 , respectively , in our analysis .
\(2 ) * he 10051439 and he
1447 + 0102 * : a large discrepancy between @xmath44 and those estimated from the other three color indices is found in stars with strong c@xmath0 and cn molecular bands ; in general , the temperature derived from @xmath33 is lower ( the effect of molecular bands on @xmath36 color is significant ) , as was also found for cs 22948027 and cs 29497034 .
however , the opposite results are obtained for he 10051439 and he 1447 + 0102 , for which quite high temperatures are derived from the @xmath33 colors .
we suspect there is some problem in the estimate of the temperatures from this color for these objects , and adopt the lower values .
\(3 ) * cs 29528028 * : the effective temperature of cs 29528028 from the @xmath33 and other colors is higher than the values found in most low - metallicity main - sequence turn - off stars . in such stars , the effect of molecular absorption is almost negligible in the broad - band photometry , and the @xmath45 color also provides useful information .
the effective temperature of cs 29528028 from the @xmath45 calibration @xcite is 6580 k , significantly lower than that from @xmath33 .
we adopted 6800 k for this object . for the above five stars we assumed larger uncertainties in the estimate of errors of abundance determination due to the uncertainties in their adopted atmospheric parameters ( see [ sec : err ] ) .
we perform an abundance analyses of our program stars in the standard manner for the measured equivalent widths of fe i and fe ii lines .
the microturbulent velocity ( @xmath46 ) is determined from the fe i lines by demanding that there exist no dependence of the derived abundance on equivalent widths . for he 05071653
, he 11570518 , he 13191935 , and he 22210453 , however , no or only a few weak fe i lines were available , hence reliable microturbulent velocities could not be directly determined . in these cases
we assumed a microturbulent velocity of 2.0 km s@xmath21 , close to the average of the values of the other cemp giants in our sample .
a larger uncertainty on the microturbulent velocity is assumed for these three stars in calculations of their abundance errors ( [ sec : err ] ) .
surface gravities ( @xmath47 ) are determined from the ionization balance between fe i and fe ii .
the final atmospheric parameters are reported in table [ tab : param ] .
figure [ fig : teffg ] shows the correlation between the effective temperatures and gravities for our sample ( filled circles ) and also for other stars studied by previous work ( see [ sec : sample ] for references ) .
a clear correlation is found for stars with @xmath48 k , as most stars in our sample lie along the halo red - giant branch .
our sample also includes a number of main - sequence turn - off and subgiant stars .
we investigated the correlation between the lower excitation potential and the derived fe abundances from individual lines .
we found that eight objects among the 26 stars show statistically significant dependence of the derived abundances on the excitation potential .
the dependence disappears by assuming lower effective temperatures by 150250 k than derived from the colors .
this result possibly implies that our calibration based on color indices systematically overestimates the effective temperature of cemp stars .
however , given the agreement of the results for similar number of objects for which sufficient number of lines are measured , the systematic error is unlikely to be as large as 200 k. it should be noted that the five objects to which we applied special corrections of effective temperatures show no significant correlation between the excitation potential and the abundances from individual lines . for the stars for which na
abundance measurements are made , we adopt the atmospheric parameters determined by @xcite , @xcite , and @xcite for cs 22957027 , for cs 22183015 , and for the other seven stars , respectively . for the ch 4323 band and/or the c@xmath0 swan bands , carbon abundances are determined using a spectrum - synthesis approach .
the ch band is used for stars that have relatively weak absorption features of molecules due to their high temperature and/or small excesses of carbon . the c@xmath0 swan 00 band ( 5167 )
is applied to the analysis for stars in which the ch band is highly saturated . for cool objects with significant enhancements of carbon , even the c@xmath0 swan 00 band is severely saturated . in these instances the c@xmath0 swan 01 band at 5635 is used . the results are listed in table [ tab : res ] , where the absorption band used for the analysis is also presented for each star .
there exist some uncertainties in the transition probabilities of these molecular bands , which may possibly cause systematic errors in the carbon abundance determinations .
the abundance analysis for the cemp subgiant lp 62544 by @xcite , using these same three bands , and for which the same line lists as those used in the present study were adopted , demonstrated that the derived carbon abundance from the c@xmath0 bands is 0.2 dex higher than that from the ch band .
thus , a small systematic error in the carbon abundance determination possibly exists , depending on the molecular bands applied to the analysis .
the carbon isotope ratio ( @xmath12c/@xmath11c ) is measured using the c@xmath0 swan 10 bands at 4736 ( @xmath12c@xmath0 ) and 4745 ( @xmath12c@xmath11c ) for 12 objects in which these bands are detected .
the line list and analysis technique used by @xcite are adopted .
the results are listed in table [ tab : ciso ] .
most stars have @xmath12c/@xmath11c ratios around 10 , a typical value found in classical ch stars ( e.g. , * ? ? ?
* ) , while quite high values of this ratio are found for three of our program stars .
the nitrogen abundances are estimated from the cn 4215 band .
for some of our stars the band appears clearly , and without severe saturation . in these cases , a reliable fit of synthetic spectra to the observed
one is obtained .
however , this band is not detected in seven of our program stars , and only upper limits on nitrogen abundances are determined ( see table [ tab : res ] ) .
for some cool objects the cn band is
so severely saturated that the fitting of the synthetic spectra is quite uncertain .
we compare the observed spectrum and the synthetic ones calculated changing the nitrogen abundances , and estimate the fitting errors by eye . in the worst case the uncertainty reaches 0.5 dex .
the abundances of other elements are measured by standard analyses for the measured equivalent widths .
the effects of hyperfine splitting on abundance determination are significant in the analysis of ba lines if the abundances of isotopes with odd mass numbers ( @xmath49ba and @xmath50ba ) are high @xcite .
this is the case when the contribution of the r - process is large , while the s - process mostly yields @xmath51ba .
hence , the derived abundances are dependent on the processes of ba production that are assumed in the analysis .
we first determine the ba abundances with a single - line approximation , and classify our sample into `` ba - enhanced '' and `` ba - normal '' stars . here
, a ba - enhanced star is defined as having [ ba / fe ] @xmath2 ( see [ sec : ba ] ) . for ba - enhanced cemp stars , we assume that the origin of barium is the s - process , and re - calculate the abundances including hyperfine splitting , assuming the ba isotope ratios estimated by @xcite . the effect of hyperfine splitting on the abundances derived from the two resonance lines is on the order of 0.1 dex or smaller . for ba - normal stars , we assume the ba isotope ratios of the r - process component in solar - system material , and derive the final abundance including the effects of hyperfine splitting .
the result of including hyperfine splitting is 0.10.2 dex , depending on the strengths of the ba lines . for cs 29528028 , for which a blue spectrum was obtained in the present study ,
the abundances of neutron - capture elements other than ba are measured ( table [ tab : heavy ] ) .
the effect of hyperfine splitting is included in the analysis of la lines , using the line data by @xcite .
the overabundances of the heavy neutron - capture elements ( ba nd ) are extremely high , while the excesses of the light neutron - capture elements ( sr and zr ) are relatively small .
this abundance pattern is similar to those found in other cemp stars with excesses of s - process elements ( e.g. , aoki et al .
2002d ) .
sodium abundances are determined for 30 stars , including nine objects for which abundances of other elements were previously studied , using the d lines ( @xmath525889 , 5895 ) for most of the stars with available red spectra .
these lines are known to be significantly affected by non - lte effects .
the abundances derived from these lines are corrected for these effects using the results of @xcite , who provide the non - lte corrections for the 5895 line in their figure 2 .
we assume the same corrections apply to the 5889 line .
table [ tab : nares ] presents the results of the na abundance measurements .
the results of an lte analysis are listed in the third column , while the non - lte corrections applied to the results from the d lines , which are dependent on the line strength , are listed in the fourth column . for several stars , the subordinate lines at 5682 and 5688 are used to measure the na abundances .
the non - lte effects on these weak lines are estimated to be quite small @xcite .
the final results are obtained from the non - lte corrected na abundances from the d lines and lte results from the subordinate lines , and are listed in the fifth and sixth columns of table [ tab : nares ] .
the na abundances of four stars are determined using all four of these lines . for cs 29528028 and he 0400 - 2030 , the non - lte corrected values derived from the d lines agree well with the results from the subordinate lines .
however , for cs 22957027 , the results from the d lines are higher by 0.5 dex than those from the other lines , even though a large non - lte correction ( 0.5 dex ) is applied to the results from the d lines . by contrast , for he 05071653 , the results from the subordinate lines are higher by 0.3 dex than the non - lte corrected values from the d lines .
these discrepancies imply that either ( 1 ) the non - lte corrections applied to these stars are inappropriate , or ( 2 ) measurements of lte abundances from the very strong d lines and/or very weak subordinate lines are uncertain .
we note , however , that the abundances derived from the two subordinate lines agree well in cs 22957027 and he 05071653 .
this suggests that either errors exist in the abundance measurements for the d lines , or in the non - lte corrections , or in both .
the random errors in abundance measurements obtained from analysis of molecular bands are estimated by fitting with synthetic spectra calculated by changing the assumed abundances of carbon and nitrogen , as mentioned in [ sec : cn ] .
random abundance errors in the analysis are estimated from the standard error of the abundances for each species .
these values are sometimes unrealistically small , however , when only a few lines are detected . for this reason
, we adopted the larger of ( 1 ) the value for the listed species , or ( 2 ) the standard error derived using the standard deviation of the abundance from individual fe i lines and the number of lines used for the species as estimates of the random errors .
typical random errors are on the order of 0.10.15 dex .
errors arising from uncertainties in the atmospheric parameters are evaluated by adopting @xmath53 k , @xmath54 dex , and @xmath55 km s@xmath21 for four stars , taking the evolutionary status and strength of the molecular bands into consideration : cs 22174007 , a giant with relatively weak molecular absorption ; cs 22948027 , a giant with strong molecular absorption ; cs 30322023 , a cool giant ; and cs 29503010 , a main - sequence star .
the uncertainties estimated for each of our program stars are taken to be equal to those associated with the one of these four templates that is closest in evolutionary status and in strength of the molecular bands . for stars
whose effective temperatures are more uncertain ( see [ sec : param ] ) the abundance uncertainties are estimated assuming @xmath56 k. for stars for which the microturbulent velocities are assumed to be 2.0 km s@xmath21 , @xmath57 km s@xmath21 is assumed in obtaining estimates of abundance uncertainties .
finally , we derive the total uncertainty by adding , in quadrature , the individual errors , and list them in table [ tab : res ]
. the above error estimates do not include possible systematic errors arising from non - lte effects .
the non - lte correction for iron abundances is on the order of + 0.2 dex in very metal - poor stars , according to @xcite , although @xcite find no strong effects ( @xmath58 dex ) ) for dwarfs and red giants cooler than 5000 k. the effect on the abundance ratios of important elements in this work ( e.g. ba , mg ) is also 0.10.2 dex , depending on the spectral lines used for the analyses @xcite .
the effects of inhomogeneity of the stellar atmosphere ( the so - called 3d effects ) are also thought to have a significant influence on the determination of c and n abundances from molecular features @xcite .
although quantitative estimates for 3d effects , coupling with non - lte ones , are not available at present , the corrections are possibly on the order of @xmath59 dex or more ( in the negative direction ) .
we do not apply any correction for these effects in this paper .
instead , we compare below our results with other studies that derived carbon and nitrogen abundances from molecular bands for carbon - enhanced and non carbon - enhanced stars .
the chemical composition of the extremely metal - poor ( [ fe / h]@xmath60 ) star bs 16929005 was determined by @xcite . although the wavelength coverage is different between their study and ours , the results for elements analyzed in common agree to within 0.15 dex .
an exception is ca , for which our analysis yields a [ ca / fe ] lower by 0.3 dex .
there is no line analyzed in common by the two studies .
our measurement is based on only two lines , at 4226 and 6162 . @xcite and
@xcite found the first of these lines yields abundances that are systematically lower than other ca i lines by 0.18 dex .
similar results are found for cs 30312100 , which has similar stellar parameters to bs 16929005 , and for which the ca abundance is determined from 10 lines , including the two mentioned above .
the abundances derived from the lines at 4226 and 6162 are 0.1 dex lower than the average of the result provided by the other available ca i lines .
given this , and the small number of lines measured by the present analysis ( two lines ) , the discrepancy of 0.3 dex is possibly a result of random errors .
two stars from our program with very large carbon enhancements , cs 22948027 ( [ c / fe ] = + 2.1 ) and cs 29497034 ( [ c / fe ] = + 2.7 ) , have been studied in detail by @xcite and , more recently , by @xcite .
here we compare our results with those reported by @xcite .
the iron and carbon abundances of cs 22948027 derived by our present analysis are about 0.2 dex higher than theirs .
this result is well explained by the difference of effective temperatures adopted in the analyses ( our effective temperature is 200 k higher ) .
the nitrogen abundance determined by our study is significantly ( about 0.7 dex ) higher than theirs .
this could also , at least in part , be explained by the difference in the atmospheric parameters adopted by the two studies .
note that @xcite used cn lines of the red system for the abundance analysis , while we analyze a band of the cn violet system , which could also contribute to the discrepancy in the derived n abundances .
our na abundance is also 0.35 dex higher than their lte abundance determination for this element .
the abundances of other elements studied by both works show fairly good agreement . for cs 29497034 , the results of our study agree well with those of @xcite .
the exceptions are for na and mg , for which we derive abundances that are 0.4 and 0.5 dex higher for these elements , respectively .
the small differences in the atmospheric parameters adopted by the two studies can not explain these discrepancies .
the abundances of these two elements for cs 29497034 , as well as the na abundance for cs 22948027 , are determined from only two very strong lines , and small measurement errors in equivalent widths could result in large abundance errors .
we note , however , that the excesses of na and mg in cs 29497034 are clearly found in the results of @xcite , as well as in our analysis . for the following discussion we combined the results of recent abundance studies for cemp stars @xcite with those of the present work . a total of 42 stars , including nine for which na abundances are determined by our analysis , are compiled .
it should be noted that , although cs 22892052 is an cemp star having a large ba overabundance , the origin should be quite different from that for other ba - enhanced cemp stars .
hence , this object is not included in our sample .
details of the selection of the results for objects for which more than one paper reported chemical abundances are discussed in the appendix . the chemical composition and atmospheric parameters for the extended sample of cemp stars ( 64 objects , including 22 stars studied by the present work ) are listed in table [ tab : cempall ] .
figure [ fig : cfe ] shows the carbon abundance ratios ( [ c / fe ] ) as a function of iron abundances ( [ fe / h ] ) for our sample ( filled circles ) and for other cemp stars studied by previous work ( filled squares ) .
the carbon abundances of these stars are determined from the strengths of the molecular bands of ch and/or c@xmath0 .
the carbon abundances for other stars that were not regarded as carbon - enhanced objects by previous studies are indicated with crosses @xcite .
the carbon abundances for these stars were measured from the ch band .
we caution that figure [ fig : cfe ] can not be used to constrain the fraction of cemp stars as compared to other metal - poor stars , because of bias in the sample selection .
lucatello et al .
( 2006 ) provides a discussion of these phenomena based on a sample with better - controlled selection criteria ( the heres sample of barklem et al .
2005 ) .
our program stars distribute in metallicity from [ fe / h]@xmath61 to @xmath62 , providing a useful sample to investigate the metallicity dependence of the abundance trends of cemp stars .
there are 21 stars in our sample with remarkable excesses of carbon ( [ c / fe ] @xmath63 ) , while the [ c / fe ] values of the five other newly - analyzed stars are not clearly distinguishable from most stars which were _ not _ identified as carbon - enhanced objects by previous work . before proceeding with our inspection of elemental abundance patterns , we first consider an appropriate definition for the carbon - enhancement phenomenon in metal - poor stars . the carbon abundance at the surface of a star is expected to decrease when the star evolves to become a red giant , due to mixing with internal material that has been affected by the cno cycle ( the first dredge - up ) . a further decrease of carbon , accompanied with an increase of nitrogen , has been observed in highly evolved metal - poor red giants ( e.g. , * ? ? ?
* ) , indicating the influence of extra mixing during evolution along the red - giant branch @xcite .
as a result , the surface abundance of carbon in late - type stars will in general be lower than it was when the star was in an earlier evolutionary stage .
these evolutionary effects should thus be included in the definition of cemp stars . for this purpose
we calculate the luminosities of the stars in our present sample , and for other stars studied by previous work , using the relation @xmath64 @xmath65 @xmath66 @xmath67 @xmath68 assuming the mass of the stars to be 0.8 m@xmath69 , following @xcite and @xcite .
figure [ fig : cl ] depicts [ c / fe ] as a function of luminosity .
the [ c / fe ] values of most stars studied in previous works are around + 0.4 for @xmath70l@xmath71 , while the value decreases with increasing luminosity .
a similar trend was obtained by @xcite for their metal - poor field giants ( figure 10 of their paper ) .
we note that their carbon abundance ratios ( [ c / fe ] ) are systematically lower than ours by about 0.4 dex , although their abundances were determined from the ch molecular bands as in our present analyses .
this discrepancy might arise because of the difference in the metallicity ranges studied : the gratton et al .
sample mostly consists of stars with [ fe / h ] @xmath72 , while the non carbon - enhanced stars shown in figure [ fig : cl ] have [ fe / h ] @xmath73 .
the difference of metallicity could result in differences in the intrinsic carbon abundances , or in a difference in the 3d effects in stellar atmospheres .
@xcite also found decreases in [ li / h ] , @xmath12c/@xmath11c , and an increase in [ n / fe ] , at @xmath74 ( @xmath75 ) @xmath76 , which are expected to accompany a decrease of carbon during evolution along the red - giant branch .
we note that some systematic errors possibly exist in our estimates of luminosity ( e.g. , nlte effects on the determination of surface gravities ) and also in carbon abundances ( e.g. , 3d effects on the molecular bands )
. however , such systematic errors do not directly affect the empirical definition of cemp stars adopted here . from inspection of figure
[ fig : cl ] we define the stars that satisfy the following criteria as cemp stars : 1 .
@xmath77c / fe@xmath78\geq + 0.7 $ ] for stars with @xmath74 ( @xmath75 ) @xmath79 2 .
@xmath77c / fe@xmath78\geq + 3.0 - \log(l/$]l@xmath69 ) for stars with @xmath80 2.3 the division between cemp and non - cemp stars , using the above definition , is indicated by the dotted line in figure [ fig : cl ] .
four stars with [ c / fe ] @xmath81 ( cs 22174007 , cs 22886042 , cs 30312100 , and cs 31062041 ) are consequently dropped from the sample of cemp stars . on the other hand ,
the cool giant cs 30322023 ( @xmath82)=2.9 ) is included in the sample of cemp stars , although its [ c / fe ] value is only + 0.56 .
another example of a cool giant with relatively low [ c / fe ] that meets our revised definition of cemp stars is cs 30314067 ( [ c / fe]=+0.5 and @xmath82)=3.2 ) , which was studied by @xcite .
both cool giants have quite high nitrogen abundances ( [ n / fe ] = + 2.5 and + 1.2 for cs 30322023 and cs 30314067 , respectively ) , suggesting that a significant amount of carbon was converted to nitrogen during their red - giant evolution . with the application of the above criteria , a total of 22 stars in our sample are classified as cemp stars .
the following discussion is based on these 22 stars and 42 additional stars studied in previous work ( shown by filled squares in figure [ fig : cl ] ) that meet our revised definition of cemp stars .
it should be noted that figure [ fig : cfe ] shows that there is an approximate limit to the observed [ c / fe ] values at a given metallicity .
the limit corresponds to [ c / h ] @xmath8 .
this point is discussed in more detail in
[ sec : ch ] .
barium is a key element for identifying the origin of carbon in cemp stars ( see [ sec : intro ] ) .
@xcite separated their sample of cemp stars into two groups of ba - enhanced and ba - normal stars .
figure [ fig : bafe ] shows the barium abundance ratio ( [ ba / fe ] ) as a function of [ fe / h ] for our expanded sample .
a large scatter of barium abundances is known to exist for very metal - deficient stars ( e.g. , * ? ? ?
* ; * ? ? ?
however , stars with [ ba / fe ] @xmath9 are quite rare among non carbon - enhanced objects . here
we define stars with [ ba / fe ] @xmath2 as `` ba - enhanced '' objects , following @xcite @xmath83 ) vs. cemp - s stars ( [ ba / fe ] @xmath84 ; [ ba / eu ] @xmath2 ) . ] . in figure
[ fig : bafe ] the four stars that are excluded from our sample of cemp stars are shown by open circles the barium abundances of these stars are normal .
one clear result seen from inspection of figure [ fig : bafe ] is that 36 of the 37 cemp stars with [ fe / h ] @xmath1 are ba - enhanced objects .
the exception is he 1410 + 0213 @xcite .
the iron abundance of this star , determined from lines , is [ fe / h ] @xmath85 .
however , note that the [ fe / h ] derived from lines for this object is 0.41 dex lower than that from .
thus , the iron abundance of this star is possibly lower than that shown in the figure , and closer to [ fe / h ] @xmath86 .
our first conclusion is that a majority of , if not all , cemp stars with relatively high metallicity ( [ fe / h ] @xmath1 ) also exhibit excesses of the neutron - capture element ba .
the [ ba / fe ] abundance ratios for the stars in our expanded sample are shown as a function of [ c / fe ] in figure [ fig : cba ] ( the upper panel ) . a correlation between the barium and carbon abundance ratios
is seen for the ba - enhanced stars .
the solid line in this figure indicates the correlation for stars with [ ba / fe]@xmath87 ( 54 stars ) . the null hypothesis that there is no correlation between the two values
is rejected by the regression analysis , as well as by the spearman rank correlation test , at the 99.9% confidence level .
this suggests that the excesses of both the barium and carbon originate in the same astrophysical site , which is most likely to be agb progenitor stars .
as noted above , the carbon abundances of giants are possibly affected by internal processes , including the cn - cycle .
the total abundance of carbon and nitrogen might thus better represent the initial enrichment of carbon by progenitor objects . for this purpose ,
we also show [ ba / fe ] as a function of [ ( c+n)/fe ] in the lower panel of figure [ fig : cba ] . as described above , the cn band is not detected in seven stars , and only upper limits on nitrogen abundances are determined . for these stars the upper limit of [ ( c+n)/fe ] is determined by adopting the upper limit of the nitrogen abundance , while the lower limit of [ ( c+n)/fe ] is determined by assuming the nitrogen abundance to be zero .
the values are listed in table [ tab : ciso ] .
the correlation between [ ba / fe ] and [ ( c+n)/fe ] of 44 objects for which [ ( c+n)/fe ] values are determined is shown by the dashed line in the figure .
although the correlation appears weaker than that found for the [ c / fe ] case , the significance level of the rejection of the no - correlation null hypothesis is still as high as 99% .
the reason for the somewhat weaker correlation is the existence of strongly n - enhanced stars among very metal - poor objects ( e.g. cs 22960053 , cs30322023 ) .
here we separate the sample into two groups with [ fe / h ] @xmath1 ( 29 objects ) and [ fe / h ] @xmath88 ( 15 objects ) . in figure
[ fig : cba ] , stars with [ fe / h ] @xmath88 are shown by symbols with open circles , and the correlation is exhibited by the solid line . a clear correlation between [ ( c+n)/fe ] and [ ba / fe ] is seen for stars with higher metallicity , confirming the above suggestion of the correlation between [ ba / fe ] and [ c / fe ] .
the slope of the solid line ( 0.84 dex per dex ) is also interesting . if the [ ba / fe ] and [ c / fe ] values are determined by the enrichment of agb stars and the dilution in the secondary objects , a slope of unity is expected .
the regression analysis indicates that this possibility can not not rejected ( at the 70% confidence level ) .
a discussion of the stars with lower metallicity can be found below . as can be seen in figure [ fig : bafe ] , the 27 stars with [ fe / h ]
@xmath89 exhibit a large range of ba abundances ( @xmath3 [ ba / fe ] @xmath90 ) .
nine of them are ba - normal stars according to our definition .
figure [ fig : histfe ] shows the [ fe / h ] distribution for ba - enhanced and ba - normal cemp stars , respectively .
the ba - normal stars are only found in the lowest metallicity range , while ba - enhanced stars are distributed over a range in [ fe / h ] above @xmath62 .
the null hypothesis that the samples are drawn from the same parent populations in [ fe / h ] is rejected by the rank - sum ( mann - whitney ) test at the 99.9% confidence level .
our second conclusion is thus that some source(s ) of carbon enhancement that are different from `` standard '' agb nucleosynthesis must operate efficiently at low metallicity . in the smaller sample discussed by @xcite , cemp stars are clearly distinguished into two groups according to their barium abundances .
however , figure [ fig : bafe ] shows that the cemp stars with [ fe / h ] @xmath91 rather exhibit a continuous distribution in barium abundances , owing partly to the two cemp stars having @xmath92 [ ba / fe ] @xmath93 added by the present work ( cs 22960053 and cs 30322023 ) .
these two stars have very high nitrogen abundance ratios , and their values of [ ba / fe ] and [ ( c+n)/fe ] deviate from the correlation found for stars with higher metallicity ( figure [ fig : cba ] ) .
further observations of stars in this metallicity range are clearly desirable , in order to derive some more definitive conclusion on the distribution of ba abundance ratios at the lowest metallicity . as discussed in
[ sec : cfe ] , a limit on the carbon excess is seen in the distribution of carbon abundance ratios ( figure [ fig : cfe ] ) .
this is more clearly seen in figure [ fig : histc ] , where the distributions of carbon abundances ( [ c / h ] ) are shown for ba - enhanced and ba - normal stars separately .
most of the ba - enhanced stars have @xmath94 [ c / h ] @xmath95 , and there is a cutoff of [ c / h ] values at around zero ( i.e. , solar abundance ) .
we note that the four stars having [ c / h ] @xmath96 are cs 29503010 , he 02061916 , he 1447 + 0102 , and cs 22887048 . the [ c / h ] values of these stars are , however , at most + 0.14 ( see table [ tab : cempall ] ) , and hence close to solar .
@xcite recently stated that the carbon abundances of their sample of cemp stars are approximately constant at about 1/5th of the solar value .
most ba - enhanced cemp stars in our sample have carbon abundances that are higher than this value . however , their sample includes ba - normal cemp stars , which could reduce the average of the carbon abundances . given this point , and the relatively small sample size of @xcite , the distribution of carbon abundances derived here is compatible with their conclusion .
the enhancement of carbon in the ba - enhanced cemp stars is primarily determined by the process of agb nucleosynthesis in the binary mass - transfer scenario .
the material containing enhanced carbon is transferred to the surface of the companion , and is expected to be diluted during the evolution of the companion star by mixing with non - polluted material from the companion s interior .
model calculations for the nucleosynthesis in agb stars at low metallicity have shown that the final predicted yield of carbon is on the order of [ c / h ] = 0 .
for instance , @xcite provide the total yields of cno isotopes in their table 5 .
the @xmath12c yields in mass are predicted to be 2.814.95@xmath97 m@xmath69 , which corresponds to [ c / h ] = @xmath98 @xmath99 , for stars with initial mass in the range 1.33.0 m@xmath69 and metallicity of @xmath100 ( [ fe / h]@xmath101 ) . @xcite
derive a similar value ( [ c / h ] = @xmath102 ) for a star with 2.5 m@xmath69 and @xmath103 ( [ fe / h ] = @xmath104 ) .
the chemical compositions of the ba - enhanced cemp stars presumably represent the yields of such low - mass , metal - poor agb stars .
we note that the carbon yields rapidly decrease with increasing mass ( @xmath105m@xmath69 ) , due to hot bottom burning . given these model yields of [ c / h ] @xmath106 0
, the simplest explanation of the upper cutoff of [ c / h ] @xmath106 0 in figures [ fig : cfe ] and [ fig : histc ] is that the observed upper envelope represents objects in which the bulk of the material originated in the agb star and was transferred when the secondary was on or near the main sequence .
the amount of mass transferred was so large that subsequent mixing / dilution in the convection envelope of these objects during giant branch evolution has not greatly changed [ c / h ] from the value that existed in material transferred from the agb donor .
although the [ c / h ] values of most ba - enhanced cemp stars are found in a rather narrow range , a correlation between the [ c / h ] and luminosity is found .
the top panel of figure [ fig : chcnhl ] shows [ c / h ] as a function of luminosity for the ba - enhanced cemp stars ( filled symbols ) and the ba - normal ones ( open symbols ) .
a least - square fit to the data for the ba - enhanced stars is shown by the solid line . the null hypothesis that no correlation exists between these two quantities
is rejected by the regression analysis and by the spearman rank correlation test , both at the 99.9% confidence level .
the [ c / h ] value decreases by 0.6 dex ( i.e. , a factor of four ) with increasing luminosity from @xmath107 0 to 3 .
if this correlation reflects the decrease of carbon abundance at the surface of the object during evolution along the red giant branch , two mechanisms to decrease carbon might be considered .
one is the dilution of the carbon - rich material provided by the donor agb star with the non carbon - rich material of the secondary we are currently observing .
the other is the cn cycle inside the star through the first dredge - up and extra mixing .
however , the bottom panels of figure [ fig : chcnhl ] exhibits a similar correlation between the [ ( c+n)/h ] values and luminosity to that found for [ c / h ] , although the confidence level of the null - hypothesis rejection is only 90% in this case .
that is , the cn cycle is unlikely the primary reason for the relatively low carbon abundances found in some high luminosity objects .
an exception is cs 30322023 ( [ c / h]@xmath108 and [ ( c+n ) /fe]@xmath109 ) , whose surface chemical composition should be severely affected by the cn cycle .
we also shows the plot for [ n / h ] in the middle panel for completeness , where no statistically significant correlation is found between the two values . in summary ,
some dilution of the envelope material in the receiving companion is suggested for ba - enhanced cemp stars from the correlation between [ c / h ] ( and [ ( c+n)/h ] ) and luminosity .
most ba - enhanced cemp stars in our sample show , however , that the decrease of [ c / h ] during the stellar evolution is less than 1 dex .
given the fact that the convective layer becomes about two orders of magnitude deeper when a star evolves to become a red giant , the small dilution of carbon found in ba - enhanced cemp stars suggests that a significant amount of carbon - enhanced material has been accreted from the donor agb star .
in contrast to the ba - enhanced objects , the [ c / h ] values of ba - normal cemp stars exhibit a very wide distribution , with an average value of [ c / h ] that is significantly lower than that of ba - enhanced stars .
the difference of the [ c / h ] distributions between the two groups are clearly seen in figure [ fig : histc ] . the null hypothesis that the samples are drawn from the same parent populations in [ c / h ]
is rejected by the rank - sum test at the 99.9% confidence level .
this result suggests again a difference in the mechanisms that enrich carbon between ba - normal and ba - enhanced stars .
we note that , since the average metallicity of ba - normal cemp stars is much lower than that of ba - enhanced stars ( fig .
[ fig : histfe ] ) , their carbon abundance ratios ( [ c / fe ] ) are higher than the criterion for cemp stars ( 4 ) , although their [ c / h ] values are not outstandingly high .
we also note that apparent correlations are found between the luminosity and abundance distributions of c , n , and c+n for ba - normal stars in figure [ fig : chcnhl ] .
however , they are likely affected by observational bias .
this issue is discussed further in the next subsection .
@xcite suggested that a difference exists in the evolutionary stages between the ba - enhanced and ba - normal cemp stars . in their sample ,
the ba - normal cemp stars are found only at the top of the giant branch , while ba - enhanced stars are found throughout the h - r diagram .
the excess of unevolved ba - enhanced cemp stars , relative to ba - normal ones , was attributed by @xcite to the dilution of surface carbon enhancements in the ba - enhanced stars during the evolution from the main - sequence to the red giant branch . to examine this hypothesis for our expanded sample of cemp stars , figure [ fig : histl ]
shows the luminosity distribution of ba - enhanced and ba - normal cemp stars , respectively ( see also table [ tab : cempall ] ) .
the luminosity , as estimated from our adopted atmospheric parameters as mentioned in [ sec : cfe ] , is an indicator of the stars present evolutionary status .
the ba - enhanced stars distribute over a wide range of luminosity .
we note that the number of objects at @xmath1101.5 is relatively small .
this probably corresponds to the gap between giants and subgiant / main - sequence stars .
although no ba - normal cemp main - sequence star was included in our previous sample @xcite , two such stars were reported by @xcite and @xcite .
the suggested difference in the luminosity distributions between the ba - enhanced and ba - normal cemp stars is not clear for the present sample . indeed , according to the rank - sum test
the difference of the luminosity distributions between these two groups shown in figure [ fig : histl ] is not statistically significant ( the null hypothesis of identical luminosity distributions for the two samples can be rejected at only the 46% significance level ) .
thus , the existence of different luminosity distributions for the ba - enhanced and ba - normal cemp stars suggested by @xcite is not confirmed by our expanded sample . moreover , as discussed in
[ sec : ch ] , the carbon abundance distribution of ba - enhanced cemp stars ( figure [ fig : histc ] ) shows a concentration around [ c / h ] @xmath111 , regardless of evolutionary status , indicating that the dilution of the transferred material with carbon enhancement is not as significant as @xcite presumed . it should be kept in mind that the luminosity distribution of observed cemp stars could be affected by a selection bias , because the strengths of the molecular bands are dependent on the stellar temperature , and , therefore , on the luminosity of the star .
figure [ fig : chteff ] shows the [ c / h ] values as a function of effective temperature . in this figure the ba - normal and ba - enhanced stars are indicated by open and filled symbols , respectively .
all ba - enhanced stars but one ( cs 30322023 ) have [ c / h ] @xmath112 , while most of ba - normal stars have lower carbon abundances . the ch bands that are used in the usual abundance studies are easily detected even in main - sequence stars if the carbon abundance is as high as solar ( i.e. , [ c / h ] @xmath8 ) . however , if the abundance is lower than [ c / h ] @xmath113 , quite high - quality spectra are required for detection of the ch band . in order to estimate the detection limit of the ch g band , we calculate synthetic spectra for model atmospheres of dwarfs ( @xmath30=60007000 k and @xmath47=4.0 ) , subgiants ( @xmath30=5500 k and @xmath47=3.5 ) and giants ( @xmath42=40005000 k and @xmath47=1.02.5 ) .
the dotted line in figure [ fig : chteff ] indicates the [ c / h ] values required for the depth of the ch g band to be about 2 % , which we take as a conservative limit for detection of this feature in high - quality high - resolution spectra .
the detection limit of the ch g band in medium - resolution spectra , by which cemp stars are usually initially identified in surveys , is higher than the values shown by this line .
even if ba - normal cemp main - sequence stars having similar carbon excesses to those found in ba - normal cemp giants exist , they would not be identified as cemp stars in the surveys . on the other hand ,
carbon - enhanced giants are detected even if the carbon enhancement is not as significant as those of cemp main - sequence stars .
ba - enhanced dwarfs as well as giants are detected because of their very high carbon excesses .
this could lead to differences in the observed luminosity distributions of the two types of cemp stars .
figure [ fig : aball ] shows the abundance ratios ( [ x / fe ] ) as functions of [ fe / h ] for elements other than c , n , and ba .
the abundances of non - cemp stars are adopted from @xcite , @xcite , and @xcite . besides these sources , the results of @xcite are also shown for na abundances , while those of @xcite are shown for na , mg , ca , ti , cr , and ni abundances .
the distributions of the abundance ratios of sc , ti , cr , ni , and zn in cemp stars are quite similar to those in non carbon - enhanced stars .
a few stars appear to deviate from the observed trends .
for instance , he 13191935 ( [ fe / h ] @xmath114 ) has [ cr / fe ] @xmath115 , and cs 30338089 has [ sc / fe ] @xmath116 ( [ fe / h ] @xmath117 ) ( both stars are ba - enhanced objects ) .
however , these results are based on measurements for only one line , and no clear interpretations for these stars can be made at present .
several cemp stars exhibit quite high ca abundances compared to the typical value of non - cemp stars ( [ ca / fe ] @xmath118 ) . in our sample , three stars ( he 14290551 , he 15280409 , and he 22210453 )
have [ ca / fe ] @xmath119 .
however , the abundances are determined from only 13 lines in 55006200 .
the three objects are relatively cool and exhibit large carbon excesses ; numerous c@xmath0 and cn molecular lines exist in this wavelength range .
hence , the ca abundances of such objects are possibly overestimated due to contamination by molecular features .
five stars from the literature have [ ca / fe ] @xmath120 ( table [ tab : cempall ] ) .
one is cs 31062050 , studied by @xcite , who measured the abundance from two lines in the blue range . however , these lines are possibly affected by ch molecular lines . indeed , an analysis for 15 lines in the red spectrum of cs 31062050 , which is used for the analysis of na lines in the present work , results in [ ca / fe ] = + 0.40 ( contamination by c@xmath0 and cn molecular lines is not severe in the red spectrum of this object ) .
we conclude that our previous measurement overestimated the ca abundance for this object .
the other four objects having [ ca / fe ] @xmath120 were studied by @xcite , who measured the ca abundances for blue lines as @xcite did .
the four stars are cool giants having large excesses of carbon .
although we can not further investigate the ca abundances of these stars , because red spectra for them are not available , our re - analysis for cs 31062050 suggests that the analyses for blue lines in cemp stars possibly overestimate the ca abundances .
hence , we here do not regard the large ca overabundances found in several objects as general characteristics of cemp stars .
in contrast , a large range of abundance ratios is found for na and mg .
we find moderate overabundances of mg for most of the cemp stars in our sample , as is usually observed for non carbon - enhanced stars , but three stars in our sample exhibit large excesses of mg .
one such star is he 1447 + 0102 ( [ fe / h ] @xmath37 ) , whose mg abundance ( [ mg / fe ] @xmath121 ) is determined from the mg i 5172 line alone , with a correspondingly large uncertainty .
the mg abundances of cs 29497034 ( [ mg / fe ] @xmath122 , [ fe / h ] @xmath123 ) and cs 29528028 ( [ mg / fe ] @xmath124 , [ fe / h ] @xmath125 ) are , however , based on measurements for three lines , and the scatter of abundances derived from individual lines is not significantly large .
therefore , we suggest that these two stars have distinguishable excesses of mg from other metal - poor stars . it should be noted that @xcite and @xcite reported an excess of this element in cs 29497034 , although the mg abundance derived by their work is not as high as our present result ( see [ sec : comp ] ) .
other stars with excesses of mg studied by previous work are lp 62544 ( [ fe / h ] @xmath126 ; aoki et al .
2002a ) , cs 22949037 ( [ fe / h ] @xmath127 ; e.g. , depagne et al .
2002 ) , and cs 29498043 ( [ fe / h ] @xmath128 ; e.g. , aoki et al . 2004 ) .
the origin of the mg excesses in the latter two stars has been suggested to be `` faint supernovae '' explosions that ejected only a small amount of material surrounding the iron core @xcite . on the other hand ,
the sources of mg in other mg - enhanced cemp stars are still unknown .
we note that lp 62544 , cs 22949037 , and cs 29498043 also exhibit large overabundances of oxygen ( see references above ) .
for the two stars with mg excesses in our sample , no useful constraint on the oxygen abundance has been obtained at present .
measurements of oxygen abundances for these stars , possibly from the triplet lines at 7770 or near - infrared co lines , are strongly recommended .
the na abundances of cemp stars exhibit a large range of enhancements . although some uncertainties arising from non - lte corrections and the treatment of damping exist in the determinations of na abundances , this large scatter can not be fully explained by measurement errors alone .
previous evidence @xcite points to there being a slight increase in [ na / fe ] in more evolved , metal - poor red giants .
the former work suggests values around [ na / fe ] = @xmath129 for field subgiants , [ na / fe ] = @xmath130 for field giants , and possibly higher values in globular - cluster giants .
many ba - enhanced cemp stars deviate in their sodium abundances from the solar ratios . in order to investigate the correlation between the excesses of carbon and na , we show [ na / fe ] as a function of [ ( c+n ) /fe ] in figure [ fig : cnna ] . in this figure ,
ba - enhanced and ba - normal cemp stars are shown by filled and open symbols , respectively .
a clear correlation is found for the 14 ba - enhanced stars in our sample ( filled circles ) for which c , n , and na abundances are determined ( as confirmed by the rank correlation test at the 99% confidence level ) , while the correlation is weaker for the full sample of ba - enhanced objects .
we can at least conclude that na - enhanced stars also have high [ ( c+n)/fe ] values .
the mg - enhanced stars cs 29497034 and cs 29528028 exhibit very large enhancements of na ( [ na / fe ] = + 2.74 and + 1.78 , respectively ) .
the iron abundances of these two stars are very low ( [ fe / h ] @xmath131 ) while their excesses of c , n , and ba are very high , as compared to the rest of our sample .
these two stars can be regarded as extreme cases of the ba - enhanced cemp stars .
sodium abundances have also been determined for seven ba - normal cemp stars in the present sample ( in addition to these stars , he 01075240 and he 13272326 , the objects having [ fe / h]@xmath132 , and the dwarf carbon star g77 - 61 , are also shown by open triangles in the figure ) .
four stars among them do not deviate significantly from solar ratios , or have lower values .
the carbon excess in these stars is moderate ( [ c / fe ] @xmath133 ) . this class of objects
was first reported by @xcite , and possibly is a separate class from the other three ba - normal , extremely carbon - enhanced objects .
the two ba - normal @xmath134-element - enhanced cemp stars , cs 22949037 and cs 29498043 , both have large enhancements of na .
the other extremely carbon - rich cemp star , cs 22957027 , also exhibits a large excess of na .
although the abundances of @xmath134-elements in cs 22957027 are normal , this star could be related in some manner to the above two @xmath134-element - enhanced objects . as described in
[ sec : intro ] , the fraction of carbon - enhanced objects is significant at low metallicity , although the estimate of this fraction is dependent on the definition of carbon - enhanced stars and the methods used to identify them .
for instance , @xcite have claimed that the fraction of cemp stars in a sample of giants with [ fe / h ] @xmath13 is 14.5 @xmath135% .
lucatello et al .
( 2006 ) , based on high - resolution analyses of a much larger sample of stars , including many earlier - type stars than the cohen et al .
sample , conclude that this fraction is at least 20 @xmath136% .
an important question is what is ( are ) the origin(s ) of the carbon - enhancement in such objects .
our results indicate that the majority of cemp stars are ba - enhanced objects the fraction of ba - enhanced objects is more than 80% in our sample .
this trend is particularly remarkable for stars with relatively high metallicity , above [ fe / h ] @xmath86 .
therefore , our conclusion is that the large fraction of cemp stars at low metallicity can be principally attributed to the high efficiency of agb nucleosynthesis followed by the mass transfer to a surviving companion @xcite . at the lowest metallicities
, the contribution of ba - normal stars to the population of cemp stars becomes large .
this is another factor that leads to the increased fraction of carbon - enhanced stars at low metallicity .
the most extreme cases are the carbon - enhanced hyper metal - poor stars with [ fe / h ] @xmath137 he 13272326 @xcite and he 01075240 @xcite , for which only upper limits of ba are determined . given the wide distribution of carbon abundances among cemp stars , as well as the unique abundance patterns of several stars in this class , the cemp population of stars almost certainly comprises several astrophysical origins .
no well - established model exists to explain the ba - normal cemp stars , although @xcite suggested that they form from carbon - rich gas ( from a massive progenitor - star population ) in contrast to the agb progenitor paradigm , which explains many features of ba - enhanced objects .
for the two stars cs 22949037 and cs 29498043 , large excesses of the @xmath134-elements and oxygen are also found , suggesting the contribution of `` faint supernova '' nucleosynthesis , with the progenitor stars undergoing mixing and fallback @xcite .
a similar interpretation might be applicable to the hyper metal - poor stars with [ fe / h ] @xmath138 ( see also iwamoto et al .
2005 ) , although several other models have been proposed for these stars .
one idea is that such stars might have formed from gas polluted by mass loss from early - generation massive , rapidly - rotating stars with [ fe / h ] @xmath139 @xcite .
this model has the additional advantage of being able to account for the otherwise mysterious lack of detected li in the turnoff star he 13272326 @xcite .
piau et al .
( 2006 ) have argued that this same population might even be responsible for the alteration of primordial li from ( higher ) big bang nucleosynthesis values down to a level consistent with the spite plateau li abundance found in metal - poor turnoff stars .
if a significant population of such primordial objects did indeed exist , one might argue that they could be closely related to the astrophysical origin of the ba - normal cemp stars found at quite low metallicity .
karlsson ( 2006 ) has appealed to the same progenitors to account for the possible gap in the metallicity distribution of halo stars between [ fe / h ] @xmath140 and @xmath141 revealed by the hk and he s prism surveys . also of interest
is the recent paper by chiappini et al .
( 2006 ) , where their new chemical evolution models _ require _ the large yields of carbon and nitrogen from the models discussed by @xcite in order to account for the upturn of the observed [ n / o ] and [ c / o ] ratios in the `` unmixed '' carbon-_normal _ stars described by spite et al .
( 2005 ) .
other possibilities for the origin of ba - normal cemp stars , in particular for the objects having normal abundance ratios of the @xmath134-elements , exist .
@xcite suggested that one possible interpretation could be that the s - process in agb stars with such low metallicities did not produce any distinctive excess of barium . @xcite
have recently proposed that the ba - normal cemp stars are the result of nucleosynthesis in extremely metal - poor agb stars with masses higher than 3.5 m@xmath69 .
their models of agb stars , originally developed by @xcite , suggest that the s - process does not occur in extremely metal - poor agb stars in this mass range because of the low efficiency of radiative @xmath11c burning ( the case iv in their classification ) .
interestingly , according to their models , this class of agb stars should have metallicities of [ fe / h ] @xmath142 , which is very similar to the metallicity range over which ba - normal cemp stars are found ( [ sec : ba ] ) .
unfortunately , there is no evidence of binarity for most objects in this class .
however , according to their models , binary systems having very long orbital periods are able to produce carbon - enhancement by stellar winds from a primary agb star at very low metallicity . for such long - period binary systems ,
the present absence of variations in radial velocity is not conclusive . a statistical analysis ( e.g. , lucatello et al .
2005 ) based on large samples of ba - normal cemp stars observed over very long temporal baselines will be required in order to test such models .
in this paper we have determined chemical abundances for 26 metal - poor stars , and classified 22 stars among them as carbon - enhanced metal - poor ( cemp ) stars , based on a new definition for the carbon - enhancement phenomenon including the effects of evolution along the giant branch on observed carbon abundances .
sodium abundances were determined for another nine cemp stars for which abundances of other elements were previously studied . combining the results of previous work for 42 cemp stars and the 22 other cemp stars in our present sample , we investigated the abundance trends found in cemp stars and discussed the origins of carbon excesses in these objects .
we found that most cemp stars ( more than 80% ) exhibit excesses of neutron - capture elements ( e.g. , ba ) , suggesting significant contributions of agb nucleosynthesis .
however , the fraction of ba - normal stars increases at lower metallicity .
the difference in the metallicity distribution of ba - enhanced and ba - normal cemp stars found in the present work should be a key to understanding the processes responsible for the ba - normal cemp stars .
another difference between the ba - enhanced and ba - normal stars is found in their carbon abundance distributions . while the ba - enhanced stars have quite high carbon abundances ( @xmath143 [ c / h ] @xmath144 ) , independent of their metallicity , the [ c / h ] values of ba - normal cemp stars distribute over a wide range , while their average value is relatively low .
this difference also suggests that the processes responsible for the ba - normal cemp stars are quite different from the agb nucleosynthesis that can explain the chemical abundances of ba - enhanced cemp stars .
some possible interpretations are proposed for the origin of ba - normal cemp stars , in particular for the @xmath134-element - enhanced stars . moreover , these stars are likely to be closely related to the two known hyper metal - poor stars with [ fe / h ] below @xmath145 . however , the sample of such stars studied with high - resolution spectroscopy is still quite small ( ten objects in the present sample ) .
further observations for this class of objects are strongly recommended .
measurements of abundances for ba - enhanced cemp stars based on high - resolution spectroscopy have been rapidly increasing in recent years .
although the abundances of these objects are usually interpreted within the paradigm of agb nucleosynthesis and mass transfer in binary systems , their detailed abundance patterns are not well explained by current models .
one example is the abundance pattern found for objects with very large excesses of neutron - capture elements , which can not be explained by the present agb models .
several cemp stars are known to have enhancements of both r- and s - process elements , the so - called cemp - r / s stars .
more detailed abundance studies , as well as the monitoring of radial velocities to investigate their binarity , are strongly desired .
we plan to investigate these problems in future papers in this series .
w.a . is grateful for useful discussions on the nucleosynthesis and evolution of cemp stars with drs .
m. y. fujimoto and t. suda . j.e.n . acknowledges support from australian research council grant dp0342613 . t.c.b . acknowledges partial support from a series of grants awarded by the us national science foundation , most recently , ast 04 - 06784 , as well as from grant phy 02 - 16783 ; physics frontier center / joint institute for nuclear astrophysics ( jina ) .
n.c . acknowledges support from deutsche forschungsgemeinschaft under grants ch 214/3 and re 353/44 .
in this paper we discussed the abundances and evolutionary status for 22 cemp stars analyzed by the present work and 42 cemp stars that have been previously studied .
table [ tab : cempall ] lists the abundances and atmospheric parameters of cemp stars discussed in [ sec : cfe ] and [ sec : disc ] .
references are provided in the last column of the table .
the chemical compositions of the following objects were reported by more than one paper . here
we provide a few comments for these objects and justify our adopted results . * * cs 22948027 and cs 29497034 * : these objects were studied in considerable detail by @xcite and @xcite .
comparisons of their measurements and ours are discussed in [ sec : comp ] .
@xcite studied binarity , as well as abundances , for cs 22948027 . in the discussion we adopted the results from the present work . * * cs 22957027 * : @xcite and @xcite reported abundances for this star .
@xcite studied its binarity , as well as its elemental abundances .
we adopted the results of our previous work @xcite , based on spectra of higher quality than previously available .
the na abundance of this object was determined by the present work .
@xcite also reported a large excess of na , although their value of [ na / fe ] is 0.5 dex lower than ours .
we note that our result is based on abundances from subordinate lines and the non - lte corrected abundances from d lines ( see [ sec : naba ] ) . * * cs 22898027 * : @xcite reported the abundances of this object for the first time .
its binarity and abundances were studied by @xcite .
we adopted the results of our previous work @xcite , based on spectra with remarkably higher quality than available previously .
the na abundance was determined in the present work . * * cs 22942019 * : we adopted the results of our previous studies @xcite .
this star was also analyzed by @xcite , who derived results similar to ours .
we adopted their na abundance . * * cs 31062050 * : the abundances determined by our previous work @xcite were adopted .
this object was also studied by @xcite , based on high quality spectra .
their results for elements discussed in the present paper agree within the errors with our previous work .
an exception is the ba abundance ratio , for which @xcite derived a value 0.5 dex higher than ours . * * cs 22880 - 074 * : the results of our previous studies @xcite and the present work for the na abundance were adopted .
this object was also studied by @xcite .
the abundances derived by them agree within 0.3 dex with our results .
an exception is the mg abundance , for which @xcite derived a quite low value ( [ mg / fe ] = 0.06 ) compared to other metal - poor stars . * * cs 29502092 * : the abundances determined by @xcite were adopted .
this object was more recently studied by @xcite , who compared in some detail their results with those of @xcite . *
* cs 29497030 * : this object was studied by @xcite and @xcite .
more recently , @xcite made a more comprehensive abundance analysis using higher quality spectra .
we adopted the results of @xcite . * * cs 22949037 * : the abundances of this object were reported by @xcite for the first time .
this object was re - investigated by @xcite , @xcite , and @xcite .
we adopted the results by @xcite , who made the most comprehensive abundance analysis .
an exception is the na abundance , for which we adopted the result of @xcite , who corrected the value of @xcite using the same spectra . *
* cs 22183015 * : this object was studied by @xcite , in particular for the neutron - capture elements .
abundances of this object , including those of light elements not measured by @xcite , were recently reported by @xcite .
here we adopted the results by @xcite .
hirschi , r. 2006 , in proceedings of origins of matter and evolution of galaxies ( omeg05 ) : new horizons of nuclear astrophysics and cosmlogy , " ( ed .
s. kubono , w. aoki , t. kajino , t. motobayashi , k. nomoto ) , aip conf .
847 , 71 ( astro - ph/0601498 ) llccrrrrr 1 & bs 16929005 & 13:03:29.4 & + 33:51:06 & 600 & 1800 & may 25 , 2003 & @xmath14650.4 & 0.13 + 2 & cs 22174007 & 01:14:06.7 & @xmath14611:02:32 & 600 & 6700 & oct 26 , 2002 & @xmath14670.76 & 0.03 + 3 & cs 22886042 & 22:20:25.9 & @xmath14610:23:20 & 600 & 2300 & oct 26 , 2002 & @xmath146221.4 & 0.05 + 4 & cs 22948027 & 21:37:45.4 & @xmath14639:27:20 & 600 & 3100 & oct 26 , 2002 & @xmath14662.49 & 0.09 + 5 & cs 22960053 & 22:16:17.0 & @xmath14643:54:21 & 2400 & 1900 & oct 26 , 2002 & 282.26 & 0.08 + 6 & cs 29503010 & 00:04:55.4 & @xmath14624:24:19 & 900 & 2500 & oct 26 , 2002 & @xmath14624.28 & 0.04 + 7 & cs 29528028 & 02:28:26.5 & @xmath14621:33:00 & 1200 & 1500 & oct 26 , 2002 & & + & cs 29528028 & & & 3000 & 3000 & oct 28 , 2002 & 58.23 & 0.34 + 8 & cs 30312100 & 15:43:31.8 & + 02:01:18 & 150 & 1000 & may 25 , 2003 & @xmath146122.66 & 0.12 + 9 & cs 30322023 & 21:30:32.0 & @xmath14646:16:31 & 600 & 3200 & oct 26 , 2002 & 116.47 & 0.07 + 10 & cs 30338089 & 23:15:50.1 & + 10:19:26 & 900 & 2000 & oct 26 , 2002 & @xmath146113.06 & 0.10 + 11 & cs 31062041 & 00:35:02.9 & @xmath14615:54:36 & 1200 & 2500 & oct 26 , 2002 & @xmath146172.45 & 0.05 + 12 & cs 29497034 & 00:41:39.7 & @xmath14626:18:53 & 1200 & 7000 & oct 26 , 2002 & @xmath14649.52 & 0.10 + 13 & he 02061916 & 02:09:19.6 & @xmath14619:01:56 & 1200 & 2200 & oct 26 , 2002 & @xmath146200.01 & 0.09 + 14 & he 04002030 & 04:02:14.8 & @xmath14620:21:53 & 900 & 1900 & oct 26 , 2002 & 221.83 & 0.05 + 15 & he 04410652 & 04:43:29.8 & @xmath14606:46:54 & 1800 & 3000 & oct 26 , 2002 & @xmath14629.77 & 0.05 + 16 & he 05071653 & 05:09:16.5 & @xmath14616:50:05 & 1200 & 10500 & oct 26 , 2002 & 353.01 & 0.07 + 17 & he 10051439 & 10:07:52.3 & @xmath14614:54:20 & 1800 & 1250 & oct 26 , 2002 & 86.87 & 0.20 + 18 & he 11570518 & 12:00:18.1 & @xmath14605:34:43 & 1800 & 1050 & may 25 , 2003 & 142.91 & 0.26 + 19 & he 13191935 & 13 22 38.7 & @xmath14619 51 11 & 900 & 1000 & may 25 , 2003 & 368.98 & 0.11 + 20 & he 14290551 & 14:32:31.3 & @xmath14606:05:00 & 600 & 2600 & may 25 , 2003 & @xmath14644.94 & 0.11 + 21 & he 1447 + 0102 & 14:50:15.1 & + 00:50:15 & 1500 & 1800 & may 25 , 2003 & @xmath14661.22 & 0.21 + 22 & he 15231155 & 15:26:41.0 & @xmath14612:05:43 & 600 & 3500 & may 26 , 2003 & @xmath14645.03 & 0.12 + 23 & he 15280409 & 15 30 54.3 & @xmath14604 19 40 & 1500 & 1300 & may 25 , 2003 & @xmath146157.48 & 0.25 + 24 & he 22210453 & 22:24:25.7 & @xmath14604:38:02 & 900 & 3000 & oct 26 , 2002 & @xmath146189.91 & 0.13 + 25 & he 22280706 & 22:31:24.5 & @xmath14606:50:51 & 900 & 4000 & oct 26 , 2002 & @xmath14611.69 & 0.12 + 26 & he 23300555 & 23:32:54.8 & @xmath14605:38:50 & 1200 & 1500 & oct 26 , 2002 & @xmath146235.03 & 0.09 + llcccccccccr 1 & bs 16929005 & 14.228 & 13.610 & 13.208 & 12.766 & 12.170 & 11.750 & 11.670 & 0.01 & 5257 & 31 + 2 & cs 22174007 & 13.103 & 12.409 & 11.961 & 11.509 & 10.885 & 10.404 & 10.318 & 0.08 & 5309 & @xmath14621 + 3 & cs 22886042 & 14.106 & 13.275 & 12.763 & 12.269 & 11.511 & 11.028 & 10.940 & 0.03 & 4830 & @xmath14636 + 4 & cs 22948027 & 13.790 & 12.660 & 12.160 & 11.760 & 10.979 & 10.534 & 10.427 & 0.08 & 5135 & @xmath146298 + 5 & cs 22960053 & 15.705 & 14.950 & 14.537 & 14.046 & 13.298 & 12.833 & 12.761 & 0.07 & 5160 & @xmath146121 + 6 & cs 29503010 & 14.209 & 13.737 & 13.430 & 13.128 & 12.726 & 12.468 & 12.433 & 0.08 & 6532 & @xmath14633 + 7 & cs 29528028 & 14.862 & 14.510 & 14.292 & 14.015 & 13.755 & 13.535 & 13.537 & 0.03 & 7046 & 248 + 8 & cs 30312100 & 13.629 & 12.916 & 12.444 & 11.995 & 11.390 & 10.946 & 10.879 & 0.05 & 5256 & 105 + 9 & cs 30332023 & 13.460 & 12.200 & 11.498 & 10.908 & 10.050 & 9.479 & 9.313 & 0.03 & 4344 & 2 + 10 & cs 30338089 & 15.107 & 14.144 & 13.632 & 13.176 & 12.558 & 12.046 & 12.000 & 0.01 & 4985 & 72 + 11 & cs 31062041 & 14.754 & 13.934 & 13.476 & 12.984 & 12.389 & 11.926 & 11.821 & 0.03 & 5066 & 104 + 12 & cs 29497034 & 13.338 & 12.221 & 11.704 & 11.321 & 10.571 & 10.115 & 9.996 & 0.05 & 5037 & @xmath146340 + 13 & he 02061916 & 15.133 & 13.999 & 13.458 & 12.987 & 12.243 & 11.765 & 11.660 & 0.13 & 5161 & @xmath14676 + 14 & he 04002030 & 14.921 & 14.033 & 13.605 & 13.164 & 12.494 & 12.064 & 11.975 & 0.13 & 5519 & @xmath14674 + 15 & he 04410652 & 15.250 & 14.228 & 13.678 & 13.106 & 12.354 & 11.828 & 11.706 & 0.11 & 4864 & 44 + 16 & he
05071653 & 13.633 & 12.507 & 12.002 & 11.579 & 10.883 & 10.430 & 10.315 & 0.03 & 4980 & @xmath14671 + 17 & he 10051439 & 14.440 & 13.524 & 13.054 & 12.515 & 12.240 & 11.754 & 11.696 & 0.03 & 5512 & 651 + 18 & he 11570518 & 16.273 & 15.120 & 14.635 & 14.128 & 13.418 & 12.917 & 12.846 & 0.03 & 4875 & 13 + 19 & he 13191935 & 15.698 & 14.198 & 13.630 & 13.167 & 12.334 & 11.769 & 11.681 & 0.03 & 4635 & @xmath146175 + 20 & he 14290551 & 14.015 & 12.606 & 11.995 & 11.537 & 10.734 & 10.272 & 10.066 & 0.05 & 4681 & @xmath146191 + 21 & he 1447 + 0102 & 15.632 & 14.614 & 14.121 & 13.676 & 13.207 & 12.760 & 12.682 & 0.02 & 5285 & 274 + 22 & he 15231155 & 14.571 & 13.225 & 12.622 & 12.138 & 11.372 & 10.846 & 10.748 & 0.05 & 4750 & @xmath14656 + 23 & he
15280409 & 16.003 & 14.757 & 14.159 & 13.681 & 12.945 & 12.455 & 12.358 & 0.10 & 4994 & @xmath1467 + 24 & he 22210453 & 15.098 & 13.650 & 12.942 & 12.376 & 11.524 & 10.997 & 10.815 & 0.04 & 4378 & 15 + 25 & he 22280706 & 14.353 & 13.332 & 12.797 & 12.353 & 11.692 & 11.289 & 11.192 & 0.05 & 5130 & 87 + 26 & he 23300555 & 15.412 & 14.559 & 14.158 & 13.612 & 12.850 & 12.395 & 12.300 & 0.03 & 4915 & @xmath14674 + llccrrrcc na1 & hd 196944 & 30 & 17500 & sep 13 , 2005 + na2 & lp 7067 & 180 & 2500 & sep 18 , 2005 + na3 & cs 22183015 & 600 & 2200 & june 17 , 2005 + na4 & cs 22880074 & 3000 & 5500 &
june 18 , 2005 + na5 & cs 22898027 & 3000 & 15000 & july 3 , 2000 + na6 & cs 22957027 & 5330 & 17800 & june 18 , 2005 + na7 & cs 29526110 & 720 & 3500 & sep 18 , 2005 + na8 & cs 30301015 & 300 & 1900 & sep 13 , 2005 + na9 & cs 31062050 & 3600 & 22200 & june 29 , 2005 + lcccccccccccc na i & 5688.20 & 2.10 & @xmath147 & ... & ... & ... & ... & ... & 5.3 & 38.8 & ... & 11.1 + na i & 5889.95 & 0.00 & @xmath148 & 95.5 & 161.9 & 164.8 & 195.8 & ... & 153.7 & 194.6 & 147.8 & ...
+ na i & 5895.92 & 0.00 & @xmath149 & 64.2 & 141.5 & 137.6 & 188.5 & 172.0x & 147.2 & 183.8 & 126.0 & ... + mg i & 4571.10 & 0.00 & @xmath150 & ... & 54.6 & 58.1 & ... & ... & 26.9 & ... & 38.5 & 99.5
+ mg i & 4702.99 & 4.35 & @xmath151 & 28.6 & 67.1 & 70.5 & ... & ... & ... & 55.8 & 63.2 & ... + mg i & 5172.69 & 2.71 & @xmath152 & 128.5 & 191.9 & 174.9 & 191.1 & 130.4 & ... & 141.4 & 180.3 & 190.5 + mg i & 5183.60 & 2.72 & @xmath153 & 143.1 & 215.5 & 190.2 & ... & 146.1 & ... & 157.3 & 202.4 & ... + mg i & 5528.40 & 4.35 & @xmath154 & 28.0 & 69.0 & 66.9 & ... & 41.8 & 103.6 & ... & 64.7 & 71.0 + ca i & 4226.73 & 0.00 & @xmath155 & 124.9 & ... & ... & ... & ... & ... & 87.7 & 188.4 & ... + ca i & 4318.65 & 1.90 & @xmath156 & ... & ... & ... & ... & ... & ... & ... & 38.8 & ... + ca i & 4425.44 & 1.88 & @xmath157 & ... & 53.3 & 49.4 & ... & ... & 72.3 & ... & 42.2 & ... + ca i & 4434.96 & 1.89 & @xmath158 & ... & ... & ... & ... & ... & 89.8 & ... & ... & ... + ca i & 4455.89 & 1.90 & @xmath159 & ... & 42.7 & 32.8 & ... & ... & 59.0 & ... & 32.5 & ... + ca i & 5265.56 & 2.52 & @xmath160 & ... & 27.9 & 27.5 & ... & ... & 51.2 & ... & ... & 16.3 + ca i & 5581.97 & 2.52 & @xmath161 & ... & ... & 16.9 & ... & ... & 29.9 & ... & ... & ... + ca i & 5588.76 & 2.53 & @xmath162 & ... & 49.5 & 41.8 & 61.3 & ... & 74.4 & ... & 35.6 & 24.0 + ca
i & 5594.47 & 2.52 & @xmath148 & ... & 40.1 & 31.5 & 79.0 & ... & 62.4 & ... & 27.0 & 25.2 + ca i & 5601.29 & 2.53 & @xmath151 & ... & 15.2 & ... & 122.5 & ... & 30.5 & ... & ... & ... + ca i & 5857.45 & 2.93 & @xmath155 & ... & 26.7 & 21.7 & 64.2 & ... & 46.2 & ... & 25.4 & ... + ca i & 6102.72 & 1.88 & @xmath163 & ... & 34.5 & 29.9 & 51.2 & ... & 48.7 & ... & 21.9 & ...
+ ca i & 6122.22 & 1.89 & @xmath164 & ... & 59.5 & 54.9 & 91.4 & 23.8 & 75.5 & ... & 47.6 & 51.7 + ca i & 6162.17 & 1.90 & @xmath165 & 22.0 & 72.1 & 67.3 & ... & 30.3 & 83.9 & ... & 67.0 & 68.8 + ti
i & 4533.24 & 0.85 & @xmath166 & ... & 49.3 & 45.7 & ... & ... & 42.3 & ... & 40.5 & 35.0 + ti i & 4981.73 & 0.85 & @xmath167 & ... & 55.6 & 51.5 & ... & ... & 46.6 & ... & 45.4 & 47.0 + ti i & 4991.07 & 0.84 & @xmath168 & ... & 51.0 & 47.9 & ... & ... & 58.0 & ... & 44.6 & 43.8 + ti i & 4999.50 & 0.83 & @xmath169 & ... & 43.8 & 49.9 & ... & ... & 49.7 & ... & 35.3 & 43.3 + ti i & 5007.21 & 0.82 & @xmath170 & ... & 47.5 & 45.5 & ... & ... & 61.4 & ... & 47.5 & 32.2 + ti i & 5173.74 & 0.00 & @xmath171 & ... & 24.9 & 17.2 & 54.8 & ... & ... & ... & 21.4 & 24.1 + ti
i & 5192.97 & 0.02 & @xmath172 & ... & 29.6 & 21.9 & 68.7 & ... & ... & ... & 21.2 & ... + ti i & 5206.08 & 2.49 & @xmath173 & 22.9 & 68.9 & ... & ... & ... & ... & ... & ... & ... + ti i & 5210.38 & 0.05 & @xmath174 & ... & 34.9 & 25.2 & 45.5 & ... & ... & ... & 31.2 & 28.2 + cr i & 4616.14 & 0.98 & @xmath175 & ... & 13.9 & ... & ... & ... & ... & ... & ... & ... + cr i & 4626.19 & 0.97 & @xmath176 & ... & ... & ... & ... & ... & ... & ... & 21.1 & ... + cr i & 4646.17 & 1.03 & @xmath177 & ... & 29.5 & 31.1 & ... & ... & 37.0 & ... & 39.1 & ... + cr i & 4651.29 & 0.98 & @xmath178 & ... & 9.1 & ... & ... & ... & ... & ... & ... & ... + cr i & 4652.16 & 1.00 & @xmath179 & ... & 19.7 & ... & ... & ... & ... & ... & ... & ... + cr i & 5208.44 & 0.94 & @xmath180 & 30.3 & 88.1 & 87.7 & 100.8 & 38.7 & ... & ... & 94.6 & 95.8 + cr
i & 5298.27 & 0.98 & @xmath181 & ... & 18.4 & ... & ... & ... & ... & ... & ... & ... + cr i & 5409.79 & 1.03 & @xmath182 & ... & 32.6 & 33.4 & 67.4 & ... & 41.9 & ... & ... & 33.5 + fe i & 4383.54 & 1.49 & @xmath183 & ... & ... & ... & ... & ... & ... & ... & 161.4 & ... + fe i & 4407.71 & 2.18 & @xmath184 & ... & 38.5 & 50.1 & ... & ... & 43.5 & ... & ... & ... + fe i & 4408.41 & 2.20 & @xmath185 & ... & 43.6 & 41.0 & ... & ... & 44.2 & ... & ... & ... + fe i & 4415.12 & 1.61 & @xmath186 & ... & ... & ... & ... & ... & ... & 38.6 & ... & ... + fe i & 4422.57 & 2.85 & @xmath187 & ... & 36.3 & 40.2 & ... & ... & ... & ... & ... & ... + fe i & 4430.61 & 2.22 & @xmath188 & 8.9 & 40.8 & 44.1 & ... & ... & 42.3 & ... & 35.9 & ... + fe i & 4433.22 & 3.65 & @xmath189 & ... & 16.3 & ... & ... & ... & ... & ... & ... & ... + fe i & 4442.34 & 2.20 & @xmath190 & ... & 65.9 & 65.7 & ... & ... & 74.7 & ... & 56.3 & 86.5 + fe i & 4443.19 & 2.86 & @xmath191 & ... & 31.7 & 27.0 & ... & ... & 47.6 & ... & ... & 20.5 + fe i & 4447.72 & 2.22 & @xmath192 & 13.0 & 57.2 & 67.6 & ... & ... & 63.5 & ... & 54.6 & 68.1 + fe i & 4461.65 & 0.09 & @xmath1463.21 & 39.0 & 88.8 & 97.0 & ... & ... & 60.6 & ... & 81.5 & 150.4 + fe i & 4466.55 & 2.83 & @xmath1460.60 & 16.2 & 67.0 & 63.8 & ... & ... & 67.2 & ... & 57.9 & 81.3 + fe i & 4476.02 & 2.85 & @xmath1460.82 & ... & 52.4 & 52.8 & ... & ... & ... & ... & 48.8 & ... + fe i & 4484.22 & 3.60 & @xmath1460.86 & ... & 15.6 & 18.9 & ... & ... & ... & ... & ... & ... + fe i & 4489.74 & 0.12 & @xmath1463.97 & 12.1 & 54.0 & 66.6 & ... & ... & 31.3 & ... & 39.2 & 106.5 + fe i & 4494.56 & 2.20 & @xmath1461.14 & 20.8 & 67.2 & 71.7 & ... & ... & 65.8 & ... & 62.1 & ... + fe i & 4528.61 & 2.18 & @xmath1460.82 & 33.8 & 86.3 & 83.2 & ... & ... & ... & ... & 80.2 & 104.7 + fe i & 4531.15 & 1.49 & @xmath1462.15 & 14.6 & 60.2 & 65.5 & ... & ... & 57.4 & ... & 55.7 & 87.1 + fe i & 4592.65 & 1.56 & @xmath1462.45 & ... & 43.1 & 51.3 & ... & ... & 43.3 & ... & 32.8 & 50.3 + fe i & 4602.94 & 1.49 & @xmath1462.21 & 16.3 & 60.5 & 71.1 & ... & ... & 49.4 & ... & 45.0 & 75.2 + fe i & 4733.59 & 1.49 & @xmath1462.99 & ... & 22.0 & 27.0 & ... & ... & 23.4 & ... & ... & ... + fe i & 4736.77 & 3.21 & @xmath1460.75 & ... & 35.2 & 34.4 & ... & ... & 48.0 & ... & 35.6 & ... + fe i & 4871.32 & 2.87 & @xmath1460.36 & 28.3 & 66.2 & 72.5 & ... & 23.3 & 77.9 & ... & 59.5 & 62.7 + fe i & 4872.14 & 2.88 & @xmath1460.57 & 14.5 & 54.0 & 49.4 & ... & ... & 68.1 & ... & 48.8 & 48.6 + fe i & 4890.75 & 2.88 & @xmath1460.39 & 18.5 & 65.6 & 69.0 & ... & 28.6 & 87.9 & ... & 65.6 & 68.8 + fe i & 4891.49 & 2.85 & @xmath1460.11 & 33.1 & 80.9 & 80.2 & ... & 41.3 & 93.3 & ... & 74.3 & 74.8 + fe i & 4903.31 & 2.88 & @xmath1460.93 & ... & 40.6 & 46.6 & ... & ... & 57.8 & ... & 32.9 & 31.2 + fe i & 4918.99 & 2.87 & @xmath1460.34 & 21.3 & 68.4 & 67.6 & ... & ... & 82.6 & ... & 71.2 & 63.3 + fe i & 4920.50 & 2.83 & 0.07 & 39.5 & 86.9 & 91.2 & ... & 42.6 & ... & 21.5 & 86.4 & ... + fe i & 4924.77 & 2.28 & @xmath1462.26 & ... & 19.0 & 16.8 & ... & ... & 22.7 & ... & ... & 21.8 + fe i & 4938.81 & 2.88 & @xmath1461.08 & ... & 32.7 & 32.0 & ... & ... & 49.5 & ... & 38.2 & 17.7 + fe i & 4939.69 & 0.86 & @xmath1463.34 & ... & 41.3 & 51.5 & ... & ... & 28.8 & ... & 44.9 & 70.8 + fe i & 4994.13 & 0.92 & @xmath1462.96 & ... & 52.0 & 56.3 & ... & ... & 36.4 & ... & 43.3 & 79.5 + fe i & 5006.12 & 2.83 & @xmath1460.61 & ... & 60.6 & 59.9 & ... & ... & 72.4 & ... & 52.7 & 56.4 + fe i & 5012.07 & 0.86 & @xmath1462.64 & 27.2 & 79.0 & 87.2 & ... & 37.2 & 59.8 & ... & 69.4 & 125.2 + fe i & 5041.76 & 1.49 & @xmath1462.20 & ... & 67.7 & 69.1 & ... & ... & ... & ... & 47.8 & 85.1 + fe i & 5049.82 & 2.28 & @xmath1461.34 & ... & 54.0 & 58.5 & ... & 14.9 & 61.5 & ... & 48.4 & 57.5 + fe i & 5051.63 & 0.92 & @xmath1462.80 & 17.6 & 68.7 & 81.1 & ... & ... & 53.4
& ... & 63.1 & 105.9 + fe i & 5068.77 & 2.94 & @xmath1461.04 & ... & 31.2 & 30.8 & ... & ... & 45.8 & ... & 35.3 & ... + fe i & 5083.34 & 0.96 & @xmath1462.96 & ... & 56.7 & 59.1 & ... & ... & 44.5 & ... & 44.0 & 96.4 + fe i & 5110.41 & 0.00 & @xmath1463.76 & ... & 77.1 & 85.3 & ... & ... & 50.8 & ... & ... & 144.0 + fe
i & 5123.72 & 1.01 & @xmath1463.07 & ... & 48.5 & 61.8 & ... & ... & 47.8 & ... & 31.8 & 85.3 + fe i & 5127.36 & 0.92 & @xmath1463.31 & ... & 41.2 & 46.0 & ... & ... & 33.1 & ... & 29.6 & 68.9 + fe i & 5133.69 & 4.18 & 0.14 & ... & 28.4 & 21.9 & ... & ... & 59.6 & ... & 21.0 & ... + fe i & 5150.84 & 0.99 & @xmath1463.00 & ... & 45.9 & 53.5 & ... & ... & 30.2 & ... & 37.9 & 69.1 + fe i & 5151.91 & 1.01 & @xmath1463.32 & ... & 36.6 & 37.7 & ... & ... & 23.5 & ... & 34.1 & 54.8 + fe i & 5162.27 & 4.18 & 0.02 & ... & 25.9 & 17.9 & ... & ... & 46.3 & ... & ... & ... + fe i & 5166.28 & 0.00 & @xmath1464.20 & ... & 51.2 & 67.5 & 72.8 & ... & 26.4 & ... & 41.1 & 111.2 + fe i & 5171.60 & 1.49 & @xmath1461.79 & 32.5 & 80.5 & 91.1 & 96.1 & 37.3 & 72.9 & ... & 74.8 & 112.0 + fe i & 5191.46 & 3.04 & @xmath1460.55 & ... & 50.1 & 53.4 & ... & ... & 70.5 & ... & 37.0 & 49.5 + fe i & 5192.34 & 3.00 & @xmath1460.42 & ... & 59.9 & 61.3 & 79.5 & ... & 73.0 & ... & 47.7 & 49.2 + fe i & 5194.94 & 1.56 & @xmath1462.09 & ... & 60.8 & 67.9 & 75.2 & ... & 53.3 & ... & 56.9 & ... + fe i & 5216.27 & 1.61 & @xmath1462.15 & 16.6 & 51.8 & 65.9 & 65.7 & 21.0 & 50.1 & ... & 51.7 & 75.1 + fe i & 5232.94 & 2.94 & @xmath1460.06 & 29.7 & 74.5 & 78.5 & 90.5 & 41.9 & 89.4 & ... & 72.8 & 75.7 + fe i & 5247.05 & 0.09 & @xmath1464.95 & ... & ... & 19.4 & ... & ... & ... & ... & ... & 35.5 + fe i & 5254.96 & 0.11 & @xmath1464.76 & ... & 20.8 & 32.8 & 30.7 & ... & ... & ... & ... & 58.7 + fe i & 5266.56 & 3.00 & @xmath1460.39 & ... & 59.9 & 62.9 & 71.7 & 20.2 & 72.0 & ... & 54.9 & 51.6 + fe i & 5269.54 & 0.86 & @xmath1461.32 & ... & 133.8 & 138.5 & 149.1 & 78.8 & 118.5 & ... & 130.5 & 186.9 + fe i & 5281.79 & 3.04 & @xmath1460.83 & ... & 34.7 & 34.4 & 41.5 & ... & 50.8 & ... & 27.1 & ... + fe i & 5302.30 & 3.28 & @xmath1460.72 & ... & 26.2 & 19.8 & ... & ... & 41.3 & ... & ... & ... + fe i & 5324.18 & 3.21 & @xmath1460.10 & ... & 56.7 & 59.0 & 74.1 & ... & 75.4 & ... & 60.4 & 56.1 + fe i & 5328.04 & 0.92 & @xmath1461.47 & ... & 120.7 & 130.6 & 131.2 & 86.8 & 111.8 & 17.0 & 114.1 & 172.9 + fe i & 5332.90 & 1.56 & @xmath1462.78 & ... & 27.0 & 27.5 & ... & ... & ... & ... & ... & 38.6 + fe i & 5339.93 & 3.27 & @xmath1460.65 & ... & 29.6 & 29.3 & 44.5 & ... & 47.9 & ... & 33.0 & ... + fe i & 5341.02 & 1.61 & @xmath1461.95 & ... & 66.9 & 79.1 & 67.1 & ... & 60.3 & ... & 58.0 & ... + fe i & 5367.47 & 4.41 & 0.44 & ... & 25.2 & 19.2 & 50.7 & ... & 44.6 & ... & ... & ... + fe i & 5369.96 & 4.37 & 0.54 & ... & 29.5 & 26.3 & 50.3 & ... & 49.1 & ... & ... & ... + fe i & 5371.49 & 0.96 & @xmath1461.64 & ... & 109.5 & 123.8 & 134.0 & 81.6 & 102.8 & ... & ... & 167.4 + fe i & 5383.37 & 4.31 & 0.64 & ... & 31.9 & 34.6 & 50.6 & ... & 60.9 & ... & ... & ... + fe i & 5397.13 & 0.92 & @xmath1461.99 & ... & 107.4 & ... & 101.8 & 72.7 & 88.5 & ... & ... & 156.2 + fe i & 5404.15 & 4.43 & 0.52 & ... & 30.0 & ... & ... & ... & 68.0 & ... & ... & ... + fe i & 5405.77 & 0.99 & @xmath1461.84 & ... & 106.6 & 110.0 & 122.3 & 66.0 & 93.6 & ... & ... & 155.4 + fe i & 5415.20 & 4.39 & 0.64 & ... & 29.2 & 23.3 & 51.9 & ... & 51.2 & ... & ... & ... + fe i & 5424.07 & 4.32 & 0.52 & ... & 37.6 & 33.5 & ... & ... & 61.8 & ... & ... & ... + fe i & 5429.70 & 0.96 & @xmath1461.88 & 73.5 & 106.6 & 117.7 & 108.9 & ... & 89.5 & ... & ... & 161.3 + fe i & 5434.52 & 1.01 & @xmath1462.12 & 48.6 & 91.4 & 98.1 & 100.7 & ... & 85.3 & ... & ... & ... + fe i & 5569.62 & 3.42 & @xmath1460.54 & ... & 23.6 & 28.0 & ... & ... & 47.2 & ... & 17.8 & 20.9 + fe i & 5572.84 & 3.40 & @xmath1460.28 & ... & 44.8 & 38.9 & ... & ... & 59.4 & ... & 36.4 & 32.8 + fe i & 5576.09 & 3.43 & @xmath1461.00 & ... & 21.4 & 16.4 & ... & ... & 29.7 & ... & ... & ... + fe i & 5586.75 & 3.37 & @xmath1460.10 & 13.6 & 44.8 & 53.0 & 72.2 & ... & 70.8 & ... & 43.2 & ... + fe i & 5615.64 & 3.33 & 0.05 & 30.8 & 59.2 & 66.6 & ... & 21.3 & 81.9 & ... & 62.0 & 57.5 + fe i & 5624.54 & 3.42 & @xmath1460.75 & ... & 19.2 & 21.9 & ... & ... & 33.2 & ... & 15.7 & ... + fe i & 6065.48 & 2.61 & @xmath1461.53 & ... & 30.0 & 31.0 & ... & ... & 31.6 & ... & ... & 33.4 + fe i & 6136.61 & 2.45 & @xmath1461.40 & ... & ... & 46.4 & ... & ... & ... & ... & 44.1 & ... + fe i & 6137.69 & 2.59 & @xmath1461.40 & ... & ... & ... & ... & ... & ... & ... & 29.7 & ... + fe i & 6191.56 & 2.43 & @xmath1461.42 & ... & 42.4 & 48.7 & ... & ... & 47.9 & ... & 36.4 & ... + fe i & 6213.43 & 2.22 & @xmath1462.48 & ... & 10.1 & 10.5 & ... & ... & ... & ... & ... & ... + fe i & 6219.28 & 2.20 & @xmath1462.43 & ... & 14.0 & 21.6 & ... & ... & ... & ... & ... & ... + fe i & 6230.72 & 2.56 & @xmath1461.28 & ... & 46.7 & 48.3 & 60.2 & ... & 53.9 & ... & 41.2 & 52.2 + fe i & 6393.60 & 2.43 & @xmath1461.43 & ... & 42.3 & 43.8 & ... & ... & 44.1 & ... & ... & ... + ni i & 3858.30 & 0.42 & @xmath1460.95 & ... & ... & ...
& ... & ... & ... & 26.2 & ... & ... + ni i & 4714.42 & 3.38 & 0.23 & ... & 25.6 & 27.3 & ... & ... & 43.5 & ... & 29.6 & ... + ni i & 4980.16 & 3.61 & @xmath1460.11 & ... & 12.8 & ... & ... & ... & 23.9 & ... & 13.1 & ... + zn i & 4722.15 & 4.03 & @xmath1460.39 & ... & 10.7 & ... & ... & ... & 44.7 & ... & 20.4 & ... + zn i & 4810.53 & 4.08 & @xmath1460.17 & ... & 14.7 & 21.4 & 33.3 & ... & 26.4 & ... & 28.8 & ... + sc ii & 4246.82 & 0.31 & 0.24 & ... & ... & ... & ... & ... & ... & 35.2 & ... & ... + sc ii & 4320.75 & 0.61 & @xmath1460.25 & ... & ... & ... & ... & ... & ... & ... & 53.3 & ... + sc ii & 4400.39 & 0.61 & @xmath1460.54 & ... & ... & ... & ... & ... & ... & ... & 36.1 & ... + sc ii & 4415.54 & 0.60 & @xmath1460.67 & ... & 51.8 & 57.4 & ... & 33.1 & 55.1 & ... & 39.4 & 63.7 +
sc ii & 5031.01 & 1.36 & @xmath1460.40 & ... & 23.1 & 26.7 & ... & ... & 29.7 & ... & ... & 37.1 + sc ii & 5526.79 & 1.77 & 0.02 & ... & 21.8 & 30.0 & ... & ... & 39.5 & ... & ... & 36.4 + ti ii & 3913.48 & 1.12 & @xmath1460.53 & ... & ... & ... & ... & ... & ... & 46.8 & ... & ... + ti ii & 4417.71 & 1.16 & @xmath1461.43 & 36.9 & 80.7 & 88.6 & ... & 48.3 & 71.1 & ... & 74.5 & 89.4 + ti ii & 4418.31 & 1.24 & @xmath1461.99 & ... & 38.0 & 47.9 & ... & ... & 28.9 & ... & 31.6 & 31.4 + ti ii & 4441.73 & 1.18 & @xmath1462.41 & ... & 26.7 & ... & ... & ... & 29.7 & ... & ... & ... + ti ii & 4443.77 & 1.08 & @xmath1460.70 & 59.5 & 103.8 & 99.9 & ... & 76.1 & 102.2 & 41.5 & 89.8 & 128.6 + ti ii & 4444.54 & 1.12 & @xmath1462.21 & ... & 32.2 & 31.1 & ... & ... & ... & ... & ... & 48.3 + ti ii & 4450.50 & 1.08 & @xmath1461.51 & 24.7 & 73.8 & 77.1 & ... & 45.8 & 63.9 & ... & 65.5 & 86.6 + ti ii & 4464.46 & 1.16 & @xmath1462.08 & ... & 54.8 & 61.5 & ... & ... & 37.0 & ... & 49.0 & ... + ti
ii & 4468.52 & 1.13 & @xmath1460.60 & 60.8 & 102.7 & 111.7 & ... & 88.9 & 99.4 & ... & 104.5 & 142.2 + ti ii & 4501.27 & 1.12 & @xmath1460.76 & 56.0 & 97.1 & 103.3 & ... & 71.4 & 93.4 & ... & 93.6 & 116.3 + ti ii & 4529.48 & 1.57 & @xmath1462.03 & ... & 35.0 & 40.6 & ... & ... & 31.1 & ... & 25.7 & ... + ti ii & 4533.97 & 1.24 & @xmath1460.77 & 53.4 & 103.6 & 108.0 & ... & 78.7 & 101.0 & 33.4 & 95.9 & 120.6 + ti ii & 4563.77 & 1.22 & @xmath1460.96 & 48.0 & 91.3 & 99.5 & ... & 63.6 & 88.4 & 27.6 & 89.2 & 99.1 + ti ii & 4571.96 & 1.57 & @xmath1460.53 & 46.9 & 96.2 & 102.5 & ... & 62.8 & 100.6 & 29.4 & 87.8 & 119.8 + ti ii & 4589.92 & 1.24 & @xmath1461.79 & ... & 58.0 & 56.4 & ... & ... & 51.1 & ... & 41.8 & 58.6 + ti ii & 4779.98 & 2.05 & @xmath1461.37 & ... & 29.1 & 29.9 & ... & ... & 29.8 & ... & 25.9 & ... + ti ii & 4798.51 & 1.08 & @xmath1462.67 & ... & 20.7 & 25.6 & ... & ... & ... & ... & ... & ... + ti
ii & 4805.09 & 2.06 & @xmath1461.10 & ... & 41.2 & 43.3 & ... & ... & 45.9 & ... & 28.1 & 32.3 + ti ii & 5129.15 & 1.89 & @xmath1461.39 & ... & 33.0 & 32.4 & ... & ... & 33.1 & ... & ... & 30.6 + ti ii & 5185.90 & 1.89 & @xmath1461.35 & ... & 26.0 & 23.4 & 46.8 & ... & 27.9 & ... & 28.1 & ... + ti ii & 5336.78 & 1.58 & @xmath1461.63 & ... & 37.8 & 35.4 & 78.4 & ... & 34.5 & ... & 29.4 & 37.5 + ti ii & 5381.01 & 1.57 & @xmath1461.97 & ... & 22.7 & 20.9 & 58.5 & ... & 24.9 & ... & ... & ... + ti ii & 5418.75 & 1.58 & @xmath1462.11 & ... & 18.5 & 14.2 & 66.0 & ... & 13.8 & ... & ... & ... + fe ii & 4416.82 & 2.78 & @xmath1462.60 & ... & 34.6 & 36.5 & ... & ... & 55.9 & ... & 25.8 & ... + fe ii & 4508.28 & 2.86 & @xmath1462.58 & ... & 42.6 & 43.3 & ... & ... & 48.1 & ... & 38.6 & ... + fe ii & 4515.34 & 2.84 & @xmath1462.48 & ... & 29.8 & 35.3 & ... & ... & 40.9 & ... & 23.6 & ... + fe ii & 4520.23 & 2.81 & @xmath1462.60 & ... & 30.4 & 34.7 & ... & ... & 42.1 & ... & 24.3 & ... + fe ii & 4541.52 & 2.86 & @xmath1463.05 & ... & 17.7 & ... & ... & ... & ... & ... & ... & ... + fe ii & 4582.83 & 2.84 & @xmath1463.10 & ... & 11.2 & ... & ... & ... & ... & ... & ... & ... + fe ii & 4583.83 & 2.81 & @xmath1462.02 & 20.2 & 65.0 & 67.4 & ... & 34.4 & 77.2 & 12.0 & 53.8 & 61.4 + fe ii & 4731.44 & 2.89 & @xmath1463.36 & ... & 13.1 & ... & ... & ... & 23.4 & ... & ... & ... + fe ii & 4923.93 & 2.89 & @xmath1461.32 & 39.5 & 83.2 & 93.5 & ... & 43.7 & 93.1 & 23.6 & 77.1 & 88.1 + fe ii & 5018.45 & 2.89 & @xmath1461.22 & 48.7 & 92.5 & 97.3 & ... & 69.6 & 100.0 & ... & 93.5 & 97.5 + fe ii & 5169.00 & 2.89 & @xmath1460.87 & ... & ... & ... & ... & ... & ... & 37.2 & ... & ... + fe ii & 5197.56 & 3.23 & @xmath1462.10 & ... & ... & 32.8 & ... & ... & ... & ... & 21.3 & ... + fe ii & 5234.62 & 3.22 & @xmath1462.27 & ... & 32.2 & 33.7 & 45.7 & ... & 44.6 & ... & 29.6 & ... + fe ii & 5276.00 & 3.20 & @xmath1461.94 & ... & 36.7 & 38.1 & 57.0 & ... & 53.3 & ... & 44.0 & 27.4 + fe ii & 5534.83 & 3.24 & @xmath1462.93 & ... & ... & ... & ... & ... & 19.9 & ... & ... & ... + ba ii & 4130.65 & 2.72 & 0.56 & ... & ... & ... & ... & ... & ... & 54.2 & ... & ... + ba ii & 4554.03 & 0.00 & 0.16 & 43.8 & 112.1 & 125.1 & ... & 111.9 & 209.2 & 179.8 & 78.5 & ... + ba ii & 4934.09 & 0.00 & @xmath1460.16 & 28.6 & 106.5 & 121.4 & ... & 120.5 & 177.4 & 170.1 & 67.1 & 197.4 + ba ii & 5853.67 & 0.60 & @xmath1461.02 & ... & 13.9 & 24.1 & 151.2 & 29.1 & 80.7 & 68.2 & ... & 98.7 + ba ii & 6141.70 & 0.70 & @xmath1460.07 & ... & ... & ... & ... & ... & ... & ... & 21.0 & ... + ba ii & 6496.90 & 0.60 & @xmath1460.37 & ... & 48.4 & 57.0 & 239.9 & 77.3 & 131.9 & 113.0 & ... & 161.8 + lcccccccccccc na i & 5682.63 & 2.10 & @xmath1460.70 & ... & ... & ... & ... & ... & ... & 36.0 & ... & ... + na i & 5688.20 & 2.10 & @xmath1460.46 & 14.9 & ... & ... & ... & 34.2 & ... & 60.6 & ... & ... + na i & 5889.95 & 0.00 & 0.10 & 156.6 & 166.8 & 218.6 & 200.6 & 225.6 & 186.9 & 295.2 & 154.9 & 202.6 + na i & 5895.92 & 0.00 & @xmath1460.20 & 151.3 & 151.6 & 202.3 & 178.0 & 216.0 & 169.0 & 246.3 & 143.2 & 183.4 + mg i & 4571.10 & 0.00 & @xmath1465.69 & 37.0 & 64.3 & ... & ... & 80.1 & 55.9 & 103.4 & ... & ... + mg i & 4702.99 & 4.35 & @xmath1460.52 & ... & 73.3 & ... & ... & ... & 46.0 & ... & ... & ... + mg i & 5172.69 & 2.71 & @xmath1460.38 & 177.0 & 180.1 & 177.6 & 175.3 & ... & 173.2 & ... & 133.1 & 189.0 + mg i & 5183.60 & 2.72 & @xmath1460.16 & ... & ... & 229.4 & 236.9 & ... & ... & ... & 144.5 & ... + mg i & 5528.40 & 4.35 & @xmath1460.49 & ... & 77.6 & ... & 110.0 & 102.0 & 77.1 & 163.4 & 44.5 & ... + ca i & 4425.44 & 1.88 & @xmath1460.36 & ... & 49.5 & ... & ... & 71.9 & ... & ... & ... & ... + ca i & 4455.89 & 1.90 & @xmath1460.53 & ... & 42.5 & ... & ... & 81.3 & 59.4 & ... & ... & ... + ca i & 5265.56 & 2.52 & @xmath1460.11 & ... & 23.4 & ... & ... & ... & 27.8 & ... & ... & ... + ca i & 5581.97 & 2.52 & @xmath1460.56 & ... & ... & ... & ... & 20.4 & ... & ... & ... & ... + ca i & 5588.76 & 2.53 & 0.36 & ... & 43.0 & ... & 55.1 & 62.0 & 48.9 & 117.2 & ... & 50.2 + ca i & 5594.47 & 2.52 & 0.10 & 54.0 & 34.6 & 30.0 & 54.6 & 63.2 & 38.3 & 128.2 & ... & 46.5 + ca i & 5601.29 & 2.53 & @xmath1460.52 & ... & 12.2 & ... & ... & 16.8 & 17.5 & ... & ... & ... + ca i & 5857.45 & 2.93 & 0.24 & 28.1 & ... & ... & 32.0 & 39.0 & 24.8 & 98.2 & ... & ... + ca i & 6102.72 & 1.88 & @xmath1460.77 & ... & ... & ... & 24.9 & ... & 40.5 & ... & ... & ... + ca i & 6122.22 & 1.89 & @xmath1460.32 & 67.2 & 61.7 & ... & ... & 70.5 & 63.8 & 137.7 & ... & ... + ca i & 6162.17 & 1.90 & @xmath1460.09 & 78.1 & 73.1 & ... & ... & 99.1 & 73.0 & 159.2 & 45.3 & ... + ti i & 4533.24 & 0.85 & 0.53 & ... & 52.6 & ... & ... & 50.4 & 47.8 & 84.9 & ... & ... + ti i & 4981.73 & 0.85 & 0.56 & ... & 54.2 & ... & ... & 59.2 & 55.2 & 128.3 & 28.0 & ... + ti i & 4991.07 & 0.84 & 0.44 & 49.8 & 44.5 & ... & ... & 63.7 & 54.0 & 126.1 & ... & ... + ti i & 4999.50 & 0.83 & 0.31 & ... & 39.2 & ... & ... & ... & 57.2 & ... & ... & ... + ti i & 5007.21 & 0.82 & 0.17 & ... & 39.0 & ... & ... & 59.8 & 49.8 & 130.6 & ... & ... + ti i & 5016.16 & 0.85 & @xmath1460.52 & ... & ... & ... & ... & 18.6 & ... & ... & ... & ... + ti i & 5173.74 & 0.00 & @xmath1461.06 & ... & 18.9 & ... & ... & 34.9 & 37.7 & ... & ... & ... + ti i & 5192.97 & 0.02 & @xmath1460.95 & ... & 25.9 & 25.5 & 39.2 & ... & 52.0 & ... & ... & ... + ti i & 5206.08 & 2.49 & 1.07 & ... & ... & 60.7 & ... & ... & ... & ... & ... & ... + ti i & 5210.38 & 0.05 & @xmath1460.83 & 33.2 & 29.5 & 23.4 & 46.6 & 37.7 & 35.2 & 102.3 & ... & 34.3 + cr i & 4616.14 & 0.98 & @xmath1461.19 & ... & ... & ... & ... & 29.2 & ... & ... & ... & ... + cr i & 4626.19 & 0.97 & @xmath1461.32 & ... & ... & ... & ... & 25.2 & ... & ... & ... & ... + cr i & 4646.17 & 1.03 & @xmath1460.70 & ... & ... & ... & ... & 50.9 & 29.4 & ... & ... & ... + cr i & 4651.29 & 0.98 & @xmath1461.46 & ... & ... & ... & ... & 27.3 & ... & ... & ... & ... + cr i & 4652.16 & 1.00 & @xmath1461.03 & ... & ... & ... & ... & 31.9 & 22.3 & ... & ... & ... + cr i & 5208.44 & 0.94 & 0.16 & 88.9 & ... & 74.0 & 102.3 & 109.9 & ... & ... & 39.3 & 93.3 + cr i & 5298.27 & 0.98 & @xmath1461.15 & ... & ... & ... & ... & 25.1 & ... & ... & ... & ... + cr i & 5409.79 & 1.03 & @xmath1460.72 & 41.2 & 35.7 & ... & 57.1 & 56.0 & 38.6 & 111.6 & ... & ... + fe i & 4408.41 & 2.20 & @xmath1461.77 & ... & 32.3 & ... & ... & ... & ... & ... & ... & ... + fe i & 4422.57 & 2.85 & @xmath1461.11 & ... & 29.9 & ... & ... & ... & ... & ... & ... & ... + fe i & 4430.61 & 2.22 & @xmath1461.66 & ... & 43.4 & ... & ... & 63.2 & ... & ... & ... & ... + fe i & 4442.34 & 2.20 & @xmath1461.25 & ... & 64.6 & ... & ... & ... & ... & ... & ... & ... + fe i & 4443.19 & 2.86 & @xmath1461.04 & ... & 33.8 & ... & ... & ... & ... & ... & ... & ... + fe i & 4447.72 & 2.22 & @xmath1461.34 & ... & 53.7 & ... & ... & 83.0 & ... & ... & ... & ... + fe i & 4461.65 & 0.09 & @xmath1463.21 & ... & 92.9 & ... & ... & 96.7 & 106.7 & ... & ... & ... + fe i & 4466.55 & 2.83 & @xmath1460.60 & 60.4 & 58.8 & ... & ... & 81.0 & 58.8 & 109.7 & ... & ... + fe i & 4476.02 & 2.85 & @xmath1460.82 & 54.9 & 54.3 & ... & ... & 88.1 & 53.1 & 127.8 & ... & ... + fe i & 4484.22 & 3.60 & @xmath1460.86 & ... & ... & ... & ... & ... & ... & 62.5 & ... & ... + fe i & 4489.74 & 0.12 & @xmath1463.97 & ... & 57.5 & ... & ... & 60.8 & 66.3 & 104.0 & ... & ... + fe i & 4494.56 & 2.20 & @xmath1461.14 & 71.7 & 71.5 & ... & ... & 83.7 & 75.5 & ... & ... & ... + fe i & 4528.61 & 2.18 & @xmath1460.82 & ... & 92.2 & ... & ... & ... & 111.4 & ... & 71.7 & ... + fe i & 4531.15 & 1.49 & @xmath1462.15 & ... & 64.8 & ... & ... & ... & 70.1 & ... & 47.0 & ... + fe i & 4592.65 & 1.56 & @xmath1462.45 & 33.4 & 43.2 & ... & ... & 71.1 & 56.3 & 111.7 & ... & ... + fe i & 4602.94 & 1.49 & @xmath1462.21 & ... & 59.3 & ... & ... & 68.4 & 63.7 & 101.9 & ... & ... + fe i & 4733.59 & 1.49 & @xmath1462.99 & ... & ... & ... & ... & 38.0 & ... & ... & ... & ... + fe i & 4736.77 & 3.21 & @xmath1460.75 & ... & 34.8 & ... & ... & 54.7 & ... & ... & ... & ... + fe i & 4871.32 & 2.87 & @xmath1460.36 & 82.7 & 68.7 & ... & ... & 82.5 & 71.8 & 133.3 & ... & ... + fe i & 4872.14 & 2.88 & @xmath1460.57 & ... & 55.4 & ... & ... & 70.8 & 68.7 & 142.4 & ... & ... + fe i & 4890.75 & 2.88 & @xmath1460.39 & ... & 65.0 & ... & ... & 96.2 & 77.3 & ... & ... & ... + fe i & 4891.49 & 2.85 & @xmath1460.11 & ... & 86.6 & ... & ... & 86.8 & 95.3 & 158.4 & 54.2 & ... + fe i & 4903.31 & 2.88 & @xmath1460.93 & ... & 36.6 & ... & ... & 59.3 & 43.0 & 92.4 & ... & ... + fe i & 4918.99 & 2.87 & @xmath1460.34 & ... & 61.5 & ... & ... & 79.7 & 77.8 & 134.1 & ... & ... + fe i & 4920.50 & 2.83 & 0.07 & ... & 92.5 & ... & ... & ... & ... & 197.9 & 53.6 & ... + fe i & 4924.77 & 2.28 & @xmath1462.26 & ... & 17.2 & ... & ... & 28.2 & 33.0 & 83.3 & ... & ... + fe i & 4938.81 & 2.88 & @xmath1461.08 & ... & 39.6 & ... & ... & 48.3 & 34.6 & 85.7 & ... & ... + fe i & 4939.69 & 0.86 & @xmath1463.34 & ... & 51.8 & ... & ... & 44.1 & 68.8 & ... & ... & ... + fe i & 4994.13 & 0.92 & @xmath1462.96 & 60.3 & 52.8 & ... & ... & 64.5 & 64.0 & 100.1 & ... & ... + fe i & 5006.12 & 2.83 & @xmath1460.61 & ... & 57.5 & ... & ... & 71.6 & 62.0 & ... & 48.2 & ... + fe i & 5012.07 & 0.86 & @xmath1462.64 & 64.6 & 87.1 & ... & ... & 79.6 & 79.5 & 131.6 & ... & ... + fe i & 5041.76 & 1.49 & @xmath1462.20 & ... & 73.3 & ... & ... & 81.9 & 78.2 & ... & 16.7 & ... + fe i & 5049.82 & 2.28 & @xmath1461.34 & ... & 56.6 & ... & ... & 73.7 & 60.3 & ... & ... & ... + fe i & 5051.63 & 0.92 & @xmath1462.80 & ... & 76.4 & ... & ... & 77.3 & 77.0 & 121.0 & 26.9 & ... + fe i & 5068.77 & 2.94 & @xmath1461.04 & ... & 29.0 & ... & ... & 54.9 & 31.2 & ... & ... & ... + fe i & 5083.34 & 0.96 & @xmath1462.96 & ... & 63.8 & ... & ... & 77.2 & 66.1 & ... & ... & ... + fe i & 5110.41 & 0.00 & @xmath1463.76 & ... & 91.8 & ... & ... & ... & 88.6 & ... & ... & ... + fe i & 5123.72 & 1.01 & @xmath1463.07 & ... & 51.5 & ... & ... & ... & 75.1 & ... & ... & ... + fe i & 5127.36 & 0.92 & @xmath1463.31 & ... & 49.5 & ... & ... & 52.2 & 44.4 & ... & ... & ... + fe i & 5133.69 & 4.18 & 0.14 & ... & 24.3 & ... & ... & 51.4 & 36.8 & ... & ... & ... + fe i & 5150.84 & 0.99 & @xmath1463.00 & ... & 53.5 & ... & ... & 48.4 & 50.6 & ... & ... & ... + fe i & 5151.91 & 1.01 & @xmath1463.32 & ... & 40.4 & ... & ... & 45.3 & 48.4 & ... & ... & ... + fe i & 5162.27 & 4.18 & 0.02 & ... & 27.7 & ... & ... & 43.9 & 17.1 & ... & ... & ... + fe i & 5166.28 & 0.00 & @xmath1464.20 & 49.0 & 58.5 & ... & 60.0 & 58.8 & 61.8 & 104.9 & ... & 77.2 + fe i & 5171.60 & 1.49 & @xmath1461.79 & 73.3 & 84.5 & 69.4 & 81.8 & 94.3 & 89.8 & 127.5 & 69.3 & 91.2 + fe i & 5191.46 & 3.04 & @xmath1460.55 & ... & 54.7 & ... & ... & 100.2 & 68.9 & ... & ... & ... + fe i & 5192.34 & 3.00 & @xmath1460.42 & 55.1 & 57.9 & 44.1 & 75.9 & ... & 71.4 & ... & 28.1 & ... + fe i & 5194.94 & 1.56 & @xmath1462.09 & 54.0 & 73.7 & 42.5 & 78.2 & ... & 71.7 & ... & ... & ... + fe i & 5216.27 & 1.61 & @xmath1462.15 & 53.6 & 55.9 & 34.9 & 57.8 & 68.1 & 68.2 & 118.3 & ... & 51.7 + fe i & 5232.94 & 2.94 & @xmath1460.06 & 71.6 & 70.7 & 64.5 & 79.9 & ... & 78.7 & 139.7 & 43.2 & 105.0 + fe i & 5247.05 & 0.09 & @xmath1464.95 & ... & ... & ... & ... & ... & 16.3 & 53.1 & ... & ... + fe i & 5254.96 & 0.11 & @xmath1464.76 & 12.6 & 24.7 & ... & ... & ... & 30.5 & 81.4 & ... & ... + fe i & 5266.56 & 3.00 & @xmath1460.39 & 53.0 & 56.7 & 44.2 & 66.8 & 81.0 & 65.1 & 121.2 & 28.0 & 71.3 + fe i & 5269.54 & 0.86 & @xmath1461.32 & 115.8 & 138.4 & 123.3 & 139.8 & 143.3 & 139.5 & ... & 88.0 & 142.9 + fe i & 5281.79 & 3.04 & @xmath1460.83 & 30.8 & 37.3 & ... & 43.2 & 52.4 & 36.7 & 96.8 & ... & 50.3 + fe i & 5302.30 & 3.28 & @xmath1460.72 & ... & 26.9 & ... & ... & 42.5 & 30.5 & ... & ... & ... + fe i & 5324.18 & 3.21 & @xmath1460.10 & 60.5 & 58.7 & 47.1 & ... & 83.4 & 60.9 & 122.3 & ... & 74.6 + fe i & 5328.04 & 0.92 & @xmath1461.47 & 107.5 & 127.1 & 108.7 & 126.2 & 128.8 & 130.1 & 181.8 & 81.1 & ... + fe i & 5332.90 & 1.56 & @xmath1462.78 & ... & 34.2 & ... & ... & 40.3 & 34.8 & 80.2 & ... & ... + fe i & 5339.93 & 3.27 & @xmath1460.65 & 29.7 & 35.8 & ... & 49.3 & ... & 33.6 & ... & ... & ... + fe i & 5341.02 & 1.61 & @xmath1461.95 & 55.6 & 76.3 & ... & 71.2 & ... & 73.9 & 133.5 & ... & ... + fe i & 5367.47 & 4.41 & 0.44 & 33.4 & ... & ... & 44.9 & 37.6 & 24.9 & ... & ... & ... + fe i & 5369.96 & 4.37 & 0.54 & 37.5 & 26.7 & ... & 43.5 & 40.5 & 27.5 & 100.7 & ... & ... + fe i & 5371.49 & 0.96 & @xmath1461.64 & 111.5 & 125.9 & 115.4 & 132.3 & 122.9 & 118.7 & ... & 73.8 & ... + fe i & 5383.37 & 4.31 & 0.64 & ... & 33.5 & ... & ... & 54.2 & 31.1 & ... & ... & ... + fe i & 5397.13 & 0.92 & @xmath1461.99 & ... & ... & 90.0 & ... & 100.6 & 114.7 & 144.7 & 68.4 & ... + fe i & 5404.15 & 4.43 & 0.52 & ... & 43.5 & ... & ... & 56.3 & 35.2 & ... & ... & ... + fe i & 5405.77 & 0.99 & @xmath1461.84 & 99.9 & 107.8 & 98.3 & 111.9 & 122.8 & 112.1 & 150.9 & 55.1 & ... + fe i & 5415.20 & 4.39 & 0.64 & 41.1 & 29.8 & ... & 50.1 & 40.7 & 30.1 & 95.1 & ... & ... + fe i & 5424.07 & 4.32 & 0.52 & ... & 41.6 & ... & ... & 55.2 & 39.7 & ... & ... & ... + fe i & 5429.70 & 0.96 & @xmath1461.88 & 88.9 & 123.0 & 88.3 & 104.1 & 109.5 & 113.0 & ... & 50.5 & ... + fe i & 5434.52 & 1.01 & @xmath1462.12 & 85.3 & 103.8 & 91.2 & 112.4 & ... & 100.9 & ... & ... & ... + fe i & 5569.62 & 3.42 & @xmath1460.54 & ... & 28.8 & ... & ... & 46.8 & 31.1 & ... & ... & ... + fe i & 5572.84 & 3.40 & @xmath1460.28 & ... & 41.8 & ... & ... & 64.2 & 43.6 & ... & ... & ... + fe i & 5576.09 & 3.43 & @xmath1461.00 & ... & ... & ... & ... & 35.0 & 25.6 & ... & ... & ... + fe i & 5586.75 & 3.37 & @xmath1460.10 & 48.2 & 49.7 & ... & ... & 74.3 & 51.7 & 109.4 & ... & ... + fe i & 5615.64 & 3.33 & 0.05 & ... & 59.1 & ... & ... & 76.9 & 64.5 & 149.5 & 29.6 & ... + fe i & 5624.54 & 3.42 & @xmath1460.75 & ... & ... & ... & ... & 37.7 & 16.6 & ... & ... & ... + fe i & 6065.48 & 2.61 & @xmath1461.53 & ... & ... & ... & 51.5 & 61.1 & 38.7 & 102.9 & ... & ... + fe i & 6136.61 & 2.45 & @xmath1461.40 & 47.2 & 52.2 & ... & ... & ... & ... & ... & ... & ... + fe i & 6191.56 & 2.43 & @xmath1461.42 & 38.4 & 49.1 & ... & ... & 54.2 & 53.1 & ... & ... & ... + fe i & 6213.43 & 2.22 & @xmath1462.48 & ... & ... & ... & ... & 32.3 & 15.6 & 75.8 & ... & ... + fe i & 6219.28 & 2.20 & @xmath1462.43 & 18.6 & ... & ... & ... & 46.7 & 23.2 & 78.7 & ... & ... + fe i & 6230.72 & 2.56 & @xmath1461.28 & 44.9 & 51.8 & 34.8 & 48.5 & 60.8 & 51.1 & 108.4 & ... & ... + fe i & 6393.60 & 2.43 & @xmath1461.43 & 46.0 & 50.5 & ... & 74.1 & 74.6 & 44.2 & 123.6 & 19.5 & ... + ni i & 4714.42 & 3.38 & 0.23 & ... & 25.5 & ... & ... & 40.0 & 31.1 & ... & ... & ... + ni i & 4980.16 & 3.61 & @xmath1460.11 & ... & ... & ... & ... & ... & 14.8 & 67.3 & ... & ... + zn i & 4722.15 & 4.03 & @xmath1460.39 & ... & 17.2 & ... & ... & ... & 16.8 & ... & ... & ... + zn i & 4810.53 & 4.08 & @xmath1460.17 & 27.3 & 15.1 & ... & ... & 16.1 & 20.1 & 62.6 & ... & ... + sc ii & 4415.54 & 0.60 & @xmath1460.67 & 100.4 & 64.1 & ... & ... & 66.1 & 93.2 & 106.7 & ... & ... + sc ii & 5031.01 & 1.36 & @xmath1460.40 & ... & 31.6 & ... & ... & 37.2 & 31.4 & ... & ... & ... + sc ii & 5526.79 & 1.77 & 0.02 & ... & 29.3 & ... & ... & 61.2 & 35.0 & ... & ... & ... + ti ii & 4417.71 & 1.16 & @xmath1461.43 & ... & 96.2 & ... & ... & 94.6 & ... & ... & ... & ... + ti ii & 4418.31 & 1.24 & @xmath1461.99 & ... & 44.3 & ... & ... & 55.0 & 61.1 & ... & ... & ... + ti ii & 4441.73 & 1.18 & @xmath1462.41 & ... & 36.3 & ... & ... & ... & 43.5 & ... & ... & ... + ti ii & 4443.77 & 1.08 & @xmath1460.70 & 106.7 & 110.4 & ... & ... & 114.5 & 118.5 & 151.5 & 65.0 & ... + ti ii & 4444.54 & 1.12 & @xmath1462.21 & ... & 33.9 & ... & ... & ... & ... & ... & ... & ... + ti ii & 4450.50 & 1.08 & @xmath1461.51 & ... & 83.1 & ... & ... & 73.0 & 83.0 & ... & ... & ... + ti ii & 4464.46 & 1.16 & @xmath1462.08 & 71.2 & 57.4 & ... & ... & 61.1 & 64.4 & ... & ... & ... + ti ii & 4468.52 & 1.13 & @xmath1460.60 & ... & 100.3 & ... & ... & 116.4 & 114.4 & ... & ... & ... + ti ii & 4501.27 & 1.12 & @xmath1460.76 & 99.0 & 109.2 & ... & ... & 103.8 & 104.7 & ... & 65.2 & ... + ti ii & 4529.48 & 1.57 & @xmath1462.03 & ... & 37.1 & ... & ... & 53.9 & ... & ... & ... & ... + ti ii & 4533.97 & 1.24 & @xmath1460.77 & ... & 104.4 & ... & ... & 100.7 & 128.9 & ... & ... & ... + ti ii & 4563.77 & 1.22 & @xmath1460.96 & 109.0 & 98.4 & ... & ... & 99.9 & 109.2 & 142.7 & 62.3 & ... + ti ii & 4571.96 & 1.57 & @xmath1460.53 & ... & 98.8 & ... & ... & 109.5 & 111.1 & ... & ... & ... + ti ii & 4589.92 & 1.24 & @xmath1461.79 & 43.9 & 61.1 & ... & ... & 67.5 & 58.0 & 78.5 & ... & ... + ti ii & 4779.98 & 2.05 & @xmath1461.37 & 44.7 & 31.7 & ... & ... & 36.4 & ... & 74.3 & ... & ... + ti ii & 4798.51 & 1.08 & @xmath1462.67 & ... & 23.6 & ... & ... & 25.1 & ... & ... & ... & ... + ti ii & 4805.09 & 2.06 & @xmath1461.10 & 68.1 & 38.0 & ... & ... & 50.8 & 48.2 & 100.0 & ... & ... + ti ii & 5129.15 & 1.89 & @xmath1461.39 & ... & ... & ... & ... & ... & 43.2 & ... & ... & ... + ti ii & 5185.90 & 1.89 & @xmath1461.35 & ... & 33.7 & 26.4 & 41.6 & 35.7 & 33.1 & 68.9 & ... & 29.2 + ti ii & 5336.78 & 1.58 & @xmath1461.63 & ... & 38.8 & ... & ... & 47.2 & 46.3 & ... & ... & ... + ti ii & 5381.01 & 1.57 & @xmath1461.97 & ... & ... & ... & 47.5 & 40.4 & 32.1 & ... & ... & ... + ti ii & 5418.75 & 1.58 & @xmath1462.11 & ... & ... & ... & ... & ... & 27.0 & 82.0 & ... & ... + fe ii & 4416.82 & 2.78 & @xmath1462.60 & ... & 37.7 & ... & ... & 66.1 & 56.5 & ... & ... & ... + fe ii & 4508.28 & 2.86 & @xmath1462.58 & ... & 47.0 & ... & ... & 43.9 & 48.3 & 78.1 & ... & ... + fe ii & 4515.34 & 2.84 & @xmath1462.48 & ... & 26.9 & ... & ... & ... & ... & ... & ... & ... + fe ii & 4520.23 & 2.81 & @xmath1462.60 & ... & ... & ... & ... & 37.0 & ... & ... & ... & ... + fe ii & 4582.83 & 2.84 & @xmath1463.10 & ... & ... & ... & ... & 23.4 & ... & ... & ... & ... + fe ii & 4583.83 & 2.81 & @xmath1462.02 & 53.9 & 79.7 & ... & ... & 76.2 & 74.6 & ... & 51.1 & ... + fe ii & 4731.44 & 2.89 & @xmath1463.36 & ... & ... & ... & ... & 17.5 & 24.9 & 44.9 & ... & ... + fe ii & 4923.93 & 2.89 & @xmath1461.32 & 89.0 & 93.8 & ... & ... & 91.6 & 92.8 & 132.8 & 53.0 & ... + fe ii & 5018.45 & 2.89 & @xmath1461.22 & ... & 102.0 & ... & ... & 104.8 & 101.0 & ... & 57.5 & ... + fe ii & 5197.56 & 3.23 & @xmath1462.10 & ... & 29.7 & ... & 37.8 & ... & ... & 68.7 & ... & 25.6 + fe ii & 5234.62 & 3.22 & @xmath1462.27 & 29.4 & 39.2 & 17.6 & 38.2 & 48.5 & 38.6 & 73.9 & ... & 52.2 + fe ii & 5276.00 & 3.20 & @xmath1461.94 & 38.2 & 43.6 & 35.6 & ... & 49.7 & 52.9 & 98.9 & ... & 52.5 + fe ii & 5534.83 & 3.24 & @xmath1462.93 & ... & ... & ... & ... & 16.9 & 16.7 & ... & ... & ... + ba ii & 4554.03 & 0.00 & 0.16 & ... & 110.3 & ... & ... & ... & 204.3 & ... & 129.3 & ... + ba ii & 4934.09 & 0.00 & @xmath1460.16 & ... & 104.5 & 222.3 & 259.6 & ... & 222.3 & ... & 141.4 & 275.0 + ba ii & 5853.67 & 0.60 & @xmath1461.02 & 118.3 & 19.4 & 122.5 & 122.4 & 113.5 & 96.2 & 188.1 & 52.5 & 153.3 + ba ii & 6141.70 & 0.70 & @xmath1460.07 & ... & ... & ... & ... & 169.3 & ... & 327.0 & ... & 201.1 + ba ii & 6496.90 & 0.60 & @xmath1460.37 & 181.9 & 48.8 & 212.3 & 197.8 & 182.1 & 146.5 & 260.1 & 80.5 & 241.8 + lccccccccccc na
i & 5889.95 & 0.00 & 0.10 & ... & 228.5 & 171.7 & ... & 170.8 & ... & 292.6 & 166.6 + na i & 5895.92 & 0.00 & @xmath1460.20 & 246.6 & 211.5 & 149.7 & 213.9 & 162.2 & 246.5 & 275.9 & 142.8 + mg i & 4571.10 & 0.00 & @xmath1465.69 & ... & ... & ... & ... & ... & ... & ... & 33.0 + mg i & 4702.99 & 4.35 & @xmath1460.52 & ... & ... & ... & ... & ... & ... & ... & 67.0 + mg
i & 5172.69 & 2.71 & @xmath1460.38 & 297.7 & 187.9 & 192.1 & 216.7 & 170.4 & 272.2 & 183.4 & 163.8 + mg i & 5183.60 & 2.72 & @xmath1460.16 & ... & ... & ... & ... & ... & ... & ... & 182.5 + mg i & 5528.40 & 4.35 & @xmath1460.49 & ... & ... & ... & ... & ... & ... & 114.0 & 53.5 + ca i & 5588.76 & 2.53 & 0.36 & 119.2 & ... & ... & 80.1 & ... & ... & 52.3 & 31.8 + ca i & 5594.47 & 2.52 & 0.10 & 119.9 & 48.7 & ... & 72.5 & ... & 90.5 & 56.1 & 30.4 + ca i & 5601.29 & 2.53 & @xmath1460.52 & ... & ... & ... & ... & ... & ... & ... & 28.9 + ca i & 5857.45 & 2.93 & 0.24 & ... & ... & 37.5 & 63.0 & ... & 81.5 & 35.6 & 22.2 + ca i & 6102.72 & 1.88 & @xmath1460.77 & 98.1 & ... & ... & 54.2 & ... & ... & 30.4 & 15.0 + ca
i & 6122.22 & 1.89 & @xmath1460.32 & ... & ... & ... & ... & ... & ... & 66.1 & 45.3 + ca i & 6162.17 & 1.90 & @xmath1460.09 & ... & 123.1 & 95.7 & ... & 94.5 & 149.1 & 68.9 & 58.9 + ti
i & 4981.73 & 0.85 & 0.56 & ... & ... & ... & ... & ... & ... & ... & 44.0 + ti i & 5192.97 & 0.02 & @xmath1460.95 & ... & 43.7 & ... & 56.8 & ... & 105.5 & 44.0 & ... + ti i & 5206.08 & 2.49 & 1.07 & 126.1 & 87.0 & 70.2 & ... & 65.7 & ... & ... & ... + ti i & 5210.38 & 0.05 & @xmath1460.83 & 103.1 & 49.2 & 31.7 & 64.2 & ... & 122.6 & 40.1 & ... + cr i & 5208.44 & 0.94 & 0.16 & ... & ... & 102.8 & ... & 95.1 & ... & 86.4 & 75.8 + cr
i & 5409.79 & 1.03 & @xmath1460.72 & ... & ... & ... & ... & ... & ... & 51.1 & ... + fe i & 4461.65 & 0.09 & @xmath1463.21 & ... & ... & ... & ... & ... & ... & ... & 88.7 + fe i & 4489.74 & 0.12 & @xmath1463.97 & ... & ... & ... & ... & ... & ... & ... & 46.7 + fe i & 4528.61 & 2.18 & @xmath1460.82 & ... & ... & ... & ... & ... & ... & ... & 72.3 + fe i & 4871.32 & 2.87 & @xmath1460.36 & ... & ... & ... & ... & ... & ... & ... & 61.7 + fe i & 4890.75 & 2.88 & @xmath1460.39 & ... & ... & ... & ... & 77.3 & ... & ... & 58.4 + fe i & 4891.49 & 2.85 & @xmath1460.11 & ... & ... & ... & ... & ... & ... & ... & 74.3 + fe i & 4920.50 & 2.83 & 0.07 & ... & ... & ... & ... & ... & ... & ... & 73.9 + fe i & 4924.77 & 2.28 & @xmath1462.26 & ... & ... & ... & ... & ... & ... & ... & 20.8 + fe i & 4994.13 & 0.92 & @xmath1462.96 & ... & ... & ... & ... & ... & ... & ... & 37.5 + fe i & 5012.07 & 0.86 & @xmath1462.64 & ... & ... & ... & ... & ... & ... & ... & 60.8 + fe i & 5051.63 & 0.92 & @xmath1462.80 & ... & ... & ... & ... & ... & ... & ... & 45.0 + fe i & 5110.41 & 0.00 & @xmath1463.76 & ... & ... & ... & ... & ... & ... & ... & 60.2 + fe i & 5166.28 & 0.00 & @xmath1464.20 & 144.8 & 82.3 & ... & 85.3 & 44.5 & 146.1 & 37.0 & 24.1 + fe i & 5171.60 & 1.49 & @xmath1461.79 & 140.0 & 97.6 & 66.9 & 97.3 & 87.4 & 146.1 & 67.5 & 58.4 + fe i & 5191.46 & 3.04 & @xmath1460.55 & ... & 89.0 & ... & ... & ... & ... & ... & 36.1 + fe i & 5192.34 & 3.00 & @xmath1460.42 & ... & 76.3 & ... & 87.3 & 49.9 & 105.2 & 54.7 & 51.6 + fe i & 5194.94 & 1.56 & @xmath1462.09 & 150.5 & 77.1 & 59.9 & 80.6 & ... & 126.8 & 51.1 & 49.0 + fe i & 5216.27 & 1.61 & @xmath1462.15 & 126.5 & 71.0 & 37.5 & 68.3 & 44.4 & 106.7 & 44.1 & 43.7 + fe i & 5232.94 & 2.94 & @xmath1460.06 & 129.9 & 94.9 & 61.5 & 93.8 & 53.8 & 121.8 & 68.3 & 59.3 + fe i & 5247.05 & 0.09 & @xmath1464.95 & 62.3 & 23.1 & ... & 19.9 & ... & 61.5 & ... & ... + fe i & 5254.96 & 0.11 & @xmath1464.76 & 93.8 & ... & ... & ... & ... & 85.3 & 15.6 & ... + fe i & 5266.56 & 3.00 & @xmath1460.39 & 112.5 & 66.9 & 59.2 & 75.6 & 53.2 & 100.0 & 53.5 & 50.6 + fe i & 5269.54 & 0.86 & @xmath1461.32 & ... & 151.4 & ... & 154.8 & 113.1 & 189.0 & 116.2 & 108.8 + fe i & 5281.79 & 3.04 & @xmath1460.83 & 87.3 & 56.2 & 31.7 & 60.4 & ... & 81.4 & 29.7 & 23.1 + fe i & 5302.30 & 3.28 & @xmath1460.72 & ... & ... & ... & ... & ... & ... & 41.4 & ... + fe i & 5324.18 & 3.21 & @xmath1460.10 & 122.9 & 71.5 & 53.5 & 87.5 & 44.3 & 111.3 & 51.3 & 45.9 + fe i & 5328.04 & 0.92 & @xmath1461.47 & 207.3 & 149.3 & 107.8 & 145.9 & 102.4 & ... & 106.4 & 108.5 + fe i & 5339.93 & 3.27 & @xmath1460.65 & ... & ... & ... & ... & 29.6 & ... & 42.0 & 26.3 + fe i & 5341.02 & 1.61 & @xmath1461.95 & ... & 79.0 & ... & ... & ... & 116.3 & 56.1 & 51.8 + fe i & 5367.47 & 4.41 & 0.44 & ... & ... & ... & ... & ... & ... & 42.1 & ... + fe i & 5369.96 & 4.37 & 0.54 & ... & ... & ... & ... & ... & ... & 31.5 & 25.5 + fe i & 5371.49 & 0.96 & @xmath1461.64 & ... & ... & ... & ... & ... & ... & 103.2 & 102.5 + fe i & 5383.37 & 4.31 & 0.64 & ... & ... & ... & ... & ... & ... & 32.4 & 26.2 + fe i & 5397.13 & 0.92 & @xmath1461.99 & ... & ... & ... & ... & ... & ... & 91.3 & ... + fe i & 5405.77 & 0.99 & @xmath1461.84 & ... & ... & ... & ... & ... & 170.3 & 100.5 & 93.3 + fe i & 5415.20 & 4.39 & 0.64 & ... & ... & ... & ... & ... & ... & 41.2 & ... + fe i & 5424.07 & 4.32 & 0.52 & ... & ... & ... & ... & ... & ... & ... & 31.8 + fe i & 5429.70 & 0.96 & @xmath1461.88 & ... & ... & ... & ... & 83.0 & ... & 77.5 & 84.2 + fe i & 5434.52 & 1.01 & @xmath1462.12 & 202.5 & ... & ... & ... & ... & ... & 76.6 & 76.4 + fe i & 5572.84 & 3.40 & @xmath1460.28 & ... & ... & ... & ... & ... & ... & ... & 40.7 + fe i & 5586.75 & 3.37 & @xmath1460.10 & 118.7 & 62.1 & ... & ... & ... & 92.1 & 46.0 & 42.1 + fe i & 5615.64 & 3.33 & 0.05 & ... & ... & ... & ... & ... & ... & ... & 58.1 + fe i & 6065.48 & 2.61 & @xmath1461.53 & 129.3 & ... & ... & ... & ... & ... & 27.6 & 19.8 + fe i & 6136.61 & 2.45 & @xmath1461.40 & 132.5 & 69.6 & 39.6 & ... & ... & 117.6 & ... & 38.6 + fe i & 6137.69 & 2.59 & @xmath1461.40 & 118.2 & 62.5 & 32.8 & 71.1 & 35.7 & 113.6 & ... & 35.0 + fe i & 6191.56 & 2.43 & @xmath1461.42 & ... & ... & ... & ... & ... & 111.4 & 29.6 & 27.8 + fe i & 6213.43 & 2.22 & @xmath1462.48 & ... & ... & ... & 47.2 & ... & ... & 15.7 & ... + fe i & 6219.28 & 2.20 & @xmath1462.43 & ... & ... & ... & 58.9 & ... & ... & ... & ... + fe i & 6230.72 & 2.56 & @xmath1461.28 & 127.1 & ... & 42.5 & 65.6 & ... & 98.7 & 38.5 & 37.1 + fe i & 6393.60 & 2.43 & @xmath1461.43 & ... & ... & 46.4 & ... & 45.1 & ... & 42.6 & 37.4 + zn i & 4722.15 & 4.03 & @xmath1460.39 & ... & ... & ... & ... & ... & ... & ... & ... + zn i & 4810.53 & 4.08 & @xmath1460.17 & ... & ... & ... & ... & ... & 59.5 & 22.7 & ... + sc ii & 4415.54 & 0.60 & @xmath1460.67 & ... & ... & ... & ... & ... & ... & ... & 59.3 + sc ii & 5526.79 & 1.77 & 0.02 & ... & ... & ... & ... & ... & ... & ... & 20.9 + ti ii & 4443.77 & 1.08 & @xmath1460.70 & ... & ... & ... & ... & ... & ... & ... & 74.4 + ti ii & 4468.52 & 1.13 & @xmath1460.60 & ... & ... & ... & ... & ... & ... & ... & 99.0 + ti ii & 4533.97 & 1.24 & @xmath1460.77 & ... & ... & ... & ... & ... & ... & ... & 94.3 + ti ii & 4563.77 & 1.22 & @xmath1460.96 & ... & ... & ... & ... & ... & ... & ... & 81.8 + ti ii & 4571.96 & 1.57 & @xmath1460.53 & ... & ... & ... & ... & ... & ... & ... & 93.1 + ti ii & 4589.92 & 1.24 & @xmath1461.79 & ... & ... & ... & ... & ... & ... & ... & 38.0 + ti ii & 4779.98 & 2.05 & @xmath1461.37 & ... & ... & ... & ... & ... & ... & ... & 21.0 + ti ii & 5185.90 & 1.89 & @xmath1461.35 & 85.2 & 50.7 & 31.7 & 55.7 & ... & 76.7 & 21.1 & ... + ti ii & 5336.78 & 1.58 & @xmath1461.63 & ... & ... & ... & ... & ... & ... & 56.5 & ... + ti ii & 5381.01 & 1.57 & @xmath1461.97 & ... & ... & ... & ... & ... & ... & 35.3 & ... + fe ii & 4508.28 & 2.86 & @xmath1462.58 & ... & ... & ... & ... & ... & ... & ... & 27.4 + fe ii & 4583.83 & 2.81 & @xmath1462.02 & ... & ... & ... & ... & ... & ... & ... & 51.2 + fe ii & 4923.93 & 2.89 & @xmath1461.32 & ... & ... & ... & ... & ... & ... & ... & 84.7 + fe ii & 5018.45 & 2.89 & @xmath1461.22 & ... & ... & ... & ... & ... & ... & ... & 84.6 + fe ii & 5197.56 & 3.23 & @xmath1462.10 & 96.2 & 39.5 & 37.4 & 49.9 & ... & 66.5 & ... & 21.1 + fe ii & 5234.62 & 3.22 & @xmath1462.27 & 69.5 & 48.9 & 28.9 & 57.9 & ... & 63.9 & 25.5 & ... + fe ii & 5276.00 & 3.20 & @xmath1461.94 & 92.9 & 47.7 & 51.3 & 62.0 & 41.2 & ... & 28.7 & 29.1 + ba ii & 4554.03 & 0.00 & 0.16 & ... & 252.5 & ... & ... & 267.1 & ... & ... & 169.5 + ba ii & 4934.09 & 0.00 & @xmath1460.16 & ... & 261.0 & ... & 267.3 & ... & ... & ... & 168.8 + ba ii & 5853.67 & 0.60 & @xmath1461.02 & 219.4 & 141.3 & 136.8 & 143.0 & ... & ... & 128.8 & 64.2 + ba ii & 6141.70 & 0.70 & @xmath1460.07 & 285.6 & 164.4 & 233.5 & 189.8 & 182.2 & ... & ... & ... + ba ii & 6496.90 & 0.60 & @xmath1460.37 & ... & ... & 219.9 & ... & 188.6 & 256.6 & 182.6 & 111.7 + lcccc fe i & 3618.77 & 0.99 & 0.00 & 68.4 + fe i & 3631.46 & 0.96 & @xmath1460.04 & 47.5 + fe i & 3745.56 & 0.09 & @xmath1460.77 & 65.0 + fe i & 3758.23 & 0.96 & @xmath1460.03 & 52.2 + fe i & 3815.84 & 1.49 & 0.23 & 65.1 + fe i & 3820.43 & 0.86 & 0.12 & 92.4 + fe i & 3824.44 & 0.00 & @xmath1461.36 & 44.0 + fe i & 3825.88 & 0.92 & @xmath1460.04 & 76.5 + fe i & 3827.82 & 1.56 & 0.06 & 48.2 + fe i & 3849.97 & 1.01 & @xmath1460.87 & 39.5 + fe i & 3856.37 & 0.05 & @xmath1461.29 & 40.1 + fe i & 3859.91 & 0.00 & @xmath1460.71 & 80.3 + fe i & 3865.52 & 1.01 & @xmath1460.98 & 33.3 + fe i & 3899.71 & 0.09 & @xmath1461.53 & 45.3 + fe i & 3920.26 & 0.12 & @xmath1461.75 & 36.4 + fe i & 3922.91 & 0.05 & @xmath1461.65 & 36.5 + fe i & 3930.30 & 0.09 & @xmath1461.49 & 37.7 + fe i & 4005.24 & 1.56 & @xmath1460.61 & 30.3 + fe i & 4045.81 & 1.49 & 0.28 & 69.3 + fe i & 4071.74 & 1.61 & @xmath1460.02 & 53.5 + fe i & 4132.06 & 1.61 & @xmath1460.67 & 20.9 + fe i & 4143.87 & 1.56 & @xmath1460.51 & 31.4 + fe i & 4202.03 & 1.49 & @xmath1460.71 & 22.3 + fe i & 4260.47 & 2.40 & 0.08 & 30.6 + fe i & 4325.76 & 1.61 & 0.01 & 56.7 + la ii & 3849.02 & 0.00 & @xmath1460.45 & 33.4 + la ii & 3949.10 & 0.40 & 0.49 & 82.5 + la ii & 3988.52 & 0.40 & 0.21 & 53.4 + la ii & 3995.75 & 0.17 & @xmath1460.06 & 36.7 + la ii & 4123.23 & 0.32 & 0.13 & 32.0 + la ii & 4286.97 & 1.95 & 0.21 & 36.0 + la ii & 4333.74 & 0.17 & @xmath1460.06 & 43.3 + ce ii & 4186.60 & 0.86 & 0.90 & 35.3 + nd ii & 4156.08 & 0.18 & 0.10 & 29.7 + lcccccccccccc na i & 5682.63 & 2.10 & @xmath1460.70 & ... & ... & ... & ... & ... & 3.4 & ... & ... & ... + na i & 5688.20 & 2.10 & @xmath1460.46 & ... & ... & ... & ... & ... & 8.6 & ... & ... & 6.1 +
na i & 5889.95 & 0.00 & 0.10 & 186.7 & 146.3 & 116.6 & 151.4 & 128.9 & 171.3 & 84.1 & 220.3 & ...
+ na i & 5895.92 & 0.00 & @xmath1460.20 & 165.2 & 122.3 & 100.8 & 133.8 & 99.9 & 155.0 & ... & 197.8 & ... + llcccc 1 & bs 16929005 & 5250 & 2.8 & @xmath1463.2 & 2.3 + 2 & cs 22174007 & 5300 & 2.7 & @xmath1462.3 & 2.1 + 3 & cs 22886042 & 4800 & 1.6 & @xmath1462.5 & 2.1 + 4 & cs 22948027 & 5000 & 1.9 & @xmath1462.3 & 2.0 + 5 & cs 22960053 & 5200 & 2.1 & @xmath1463.2 & 2.1 + 6 & cs 29503010 & 6500 & 4.5 & @xmath1461.2 & 1.9 + 7 & cs 29528028 & 6800 & 4.0 & @xmath1462.9 & 1.9 + 8 & cs 30312100 & 5300 & 2.8 & @xmath1462.4 & 2.2 + 9 & cs 30322023 & 4300 & 1.0 & @xmath1463.3 & 3.3 + 10 & cs 30338089 & 5000 & 2.1 & @xmath1462.5 & 1.7 + 11 & cs 31062041 & 5100 & 2.2 & @xmath1462.5 & 2.3 + 12 & cs 29497034 & 4900 & 1.5 & @xmath1463.0 & 2.2 + 13 & he 02061916 & 5200 & 2.7 & @xmath1462.2 & 2.0 + 14 & he 04002030 & 5600 & 3.5 & @xmath1461.8 & 2.2 + 15 & he 04410652 & 4900 & 1.4 & @xmath1462.5 & 2.3 + 16 & he 05071653 & 5000 & 2.4 & @xmath1461.5 & 2.0@xmath193 + 17 & he 10051439 & 5000 & 1.9 & @xmath1463.2 & 2.0 + 18 & he 11570518 & 4900 & 2.0 & @xmath1462.4 & 2.0@xmath193 + 19 & he 13191935 & 4600 & 1.1 & @xmath1461.8 & 2.0@xmath193 + 20 & he 14290551 & 4700 & 1.5 &
@xmath1462.5 & 2.0 + 21 & he 1447 + 0102 & 5100 & 1.7 & @xmath1462.5 & 1.8 + 22 & he 15231155 & 4800 & 1.6 & @xmath1462.2 & 1.8 + 23 & he 15280409 & 5000 & 1.8 & @xmath1462.6 & 1.8 + 24 & he 22210453 & 4400 & 0.4 & @xmath1462.2 & 2.0@xmath193 + 25 & he 22280706 & 5100 & 2.6 & @xmath1462.5 & 1.6 + 26 & he 23300555 & 4900 & 1.7 & @xmath1462.8 & 1.8 + lcccccccccccccc @xmath194 & 7.45 & 7.45 & 8.39 & 7.78 & 6.17 & 7.53 & 6.31 & 3.05 & 4.90 & 4.90 & 5.64 & 6.23 & 4.60 & 2.17 + bs 16920005 & & & & & & & & & & & & & & + @xmath195 & 4.28 & 4.27 & 6.3 & @xmath196 & 2.94 & 4.85 & 3.27 & ... & ... & 2.17 & 2.14 & ... & ... & @xmath1461.48 + [ x / fe ] & @xmath1463.17 & @xmath1463.18 & 1.08 & @xmath1972.39 & @xmath1460.06 & 0.49 & 0.13 & ... & ... & 0.44 & @xmath1460.34 & ... & ... & @xmath1460.48 + n & 24 & 3 & ch4322 & cn4215 & 2 & 4 & 2 & ... & ... & 8 & 1 & ... & ... & 2 + s.e.([x / fe ] ) & 0.14 & 0.13 & 0.16 & 0.07 & 0.17 & 0.10 & 0.09 & ... & ... & 0.14 & 0.09 & ... & ... & 0.12 + cs 22174007 & & & & & & & & & & & & & & + @xmath195 & 5.26 & 5.26 & 6.6 & @xmath196 & 4.36 & 5.82 & 4.35 & 0.96 & 3.09 & 3.17 & 3.26 & 4.07 & 2.46 & @xmath1460.59 + [ x / fe ] & @xmath1462.20 & @xmath1462.20 & 0.41 & @xmath1971.42 & 0.39 & 0.49 & 0.24 & 0.11 & 0.39 & 0.46 & @xmath1460.19 & 0.03 & 0.06 & @xmath1460.57 + n & 85 & 12 & ch4322 & cn4215 & 2 & 5 & 10 & 3 & 8 & 22 & 7 & 2 & 2 & 4 + s.e.([x / fe ] ) & 0.14 & 0.13 & 0.16 & 0.07 & 0.17 & 0.10 & 0.10 & 0.14 & 0.09 & 0.14 & 0.09 & 0.11 & 0.13 & 0.13 + cs 22886042 & & & & & & & & & & & & & & + @xmath195 & 4.84 & 4.84 & 5.9 & @xmath198 & 3.91 & 5.43 & 3.96 & 0.48 & 2.45 & 2.63 & 2.74 & 3.65 & 2.28 & @xmath1460.94 + [ x / fe ] & @xmath1462.61 & @xmath1462.61 & 0.12 & @xmath1972.03 & 0.35 & 0.51 & 0.26 & 0.04 & 0.16 & 0.35 & @xmath1460.29 & 0.03 & 0.29 & @xmath1460.50 + @xmath199 & 84 & 10 & ch4322 & cn4215 & 2 & 5 & 10 & 3 & 8 & 21 & 3 & 1 & 1 & 4 + s.e.([x / fe ] ) & 0.14 & 0.13 & 0.16 & 0.07 & 0.17 & 0.10 & 0.10 & 0.14 & 0.09 & 0.13 & 0.09 & 0.11 & 0.13 & 0.13 + cs 22948027 & & & & & & & & & & & & & & + @xmath195 & 5.24 & 5.22 & 8.3 & 8.0 & 4.88 & 5.87 & 4.67 & ... & 3.28 & 3.37 & 3.31 & ... & 2.69 & 2.27 + [ x / fe ] & @xmath1462.21 & @xmath1462.23 & 2.12 & 2.43 & 0.92 & 0.55 & 0.57 & ... & 0.60 & 0.68 & @xmath1460.12 & ... & 0.31 & 2.31 + @xmath199 & 25 & 2 & cc5635 & cn4215 & 2 & 1 & 5 & ... & 3 & 4 & 2 & ... & 1 & 2 + s.e.([x / fe ] ) & 0.28 & 0.16 & 0.20 & 0.35 & 0.13 & 0.14 & 0.19 & ... & 0.13 & 0.26 & 0.12 & ... & 0.21 & 0.15 + cs 22960053 & & & & & & & & & & & & & & + @xmath195 & 4.31 & 4.30 & 7.3 & 7.7 & ... & 5.04 & 3.61 & 0.40 & ... & 2.30 & 2.27 & ... & ... & @xmath1460.11 + [ x / fe ] & @xmath1463.14 & @xmath1463.15 & 2.05 & 3.06 & ... & 0.65 & 0.43 & 0.49 & ... & 0.54 & @xmath1460.23 & ... & ... & 0.86 + @xmath199 & 16 & 3 & ch4322 & cn4215 & ... & 3 & 2 & 1 & ... & 8 & 1 & ... & ... & 4 + s.e.([x / fe ] ) & 0.15 & 0.15 & 0.16 & 0.21 & ... & 0.12 & 0.12 & 0.15 & ... & 0.15 & 0.11 & ... & ... & 0.14 + cs 29503010 & & & & & & & & & & & & & & + @xmath195 & 6.39 & 6.34 & 8.4 & @xmath200 & 5.23 & 6.83 & 5.41 & 2.30 & 4.23 & 4.23 & 4.46 & 5.18 & 3.85 & 2.61 + [ x / fe ] & @xmath1461.06 & @xmath1461.11 & 1.07 & @xmath1971.28 & 0.12 & 0.36 & 0.16 & 0.31 & 0.39 & 0.39 & @xmath1460.12 & 0.01 & 0.32 & 1.50 + @xmath199 & 74 & 11 & ch4322 & cn4215 & 2 & 2 & 12 & 3 & 5 & 20 & 2 & 2 & 1 & 4 + s.e.([x / fe ] ) & 0.19 & 0.17 & 0.19 & 0.09 & 0.15 & 0.18 & 0.15 & 0.14 & 0.22 & 0.17 & 0.10 & 0.12 & 0.11 & 0.16 + cs 29528028 & & & & & & & & & & & & & & + @xmath195 & 4.59 & 4.57 & 8.3 & @xmath1978.5 & 6.05 & 6.36 & 3.91 & 0.78 & ... & 2.91 & ... & 3.63 & ... & 2.58 + [ x / fe ] & @xmath1462.86 & @xmath1462.88 & 2.77 & @xmath1973.58 & 2.74 & 1.69 & 0.46 & 0.59 & ... & 0.87 & ... & 0.26 & ... & 3.27 + @xmath199 & 28 & 3 & ch4322 & cn4215 & 3 & 3 & 1 & 1 & ... & 5 & ... & 1 & ... & 5 + s.e.([x / fe ] ) & 0.22 & 0.18 & 0.21 & 0.16 & 0.19 & 0.21 & 0.19 & 0.18 & ... & 0.20 & ... & 0.16 & ... & 0.20 + cs 30312100 & & & & & & & & & & & & & & + @xmath195 & 5.12 & 5.12 & 6.6 & @xmath201 & 4.02 & 5.66 & 4.16 & 0.69 & 2.96 & 3.03 & 3.43 & 4.11 & 2.82 & @xmath1460.94 + [ x / fe ] & @xmath1462.33 & @xmath1462.33 & 0.54 & @xmath1971.85 & 0.18 & 0.47 & 0.18 & @xmath1460.03 & 0.40 & 0.46 & 0.12 & 0.21 & 0.55 & @xmath1460.78 + @xmath199 & 59 & 10 & ch4322 & cn4215 & 2 & 5 & 10 & 3 & 8 & 16 & 3 & 2 & 2 & 3 + s.e.([x / fe ] ) & 0.14 & 0.13 & 0.16 & 0.07 & 0.17 & 0.10 & 0.10 & 0.14 & 0.09 & 0.16 & 0.09 & 0.11 & 0.13 & 0.13 + cs 30322023 & & & & & & & & & & & & & & + @xmath195 & 4.20 & 4.18 & 5.7 & 7.0 & 3.96 & 4.82 & 3.36 & 0.08 & 1.72 & 2.00 & 2.04 & ... & ... & @xmath1460.49 + [ x / fe ] & @xmath1463.25 & @xmath1463.28 & 0.56 & 2.47 & 1.04 & 0.54 & 0.30 & 0.28 & 0.07 & 0.35 & @xmath1460.36 & ... & ... & 0.59 + @xmath199 & 53 & 4 & ch4322 & cn4215 & 1 & 3 & 5 & 3 & 7 & 14 & 2 & ... & ... & 3 + s.e.([x / fe ] ) & 0.23 & 0.18 & 0.12 & 0.33 & 0.18 & 0.26 & 0.16 & 0.18 & 0.18 & 0.23 & 0.15 & ... & ... & 0.17 + cs 30338089 & & & & & & & & & & & & & & + @xmath195 & 5.01 & 5.00 & 8.0 & 6.6 & 4.45 & 5.56 & 4.38 & 1.86 & 2.79 & 3.22 & 3.19 & ... & 2.62 & 1.94 + [ x / fe ] & @xmath1462.45 & @xmath1462.45 & 2.06 & 1.27 & 0.73 & 0.48 & 0.51 & 1.26 & 0.33 & 0.77 & @xmath1460.01 & ... & 0.46 & 2.22 + @xmath199 & 34 & 4 & cc5635 & cn4215 & 3 & 2 & 4 & 1 & 2 & 7 & 2 & ... & 2 & 2 + s.e.([x / fe ] ) & 0.20 & 0.15 & 0.19 & 0.33 & 0.13 & 0.14 & 0.11 & 0.10 & 0.13 & 0.25 & 0.12 & ... & 0.17 & 0.15 + cs 31062041 & & & & & & & & & & & & & & + @xmath195 & 5.08 & 5.07 & 6.6 & @xmath196 & 4.23 & 5.70 & 4.15 & 0.87 & 2.78 & 2.95 & 3.07 & 3.82 & 2.42 & @xmath1460.43 + [ x / fe ] & @xmath1462.37 & @xmath1462.38 & 0.58 & @xmath1971.59 & 0.43 & 0.54 & 0.21 & 0.19 & 0.25 & 0.43 & @xmath1460.20 & @xmath1460.04 & 0.20 & @xmath1460.23 + @xmath199 & 75 & 9 & ch4322 & cn4215 & 2 & 4 & 8 & 3 & 8 & 19 & 1 & 1 & 2 & 4 + s.e.([x / fe ] ) & 0.15 & 0.15 & 0.16 & 0.07 & 0.18 & 0.12 & 0.11 & 0.15 & 0.11 & 0.17 & 0.11 & 0.12 & 0.14 & 0.14 + cs 29497034 & & & & & & & & & & & & & & + @xmath195 & 4.54 & 4.56 & 8.2 & 7.5 & 5.03 & 5.93 & 3.91 & ... & 2.51 & 2.40 & 2.57 & ... & ... & 1.49 + [ x / fe ] & @xmath1462.91 & @xmath1462.89 & 2.72 & 2.63 & 1.78 & 1.31 & 0.52 & ... & 0.52 & 0.41 & @xmath1460.15 & ... & ... & 2.23 + @xmath199 & 15 & 2 & cc5635 & cn4215 & 2 & 3 & 1 & ... & 2 & 1 & 1 & ... & ... & 3 + s.e.([x / fe ] ) & 0.27 & 0.14 & 0.20 & 0.53 & 0.12 & 0.12 & 0.12 & ... & 0.11 & 0.21 & 0.10 & ... & ... & 0.13 + he 02061916 & & & & & & & & & & & & & & + @xmath195 & 5.36 & 5.38 & 8.4 & 7.3 & 4.82 & 5.96 & 4.33 & ... & 3.22 & 3.40 & 3.55 & ... & ... & 2.05 + [ x / fe ] & @xmath1462.09 & @xmath1462.08 & 2.10 & 1.61 & 0.74 & 0.52 & 0.11 & ... & 0.41 & 0.58 & 0.00 & ... & ... & 1.97 + @xmath199 & 22 & 2 & cc5635 & cn4215 & 2 & 3 & 4 & ... & 2 & 2 & 2 & ... & ... & 3 + s.e.([x / fe ] ) & 0.20 & 0.16 & 0.19 & 0.33 & 0.14 & 0.15 & 0.13 & ... & 0.14 & 0.19 & 0.13 & ... & ... & 0.16 + he 04002030 & & & & & & & & & & & & & & + @xmath195 & 5.72 & 5.70 & 7.8 & 8.8 & 5.20 & 6.42 & 4.76 & 1.63 & 3.44 & 3.58 & 3.82 & 4.40 & 2.67 & 2.07 + [ x / fe ] & @xmath1461.73 & @xmath1461.76 & 1.14 & 2.75 & 0.76 & 0.62 & 0.18 & 0.32 & 0.27 & 0.41 & @xmath1460.08 & @xmath1460.10 & @xmath1460.20 & 1.64 + @xmath199 & 66 & 11 & ch4322 & cn4215 & 3 & 2 & 9 & 3 & 7 & 18 & 8 & 1 & 1 & 3 + s.e.([x / fe ] ) & 0.17 & 0.18 & 0.16 & 0.31 & 0.20 & 0.14 & 0.14 & 0.17 & 0.14 & 0.17 & 0.14 & 0.15 & 0.16 & 0.16 + he 04410652 & & & & & & & & & & & & & & + @xmath195 & 4.98 & 4.97 & 7.3 & 6.2 & 4.42 & 5.40 & 4.15 & 0.72 & 2.77 & 2.71 & 2.90 & 3.88 & 2.29 & 0.81 + [ x / fe ] & @xmath1462.47 & @xmath1462.48 & 1.38 & 0.89 & 0.72 & 0.35 & 0.31 & 0.14 & 0.34 & 0.29 & @xmath1460.26 & 0.12 & 0.17 & 1.11 + @xmath199 & 74 & 9 & cc5165 & cn4215 & 2 & 4 & 9 & 3 & 0 & 17 & 3 & 2 & 2 & 4 + s.e.([x / fe ] ) & 0.20 & 0.21 & 0.19 & 0.24 & 0.14 & 0.15 & 0.13 & 0.11 & 0.14 & 0.20 & 0.13 & 0.11 & 0.18 & 0.24 + he 05071653 & & & & & & & & & & & & & & + @xmath195 & 6.07 & 6.06 & 8.3 & 7.2 & 5.22 & 6.34 & 5.51 & 1.87 & 3.99 & 3.90 & 4.33 & 5.09 & 3.34 & 2.68 + [ x / fe ] & @xmath1461.38 & @xmath1461.39 & 1.29 & 0.80 & 0.43 & 0.19 & 0.58 & 0.20 & 0.47 & 0.38 & 0.07 & 0.25 & 0.12 & 1.89 + @xmath199 & 28 & 5 & cc5635 & cn4215 & 4 & 1 & 5 & 1 & 5 & 7 & 1 & 1 & 1 & 3 + s.e.([x / fe ] ) & 0.19 & 0.17 & 0.19 & 0.32 & 0.29 & 0.12 & 0.12 & 0.17 & 0.12 & 0.15 & 0.12 & 0.15 & 0.16 & 0.14 + he 10051439 & & & & & & & & & & & & & & + @xmath195 & 4.28 & 4.30 & 7.7 & 6.4 & 4.19 & 4.97 & 3.71 & ... & 2.21 & 1.92 & 2.10 & ... & ... & 0.07 + [ x / fe ] & @xmath1463.17 & @xmath1463.15 & 2.48 & 1.79 & 1.19 & 0.60 & 0.57 & ... & 0.48 & 0.19 & @xmath1460.38 & ... & ... & 1.06 + @xmath199 & 19 & 3 & cc5170 & cn4215 & 2 & 3 & 1 & ... & 1 & 3 & 1 & ... & ... & 4 + s.e.([x / fe ] ) & 0.32 & 0.22 & 0.20 & 0.27 & 0.21 & 0.21 & 0.21 & ... & 0.21 & 0.28 & 0.20 & ... & ... & 0.22 + he 11570518 & & & & & & & & & & & & & & + @xmath195 & 5.11 & 5.10 & 8.2 & 7.0 & 4.74 & 5.69 & 4.13 & ... & 2.66 & 2.65 & 3.02 & ... & ... & 1.97 + [ x / fe ] & @xmath1462.34 & @xmath1462.35 & 2.15 & 1.56 & 0.92 & 0.50 & 0.16 & ... & 0.10 & 0.09 & @xmath1460.28 & ... & ... & 2.14 + @xmath199 & 8 & 3 & cc5635 & cn4215 & 2 & 1 & 2 & ... & 1 & 1 & 1 & ... & ... & 4 + s.e.([x / fe ] ) & 0.31 & 0.19 & 0.29 & 0.56 & 0.18 & 0.16 & 0.16 & ... & 0.20 & 0.23 & 0.18 & ... & ... & 0.24 + he 13191935 & & & & & & & & & & & & & & + @xmath195 & 5.71 & 5.72 & 8.1 & 6.5 & 4.95 & 6.26 & 5.06 & ... & 3.26 & 3.35 & 4.65 & ... & ... & 2.31 + [ x / fe ] & @xmath1461.74 & @xmath1461.73 & 1.45 & 0.46 & 0.52 & 0.47 & 0.50 & ... & 0.10 & 0.19 & 0.75 & ... & ... & 1.89 + @xmath199 & 15 & 3 & cc5635 & cn4215 & 1 & 1 & 3 & ... & 1 & 1 & 1 & ... & ... & 2 + s.e.([x / fe ] ) & 0.33 & 0.22 & 0.29 & 0.56 & 0.21 & 0.20 & 0.20 & ... & 0.23 & 0.25 & 0.21 & ... & ... & 0.26 + he 14290551 & & & & & & & & & & & & & & + @xmath195 & 4.98 & 4.97 & 8.2 & 6.7 & 4.84 & 5.58 & 4.55 & ... & 2.66 & 2.80 & ... & ... & ... & 1.27 + [ x / fe ] & @xmath1462.47 & @xmath1462.48 & 2.28 & 1.39 & 1.14 & 0.52 & 0.70 & ... & 0.23 & 0.37 & ... & ... & ... & 1.57 + @xmath199 & 17 & 3 & cc5635 & cn4215 & 2 & 1 & 2 & ... & 2 & 1 & ... & ... & ... & 4 + s.e.([x / fe ] ) & 0.20 & 0.16 & 0.19 & 0.52 & 0.14 & 0.15 & 0.12 & ... & 0.13 & 0.18 & ... & ... & ... & 0.17 + he 1447 + 0102 & & & & & & & & & & & & & & + @xmath195 & 4.98 & 5.00 & 8.4 & 6.7 & 4.77 & 6.50 & 4.73 & ... & 2.89 & 2.71 & ... & ... & ... & 2.40 + [ x / fe ] & @xmath1462.47 & @xmath1462.45 & 2.48 & 1.39 & 1.07 & 1.43 & 0.88 & ... & 0.46 & 0.27 & ... & ... & ... & 2.70 + @xmath199 & 12 & 3 & cc5635 & cn4215 & 2 & 1 & 2 & ... & 1 & 1 & ... & ... & ... & 3 + s.e.([x / fe ] ) & 0.27 & 0.14 & 0.20 & 0.44 & 0.12 & 0.13 & 0.13 & ... & 0.11 & 0.22 & ... & ... & ... & 0.14 + he 15231155 & & & & & & & & & & & & & & + @xmath195 & 5.30 & 5.28 & 8.1 & 7.3 & 5.21 & 6.00 & 4.63 & ... & 3.07 & 3.02 & ... & ... & ... & 1.74 + [ x / fe ] & @xmath1462.15 & @xmath1462.17 & 1.86 & 1.67 & 1.19 & 0.62 & 0.46 & ... & 0.32 & 0.27 & ... & ... & ... & 1.72 + @xmath199 & 16 & 3 & cc5635 & cn4215 & 1 & 1 & 4 & ... & 2 & 1 & ... & ... & ... & 3 + s.e.([x / fe ] ) & 0.19 & 0.14 & 0.19 & 0.52 & 0.12 & 0.15 & 0.11 & ... & 0.11 & 0.17 & ... & ... & ... & 0.14 + he 15280409 & & & & & & & & & & & & & & + @xmath195 & 4.85 & 4.86 & 8.2 & 7.2 & 4.69 & 5.76 & 4.74 & ... & ... & ... & ... & ... & ... & 1.87 + [ x / fe ] & @xmath1462.61 & @xmath1462.59 & 2.42 & 2.03 & 1.13 & 0.83 & 1.04 & ... & ... & ... & ... & ... & ... & 2.30 + @xmath199 & 15 & 1 & cc5635 & cn4215 & 2 & 1 & 1 & ... & ... & ... & ... & ... & ... & 3 + s.e.([x / fe ] ) & 0.19 & 0.14 & 0.19 & 0.52 & 0.12 & 0.13 & 0.10 & ... & ... & ... & ... & ... & ... & 0.14 + he 22210453 & & & & & & & & & & & & & & + @xmath195 & 5.23 & 5.23 & 8.0 & 6.4 & 4.86 & 6.10 & 4.91 & ... & 3.21 & 2.96 & ... & ... & 2.87 & 1.69 + [ x / fe ] & @xmath1462.22 & @xmath1462.23 & 1.83 & 0.84 & 0.92 & 0.80 & 0.82 & ... & 0.54 & 0.28 & ... & ... & 0.49 & 1.75 + @xmath199 & 19 & 2 & cc5635 & cn4215 & 1 & 1 & 3 & ... & 2 & 1 & ... & ... & 1 & 1 + s.e.([x / fe ] ) & 0.31 & 0.20 & 0.31 & 0.48 & 0.18 & 0.17 & 0.17 & ... & 0.21 & 0.24 & ... & ... & 0.26 & 0.24 + he 22280706 & & & & & & & & & & & & & & + @xmath195 & 5.04 & 5.04 & 8.3 & 6.5 & 5.43 & 5.79 & 4.37 & ... & 3.13 & 3.15 & 3.39 & ... & 2.63 & 2.26 + [ x / fe ] & @xmath1462.41 & @xmath1462.42 & 2.32 & 1.13 & 1.67 & 0.67 & 0.47 & ... & 0.64 & 0.66 & 0.16 & ... & 0.44 & 2.50 + @xmath199 & 30 & 2 & cc5635 & cn4215 & 2 & 1 & 6 & ... & 2 & 3 & 2 & ... & 1 & 2 + s.e.([x / fe ] ) & 0.20 & 0.16 & 0.19 & 0.33 & 0.14 & 0.15 & 0.13 & ... & 0.14 & 0.19 & 0.13 & ... & 0.18 & 0.16 + he 23300555 & & & & & & & & & & & & & & + @xmath195 & 4.68 & 4.69 & 7.7 & 6.0 & 4.37 & 5.42 & 3.96 & 0.53 & 2.44 & 2.51 & 2.77 & ... & ... & 0.61 + [ x / fe ] & @xmath1462.78 & @xmath1462.76 & 2.09 & 1.00 & 0.97 & 0.67 & 0.42 & 0.25 & 0.31 & 0.39 & @xmath1460.10 & ... & ... & 1.22 + @xmath199 & 40 & 6 & cc5635 & cn4215 & 2 & 5 & 7 & 2 & 1 & 7 & 1 & ... & ... & 4 + s.e.([x / fe ] ) & 0.19 & 0.14 & 0.19 & 0.42 & 0.12 & 0.17 & 0.19 & 0.08 & 0.11 & 0.23 & 0.10 & ... & ... & 0.25 + lccccc bs 16920005 & 1.46 & 0.981.76 & ... & @xmath202 & ... + cs 22174007 & 1.57 & 0.310.86 & ... & @xmath203 & ... + cs 22886042 & 2.50 & 0.021.35 & ... & @xmath204 & ... + cs 22948027 & 2.27 & 2.20 & @xmath205 & 2.00 & 0.20 + cs 22960053 & 1.87 & 2.50 & ... & 0.40 & ... + cs 29503010 & 0.13 & 0.971.12 & ... & @xmath206 & ... + cs 29528028 & 0.71 & 2.673.09 & ... & @xmath207 & ... + cs 30312100 & 1.47 & 0.441.22 & ... & @xmath202 & ... + cs 30322023 & 2.91 & 1.79 & ... & 0.05 & ... + cs 30338089 & 2.07 & 1.98 & @xmath208 & 25.12 & 2.09 + cs 31062041 & 2.01 & 0.481.03 & ... & @xmath203 & ... + cs 29497034 & 2.64 & 2.70 & @xmath208 & 5.01 & 0.42 + he 02061916 & 1.54 & 2.04 & @xmath209 & 12.59 & 0.84 + he 04002030 & 1.14 & 2.09 & ... & 0.10 & ... + he 04410652 & 2.74 & 1.32 & ... & 12.59 & ... + he 05071653 & 1.77 & 1.23 & @xmath210 & 12.59 & 0.31 + he 10051439 & 2.27 & 2.41 & ... & 19.95 & ... + he 11570518 & 2.14 & 2.08 & @xmath209 & 15.85 & 1.06 + he 13191935 & 2.93 & 1.37 & @xmath211 & 39.81 & 4.98 + he 14290551 & 2.56 & 2.20 & @xmath212 & 31.62 & 1.05 + he
1447 + 0102 & 2.51 & 2.39 & @xmath213 & 50.12 & 2.00 + he 15231155 & 2.50 & 1.83 & ... & 6.31 & ... + he 15280409 & 2.37 & 2.37 & @xmath214 & 10.00 & 0.83 + he 22210453 & 3.55 & 1.75 & @xmath215 & 39.81 & 3.98 + he 22280706 & 1.61 & 2.23 & @xmath216 & 63.10 & 4.21 + he 23300555 & 2.44 & 2.00 & ... & 50.12 & ... + llcccc 1 & bs 16929005 & 2.94 & @xmath1460.3 & 2.64 & @xmath1460.36 + 2 & cs 22174007 & 4.27 & @xmath1460.4 & 3.87 & @xmath1460.10 + 3 & cs 22886042 & 3.91 & @xmath1460.4 & 3.51 & @xmath1460.05 + 4 & cs 22948027 & 4.88 & @xmath1460.4 & 4.48 & 0.52 + 5 & cs 22960053 & ... & ... & ... & ... + 6 & cs 29503010 & 5.27 & ... & 5.27 & 0.16 + 7 & cs 29528028 & 5.99 & @xmath1460.7 & 5.64 & 2.33 + 8 & cs 30312100 & 4.02 & @xmath1460.4 & 3.62 & @xmath1460.23 + 9 & cs 30322023 & 3.96 & ... & 3.96 & 1.04 + 10 & cs 30338089 & 4.45 & @xmath1460.4 & 4.18 & 0.46 + 11 & cs 31062041 & 4.23 & @xmath1460.4 & 3.83 & 0.03 + 12 & he 00392635 & 5.03 &
@xmath1460.4 & 4.63 & 1.37 + 13 & he 02061916 & 4.82 & @xmath1460.4 & 4.42 & 0.34 + 14 & he 04002030 & 5.15 & @xmath1460.4 & 4.95 & 0.51 + 15 & he 04410652 & 4.42 & @xmath1460.4 & 4.02 & 0.32 + 16 & he 05071653 & 5.22 & @xmath1460.4 & 5.02 & 0.23 + 17 & he 10051439 & 4.19 & @xmath1460.4 & 3.79 & 0.79 + 18 & he 11570518 & 4.57 & @xmath1460.4 & 4.17 & 0.34 + 19 & he 13191935 & ... & ... & ... & ... + 20 & he 14290551 & 4.65 & @xmath1460.3 & 4.35 & 0.65 + 21 & he 1447 + 0102 & 4.77 & @xmath1460.4 & 4.37 & 0.67 + 22 & he 15231155 & ... & ... & ... & ... + 23 & he 15280409 & 4.69 & @xmath1460.4 & 4.29 & 0.73 + 24 & he 22210453 & ... & ... & ... & ... + 25 & he 22280706 & ... & ... & ... & ... + 26 & he 23300555 & 4.37 & @xmath1460.4 & 3.97 & 0.58 + na1 & hd 196944 & 5.18 & @xmath1460.4 & 4.78 & 0.86 + na2 & lp 7067 & 4.92 & @xmath1460.7 & 4.22 & 0.6 + na3 & cs 22183015 & 4.03 & @xmath1460.5 & 3.53 & 0.11 + na4 & cs 22880074 & 4.89 & @xmath1460.7 & 4.19 & @xmath1460.05 + na5 & cs 22898027 & 4.74 & @xmath1460.5 & 4.24 & 0.33 + na6 & cs 22957027 & 4.56 & @xmath1460.4 & 4.36 & 1.31 + na7 & cs 29526110 & 4.02 & @xmath1460.3 & 3.72 & @xmath1460.07 + na8 & cs 30301015 & 4.92 & @xmath1460.3 & 4.62 & 1.09 + na9 & cs 31062050 & 4.2 & ... & 4.2 & 0.34 + lrrrrrrrrrrrrrrrl bs 16920005 & @xmath1463.17 & 1.08 & @xmath1972.39 & @xmath1460.36 & 0.49 & 0.13 & & 0.44 & @xmath1460.34 & & & @xmath1460.48 & 5250 & 2.80 & 1.46 & + cs 22948027 & @xmath1462.21 & 2.12 & 2.43 & 0.52 & 0.55 & 0.57 & & 0.64 & @xmath1460.12 & & 0.31 & 2.31 & 5000 & 1.90 & 2.27 & + cs 22960053 & @xmath1463.14 & 2.05 & 3.06 & & 0.65 & 0.43 & 0.49 & 0.54 & @xmath1460.23 & & & 0.86 & 5200 & 2.10 & 2.14 & + cs 29503010 & @xmath1461.06 & 1.07 & @xmath1971.28 & 0.16 & 0.36 & 0.16 & 0.31 & 0.39 & @xmath1460.12 & 0.01 & 0.32 & 1.50 & 6500 & 4.50 & 0.13 & + cs 29528028 & @xmath1462.86 & 2.77 & @xmath1973.58 & 2.33 & 1.69 & 0.46 & 0.59 & 0.87 & & 0.26 & & 3.27 & 6800 & 4.00 & 0.71 & + cs 30322023 & @xmath1463.25 & 0.56 & 2.47 & 1.04 & 0.54 & 0.30 & 0.28 & 0.21 & @xmath1460.36 & & & 0.59 & 4300 & 1.00 & 2.91 & + cs 30338089 & @xmath1462.45 & 2.06 & 1.27 & 0.46 & 0.48 & 0.51 & 1.26 & 0.55 & @xmath1460.01 & & 0.46 & 2.22 & 5000 & 2.10 & 2.07 & + he 00392635 & @xmath1462.91 & 2.72 & 2.63 & 1.37 & 1.31 & 0.52 & & 0.46 & @xmath1460.15 & & & 2.23 & 4900 & 1.50 & 2.64 & + he 02061916 & @xmath1462.09 & 2.10 & 1.61 & 0.34 & 0.52 & 0.11 & & 0.50 & 0.00 & & & 1.97 & 5200 & 2.70 & 1.54 & + he 04002030 & @xmath1461.73 & 1.14 & 2.75 & 0.51 & 0.62 & 0.18 & 0.32 & 0.34 & @xmath1460.08 & @xmath1460.10 & @xmath1460.20 & 1.64 & 5600 & 3.50 & 0.87 & + he 04410652 & @xmath1462.47 & 1.38 & 0.89 & 0.32 & 0.35 & 0.31 & 0.14 & 0.31 & @xmath1460.26 & 0.12 & 0.17 & 1.11 & 4900 & 1.40 & 2.74 & + he 05071653 & @xmath1461.38 & 1.29 & 0.80 & 0.23 & 0.19 & 0.58 & 0.20 & 0.43 & 0.07 & 0.25 & 0.12 & 1.89 & 5000 & 2.40 & 1.77 & + he 10051439 & @xmath1463.17 & 2.48 & 1.79 & 0.79 & 0.60 & 0.57 & & 0.33 & @xmath1460.38 & & & 1.06 & 5000 & 1.90 & 2.27 & + he 11570518 & @xmath1462.34 & 2.15 & 1.56 & 0.34 & 0.50 & 0.16 & & 0.09 & @xmath1460.28 & & & 2.14 & 4900 & 2.00 & 2.14 & + he 13191935 & @xmath1461.74 & 1.45 & 0.46 & & 0.47 & 0.50 & & 0.14 & 0.75 & & & 1.89 & 4600 & 1.10 & 2.93 & + he 14290551 & @xmath1462.47 & 2.28 & 1.39 & 0.65 & 0.52 & 0.70 & & 0.30 & & & & 1.57 & 4700 & 1.50 & 2.56 & + he
1447 + 0102 & -2.47 & 2.48 & 1.39 & 0.67 & 1.43 & 0.88 & & 0.37 & & & & 2.70 & 5100 & 1.70 & 2.51 & + he 15231155 & @xmath1462.15 & 1.86 & 1.67 & & 0.62 & 0.46 & & 0.30 & & & & 1.72 & 4800 & 1.60 & 2.50 & + he 15280409 & @xmath1462.61 & 2.42 & 2.03 & 0.73 & 0.83 & 1.04 & & & & & & 2.30 & 5000 & 1.80 & 2.37 & + he 22210453 & @xmath1462.22 & 1.83 & 0.84 & & 0.80 & 0.82 & & 0.41 & & & 0.49 & 1.75 & 4400 & 0.40 & 3.55 & + he 22280706 & @xmath1462.41 & 2.32 & 1.13 & & 0.67 & 0.47 & & 0.65 & 0.16 & & 0.44 & 2.50 & 5100 & 2.60 & 1.61 & + he 23300555 &
@xmath1462.78 & 2.09 & 1.00 & 0.58 & 0.67 & 0.42 & 0.25 & 0.35 & @xmath1460.10 & & & 1.22 & 4900 & 1.70 & 2.44 & + cs 29498043 & @xmath1463.54 & 2.09 & 2.27 & 1.47 & 1.75 & 0.16 & @xmath1460.03 & 0.30 & @xmath1460.38 & @xmath1460.36 & & @xmath1460.46 & 4600 & 1.20 & 2.83 & 1 + cs 22957027 & @xmath1463.12 & 2.37 & 1.62 & 1.31 & 0.69 & 0.14 & & 0.35 & @xmath1460.21 & @xmath1460.25 & & @xmath1461.23 & 5100 & 1.90 & 2.31 & 2 + cs 29526110 & @xmath1462.38 & 2.20 & 1.40 & @xmath1460.07 & 0.30 & 0.36 & & 0.12 & @xmath1460.33 & @xmath1460.11 & & 2.11 & 6500 & 3.20 & 1.43 & 2,3 + cs 22898027 & @xmath1462.26 & 2.20 & 0.90 & 0.33 & 0.41 & 0.40 & & 0.41 & @xmath1460.10 & 0.02 & & 2.23 & 6250 & 3.70 & 0.86 & 2,3 + cs 31062012 & @xmath1462.55 & 2.10 & 1.20 & 0.60 & 0.45 & 0.29 & & 0.36 & @xmath1460.23 & @xmath1460.02 & & 1.98 & 6250 & 4.50 & 0.06 & 2,3 + cs 22880074 & @xmath1461.93 & 1.30 & @xmath1460.10 & @xmath1460.09 & 0.46 & 0.67 & & 0.41 & 0.20 & 0.13 & & 1.31 & 5850 & 3.80 & 0.64 & 2,3 + cs 31062050 & @xmath1462.31 & 2.00 & 1.20 & 0.34 & 0.60 & 0.85 & & 0.42 & @xmath1460.02 & & & 2.30 & 5600 & 3.00 & 1.37 & 2,3 + hd 196944 & @xmath1462.25 & 1.20 & 1.30 & 0.86 & 0.42 & 0.33 & & 0.27 & @xmath1460.13 & 0.02 & 0.09 & 1.10 & 5250 & 1.80 & 2.46 & 2,3 + cs 22942019 & @xmath1462.64 & 2.00 & 0.80 & & 0.58 & 0.68 & & 0.39 & & & 0.41 & 1.92 & 5000 & 2.40 & 1.77 & 2,3 + cs 30301015 & @xmath1462.64 & 1.60 & 1.70 & 1.09 & 0.86 & 0.77 & & 0.23 & & & 0.33 & 1.45 & 4750 & 0.80 & 3.28 & 2,3 + cs 30314067 & @xmath1462.85 & 0.50 & 1.20 & @xmath1460.08 & 0.42 & 0.22 & 0.02 & 0.12 & @xmath1460.38 & @xmath1460.10 & & @xmath1460.57 & 4400 & 0.70 & 3.25 & 4 + cs 29502092 & @xmath1462.76 & 1.00 & 0.70 & @xmath1460.01 & 0.37 & 0.27 & @xmath1460.02 & 0.19 & & 0.16 & & @xmath1460.82 & 5000 & 2.10 & 2.07 & 4 + cs 22877001 & @xmath1462.72 & 1.00 & 0.00 & @xmath1460.24 & 0.29 & 0.42 & @xmath1460.04 & 0.15 & @xmath1460.29 & @xmath1460.25 & & @xmath1460.49 & 5100 & 2.20 & 2.01 & 4 + lp 62544 & @xmath1462.72 & 2.25 & 0.95 & 1.75 & 1.12 & 0.35 & 0.38 & 0.30 & @xmath1460.21 & @xmath153 & & 2.81 & 5500 & 2.50 & 1.84 & 5 + cs 29497030 & @xmath1462.57 & 2.47 & 2.12 & 0.58 & 0.44 & 0.47 & 0.67 & 0.64 & 0.03 & 0.04 & & 2.32 & 7000 & 4.10 & 0.66 & 6 + cs 22881036 & @xmath1462.06 & 1.96 & 1.00 & 0.16 & 0.40 & 0.62 & & 0.33 & & & & 1.93 & 6200 & 4.00 & 0.55 & 7 + cs 21481247 & @xmath1462.32 & 1.91 & 1.65 & & 0.50 & 0.45 & 0.59 & 0.55 & @xmath1460.35 & 0.06 & & 2.36 & 6380 & 3.90 & 0.69 & 8 + cs 00242523 & @xmath1462.72 & 2.60 & 2.10 & @xmath1460.17 & 0.73 & 0.66 & 0.37 & 0.85 & @xmath1460.41 & & & 1.46 & 6625 & 4.30 & 0.36 & 9 +
cs 22949037 & @xmath1463.99 & 1.17 & 2.57 & 1.44 & 1.58 & 0.35 & 0.10 & 0.35 & @xmath1460.41 & @xmath1460.07 & 0.70 & @xmath1460.58 & 4900 & 1.50 & 2.60 & 10 + he 00071832 & @xmath1462.65 & 2.55 & 1.87 & & 0.76 & 0.32 & & 0.39 & & 0.02 & & 0.16 & 6515 & 3.80 & 0.83 & 11 + he 03383945 & @xmath1462.42 & 2.13 & 1.55 & 0.36 & 0.30 & 0.38 & 0.53 & 0.37 & @xmath1460.12 & 0.01 & & 2.41 & 6160 & 4.13 & 0.40 & 12 + bs16080175 & @xmath1461.86 & 1.75 & & & & & & 0.47 & & & 0.20 & 1.55 & 6240 & 3.70 & 0.86 & 13 + bs17436058 & @xmath1461.78 & 1.50 & 1.25 & & & & & 0.33 & & & 0.01 & 1.61 & 5690 & 2.68 & 1.72 & 13 + cs 22887048 & @xmath1461.70 & 1.84 & & & & & & 0.48 & & & & 2.00 & 6500 & 3.35 & 1.28 & 13 + he 01313953 & @xmath1462.71 & 2.45 & & & 0.30 & 0.37 & 0.59 & 0.13 & @xmath1460.21 & @xmath1460.28 & & 2.20 & 5928 & 3.80 & 0.67 & 14 + he 1105 + 0027 & @xmath1462.42 & 2.00 & & & 0.47 & 0.47 & 0.28 & 0.32 & 0.05 & @xmath1460.29 & & 2.45 & 6132 & 3.50 & 1.03 & 14 + he 02022204 & @xmath1461.98 & 1.16 & & & @xmath1460.01 & 0.26 & @xmath1460.04 & @xmath1460.06 & @xmath1460.40 & 0.04 & 0.03 & 1.41 & 5280 & 1.70 & 2.57 & 14 + he 1135 + 0139 & @xmath1462.33 & 1.19 & & & 0.33 & 0.38 & 0.15 & 0.14 & @xmath1460.24 & 0.01 & & 1.13 & 5487 & 1.80 & 2.53 & 14 + cs 22183015 & @xmath1462.75 & 1.92 & 1.77 & 0.11 & 0.54 & 0.96 & 0.56 & 0.56 & @xmath1460.35 & @xmath1460.33 & & 2.04 & 5620 & 3.40 & 0.97 & 15 + he 01430441 & @xmath1462.31 & 1.98 & 1.73 & & 0.63 & 0.43 & 0.67 & 0.40 & @xmath1460.38 & @xmath1460.31 & 0.46 & 2.32 & 6240 & 3.70 & 0.86 & 15 + he 00121441 & @xmath1462.52 & 1.59 & 0.64 & & 0.91 & 0.42 & & 0.08 & @xmath1460.18 & & & 1.15 & 5730 & 3.50 & 0.91 & 15 + he 02120557 & @xmath1462.27 & 1.74 & 1.09 & & 0.04 & 0.14 & & 0.23 & @xmath1460.09 & & & 2.18 & 5075 & 2.15 & 2.05 & 15 + he 0336 + 0113 & @xmath1462.68 & 2.25 & 1.60 & & 1.04 & 0.42 & & 0.25 & @xmath1460.26 & & & 2.63 & 5700 & 3.50 & 0.90 & 15 + he 10310020 & @xmath1462.86 & 1.63 & 2.48 & & 0.50 & 1.12 & & 0.63 & @xmath1460.38 & & & 1.21 & 5080 & 2.20 & 2.00 & 15 + he 11500428 & @xmath1463.30 & 2.37 & 2.52 & & 0.34 & 1.03 & & 0.49 & @xmath1460.70 & & & @xmath1460.61 & 5200 & 2.55 & 1.69 & 15 + he 1410 + 0213 & @xmath1462.16 & 1.73 & 1.78 & & 0.18 & & & 0.33 & @xmath1460.55 & & & 0.07 & 5605 & 3.50 & 0.87 & 15 + he 14100004 & @xmath1463.02 & 1.99 & & 0.48 & 0.58 & 0.11 & & 0.30 & @xmath1460.19 & & & 1.06 & 4985 & 2.00 & 2.17 & 15 + he 14341442 & @xmath1462.39 & 1.95 & 1.40 & 0.03 & 0.30 & 0.24 & 0.18 & 0.45 & @xmath1460.30 & 0.76 & & 1.23 & 5420 & 3.15 & 1.16 & 15 + he 1443 + 0113 & @xmath1462.07 & 1.84 & & 0.37 & 0.37 & & & & & & & 1.40 & 4945 & 1.85 & 2.30 & 15 + he 15090806 & @xmath1462.91 & 1.98 & 2.23 & & 0.64 & 0.55 & & 0.62 & @xmath1460.08 & & & 1.93 & 5185 & 2.50 & 1.73 & 15 + he 21580348 & @xmath1462.70 & 1.87 & 1.52 & & 0.68 & 0.82 & & 0.54 & @xmath1460.66 & & & 1.59 & 5215 & 2.50 & 1.74 & 15 + he 22320603 & @xmath1461.85 & 1.22 & 0.47 & & 0.85 & 0.35 & 0.05 & 0.23 & @xmath1460.06 & @xmath1460.65 & & 1.41 & 5750 & 3.50 & 0.91 & 15 + bands used for carbon abundance determinations for cs 22960053 and he 14290551 , respectively .
the solid lines indicate the synthetic spectra for adopted carbon abundances , while other lines show those for carbon abundances changed by @xmath2170.3 dex and @xmath2170.2 dex for cs 22960053 and he 14290551 , respectively .
, title="fig:",width=321 ] bands used for carbon abundance determinations for cs 22960053 and he 14290551 , respectively .
the solid lines indicate the synthetic spectra for adopted carbon abundances , while other lines show those for carbon abundances changed by @xmath2170.3 dex and @xmath2170.2 dex for cs 22960053 and he 14290551 , respectively . ,
title="fig:",width=321 ] ) as a function of [ fe / h ] .
the filled circles ( with error bars ) indicate stars from our sample , while filled squares indicate other cemp stars .
non carbon - enhanced objects studied by previous works @xcite are shown by crosses.,width=321 ] ) as a function of luminosity estimated from measured effective temperatures and gravities for stars in our expanded sample .
the meanings of symbols are the same as in figure [ fig : cfe ] .
the results for non carbon - enhanced stars from @xcite @xcite and @xcite are shown here .
the dashed line indicates the dividing line between stars that we define as cemp and the carbon - normal metal - poor stars ( see text ) , width=321 ] , but for [ ba / fe ] . the results for non carbon - enhanced stars from @xcite , @xcite @xcite and @xcite are shown here .
the non carbon - enhanced objects in our sample ( see text ) are shown by open circles here .
the upper limits of [ ba / fe ] are shown for he 13272326 , he 01075240 , and g 7761 .
the dotted lines indicate [ ba / fe]@xmath218 and [ fe / h]@xmath86 .
, width=321 ] ratio as a function of [ c / fe ] for stars in our expanded sample ( upper panel ) .
the meanings of the symbols are the same as in figure [ fig : bafe ] .
stars with [ fe / h ] @xmath219 are indicated by symbols with large open circles .
the solid line indicates the least square fit for stars with [ ba / fe]@xmath87 ( 54 stars ) .
the lower panel is the same as the upper one , but for [ ba / fe ] as a function of [ ( c+n)/fe ] instead of [ c / fe ] . the upper and lower limits of [ ( c+n)/fe ] are connected by lines for stars for which only upper limits to the n abundance are determined ( see text ) .
the dashed line indicates the least square fit for stars with [ ba / fe]@xmath87 for which n abundance is determined ( 44 stars ) , while the solid line shows that for stars having [ fe / h]@xmath1 among them ( 29 stars).,title="fig:",width=321 ] ratio as a function of [ c / fe ] for stars in our expanded sample ( upper panel ) .
the meanings of the symbols are the same as in figure [ fig : bafe ] .
stars with [ fe / h ] @xmath219 are indicated by symbols with large open circles .
the solid line indicates the least square fit for stars with [ ba / fe]@xmath87 ( 54 stars ) .
the lower panel is the same as the upper one , but for [ ba / fe ] as a function of [ ( c+n)/fe ] instead of [ c / fe ] . the upper and lower limits of [ ( c+n)/fe ] are connected by lines for stars for which only upper limits to the n abundance are determined ( see text ) .
the dashed line indicates the least square fit for stars with [ ba / fe]@xmath87 for which n abundance is determined ( 44 stars ) , while the solid line shows that for stars having [ fe / h]@xmath1 among them ( 29 stars).,title="fig:",width=321 ] distribution for cemp stars in our expanded sample .
the histogram with the `` open '' bars indicates the ba - enhanced stars , while that with the hatched area is for the ba - normal stars .
there appears to be a clear difference in the distributions of [ fe / h ] for these two classes of stars ( see text).,width=321 ] , but for the distribution of [ c / h ] for the cemp stars in our expanded sample .
again , there appears a clear difference in the distribution of [ c / h ] for the ba - enhanced and ba - normal stars ( see text).,width=321 ] ( top ) , [ n / h ] ( middle ) and [ ( c+n)/h ] ( bottom ) as a function of luminosity for ba - enhanced cemp stars ( filled symbols ) and ba - normal cemp stars ( open symbols ) .
the solid lines indicate the least square fit to the data of ba - enhanced cemp stars . in the fitting to the [ n / h ] and [ ( c+n)/h ] data , the stars for which only upper limits of n abundances are determined are ignored .
note that , on average , there exists an alteration of the surface abundances of c and c+n by about 0.6 dex , as one progresses from relatively unevolved to evolved stars at higher luminosity.,width=321 ] ratio as a function of effective temperature for stars in our expanded sample .
the circles and squares indicate cemp stars identified by the present and previous work , respectively .
the filled symbols are ba - enhanced cemp stars , while the open symbols are ba - normal cemp stars .
the cross indicates non carbon - enhanced stars .
the dotted line indicates the value of [ c / h ] for which the depth of the ch g band reaches to about 2 % below the local continuum , as a function of temperature , estimated from synthetic spectra .
stars with [ c / h ] below this line would not be detectable in high - resolution spectra of the quality used for this and most previous studies ( see text ) .
, width=321 ] ratios as a function of [ ( c+n)/fe ] for stars in our expanded sample .
the filled and open circles indicate ba - enhanced and ba - normal cemp stars studied by the present work , respectively , while the filled and open squares exibit ba - enhanced and ba - normal cemp stars previously studied , respectively . , | the chemical compositions of 26 metal - poor stars that exhibit strong ch and/or c@xmath0 molecular bands are determined based on high - resolution spectroscopy .
we define carbon - enhanced stars taking account of the carbon abundance ratio ( [ c / fe ] ) and the evolutionary status , which is a slight modification over previous definitions .
twenty two stars in our sample satisfy our modified definition for carbon - enhanced metal - poor ( cemp ) stars .
in addition , we measure na abundances for nine other carbon - enhanced stars for which abundances of other elements have been previously reported . combining our new sample with the results of previous work , we investigate the abundance and evolutionary status of a total of 64 cemp stars .
the following results are obtained : ( 1 ) all but one of the 37 stars with [ fe / h ] @xmath1 exhibit large excesses of barium ( [ ba / fe ] @xmath2 ) , while the other 27 stars with lower metallicity exhibit a large scatter in their barium abundance ratios ( @xmath3 [ ba / fe]@xmath4 + 3.3 ) .
( 2 ) a correlation between the carbon and barium abundance ratios ( [ c / fe ] and [ ba / fe ] ) is found in ba - enhanced objects ( comprising 54 stars ) , suggesting that the origin of the observed carbon excess in ba - enhanced stars is nucleosynthesis in asymptotic giant branch ( agb ) stars , where the main @xmath5-process occurs .
the correlation between the barium abundance ratio and that of carbon plus nitrogen ( [ ( c+n)/fe ] ) is relatively weak , because of the large excesses of nitrogen in some extremely metal - poor stars .
( 3 ) the majority of the ba - enhanced stars have @xmath6 [ c / h ] @xmath7 , and a clear cutoff exists at [ c / h ] @xmath8 , which we take as the limit of carbon - enrichment by metal - poor agb stars . within the above range ,
the [ c / h ] of the ba - enhanced cemp stars decreases , on average , by up to 0.6 dex from the main - sequence turnoff up the red - giant branch , suggesting some dilution of carbon enhancement during their evolution .
the [ c / h ] values of ba - normal stars are relatively low , with a wide distribution .
( 4 ) the difference in the distributions of evolutionary status between ba - enhanced and ba - normal cemp stars suggested by our previous work is not statistically confirmed by the present , enlarged sample .
( 5 ) excesses of na are found in stars with extremely large enhancements of c , n and ba , suggesting efficient production of this element by agb nucleosynthesis .
the implications of these results on the origins of carbon in cemp stars , in particular for ba - normal stars , are discussed . |
there has been considerable interest in understanding the interplay among the crystalline - electric - field ( cef ) effect , the indirect exchange [ ruderman - kittel - kasuya - yoshida ( rkky ) ] interaction among the 4@xmath9 magnetic moments and the kondo effect in ce compounds , since these are the decisive factors of the physical properties in these compounds .
it is therefore worthwhile to carry out careful investigation in new ce compounds . with this motivation ,
we report here the results of magnetic susceptibility ( @xmath10 ) , magnetization ( @xmath11 ) , heat - capacity ( @xmath12 ) , electrical - resistivity ( @xmath13 ) , thermoelectric - power ( @xmath14 ) , and hall - coefficient ( @xmath15 ) measurements on single - crystalline ce@xmath0pdsi@xmath1 , grown for the first time .
this compound has been reported to form in an alb@xmath0-derived hexagonal crystal structure and to exhibit kondo effect.@xcite the intensity of investigation in the re@xmath16si@xmath1 ( re @xmath17 rare earth , @xmath18 @xmath17 transition metal ) series , crystallizing in the above - mentioned structure,@xcite increased only in the recent years and these compounds have been reported to exhibit many unusual features in the magnetic , thermal , and transport properties ( see , for instance , refs .
110 and references therein ) .
gd@xmath0pdsi@xmath1 exhibits kondo - lattice - like anomalies , e.g. , a resistivity minimum above @xmath3 accompanied by a large negative magnetoresistance.@xcite these features , presumably due to a novel mechanism , are not common to gd compounds .
ce@xmath0cosi@xmath1 is a mixed - valent compound , a small la substitution for ce induces a non - fermi - liquid behavior in @xmath13.@xcite eu@xmath0pdsi@xmath1 exhibits two distinct magnetic transitions , with the possibility of quasi - one - dimensional magnetism for the high - temperature transition@xcite and unusual magnetic characteristics.@xcite particularly considering that the gd - based compound in the pd series has been found to show many interesting anisotropic features,@xcite it is tempting to carry out detailed studies on the single - crystalline ce@xmath0pdsi@xmath1 as well .
previous magnetic , electrical - resistance , and heat - capacity measurements on this compound were performed only in the polycrystalline form@xcite and no clear magnetic ordering could be detected .
thus the present studies extended to much lower temperatures , particularly on single crystals , serve as a first thorough characterization of the bulk properties of this compound .
single crystals of ce@xmath0pdsi@xmath1 have been prepared by the czochralski pulling method using a tetra - arc furnace in an argon atmosphere .
the single - crystalline nature has been confirmed by back - reflection laue technique .
the magnetic measurements were carried out with a quantum design superconducting quantum interference device ( squid ) magnetometer .
the heat capacity was measured by a quasiadiabatic heat- pulse method using a dilution refrigerator .
the electrical - resistivity and hall - effect measurements have been performed by a conventional dc four - probe method in the temperature interval of 0.5300 k. the thermoelectric - power data have been taken by the differential method using a au - fe ( 0.07%)-chromel thermocouple .
figure 1(a ) shows the temperature dependence of the inverse magnetic susceptibility @xmath10@xmath19 , measured in a magnetic field @xmath20 1 koe for both @xmath21 $ ] and @xmath22 $ ] . there is a large difference in the absolute values of @xmath10 for two geometries , apparently due to cef effect . the effective magnetic moment ( @xmath23 ) and the paramagnetic curie temperature @xmath24
are estimated from the high - temperature linear region to be about 2.60@xmath25/ce = 8.6 cm and 3.6 k for @xmath21 $ ] and 2.65@xmath25/ce and @xmath8107 k for @xmath22 $ ] , respectively .
these values of @xmath23 are very close to that expected for a free trivalent ce ion ( 2.54@xmath25 ) . the large negative value of @xmath24 for @xmath22 $ ] and a small positive value for @xmath21 $ ] are presumably due to cef effect.@xcite the large anisotropy in @xmath24 may also indicate the existence of anisotropy in the exchange interaction which depends on the cef level scheme .
the expanded view of the temperature dependence of @xmath10 is shown in the inset of fig .
there is no difference between the field - cooled and the zero - field - cooled measurements of @xmath10 down to 1.9 k within the experimental accuracy , indicating the absence of any spin - glass - like behavior .
this fact is in contrast to the formation of spin - glass state in the isostructural u@xmath26pdsi@xmath27.@xcite there is a peak in @xmath28 at around 2.8 k for @xmath21 $ ] and at 2.5 k for @xmath22 $ ] .
these peaks can be ascribed to the antiferromagnetic ordering .
the occurrence of the peak at slightly different temperatures for two directions might be due to the anisotropic field dependence of @xmath6 for two directions , since the heat - capacity measurement in absence of a magnetic field shows a peak at around 3 k as shown below .
recent neutron - diffraction data on polycrystals also suggests the occurrence of antiferromagnetic ( af ) ordering below 2.5 k in a sinusoidally modulated af structure.@xcite we have tried to analyse the @xmath28 data using a cef model , considering hexagonal site symmetry@xcite of ce in ce@xmath0pdsi@xmath1 . according to the hutchings notation,@xcite the cef hamiltonian for @xmath29 ion with the hexagonal point symmetry
is given by @xmath30 where @xmath31 and @xmath32 represent the cef parameters and the steven s equivalent operators , respectively .
the results of this cef analysis using eq .
( 1 ) , employing the @xmath28 data at the paramagnetic region , leads to @xmath33 10.1 k , and @xmath34 0.11 k. this set of parameters corresponds to a crystal - field level scheme with the three doublets @xmath5 , @xmath35 and @xmath36 at around 0 , 28 k ( @xmath37 ) , and 130 k ( @xmath38 ) , respectively .
accordingly , the calculated values of @xmath10@xmath39 are shown by the solid lines in fig .
1(a ) , which indicates that the anisotropy in @xmath28 is mainly induced by the cef effect . however , there is a deviation of the calculated @xmath10@xmath39 from the experimental values , particularly at low temperatures .
the following explanations can be offered to this deviation : according to szytula _
et al_,@xcite pd and si atoms are random in this crystal structure ( space group @xmath40 ) .
therefore this randomness or disorder between pd and si sites may produce local modification of the cef effects due to the distribution of the cef parameters .
alternatively , if pd and si are well ordered ( space group @xmath41 , see ref .
10 ) , there are two different crystallographic environments for ce ions,@xcite in which case the cef effect may be different for these two sites .
all these factors are neglected in the present cef calculations .
the isothermal magnetization at 2 k is shown in fig .
@xmath11 varies distinctly in different ways with the applied magnetic field for @xmath21 $ ] and @xmath22 $ ] .
this anisotropy in @xmath11 is presumably due to the cef effect .
the magnetic moments at @xmath20 50 koe for two directions are @xmath42 1.18 and @xmath42 0.25@xmath25/ce for @xmath21 $ ] and @xmath22 $ ] , respectively .
the larger magnetic moment for @xmath21 $ ] indicates the @xmath43-@xmath44 plane as the easy plane of magnetization , and the @xmath5 doublet as the ground state , in agreement with the cef analysis from @xmath28 described above .
these two facts are also consistent with the proposed magnetic structure of ce@xmath0pdsi@xmath1 based on neutron - diffraction experiment , i.e. , the ce magnetic moments lie in the @xmath43-@xmath44 plane.@xcite = 8.6 cm figure 2 shows the temperature dependence of heat capacity , @xmath45 , for ce@xmath0pdsi@xmath1 .
a peak in @xmath45 at around 3 k indicates the existence of the antiferromagnetic ordering below @xmath46 3 k , supporting the conclusion from the @xmath10@xmath47 behavior .
the transition is , however , not very sharp , which might be due to the presence of site disorders .
the inset shows the @xmath48 vs @xmath49 plot at temperatures far below @xmath6 where phonon contribution is negligible .
the linear coefficient of specific heat ( @xmath50 ) , named as the sommerfeld coefficient , estimated from this plot at such low temperatures is 108 mj / k@xmath4 mol ce . at this temperature
range the specific heat can be expressed as @xmath51 with @xmath52 862 mj / k@xmath53 mol ce and the parameter @xmath54 mainly comes from the contribution of an antiferromagnetic - magnon part . for y@xmath0pdsi@xmath1 , the non - magnetic compound serving as a reference for phonon contribution ,
the @xmath50 value is 4.5 mj / k@xmath4 moly.@xcite the moderately large value of @xmath50 for ce@xmath0pdsi@xmath1 suggests that even in the magnetically - ordered state ce@xmath0pdsi@xmath1 may be classified as a heavy fermion . the solid line in fig .
2 shows the magnetic entropy ( @xmath55 ) estimated from the 4@xmath9 contribution ( @xmath56 ) to @xmath12 .
@xmath56 is obtained by employing the @xmath12 values of y@xmath0pdsi@xmath1 ( ref .
1 ) as a reference for the lattice contribution taking into account the difference of debye temperatures of the two compounds by using the procedure suggested in ref . 14 . at @xmath3 , @xmath57 2.3 j / k mol ce is @xmath42 40% of @xmath58 ln 2 expected for a complete removal of the two fold degeneracy of a cef ground - state doublet .
this reduced entropy value might be due to the substantial kondo - derived reduction of the ce moments and/or a presence of short - range correlations above @xmath3.@xcite tentatively assuming a kondo - derived reduction , the magnetic entropy @xmath55 ( @xmath59 0.4 @xmath58 ln 2 ) at @xmath3 yields the kondo temperature @xmath60 8 k , according to the bethe - ansatz for a spin-@xmath61 kondo model ( see refs .
16 and 17 ) .
the theoretical calculations of @xmath45 using @xmath60 8 k , however , can not reproduce the experimental curve above @xmath3 .
this deviation indicates the presence of short - range - af correlations above @xmath3 ; the presence of these correlations is , in fact , detected in a recent neutron - scattering experiment,@xcite and a possible contribution from an excited cef level ( 28 k ) even at the measured temperature range . the temperature dependence of resistivity @xmath62 for ce@xmath0pdsi@xmath1 with the current @xmath63 $ ] and @xmath64 $ ] as well as in polycrystalline y@xmath0pdsi@xmath1 is shown in fig .
@xmath13 for both current directions gradually decreases with decreasing temperature down to about 20 k showing a broad hump around 100 k ; below 20 k , there is a weak upturn giving rise to a minimum at around 20 k , followed by a drop below 8 k. the @xmath62 in y@xmath0pdsi@xmath1 shows usual metallic behavior , however there is a drop in @xmath13 below 6 k , presumably attributable to the presence of traces of the superconducting phase ypdsi.@xcite as known for many other ce alloys , the broad hump in ce@xmath0pdsi@xmath1 can be ascribed to the combined effect of cef and kondo effect . the magnetic contribution to the resistivity @xmath65@xmath66(ce@xmath0pdsi@xmath1)@xmath67(y@xmath0pdsi@xmath1 ) ,
obtained by using @xmath62 data for @xmath64 $ ] in ce@xmath0pdsi@xmath1 , is shown in the inset ; the ratio of the slopes of the @xmath65 vs ln @xmath68 plot at high to low temperature turns out close to 35/3 confirming that the ground state is a doublet ( see ref . 19 ) .
it clearly reveals the presence of a peak at around 50 k which might be related with @xmath69 enhanced by cef effect as proposed by hanzawa _
et al_.@xcite correspondingly , there is also a broad hump in @xmath70 ( see fig .
the minimum in @xmath62 around 20 k is either due to the kondo effect or a consequence of magnetic - precursor effect as noted for some gd alloys.@xcite it may be noted that the drop in @xmath13 sets in at 8 k , much above the @xmath3 , similar to cept@xmath0ge@xmath0,@xcite = 8.6 cm and such a feature in magnetically ordered kondo lattices arises from a combination of indirect exchange interaction and the kondo effect.@xcite there is a small difference in absolute values for two geometries , which might be due to the combined effect of anisotropy in the fermi surface and preferably oriented microcracks . it may also be remarked that the residual resistivity is large even for the single crystal , which might be due to a presence of crystallographic ( pd - si ) disorder or a combined effect of disorder and a dominance of kondo contribution even in the magnetically ordered state .
figure 4 shows the temperature dependence of thermoelectric power ( @xmath14 ) in ce@xmath0pdsi@xmath1 as well as in y@xmath0pdsi@xmath1 ( polycrystal ) . in ce@xmath0pdsi@xmath1 , for the temperature gradient
@xmath71@xmath72 $ ] , @xmath14 is positive at room temperature , then it gradually increases with decreasing temperature and shows a broad hump around 100 k. there is a change of sign around 50 k with a minimum around 17 k. on the other hand , for @xmath71@xmath73 $ ] , @xmath14 has a large negative value at room temperature and decreases with decreasing = 8.6 cm temperature showing a broad hump around 100 k. @xmath70 for this direction also shows a minimum , however , at a temperature slightly higher than that for @xmath71@xmath72 $ ] .
thus @xmath70 in ce@xmath0pdsi@xmath1 is highly anisotropic with the directions of the thermal gradient .
the anisotropy in the fermi surface might be one reason behind this anisotropy , since @xmath70 in the isostructural gd@xmath0pdsi@xmath1 is also anisotropic.@xcite for nonmagnetic y@xmath0pdsi@xmath1 , @xmath70 has a large negative value at room temperature and decreases gradually with temperature . in ce@xmath0pdsi@xmath1
, the broad hump can be attributed to the interplay between cef and kondo effect.@xcite the origin of the minimum at @xmath74 @xmath59 17 k is not clear yet , however , the possible explanation may be the kondo scattering in the cef ground state or the growth of af correlations as in the case of ceaual@xmath1.@xcite tentatively assuming the kondo - derived origin , @xmath69 @xmath59 8 k would be obtained using the relation @xmath69
@xmath59 @xmath75 that holds for ceal@xmath0 and cecu@xmath26 ( see ref .
however , the temperature dependence of @xmath12 above @xmath3 suggests that this estimation of @xmath69 is a rough one , indicating that both kondo effect and af correlations may play a role for this minimum . for @xmath71@xmath72 $ ]
, @xmath70 is similar to the behavior in the typical magnetically ordered heavy - kondo compounds , e.g. , cecu@xmath0 and ceal@xmath0,@xcite though the overall temperature dependence of @xmath14 in ce@xmath0pdsi@xmath1 is rather weaker .
if the crystallographic disorder is the dominant origin of the large residual resistivity , the kondo contribution to @xmath70 could be suppressed .
since , according to the gorter - nordheim rule , the thermoelectric power for more than one scattering mechanisms can be expressed as @xmath76/\rho$ ] , where the subscripts 1 and 2 correspond to different scattering mechanism:@xcite in the present case , 1 represents the kondo scattering and 2 represents the other scatterings .
therefore a large @xmath77 can suppress the kondo contribution @xmath78 .
the large negative thermoelectric power in y@xmath26pdsi@xmath27 might arise from the 4@xmath79 band of pd , as in the case of the 3@xmath79 band of co in yco@xmath0.@xcite = 8.6 cm the @xmath70 curve for @xmath80//[0001 ] at high temperatures is almost parallel to that of y@xmath0pdsi@xmath1 , therefore a significant effect of the pd-4@xmath79 band on @xmath70 even in ce@xmath0pdsi@xmath1 can not be ruled out .
the temperature dependence of hall coefficient ( @xmath15 ) for @xmath20 15 koe , shown in fig .
5(a ) , also reflects the anisotropic nature of this material . for both geometries ( as labeled in the figure )
@xmath15 is positive at room temperature and increases gradually with decreasing temperature . at low temperatures
@xmath15 becomes highly anisotropic and shows a positive peak for both geometries , in the vicinity of @xmath3 ( see the inset ) .
the large anisotropy observed in @xmath15 is also reflected in the @xmath28 data taken at @xmath20 15 koe ( not shown ) , indicating that the anisotropy in @xmath15 is of magnetic origin .
the positive value of @xmath81 at all temperatures and a positive peak at the vicinity of @xmath3 are similar to those in the antiferromagnetic kondo - lattice compound ceal@xmath0.@xcite in contrast , @xmath15 ( @xmath59 0.9@xmath8210@xmath83 m@xmath84/c at 300 k ) in y@xmath0pdsi@xmath1 is almost temperature independent .
clearly , there is a dominant 4@xmath9 contribution in ce@xmath26pdsi@xmath27 .
the hall coefficient in magnetic materials like those in ce compounds is generally a sum of two terms ; an ordinary hall coefficient ( @xmath85 ) due to lorentz force and an anomalous part arising from magnetic scattering ( skew scattering);@xcite in the paramagnetic state @xmath86 , where @xmath87 is a constant . using this relation ,
@xmath85 is estimated by plotting @xmath15 versus @xmath88 [ fig .
5(b ) ] . from fig .
5(b ) , it is obvious that the plot is linear for both @xmath21 $ ] and @xmath22 $ ] in the paramagnetic state with a value of @xmath893.2@xmath82 10@xmath83 m@xmath84/c and 1.0@xmath8210@xmath90 m@xmath84/c , and @xmath91 7.4@xmath8210@xmath92 mol / c and 4.8@xmath8210@xmath92 mol / c , respectively .
this linear behavior indicates the presence of dominant skew scattering in ce@xmath0pdsi@xmath1 . in the vicinity of @xmath3 , however , the data deviate from the high - temperature linear variation .
@xmath85 of different sign with the anisotropic values for two geometries indicates the presence of anisotropy in the fermi surface , in agreement with the @xmath70 data .
summarizing , we have investigated the magnetic behavior of recently synthesized ce@xmath0pdsi@xmath1 in the single - crystalline form and the results show strong anisotropic behavior of the measured properties .
the paramagnetic curie temperature for @xmath21 $ ] is positive , however , the value is negative for @xmath22 $ ] .
the sign of the thermoelectric power is different for the two measured crystallographic orientations at high temperatures .
distinct features due to an interplay between cef and kondo effect have also been observed in the thermoelectric power and resistivity data .
the ordinary hall - coefficient is anisotropic with opposite sign for the two measured geometries .
the results establish that this compound is an antiferromagnetic kondo lattice with @xmath3= 3 k. the magnitude of @xmath69 is also estimated to be of same order as @xmath3 and this fact suggests a delicate competition between the kondo effect and indirect exchange interaction
. therefore it would be of interest to investigate this compound under high pressure . c. s. garde and j. ray , phys .
b * 51 * , 2960 ( 1995 ) ; j. sakurai , _ transport and thermal properties of f - electron systems _ , edited by g. oomi , h. fujii and t. fujita ( plenum press , new york , 1993 ) , p. 165 ; a. k. bhattacharjee and b. coqblin , phys . rev .
b * 13 * , 3441 ( 1976 ) . | magnetization , heat capacity , electrical resistivity , thermoelectric power , and hall effect have been investigated on single - crystalline ce@xmath0pdsi@xmath1 .
this compound is shown to order antiferromagnetically below nel temperature @xmath2 3 k. the sommerfeld coefficient far below @xmath3 is found to be about 110 mj / k@xmath4 mol ce , which indicates the heavy - fermion character of this compound .
the transport and magnetic properties exhibit large anisotropy with an interplay between crystalline - electric - field ( cef ) and kondo effects .
the sign of thermoelectric power is opposite for different directions at high temperatures and the ordinary hall coefficient is anisotropic with opposite sign for different geometries , indicating the anisotropic fermi surface .
the cef analysis from the temperature dependence of magnetic susceptibility suggests that the ground state is @xmath5 .
the first and the second excited cef doublet levels are found to be located at about 30 and 130 k , respectively . the kondo temperature is estimated to be the same order as @xmath6 , indicating the presence of a delicate competition between the kondo effect and magnetic order . @eqnnum @eqnnum definecounterequation addtoresetequationsection @prefix .
@prefix 0.5 cm pacs number(s ) : 75.30.mb , 71.27.@xmath7a , 71.70.ch , 75.40.@xmath8s |
we have produced several batches of the solid h@xmath4-h@xmath4o compound in the c@xmath0 phase and recovered the samples at room pressure and liquid nitrogen temperature , by application of a standardized and reproducible procedure .
the structure of some specimens has been examined by x - ray diffraction at the crist laboratory of the university of firenze .
the sample examined presented a considerable texture , which prevented a rietvield analysis , but the measured diffraction pattern can be used to discriminate among several proposed structures .
a le bail fit of the diffraction pattern is presented in fig . 2 , assuming the c@xmath0-ii structure , space group @xmath7 .
lattice constant are @xmath12 and @xmath13 .
the other structures considered , namely st@xmath11 ( @xmath14 ) , ih - c@xmath0 ( @xmath15 ) @xcite and ice i@xmath16 ( @xmath17 ) , give diffraction patterns in evident disagreement with the experimental one .
details of the synthesis procedure and of the x - rays measurements are given in the methods section . by using our cryogenic raman apparatus @xcite
we have measured spectra during different @xmath18 cycles , observing the crystal lattice excitations and the rotations of the h@xmath4 molecule ( 150 - 650 @xmath19 ) , the oh stretching mode of the water molecule ( 3000 - 3400 @xmath19 ) , and the vibron of the h@xmath4 molecule ( 4100 - 4200 @xmath19 ) .
our raman spectra ( see fig . 3 ) and overall results add important information on the structure , which are consistent with only one of the proposed structures , namely the c@xmath0-ii one @xcite with space group @xmath7 depicted in fig . 1 .
in this structure the water molecules form spiraling channels with a free bore hole along the @xmath20 axis of about 5.26 and with a diameter of 6.10 , which can accommodate the h@xmath4 molecules . observing the spectra ( fig .
3a ) , we notice that the lattice phonon band ( black line ) has a broad smooth shape , showing similarities with the same band in both ice ih ( blue line ) and sii clathrates ( red line ) .
the absence of sharp lines rules out the possibility that the water lattice in the c@xmath0 phase might be proton ordered .
this is at variance with the c@xmath6 filled ice , which exhibits proton order and , consequently , a lattice phonon spectrum with evident sharp lines @xcite .
the rotational spectrum of the hydrogen molecules ( fig .
3c ) presents well distinct lines , proving that the h@xmath4 molecules rotate almost freely .
the s@xmath0(0 ) rotational band is split in probably three components , with a larger splitting than in clathrate - hydrates .
this is an indication of a more intense interaction of the hydrogen molecule with the environment and a larger perturbation of the rotational motion . according to the proposed structure ,
the hydrogen molecules are arranged in the channels , probably in a spiraling configuration @xcite , at a distance from water oxygen atoms of about 3.1 and at about 2.95 between each other , the same as in solid hydrogen at about 1.5 gpa .
the larger splitting of the rotational band components is accountable on the basis of the shortest distance between h@xmath4 and h@xmath4o molecules in the @xmath7 structure , and indirectly confirms it . in addition , at such a short distance the h@xmath4-h@xmath4 anisotropic interaction may sustain collective rotational excitations , as it happens in solid h@xmath4 @xcite and ar(h@xmath4)@xmath4 high pressure compound @xcite , which may contribute to the width of the rotational lines . the spectrum of the h@xmath4 vibron region ( fig .
3d ) consists of only one doublet ( q@xmath6(0 ) and q@xmath6(1 ) lines ) , and this is in accordance with the occupation of a single crystallographic site for the hydrogen molecules , as in the proposed structure @xcite .
the oh stretching band ( fig .
3b ) has the main peak at 3123 @xmath19 , a frequency higher than that of both sii clathrate and ice ih .
the oh vibrational frequency is known to decrease while decreasing the o - o distance in different ices .
this is observed both as a function of the considered compound @xcite or , for the same structure , as a function of pressure , as reported for ice vii @xcite and c@xmath6 filled ice @xcite
. the higher frequency of the oh stretching mode observed for the c@xmath0 filled ice is in accordance with the larger o - o distance in the structure @xmath7 ( @xmath21 ) than in sii clathrate ( @xmath22 ) and in ice ih ( @xmath23 ) .
we derive the hydrogen molar fraction @xmath24 from the intensity ratio of the hydrogen rotational lines and the lattice phonon band , @xmath25 .
this analysis is done thanks to the assumption , that sounds obvious , that the raman intensity of a band is proportional to the number of molecules giving rise to it .
this is anyhow an approximation , since difference of the polarisability of the same molecules in different environments may lead to differences in intensity .
this matter has been considered in sec .
iiia of ref .
@xcite , and we refer to that paper for a more thorough discussion .
we calculate the calibration factor by fitting data arising from two independent experimental methods .
a first set of data is obtained by using similar spectra of sii clathates , for which the hydrogen content is calculated by counting the number of molecules in the large and small cages @xcite .
the spectra of these samples in the 200 - 900 @xmath19 range ( not reported in ref .
@xcite ) are used to calculate @xmath25 .
to these data we add the information we derive from new volumetric measurements performed on our c@xmath0 sample at three temperatures , namely 20 , 50 and 80 k. the procedure is explained in detail in the methods section .
for the samples examined just after synthesis , whose typical spectra are represented in fig . 3
, we obtain @xmath26% .
we discover that the c@xmath0 structure is metastable at room pressure even when emptied , and is a new form of ice , namely ice xvii .
the mechanical stability of the sample is tested by increasing the temperature while keeping the sample under dynamic vacuum .
we then observe the gradual release of the hydrogen from the sample , up to the complete undetectability of the hydrogen rotational lines obtained after pumping for 1 - 2 hours at a temperature of about 120 k. considering the sensitivity and the signal to noise ratio of our detection system , we can ascertain that the h@xmath4/h@xmath4o molar fraction in the emptied sample is less then 0.5% . during the heating process
, we do not observe any abrupt change of the lattice phonon and oh stretching bands , demonstrating that no structural phase transitions have occurred .
decreasing again the temperature of ice xvii and comparing the spectra measured at the same low temperature we notice after the annealing an increase in the frequency of about 7 @xmath19 for the lattice phonon bands
4a ) and a decrease of about 20 @xmath19 for the oh mode ( fig .
this indicates a decrease of the average o - o distance . the same indication we obtain from the concurrent increase of the lattice modes , indicating a stronger binding .
however , the overall similarity of the band shape before and after annealing leads us to believe that the structure of the filled c@xmath0 ice and that of ice xvii are essentially the same .
the situation for the c@xmath0 structure is different from previous observation in sii , where upon cages emptying , the lattice constant , and consequently the o - o distance increases @xcite .
the contraction of the water framework upon inclusion of ne in the cages of sii clathrate may be imputed to a attractive interaction between this atom and water .
it is interesting to compare also with the situation of the he - hydrate , which become ice ii upon emptying @xcite . here
the phenomenology is more complex , but the changes in the structure can be related to the mainly repulsive interaction between the he and o atoms .
the precise nature of the effect of the annealing on the structure of the c@xmath0 samples is not clear yet .
the increased sharpness of the spectral features measured after the annealing suggests the decrease of defects in the structure .
we recall that recently some authors @xcite , in order to explain the non perfect refinement of the x - rays diffraction pattern with the @xmath7 structure , have hypothesized that some nitrogen molecules , instead of water molecules , may have been trapped in the channels during the recovering of the sample in liquid nitrogen
. a possible effect of the annealing might be the removal of these nitrogen molecules . to check this hypothesis ,
we have measured raman spectra of our samples , before and after this annealing process , in the region of the n@xmath4 stretching mode , observing , at 20 k before the annealing , a quite evident and sharp line at a frequency @xmath27 @xmath19 . by virtue of their vibration frequency ,
sensibly lower than both that of n@xmath4 gas ( 2329.917 @xmath19 ) @xcite , and that of solid n@xmath4 in the @xmath10-phase ( a doublet at 2327.5 and 2328.5 @xmath19 ) @xcite , we can establish that the n@xmath4 molecules are trapped inside the structure .
the raman line essentially disappears ( it becomes 1000 times weaker ) after the annealing process .
the evident spectroscopic signatures of structural modification described above are a combined effects of the removal of both h@xmath4 and n@xmath4 from the sample .
the unexpected property of ice xvii is that , when exposed to even a low pressure of hydrogen gas , it adsorbs the molecules up to an amount which is pressure dependent , but may grows larger than that initially present just after synthesis . as a matter of fact
, we generally obtain @xmath28% for pristine samples , possibly because of the presence of adsorbed nitrogen in the channel and/or of the handling of the sample for the insertion of the raman apparatus .
the adsorption process has initially a fast kinetics , even at 15 k. the rotational raman spectra measured at 40 k after subsequent partial refilling steps are shown in fig .
maximum gas pressure is in this case only 250 mbar , but the rotational intensity grows higher than that in the pristine c@xmath0 sample .
the shape of the rotational lines does not change significantly , demonstrating that h@xmath4 is penetrating again in the same positions in the channels as those initially occupied at the time of synthesis .
normalizing the rotational raman intensity with respect to the lattice phonon band and applying the calculated conversion constant , we estimate the h@xmath4/h@xmath4o molar fraction .
we notice that , in some instances , as for the most intense spectrum shown in fig 4c , the measured h@xmath4/h@xmath4o molar ratio in the sample exceeds 40 % , reaching higher values than the established theoretical maximum ( @xmath29 ) for sii hydrates .
we have measured then several adsorption isotherms , deriving the amount of adsorbed h@xmath4 as a function of pressure , at several different temperatures between 15 to 80 k. results for four temperatures are presented in fig .
pressure has been increased gradually ( squares in fig .
5 ) and , when possible , measurements have been taken after the same delay @xmath30 from pressure increase , which is indicated in each figure .
the pressure at which saturation is reached is strongly dependent on temperature and ranges between a few millibar at 15 k to several bar at 80 k. by measuring with the same method , we have also examined the gas desorption while releasing pressure .
a large hysteresis and other kinetics effects are evident at the lower temperatures . at 15 k , after decreasing pressure down to zero , the sample does not sensibly release hydrogen in @xmath31 min .
the same effect is observed at 40 k , the filling remaining almost constant for at least 36 min . at this temperature
, the increase in the adsorbed gas observed at about 240 mbar , after a 16.5 hour long exposure to hydrogen , indicates that at least two time scales dominate the phenomenon : one fast adsorption in the outermost sites of the channels and a slower one , limited by molecular diffusion inside the channels . at higher temperatures , 65 and 80 k
, we observe a very small hysteresis , indicating that the time needed to hydrogen to diffuse for distance of the order of the linear dimension of the sample grains is quite shorter than the time interval between measurements . at temperature above 50 k ,
the desorption process is quick ( compared to sii clathrates ) .
this is not surprising at all , since the channels in ice xvii are much wider than the free space inside both the 6-membered and 5-membered rings .
from the isotherm hydrogen uptake as a function of the pressure presented in fig .
5 , it is possible to estimate , by means of the van@xmath32 t hoff equation , the enthalpy of adsorption , @xmath33 , for a given hydrogen uptake @xmath24 : @xmath34 where @xmath35 is the gas constant .
@xcite for this calculation we use the data collected at 50 , 65 and 80 k , estimating the derivative with a straight line fit .
the resulting values of @xmath36 decrease with increasing molar ratio @xmath24 starting from about 5 kj / mol at @xmath37 down to about 2 kj / mol at @xmath38 , in agreement with typical values for physisorption .
this work is the first characterization of the dynamics of the c@xmath0 filled ice .
moreover , two remarkable results arise from this study . first , we discovered that the c@xmath0 structure is stable when emptied . due to its large range of stability ( up to 130 k at least ) ,
it should be counted as a new form of ice and named ice xvii .
second , we found that ice xvii is capable to adsorb hydrogen gas in large quantities ( presumably , on the basis of its structure up to @xmath39 ) , without substantial change in its structure .
we believe that , more than the maximum estimated hydrogen uptake , it is the fast reversibility of gas adsorption , the theoretically infinite number of possible cycles ( there is no chemical change of the substance involved ) and the relatively modest pressures at which the process occurs , at temperature close to that of liquid nitrogen , that make ice xvii really appealing for hydrogen storage applications .
we have produced several instances of solid h@xmath4-h@xmath4o compound by application of a standardized procedure .
the check of the recovered samples with raman scattering confirms that we have always obtained the sample in the same c@xmath0 phase . to synthesize the samples ,
we introduce a few grams of finely ground ice into a beryllium - copper autoclave , and we expose the sample to hydrogen gas at a pressure in the range or above 430 mpa , at a temperature of about 255 k ( -18@xmath2c ) .
we prudentially leave the sample under pressure for a few days , even though , while rising pressure during synthesis , we observe at about 360 mpa an abrupt decrease of the rate of pressure rise , which is a clear indication of large hydrogen adsorption by the ice . then we quench the autoclave in liquid nitrogen when still under pressure , and finally , after releasing the pressure , we recover the sample at liquid nitrogen temperature , in the form of a fine powder .
the same procedure is applied for the synthesis of the samples to be investigated by x - rays diffraction , except that in this case a drop of water is frozen from the beginning inside a few common x - rays glass capillaries , each of them 10 - 15 mm long .
the frozen water occupies a capillary length of about 1 mm .
the capillaries are inserted in the autoclave where they undergo the same standardized procedure , and are then recovered and handled at liquid nitrogen temperature . for raman measurements , a small amount of sample is inserted in our optical cell , in contact with the cold finger of a closed - cycle he cryostat @xcite .
the transfer process has to be accomplished at almost liquid nitrogen temperature and in a dry - nitrogen atmosphere , exerting particular care to avoid sample heating .
once the raman cell is filled with the sample , it is sealed with the optical window and purged with helium gas .
the sample cell can then be set to any temperature between room temperature and @xmath40 k. gas pressure in the cell can be varied and measured with an accuracy of about 1 % .
raman spectra are excited by means of an ar ion laser at 514.5 nm , with a power of about 30 mw on the sample , focused on a spot of about 30 @xmath41 m .
scattered light is collected in an almost back - scattering geometry , focused on the entrance slit of a spex spectrograph , and recorded by a cooled ccd detector ( andor ) .
the maximum resolution of the instrument is 0.4 @xmath19 .
similar care has to be exerted to mount the capillary on the x - ray diffractometer ( four circle oxford diffraction xcaliburpx ) the capillary is maintained at low temperature ( about 90 - 100 k ) by a cold nitrogen blow provided by a oxford cryosteam system .
diffractograms are collected using a copper x - ray source , and recorded by a 165 diameter ccd detector .
total collection time is of the order of 7 minutes . during collection ,
the capillary is turning rapidly .
we measure the molar fraction of hydrogen molecules contained in the samples of filled ice by computing the ratio of the intensity of the s@xmath0(0 ) and s@xmath0(1 ) rotational lines , weighted by the inverse of the respective cross section with the intensity of the lattice phonon band . specifically ,
if we indicate with @xmath42 and @xmath43 the raman cross section for the s@xmath0(0 ) and s@xmath0(1 ) rotational lines , and with @xmath44 and @xmath45 their intensity , respectively , the quantity @xmath46 is proportional to the number of h@xmath4 molecules in the lowest two @xmath47 and @xmath48 rotational states , that is , at low temperature , to the number of h@xmath4 molecules giving raman scattering . dividing @xmath49 by the intensity of the water lattice phonon band , @xmath50
, we obtain a quantity which is independent of laser intensity and detector efficiency , and proportional to the h@xmath4/h@xmath4o molar ratio in the sample .
this procedure needs a calibration , that we have accomplished fitting data obtained with two independent methods . for several samples of sii clathates studied in our laboratory ,
the hydrogen content has been indirectly derived , by counting the molecules in the small and large cages @xcite , while the spectra of the same samples in the 200 - 900 @xmath19 range are in our records , even if not reported in ref .
this allows us to draw some calibration points ( black squares ) in fig .
we have performed additional volumetric measurements on our c@xmath0 sample while measuring , with raman scattering , the adsorption isotherms at three temperatures , namely 20 , 50 and 80 k. these results are reported as colored dots in fig .
the estimated uncertainty on these experimental points depends critically on the precision of the different calibrated volumes , of the pressure transducer used and on the pressure of the measurement .
the cumulative linear fit gives a calibration factor of @xmath51 , that is the value we have used in this work .
huang , y. _ et al .
_ a new phase diagram of water under negative pressure : the rise of the lowest - density clathrate s - iii .
_ science advances _ * 2 , * e1501010 ( 2016 ) .
falenty , a. , hansen , t. c. & kuhs , w. f. formation and properties of ice xvi obtained by emptying a type sii clathrate hydrate .
_ nature _ * 516 , * 231 - 233 ( 2014 ) .
ulivi , l. _ et al .
_ quantum rattling of molecular hydrogen in clathrate hydrate nanocavities .
b _ * 76 , * 161401 ( 2007 ) . xu , m. , ulivi , l. , celli , m. , colognesi , d. & bai ' c , z. quantum calculation of inelastic neutron scattering spectra of a hydrogen molecule inside a nanoscale cavity based on rigorous treatment of the coupled translation - rotation dynamics .
b _ * 83 , * 241403 ( 2011 ) . xu , m. , ulivi , l. , celli , m. , colognesi , d. & bai , z. rigorous quantum treatment of inelastic neutron scattering spectra of a heteronuclear diatomic molecule in a nanocavity : hd in the small cage of structure ii clathrate hydrate . _
* 563 , * 1 ( 2013 ) .
celli , m. _ et al .
_ experimental inelastic neutron scattering spectrum of hydrogen hexagonal clathrate - hydrate compared with rigorous quantum simulations .
_ j. chem .
phys . _ * 139 , * 164507 ( 2013 ) .
colognesi , d. , celli , m. , ulivi , l. , xu , m. & bai , z. neutron scattering measurements and computation of the quantum dynamics of hydrogen molecules trapped in the small and large cages of clathrate hydrates .
_ j. phys .
a _ * 117 , * 7314 - 7326 ( 2013 ) .
giannasi , a. , celli , m. , ulivi , l. & zoppi , m. low temperature raman spectra of hydrogen in simple and binary clathrate hydrates .
_ j. chem .
* 129 , * 084705 ( 2008 ) .
strobel , t. a. , sloan , e. d. & koh , c. a. raman spectroscopic studies of hydrogen clathrate hydrates .
_ j. phys .
chem . _ * 130 * , 014506 1 - 10 ( 2009 ) , giannasi , a. , celli , m. , zoppi , m. , moraldi , m. & ulivi , l. experimental and theoretical analysis of the rotational raman spectrum of hydrogen molecules in clathrate hydrates .
_ j. chem .
_ * 135 , * 054506 ( 2011 ) .
zaghloul , m. a. , celli , m. , salem , n. m. , elsheikh , s. m. & ulivi , l. high pressure synthesis and in situ raman spectroscopy of h@xmath4 and hd clathrate hydrates .
_ j. chem .
phys . _ * 137 , * 164320 ( 2012 ) .
vos , w. l. , finger , l. w. , hemley , r. j. & mao , h. k. novel h@xmath4+h@xmath4o clathrates at high pressures .
lett . _ * 71 , * 3150 - 3153 ( 1993 )
. strobel , t. a. , somayazulu , m. & hemley , r. j. phase behavior of h@xmath4+h@xmath4o at high pressures and low temperatures .
_ j. phys .
c _ * 115 , * 4898 - 4903 ( 2011 ) .
lokshin , k. a. & zhao , y. fast synthesis method and phase diagram of hydrogen clathrate hydrate .
_ * 2006 * , _ 88 _ , 131909 .
doi : 10.1063/1.2190273 efimchenko , v. s. , antonov , v. e. , barkalov , o. i. , beskrovnyy , a. i. , fedotov , v. k. & klyamkin , s. n. phase transitions and equilibrium hydrogen content of phases in the water hydrogen system at pressures to 1.8 kbar .
_ high press .
_ , * 26 * 439 - 443 , ( 2006 ) .
doi : 10.1080/08957950601105390 choukroun , m. & grasset , o. thermodynamic model for water and high - pressure ices up to 2.2 gpa and down to the metastable domain .
_ j. chem .
phys . _ * 127 , * 124506 ( 2007 ) .
smirnov , g. s. & stegailov , v. v. toward determination of the new hydrogen hydrate clathrate structures .
_ j. phys .
_ * 4 , * 3560 - 3564 ( 2013 ) .
qian , g. r. , lyakhov , a. o. , zhu , q. , oganov , a. r. & dong , x. novel hydrogen hydrate structures under pressure .
. reports _ * 4 , * 5606 ( 2014 ) .
giannasi , a. , celli , m. , grazzi , f. , ulivi , l. & zoppi , m. an apparatus for simultaneous thermodynamic and optical measurements , with large temperature excursions .
instrum . _ * 79 , * 013105 ( 2008 ) .
van kranendonk , j. _ solid hydrogen _
( plenum press , new york , 1983 ) .
grazzi , f. , santoro , m. , moraldi , m. , & ulivi , l. anisotropic interactions of hydrogen molecules from the pressure dependence of the rotational spectrum in the ar(h@xmath4)@xmath4 compound .
lett . _ * 87 , * 125506 ( 2001 ) .
grazzi , f. , santoro , m. , moraldi , m. & ulivi , l. roton excitations of the hydrogen molecule in the ar(h@xmath4)@xmath4 compound .
b _ * 66 , * 144303 ( 2002 ) . pimentel , g. c. & sederholm , c. h. correlation of infrared stretching frequencies and hydrogen bond distances in crystals .
_ j. chem .
* 24 , * 639 - 641 ( 1956 ) .
ouillon , r. , turc , c. , lemaistre , j. p. & ranson , p. high resolution raman spectroscopy in the @xmath10 and @xmath54 crystalline phases of n@xmath4 .
_ j. chem .
_ , * 93 , * 3005 ( 1990 ) .
doi : 10.1063/1.458887 . | the properties of some forms of water ice reserve still intriguing surprises .
we report here on the direct observation of a new form of ice , metastable at atmospheric pressure and having remarkable molecular adsorption ability .
we study the crystalline solid compound of water and molecular hydrogen , called filled ice and indicated with c@xmath0 , that is formed at about 400 mpa and can be recovered at room pressure and low temperature , still containing a large fraction of molecular hydrogen .
we perform x - rays diffraction to check the structure of the recovered sample . by means of raman spectroscopy
, we measure the hydrogen release at different temperatures , and succeed in rapidly removing all the hydrogen molecules , obtaining a new form of ice ( ice xvii ) which tolerate heating under vacuum up to about 120 k. of paramount interest is the fact that the emptied crystal can adsorb again hydrogen and release it repeatedly , showing a temperature dependent hysteresis .
we present here the first characterization of this new material , which can contain molecular hydrogen in an amount larger than clathrate - hydrates , and adsorbs and release it incredibly faster without changing its structure .
water molecules in the solid state may give rise to more than fifteen different forms of ices , depending on temperature and pressure @xcite .
in addition , when water freezes in the presence of some other molecular substance , it may form peculiar crystal structures , known as clathrate - hydrates , trapping the guest molecules inside cages of different geometries @xcite .
until recently it has been believed that caged guest molecules are essential for the stability of the crystal , so that the water skeleton would collapse without them , at least at positive pressure , while , at negative pressure , existing or hypothetical empty clathrate structures are actually believed to be the stable phase of solid water @xcite .
one experimental study has recently demonstrated , however , that at least one of these clathrate structures exists in its metastable state if preserved at low temperature , even in absence of guest molecules @xcite .
water and molecular hydrogen at high pressure form four known solid compounds .
the first in order of increasing pressure , stable at @xmath1 mpa and at 4 - 6 @xmath2c below zero , is a clathrate - hydrate @xcite having the so called cubic sii structure , common to other clathrates with different molecular guests @xcite . in this non - stoichiometric compound hydrogen molecules
are trapped in two types of cages , with a total hydrogen molar fraction ( @xmath3 mol(h@xmath4)/mol(h@xmath4o ) ) up to about 35 % .
the hydrogen molecules in the cages perform a peculiar quantum rattling and rotational motion , which has been studied efficiently with inelastic neutron scattering @xcite and raman scattering @xcite .
the latter technique is a very powerful and convenient one to identify the interactions of the molecules with the water environment , and to measure the composition of the sample . besides sii clathrates ,
the existence of two other stable phases for the h@xmath4-h@xmath4o solid mixture at higher pressure ( @xmath5 700 mpa ) is known since 1993 @xcite .
these two structures , indicated with c@xmath6 and c@xmath4 , do not possess the typical cage structure of clathrates , and are usually and more properly named filled ices .
recently , only two experimental studies @xcite have investigated the phase diagram of the h@xmath4-h@xmath4o compounds at intermediate pressures and have demonstrated the presence of a further stable phase , named c@xmath0 , at temperatures 100 - 270 k and pressures 360 - 700 mpa , that is intermediate between the stability region of the sii clathrate and c@xmath6 phase ( see fig . 1 ) . up to now , a definitive consensus on the structure of the c@xmath0 phase has not been reached . the structural model ( later referred to as c@xmath0-i ) originally proposed in ref .
@xcite ( space group @xmath7 or @xmath8 ) after ex - situ x - ray diffraction measurements at room pressure and 80 k , assumes the presence of water molecules in sites @xmath9 with 0.5 occupancy .
these water molecules would not present correct hydrogen bonds with the other molecules of the lattice .
strobel @xcite suggests two possible structures ( indicated as @xmath10-quartz and st@xmath11 ) after the analysis of x - ray diffraction collected from one sample pressurized in a diamond anvil cell .
these models were later examined theoretically by smirnov and stegailov @xcite , who considered also a variant of the c@xmath0-i structure ( indicated by c@xmath0-ii ) , and suggested c@xmath0-ii and st@xmath11 both as viable candidates . actually , the @xmath10-quartz structure , with the proposed parameters @xcite gives rise to atypical , very short , o - o distances .
after this , oganov and coworkers @xcite used dft - based structure searching to predict several phases of the h@xmath4-h@xmath4o mixture , but these searching methods are restricted in the number of atoms , so any large clathrate - sized unit cells can easily be missed . |
accreting black hole binaries show variability on a wide range of timescales from milli - seconds to months and years . while the long term variability typically arises due to the transient nature or the state transitions , the short term ( sub - second ) variability is generally attributed to the processes in the inner regions of accretion disk .
a perplexing feature of such a short term variability is the presence of quasi periodic oscillations , characterized by a narrow peak superimposed over a broad band noise in the power density spectrum ( pds ) .
qpos have been observed in a frequency range of less than hz to hundreds of hz @xcite and they are generally classified as low frequency qpos ( lfqpos , 0.1 - 30 hz ) and high frequency qpos ( hfqpos , 40 - 450 hz ) .
frequency of hfqpos does not vary with a sizeable change in luminosity suggesting that it depends on the fundamental system parameters such as black hole mass and spin @xcite .
it is evident that the lfqpos are not linked with the orbital motion of the material in the disk because the observed frequencies correspond to the far outer regions of the accretion disk .
the properties of these qpos ( e.g. low frequency break , centroid frequency , rms amplitude ) are strongly coupled with the changes in spectrum and luminosity ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , suggesting that these qpos originate in the same inner region of the accretion disk . despite the knowledge of empirical correlations of qpo properties with spectral parameters ( e.g. * ? ? ?
* ; * ? ? ?
* ) , the exact physics is not very well understood .
there are several models to explain the origin and properties of lfqpos . in one of the most promising models , the lfqpo originates as a result of relativistic lense - thirring ( lt ) precession of the hot inner flow @xcite .
the model successfully explains the variation of frequencies of qpo centroid and the low frequency break in terms of variation in the truncation radius of the inner disk .
another model based on the accretion ejection instability suggests that a qpo is produced by a spiral pattern rotating at a frequency of a few tenths of the keplerian frequency at the inner edge of the disk @xcite .
however , according to the shock oscillation model , wherein a shock is formed by a centrifugal barrier in a region of lower viscosity , lfqpos are originated due to oscillation of the post - shock region @xcite .
most of the available qpo models explain the frequencies and some of the observed correlations along with other timing and spectral properties .
however , in many cases , the models are silent about the energy dependence of the qpo or the exact mechanism of origin of the qpo .
observing a qpo as a direct oscillation of one or more spectral parameter could provide much deeper insights in the modulation mechanism and thus origin of the qpo . in this context , here we attempt to investigate the observed energy dependence of the lfqpo in grs 1915 + 105 in terms of modulation of the spectral parameters .
grs 1915 + 105 is one of the most enigmatic black hole binary system famous for its relativistic jets exhibiting superluminal motion @xcite and a variety of variability patterns @xcite .
its lfqpos and their correlations with various other parameters have been studied extensively ( * ? ? ?
? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
however , another intriguing aspect of this system is the spin of the black hole .
various reports have found spin of grs 1915 + 105 ranging from 0.56 @xcite to @xmath00.98 @xcite .
probably , the most accurate spin estimate for grs 1915 + 105 is available from the nustar observation ( * ? ? ?
* hereafter m13 ) , where the lower values of spin are ruled out at a high level of confidence based on the accurate modeling of relativistically blurred reflection .
this observation belongs to a special state of grs 1915 + 105 known as ` plateau ' state and it shows a qpo with frequency of 1.5 hz .
properties of this lfqpo have been studied in detail by ( * ? ? ?
* hereafter z15 ) where the authors show that the rms amplitude of the qpo increases with energy with a characteristic flattening between 10@xmath120 kev .
this suggests the possibility of an interplay between independently varying spectral parameters .
thus investigation of the spectral parameter(s ) , which lead to the observed energy dependence of the qpo power can provide significant clues on the origin of the qpo . in this context
, here we delve into the energy dependence of the qpo power by means of simulating light curves by varying different spectral parameter(s ) . in the next section ,
we discuss the nustar observation and its analysis to reproduce the results of m13 and z15 .
the results of these spectral and timing analysis are used to generate the simulated light curve as explained in the following section , with the discussion on our findings and preliminary conclusions in the subsequent sections . [ cols="<,^,^",options="header " , ]
the nustar observation of grs 1915 + 105 carried out on july 03 , 2012 has been used for the present study .
the same observation has been analyzed in detail by m13 and z15 for spectroscopic and timing studies respectively .
we extract the spectrum following m13 and adopt their best fit model , tbabs@xmath2(kerrconv@xmath3reflionx_hc + cutoffpl ) where the reflection of the primary cut - off power law is modeled by reflionx_hc ( a variant of reflionx model ) , which assumes the incident spectrum as cut - off power law . with the same spectral model ,
we ensure that the best fit results are almost identical , which are then used in our simulations as explained in the next section .
we also carried out the timing analysis on the line of z15 .
the light curves are extracted in an energy range of 3 - 80 kev with a bin size of 100 ms .
pds is generated for each interval of 1024 bins and co - added for all the intervals .
the pds from both fpma and fpmb show a very strong type - c qpo at @xmath01.5 hz .
the pds also shows strong spikes at every 1 hz , however , it is known that these are instrumental artifacts in the early nustar observations and hence we ignore these bins during fitting in xspec .
the light curves from fpma and fpmb are added together in order to improve the statistics and the pds is generated from the combined light curve ( fig .
[ pds - spec ] left panel ) .
we fit the pds using multiple lorentzians and a constant , which takes care of white noise level in the pds .
it is obvious that the pds has many complex features such as broad band noise , a small peak at frequency of @xmath00.75 hz and a strong qpo at @xmath01.5 hz .
here we concentrate only on the @xmath01.5 hz qpo feature because it is known to have systematic dependence on energy ( fig .
[ pds - spec ] right panel ) .
the small peak at @xmath00.75 hz has been described as a sub - harmonic by z15 .
however for the purpose of the present work , we consider it as a part of broad band noise mainly because of very low q value of @xmath03.6 .
z15 have reported that the rms amplitude of the qpo increases with energy .
we verify this from the light curves in different energy bands ( 3@xmath14 kev , 4@xmath15 kev , 5@xmath16 kev , 6@xmath17 kev , 7@xmath19 kev , 9@xmath112 kev , 12@xmath115 kev and 15@xmath130 kev ) .
we find an increasing trend with a shallow dip ( or flattening ) in 10@xmath120 kev as reported in z15 ( fig .
[ pds - spec ] right panel ) .
they have also reported that this observation shows increase in count rate as well as qpo frequency during the later part ( after @xmath040 ks of the start of the observation ) .
therefore , we restrict both the spectral and timing analysis to the first 40 ks of data , where the count rate and qpo frequency are quite stable .
effective exposure times for our analysis are 9.71 ks and 10.01 ks for fpma and fpmb respectively .
in many cases , the observed spectrum can be explained with different models with same statistical significance . in such cases , the compatibility of a spectral model with the observed timing characteristic can be helpful to solve the degenracy .
the main objective of the present work is to investigate the compatibility of the observed qpo properties with the spectral model .
one possibility to achieve this is by comparing the amplitude of the oscillations across multiple energy bands .
another possibility is by comparing the phase lags in a similar manner .
black hole binaries are known for showing phase lags between different energy bands .
phase lags for grs 1915 + 105 has also been reported by several authors ( e.g * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , wherein it is generally found that the hard photons lag the soft photons .
we also calculated phase lags between different energy bands using nustar data , however because of the relatively poor statistics ( compared to e.g. rxte - pca ) , the phase lags can not be constrained ( obtained values are consistent with zero with larger error bars ) with the present data .
hence we adopt the alternate method by comparing the observed energy dependence of the qpo rms amplitude with the simulated energy dependence .
an important feature of the present best fit spectral model is that it consists of only one primary continuum component , cut - off power law , covering the full energy range .
hence , it does not allow interpretation of the qpo in terms of oscillation of one of the spectral component as suggested by @xcite ( also corroborated by * ? ? ?
* ; * ? ? ?
thus observed energy dependence of the qpo power must result from the variation of one or more spectral parameters . here
we assume that a particular spectral parameter varies with a frequency close to the observed qpo frequency .
we consider four possible parameters of the spectral model , viz . the spectral index , normalization of the cut - off power law , normalization of the reflection component and the ionization parameter ; which can vary with the phases of the qpo .
though the high energy cut - off can vary , we keep it fixed at the best fit value because its variation does not affect the model in the energy range of interest .
all parameters of kerrconv are assumed not to vary with the phases of the qpo .
we then simulate the light curve with a bin size of 100 ms .
initially we simulate the light curve for variation of only one parameter . as a zeroth order approximation
, we assume that the parameter is varying sinusoidally with qpo phases in a range centered at the best fit value .
the assumption of sinusoidal oscillation is clearly an over simplification and the actual oscillation profile could be more complex .
however , the energy dependence of rms amplitude is not likely to have strong dependence on the exact shape of oscillations .
we also include a small random variation over the sinusoidal variation to be a bit more realistic .
later we also simulate the light curve for multiple parameters varying simultaneously . in either case the basic algorithm for simulation of light curve is as follows- it is assumed that a parameter @xmath4 obtained from the best - fit spectral model varies at a frequency close to the observed qpo frequency ( @xmath5 ) .
the amplitudes of the sinusoidal variation and random fluctuation are @xmath6 and @xmath7 respectively such that @xmath8 = r @xmath6 .
the simulation is carried out for r=0.1 , 0.2 and 0.3 and it is found that the results remain unaffected .
therefore , we present results hereafter for r=0.1 ( i.e. 10% random variations superimposed over the sinusoidal oscillations ) .
* let @xmath9 be the best fit value of parameter @xmath4 . *
the spectral model is evaluated in xspec with the parameter @xmath4 assuming values in the range @xmath10 , keeping the rest of the parameters at their best fit values .
model counts , including the effects of detector response matrix as well as effective area , are obtained in each of the energy bins and are tabulated . *
the light curve is simulated with a bin size of 100 ms for a duration of @xmath0 10 ks to match the exposure time of the observation . * for each interval ( @xmath0 2 s ) a random frequency of oscillation ( @xmath11 ) is drawn from a lorentzian distribution with the mean @xmath5 ( = 1.5 hz ) and the fwhm ( = 0.089 hz ) values same as those for the observed qpo .
the frequency remains the same throughout an interval .
we have also verified that the overall results remain unaffected if the frequency is changed more often within one interval .
it is assumed here that the quasi periodic nature arises due to frequency modulation ( see * ? ? ?
* ) . * the phase of oscillation for each time bin @xmath12
is obtained by @xmath13 where @xmath14 is the bin size and @xmath15 is the phase for previous time bin .
continuity of the phases across intervals is also maintained in the same manner . *
the value assumed by parameter @xmath4 for the @xmath12-th timebin is calculated as @xmath16 where @xmath17 is a standard normal variate ( see fig [ simresult-3par ] , left panel ) .
* mean counts in the time bin @xmath12 for a given energy bin @xmath18 ( @xmath19 ) corresponding to value @xmath20 is obtained from the table generated in the first step .
the energy bin @xmath18 is kept same as that used to extract the observed light curves . *
random number is generated from poisson distribution with mean @xmath19 and is recorded as counts in the @xmath12-th time bin of the light curve in energy bin @xmath18 .
* power density spectra are generated for these simulated light curves using powspec and are fitted with xspec as usual .
fig ( [ simlc ] ) shows a section of the simulated light curve and its pds in an energy range of 9@xmath112 kev .
it should be noted that we have not included low frequency noise in our simulations because our primary objective here is to investigate the energy dependence of the qpo power .
fig ( [ simresult-1par ] ) summarizes the overall results of our simulations .
the four panels show the simulated energy dependence of the qpo power for variation of the four spectral parameters .
the observed energy dependence is shown with black solid line in all the panels .
for each parameter , multiple lines are shown corresponding to the three different range of variation of parameters .
it can be seen that only the variation of spectral index can reproduce the overall trend of the increasing qpo power with energy .
variation of other three parameters can not reproduce the observed trend . even in the case of the spectral index variation , the exact shape of the energy dependence curve ,
particularly the dip or flattening in the energy range of 10@xmath120 kev is not reproduced .
this suggests that , while the variation of spectral index during qpo phases is essential it is not the only source of the variability .
hence we test the hypothesis by simulationg the light curves considering the variation of more than one parameter simultaneously .
the simulation was repeated for the variation of two and three parameters and the results are shown in fig ( [ simresult-2par ] ) and ( [ simresult-3par ] ) respectively . since the reflection component represents the comptonization component reflected from the disk , it is expected that the former would be stronger for a harder spectrum . or in other words , the reflection component should be anti - correlated with the spectral index .
therefore , the parameters of reflection component ( @xmath21 and @xmath22 ) were varied out of phase with spectral index in the simulation of light curves .
panel a of fig ( [ simresult-2par ] ) shows the energy dependence obtained from the variation of reflection normalization ( @xmath21 ) and spectral index and panel b shows the results for ionization parameter ( @xmath22 ) and spectral index .
a comparison of panel a of fig ( [ simresult-1par ] ) and fig ( [ simresult-2par ] ) shows that the variation of spectral index and normalization of reflection reproduce the observed energy dependence more closely than the variation of spectral index alone .
we also studied the energy dependence with the light curves simulated by varying 1 ) normalization of cut - off power law and normalization of reflection component and 2 ) normalization of cut - off power law and ionization parameter .
the variation of the two set of parameters does not reproduce the observed increasing trend of qpo power .
we have further attempted to reproduce the observed pattern by varying three parameters simultaneously .
the left panel of fig ( [ simresult-3par ] ) shows the variation of spectral parameters with time .
representative results obtained from the variation of three parameters are shown in the right panel of fig ( [ simresult-3par ] ) .
it can be seen that the results start resembling the observed pattern more closely .
we avoid attempts to exactly reproduce the observed pattern as this will require detailed fine tuning of the specifics such as number of varying parameters , amplitudes of variation , phase lags etc .
more importantly , a relook into the spectral model itself might also be necessary which is beyond the scope of this paper .
the main objective here is to convey that the present best fit spectral model is compatible with the observed trend of the increasing qpo power with energy when the spectral index is assumed to be varying .
however , it might be necessary to look beyond this model in order to reproduce the exact energy dependence of the qpo power .
increasing trend of the lfqpo power with energy is well known ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the trend can also be seen in the results shown by @xcite .
however , origin of this trend in terms of variation of the spectral components has not been investigated so far . here
we attempt to explain the energy dependence of lfqpo observed during the first nustar observation of grs 1915 + 105 .
we simulate the light curve for variation of various spectral parameters and compare the resultant energy dependence with the observed one .
we find that only variation of the spectral index reproduces the overall trend of the increasing qpo power with energy .
variation of any other parameter does not reproduce the observed trend .
this has significant implications on the feasibility of various models proposed to explain the origin of the qpo .
for example , the fact that the variation of reflection normalization does not reproduce the observed trend suggests that the qpo models based on geometric modulation @xcite may not be realistic . among the other existing models for the low frequency qpos , the model based on lense - thirring precession is considered the most successful model .
it successfully explains the overall shape of the pds and the variation of qpo frequency with the changes in spectral hardness , which is further linked with the truncation radius of the inner disk @xcite .
however , the energy dependence of qpo is not apparently clear .
the similar situation exists with a magnetohydrodynamic model for qpo based on the accretion ejection instability @xcite .
the shock oscillation model @xcite does produce the spectral variation and it could be compatible with the energy dependent qpo power .
however , the simulation results obtained from the present spectral model can not be directly compared with the predictions of shock oscillation model .
it might be necessary to perform similar simulation with appropriate model and then verify the results .
another observation is the flattening of qpo spectrum at 10 - 20 kev which can not be reproduced with variation of only spectral index .
the flattening seems to be more common as it has been observed at many other occasions ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
similar feature is also found in few other sources @xcite .
we have verified that the same feature is found in the rxte observations of grs 1915 + 105 during the ` plateau ' states .
we fit the spectra during these observations with the present spectral model consisting of cut - off power law and relativistically blurred reflection .
the best fit spectral parameters are similar to that found for the present nustar spectrum .
we performed similar simulations using rxte spectra and could reproduce the overall trend of increasing power of qpo .
however , the flattening is not found in the simulation results .
the observed flattening might be an indicator of the presence of more than one continuum component as suggested by @xcite .
overall our simulation results can put significant constraints on the modulation mechanism assumed in various present theoretical models of qpo .
recently , @xcite have made an attempt to understand qpos with frequency resolved spectroscopy .
we suggest that the qpo phase - resolved spectroscopy , similar to that shown by @xcite can be another promising way to validate the spectral models , but with full energy resolution spectra .
recently launched indian astronomy mission astrosat @xcite , with large effective area of the laxpc instrument , will provide ample opportunities for such qpo phase - resolved spectroscopy owing to its event mode data over broad energy range .
this research has made use of data obtained from high energy astrophysics science archive research center ( heasarc ) , provided by nasa s goddard space flight center .
research at physical research laboratory is supported by department of space , govt . of india .
we thank a. c. fabian and m. parker for providing reflionx_hc model .
we acknowledge the anonymous referee for his / her valuable comments . , k. p. , tandon , s. n. , agrawal , p. c. , et al .
2014 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol . 9144 , society of photo - optical instrumentation engineers ( spie ) conference series , 1 | the low frequency quasi periodic oscillations ( qpos ) are commonly observed during hard states of black hole binaries .
several studies have established various observational / empirical correlations between spectral parameters and qpo properties , indicating a close link between the two . however , the exact mechanism of generation of qpo is not yet well understood . in this paper , we present our attempts to comprehend the connection between the spectral components and the low frequency qpo observed in grs 1915 + 105 using the data from nustar .
detailed spectral modeling as well as the presence of the low frequency qpo and its energy dependence during this observation have been reported by @xcite and @xcite respectively .
we investigate the compatibility of the spectral model and energy dependence of the qpo by simulating light curves in various energy bands for small variation of the spectral parameters .
the basic concept here is to establish connection , if any , between the qpo and the variation of either a spectral component or a specific parameter , which in turn can shed some light on the origin of the qpo .
we begin with the best fit spectral model of @xcite and simulate the light curve by varying the spectral parameter at frequencies close to the observed qpo frequency in order to generate the simulated qpo .
further we simulate similar light curves in various energy bands in order to reproduce the observed energy dependence of rms amplitude of the qpo .
we find that the observed trend of increasing rms amplitude with energy can be reproduced qualitatively if the spectral index is assumed to be varying with the phases of the qpo .
variation of any other spectral parameter does not reproduce the observed energy dependence . |
certain operations permitted in classical physics can not be done perfectly in quantum mechanics .
this is best exemplified by the celebrated no - cloning theorem @xcite which forbids to create two perfect copies of an unknown quantum state .
the no - cloning theorem is closely related to another no - go theorem stating that one can not gain some information on an unknown quantum state without disturbing it .
namely , if this would be possible one would be able to prepare two approximate replicas of this state which would be better than the best ones allowed by quantum mechanics @xcite .
therefore , in quantum mechanics any operation on an unknown quantum state giving some information on the state inevitably disturbs the state and in addition , the more information it extracts the larger is the disturbance .
this fundamental property of quantum operations is reflected in the plane of values of quantities quantifying the information gain and the state disturbance by a certain optimal trade - off curve that can not be overcome by any quantum operation . among all quantum operations
particularly interesting are those which lie on this curve since they to the best possible extent as quantum mechanics allows approximate ideal disturbance - free measurement device .
these operations , conventionally denoted as minimal disturbance measurements ( mdms ) , in general depend on the set of input states , their a priori distribution and also on the quantities quantifying the information gain and the state disturbance @xcite .
the most successful approach to the finding of the optimal trade - offs and corresponding mdms proved to be that based on the quantification of the information gain by the mean estimation fidelity @xmath2 and the state disturbance by the mean output fidelity @xmath3 @xcite . using this approach it was possible to derive analytically the optimal trade - offs between @xmath2 and @xmath3 and to find the mdms for a single copy of a completely unknown pure state of a @xmath1-level system @xcite , a completely unknown pure state of a @xmath1-level system produced by @xmath1 independent phase shifts of some reference state @xcite and a completely unknown maximally entangled state of two @xmath1-level systems @xcite . besides , the first two above mentioned mdms were also demonstrated experimentally for @xmath4 ( qubit ) @xcite .
the studies on mdm were not restricted to finite - dimensional quantum systems and also mdm for a completely unknown coherent state was found and realized experimentally in @xcite .
multi - copy mdms were introduced by banaszek and devetak who considered partial measurement on @xmath0 identical copies of a pure qubit state .
they assumed mdms which output @xmath0 disturbed quantum copies of the state and a classical estimate and they _ numerically _ found the optimal fidelity trade - off for this scenario @xcite .
the mdms are not only of fundamental importance but they can be also applied to increase transmission fidelity of certain lossy and noisy channels @xcite . in this paper we further investigate the minimum disturbance measurement on several copies of the state .
in contrast to ref . @xcite we assume operations which output only a _ single _ quantum copy of the input state .
we derive _ analytically _ the optimal trade - off between the mean estimation fidelity @xmath2 and the mean output fidelity @xmath3 for an ensemble of @xmath0 identical pure qubits which is given by the formula @xmath5 moreover , we also conjecture that the optimal fidelity trade - off for an ensemble of @xmath0 identical pure states of a @xmath1-level system has the form @xmath6 the paper is organized as follows . the general formalism allowing to determine the mdm
is presented in sec .
[ sec_1 ] . in sec .
[ sec_qubits ] we find the optimal fidelity trade - off and the corresponding mdm for @xmath0 identical qubits . in sec .
[ sec_2 ] we present a conjecture of the optimal fidelity trade - off for @xmath0 identical @xmath1-level systems .
finally , sec .
[ sec_3 ] contains conclusions .
let us investigate a general mdm for @xmath0 identical pure states of a @xmath1-level system ( qudit ) .
such states are represented by vectors in a @xmath1-dimensional hilbert space @xmath7 with an orthonormal basis @xmath8 .
the qudits form an orbit of the group @xmath9 of @xmath10 unitary matrices with determinant @xmath11 , @xmath12 , where @xmath13 , @xmath14 , is a unitary representation of @xmath9 on @xmath7 .
we consider here quantum operations on @xmath0 identical qudits @xmath15 - and also yields a classical estimate of @xmath16 . without loss of generality , these estimates can be labeled by the elements of the group @xmath9 . note that the input hilbert space of the operation is the symmetric subspace of the hilbert space of @xmath0 qudits , @xmath17 and the output hilbert space is the space of a single qudit , @xmath18 .
our task is to find an operation which exhibits the best possible performance in the following protocol @xcite . in each run of the protocol
, the operation is applied on the quantum state ( [ nquditstate ] ) .
we assume that @xmath19 is chosen randomly with uniform a priori distribution from the set of states @xmath20 .
if the outcome @xmath21 is detected the operation produces a single qudit output state @xmath22 .
this state is not normalized and its trace @xmath23 $ ] is the probability density of obtaining the outcome @xmath24 on the state ( [ nquditstate ] ) .
the information on the state @xmath16 contained in the measurement outcome @xmath24 is converted into a guess of the state which is in our case a single qudit state @xmath25 .
the performance of this procedure can be quantified by two mean fidelities : mean output fidelity @xmath3 defined as @xmath26 which quatifies the average state disturbance and the mean estimation fidelity @xmath2 defined by the formula @xmath27 which quantifies the average information gain . here
, the integrals are taken over the whole group @xmath9 and @xmath28 is the normalized invariant haar measure on the group .
quantum mechanics sets a fundamental bound on the maximum value of the fidelity @xmath3 that can be attained for a given value of the fidelity @xmath2 for any considered quantum operation .
the bound can be expressed in the form of a nontrivial optimal trade - off relation between @xmath3 and @xmath2 and the mdm is defined as a quantum operation for which the fidelities @xmath2 and @xmath3 satisfy the trade - off .
two extreme cases of the trade - off are well - known .
first , if @xmath2 is the optimal estimation fidelity of the qudit state from @xmath0 identical copies , i.e. @xmath29 @xcite , then @xmath3 can be at most equal to @xmath30 .
second , if @xmath31 then @xmath2 can not be larger than the optimal estimation fidelity of a qudit from @xmath32 identical copies , i.e. @xmath33 . to find the whole optimal trade - off we can use the method developed in @xcite . with the help of jamiolkowski - choi isomorphism
@xcite we can represent the completely positive map corresponding to each particular outcome @xmath24 by a positive - semidefinite operator @xmath34 acting on the tensor product of the input and output hilbert spaces @xmath35 .
it holds that @xmath36 $ ] where @xmath37 .
as shown in @xcite the optimal partial measurement can be assumed to be covariant which means that @xmath34 are generated from a single properly normalized operator @xmath38 , @xmath39 \,\chi_n^{(d ) } \,[u_d^{t \otimes n}(h ) \otimes u_d^\dagger(h ) ] .
\label{chicovariant}\ ] ] the overall operation must be trace - preserving which imposes the constraint , @xmath40 d h = \openone_{\mathrm{in } } , \label{chinorm}\ ] ] where @xmath41 stands for the partial trace over the output single - qudit hilbert space and @xmath42 denotes the identity operator on the input space @xmath43 .
the formula ( [ chinorm ] ) expresses the completeness of the measurement carried on the input state .
the unitary representation @xmath44 of @xmath9 acts irreducibly on @xmath43 . for the covariant map ( [ chicovariant ] )
the integral in eq .
( [ chinorm ] ) can thus be evaluated with the help of schur s lemma and we get @xmath45\openone_{\mathrm{in}}$ ] where @xmath46 is the dimension of the symmetric hilbert space @xmath43 .
the trace - preservation condition ( [ chinorm ] ) thus boils down to the proper normalization of the map that should read @xmath47=d(n , d)$ ] .
the operator @xmath38 generating the optimal partial measurement is proportional to a rank - one projector and can be written as @xmath48 where @xmath49 is the eigenvector of a positive - semidefinite operator @xmath50\ ] ] corresponding to its maximum eigenvalue @xcite . here
\psi(g ) dg,\label{rf}\\ r^{(d)}_{g}&=&\mathrm{tr}_{\rm out}[r^{(d)}_f \openone_{\rm in } \otimes \psi(0)]\otimes\openone_{\rm out}\label{rg}.\end{aligned}\ ] ] using the map @xmath38 the fidelities @xmath3 and @xmath2 can be expressed as @xmath52 , \qquad g=\mathrm{tr}[\chi^{(d)}_{n}r^{(d)}_{g}].\ ] ] the operator @xmath53 can be easily evaluated using schur s lemma and after some algebra we arrive at @xcite @xmath54 where @xmath55 stands for the partial transposition with respect to the first @xmath0 qudits and @xmath56 is the projector onto the subspace @xmath57 . in what follows it
is convenient to work with the occupation number basis @xmath58 that forms an orthonormal basis in the subspace @xmath59 . here
@xmath60 is the symmetrization operator for @xmath0 qudits , the symbol @xmath61 stands for summation over all @xmath62 permutations of @xmath0 qudits and @xmath63 denotes the permutation operator of @xmath0 qudits ; the integers @xmath64 , @xmath65 , @xmath66 are the numbers of qudits in the states @xmath67 , @xmath66 that satisfy the constraint @xmath68 . making use of the occupation number basis the operator
@xmath56 can be expressed as @xmath69 to find the desired mdm for @xmath0 identical qudits we have to diagonalize a large matrix @xmath70 .
for a general @xmath1 this is a complex task which can be solved numerically .
however , if we resort to the qubit case ( @xmath4 ) we can find the optimal fidelity trade - off and the mdm analytically .
the obtained result then can be used to make at least a conjecture about the optimal fidelity trade - off for @xmath0 qudits .
for qubits the operator ( [ pi ] ) reads as @xmath71 where @xmath72 is a completely symmetric state of @xmath0 qubits in which @xmath73 qubits are in the basis state @xmath74 and the remaining @xmath75 qubits are in the basis state @xmath76 .
hence , making use of the formula @xmath77 and eq .
( [ rfd ] ) one finds that @xmath78.\end{aligned}\ ] ] further , substitution of the obtained expression into eq .
( [ rg ] ) gives the operator @xmath79 in the form @xmath80 in order to determine the optimal @xmath81 we have to find the maximum eigenvalue and the corresponding eigenvector of the matrix @xmath82 .
the matrix @xmath82 has a block diagonal structure with two one - dimensional blocks and @xmath0 two - dimensional blocks .
the elements of the one - dimensional blocks are the eigenvalues @xmath83 and @xmath84 with the characteristic subspaces spanned by the basis vectors @xmath85 and @xmath86 , respectively .
the two - dimensional blocks correspond to the invariant subspaces spanned by the basis vectors @xmath87 , @xmath88 and have the form @xmath89 , where @xmath90 the matrix @xmath91 possesses two eigenvalues @xmath92 from which one obtains the remaining eigenvalues of the matrix @xmath93 as @xmath94 .
the larger eigenvalue @xmath95 is a decreasing function of @xmath73 attaining maximum for @xmath96 .
obviously , @xmath97 . moreover , using in eq . (
[ lambda ] ) the inequality @xmath98 following from the inequality @xmath99 one can show that also @xmath100 holds and therefore @xmath101 is the maximum eigenvalue of the matrix @xmath93 .
the eigenvalue is non - degenerate and its eigenvector determining the optimal map @xmath81 reads as @xmath102 where @xmath103 and @xmath104 are nonnegative real numbers satisfying the condition @xmath105 . on inserting @xmath106 into eqs .
( [ fg ] ) one arrives after some algebra at the optimal fidelities @xmath107,\label{qubitf}\\ g&=&\frac{1}{n+2}\left(n+\alpha^{2}\right)\label{qubitg}.\end{aligned}\ ] ] expressing now the parameters @xmath108 using eq .
( [ qubitg ] ) and the normalization condition @xmath105 and substituting the obtained formulas into eq .
( [ qubitf ] ) we finally obtain the optimal fidelity trade - off for @xmath0 identical qubits ( [ nqubit ] ) .
the trade - off is depicted for several numbers of copies @xmath0 in fig .
[ fig1 ] .
the specific feature of the optimal map ( [ nqubitmap1 ] ) is that it can be rewritten as the following coherent superposition of two maps : @xmath109 where @xmath110 , @xmath111 and @xmath112 . the first map in the superposition is described by the vector @xmath113 and corresponds to the choice @xmath114 ( @xmath115 ) .
since in this case @xmath116 the map apparently realizes optimal estimation of a qubit from @xmath0 identical copies @xcite .
the second map is obtained by choosing @xmath117 ( @xmath118 ) and it is represented by the second vector on the right hand side of eq .
( [ nqubitmap2 ] ) .
it gives @xmath31 and @xmath119 which corresponds to optimal estimation of a qubit from @xmath32 copies while one copy is left intact by the map .
interestingly , the fact that one can create a mdm as a coherent superposition of the two extreme maps is a general property of the mdms that is valid not only for the present case of @xmath0 qubits , but it holds also for a single phase covariant qudit @xcite , two maximally entangled qudits @xcite or a single completely unknown qudit when @xmath120 is the maximally entangled state of two qudits and @xmath121 satisfy the condition @xmath122 .
thus although we are not able to solve analytically the above optimization task of finding the mdm for @xmath0 qudits for a general @xmath1 we can use the superposition principle together with eqs .
( [ nqubitmap2 ] ) and ( [ quditmap ] ) to guess the desired optimal map to read @xmath123 where @xmath124 and @xmath125 in order to facilitate the following calculations we rearrange the terms on the right hand side of the map ( [ nquditmap1 ] ) and rewrite it in the form : @xmath126 where @xmath127 fulfill the condition @xmath105 and where we have used the short hand notation @xmath128 for a completely symmetric state of @xmath0 qudits containing @xmath32 qudits in the basis state @xmath76 and a single qudit in the basis state @xmath129 .
the fidelities @xmath3 and @xmath2 for this map can be again calculated with the help of eq .
( [ fg ] ) . substituting eq .
( [ nquditmap2 ] ) into eq .
( [ fg ] ) and taking into account the symmetry of the projector @xmath56 that implies @xmath130 the problem of finding @xmath3 and @xmath2 reduces to the calculation of the following scalar products @xmath131 where we have used the short hand notation @xmath132 for a completely symmetric state of @xmath133 qudits containing @xmath32 qudits in the basis state @xmath76 , a single qudit in the state @xmath134 and a single qudit in the state @xmath129 .
the scalar products can be easily evaluated using eq .
( [ occupation ] ) as @xmath135 , @xmath136 and @xmath137 .
hence , one obtains @xmath138,\nonumber\\ g&=&\frac{1}{n+d}\left(n+\alpha^{2}\right).\end{aligned}\ ] ] eliminating now the parameters @xmath103 and @xmath104 from these equations using the same procedure as in the qubit case we arrive finally at the fidelity trade - off ( [ nqudit ] ) .
although the found trade - off was not shown to be optimal here , there are several indications supporting our conjecture that it is really optimal .
first , for @xmath139 we obtain @xmath140 using eqs .
( [ quditfg ] ) and therefore these optimal fidelities satisfy our trade - off .
second , by putting @xmath141 and @xmath142 one finds that @xmath31 and @xmath143 which means that also the second extreme case is fulfilled .
finally , for @xmath4 the trade - off reduces to the optimal trade - off for @xmath0 identical qubits ( [ nqubit ] ) while for @xmath144 it boils down to the optimal trade - off for a single completely unknown qudit @xcite .
the trade - off ( [ nqudit ] ) is depicted in fig .
[ fig2 ] for @xmath145 and @xmath146 .
in summary , in the present paper we have derived analytically the optimal trade - off between the mean estimation fidelity and the mean output fidelity for partial measurements on @xmath0 identical pure qubits .
furthermore , based on the structure of the optimal map saturating the trade - off we have made a conjecture about the optimal fidelity trade - off for partial measurements on @xmath0 identical pure qudits .
the obtained results provide an insight into the generic structure and properties of mdms .
the optimal partial measurements saturate the fundamental bound on conversion of quantum information onto classical information and may thus find applications in quantum communication and information processing .
the research has been supported by the research projects `` measurement and information in optics , '' ( msm 6198959213 ) and center of modern optics ( lc06007 ) of the czech ministry of education .
partial support by the secoqc ( ist-2002 - 506813 ) project of the sixth framework program of eu is also acknowledged . | we derive analytical formula for the optimal trade - off between the mean estimation fidelity and the mean fidelity of the qubit state after a partial measurement on @xmath0 identically prepared qubits .
we also conjecture analytical expression for the optimal fidelity trade - off in case of a partial measurement on @xmath0 identical copies of a @xmath1-level system . |
the optical dipole trap ( odt ) is created by the focus of a single far red - detuned gaussian laser beam . for a given laser power @xmath55
, the potential along the propagation direction of the light , which we denote the longitudinal axis , can be written as : @xmath56 here , @xmath57 is the initial depth of the optical potential at a laser power of @xmath58 , @xmath59 is the trap depth parameter and @xmath60 is the rayleigh range of the optical trapping beam with minimal waist @xmath61 and wavelength @xmath62 . by changing the laser power @xmath55 , we can adjust the trap depth during our experiments . in a harmonic approximation
the trap frequencies of the dipole trap are @xmath63 along the longitudinal direction and @xmath64 in the radial direction , resulting in an aspect ratio of @xmath65 . to allow atoms to tunnel out of the trap
, we create a potential barrier by superimposing the optical dipole trap with a linear magnetic gradient along the longitudinal axis [ see fig .
1(c ) of main text ] .
this adds the magnetic part @xmath66 to the potential , where @xmath67 is the magnetic moment of the atoms and @xmath68 is the strength of the magnetic field gradient . for detailed information on the determination of the potential parameters ,
see ref .
in several parts of our experimental sequence , we tilt the potential of the optical dipole trap ( odt ) to spill atoms from the trap [ fig .
1(c ) of the main text ] .
to do this , we apply a magnetic field gradient of @xmath68 , which adds a linear potential along the longitudinal axis of the odt ( section i ) .
the strength of this linear magnetic potential depends on the magnetic moment of the atoms in the trap , which is shown in figure [ fig - mu](a ) as a function of the magnetic offset field . by tilting the odt
, we create a potential barrier that separates the in - trap states from the continuum of states outside of the trap .
we reduce the height of the potential barrier by lowering the optical power in the odt ( @xmath69 ) .
atoms now leave the trap with rates that strongly depend on their energy .
this allows us to remove atoms with energies above or marginally below the potential barrier from the trap while keeping atoms with significantly lower energies . after a certain time we ramp the optical power in the odt back up to its initial value and thereby
stop the spilling process .
when performing this spilling technique in the paschen - back regime , the magnetic moments of spin - up and spin - down atoms are approximately equal [ fig .
[ fig - mu](b ) ] .
therefore , the potential has the same shape for the spin - up and spin - down atoms .
all tunneling measurements of the main text are performed deep in the paschen - back regime ( @xmath70 ) , where the magnetic moments of spin - up and spin - down atoms differ by less than @xmath71 .
we use the spilling technique to let exactly one atom ( for the ( 2,1 ) and the ( 3,1 ) system ) or two atoms ( for the ( 2,2 ) system ) tunnel out of the trap ( section vii ) .
to prepare the initial few - particle systems , we start with a degenerate gas of about 600 noninteracting @xmath8 atoms in their two lowest hyperfine states in the odt . at a magnetic offset field of @xmath72 ,
where the atomic sample is noninteracting and the magnetic moments of spin - up and spin - down atoms are approximately equal , we spill all atoms above a certain trap level from the trap ( section ii ) .
this allows us to prepare spin - balanced ground - state systems of up to 10 atoms with a fidelity of @xmath73 [ ] . to prepare spin - imbalanced systems
, we first prepare a spin - balanced system and then perform a second spilling process at a magnetic offset field of @xmath74 . at this field ,
the magnetic moment of the spin - up atoms ( @xmath75 ) is negligible [ see inset of figure [ fig - mu](a ) ] , which allows us to selectively spill spin - down atoms ( @xmath76 ) from the trap [ ] . in a single experimental run , we can either measure the total atom number @xmath77 or the number of spin - up atoms @xmath0 in the odt . to measure @xmath77 , we release the atoms from the odt , recapture them in a magneto - optical trap and measure their fluorescence signal . to measure @xmath0
, we selectively spill all spin - down atoms from the trap and then measure the number of remaining atoms .
a more detailed description of the preparation of both balanced and imbalanced few - fermion systems and of the detection of atom numbers in our experiment can be found in ref .
in a gas of ultracold @xmath8 atoms , the interactions between the atoms are well described by contact s - wave scattering . in a one - dimensional system
, the interaction potential between two atoms in different hyperfine states can be written as : @xmath78 where @xmath79 and @xmath80 are the positions of the two atoms , @xmath11 is the 1d coupling constant , and @xmath81 is the delta function [ ] .
contact interactions between identical fermions are forbidden .
starting from the 3d s - wave scattering length @xmath12 [ ] , we calculate @xmath11 as : @xmath82 where @xmath16 is the mass of a @xmath8 atom , @xmath83 is the harmonic oscillator length of the radial confinement and @xmath84 with the riemann zeta function @xmath85 [ ] .
figure [ fig - g_1d ] shows the calculated 1d coupling constant @xmath11 between two atoms in the two lowest hyperfine states of @xmath8 , which are trapped in our optical dipole trap with trap depth parameter @xmath86 .
all values of @xmath11 are given in units of @xmath87 , where @xmath88 is the harmonic oscillator length in longitudinal direction .
in the tilted potential used for the tunneling measurements [ fig .
1(c ) of the main text ] , we calculate @xmath15 by expanding the combined potential @xmath89 ( eq . [ dipole_trap ] and eq .
[ magnetic_gradient ] ) around its minimum .
to measure the mean occupation numbers on single - particle trap levels , we perform a series of measurements where we spill all population above a certain trap level @xmath90 from the trap ( section ii ) and count the number of remaining atoms ( section iii ) .
the mean number @xmath91 of remaining atoms corresponds to the sum of the populations on trap levels 0 to @xmath90 . for the ground state , the mean occupancy
is directly given by @xmath92 .
for excited states , we obtain the mean occupancies by subtracting @xmath93 from @xmath91 . when measuring the occupation - number distribution of a single spin - down atom in the trap ( fig . 4 of the main text )
, we correct for the finite fidelity of our experiment . during each experimental run
, we measure the probability @xmath94 to detect zero instead of one spin - down atom in a trap that still contains many trap levels . in an ideal experiment
, we would expect this probability to be zero .
due to the finite fidelity of both the preparation of the initial state and the detection of the atom number , we measure values of @xmath95 in all experimental runs .
we divide each measured mean atom number @xmath91 by @xmath96 to correct the data for this finite experimental fidelity .
figure [ fig - occupation_correction ] shows the effect of this correction for the data of fig . 4 of the main text .
our tunneling model predicts that above the cir tunneling into the fm final states becomes predominant . to confirm this prediction
, we show that both the initial @xmath5 and @xmath6 systems end up in a fm @xmath97-particle state after the tunneling of one particle deep in the super - tonks regime . for the initial @xmath6 [ @xmath5 ] system ,
the tunneling process is performed at an interactions strength of @xmath99 [ @xmath100 and a trap depth parameter @xmath101 [ @xmath102 .
we ramp the final @xmath97-particle system back through the fermionization regime to zero interaction strength .
then , we measure the occupation number distributions on the few lowest single - particle trap levels ( section v ) as shown in fig .
[ fig - fm_state ] . within our experimental fidelity ,
both distributions coincide with the expectation for a fm state .
note that in this measurement , the trap for the @xmath6 system was significantly deeper than in the measurements for fig .
3(b ) of the main text ( see section vii ) .
this could explain the deviation between the data of fig .
3(b ) of the main text and the expected limiting value of @xmath103 in the super - tonks regime .
the spatial wave function of a system of fermions in a ferromagnetic state is completely antisymmetric , which prohibits any two particles to be at the same place .
hence , a system of ultracold fermionic atoms with contact interactions in a fm state has to be noninteracting .
we confirm this for the final two - particle state after letting one atom tunnel from the @xmath5 system above the cir by measuring that independent of the interaction strength always one atom occupies the ground state of the trap ( fig .
[ fig - gs_occupation ] ) .
to measure the probability of spin - down tunneling ( fig . 3 of the main text ) , we let a specific number of atoms ( one atom in the @xmath5 and the @xmath6 system , two atoms in the @xmath17 system ) tunnel from the trap and measure the probability of spin - down tunneling . to do this ,
we superimpose the optical dipole trap ( eq . [ dipole_trap ] ) with a magnetic field gradient ( eq . [ magnetic_gradient ] ) to tilt the trap .
thereby we create a potential barrier between the in - trap states and the continuum of states outside the trap . to initiate the tunneling process
, we ramp down the intensity of the trapping light within a time @xmath104 and thereby lower the trap depth parameter @xmath105 ( eq . [ dipole_trap ] ) from its initial value @xmath106 to a final value @xmath107 . after a time @xmath108 , during which atoms can tunnel out of the trap , we ramp the power back to its original value to stop the tunneling process . in table
[ fig - spin_down_table]a , b , and c the parameters for the tunneling measurements on the @xmath5 , the @xmath6 , and the @xmath17 system are listed . for the @xmath5 and the @xmath6 systems ,
the parameters lead to the tunneling of exactly one atom , while for the @xmath17 system two atoms tunnel out of the trap . throughout the experimental runs ,
we perform control measurements in which we detect the number of atoms ( @xmath109 ) directly after the tunneling process ( see fig .
[ fig - spin_down_raw](a ) , ( b ) , and ( c ) for the control measurements on the @xmath5 , the @xmath6 , and the @xmath17 system ) .
the blue circles show the probability to measure the expected @xmath109 ( 2 atoms for the @xmath5 system , 3 atoms for the @xmath6 system , and 2 atoms for the @xmath17 system ) .
this probability is always smaller than one due to the finite experimental fidelity of our measurements and due to three and four - body losses .
since our experimental fidelity is on the order of @xmath110 , we can assume that shots with two or more atoms missing [ black squares in fig .
[ fig - spin_down_raw](a ) , ( b ) , ( c ) ] are primarily a result of three and four - body losses . for the @xmath5 system , three - body losses mostly appear for tunneling close to the resonance [ black squares in fig . [ fig - spin_down_raw](a ) ] , while for the @xmath6 and the @xmath17 systems trap losses can be observed for tunneling at all interaction strengths [ black squares in fig .
[ fig - spin_down_raw](b ) , ( c ) ] .
after the initial tunneling process , we selectively remove the spin - down component from the trap and measure the number of spin - up atoms ( @xmath111 ) in the trap [ fig .
[ fig - spin_down_raw](d ) , ( e ) , and ( f ) for the @xmath5 , the @xmath6 and the @xmath17 systems ] . in these figures ,
the blue circles correspond to the probability of spin - down tunneling .
we correct the probability of spin - down tunneling for three and four - body losses , by subtracting the probabilities of trap losses [ black squares in fig .
[ fig - spin_down_raw](a ) , ( b ) , ( c ) ] from the respective probabilities to detect 0 spin - up atoms ( 0 , or 1 spin - up atoms in the @xmath6 system ) [ black squares in fig . [ fig - spin_down_raw](d ) , ( e ) , ( f ) ] .
since the sum of all probabilities has to be normalized , this correction increases the values of both the red and blue symbols in fig .
[ fig - spin_down_raw](d ) , ( e ) , and ( f ) , while keeping their respective ratio constant .
the corrected probabilities are shown in fig .
[ fig - spin_down_correction ] .
the blue circles in this figure are the corrected probabilities of spin - down tunneling that are also shown in fig . 3 of the main text .
in the fermionization regime of diverging interaction strength , a 1d multicomponent system of @xmath77 fermionic atoms acquires the density distribution of @xmath77 identical noninteracting fermions . in this limit ,
the ordering of the atoms is fixed and a one - to - one correspondence between the full wave function of space and spin and a pure spin wave function can be established [ ] .
this implies that all observables can be mapped from the full continuous hilbert space into a discrete spin space and vice versa .
this mapping therefore allows to treat a 1d system in the fermionization regime with a spin - chain model and therefore constitutes a substantial simplification of its theoretical description . for a spin state @xmath112 consisting of @xmath77 spin-1/2 particles
, we will use the notation @xmath113 where the spin state @xmath114 of the @xmath90th particle in the spin chain can either be up @xmath115 or down @xmath116 .
the spin - chain hamiltonian of @xmath77 spins is given by [ ] @xmath117 here , @xmath118 is the energy of @xmath77 spinless noninteracting fermions in 1d , @xmath90 refers to the @xmath90th particle , @xmath119 permutes neighboring spins and @xmath21 is the exchange coupling between neighboring spins .
@xmath21 is given by the formula @xmath120 here , @xmath121 is the ground state of @xmath77 spinless noninteracting fermions in 1d and the function @xmath122 is 1 if @xmath123 and zero otherwise .
it is important to note that @xmath21 is proportional to the inverse interaction strength ( @xmath124 ) and that it is approximately proportional to the local density cubed and therefore depends on the trap geometry . to determine the exchange couplings @xmath21 for our tunneling measurements , we calculate the quasi - bound states @xmath121 in the tilted trap ( fig .
1c of main text ) as follows : first , we approximate the potential by the taylor series around its local minimum .
we adjust the maximum order of the taylor series such that the potential is well approximated up to the barrier maximum and that the taylor series is larger than the barrier maximum outside the quasi - bound region .
then , we calculate the spectrum and the eigenfunctions of the approximate potential using the eigenfunctions of the harmonic oscillator at the local minimum . in our model , we consider tunneling of one atom from the trap .
hence , the model connects an initial spin state @xmath125 , which is an eigenstate of @xmath126 , with a final spin state @xmath127 . in the final state
, @xmath128 is an eigenstate of @xmath129 and describes the @xmath130 particles that remain in the trap and @xmath131 indicates the spin orientation of the particle that left the trap .
the energy @xmath132 of the tunneling particle is equal to the difference between the energies of the initial and final in - trap states .
close to the cir , the energy shifts of all states are linear in @xmath133 ( fig
. 2 of the main text ) and the eigenenergies of @xmath125 and @xmath128 are well approximated by the solutions of the spin - chain model . in this limit
, their difference is @xmath134 further away from the cir , we interpolate between the linear energy shifts of the spin - chain model and the corresponding saturation energies at @xmath135 using the rescaled energy shifts of the @xmath136 system in the ground state [ ] .
the weight of each tunneling channel @xmath137 for tunneling from @xmath125 to @xmath138 is approximated by @xmath139 that means that it is proportional to the squared spin overlap ( @xmath140 ) , the energy of the tunneling particle ( @xmath132 ) and the probability ( @xmath141 ) that a particle with energy @xmath132 tunnels through the barrier .
the energy - dependent tunneling parameter @xmath142 is given by @xmath143 } .\ ] ] here , @xmath79 and @xmath80 are the two largest solutions of @xmath144 .
the probability to tunnel from state @xmath125 to state @xmath128 is given by @xmath145 where the sum is over all possible final states @xmath146 .
exactly on resonance , @xmath147 , the energy differences @xmath132 are equal for all final states ( fig . 2 of the main text ) .
therefore , the energy - dependent term @xmath148 in eq .
[ eq : tunnelrate ] is equal for all tunneling channels and can be canceled from eq .
[ eq : tunnelprob ] . in this case
the probability to tunnel from state @xmath125 to state @xmath138 simplifies to @xmath149 since @xmath150 .
the limiting behavior of @xmath151 far left and right from the cir is determined by the predominant tunneling channels of these regimes . for small repulsive interactions ( left blue arrow in fig .
2 of the main text ) , tunneling into the antiferromagnetic @xmath97-particle in - trap state is predominant since its energy is much smaller than the energies of all other @xmath97-particle states . for small attractive interactions ( right blue arrow in fig . 2 of the main text ) ,
tunneling into the fm @xmath97-particle in - trap state is predominant since the ordering of the energy levels is inverted .
this is also confirmed by the measurements of section vi .
using eq . [ eq : tunnelprob ] , we calculate the probabilities for spin - down tunneling @xmath20 as a function of interaction strength . here , @xmath20 is the probability of tunneling from an initial state @xmath125 into the final state @xmath152 , where the in - trap state @xmath128 only contains spin - up atoms and @xmath131 only contains spin - down atoms .
the initial and final in - trap states ( @xmath125 and @xmath128 ) of our tunneling model are always eigenstates of the spin - chain hamiltonian ( eq .
[ eq : spin - chain_hamiltonian ] ) .
in particular , the initial state will always be the antiferromagnetic ground state @xmath153 of the repulsive spin - chain model with @xmath77 atoms .
we will solve the @xmath5 and @xmath17 systems analytically to make the physics more transparent . to achieve this
, we will first determine the eigenstates of the @xmath5 system in a symmetric trap and then show how they mix in a tilted trap ( see fig .
1(c ) of the main text and section vii ) . in the @xmath17 case
, we will first determine the two eigenstates of @xmath154 with @xmath155 .
again , we will calculate the spin - chain hamiltonian within the subspace of these two states and show how they couple in a tilted trap .
we will show that both systems are closely related , since the @xmath17 system ends up in the afm states of the three - particle system with almost 100% probability after the first tunneling process .
within the spin basis @xmath156 , @xmath157 , and @xmath158 the spin - chain hamiltonian reads @xmath159 it has in a symmetric trap ( @xmath160 ) the eigenstates [ ] @xmath161 @xmath162 @xmath163 , @xmath164 , @xmath165 . in this eigenbasis of the symmetric trap
, the spin - chain hamiltonian reads @xmath166 therefore , an imbalance of @xmath167 and @xmath168 coherently mixes the afm with the i m state , while the fm state is decoupled .
this can be understood , since both the afm and the i m state have an absolute value of the total spin of @xmath169 , while the fm state has @xmath170 .
the eigenstates of @xmath171 in a nonharmonic trap can hence be written as @xmath172 , @xmath173 with the mixing angle @xmath174 .
the @xmath5 system is initially prepared in state @xmath175 .
exactly on resonance , @xmath147 , the ordering of the particles is fixed by their mutual repulsion such that only the rightmost particle can leave the trap ( fig .
1(c ) of the main text ) .
the probability for spin - down tunneling , @xmath36 , is hence just the probability of the rightmost spin of the @xmath175 state to point downwards , @xmath176 in the symmetric trap @xmath177 , the probability for spin - down tunneling is @xmath178 . in the tilted trap ,
@xmath179 is positive , since the density between the left particle and the particle in the middle is larger than the density between the particle in the middle and the right particle ( fig .
1(c ) of the main text ) .
@xmath180 is hence lower than @xmath181 in the tilted trap .
we obtain for our experimental trap parameters during the tunneling process a mixing angle of @xmath182 resulting in @xmath183 for the @xmath175 state ( fig .
3(a ) of the main text ) . away from resonance , @xmath184
, one has to calculate the weights of all the possible tunneling channels .
the eigenstates of @xmath185 and @xmath186 are @xmath187 , @xmath188 , and @xmath189 .
these eigenstates have the eigenenergies @xmath190 and @xmath191 ( now calculated with @xmath167 of the two - particle in - trap system ) .
the probability for spin - down tunneling , @xmath36 , is the probability to tunnel from the initial state @xmath175 into the final state @xmath187 .
far left from the cir , tunneling into @xmath192 is predominant and one obtains the limiting value @xmath193 far right from the cir , tunneling into @xmath194 and @xmath195 is predominant and one obtains the limiting value @xmath196 ( see fig .
3(a ) of the main text ) .
interestingly , this limiting value is independent of the mixing angle @xmath197 and hence the tilting of the trap . in between
, @xmath198 is calculated using the parameters @xmath107 and @xmath133 of table [ fig - spin_down_table](a ) .
this leads to the red solid curve in fig .
3(a ) of the main text .
similar calculations for the i m and fm initial states lead to the dotted and dashed curves of fig .
3(a ) of the main text , respectively .
we assumed for the above calculations that the trap is tilted so slowly that the spin configuration can follow the change of @xmath167 and @xmath168 induced by the change of the total density .
this assumption is wrong close to the cir , where the spin configuration of the initially deeper trap may be frozen .
we would hence expect a larger @xmath36 close to the cir .
a full dynamical calculation showed , however , that this effect is not sufficiently strong to explain the resonance feature of @xmath36 close to the cir . within the spin basis @xmath199 , @xmath200 , @xmath201 , and
@xmath202 the spin - chain hamiltonian reads @xmath203 the @xmath6 system is initially prepared in @xmath204 , which is the ground state of this hamiltonian for @xmath205 .
we determine this state numerically .
we have to consider tunneling into the final states @xmath206 , @xmath207 , @xmath208 , and @xmath209 , where @xmath210 is given by eq .
( [ eq - fm ] ) , @xmath211 is given by eq .
( [ eq - im ] ) , and @xmath212 is given by eq .
( [ eq - afm ] ) .
exactly on resonance , the probability for spin - down tunneling is just given by @xmath213 we obtain for the harmonic trap a value of @xmath214 [ ] . this value is again lowered by the tilting and we obtain for the experimental parameters at the cir ( see table [ fig - spin_down_table](b ) ) a value of @xmath215 .
far left from the cir , tunneling into @xmath216 is predominant and one obtains again the limiting value @xmath217 .
far right from the cir , tunneling into the two fm final states , @xmath218 and @xmath219 , is predominant and one obtains the limiting value @xmath220 which is again independent of the tilting .
the experimentally measured value for @xmath36 is , however , only @xmath221 in the super - tonks regime ( see fig .
3(b ) of the main text ) .
the deviation from the theoretical prediction ( @xmath222 ) is presumably caused by a stronger contribution of the i m tunneling channel in the experiment .
indeed , the i m tunneling channel contributes stronger if the lowering of the potential barrier is assumed to be nonadiabatic , as outlined in the following paragraphs . in the fermionization regime ,
the energy of the energetically highest particle is given by the @xmath77th trap level .
this energy grows linearly with @xmath133 in the fermionization regime and it saturates in the weakly - interacting regimes , @xmath135 , at values , which are again given by the ( quasi - bound ) trap levels [ ] . the saturated energy shifts can hence never exceed the barrier maximum .
this situation applies to an adiabatic lowering of the potential barrier before the tunneling process .
this condition is fulfilled , if the tilted trap is deep enough or if the barrier is lowered slowly when the corresponding trap level approaches the barrier maximum .
if the barrier height is nonadiabatically lowered , the saturated energy shifts may exceed the barrier maximum . in this case the saturated shifts are multiples of the level spacing at the fermi edge and they are not limited by the barrier maximum .
the energy of the tunneling particle may now exceed the barrier maximum for the energetically most favorable tunneling channels . in that case
the particle tunnels out with a probability of 1 . moreover , the tunneling energies of the intermediate channels may also be slightly below or above the barrier maximum .
this is the case for the i m tunneling channel of the @xmath6 system in the super - tonks regime . as a result ,
the weight of the i m tunneling channel is much stronger than in the adiabatic approximation , which leads to a substantial reduction of @xmath36 in the super - tonks regime .
we obtain @xmath223 and @xmath224 for the @xmath6 system in the super - tonks regime , which is in much better agreement with the experimental data .
we finally note that , in contrast to the @xmath6 system , both models lead to nearly the same results for the @xmath5 and @xmath17 systems . within the spin basis @xmath225 , @xmath226 , @xmath227 , @xmath228 , @xmath229 , and @xmath230 the spin - chain hamiltonian reads @xmath231 the @xmath17 system is initially prepared in @xmath232 , which is the ground state of this hamiltonian for @xmath205 .
it can be determined analytically since only the ground state and another excited have an absolute value of the total spin of @xmath155 .
the two eigenstates with @xmath155 are @xmath233 and @xmath234 in the subspace of these two states , the spin - chain hamiltonian attains the form @xmath235 the afm ground state is hence @xmath236 with the mixing angle @xmath237 and the energy @xmath238 .
exactly on resonance , the probability for the tunneling of two spin - down particles is therefore given by @xmath239 we obtain for the experimental parameters at the cir ( see table [ fig - spin_down_table](c ) ) an angle of @xmath240 and hence @xmath241 .
away from resonance , we have to consider the sequential tunneling of two particles from the initial @xmath232 state . in the following ,
we discuss only the case that a spin - down particle has tunnelled in the first tunneling process ( the spin - up case is analogous ) .
therefore , we have to consider the tunneling into the final states @xmath242 , @xmath243 , and @xmath244 , where @xmath210 is given by eq .
( [ eq - fm ] ) , @xmath211 is given by eq .
( [ eq - im ] ) , and @xmath212 is given by eq .
( [ eq - afm ] ) .
it follows from eqs .
( [ eq - fm ] ) and ( [ eq - afm4 ] ) that the spin overlap of @xmath232 with @xmath245 is exactly zero independent of the tilting of the trap .
moreover , the squared spin overlap with the i m state , @xmath246 reaches at most @xmath247 , while the squared spin overlap with the afm state , @xmath248 is at least @xmath249 ( the remaining @xmath250 belong to the spin - up case ) .
therefore , even in the super - tonks regime , where tunneling into the i m state is energetically favored , the particle tunnels with at most @xmath251 into the i m state due to the small spin overlap . we conclude that the system ends up in the afm three - particle in - trap state with @xmath221 after the first tunneling process .
the second tunneling process is then the tunneling from the afm state of a ( 2,1 ) system , which has been discussed above .
it follows that the probability for spin - down tunneling , @xmath252 , is for the ( 2,2 ) system half as large as for the ( 2,1 ) system in the whole interaction regime .
this is indeed the case , as can be seen in fig .
3(c ) of the main text .
the occupation - number distributions of fig . 4 of the main text have been calculated by means of an exact diagonalization of the full continuous model .
the second - quantized hamiltonian of the full model reads @xmath253 with the fermionic field operators @xmath254 and the angular frequency @xmath15 of the harmonic oscillator at the local minimum .
we expand the field operators in the energy eigenbasis of the harmonic oscillator , @xmath255 . here
, @xmath90 refers to the @xmath90th energy level , @xmath256 is the @xmath90th energy eigenstate of the harmonic oscillator , and @xmath257 annihilates a fermion in energy eigenstate @xmath90 and spin state @xmath258 .
the hamiltonian is then numerically calculated and diagonalized in the many - body energy eigenbasis of the noninteracting problem . here
, the finite size of the computer memory restricts the calculation to all noninteracting eigenstates below a certain cutoff energy .
this allows for an accurate numerical calculation of the low - energy eigenstates of the interacting few - body system ( for not too many particles and not too strong interactions , @xmath259 ) .
the mean occupancy of single - particle level @xmath260 is the expectation value @xmath261 of the corresponding eigenstate of the interacting system . | we report on the deterministic preparation of antiferromagnetic heisenberg spin chains consisting of up to four fermionic atoms in a one - dimensional trap .
these chains are stabilized by strong repulsive interactions between the two spin components without the need for an external periodic potential .
we independently characterize the spin configuration of the chains by measuring the spin orientation of the outermost particle in the trap and by projecting the spatial wave function of one spin component on single - particle trap levels .
our results are in good agreement with a spin - chain model for fermionized particles and with numerically exact diagonalizations of the full few - fermion system .
the high control and tunability of ultracold atomic systems offer the fascinating possibility to simulate quantum magnetism @xcite , a topic of fundamental importance in condensed matter physics @xcite .
systems of spin-1/2 fermions with antiferromagnetic ( afm ) correlations are of particular interest due to the observation of high - temperature superconductivity in cuprates with afm correlations @xcite . the experimental implementation of the necessary exchange couplings is usually realized by superexchange processes of neighboring atoms in the mott - insulating state of a deep optical lattice .
superexchange couplings were measured in both bosonic @xcite and fermionic double - well systems @xcite and short - range afm correlations of fermionic atoms were detected in various lattice geometries @xcite .
furthermore , superexchange processes were used to study the dynamics of spin impurities above the ferromagnetic ( fm ) ground state of bosons in the mott - insulating state of a one - dimensional ( 1d ) lattice @xcite .
bosonic atoms were also used to simulate afm ising spin chains in a tilted optical lattice @xcite
. however , the afm ground state of spin-1/2 fermions in a deep optical lattice has so far not been realized due to the very low energy scale associated with the superexchange coupling .
this problem can be circumvented in 1d systems , where quantum magnetism can be simulated without an optical lattice @xcite . in the regime of strong interactions ,
the spatial wave function of both fermions @xcite and bosons @xcite can be mapped on the wave function of spinless noninteracting fermions [ fig .
[ fig - sketch](a ) ] .
in this so called fermionization limit , the strong interactions lead to the formation of a wigner - crystal - like state @xcite , which has a highly degenerate ground state when the particles have multiple internal degrees of freedom [ fig .
[ fig - sketch](b ) ] @xcite .
close to the limit of fermionization , the structure of the quasi - degenerate ground - state multiplet @xcite is determined by an effective sutherland spin - chain hamiltonian , which for two - component systems becomes a heisenberg model @xcite . in this paper , we report on the realization of heisenberg spin chains of @xmath0 spin - up and @xmath1 spin - down particles with @xmath2 , @xmath3 , and @xmath4 .
we show that under an adiabatic change of the interaction strength the noninteracting ground states of these systems evolve into the respective afm states in the limit of infinitely strong repulsion @xcite .
we identify the afm states by two independent measurements .
first , we use a tunneling technique to measure the spin orientation of the outermost particle of the spin chain .
second , we probe the spatial wave function of the spin - down atom in the @xmath5 and @xmath6 system by projecting it on single - particle trap levels . in our experiments ,
we realize a spin-@xmath7 system by trapping ultracold @xmath8 atoms in an elongated optical dipole trap @xcite in their two lowest hyperfine states @xmath9 and @xmath10 . as
the energy of the atoms is much smaller than the lowest transverse excitation energy in the trap , their dynamics are restricted to the longitudinal axis of the trap .
in such a quasi-1d system , the interaction strength between ultracold atoms of opposite spin is determined by the 1d coupling constant , @xmath11 , which diverges at a confinement - induced resonance ( cir ) when the 3d scattering length , @xmath12 , approaches the harmonic oscillator length of the radial confinement @xcite .
we use a magnetic feshbach resonance to control @xmath12 and therefore are able to smoothly tune @xmath11 across the cir . at the same time
, scattering between fermionic atoms of the same spin component is forbidden . throughout this paper , @xmath11 will be given in units of @xmath13 , where @xmath14 and @xmath15 are the harmonic oscillator length and the trap frequency in longitudinal direction and @xmath16 is the mass of a @xmath8 atom .
we start our experiments by preparing a @xmath5 , @xmath6 , or @xmath17 system in the noninteracting many - particle ground state of the trap @xcite . by changing the magnetic offset field with a constant rate
, we ramp the system into the fermionization regime close to the cir ( fig .
[ fig - tunneling_energies ] ) , where it forms a spin chain . below the cir
, @xmath11 is large and positive and the system is in the tonks regime of strong repulsion @xcite . when crossing the cir , @xmath11 changes sign from @xmath18 to @xmath19 while the system continuously follows the so called upper branch @xcite into the super - tonks regime of strong attraction @xcite ( fig .
[ fig - tunneling_energies ] ) . in the super - tonks regime ,
the system is in an excited state , which is metastable against decay into bound states . in a first set of measurements ,
we identify the states of the spin chains by probing the spin distributions in the trap . here
, we make use of the fact that in the fermionization regime the atoms become impenetrable and therefore their ordering along the longitudinal axis of the trap is fixed .
this allows us to determine the spin orientation of the outermost particle in the trap in a tunneling measurement . to do this
, we tilt the trap as shown in fig .
[ fig - sketch](c ) and thereby allow atoms to tunnel out of the trap .
we carefully adjust the trap parameters during the tunneling process , to let exactly one atom ( for the @xmath5 and the @xmath6 systems ) or two atoms ( for the @xmath17 system ) tunnel @xcite .
finally , we measure the number of spin - up atoms in the final state to determine the spin of the atoms that left the trap during the tunneling process @xcite .
we define spin - down tunneling as the process in which all spin - down atoms tunnel out . by repeating this measurement at different magnetic offset fields ,
we deduce the probability of spin - down tunneling , @xmath20 , as a function of the inverse 1d coupling constant as shown in fig .
[ fig - tunneling ] . as shown in fig .
[ fig - sketch](b ) for a @xmath5 system in a harmonic trap , the different states of the spin chain can be uniquely identified by their spin densities @xcite , specifically by the probability of the outermost spin to point downwards . since in the fermionization regime
the ordering of the atoms in the trap is fixed only the outermost atom can escape during the tunneling process .
exactly at the cir , the probability of spin - down tunneling should therefore directly reveal the state of the spin chain @xcite .
away from resonance , the probability of spin - down tunneling is also influenced by the energy of the final in - trap states , favoring final states with lower energy as indicated by the blue arrows in fig .
[ fig - tunneling_energies ] . to identify the spin states throughout the entire spin - chain regime
, we compare our data to the results of a tunneling model , which in the following section is explained for a @xmath5 system . in our tunneling model ,
the initial states are eigenstates of a heisenberg spin - chain hamiltonian @xcite , where the exchange couplings @xmath21 between neighboring spins depend on the trap geometry and on the inverse 1d coupling constant @xcite . for the @xmath5 system with repulsive interactions and a symmetric trap @xmath22 ,
these eigenstates are the antiferromagnetic ( afm ) ground state , the intermediate ( i m ) state , and the ferromagnetic ( fm ) state as shown in fig .
[ fig - sketch](b ) . during the tunneling process
the trap is tilted as shown in fig .
[ fig - sketch](c ) and therefore the density is not symmetric .
hence , the exchange couplings are not identical anymore @xmath23 , which leads to a coherent mixing of the afm and i m state during the tunneling process @xcite .
we calculate a probability of approximately @xmath24 for the rightmost spin of the afm state in the tilted trap to point downwards .
this is in good agreement with the blue data points in fig .
[ fig - tunneling](a ) that cross the cir at @xmath25 .
away from the cir , the eigenstates of both the three - particle and the two - particle spin chains are nondegenerate ( fig .
[ fig - tunneling_energies ] ) . in this case , the energies of the initial three - particle state @xmath26 and the final two - particle state @xmath27 involved in the tunneling process are important , since their difference determines the energy @xmath28 of the tunneling particle .
the tunneling rate of the particle that leaves the trap is strongly affected by its energy and can be calculated as @xmath29 where @xmath30 with @xmath31 indicating the spin orientation of the tunneling particle .
the tunneling parameter @xmath32 is determined by means of a wkb calculation @xcite .
the probability to tunnel from state @xmath26 to state @xmath27 is given by @xmath33 where the sum is over all possible final states @xmath34 .
using eq .
( [ eq - tunnel_probability ] ) , we calculate the probabilities @xmath35 of tunneling into the spin - polarized final state [ red lines in fig .
[ fig - tunneling](a ) ] , which is equivalent to the probability of spin - down tunneling ( @xmath36 ) . far below the cir ,
the energy dependent term @xmath37 dominates the outcome of the tunneling rates ( eq . [ eq - tunnel_rate ] ) .
therefore , tunneling into the afm two - particle ground state @xmath38 is strongly favored if its spin overlap to the initial state is not zero .
this leads to a limiting value of @xmath39 for initial afm and i m states . above the resonance , the energy ordering of the two - particle fm and afm states
is reversed and tunneling into the fm states is predominant ( fig . [ fig - tunneling_energies ] ) attainable in our experiment , the spatial wave function overlap between the spin - chain states and the molecular ground state is negligible .
we expect that for a small negative @xmath11 tunneling into the molecular state is predominant . ] . here , @xmath36 is determined by the ratio of the spin overlaps between the first two spins of the initial states and the fm two - particle states @xmath40 and @xmath41 .
the comparison of the theoretically predicted @xmath36 of the afm state in the tilted trap [ red solid line in fig [ fig - tunneling](a ) ] with the experimental data ( blue points ) shows good agreement , while the fm ( red dashed line ) and i m ( red dotted line ) states are clearly excluded .
we therefore conclude that before tunneling both below and above the cir the system is in the afm state .
the gray points at @xmath42 indicate a narrow resonance effect that couples the afm state to the i m state of the spin chain .
since this resonance is accompanied by strongly enhanced three - body losses @xcite , we suspect it to be caused by a coupling of the afm and the i m states via a molecular state with center - of - mass excitation . the coupling to such molecular states is strongly enhanced by the anharmonicity of our tilted trap @xcite . for the afm state of the @xmath6 system , a similar calculation predicts @xmath43 on resonance and a saturation value of @xmath44 deep in the super - tonks regime @xcite .
as shown in fig [ fig - tunneling](b ) , the general trend of our measurements agrees with this prediction for the afm state , but in the super - tonks regime , there is a significant deviation .
the reason for this deviation is that the calculation assumes an adiabatic lowering of the potential barrier . as a result ,
the tunneling energies of all tunneling channels are always well below the barrier maximum .
we believe that this condition is not fulfilled for the @xmath6 system in the super - tonks regime , where an especially low potential barrier was used for the tunneling measurement .
indeed , if we model a nonadiabatic lowering of the potential barrier , the contribution from tunneling into the i m state reduces @xmath36 to values that are compatible with the experimental results @xcite . in order to study the spin configuration of the balanced @xmath17 system ,
we adapt the previous procedure and let two atoms tunnel out of the trap . here ,
@xmath45 is defined as the probability to end up in state @xmath40 , where both spin - down atoms tunneled out of the trap .
again , the predicted @xmath46 on resonance and the limiting value of @xmath47 in the super - tonks regime are in good agreement with the experiment as shown in fig .
[ fig - tunneling](c ) . to independently confirm the results of our measurement of the spin distribution
, we perform a second set of measurements that directly probes the spatial wave function of the system . as shown in fig .
[ fig - sketch](a ) , the relative spatial wave function between identical spins always exhibits a smooth zero crossing , while between distinguishable spins with strong interactions it can exhibit a cusp . the cusps lead to occupancies of high - energy trap levels , while the zero crossings require only the occupation of the lowest trap levels . in general , the more symmetric the spatial wave function of a state is , the more cusps it will contain . therefore , the occupation - number distribution on single - particle trap levels directly reveals the spin configuration of the system .
to probe this distribution , we prepare an interacting ( 2,1 ) or ( 3,1 ) system and remove all atoms of the spin - up component from the trap with a short pulse of light .
the light is @xmath48 polarized and resonant to the d2 transition of the spin - up atoms @xmath49 to @xmath50 .
we confirm that within our experimental fidelity all spin - up atoms are removed from the trap by the light pulse , while only @xmath51 of the population of spin - down atoms is lost .
with @xmath52 the duration of the light pulse is significantly shorter than the inverse longitudinal trap frequency of approximately @xmath53 , which sets the timescale of redistribution along the spin chain .
this process therefore projects the spin - down component of the wave function of the interacting ( @xmath54)-particle system on single - particle trap levels .
finally , we measure the mean occupancies on the single - particle trap levels @xcite . in fig .
[ fig - occupancies ] we compare the mean occupancies of the spin - down atom for the @xmath5 and the @xmath6 systems in the super - tonks regime with the theoretical prediction that we obtained by numerically diagonalizing the many - body hamiltonian for these systems .
the comparison shows that both systems are in the afm spin state and thereby confirms that our systems follow this state throughout the fermionization regime . in conclusion
, we have prepared antiferromagnetic heisenberg spin chains of up to four atoms in a one - dimensional trap and independently probed the spin distributions and spatial wave functions of the systems .
this constitutes a direct observation of quantum magnetism beyond two - particle correlations in a system of ultracold fermionic atoms . by using the methods developed in ref .
[ ] , multiple spin chains can be realized and coupled , which offers a new approach to studying two and three - dimensional quantum magnetism .
+ + we thank d. blume , i. brouzos , g. m. bruun , g. conduit , j. c. cremon , s. e. gharashi , c. greene , m. rontani , a. n. wenz and n. t. zinner for helpful discussions and input .
this work was supported by the european research council starting grant no .
279697 , the heidelberg center for quantum dynamics , the dfg ( project sa 1031/7 - 1 and rtg 1729 ) , the cluster of excellence quest , the swedish research council , and nanolund .
s.m . and f.d .
contributed equally to this work .
m. lewenstein , a. sanpera , v. ahufinger , b. damski , a. sen de , and u. sen , ultracold atomic gases in optical lattices : mimicking condensed matter physics and beyond , http://dx.doi.org/10.1080/00018730701223200[_adv .
phys . _ * 56 * , 243 ( 2007 ) ] .
a. auerbach , _ interacting electrons and quantum magnetism , _ ( springer , new york , 1994 ) .
m. r. norman , the challenge of unconventional superconductivity , http://dx.doi.org/10.1126/science.1200181 [ _ science _ * 332 * , 196 ( 2011 ) ] .
s. trotzky , p. cheinet , s. flling , m. feld , u. schnorrberger , a. m. rey , a. polkovnikov , e. a. demler , m. d. lukin , and i. bloch , time - resolved observation and control of superexchange interactions with ultracold atoms in optical lattices , http://dx.doi.org/10.1126/science.1150841[_science_ * 319 * , 295 ( 2008 ) ] .
s. murmann , a. bergschneider , v. m. klinkhamer , g. zrn , t. lompe , and s. jochim , two fermions in a double well : exploring a fundamental building block of the hubbard model , http://dx.doi.org/10.1103/physrevlett.114.080402[_phys .
rev .
lett . _ * 114 * , 080402 ( 2015 ) ] .
d. greif , t. uehlinger , g. jotzu , l. tarruell , and t. esslinger , short - range quantum magnetism of ultracold fermions in an optical lattice , http://dx.doi.org/10.1126/science.1236362[_science_ * 340 * , 1307 ( 2013 ) ] .
r. a. hart , p. m. duarte , t .- l .
yang , x. liu , t. paiva , e. khatami , r. t. scalettar , n. trivedi , d. a. huse , and r. g. hulet , observation of antiferromagnetic correlations in the hubbard model with ultracold atoms , http://dx.doi.org/10.1038/nature14223[_nature_ * 519 * , 211 ( 2015 ) ] .
m. messer , r. desbuquois , t. uehlinger , g. jotzu , s. huber , d. greif , and t. esslinger , exploring competing density order in the ionic hubbard model with ultracold fermions , http://dx.doi.org/10.1103/physrevlett.115.115303[_phys .
rev .
lett . _ * 115 * , 115303 ( 2015 ) ] .
t. fukuhara , a. kantian , m. endres , m. cheneau , p. schau , s. hild , d. bellem , u. schollwck , t. giamarchi , c. gross , i. bloch , and s. kuhr , quantum dynamics of a mobile spin impurity , http://dx.doi.org/10.1038/nphys2561[_nat .
phys . _ * 9 * , 235 ( 2013 ) ]
. j. simon , w. s. bakr , r. ma , m. e. tai , p. m. preiss , and m. greiner , quantum simulation of antiferromagnetic spin chains in an optical lattice , http://dx.doi.org/10.1038/nature09994[_nature_ * 472 * , 307 ( 2011 ) ] .
f. meinert , m. j. mark , e. kirilov , k. lauber , p. weinmann , a. j. daley , and h .- c .
ngerl , quantum quench in an atomic one - dimensional ising chain , http://dx.doi.org/10.1103/physrevlett.111.053003[_phys .
rev .
lett . _ * 111 * , 053003 ( 2013 ) ] .
f. deuretzbacher , d. becker , j. bjerlin , s. m. reimann , and l. santos , quantum magnetism without lattices in strongly interacting one - dimensional spinor gases , http://dx.doi.org/10.1103/physreva.90.013611[_phys .
rev .
a _ * 90 * , 013611 ( 2014 ) ] .
m. e. beverland , g. alagic , m. j. martin , a. p. koller , a. m. rey , and a. v. gorshkov , realizing exactly solvable su(n ) magnets with thermal atoms , http://arxiv.org/abs/1409.3234[arxiv:1409.3234 ] .
g. pagano , m. mancini , g. cappellini , p. lombardi , f. schfer , h. hu , x .-
j .
liu , j. catani , c. sias , m. inguscio , and l. fallani , a one - dimensional liquid of fermions with tunable spin , http://dx.doi.org/10.1038/nphys2878[nat .
phys .
* 10 * , 198 ( 2014 ) ] .
g. zrn , f. serwane , t. lompe , a. n. wenz , m. g. ries , j. e. bohn , and s. jochim , fermionization of two distinguishable fermions , http://dx.doi.org/10.1103/physrevlett.108.075303[_phys .
rev .
lett .
_ * 108 * , 075303 ( 2012 ) ] .
m. girardeau , relationship between systems of impenetrable bosons and fermions in one dimension , http://dx.doi.org/10.1063/1.1703687[_j .
math .
phys . _ * 1 * , 516 ( 1960 ) ] .
b. paredes , a. widera , v. murg , o. mandel , s. flling , i. cirac , g. v. shlyapnikov , t. w. hnsch , and i. bloch , tonks-girardeau gas of ultracold atoms in an optical lattice , http://dx.doi.org/10.1038/nature02530[_nature_ * 429 * , 277 ( 2004 ) ] .
t. kinoshita , t. wenger , and d. s. weiss , observation of a one - dimensional tonks - girardeau gas , http://dx.doi.org/10.1126/science.1100700[_science_ * 305 * , 1125 ( 2004 ) ] .
k. a. matveev , conductance of a quantum wire in the wigner - crystal regime , http://dx.doi.org/10.1103/physrevlett.92.106801[_phys .
rev .
lett . _ * 92 * , 106801 ( 2004 ) ] ; k. a. matveev , conductance of a quantum wire at low electron density , http://dx.doi.org/10.1103/physrevb.70.245319[_phys .
rev .
b _ * 70 * , 245319 ( 2004 ) ] . f. deuretzbacher , k. fredenhagen , d. becker , k. bongs , k. sengstock , and d. pfannkuche , exact solution of strongly interacting quasi - one - dimensional spinor bose gases , http://dx.doi.org/10.1103/physrevlett.100.160405[_phys . rev . lett . _ * 100 * , 160405 ( 2008 ) ] .
k. a. matveev and a. furusaki , spectral functions of strongly interacting isospin-1/2 bosons in one dimension , http://dx.doi.org/10.1103/physrevlett.101.170403[_phys .
rev .
lett . _ * 101 * , 170403 ( 2008 ) ] .
m. d. girardeau and a. minguzzi , soluble models of strongly interacting ultracold gas mixtures in tight waveguides , http://dx.doi.org/10.1103/physrevlett.99.230402[_phys .
rev .
lett . _ * 99 * , 230402 ( 2007 ) ] .
l. guan , s. chen , y. wang , and z .- q .
ma , exact solution for infinitely strongly interacting fermi gases in tight waveguides , http://dx.doi.org/10.1103/physrevlett.102.160402[_phys .
rev .
lett . _ * 102 * , 160402 ( 2009 ) ] .
s. e. gharashi and d. blume , correlations of the upper branch of 1d harmonically trapped two - component fermi gases , http://dx.doi.org/10.1103/physrevlett.111.045302[_phys .
rev .
lett . _ * 111 * , 045302 ( 2013 ) ] .
p. o. bugnion and g. j. conduit , ferromagnetic spin correlations in a few - fermion system , http://dx.doi.org/10.1103/physreva.87.060502[_phys .
rev .
a _ * 87 * , 060502(r ) ( 2013 ) ] . t. sowiski , t. grass , o. dutta , and m. lewenstein , few interacting fermions in one - dimensional harmonic trap , http://dx.doi.org/10.1103/physreva.88.033607[phys .
rev .
a * 88 * , 033607 ( 2013 ) ] .
x. cui and t .- l .
ho , ground - state ferromagnetic transition in strongly repulsive one - dimensional fermi gases , http://dx.doi.org/10.1103/physreva.89.023611[_phys .
rev .
a _ * 89 * , 023611 ( 2014 ) ] . e. j. lindgren , j. rotureau , c. forssn , a. g. volosniev , and n. t. zinner , fermionization of two - component few - fermion systems in a one - dimensional harmonic trap , http://dx.doi.org/10.1088/1367-2630/16/6/063003[_new j. phys . _
* 16 * , 063003 ( 2014 ) ] .
a. g. volosniev , d. v. fedorov , a. s. jensen , m. valiente , and n. t. zinner , strongly interacting confined quantum systems in one dimension , http://dx.doi.org/10.1038/ncomms6300[_nat .
commun . _ * 5 * , 5300 ( 2014 ) ] .
n. l. harshman , spectroscopy for a few atoms harmonically trapped in one dimension , http://dx.doi.org/10.1103/physreva.89.033633[phys .
rev . a * 89 * , 033633 ( 2014 ) ] .
m. a. garca - march , b. juli - daz , g. e. astrakharchik , j. boronat , and a. polls , distinguishability , degeneracy , and correlations in three harmonically trapped bosons in one dimension , http://dx.doi.org/10.1103/physreva.90.063605[_phys .
rev .
a _ * 90 * , 063605 ( 2014 ) ] .
l. yang , l. guan , and h. pu , strongly interacting quantum gases in one - dimensional traps , http://dx.doi.org/10.1103/physreva.91.043634[phys . rev . a * 91 * , 043634 ( 2015 ) ] .
j. levinsen , p. massignan , g. m. bruun , m. m. parish , strong - coupling ansatz for the one - dimensional fermi gas in a harmonic potential , http://dx.doi.org/10.1126/sciadv.1500197[_sci . adv . _
* 1 * , e1500197 ( 2015 ) ] .
x .- w .
guan , m. t. batchelor , and j .- y .
lee , magnetic ordering and quantum statistical effects in strongly repulsive fermi - fermi and bose - fermi mixtures , http://dx.doi.org/10.1103/physreva.78.023621[_phys .
rev .
a _ * 78 * , 023621 ( 2008 ) ] .
see supplemental material for details on the experimental methods , the data evaluation and the theoretical models , which includes refs .
@xcite .
g. zrn , _ few - fermion systems in one dimension , _ phd thesis , university of heidelberg , ( 2012 ) .
g. zrn , t. lompe , a. n. wenz , s. jochim , p. s. julienne , and j. m. hutson , precise characterization of li6 feshbach resonances using trap - sideband - resolved rf spectroscopy of weakly bound molecules , http://dx.doi.org/10.1103/physrevlett.110.135301[_phys .
rev .
lett . _ * 110 * , 135301 ( 2013 ) ] .
s. e. gharashi , x. y. yin , y. yan , and d. blume , one - dimensional fermi gas with a single impurity in a harmonic trap : perturbative description of the upper branch , http://dx.doi.org/10.1103/physreva.91.013620[_phys .
rev .
a _ * 91 * , 013620 ( 2015 ) ] . f. serwane , g. zrn , t. lompe , t. b. ottenstein , a. n. wenz , and s. jochim , deterministic preparation of a tunable few - fermion system , http://dx.doi.org/10.1126/science.1201351[_science_ * 332 * , 336 ( 2011 ) ] .
m. olshanii , atomic scattering in the presence of an external confinement and a gas of impenetrable bosons , http://dx.doi.org/10.1103/physrevlett.81.938[_phys .
rev .
lett . _ * 81 * , 938 ( 1998 ) ] .
g.e .
astrakharchik , j. boronat , j. casulleras , and s. giorgini , beyond the tonks - girardeau gas : strongly correlated regime in quasi - one - dimensional bose gases , http://dx.doi.org/10.1103/physrevlett.95.190407[_phys .
rev .
lett . _ * 95 * , 190407 ( 2005 ) ] .
e. haller , m. gustavsson , m. j. mark , j. g. danzl , r. hart , g. pupillo , and h .- c .
ngerl , realization of an excited , strongly correlated quantum gas phase , http://dx.doi.org/10.1126/science.1175850[_science_ * 325 * , 1224 ( 2009 ) ] .
s. sala , g. zrn , t. lompe , a. n. wenz , s. murmann , f. serwane , s. jochim , and a. saenz , coherent molecule formation in anharmonic potentials near confinement - induced resonances , http://dx.doi.org/10.1103/physrevlett.110.203202[_phys .
rev .
lett . _ * 110 * , 203202 ( 2013 ) ] . |
the origin of gamma - ray burst ( grb ) prompt emission remains unclear .
it has been mainly observed in the energy range of @xmath2 , and most of the spectra are fitted by a simple broken power - law function ( so - called band function ) or a cutoff power - law function @xcite , the light curves being typically very variable .
many models have been proposed to explain these properties .
most of them invoke a relativistic jet that is energized by a newly born compact object .
currently three types of emission mechanisms in the relativistic jet are being actively discussed ; photospheric , leptonic synchrotron , and hadronic emission models .
the first models assume that the thermal energy stored in the jet can be radiated as prompt emission at the thomson photosphere ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the thermal energy in the jet can be produced by the particle and/or magnetic energy dissipation between the explosion center and the photosphere .
the second and third models assume the thomson - thin region as the prompt emission site . in the second models , the electrons ( and positrons ) accelerated by the shock dissipation of the kinetic energy or by the magnetic energy dissipation radiate synchrotron and synchrotron - self - compton ( ssc ) emissions ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , while in the third models the accelerated protons and secondary particles induced by the photopion cascade process produce synchrotron and inverse compton ( ic ) emissions ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
clarifying the prompt emission mechanism and the physical quantities at the emission site would help understand the nature of the relativistic jet and the compact object that energizes the jet .
for other models and more general discussions , see recent reviews @xcite , @xcite , and @xcite . grbs were only sparsely observed in the @xmath3mev energy range , until the _ fermi _ satellite was launched 2008 june 11 .
now the gbm ( @xmath4 ) and the lat ( @xmath5 ) detectors onboard _ fermi _ provide extremely broad energy coverage with good temporal resolution for grbs .
the _ fermi _ observations will put further constraints on the above three types of prompt emission models . during its first @xmath6yr routine operation , the lat has detected 14 grbs . those are summarized in @xcite .
most of their spectra are fitted by a band function even up to @xmath7gev , while at least 3 grbs ( grb 090510 , grb 090902b , and grb 090926a ) have additional distinct spectral component at @xmath1mev .
those additional components are fitted by single power - law function .
these 3 grbs are among the brightest grbs detected by _
fermi_. this suggests that such a high - energy component may be very common and we can clearly detect such a distinct high - energy component only in bright lat grbs @xcite .
_ fermi _ also revealed that the high - energy emission ( @xmath8mev ) of most lat grbs is delayed behind the onset of the mev emission , and the high - energy emission of many lat grbs lasts longer than the mev emission , showing power - law decays , which are typically detected until @xmath9s .
the delay times in the cosmological rest frame are @xmath10s for long grbs and @xmath11s for short bursts grb 081024b and grb 090510 . although some delays of the high - energy photons are just caused by the flux increases above the lat detection threshold without a spectral change , others must clearly be attributed to the spectral changes of the band component and/or the onset of the distinct spectral component .
we note that the distinct components of the 3 grbs are delayed .
the origins of the distinct spectral components and the onset delays in the high - energy range have been actively debated .
it has been proposed that the high - energy emission can be attributed to the external shock , which is made by the interaction of the jet with the ambient medium , and the onset delays can correspond to the times for the jet decelerations @xcite ( see also * ? ? ?
* ; * ? ? ?
however , the observed high - energy emission usually has a strong variability and it often correlates with the mev emission , which is at odds with an external shock origin . especially for grb 090510
, whose long - lived high - energy emission can be explained by the external shock synchrotron emission together with the optical and x - ray afterglows detected from @xmath12s after the burst trigger @xcite , it is found that the high - energy emission in the prompt phase is much brighter than that produced by the same external shock @xcite .
a similar conclusion has been obtained for grb 090902b @xcite . as for the internal shock models ,
it is not simple to explain the lat onset delays which are larger than the variability timescales ( @xmath13 ) by the delayed brightening of the ssc emission @xcite .
the spectral index of the distinct high - energy component different from the band low - energy spectral index , seen in e.g. , grb 090902b , may not be explained simply either in these models . to overcome these problems ,
we have proposed an external inverse compton ( eic ) component ( i.e. , the emission produced by up - scattering the photons incident from outside the shock ) in addition to the synchrotron and ssc components in the internal shock model @xcite ; this model , however , requires an extreme value of the microphysical parameter @xmath14 , which does not seem common .
other alternative mechanisms such as hadronic emission mechanisms require huge isotropic energies @xcite , while magnetically - dominated jets have not been explored explicitly enough for detailed comparisons with _
fermi_/lat observational results ( see * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
in this paper we concentrate on a photospheric emission model .
such models have been briefly discussed for interpreting the relation of the mev emission with the high - energy emission . in this scenario ,
the photosphere produces a band emission component which peaks around @xmath10mev , and this can not generally produce the high - energy emission because of large opacity for @xmath15 pair creation .
the high - energy emission in these models may instead arise in a dissipation region at a larger radius @xcite . in this paper
, we substantially extend this idea and discuss its consequences in greater depth .
we focus on the case in which the energy of the jet is mainly carried by photons and baryons ( where magnetic field energy is subdominant ) , and the dissipation mechanism at large radius is internal shock . we show that in typical cases the photospheric emission is efficiently compton scattered by the electrons in internal shocks outside the photosphere , and find that the up - scattered photospheric ( up ) emission is a good candidate for the observed high - energy emission in the prompt phase .
this compton scattering is a type of the eic scattering , which is thought to be important for the high - energy emission of blazars ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , and could also be important in the internal and external shocks in grbs @xcite .
we make a generic formulation of the spectral types of the radiation from the photosphere and the internal shock including synchrotron and ssc emission for the cases of the efficient eic scattering , and clarify necessary conditions for the photospheric and up components to be dominant in the energy range of @xmath16 , rather than synchrotron and ssc components .
this formulation is applied for the well - observed lat long and short grbs , grb 080916c , grb 090902b , and grb 090510 , and we typically find reasonable parameter sets for which the data can be explained by the photospheric and up emission .
our model fits indicate that the observed delayed onset of the lat emission may be interpreted as the parameter evolution of the grb jet .
the up emission ceases at the end of the prompt mev ( photospheric ) emission , and the subsequent long - lived high - energy emission may instead be related to the external shock ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
a different origin for the prompt and late lat emission is in fact implied by the change in its behavior across this transition @xcite . in what follows ,
we concentrate on the prompt emission phase .
we discuss the general temporal and spectral properties of the radiation from the photosphere and the internal shock of the grb jet in sections [ sec : ph ] , [ sec : temporal ] , and [ sec : spectral ] , and perform the case studies of grb 080916c , grb 090902b , and grb 090510 in section [ sec : case ] .
some of the implications of our results are discussed in section [ sec : summary ] .
we consider that the jet is accelerated by the thermal pressure , and the magnetic field energy is subdominant .
we derive the luminosity and temperature of the photospheric emission from the jet and the remaining kinetic luminosity of the jet above the photosphere , which is the luminosity budget for internal shocks , according to the standard fireball model @xcite .
we assume that the jet material is optically thick and the energy is dominated by thermal radiation at the base , @xmath17 , from where it expands in the roughly adiabatic condition , i.e. , without strong conversion of the kinetic energy into the thermal energy affecting the adiabatic expansion dynamics ( for a continuous strong dissipation case , see * ? ? ?
the jet material may have a lorentz factor @xmath18 at the base .
in particular , if the grb originates from a massive stellar collapse and the jet suffers strong dissipation of its kinetic energy due to interaction with the stellar envelope , the front portion of the jet may have @xmath19 cm and @xmath20 ( see recent numerical simulations by * ? ? ? * ; * ? ? ? * and discussion in section [ sec : summary ] ) . in the absence of a strong dissipation
the base radius may be as small as a few of schwarzschild radius of the central compact object , so that @xmath21 cm and @xmath22 , where @xmath23 is the mass of the central object .
the observer - frame temperature at the base is @xmath24 where @xmath25 is the isotropic equivalent luminosity of the jet , and @xmath26 is the stefan constant . in cgs units throughout this paper . ] for general cases of grbs , the adiabatic index of the jet material below the photosphere is found to be 4/3 ( see below ) .
the dynamics of the expanding jet can be summarized as @xmath27 and @xmath28 for @xmath29 , and @xmath30 and @xmath31 for @xmath32 , where the lorentz factor saturation radius @xmath33 .
here we have defined @xmath34 , where @xmath35 is the isotropic equivalent mass ejection rate .
we see that the ratio of the radiation and baryon entropy densities @xmath36 is constant throughout the evolution below the photosphere and is much larger than unity if @xmath37 ( where @xmath38 and @xmath39 is the baryon density and fluid temperature in the comoving frame ) .
thus the assumption of the adiabatic index @xmath40 is validated .
the @xmath15 pairs drop out of equilibrium at @xmath41kev , at a radius @xmath42 cm .
we consider the case in which the jet carries enough baryons to provide an electron scattering photosphere above @xmath43 ( i.e. , the case of @xmath44 ) .
then the photosphere radius @xmath45 is defined as @xmath46 , where @xmath47 is the thomson cross section and the comoving electron density is given by @xmath48 , if there is no additional @xmath15 pair creation in the jet @xcite .
if no energy dissipation takes place around the photosphere , the photons and particles are fully thermalized , typically with the comoving temperature @xmath49kev , and then the emerging radiation from the photosphere is a blackbody with the observer - frame temperature @xmath50 ( where @xmath51 is the lorentz factor at @xmath52 ) . however , some energy dissipation processes are expected to occur around the photosphere , which can make the emerging radiation spectrum deviate from a blackbody .
such processes have been demonstrated convincingly by recent analytical and numerical work : both below and above the photosphere , a fraction of the electrons can acquire high - temperature distribution ( @xmath53 ) , either due to internal shocks and/or interaction of a jet with stellar envelope @xcite , dissipation of excited plasma waves @xcite , or nuclear collisions between protons and neutrons @xcite , or in magnetically - dominated jets due to scattering with turbulent alfvn waves @xcite or heating caused by magnetic reconnection @xcite .
multiple ic scatterings of the thermal photons by mildly relativistic electrons ( with @xmath54 ) can create a power - law tail extending from the thermal peak , which saturates at the comoving - frame energy @xmath55 as the klein - nishina limit or by the direct compton cooling , which is boosted to @xmath56gev in the observer frame .
the dissipation processes also can produce a relativistic electron population , which ic - scatter the photons to even higher energies , the limitation being the @xmath57 absorption against lower energy photons . the low - energy spectral slope of the emerging photons should not be significantly changed in such processes , i.e. , the photon index @xmath58 @xcite . in our baryonic jet model , we assume that the dissipation of the kinetic energy is not so strong that the jet dynamics is described by the adiabatic evolution , and the temperature of the emerging photons is given by @xmath59 .
this implies that the energy flux in the possible non - thermal part above @xmath60 should be smaller than that in the main thermal part , leading to an upper limit on the photon index @xmath61 of the possible power - law tail .
let us simply write down the spectral shapes of the pure thermal component and the thermal plus a non - thermal tail spectrum as @xmath62 $ ] and @xmath63 $ ] , respectively , where @xmath64 , and these are normalized as @xmath65 at @xmath66.$ ] where @xmath67 for the pure thermal component since @xmath68 is the peak energy of the @xmath69 spectrum instead of the @xmath70 spectrum .
however , we obtain the same constraint @xmath71 also from this calculation . ]
then we have @xmath72 and @xmath73\varepsilon_{\rm ph}$ ] .
the condition @xmath74 provides a constraint @xmath71 .
the value of @xmath61 is expected to be determined not only by the ratio of the photon to kinetic energies but also on the electron energy distribution which depends on the type of a dominant dissipation process , which we may not specify at the current stage .
hereafter we treat @xmath61 as a free parameter ranging within the above constraint . in such
dissipation processes copious @xmath75 pairs may be created , which increase the thomson photosphere radius .
we parameterize the number density of electrons plus positrons ( which we will call ` leptons ' hereafter ) near the photosphere as @xmath76 , where @xmath77 .
typically we can have @xmath78 , in which case the inertia is still dominated by baryons .
thus the photosphere is defined as @xmath79 .
we have @xmath80 in the low baryon load case @xmath81 , where @xmath82 in this case , most of the luminosity is radiated at the photosphere .
leptons can frequently interact with photons even above the photosphere , and this compton drag acceleration determines the remaining kinetic luminosity @xcite .
the emission radius , luminosity , and peak energy of the photospheric emission and the final bulk lorentz factor and remaining kinetic luminosity of the jet above the photosphere are given by @xcite @xmath83 on the other hand , in the high baryon load case @xmath84 , we have @xmath85 . in this case
most of the initial thermal energy has been converted to the kinetic energy below the photosphere .
the corresponding quantities are given by , for which the radial spreading of shells ( with the initial width @xmath86 ) is significant , and the scaling law is different , @xmath87 @xcite . ]
@xmath88 for typical grb jets with parameters @xmath89 , the photospheric emission can be bright with @xmath90 since the jets can have @xmath91 and @xmath92 in principle .
the grb jet may be thought to consist of many successive shells with initial radial widths @xmath93 .
this implies that the photospheric emission is temporally variable . after emerging from the photospheric regions , collisions between the ejected shells
can produce internal shocks at @xmath94 . as we will show below , the leptons in the internal shock of two given shells can up - scatter their own photospheric emission , which was produced as they emerged from the photosphere .
the up emission can appear as a high - energy spectral component in the lat energy range ( see section [ sec : spectral ] ) .
the temporal properties of the emission in this model can be exemplified using a simple two - shell collision model ( see figure [ fig : is_scat ] ) . for simplicity
, we only consider the cases in which the two shells are both in the regime @xmath84 or both in the regime @xmath81 .
the duration of the observed photospheric emission from one shell is given by the light crossing time of the shell plus the angular spreading time , @xmath95 this can be rewritten as @xmath96 for @xmath84 , and @xmath97 for @xmath81 , where @xmath98 , and @xmath99 is the radius at which the compton drag acceleration ceases .
now we consider a two - shell collision .
we assume that the two shells have similar initial radial width @xmath86 , and the initial separation is @xmath100 .
first consider the high baryon load case @xmath84 .
the rapid shell with lorentz factor @xmath101 catches up with the slower shell with lorentz factor @xmath102 ( hereafter subscripts ` r ' and ` s ' denote the quantities of the rapid and slow shells , respectively ) . setting the zero time in the lab frame as the time when the rapid shell is at @xmath52 , the collision time
is given by @xmath103 \approx 2 c t_v \eta_s^2 $ ] , where we have used an approximation @xmath104 .
the dissipation of the total kinetic energy is most efficient when the masses of the two shells are similar ( c.f .
* ) . for typical cases ,
the ratio of the two lorentz factors may be @xmath105 .
then the dissipation efficiency is given by @xmath106 .
the photons which arrive at @xmath107 at time @xmath108 can be up - scattered by the energetic leptons in the internal shock .
let us define a length @xmath109 , measured from the front end of the rapid shell , as the radial scale of the region from which the photospheric photons do not interact with the leptons in the internal shock .
then we have @xmath110 , which gives @xmath111 we find that inverse compton scattering is efficient for the case @xmath112 and inefficient for the case @xmath113 . a similar argument
can be made also for the low baryon load case @xmath81 . in this case
the shells are accelerated similarly and can not collide with each other below @xmath114 .
the internal shock radius is given by @xmath115 , and we consider a case @xmath116 .
we have @xmath117 .
the second term represents a correction due to the velocity of the shell lower than @xmath118 at @xmath119 .
the second term is rewritten as @xmath120 $ ] , which can not be dominant unless @xmath121 , so that @xmath109 can typically be written as equation ( [ eq : lns ] ) also in this case .
we have two parameter regimes for the scattering efficiency : _ ( i ) efficient scattering regime , _ @xmath112 .
this includes the typical case @xmath122 . in this regime
we may observe bright up emission .
the duration of the photospheric emission can be written as @xmath123 , since we have assumed that @xmath94 for @xmath84 and @xmath124 for @xmath81 .
the onset of the up emission is correlated with that of the photospheric emission pulse released from the rapid shell and delayed from that of the photospheric emission pulse released from the slow shell by @xmath125 .
the duration of the up emission is given by the duration of the seed photons plus the angular spreading time @xmath126 the factor of @xmath127 in the second term means that the up emission has an anisotropic energy distribution in the comoving frame , being brightest at an angle @xmath128 from the line of sight in the observer frame ( c.f .
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , where @xmath129 is the lorentz factor of the merged shell . _ ( ii ) inefficient scattering regime , _ @xmath113 . in this regime , typically we may not observe bright up emission .
a caveat is that this regime includes a case @xmath130 , in which we have a third shell behind the rapid shell of the two given shells apart by @xmath131 , and the photospheric emission from the third shell can be scattered by the leptons in the internal shock of the two given shells .
this condition reduces to @xmath132 , which is satisfied when @xmath133 .
if the grb jet is in the efficient scattering regime right from the start , when it first emerges , the delay timescale between the onsets of the first photospheric emission ( in the mev energy range ) and the first up emission ( in the high - energy range ) is @xmath134 , which is comparable to @xmath135 or @xmath13 .
however , _
fermi _ observations show that the lat onset delays are much larger than the variability timescale apparent in the mev energy light curves both for long and short grbs .
therefore it is unlikely that the large delays of the high - energy emission onsets observed in many grbs are due to the above simple kinematic effect of the jet whose physical parameters do not evolve . in our model , the large delays may be interpreted as the timescale on which the physical parameters of the jet temporally change , e.g. , @xmath86 and @xmath13 change from the inefficient scattering regime into the efficient scattering regime , or @xmath25 , @xmath136 , @xmath137 , and @xmath138 change from the regime @xmath81 into the regime @xmath84 ( see the following sections for details of the latter possibility ) . in these cases
the up emission can start , being still dim compared with the photospheric emission , at a time @xmath139 after the first photospheric emission , and can become bright with the delay as observed .
as shown above , the photospheric emission of the shells can be efficiently up - scattered by the leptons in the internal shocks of the same shells . here
we derive the generic broadband spectrum of the observed radiation arising in the internal shock including synchrotron and ssc emission for the efficient scattering regime . in this case
the kinetic luminosity that can be dissipated into radiation by the internal shock is @xmath140 . at @xmath141 ,
we have @xmath142 , and the pair annihilation timescale is longer than the expansion timescale . thus the pair population freezes out and we can write @xmath143 even well above the photosphere .
hereafter we focus on the photospheric emission from the rapid shell , which is up - scattered by the leptons in the internal shock , so that @xmath137 denotes @xmath144 ( the photospheric emission from the slower shell may be much dimmer ) .
the internal shock radius is estimated as @xmath145 cm .
our assumption @xmath94 reduces to @xmath146 we assume that all the leptons participate in the non - thermal acceleration process in the internal shock and have the injected spectrum @xmath147 for @xmath148 , where @xmath149 is a constant
. then we have @xmath150 and @xmath151 , where @xmath152 is the fraction of the dissipation energy that is carried by leptons in the internal shock .
( we consider a relatively weak internal shock , i.e. , @xmath153 , where we can neglect the change of the lorentz factor of the jet for simplicity . )
these lead to @xmath154 where @xmath155 $ ] . if only a fraction of leptons are accelerated
, @xmath156 is larger than this value .
we also assume that a fraction @xmath157 of the dissipated energy is carried in the form of magnetic fields , @xmath158 .
the characteristic synchrotron energy of leptons with lorentz factor @xmath159 is given by @xmath160 the leptons emit the up emission , synchrotron emission , and ssc emission .
the comoving energy density of the synchrotron emission can be written by @xmath161 where @xmath162 is the dynamical timescale of the internal shock and @xmath163 is the lepton energy distribution averaged over the dynamical timescale .
the quantity @xmath164 is calculated as @xmath165 where @xmath166 is the cooling lorentz factor of the leptons and @xmath167 is the thomson optical depth at @xmath168 ( which can be written as @xmath169 in the case of @xmath84 ) .
we have assumed @xmath170 and defined the function @xmath171 as @xmath172 for @xmath173 , @xmath174 for @xmath175 , @xmath176 for @xmath177 and @xmath178 , and @xmath179 $ ] for @xmath177 and @xmath180 .
similarly we have the ssc emission energy density @xmath181 , where we have assumed that the klein - nishina ( kn ) effect is not significant .
this is valid for the parameters adopted in the case study of observed lat grbs in section [ sec : case ] .
we also have the up emission energy density @xmath182 , where we have assumed that the kn effect is not significant also for the up emission .
this is found to be valid for observed lat grbs in section [ sec : case ] .
we also have neglected the anisotropy of the up photon energy distribution , for simplicity .
this anisotropy leads to the reduction of the observed up luminosity averaged over a pulse than the isotropic assumption by a factor of @xmath183 @xcite , which we will neglect below . by using the relation , e.g. , @xmath184
, we have @xmath185 where we have defined @xmath186 the photospheric luminosity is given by @xmath187 ( see eq .
[ eq : ph_high_baryon ] ) . [ cols="^,^,<,<,^,^",options="header " , ] [ tab : parameters ]
next we consider the _ first _ time - bin spectrum .
if the internal shock for this component produces an electron energy distribution with @xmath188 similar to the second time - bin , the up emission would have a hard photon index @xmath189 in the gev energy range .
thus the up emission has to be very dim compared with the photopheric emission , and the overall spectrum would consist mainly of the photospheric emission
. the luminosity ratio should be @xmath190 .
the internal shock may have @xmath191 similar to the second time - bin , so that we require @xmath92 ( see equations [ eq : x_high ] and [ eq : x_low ] ) .
thus this spectrum corresponds to case 9 ( see table [ tab : cases ] ) .
as shown by the thick solid line in figure [ fig : spec_080916c ] , we can fit the spectrum mainly by the photospheric emission whose luminosity , peak energy , and high - energy spectral index are set to be @xmath192 , @xmath193mev , and @xmath194 , respectively . equations ( [ eq : ph_low_baryon ] ) determine the parameters @xmath25 and @xmath136 by @xmath195 and @xmath60 , @xmath196 and @xmath197 .
the allowed ranges of @xmath137 , @xmath138 , and @xmath13 are broad since the up spectrum is not tightly constrained .
equation ( [ eq : lup_low ] ) means that larger @xmath137 decreases the up luminosity .
the model fit shown in figure [ fig : spec_080916c ] is a case with the maximum contribution of the up emission , i.e. , a case of the minimum @xmath137 for the fixed values @xmath191 and @xmath188 .
this model fit requires the luminosity and characteristic energies of the up emission to be @xmath198 , @xmath199mev , and @xmath200mev , respectively ( corresponding to @xmath201 and @xmath202 ) .
equations ( [ eq : gamma_m ] ) and ( [ eq : gamma_c_low ] ) indicate that we can have similar values of the model parameters @xmath138 and @xmath13 as for the second time - bin , i.e. , we have @xmath203 and @xmath204 for @xmath205 and @xmath206 .
since @xmath207 , we have @xmath172 , so that equation ( [ eq : lup_low ] ) provides @xmath208 .
alternatively , we may consider a case of @xmath209 for the first time - bin .
for the case of @xmath81 , the final lorentz factors of the shells are given by @xmath210 , which may not be much different , i.e. , @xmath211 , since it depends weakly on @xmath25 , @xmath136 , and @xmath138 .
then the internal shocks may only cause a weak dissipation , with @xmath209 .
the observed first time - bin spectrum can be fitted by a photospheric emission only , with @xmath212 in this case .
the model parameters @xmath25 and @xmath136 are constrained to be the same as the above case @xmath191 . for @xmath213 as an example , we can find a set of parameter values @xmath137 , @xmath138 , and @xmath13 similar to those for the second time - bin , which are shown in table [ tab : parameters ] .
the spectral shapes of the other three time - bins are very similar to that of the second time - bin .
they can also be produced by a combination of the photospheric and up components in our model , where the luminosity ratios of the two components are required to be relatively large , i.e. , @xmath214 , which indicate case 1 , similar to the second time - bin .
we can constrain the model parameter values for the three time - bin spectra in a similar way to the second time - bin spectrum . the parameters constrained for the case of the allowed smallest luminosity budgets
are summarized in table [ tab : parameters ] .
the rightmost column represents the value of @xmath215 calculated for the adopted parameter values .
cc the @xmath15 pair - creation break energies under the isotropic photon field assumption is given by @xmath216gev for the first time - bin and @xmath217gev for the other time - bins .
since the target photon field is highly anisotropic , the real break energies may be much larger than those values ( see references listed above equation [ eq : gammagamma ] ) , so that we have neglected the pair - creation breaks in figure [ fig : spec_080916c ] .
we confirm that the second - order up emission is negligible for each time - bin .
for the second to fifth time - bin spectra , we have @xmath218 , so that the kn effect suppresses the second - order up emission . for the first time - bin ,
the second - order up emission flux is just much smaller than the other emission components .
the parameters satisfy our assumptions of @xmath124 for the first time - bin ( eq .
[ eq : cond_low ] ) and @xmath94 for the other time - bins ( eq . [ eq : cond_high ] ) .
we have shown that the parameter regime should shift from @xmath81 into @xmath84 to reproduce the observed first and second time - bin spectra in our model .
this shift is related to the large decrease of @xmath136 from @xmath219 cm in the first time - bin to @xmath220 cm in the second time - bin . at this transition
, the luminosity ratio of the photospheric and up components @xmath221 ( equations [ eq : x_low ] and [ eq : x_high ] ) increases as @xmath136 decreases ( and then @xmath210 increases ) and possibly @xmath222 increases because of the increasing differences of the final lorentz factors of the shells . as we discuss in section [ sec : summary ] , the simulations of the jet dynamics in the progenitor star @xcite suggest that @xmath136 may undergo a sudden decrease as time progresses .
thus this transition may be interpreted as the reason for the observed delay of the lat emission onset .
after the transition , the parameter sets constrained for the allowed smallest luminosities indicate that @xmath25 , @xmath137 , and @xmath13 evolve monotonically , while @xmath136 , @xmath138 , @xmath61 , and @xmath223 are stable ( note that the values of @xmath222 and @xmath224 are given to determine the other parameters ) .
the variability timescale is as small as @xmath225s , which is larger than the light crossing timescale across a schwarzschild black hole of mass @xmath226 , @xmath227s .
such a timescale ( @xmath228ms in the observer - frame ) may not be resolved by the @xmath159-ray detectors that observed this burst as far as we know .
the most sensitive detector among them for the @xmath229mev range is integral , which has a resolution down to @xmath230ms and found a variability timescale @xmath12ms @xcite .
an unsteady mass accretion from a convectively unstable torus onto the central compact object might produce such a large timescale of the flux change @xcite , or the interaction of the jet with the stellar envelope could induce a variability with light crossing timescale from the side to the axis of the jet @xmath231ms , where @xmath232 is the stellar radius and @xmath233 is the opening angle of the jet @xcite .
as seen in figure [ fig : spec_080916c ] , the low - energy portions of the spectra of the photospheric emission are much harder than the observed spectral portions below the peak energies .
this is currently a generic problem for all photospheric mission models , which will be discussed in section [ sec : summary ] .
interestingly , the observed peak energies and peak @xmath69 fluxes in the second to fifth time - bins are clearly consistent with the famous peak energy - luminosity relation ( so - called yonetoku relation ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) , which implies @xmath234 in our model .
this translates into a relation @xmath235 by equations ( [ eq : ph_high_baryon ] ) . considering that @xmath136 and @xmath138 are not much changed in our results and the dependences in the relation are weak
, the dominant relation should be @xmath236 ( or roughly @xmath237 ) .
a degree of fine tuning appears needed to satisfy this relation ; this is similar to other grb emission models , which also require fine tunings of the parameters of the jet ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
furthermore , we should clarify how tightly the parameter values related to the up component are constrained for reproducing the smooth band - like spectrum by a superposition of the two components for each time - bin .
taking the second time - bin as an example , we make different model spectra by fixing @xmath25 , @xmath136 , @xmath238 , and @xmath61 to have the same photospheric component , while changing the other parameters to see whether we obtain a smooth band - like spectrum .
the left model in figure [ fig : spec_dp ] is the result for a value of @xmath13 that is @xmath239 times larger than the value in table [ tab : parameters ] , which is marginally consistent with the data at @xmath240 level .
even larger @xmath13 , which leads to larger @xmath166 ( and then larger @xmath241 and smaller @xmath242 ) , makes the up component dimmer and then the overall spectrum significantly deviates from the data .
we also examine models for smaller @xmath138 , which leads to smaller @xmath166 ( with larger @xmath156 ) .
this case keeps @xmath243 ( and then @xmath244 ) , making the overall model spectrum somewhat bumpy . for a value of @xmath138 that is @xmath239 times smaller than the value in table [ tab : parameters ] , the bumpy model spectrum is marginally consistent with the data at @xmath240 level at @xmath245 but does not reproduce the flux at @xmath246gev because of smaller @xmath241 . even smaller @xmath138 leads to the fast - cooling case @xmath207 .
the spectrum below the up peak energy is harder than the slow - cooling case with @xmath188 , so that the model spectrum is more bumpy , although a higher @xmath247 can explain the emission at the high - energy range .
the right model in figure [ fig : spec_dp ] is the result for a value of @xmath138 that is @xmath248 times smaller than the value in table [ tab : parameters ] , which is still marginally consistent with the data at @xmath240 level .
for a smaller @xmath138 , however , the overall model spectrum significantly deviates from the data in the middle range @xmath249 .
this parametric study indicates that the parameter spaces of @xmath13 and @xmath138 are limited within a factor of @xmath250 a few for obtaining a band - like function by the superposition of the two components .
this means that in order to obtain the band - like spectra through the second to fifth time - bins , we need significant fine tuning of the model parameters @xmath13 and @xmath138 ( as well as @xmath61 and @xmath223 ) . however , we note that @xcite conclude `` compared to the null hypothesis that the data originated from a simple band grb function , adding the additional power - law component resulted in a probability of 1% that there was no additional spectral component for this ( fourth ) time bin . ''
taken at face value , this would imply that the fourth time - bin data may have an additional high - energy spectral component besides the band function although the significance level is low ( @xmath251 level ) .
furthermore , the light curve at @xmath252s in @xcite shows the abrupt steepening break of the gbm light curve at the fifth time - bin while the lat flux decays stably , which implies a two component origin at this time - bin .
@xcite claim that the background uncertainty significantly affect the results of the spectral analysis due to the low count rate at such late times .
these implications increase the plausibility of our two component model .
this burst is a long grb that occurred at a redshift @xmath253 ( which corresponds to @xmath254 cm ) .
the results of the time - resolved spectral analysis by @xcite show that all the spectra can be fitted by a band function plus a distinct power - law function .
they did not list the flux normalizations of the two components and only show an overall spectrum for the second time - bin @xmath255s , which we show by the thin line in figure [ fig : spec_090902b ] .
we apply our model only for this time - bin spectrum .
the lat emission starts @xmath256s after the gbm trigger within the first time - bin @xmath257s .
the delay time in the cosmological rest frame is @xmath258s , which is similar to that of grb 080916c .
the band and distinct high - energy power - law components of the second time - bin spectrum can be straightforwardly modeled by the photospheric and up emission , respectively , as shown by the thick line in figure [ fig : spec_090902b ] .
the observed data of _ fermi_/gbm show a clear deviation from the band function at @xmath259kev , and the thin line is obtained by assuming that the low - energy excess and the high - energy emission at @xmath260mev are the same power - law emission component @xcite . in our model
, however , the low - energy excess could be a contribution from the synchrotron component , which is separate from the high - energy up component .
if the superposition of the multiple photospheric emission from different shells can reproduce the observed low - energy power - law portion of the band component ( see section [ sec : summary ] for more discussion ) and it extends even below @xmath261kev , the contribution from the synchrotron component could explain the observed low - energy excess .
the model fit is obtained by setting the conditions @xmath262 , @xmath263mev , @xmath264 , @xmath265 , @xmath266gev , @xmath267mev , and @xmath268 ( which correspond to @xmath269 and @xmath270 ) .
for these values we have @xmath271 , and @xmath272 . these lead to @xmath84 for a reasonable assumption of @xmath273 , and
thus this spectrum corresponds to case 1 , similar to the second time - bin of grb 080916c . from these conditions and
given values @xmath191 and @xmath274 , we can constrain the model parameter values , in a similar way to that for grb 080916c . a larger @xmath242 ( for a smaller @xmath166 and a larger @xmath156 ) leads to the smaller luminosity @xmath25 , while in order to explain the low - energy excess as the synchrotron emission at @xmath275 with a reasonable range @xmath276 , we require @xmath277 ( for which @xmath278 is needed to explain the high - energy emission ) . in table [ tab : parameters ] , we show the parameter sets constrained for the case of @xmath279 , @xmath280 , and @xmath281 .
they satisfy our assumption of @xmath94 ( eq . [ eq : cond_high ] ) .
the internal shock region is found to be optically thin for @xmath15 pair creation .
the second - order up emission is confirmed to be negligible .
it has been reported that the band component of this burst can be fitted by a multi - temperature blackbody spectrum @xcite .
the observed low - energy spectral slope of the band component @xmath282 is relatively hard and near ( but still softer than ) the single - temperature thermal model @xmath283 , and the high - energy spectral slope @xmath284 is very steep .
these strongly favor a photospheric origin of the band component of this burst .
the first time - bin of grb 090902b includes a very small number of the high - energy photons , which implies larger @xmath215 .
thus the lat onset delay in this burst may be consistent with the interpretation that the parameter regime shifts from @xmath81 in the first time - bin into @xmath84 in the second time - bin , similar to the case of grb 080916c .
this burst is a short grb that occurred at a redshift @xmath285 ( corresponding to @xmath286 cm ) .
this has a precursor @xmath287s earlier than the main burst .
the analysis results of four time - binned spectra of the main burst are given in @xcite .
the first and second time - bin spectra are shown by the thin lines in figure [ fig : spec_090510 ] , and the third time - bin spectrum in figure [ fig : spec_090510_3 ] .
the first time - bin spectrum can be fitted by a band function only , while the second and third time - bin spectra can be fitted by band plus distinct power - law functions . for the last time - bin @xmath288s , the emission was detected only in the lat energy range ( @xmath289 ) ,
whose spectrum is fitted by a power - law function .
we will not fit this last time - bin spectrum with our model but we discuss a possible explanation for it in the context of our model .
the onset delay of the distinct high - energy component with respect to the band component in the cosmological rest frame is estimated to be @xmath290s .
it is remarkable that the distinct high - energy component is brighter than the band component in the third time - bin , which is unique among all the observed lat grbs .
we first consider the _ second _ time - bin spectrum . this may be explained by a combination of the photospheric and up emission components .
the model parameters for this spectrum can be found in a similar way to the case of grb 090902b .
the luminosity ratio is required to be @xmath291 , and we estimate that @xmath292 . these lead to @xmath293 for a reasonable assumption @xmath273 , and
thus this spectrum corresponds to case 1 , similar to the second time - bins of grb 080916c and grb 090902b . the best fit model for this spectrum is shown by the thick dashed line in figure [ fig : spec_090510 ] , for which @xmath294 , @xmath295mev , @xmath296 , @xmath297 , @xmath298gev , @xmath299mev , and @xmath300 are required . from these conditions and given values @xmath191 and @xmath274 , we can constrain the model parameter values in a similar way to the cases for grb 080916c and grb 090902b . the parameter sets constrained for the allowed smallest luminosity ( i.e. , for @xmath301gev , @xmath302mev , which lead to the largest @xmath242 , and @xmath303 )
are shown in table [ tab : parameters ] .
next we consider the _ first _ time - bin spectrum .
this should be produced by the photospheric emission only and the luminosity ratio is required to be @xmath304 . for a reasonable value @xmath191 , we have @xmath81 .
this spectrum corresponds to case 9 ( see table [ tab : cases ] ) .
the model fit shown by the thick solid line in figure [ fig : spec_090510 ] requires that the photospheric emission has @xmath305 , @xmath306mev , and @xmath307 .
equations ( [ eq : ph_low_baryon ] ) determine @xmath308 and @xmath309 .
these lead to a very large @xmath310 , indicating a very large @xmath311 .
the parameters for the model fit shown in figure [ fig : spec_090510 ] are shown in table [ tab : parameters ] .
we can also consider a case of @xmath81 and @xmath209 for the first time - bin , similar to the first time - bin of grb 080916c .
we show a result for a case of @xmath312 in table [ tab : parameters ] , in which @xmath313 and @xmath314 . in this case equation ( [ eq : gamma_m ] )
provides @xmath315 , which means in reality that most of the leptons have @xmath54 and only a small fraction @xmath316 of them participate in the non - thermal power - law acceleration .
the up component spectrum has a bump around @xmath60 , which makes a small contribution to the overall spectrum .
next we find the parameter values appropriate for the third time - bin spectrum . in this time - bin
the up luminosity has to be higher than the photospheric luminosity .
this may correspond to case 2 ( see table [ tab : cases ] ) , unlike the cases considered above .
the spectrum can be fitted as shown by the thick dashed line in figure [ fig : spec_090510_3 ] , where the emission component at @xmath317mev is the ssc emission ( the second - order ssc emission is hidden by the up emission ) .
the characteristic quantities required for this model fit are @xmath318 , @xmath319 , @xmath320 , @xmath321 , @xmath322gev , @xmath323mev , and @xmath324 . for a somewhat larger @xmath325
, we can constrain the model parameter values in a similar way to case 1 , and the results are shown in table [ tab : parameters ] . the value
@xmath326 is much smaller than that in the second time - bin .
this leads to @xmath327 cm , which is too small , compared to a schwarzschild radius @xmath328 cm of the central compact object of mass @xmath329 .
if we take a smaller @xmath330 and @xmath241 which are consistent with the data at @xmath240 level , i.e. , @xmath331 and @xmath332gev , we have @xmath333 cm ( and we have @xmath334 smaller than the larger @xmath330 model ) . at such a late phase of the prompt emission
, the external shock emission might contribute to the high - energy emission @xcite , but this possibility is suggested to be unlikely @xcite ( see also * ? ? ?
below we show another sets of parameter values which could explain the third time - bin emission at @xmath240 level together with the fourth time - bin emission , and have an even larger @xmath136 .
finally we discuss the fourth time - bin spectrum . in this time - bin , only the high - energy photons are detected , which requires much larger @xmath222 and/or much smaller @xmath335 .
these do not seem realistic .
here we show an example of another parameter set for the _ third _ time - bin spectrum , for which the duration of the up emission can be larger than the pulse width of the mev photospheric emission and stays bright even in the fourth time - bin without the mev emission . here
we consider the inefficient scattering regime for the third time - bin ( see section [ sec : temporal ] ; we have assumed the efficient scattering regime for the first and second time - bins , say a case of @xmath336 ) , in which the widths of the two shells @xmath86 are not much different from those for the earlier time - bins @xmath337 , but the separation of the two shells @xmath100 is much larger than those for the earlier time - bins @xmath131 . in this regime
we consider a case of @xmath338 , for which the leptons in the internal shock of the given two shells with @xmath86 and @xmath13 up - scatter the photospheric emission from a third shell at a distance @xmath131 behind these two shells . in this case the angular spreading timescale of the up emission is estimated to be @xmath339 , which is much larger than @xmath340 and @xmath341 , so that we have @xmath342 ( see eq .
[ eq : up_dur ] ) . since @xmath123 , which means that the up emission can last much longer than the mev emission .
if @xmath343 is comparable or larger than the duration of the time - bin , @xmath344s , this scenario could explain the fourth time - bin spectrum .
below we show that the model fit of the third time - bin spectrum in this scenario provides a @xmath13 consistent with the above temporal conditions .
we interpret the observed band and distinct high - energy components in the third time - bin as the photospheric emission of the third shell and the up emission produced by up - scattering of the photospheric emission of the third shell by the internal shock of the first and second shells , respectively .
the model fit is shown by the thick solid line in figure [ fig : spec_090510_3 ] , which is marginally consistent with the data at @xmath240 level .
the photospheric emission of the third shell has @xmath345 , @xmath346mev , and @xmath347 .
( hereafter a tilde denotes a quantity of the third shell . )
the up emission is required to have @xmath348 , @xmath349gev , @xmath350mev , and @xmath351 .
the condition for @xmath330 gives us @xmath334 for a given @xmath325 .
the photospheric flux of the second shell should be lower than the observed flux level .
we take @xmath352 and @xmath353mev which produces the spectral bump below the peak energy of the main photospheric emission in figure [ fig : spec_090510_3 ] and requires @xmath354 and @xmath355 through equation ( [ eq : ph_high_baryon ] ) .
equation ( [ eq : gamma_m ] ) leads to @xmath356 .
then we have @xmath357 and @xmath358 .
finally the remaining parameter @xmath13 is determined by an equation for @xmath166 , which is given by equation ( [ eq : gamma_c_high ] ) multiplied by a factor @xmath359 .
this factor leads to @xmath360 , much larger than in the above simple two - shell scenario .
the observed duration of the up emission pulse is then estimated to be @xmath361s . this is comparable to the durations of the third and fourth time - bins and roughly consistent with the condition @xmath362 , so that the high - energy emission without a corresponding mev emission in the fourth time - bin could be explained by the up emission with large angular spreading time .
the value @xmath363 cm is comparable but still somewhat smaller than a schwartzschild radius of mass @xmath329 .
a larger @xmath136 would lead to a brighter spectral bump of the photospheric emission from the second shell at @xmath364mev in figure [ fig : spec_090510_3 ] , as @xmath365 and @xmath366 in eq .
( [ eq : ph_high_baryon ] ) , violating the observed data .
this might suggest that a large fraction of the jet energy at @xmath335 is not thermal but poynting flux for this burst ( cf . * ? ? ? * ) . to summarize this scenario , the pulse widths of the up component @xmath367 are similar to those of the photospheric component @xmath123 ( as @xmath10ms ) at the first and second time - bins , while at the third time - bin the up pulse width only increases significantly to @xmath368ms , which may stay bright even at the fourth time - bin
. this may be consistent with the results of the cross - correlation function analysis by @xcite that there is no correlated variability between the gbm and lat emission . )
scenario for the third time - bin , since the third time - bin model spectrum indicates that the high - energy tail part of the photospheric emission which has sharp variability contributes to @xmath369 of ( @xmath370 of ) the observed photon numbers at the entire lat range @xmath371mev ( at @xmath372mev ) .
also , the up emission comes from a large area of the shell with solid angle of @xmath373 ( see section [ sec : temporal ] ) , which could have angular inhomogeneity leading to the sharp variability at the both third and fourth time - bins . ] in this scenario there should be a large interval @xmath374s of the photospheric emission before the third time - bin , which seems consistent with the observed quiescent time in the @xmath375mev light curve . such a variability timescale @xmath364s might arise from the mass accretion on the central compact object from an inhomogeneous torus @xcite or from the interaction of the jet with the dense environment @xcite .
the parameter sets we found for the three time - bin spectra are summarized in table [ tab : parameters ] .
we have shown that the parameter regime should shift from @xmath81 in the first time - bin into @xmath84 in the second time - bin , which is related to the large decrease of @xmath136 and corresponds to the increase of the luminosity ratio @xmath221 and the delayed onset of the distinct high - energy component , similar to the cases of grb 080916c and grb 090902b .
the second and third time - bin spectra satisfy the relation @xmath234 , similar to the case of grb 080916c , which requires a fine tuning of the parameters @xmath25 , @xmath136 , @xmath137 , and @xmath138
. the synchrotron and ssc fluxes should be lower than the observed flux levels , which put constraints on @xmath376 , which is much tighter than the cases of grb 080916c and grb 090902b .
in particular for the third time - bin , we need @xmath377 .
a typical value @xmath378 leads to a constraint @xmath379 .
this is not implausible since some external shocks driven by grbs have been suggested to have such small values of @xmath157 @xcite . the internal shock region is estimated to be optically thin for the @xmath15 pair creation for the second and third time - bins , and we find @xmath380gev for the first time - bin .
the parameter sets for all the time - bins satisfy our assumptions of equations ( [ eq : cond_high ] ) and ( [ eq : cond_low ] ) .
the second - order up emission is negligible for all the time - bins .
grb jets are thought to consist of many successive shells moving at relativistic speeds .
these naturally lead to variable photospheric emission around the mev energy range , and also to internal shocks above the photosphere .
energy dissipation near the photosphere caused by e.g. , internal shocks , excited plasma waves , and/or interaction of the jet with the dense environment , or nuclear collisions between protons and neutrons is expected to make the photospheric emission having the power - law tail spectrum above the peak energy @xmath60 , with the photon index @xmath71 under the assumption of the roughly adiabatic evolution of the jet ( see section [ sec : ph ] ) .
we have generically studied the temporal and spectral properties of the radiation from the photosphere and the internal shock in the jet which is accelerated by the thermal pressure .
we have not considered acceleration by processes related to magnetic fields .
we have shown that the photospheric emission is efficiently up - scattered to the high - energy range by the electrons ( and positrons ) accelerated in the internal shocks , for the efficient scattering regime @xmath381 which includes the typical case @xmath122 ( see section [ sec : temporal ] ) , when the radiation from the internal shocks consist of the up , synchrotron , and ssc emission components .
our generic arguments show that the ordering of the luminosities of the emission components depends on the values of @xmath238 , @xmath382 , and @xmath376 ( see section [ sec : spectral ] and table [ tab : cases ] ) . a condition @xmath383 is required in order to have the photospheric and up components dominant in the mev energy range and in the high - energy range , respectively , rather than the synchrotron and ssc components , and @xmath384 as typically observed in lat grbs .
( we can also have a case of @xmath385 for a condition @xmath386^{1/3}$ ] , which has been applied to the third time - bin spectrum of grb 090510 . ) for reasonable values of @xmath222 and @xmath376 both @xmath387 for internal shocks , that condition reduces to @xmath388 ( equation [ eq : eta _ * ] ) . in this case
the photospheric emission has luminosities @xmath389 , and its peak energy is given by @xmath390mev ( equations [ eq : t_a ] , [ eq : ph_low_baryon ] , and [ eq : ph_high_baryon ] ) , which can be consistent with the band components of the observed lat grbs as well as other grbs since we can have @xmath91 in principle ( but there are several issues to be addressed on the spectral shapes and their temporal evolution ; see below ) .
this implies that the grb jets should typically have a relatively large bulk lorentz factor at the prompt emission site , @xmath391 , in our model .
the up emission component is a good candidate for the observed high - energy emission of lat grbs in the prompt phase .
the long - lived high - energy emission after the prompt phase may be related to the external shock ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we have performed case studies of the prompt emission of grb 080916c , grb 090902b , and grb 090510 , and fitted their time - binned spectra by our model typically with reasonable sets of parameter values ( see section [ sec : case ] and table [ tab : parameters ] ) . for grb 090902b and grb 090510
, the up emission corresponds to the observed distinct high - energy component , while for grb 080916c , the up emission is dominant in the high - energy range but the photospheric plus up emission mimic the band function between @xmath7kev and @xmath7gev .
we note that fine tuning of the parameter values appears to be required for some spectra in addition to keeping @xmath92 . in order for the combined photospheric and up components to mimic a simple band function fit of the spectra of grb 080916c through the second to fifth time - bins , significant fine tuning of several parameters is required .
such a composite model is , however , compatible with the presence of an additional high - energy component , which is weakly suggested in the data of the fourth time - bin , and also the difference of the temporal behaviors of the gbm and lat emission suggests a two component origin for the fifth time - bin .
furthermore , the observed spectra of grb 080916c and grb 090510 show the temporal evolution roughly obeying @xmath234 from the second time - bins ( see also * ? ? ?
* ; * ? ? ?
* ) , which require a relation @xmath236 .
for the third and fourth time - bins of grb 090510 , when the distinct high - energy component is much brighter than the band component , we require an additional shell behind the two colliding shells and besides need @xmath363 cm close to or even smaller than a schwartzschild radius of mass @xmath329 , @xmath328 cm , and very small @xmath377 .
other lat grbs , such as grb 080825c @xcite and grb 081024b @xcite , whose redshifts are not determined , display time - binned spectra fitted by band functions with a temporal behavior of the high - energy spectral indices @xmath392 similar to those of grb 080916c , so that they also may be explained by our model . in our model the up emission should exist in all grbs if the jet is in the efficient scattering regime , but for grbs with small @xmath25 this may be below the detection threshold of lat , since the up luminosity is proportional to @xmath25 .
this is a general statement applicable to wide types of models in which @xmath25 determines the flux normalizations of both the band and the high - energy emission components , and consistent with the fact that the high - energy emission is detected by lat only in the brightest class of _ fermi _ grbs @xcite .
we have shown that if the grb jet is in the efficient scattering regime right from the start , the simple kinematic effect causes the onset of the first up emission to be delayed with respect to that of the first photospheric emission by a timescale comparable to the pulse widths or pulse separations of the photospheric emission , which should be smaller than the variability timescale apparent in the mev light curve ( see section [ sec : temporal ] ) .
this may not explain the time delays much larger than the apparent variability timescale observed in most of the lat grbs . in our model , such observed large time delays may instead be attributed to a temporal evolution of the jet parameters , @xmath25 , @xmath136 , @xmath137 , @xmath138 , and @xmath13 , a possible origin for which is discussed in the next paragraph . from the model fits of the time - binned spectra of the three lat grbs ( in section [ sec : case ] ) , we found that the parameter regime should shift from @xmath81 into @xmath84 to reproduce the observed spectra ( this shift leads to the increase of @xmath393 ; equation [ eq : x_high ] and [ eq : x_low ] ) . in this case
the up emission component starts with a small delay with respect to the photospheric emission onset by the timescale comparable to the photospheric variability time , being still dim compared with the photospheric emission while @xmath81 , but starts to be bright and detected by lat when the parameters shift into the regime of @xmath84 .
this parameter shift appears to be related to the large decrease of @xmath136 , from @xmath394 cm to @xmath395 cm , in the timescale of @xmath396s in long grbs grb 080916c and 090902b ( table [ tab : parameters ] ) .
the progenitors of long grbs are thought to be collapsing massive stars , and such a decrease of @xmath136 may be explained by the interaction of the jet with the stellar envelope before the jet breakout .
the numerical simulations by @xcite ( see also * ? ? ?
* ) show that the jet has three parts around the time when it breaks out the star ; the jet head , the collimation - shocked part , and the free expansion part .
the jet head is defined as the very thin part at the front end of the jet between the forward and reverse shocks , which typically has a mildly relativistic speed and the material inside this part escapes sideways into the cocoon ; a large fraction of the jet behind the jet head , suffering weaker dissipation by collimation shocks , has a relativistic speed ; and behind the collimation - shocked part , the jet material suffers much weaker dissipation , expanding adiabatically . while the free expansion part which follows has a saturated lorentz factor of @xmath397 . ] for the collimation - shocked part , @xmath335 may be the stellar radius and @xmath398 is given by its lorentz factor at the breakout time , while for the free expansion part , @xmath335 may be the size of the central compact object and @xmath22 .
in such a jet structure , the emission properties may be changed at the transition between the collimation - shocked part and the free expansion part .
the observed time delay would then correspond to the length scale of the collimation - shocked portion of the jet at the breakout time , @xmath399 cm .
the radius of this part , which should be slightly larger than @xmath400 , say @xmath401 cm , could be consistent with the size of the progenitor star and with our spectral modeling if @xmath402 , which is relativistic but smaller than @xmath403 .
the value @xmath404 cm for the free expansion part is what might be expected from a compact object size @xmath405 cm and @xmath22 .
this @xmath335 is a few schwarzschild radii of an object of mass @xmath406 . as for the short grb 090510
, the parameter shift appears to be related to the decrease of @xmath136 , from @xmath407 cm to @xmath408 cm , on a timescale of @xmath409s ( table [ tab : parameters ] ) .
the model fit indicates an even smaller @xmath363 cm for the third time - bin .
this might suggest a smaller size of the progenitor and the central object than those of long grbs .
in such a scenario , the delay timescale of the lat emission @xmath410 is related to the global properties of the jets and the progenitor systems , rather than the microphysical processes or the simple kinematic ( propagation ) effects , and then it might scale with the total duration of the mev emission @xmath411 .
however , in our specific model , while @xmath412 scales with the stellar radius @xmath232 ( where @xmath413 is the fraction of the collimation - shocked part at the jet breakout ) , @xmath411 may depend on the size of the central region of the star @xmath414 from which the accretion rate is very high , i.e. , @xmath415 , where @xmath23 is the mass enclosed within @xmath414 ( cf .
then @xmath411 may significantly depend on the angular velocity and density profiles of the star , which could not simply scale with @xmath410 .
observations of lat grbs with redshifts determined show that grb 080916c , grb 090902b , and grb 090926a @xcite have the durations @xmath416s , @xmath417s , and @xmath418s , respectively , while the delay timescales are all @xmath396s . in our scenario
@xmath410 is expected to be distributed more widely when we obtain a larger number of lat bursts.s ( while @xmath419s * ? ? ?
* ) . the earliest lat emission may be just the detections of the high - energy tail part of the photospheric emission while the up emission onset could correspond to the second group of the lat detections at @xmath372mev , @xmath420s after the burst trigger . ]
the shift from @xmath81 into @xmath84 makes the change of the photospheric luminosity from @xmath421 into @xmath422 , so that it would decrease if @xmath25 was roughly constant .
the observed mev emission luminosities appear not to decrease ( roughly constant for grb 080916c and grb 090510 while increase by about a factor of two for grb 090902b ) , however , so that @xmath25 should increase compensate the shift of the parameter regime of @xmath215 .
table [ tab : parameters ] shows that @xmath25 increases from the first to second time - bins by factors of @xmath127 and @xmath423 for grb 080916c and grb 090510 , respectively , while for grb 090902b , the increase factor may be estimated to be @xmath424 .
this may be another parameter tuning required in our model for lat grbs , but it might be a selection effect : grb jets with non - increasing @xmath25 could have dim up emission which is not detected by lat . for bursts whose lat emission stays bright from @xmath410 after the trigger ,
as seen in the three bursts we have studied in detail , the parameter regime is assumed to stay @xmath84 in a large fraction of the total duration in our model .
this means that the radiation efficiencies in those bursts are not so extremely large compared with the photospheric emission models in which @xmath92 is assumed over the total duration .
indeed we can roughly estimate the radiation efficiencies of the three bursts as @xmath425 for grb 080916c , @xmath426 for grb 090902b , and @xmath427 for grb 090510 .
such moderately high efficiencies of prompt emission lead to the external shock emission as bright as the prompt emission , which is consistent with the observations that all the three bursts have bright long - lived high - energy emission . for grb 090902b
, its late - time low - energy afterglow suggests the total isotropic kinetic energy after the prompt phase @xmath428erg @xcite , which is much smaller than estimated from the high - energy afterglow @xmath429erg @xcite .
this might be attributed to the two - components structure of the jet involving a narrow and bright spot at the line of sight surrounded by a wider and less energetic region @xcite .
many early x - ray afterglows observed by _ swift _
show very steep decays @xmath430 , which are not compatible with the bright high - energy afterglows of lat grbs , not decaying so rapidly .
this is another issue to be solved , and a simultaneous observation by _ fermi_/lat and _ swift_/xrt from the early times will be very helpful .
the above summary indicates that the photospheric emission models may be viable for the temporal and spectral properties of the mev and high - energy emission of the _ fermi_/lat grbs as well as the other ordinary grbs , although we need the fine tuning of several parameters ( see also the following discussion ) .
very recently @xcite showed detailed simulations of the emission processes at a dissipation region out of the photosphere of the jet in a model similar to ours , including photospheric emission , and concluded that such an emission model can explain the spectrum of grb 090902b , supporting our rough analytical model .
@xcite explores the emission from the photosphere and internal shock of the jet with a very high @xmath431 and argues that the high - energy emission of the lat grbs may be explained as synchrotron emission from the internal shock . an outstanding problem in the photospheric emission models , including
ours is that the low - energy spectrum of the photospheric emission , usually assumed to be the rayleigh - jeans part of the blackbody radiation , @xmath283 , is much harder than those of the band components of typical observed grbs widely distributed around @xmath432 @xcite as well as those of lat grbs shown in figure [ fig : spec_080916c ] , [ fig : spec_090902b ] , [ fig : spec_090510 ] , and [ fig : spec_090510_3 ] . in this paper
we have mainly considered the two - shell system like in figure [ fig : is_scat ] , but the typical duration for the spectral analysis includes many pulses ( note that our model parameters and the analysis time - bins are @xmath433s and @xmath434s for grb 080916c , @xmath435s and @xmath436s for grb 090902b , and @xmath437s and @xmath438s for grb 090510 ) , and thus the superposition of the photospheric emission from the multiple shells with different @xmath60 has the potential of reproducing the observed low - energy spectrum @xmath439 . with @xmath440 ) .
the radio synchrotron emission is thought to be highly self - absorbed ( @xmath441 ) , and the emission from the multiple shells can reproduce the observed spectral slope ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
] for example , the photospheric emissions from multiple shells with different @xmath137 and similar @xmath25 , @xmath136 , and @xmath138 have a relation @xmath442 ( not the conventional blackbody relation @xmath443 ; see equation [ eq : ph_high_baryon ] ) , so that their superposition appears as a spectrum @xmath444 , i.e. , @xmath445 .
is required in order to reproduce the observed correlation @xmath234 in the time - binned spectra of grb 080916c and grb 090510 ( see section [ sec : case ] ) , which should apply for the maximum value of @xmath137 in some range of times . ]
the parameter @xmath137 should be distributed over a range of a factor @xmath446 to reproduce the low - energy spectrum extending @xmath447 decades of energies @xcite .
however this is just a rough speculation , and we need more detailed consistency checks within the model and between the model and observations ( see also * ? ? ?
the spectral analysis of grbs with @xmath448s could reveal the intrinsic hard spectrum .
our speculation is not inconsistent with the recent analysis that the spectrum of as small time - bin as @xmath449 in grb 090902b becomes close to a blackbody , while the spectra of smaller time - bins but @xmath450 in grb 080916c keep non - thermal @xcite .
the synchrotron ( and ssc ) emission models also have a problem with the low - energy spectral index of the band component . the electrons are required to be cooled so fast that the low - energy spectral slope should be @xmath451 in simple types of these models ( e.g. , * ? ? ? * ) .
this is usually softer than the observed slopes , so that the superposition effect does not work well .
there have been many models considering the details of the microphysics in the emission sites , but there is no consensus for this problem yet ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. the spectral excesses from the band component below @xmath261kev in grb 090902b could be one of the important discoveries by _ fermi_/lat , although such an excess was not detected by other detectors extending to sufficiently low energies such as _ bepposax _ and _ hete-2 _ ( e.g. , * ? ? ?
* ) ( but see * ? ? ?
if these are real detections , they may be explained in our model as synchrotron or ssc emission from internal shocks , constraining the values of @xmath452 ( see table [ tab : parameters ] ) . on the other hand , in order for the synchrotron emission not to be prominent in the low - energy range in grb 090510 , we require @xmath453 and @xmath454 for the second and third time - bins , respectively .
such a wide range of constrained parameter @xmath157 is similar to the case in the external shock , @xmath455 , constrained from the late afterglow observations @xcite , but it is one of the fundamental problems for the shock physics to deduce the microphysical parameters @xmath157 as well as @xmath152 and @xmath223 from the first principles . the light curves dominated by the synchrotron or ssc emission should be correlated with those dominated by the up emission , since they both are produced by the same internal shocks
. such a correlation analysis would be useful to distinguish our photospheric emission model from , e.g. , other eic models @xcite .
if the high - energy emission is produced by up - scattering of soft x - ray emission from an external source , such as a cocoon ejected from a progenitor star , off the electrons in the jet , the light curve of the seed soft x - rays should not be correlated with that of the high - energy component .
we note that the kinematic arguments in section [ sec : temporal ] indicate that a fraction of the photospheric emission should also be temporally correlated with the synchrotron , ssc , and up emission in our model .
the synchrotron component could also be detected in the optical band .
for the model parameters for the second time - bin of grb 090902b shown in table [ tab : parameters ] , the synchrotron self - absorption energy in the observer frame is estimated as @xmath456ev and then we have a flux @xmath457mjy at @xmath458hz , which is a very bright optical source even though it suffers a strong self - absorption .
the bright optical prompt emission of some grbs , being often much brighter than the extrapolation of the band component ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , could be attributed to the synchrotron emission from the internal shocks in our model .
the spectra of the three lat grbs we considered may have spectral breaks at @xmath459 and at @xmath460 in the high - energy range , while a break due to the @xmath15 pair creation in the emission site is estimated to be typically much above those breaks in our model . detecting a spectral break in the high - energy range by _
fermi_/lat and/or by the future cherenkov telescope such as cta would be very helpful to constrain the models .
polarimetric observations of the prompt emission also have the potential of distinguishing the photospheric emission models from the synchrotron emission models @xcite . in the energy range @xmath461mev , which is the target of some planned missions , the photospheric emission which consists of the blackbody component and the multiple - scattered blackbody component may have very low polarization , while some of the synchrotron emission models predict detectable degrees of polarization ( see also * ? ? ?
* for general discussion of grb polarization ) .
the case studies in this paper have shown that the photosphere - internal shock model can produce three types of the spectral shapes of grb prompt emission : ( i ) the photospheric emission with the high - energy tail of the photon index @xmath71 makes the band spectrum up to @xmath462gev while a dim up emission has a small contribution or just makes the extension of that band spectrum at @xmath463gev ( like the first time - bins of grb 080916c and grb 090510 ) ; ( ii ) the up emission is bright but the combination of the up and photospheric components mimics a band function with the high - energy index of @xmath464 ( like the second to fifth time - bins of grb 080916c ) ; ( iii ) the up emission is bright and makes a high - energy component distinct from a photospheric emission ( like the second time - bin of grb 090902b and the second and third time - bins of grb 090510 ) .
the increasing number of bursts being observed in the lat field of view ( including bursts with no detections in the lat energy range ) will determine the number ratios of the band - only spectra and the band plus distinct high - energy component spectra .
extensive parametric studies would then be able to clarify whether our model with reasonably large parameter space is consistent with the high - energy parts of general grb spectra or only consistent with limited numbers of them .
a very recent analysis of 52 bright _
fermi_/gbm bursts ( @xmath465mev ) show that the number of the bursts that can be fitted by a band function with @xmath466 is 14 ( @xmath467 ) and with @xmath468 is 7 ( @xmath469 ) @xcite .
this implies that the fraction of bursts with the spectral type ( ii ) is small , and thus we might not require significant fine tuning of parameters for general grbs.mev ) show that many spectra are adequately fitted with a cutoff power - law function as well as a band function @xcite , so that they are not suitable for constraining the @xmath392 values and our model . ]
we have focused on baryonic jets which evolve roughly adiabatically and have applied a simple analytical formulation of the radiation from the photosphere and internal shock to deduce the physical parameters of the jets for each time - bin .
the fine tunings of several parameters appear to be required for the three bursts we have studied , which might suggest other possibilities , e.g. , non - adiabatic ( i.e. , significantly dissipative ) jets @xcite , or magnetically - dominated jets ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
even in those other jet models , however , the photospheric emission , its up - scattering effect in the internal shocks at large radii , and the change of the radiation properties due to the interaction of the jet with the dense environment could be important for some cases . thus , the time - binned spectra and delay timescales of the lat emission onsets could have the potential of revealing the radial structure of grb jets .
the difference in the delay timescales of long and short grbs may provide clues for understanding the differences between the jets themselves as well as their progenitors .
we thank the _ fermi _ lat / gbm group members , b. b. zhang , and t. sakamoto for useful discussions .
we also thank the anonymous referee for comments that significantly improved the paper .
we acknowledge nasa nnx09at72 g , nasa nnx08al40 g , and nsf phy-0757155 for partial support .
xfw was supported by the national natural science foundation of china ( grants 10633040 and 10921063 ) , national basic research program of china ( 973 program 2009cb824800 ) , and the special foundation for the authors of national excellent doctorial dissertations of p. r. china by chinese academy of sciences .
pm is grateful for the hospitality of the institute for advanced study , princeton , during part of this project .
99 abdo , a. a. , et al .
2009a , apj , 707 , 580 abdo , a. a. , et al .
2009b , science , 323 , 1688 abdo , a. a. , et al . 2009c , apj , 706 , l138 abdo , a. a. , et al .
2010 , apj , 712 , 558 abramowicz , m. a. , novikov , i. d. , & paczynski , b. 1991 , apj , 369 , 175 ackermann , m. , et al .
2010a , apj , 716 , 1178 ackermann , m. , et al . 2010b , apj , 717 , l127 ackermann , m. , et al .
2011 , apj in press ( arxiv:1101.2082 ) asano , k. , guiriec , s. , & mszros , p. 2009a
, apj , 705 , l191 asano , k. , inoue , s. , & mszros , p. 2009b
, apj , 699 , 953 asano , k. , & terasawa , t. 2009 , apj , 705 , 1714 band , d. l. , et al .
1993 , apj , 413 , 281 beloborodov , a. m. 2005 , apj , 618 , l13 beloborodov , a. m. 2010 mnras , 407 , 1033 bissaldi , e. , et al .
2011 , submitted ( arxiv:1101.3325 ) bttcher , m. , & dermer , c. d. 1998 , apj , 499 , l131 briggs , m. s. , et al . 1999 ,
apj , 524 , 82 brunetti , g. 2000 , astropart .
phys . , 13 , 107 cenko , s. b. , et al .
2010 , submitted ( arxiv:1004.2900 ) corsi , a. , guetta , d. , & piro , l. 2010 , apj 720 , 1008 daigne , f. , bosnjak , z. , & dubus , g. 2011 , a&a , 526 , 110 daigne , f. , & mochkovitch , r. 1998 , mnras , 296 , 275 dermer , c. , & schlickeiser , r. 1993 , apj , 416 , 458 de pasquale , m. , et al . 2010 , apj , 709 , l146 derishev , e. v. , kocharovsky , v. v. , & kocharovsky , vi .
v. 2001 , a&a , 372 , 1071 eichler , d. , & levinson , a. 2000 , apj , 529 , 146 fan , y. z. , piran , t. , narayan , r. , & wei , d. m. 2008 , mnras , 384 , 1483 fan , y. z. 2009 , mnras , 397 , 1539 fan , y. z. 2010 , mnras , 403 , 483 gao , w. h. , mao , j. r. , xu , d. , & fan , y. z. 2009 , apj , 706 , l33 ghirlanda , g. , celloti , a. , & ghisellini , g. 2002 , a&a , 393 , 409 ghirlanda , g. , et al .
2005 , mnras , 360 , l45 ghirlanda , g. , bosnjak , z. , ghisellini , g. , tavecchio , f. , & firmani , c. 2007 , mnras , 379 , 73 ghirlanda , g. , nava , l. , & ghisellini , g. 2010 , a&a , 511 , 43 ghisellini , g. , & celloti , a. 1999 , apj , 511 , l93 ghisellini , g. , celloti , a. , & lazzati , d. 2000 , mnras , 313 , l1 ghisellini , g. , & tavecchio , f. 2009 , mnras , 397 , 985 ghisellini , g. , ghirlanda g. , nava , l. , & celloti , a. 2010 , mnras , 403 , 926 giannios , d. 2006 , a&a , 457 , 763 goodman , j. 1986 , apj , 308 , l47 granot , j. 2010 , in the proceedings of `` the shocking universe - gamma - ray bursts and high energy shock phenomena '' , venice , september 14 - 18 , 2009 ( arxiv:1003.2452 ) granot , j. , cohen - tanugi , j. , & do couto e silva , e. 2008 , apj , 677 , 92 granot , j. , & guetta , d. 2003 , apj , 2003 , 598 , l11 granot , j. , komissarov , s. s. , & spitkovsky , a. 2011 , mnras , 411 , 1323 greiner , j. , et al .
2009 , a&a , 498 , 89 gruzinov , a. , & mszros , p. 2000 , apj , 539 , l21 he , h .- n . ,
wu , x .- f . , toma , k. , wang , x .- y . , & mszros , p. 2010 , submitted ( arxiv:1009.4716 ) ioka , k. 2010 , prog .
, 124 , 667 ioka , k. , murase , k. , toma , k. , nagataki , s. , & nakamura , t. 2007 , apj , 670 , l77 kaneko , y. , et al .
2006 , apjs , 166 , 298 kobayashi s. , piran t. , & sari r. 1997 , apj , 490 , 92 knigl , a. 1981 , apj , 243 , 700 kumar , p. , & barniol duran , r. 2009 , mnras , 400 , l75 kumar , p. , & barniol duran , r. 2010 , mnras , 409 , 226 kumar , p. , & mcmahon , e. 2008 , mnras , 384 , 33 kumar , p. , narayan , r. , & johnson , j. l. 2008 , mnras , 388 , 1729 lazzati , d. , & begelman , m. c. 2010 , 725 , 1137 lazzati , d. , morsony , b. , & begelman , m. c. 2009 , apj , 700 , l47 li , z. 2010 , apj , 709 , 525 lithwick , y. , & sari , r. 2001 , apj , 555 , 540 liu , r .- y . , & wang , x .- y .
2010 , apj in press ( arxiv:1009.1289 ) lyutikov , m. 2006 , new journal of physics , 8 , 119 marscher , a. 2009 , in lecture notes in physics 794 , the jet paradigm from microquasars to quasars , ed .
t. belloni ( arxiv:0909.2576 ) medvedev , m. v. 2000 , apj , 540 , 704 mszros , p. 2006 , rep .
, 69 , 2259 mszros , p. , laguna , p. , & rees , m. j. 1993 , apj , 415 , 181 mszros , p .. & rees , m. j. 1997 , apj , 476 , 232 mszros , p. , & rees , m. j. 2000 , apj , 530 , 292 mizuta , a. , nagataki , s. , & aoi , j. 2010 , submitted ( arxiv:1006.2440 ) morsony , b. j. , lazzati , d. , & begelman , m. c. 2007 , apj , 665 , 569 morsony , b. j. , lazzati , d. , & begelman , m. c. 2010 , apj , 723 , 267 murase , k. , toma , k. , yamazaki , r. , nagataki , s. , & ioka , k. 2010a , mnras , 402 , l54 murase , k. , toma , k. , yamazaki , r. , & mszros , p. 2010b
, submitted ( arxiv:1011.0988 ) nakar , e. , piran , t. , & sari , r. 2005 , apj , 635 , 516 paczynski , b. 1986 , apj , 308 , l43 paczynski , b. 1990 , apj , 363 , 218 panaitescu a. , & kumar p. 2002 , apj , 571 , 779 panaitescu , a. , & mszros , p. 2000 , apj , 544 , l17 peer , a. , mszros , p. , & rees , m. j. 2005 , apj , 635 , 476 peer , a. , zhang , b. b. , ryde , f. , et al .
2010 , submitted ( arxiv:1007.2228 ) peer , a. , & waxman , e. 2004 , apj , 603 , l1 peer , a. , & zhang , b. 2006 , apj , 653 , 454 piran , t. 2004 , rev .
76 , 1143 preece , r. d. , et al . 2000 , apjs , 126 , 19 racusin , j. l. , et al . 2008 , nature , 455 , 183 razzaque , s. , dermer , c. , & finke , j. d. 2010 , open astronomy journal , 3 , 150 rees , m. j. , & mszros , p. 1994
, apj , 430 , l93 rees , m. j. , & mszros , p. 2005 , apj , 628 , 847 rosswog , s. 2007 , mnras , 376 , l48 ryde , f. , et al .
2006 , apj , 652 , 1400 ryde , f. , et al .
2010 , apj , 709 , l172 sari , r. , & esin , a. a. 2001 , apj , 548 , 787 sekiguchi , y. , & shibata , m. 2010 , submitted ( arxiv:1009.5303 ) shemi , a. , & piran , t. 1990 , apj , 365 , l55 sikola , m. , begelman , m. c. , & rees , m. j. 1994 , apj , 421 , 153 spruit , h. c. , daigne , f. , & drenkhahn , g. , 2001 , a&a , 369 , 694 thompson , c. 1994 , mnras , 270 , 480 thompson , c. , mszros , p. , & rees , m. j. 2007 , apj , 666 , 1012 toma , k. , sakamoto , t. , zhang , b. , et al . 2009a , apj , 698 , 1042 toma , k. , wu , x. f. , & mszros , p. 2009b
, apj , 707 , 1404 vietri , m. 1997 , phys .
lett . , 78 , 4328 wang , x. y. , & mszros , p. 2006
, apj , 643 , l95 wang , x. y. , li , z. , dai , z. g. , & mszros , p. 2009
, apj , 698 , l98 yonetoku , d. , et al .
2004 , apj , 609 , 935 zhang , b. 2007 , chin . j. astron .
astrophys . ,
7 , 1 zhang , b. , & peer , a. 2009 , apj , 700 , l65 zhang , b. , & yan , h. 2011 , apj , 720 , 90 zhang , b. b. , et al .
2010 , apj in press ( arxiv:1009.3338 ) zou , y .- c . , fan , y .- z . , &
piran , t. 2011 , apj , 726 , l2 | radially inhomogeneous gamma - ray burst ( grb ) jets release variable photospheric emission and can have internal shocks occurring above the photosphere .
we generically formulate a photospheric emission model of grbs including compton up - scattered photospheric ( up ) emission off the electrons ( and positrons ) in the internal shocks , and find that the photospheric emission may correspond to the traditional ( band ) component at @xmath0mev and the up emission to the high - energy emission observed by _
fermi_/lat for some grbs at @xmath1mev .
the two components can be separate in the spectrum in some cases or can mimic a smooth broad band spectrum in other cases .
we apply our formulation to the well - studied long and short lat grbs , grb 080916c , grb 090902b , and grb 090510 , and typically find reasonable parameters for fitting the time - binned spectra , although fine tuning of several parameters is required .
the observed delays of the high - energy emission with respect to the mev emission which are large compared to the variability times are unlikely to be due to simple kinematic effects of a non - evolving jet .
these delays may instead be attributed to the temporal evolution of the physical parameters of the jet , and thus the delay timescales could provide a potential tool for investigating the structures of grb jets themselves and their progenitors .
the difference of the delay timescales of long and short grbs inferred from the _ fermi _ data might be due to the differences in the progenitors of long and short grbs .
some other properties and consequences of this model are discussed , including temporal correlations among the prompt optical , the soft x - ray , and the distinct high - energy component as well as the band component .
[ firstpage ] gamma ray : bursts ; radiation mechanisms : non - thermal |
the structure of a circumstellar disk is expected to dramatically change during the planet formation process . in a few million years
an optically thick massive protoplanetary disk will transform into an optically thin debris disk with little material ( e.g. * ? ? ? * ) .
processes like viscous accretion , photo evaporation , grain growth , and planet formation shape the structure of circumstellar disks , and in particular , the process of planet formation creates gaps , cavities , and asymmetries that can be directly observed @xcite .
circumstellar disks with gaps and/or cavities lack warm dust close to the star while still possessing a significant reservoir of cold dust in the outer disk .
such disks are classified as transitional disks , and plenty is known about their inner disk gaps / cavities .
however , it is only recently that observations have reached the sensitivity and angular resolution required to study asymmetrical features in the outer disk structure .
particularly , the only known examples of high contrast asymmetries in transitional disks at submillimeter wavelengths are hd 142527 @xcite , oph irs 48 @xcite , and lkh@xmath2 330 @xcite . to understand the process of planet formation , long wavelength observations are critical , since they are less affected by optical depth , and hence trace the bulk of the mass surface density . in this letter
, we present alma observations at 0.45 mm of two disks surrounding the stars sao 206462 and sr 21 , which exhibit large - scale asymmetries in their dust continuum emission that were not known before with high significance .
sao 206462 is an isolated herbig ae / be star ( f4ve , @xmath8 ) , at a distance of 142 pc in the sco - ob2 association .
the disk surrounding sao 206462 is classified as a transitional disk , with a large dust - depleted cavity ( @xmath9 au ) and a massive outer disk @xcite .
recently , scattered light observations of the polarized intensity at near - ir wavelengths revealed spiral - like structure in the outer disk of sao 206462 .
sr 21 is a young g3 star ( @xmath10 ) in the @xmath11-ophiuchus star forming region at a distance of 119 pc . the dust - depleted cavity ( @xmath12 au , * ?
* ) in this transition disk @xcite is not completely devoid of material , as rovibrational co line observations reveal the presence of inner disk gas @xcite .
furthermore , recent polarized intensity h - band observations of sr 21 by @xcite display emission inside the cavity radius , and unlike sao 206462 , the radial profile of emission is quite smooth .
[ cols="<,^,^,^,^,^,>,^,^,^ , > " , ]
we present maps of the dust and gas emission ( @xmath0co @xmath1 ) in figure 1 .
these observations resolve the structure of each transitional disk , revealing for the first time the striking morphology in the dust continuum emission for sao 206462 and sr 21 : a non - uniform ring with a bright asymmetry located in the south - west for sao 206462 and in the south for sr 21 . the integrated intensity map ( moment 0 ) and the intensity - weighted mean velocity map ( moment 1 ) of @xmath0co @xmath1 are presented in the middle and right panels of figure 1 .
in contrast with the dust emission , the gaseous component from the @xmath0co moment 0 map appears to be quite compact and mostly symmetric . however , significant emission ( @xmath13 ) can be found at large radii in the @xmath0co moment 1 map , just as with the dust emission .
since emission from @xmath0co is most likely optically thick , it is not a good tracer of depletion inside the cavity and other tracers must be used . however , a large depletion of the inner disk can give rise to the double - peaked structure observed in the sao 206462 moment 0 map .
further analysis of these co observations will be deferred to a future publication .
we measured the disk position angle ( pa ) from the rotation of the disk in the @xmath0co moment 1 map ; the measured pa east of north is @xmath14 and @xmath15 for sao 206462 and sr 21 , respectively .
our derived pa is consistent with measurements found in the literature . for sao 206462 measurements from molecular gas ( inside and outside the cavity ) and
polarization give a pa between @xmath16-@xmath17 @xcite .
for sr 21 , @xcite constrained its pa to @xmath18 from rovibrational co band observations of molecular gas inside the cavity , while @xcite inferred a pa of @xmath19 from dust continuum observations of the outer disk .
the contrast between the peak of emission and the opposite side of the disk is @xmath20 in sao 206462 and @xmath21 in sr 21 .
these asymmetries have lower contrast values than those found in lkh@xmath2 330 ( @xmath22 , * ? ? ?
* ) , hd 142527 ( @xmath23 , * ? ? ?
* ; * ? ? ?
* ) , and oph irs 48 ( * ? ? ?
* @xmath24 , ) , nevertheless they are significant enough to warrant further analysis .
the @xmath25 m observations from @xcite constrained the surface density at the edge of the cavity to be @xmath26 for both sao 206462 and sr 21 . assuming a dust opacity of @xmath27 at @xmath28 m @xcite , we find that the 0.45 mm optical depth of dust is @xmath29 at the peak of the emission from the ring , and it is expected that at larger radii from the disk @xmath30 should be lower .
we also computed the brightness temperature ( @xmath31 ) of the emission without the raleigh - jeans approximation .
the left panels of figure 1 show contours of @xmath31 , starting at 10 k and spaced every 5 k. at the peak of emission , @xmath32 k for sao 206462 , and @xmath33 k for sr 21 . at the distance where the asymmetry is located ( @xmath34 au for sao 206462 and@xmath35 au for sr 21 ) ,
a temperature of 30 - 35 k is a reasonable value for the dust temperature in a flared disk , consistent with @xmath36 derived above .
we expect then that these alma observations trace some of the mass surface density , since the emission is borderline between optically thin and optically thick emission .
to constrain the structure of these large - scale asymmetries , we compare the dust continuum observations of sao 206462 and sr 21 with simple morphological models that describe the emission from a ring and a vortex .
these morphological models were chosen since these are known transitional disks that can be well - described by a ring - like structure @xcite , and theoretical studies suggest that asymmetries in disk emission may arise for vortices ( e.g. , * ? ? ? * ) .
for the ring morphology we use a gaussian prescription in the radial direction : in polar coordinates . ]
the ring emission corresponds to @xmath37 , where @xmath38 is the radius where the ring peaks , @xmath39 is the flux density at @xmath38 , and @xmath40 is the ring width . in our modeling the ring emission
is assumed to be axisymmetric .
for the vortex morphology , we employ the prescription presented in @xcite , where the mass - surface density of a steady - state vortex can be described as a gaussian in the radial and azimuthal directions : @xmath41 , where @xmath42 and @xmath43 are the radius and pa of the vortex s peak , @xmath44 is the flux density at @xmath45 , and @xmath46 are the radial and azimuthal width of the vortex .
this way , the structure of a disk can be described by a ring of emission ( 3 parameters ) , a vortex ( 5 parameters ) , or the combination of both ( 8 parameters ) .
we fix the disk pa to the value measured in section 3 , and setup models of a ring and a ring with a vortex , leaving the disk center ( @xmath47 ) and inclination ( @xmath48 ) as free parameters .
each model realization is fourier transformed and sampled at the same _ uv_-positions as the observation .
the best - fit model was found by minimizing the @xmath49 statistic using the latest implementation of the affine - invariant mcmc _ emcee _
software @xcite .
table 2 presents the best - fit parameters found for a ring model and a ring+vortex model .
as expected , the @xmath49 analysis favors a ring with a vortex model , since a ring alone can not reproduce the large asymmetrical structures of these alma observations .
we note that for sr 21 , our preliminary ring+vortex model fitting favored a face - on geometry ( @xmath50 ) .
nevertheless , gas tracers show disk rotation making such geometry unlikely .
thus , we fixed the inclination to @xmath51 for sr 21 ring+vortex modeling , consistent with the @xmath0co @xmath1 line shape and motivated by the result of our ring - only model .
the best - fit model and residual maps for the two morphologies explored are presented in figures 2 and 3 for sao 206462 and sr 21 , respectively .
these figures compare alma observations ( top row , leftmost panel ) with the best - fit model and residual maps ( middle and right panels ) , where the residual emission is found by subtracting the model from the observations .
when considering only a ring , the best - fit model corresponds to the average flux of the disk over all azimuthal angles .
hence large regions of negative emission are expected since the azimuthally averaged flux also encompasses flux from the large - scale asymmetry .
the addition of a vortex prescription in the fitting improves the resulting residual map and @xmath49 of the fit , although significant emission ( @xmath52 ) remains .
the cavity radius ( @xmath53 ) found in the literature for these disks has some scatter , on average @xmath54 au for sao 206462 and @xmath55 au for sr 21 @xcite .
we find that @xmath53 from these studies is consistent with our constraints within @xmath56 for all but @xcite observations of sao 206462 , which have lower angular resolution than these alma observations .
the analysis presented above demonstrates that a ring of emission does not describe the disk structure of sao 206462 and sr 21 ; the addition of a vortex prescription better describes the large - scale asymmetries .
however , the observed asymmetries might not directly trace the mass surface density , since short - wavelength observations may not be completely optically thin .
future long - wavelength observations will confirm that these asymmetries correspond to mass rather than temperature variations .
the vortex encompasses 15% of the total flux density for sao 206462 and 28% for sr 21 .
these are quite similar to the recently studied disk around lkh@xmath2 330 ( 30% ) but much less dramatic than the oph irs 48 and hd 142527 disks , where a horseshoe shape , rather than a ring , is observed at millimeter wavelengths .
the flux density in the asymmetry corresponds to a mass ( dust+gas ) of @xmath57 for both sao 206462 and sr 21 , assuming a temperature of @xmath58 , a gas - to - dust ratio of 100 , and a dust opacity of @xmath27 at @xmath28 m @xcite .
these are lower limits on the vortex mass , given the uncertainty in optical depth at this wavelength .
compared to recent examples of large - scale asymmetries observed in lkh@xmath2 330 and oph irs 48 , the radial extent of the vortices in sao 206462 and sr 21 is also quite narrow , with a fwhm of few tens of au ( see table 2 ) .
azimuthally , the vortices are well resolved , and while the sr 21 asymmetry has a fwhm@xmath59 , similar to lkh@xmath2 330 and oph irs 48 , the azimuthal extent in sao 206462 is 30% larger , covering almost half of the disk . from our modeling
we find that the vortex aspect ratio is @xmath60 for sao 206462 and @xmath61 for sr 21 .
following the solution presented by @xcite , a steady - state vortex tends to concentrate large grains toward its center , in an amount that depends on the ratio between the dust grain stokes number ( @xmath62 ) and the gas turbulence velocity ( see also birnstiel et al .
our measurements of the vortex structure and knowledge of the stokes number allow us to constrain the gas turbulence .
dust grains with size @xmath63 mm and internal density @xmath64 g @xmath65 , embedded in a disk with a surface density of @xmath66 g @xmath67 have a stoke number @xmath68 .
the vortex aspect ratio derived for sao 206462 and sr 21 would then require turbulent velocities of about @xmath69 and @xmath70 respectively , corresponding to about 22% and 16% of the local isothermal sound speed for a gas temperature of 50 k. such values are about 5 times larger that those derived for oph irs 48 , and consistent with an active disk layer due to the magnetorotational instability @xcite .
when including the vortex prescription in the fit , there are still significant residual emission in both sao 206462 and sr 21 .
these residuals are not symmetric , and their morphology is not well described by an additional vortex since they extends outwards with increasing radius .
a spiral - like structure may be a good approximation for the residual emission seen in both disks , but the number of parameters of such a prescription is quite large ( at least 5 more parameters per spiral arm ) .
given the low signal of these features and degeneracies between such large number of parameters , we decided not to proceed with such a fit .
observations at higher angular resolution are required to study in greater detail the morphology of these structures . on the bottom left panel of figures 2 and 3 , we present polarized intensity maps at h - band of both disks @xcite .
this emission , likely optically thick , does not directly traces mass surface density . in sao 206462
two spiral - like structures are seen in h - band imaging ( figure 2 ) .
@xcite compared the shape of these structures with theoretical models for a perturbation due to low - mass companions , their best - fit spirals , s1 and s2 , are labeled in figure 2 . for the s1 spiral , the location of the unseen perturber coincides with the location of the peak of emission in the alma map ( at @xmath71 au from the star with pa of @xmath72 ) .
if the 0.45 mm observations trace a mass overdensity at this location , we may be witnessing a spiral wave pattern that arose from a density enhancement rather than a point - like perturber , as reviewed by @xcite .
figure 2 also shows the location of s1/s2 with respect to the modeling residuals .
some overlap exist between the submillimeter residuals and the near - ir emission , with the s1/s2 spiral arms generally found inward of where the submillimeter residuals are .
nevertheless , observations with better angular resolution of sao 206462 are required to investigate whether the submillimeter emission traces the spiral structures observed in the near - ir . in the case of sr 21 , @xcite showed that not only is the h - band polarized intensity image is rather smooth and axisymmetric , but it is also not compatible with the large depletion of material in the inner disk ( @xmath73 ) found by @xcite .
furthermore , radiative transfer modeling of these near - ir observations did not find a transition between the outer disk and the depleted inner disk , even for modest depletion levels of 1 - 10% @xcite .
these h - band observations of sr 21 show evidence for emission from small grains out to a 80 au radius , and compared with our alma observations , it shows no signs of the observed asymmetries ( figure 3 ) . the difference in spatial distribution of the small grains ( traced by near - ir observations ) and
the mm - sized particles ( traced by these alma observations ) , could arise from the presence of a local enhancement in the gas density .
for example disk vortices , which can be excited by rossby - wave instabilities at the outer edge of gas depleted cavities , are particularly efficient at trapping large dust grains while small grains remain coupled to the gas .
although not as dramatic as in the oph irs 48 case , the vortex in sr 21 may be driving the distribution of large and small particles within the disk .
we have observed two transitional disks , sao 206462 and sr 21 , with high sensitivity using alma .
these transitional disks display large - scale asymmetries in their dust continuum emission , just as two other recently observed with alma @xcite , and an additional disk observed with carma @xcite .
it is interesting to note that of the few such objects observed with high sensitivity and high angular resolution , most exhibit high contrast asymmetrical structures .
the lack of previous detections is likely due to phase errors introduced by the atmosphere and the lack of sensitivity of previous observations .
this may imply that the process ( or processes ) that drive these asymmetries is taking place in most transitional disks . in the case of sao 206462 and sr 21
, we have demonstrated that a ring does not describe these observations , and a vortex - like asymmetry , in addition to a ring of emission , is a better match .
such asymmetrical structures encompass a significant fraction ( @xmath7 ) of the total disk emission , although we found that their contrast is much smaller than in lkh@xmath2 330 , hd 142527 , and oph irs 48 .
asymmetries in sao 206462 and sr 21 have a narrow radial size and a broad azimuthal extent , similar to the lkh@xmath2 330 and oph irs 48 asymmetrical structures .
if these asymmetries are in fact non - transient vortices , the turbulence inside them needs to be substantial in order to produce the azimuthally - broad asymmetrical features . despite being a better fit
, the vortex prescription does not reproduce every observed feature .
significant residuals remain that are suggestive of spiral - like structure , in the case of sao 206462 some residuals roughly coincide with the spiral arms seen in h - band scattered light . to test the presence of large - scale spiral - like features , and to make sure the observed structures trace the mass surface density of the disk down to the midplane ,
observations at longer wavelengths are required .
future observations with alma at 3 mm , and/or with the very large array at 7 mm , will help in this regard , since such observations trace larger dust grains and will have lower optical depth .
we thank takayuki muto , carol grady , and katherine follette for providing the near - ir images , and we thank wladimir lyra for useful discussions .
, l.m.p acknowledge support from nsf award ast-1109334 .
the national radio astronomy observatory is a facility of the national science foundation operated under cooperative agreement by associated universities , inc .
this paper makes use of the following alma data : ads / jao.alma 2011.0.00724.s .
alma is a partnership of eso ( representing its member states ) , nsf ( usa ) and nins ( japan ) , together with nrc ( canada ) and nsc and asiaa ( taiwan ) , in cooperation with the republic of chile .
the joint alma observatory is operated by eso , aui / nrao and naoj .
andrews , s. m. , wilner , d. j. , espaillat , c. , et al .
2011 , , 732 , 42 andrews , s. m. , wilner , d. j. , hughes , a. m. , qi , c. , & dullemond , c. p. 2009
, , 700 , 1502 birnstiel , t. , dullemond , c. p. , & pinilla , p. 2013
, , 550 , l8 beckwith , s. v. w. , sargent , a. i. , chini , r. s. , & guesten , r. 1990 , , 99 , 924 brown , j. m. , blake , g. a. , dullemond , c. p. , et al . 2007 , , 664 , l107 brown , j. m. , blake , g. a. , qi , c. , et al . 2009 , , 704 , 496 brown , j. m. , pontoppidan , k. m. , van dishoeck , e. f. , et al . 2013 , , 770 , 94 bruderer , s. 2013 , , 559 , a46 casassus , s. , van der plas , g. , m , s. p. , et al . 2013 , , 493 , 191 clarke , c. j. , gendrin , a. , & sotomayor , m. 2001 , , 328 , 485 dullemond , c. p. , & dominik , c. 2005 , , 434 , 971 dutrey , a. , guilloteau , s. , pitu , v. , et al .
2008 , , 490 , l15 espaillat , c. , dalessio , p. , hernndez , j. , et al .
2010 , , 717 , 441 follette , k. b. , tamura , m. , hashimoto , j. , et al .
2013 , , 767 , 10 foreman - mackey , d. , hogg , d. w. , lang , d. , & goodman , j. 2013 , , 125 , 306 fukagawa , m. , tsukagoshi , t. , momose , m. , et al .
2013 , , 65 , l14 garufi , a. , quanz , s. p. , avenhaus , h. , et al .
2013 , , 560 , a105 hernndez , j. , hartmann , l. , megeath , t. , et al .
2007 , , 662 , 1067 ireland , m. j. , & kraus , a. l. 2008 , , 678 , l59 isella , a. , prez , l. m. , carpenter , j. m. , et al .
2013 , , 775 , 30 isella , a. , prez , l. m. , & carpenter , j. m. 2012 , , 747 , 136 lombardi , m. , lada , c. j. , & alves , j. 2008 , , 480 , 785 lyo , a .- r . , ohashi , n. , qi , c. , wilner , d. j. , & su , y .-
2011 , , 142 , 151 lyra , w. , & lin , m .- k .
2013 , , 775 , 17 mller , a. , van den ancker , m. e. , launhardt , r. , et al .
2011 , , 530 , a85 muto , t. , grady , c. a. , hashimoto , j. , et al .
2012 , , 748 , l22 paardekooper , s .- j . ,
lesur , g. , & papaloizou , j. c. b. 2010 , , 725 , 146 pontoppidan , k. m. , blake , g. a. , van dishoeck , e. f. , et al .
2008 , , 684 , 1323 regly , z. , juhsz , a. , sndor , z. , & dullemond , c. p. 2012
, , 419 , 1701 roeser , s. , demleitner , m. , & schilbach , e. 2010 , , 139 , 2440 salyk , c. , blake , g. a. , boogert , a. c. a. , & brown , j. m. 2009 , , 699 , 330 simon , j. b. , armitage , p. j. , & beckwith , k. 2011 , , 743 , 17 van der marel , n. , van dishoeck , e. f. , bruderer , s. , et al .
2013 , science , 340 , 1199 wolf , s. , & dangelo , g. 2005 , , 619 , 1114 | we present atacama large millimeter / submillimeter array ( alma ) observations in the dust continuum ( 690 ghz , 0.45 mm ) and @xmath0co @xmath1 spectral line emission , of the transitional disks surrounding the stars sao 206462 and sr 21 .
these alma observations resolve the dust - depleted disk cavities and extended gaseous disks , revealing large - scale asymmetries in the dust emission of both disks .
we modeled these disks structures with a ring and an azimuthal gaussian , where the azimuthal gaussian is motivated by the steady - state vortex solution from @xcite .
compared to recent observations of hd 142527 , oph irs 48 , and lkh@xmath2 330 , these are low - contrast ( @xmath3 ) asymmetries .
nevertheless , a ring alone is not a good fit , and the addition of a vortex prescription describes these data much better .
the asymmetric component encompasses @xmath4 and @xmath5 of the total disk emission in sao 206462 and sr 21 respectively , which corresponds to a lower limit of @xmath6 of material within the asymmetry for both disks . although the contrast in the dust asymmetry is low , we find that the turbulent velocity inside it must be large ( @xmath7 of the sound speed ) in order to drive these azimuthally wide and radially narrow vortex - like structures .
we obtain residuals from the ring and vortex fitting that are still significant , tracing non - axisymmetric emission in both disks .
we compared these submillimeter observations with recently published h - band scattered light observations . for sr 21 the scattered light emission is distributed quite differently from submillimeter continuum emission , while for sao 206462 the submillimeter residuals are suggestive of spiral - like structure similar to the near - ir emission . |
the diffuse cosmic infrared background ( cib ) consists of the cumulative energy releases in the universe that have either been redshifted , or absorbed and reradiated by dust , into the infrared ( ir ) wavelength region .
the cib therefore provides important constraints on the rates of nuclear and gravitational energy release , as well as more exotic forms of energy release , over the history of the universe . over the past few years , analyses of data obtained with the diffuse infrared background experiment ( dirbe ) and the far infrared absolute spectrophotometer ( firas ) onboard the _ cosmic background explorer _
( _ cobe _ ) satellite have provided the first measurements of the cosmic background in the far infrared to submillimeter wavelength region ( puget et al .
1996 , schlegel et al .
1998 , fixsen et al .
1998 , hauser et al .
1998 , lagache et al . 1999 ,
lagache et al .
a detailed description of the _ cobe _ instruments and the _ cobe _ mission is given by boggess et al .
( 1992 ) , silverberg ( 1993 ) , mather , fixsen , & shafer ( 1993 ) , hauser et al .
( 1997 ) , and brodd et al .
recent reviews covering the history of the quest for the cib , the current detections and limits on its spectrum , and the astrophysical implications are given by hauser and dwek ( 2001 ) , kashlinsky ( 2005 ) , and lagache , puget , and dole ( 2005 ) .
the far infrared cib measurements are limited by the accuracy to which foreground interplanetary and galactic emission can be modeled and subtracted from the _ cobe _ data .
emission from interplanetary dust is the dominant foreground below about 100 @xmath10 and emission from interstellar dust is the dominant one at longer wavelengths .
different models with different degrees of complexity have been used to remove the interplanetary dust ( ipd ) emission from the _ cobe _ maps .
puget et al . ( 1996 ) and schlegel et al . ( 1998 ) relied only on the spatial characteristics of the ipd emission and subtracted a scaled template based on the dirbe 25 @xmath9 m sky map .
the detections reported by hauser et al .
( 1998 ) and fixsen et al .
( 1998 ) used the ipd model of kelsall et al .
( 1998 ) , which was fit to the time variation of the dirbe data caused by the motion of the earth through the interplanetary dust cloud .
the uncertainty in the zero level of the emission predicted by this model makes a major contribution to the uncertainty of the related cib measurements .
wright ( 1998 ) and gorjian , wright , and chary ( 2000 ) modeled the ipd using a method similar to that of kelsall et al .
, with an added constraint that the residual 25 @xmath10 intensity after zodiacal light subtraction be zero at high galactic latitudes .
removal of the emission from interstellar dust requires an interstellar medium ( ism ) template that has a well - defined zero level and correlates well with the spatial variation of ir emission in the _ cobe _ maps . before the completion of the wisconsin h - alpha mapper ( wham ) survey , observations of galactic h i emission provided the best ism template for this purpose ( puget et al .
1996 ; schlegel et al . 1998 ; and hauser et al . 1998 ) .
for example , hauser et al . (
1998 ) determined correlations of ir intensity with galactic h i column density for selected regions at high galactic latitude and high ecliptic latitude , using data in the dirbe 100 @xmath10 , 140 @xmath10 , and 240 @xmath10 bands .
for each region and each wavelength band , the h i correlated component of the infrared emission was subtracted from the data .
careful error analysis , including estimates of systematic error in subtraction of the interplanetary and galactic foregrounds , showed that the mean residual intensity was significantly ( more than 3@xmath2 ) greater than zero at all three wavelengths .
the residual intensity passed tests for isotropy at 140 @xmath10 and 240 @xmath10 , so detection of the cib was claimed at these wavelengths .
this method of subtracting galactic foreground emission is subject to error if the ratio of galactic foreground emission to galactic h i column density varies over the region studied .
such variation could occur if there is emission from dust associated with molecular or ionized gas , and this emission is not entirely correlated with h i column density .
based on the fuse h@xmath1 absorption line study of gillmon et al .
( 2006 ) and gillmon and shull ( 2006 ) , h@xmath1 column density is expected to be negligible compared to h i column density over a large fraction of the high latitude sky , and to account for 1% to 30% of total h column density for cirrus features brighter than @xmath11 mjy sr@xmath5 in the temperature - corrected 100 @xmath10 map of schlegel , finkbeiner , and davis ( 1998 ) .
most far infrared cib determinations have used restrictions on galactic latitude , h i column density , and/or far infrared color to exclude lines of sight that may contain significant emission from dust associated with molecular gas . on the other hand , the warm ionized medium ( wim )
is known to be prevalent at high galactic latitudes .
available data indicates that h ii column density is on average about one - third of h i column density ( reynolds 1991a ) , and correlation studies using h@xmath0 as a tracer suggest that a significant fraction of the h ii is not spatially correlated with h i ( e.g. , reynolds et al .
1995 , arendt et al .
1998 , lagache et al .
the ionized gas is expected to be subject to approximately the same interstellar radiation field as the neutral gas , and the depletion study of howk and savage ( 1999 ) shows that the dust to gas mass ratio in the wim may be similar to that in the warm neutral medium .
thus , the infrared emissivity per h nucleus may be similar in the ionized and diffuse h i phases of the ism , and infrared emission from the ionized phase may have a significant effect on any cib determination that is solely based on correlation with an h i template .
the paucity of tracers of the ionized gas at high latitudes precluded any definitive measurement of the amount of ir emission from dust in this gas phase .
nevertheless , several methods were used to estimate and subtract its contribution to the foreground ism emission : ( 1 ) puget et al .
( 1996 ) and lagache et al .
( 1999 ) identified the h ii emission component with a spatially varying component obtained after subtraction of an h i correlated foreground component from _
cobe_/firas data .
this residual component exhibited a csc(@xmath12 ) dependence ( boulanger et al .
1996 ) , consistent with that expected from a plane parallel layer of ionized gas .
( 2 ) fixsen et al .
( 1998 ) used maps of h i column density and [ c ii ] 158 @xmath10 line emission as templates to model the ir emission from the neutral and ionized gas phases .
they found that essentially none of the high latitude emission observed by firas correlated with the [ c ii ] template , and their final cib spectrum was more than two times greater than the h ii emission subtracted cib spectrum of puget et al .
( 1996 ) at @xmath13 .
( 3 ) arendt et al .
( 1998 ) derived an upper limit to the 100 @xmath10 emissivity per h nucleus in the ionized medium from a correlation analysis of h@xmath0 , h i , and dirbe maps for a @xmath14 region centered at @xmath15 , @xmath16 .
the derived @xmath17 upper limit was equal to 3/4 of the 100 @xmath10 emissivity per h nucleus for the neutral atomic gas in the same region .
assuming this limit is valid for the lockman hole region , they used available h@xmath0 and pulsar dispersion measure data for the lockman hole to place an upper limit of 4 nw m@xmath4 sr@xmath5 on the possible contribution of the ionized gas phase to the 100 @xmath10 foreground emission .
this was scaled to obtain upper limits of 5 and 2 nw m@xmath4 sr@xmath5 at 140 and 240 @xmath10 , respectively , assuming the spectrum of emission from the ionized phase has the same shape as that of the neutral atomic phase .
these upper limits are comparable to the overall uncertainties in the cib determinations .
they were noted as possible errors by hauser et al .
( 1998 ) but were not included in their quoted cib uncertainties . if they had been included , their reported 240 @xmath9 m cib value would still be a 3@xmath2 detection .
the wisconsin h - alpha mapper ( wham ) northern sky survey provided the first h@xmath0 map sensitive enough to trace the ionized gas phase of the ism at high galactic latitude ( haffner et al .
lagache et al .
( 2000 ) used preliminary wham survey data to decompose the 100 to 1000 @xmath10 dirbe and firas data at high latitude into h@xmath0-correlated , h i - correlated , and isotropic components .
the regions studied cover about 2% of the sky in the range @xmath18 .
they found a significant h@xmath0-correlated component with dust temperature of 17.2 k , very similar to that of the h i - correlated component . assuming a constant electron density of 0.08 @xmath19 and an electron temperature of 8000 k for the ionized gas
, they derived an infrared emissivity per h nucleus for this phase that is similar to that of the neutral phase .
the cib spectrum they determined from analysis of the firas data is consistent with that of fixsen et al . (
the mean residual intensities they found at 100 , 140 , and 240 @xmath10 from analysis of the dirbe data are consistent with the results of hauser et al .
( 1998 ) , although their uncertainties are larger .
the agreement between these latest cib determinations is encouraging , but the disagreement between the arendt et al .
( 1998 ) and lagache et al .
( 2000 ) results for the 100 @xmath10 emissivity of the ionized medium is a matter of concern . in this paper
, we address the possible effects of the ionized medium on the hauser et al .
( 1998 ) cib results by including an h@xmath0-correlated foreground component in an analysis that is otherwise similar to the dirbe team analysis .
we make use of h@xmath0 data from the wham northern sky survey . the paper is organized as follows . in
2 we describe the data sets and sky areas used in the analysis . the decomposition method is described in 3 . from the decompositions we derive the emissivity per h nucleus for the h i and h ii phases of the ism as well as the residual emission for the different dirbe bands . in 4 we present these results and compare them with results of previous studies . in 5 we discuss the emissivity results and we discuss possible systematic errors if h@xmath0 is not a reliable tracer of far infrared emission .
our results and conclusions are summarized in 6 .
the regions of the sky analyzed in this paper are shown in figure 1 .
three of them were previously analyzed by hauser et al .
( 1998 ) and arendt et al .
( 1998 ) : the lockman hole ( lh ) , a 300 square degree region around the position of lowest h i column density at @xmath20 ; an @xmath21 region centered on the north ecliptic pole ( nep ) at @xmath22 ; and the dirbe high quality b north ( hqbn ) region at @xmath23 and @xmath24 .
these regions were originally selected because they were expected to have relatively weak galactic and interplanetary dust foregrounds , and because good quality h i observations were available for them .
the dirbe high quality b south region is not included in our analysis because it is below the declination limit of the wham northern sky survey .
in addition , we analyze the second quadrant region previously studied by lagache et al .
( 2000 ) and a new region in the first quadrant at @xmath25 .
we refer to these as the q2 region and q1 region , respectively .
the q2 region was originally selected because preliminary wham h@xmath0 data were available for it . the q1 region was selected because it is comparable to the q2 region in galactic latitude , it is at ecliptic latitude greater than 30 degrees , it has no molecular clouds detected in the co survey of hartmann , magnani , and thaddeus ( 1998 ) , and it has no cold infrared excess features that are characteristic of molecular clouds in the 100 @xmath10 infrared excess map of reach et al .
( 1998 ) .
the datasets used in our analysis are listed in table 1 .
we use dirbe 60 , 100 , 140 , and 240 @xmath10 mission - averaged skymaps from which the interplanetary foreground emission has been subtracted using the model of kelsall et al .
foreground galactic stellar emission is negligible at these wavelengths ( hauser et al .
1998 , arendt et al . 1998 ) and has not been subtracted from the data .
we use h i 21-cm line data integrated over a velocity range that includes all significant galactic emission , converted to h i column density @xmath26(h i ) assuming that the line emission is optically thin .
the h i data for the lh and nep regions are from snowden et al .
( 1994 ) and elvis , lockman , and fassnacht ( 1994 ) , and were corrected for stray radiation using the at&t bell laboratories h i survey ( stark et al .
estimated 1@xmath2 uncertainties in @xmath26(h i ) for these regions range from @xmath27 cm@xmath4 to @xmath29 cm@xmath4 .
the h i data that we use for the other regions are from the leiden - dwingeloo h i survey ( hartmann and burton 1997 ) , which has also been corrected for stray radiation ( hartmann et al .
we adopt a 1@xmath2 uncertainty of @xmath30 cm@xmath4 for each position in these regions .
this is the uncertainty of the stray radiation correction that was estimated by hartmann et al .
( 1996 ) for a reference position in the lockman hole region .
we use h@xmath0 total intensity data from the wisconsin h - alpha mapper ( wham ) northern sky survey ( haffner et al . 2003 ) , which covers the sky north of declination @xmath31 . the wham instrument has a 1 degree diameter field of view , and the survey was made on a regular galactic coordinate grid with pointings separated by @xmath32cos @xmath33 in @xmath34 and @xmath35 in @xmath33 .
the spectrum for each pointing was integrated over @xmath36 km s@xmath5 to obtain total h@xmath0 intensity .
for the regions we study , systematic errors associated with removal of geocoronal and atmospheric emission lines from the spectra can be greater than statistical measurement uncertainties .
these errors can vary from night to night , and sometimes cause @xmath37 `` blocks '' of data taken on particular nights to be noticeably offset in mean intensity relative to their surroundings .
we applied offset corrections to affected blocks in the hqbn , lockman hole , and q2 regions to remove discontinuities in h@xmath0 intensity at the block boundaries . for the hqbn and q2 regions
, most blocks appeared to be unaffected and these were assumed to set the zero level of the data .
for the hqbn region , an offset of 0.15 rayleigh was added to the data in the areas ( @xmath38 ) , ( @xmath39 ) , and ( @xmath40 ) . for the q2 region ,
an offset of 0.3 r was added in ( @xmath41 ) and an offset of 0.4 r was added in ( @xmath42 ) . for the lockman hole ,
the offset corrections are somewhat uncertain and subjective since many blocks appear to be affected , by varying amounts .
offset corrections ranged from -0.5 to 0.1 r , and the zero level is determined by the wham observations of hausen et al .
( 2002 ) for two positions in the region .
our decomposition results for the lockman hole are consistent with those for the other regions , and omitting this region from our analysis does not change our derived cib results significantly .
we adopt a 1@xmath2 uncertainty of 0.06 rayleigh for the velocity - integrated , offset - corrected h@xmath0 intensity at each wham pointing .
this is near the low end of the range of rms h@xmath0 dispersion measured within observing blocks at high galactic latitudes .
the dispersion tends to increase with increasing mean h@xmath0 intensity or decreasing latitude , presumably due to increasing rms dispersion in galactic emission . with the adopted uncertainty , the mean h@xmath0 signal - to - noise ratio is 4 , 8 , 18 , 18 , and 31 for the lh , hqbn , q1 , q2 , and nep regions , respectively .
the dirbe data and leiden h i data were interpolated to the wham pointing positions for our analysis .
the angular resolution of the elvis et al . and snowden et al .
h i data is much better that that of the other datasets , so these data were averaged over the wham field of view at each wham pointing position .
possible contamination by stellar h@xmath0 absorption was handled either by excluding positions with stars brighter than @xmath43 , or by excluding positions with h@xmath0 intensity significantly lower than their surroundings .
these two methods were compared for the lh and nep regions , and found to give consistent results . some positions with a discrete source detected in the dirbe data ( the planetary nebulae ngc 6543 , and galaxies ngc 3079 , 3310 , 3556 , 3690 , and 4102 ) were excluded from our analysis .
the planetary nebula ngc 6210 appears as a bright point source in the wham data ( reynolds et al .
2005 ) and was also excluded .
maps of the 100 @xmath10 intensity , h i column density , and h@xmath0 intensity are shown for the q1 region in figure 2 , and correlation plots are shown for this region in figure 3 .
these figures show that the correlation between @xmath44(100 @xmath10 ) and @xmath26(h i ) is tighter than that between @xmath44(100 @xmath10 ) and @xmath45(h@xmath0 ) or that between @xmath45(h@xmath0 ) and @xmath26(h i ) .
similar trends are seen for the other regions analyzed in this paper , and for the region around @xmath46 studied previously ( reynolds et al .
1995 , arendt et al .
the method of analysis follows that used by arendt et al .
( 1998 ) and is similar to that used by lagache et al .
( 2000 ) . for each dirbe wavelength band
( 60 , 100 , 140 , and 240 @xmath10 ) , the infrared intensity distribution within a given region is decomposed into a component that is correlated with h i column density , a component that is correlated with h@xmath0 intensity , and an isotropic component .
this is done by making a least squares fit of the form @xmath47 where @xmath48 is the infrared intensity after interplanetary foreground subtraction and @xmath49 , @xmath50 , and @xmath51 are fit parameters .
@xmath49 is the mean infrared emissivity per h atom for dust in the neutral atomic gas phase , @xmath50 is a measure of the mean infrared emissivity of dust in the ionized gas phase , and the intercept @xmath51 is the mean residual infrared intensity . for comparison
, a second decomposition is performed in which an h@xmath0 correlated component is not included , by making a fit of the form @xmath52 comparison of the derived @xmath51 and @xmath53 values gives the error in the galactic foreground subtraction if the h@xmath0 correlated component is neglected .
decomposition of an infrared intensity distribution into three components as in equation ( 1 ) will be successful if the spatial distibutions of @xmath26(h i ) and @xmath45(h@xmath0 ) differ significantly from each other , and also differ significantly from an isotropic distribution .
these conditions are met for each of the regions studied here .
this is illustrated for the q1 region by figures 2 and 3 .
the method of analysis also assumes that h@xmath0 intensity is a good tracer of far infrared emission from the ionized medium . in 5 , we discuss possible errors in our results if this is not the case .
extinction of the h@xmath0 emission is another potential source of error , but its effect is negligible for our regions with the data selection criteria described below .
we made fits to the 100 @xmath10 data for each region with and without a correction to @xmath45(h@xmath0 ) for extinction , and differences in the results were insignificant .
we made the worst - case assumption of pure foreground extinction . with the optical depth at h@xmath0 calculated as @xmath54(h i)/@xmath55 cm@xmath4 ] ,
the extinction correction was at most 1.22 .
the fits are made using an iterative procedure that minimizes @xmath57 calculated using measurement uncertainties in the independent and dependent variables ( press et al .
uncertainties in the fit parameters are determined from the 68% joint confidence region in parameter space , using the method of bard ( 1974 ) . for each of the regions except for hqbn , data at the highest h i column densities
are excluded from the fitting , as described by arendt et al .
( 1998 ) for the lh and nep regions . the @xmath58 _
n_(h i ) relation deviates from linearity in these regions , with excess 100 @xmath10 emission relative to _
i ) at the highest h i column densities . this type of relation has been found previously for isolated cirrus clouds and for large regions of the sky at high galactic latitude , and the excess 100 @xmath10 emission has been attributed to emission from dust associated with molecular gas or to non - negligible optical depth in the 21-cm line ( e.g. , deul and burton 1993 , reach , koo , and heiles 1994 , boulanger et al .
for the lockman hole , it is consistent with detections of co line emission toward some 100 @xmath10 brightness peaks ( heiles , reach , and koo 1988 ; stacy et al .
1991 ; reach , koo , and heiles 1994 ) .
we exclude data above h i column densities where the @xmath59 _
n_(h i ) relation begins to deviate from linearity ( see figure 8 of arendt et al .
the cut is made at @xmath60 cm@xmath4 , @xmath61 cm@xmath4 , @xmath62 cm@xmath4 , and @xmath62 cm@xmath4 for the lh , nep , q1 , and q2 regions , respectively .
measurement uncertainties are much larger for the dirbe 140 and 240 @xmath10 bands than for the 60 and 100 @xmath10 bands , so less stringent _
n_(h i ) limits were adopted for the lh and nep fits in these bands , @xmath63 cm@xmath4 for the lh and @xmath64 cm@xmath4 for the nep . with these cuts ,
the area of the sky used in the 100 @xmath10 analysis is 420 , 170 , 35 , 180 , and 380 square degrees for the hqbn , lh , nep , q1 , and q2 regions , respectively .
our analysis does not allow for possible fir emission from molecular gas that is not correlated with h i. this emission is expected to be negligible for the wham pointings that pass the @xmath26(h i ) cuts .
gillmon et al .
( 2006 ) and gillmon and shull ( 2006 ) reported results from a _ fuse _ survey of h@xmath1 absorption lines toward 45 agns at @xmath65 .
they compared their derived values of the molecular fraction , @xmath66h@xmath67h i@xmath68h@xmath69 $ ] , with values of temperature - corrected 100 @xmath10 intensity @xmath70 from the map of schlegel , finkbeiner , and davis ( 1998 ) .
( @xmath70 is proportional to galactic dust column density . )
the transition from low molecular fractions characteristic of optically thin clouds to high values characteristic of h@xmath1 self - shielded clouds was found to occur over the range @xmath71 mjy sr@xmath5 , with @xmath72 varying between @xmath73 and @xmath74 in this @xmath70 range , @xmath72 less than @xmath75 at lower @xmath70 values , and @xmath72 between @xmath76 and 0.3 at higher @xmath70 values . except for the nep region , most of the wham pointings used in our analysis ( after the @xmath26(h i )
cuts have been applied ) have @xmath70 less than 1.5 mjy sr@xmath5 , so @xmath72 is expected to be generally less than @xmath75 and the uncertainty in derived cib results due to neglect of this component is negligible . for each region except the nep , we estimate that dust associated with h@xmath1 contributes less than 0.04 , 0.05 , and 0.02 nw m@xmath4 sr@xmath5 at 100 , 140 , and 240 @xmath10 , respectively .
this assumes that @xmath26(h@xmath1 ) is constant within a region , @xmath77(h@xmath78 mean @xmath26(h i ) , and ir emissivity per h nucleus in the h@xmath1 phase is given by the slope of the ir
i ) relation . for the nep ,
most of the wham pointings that are used have @xmath70 in the transition range from 1.5 to 3.0 mjy sr@xmath5 , so @xmath72 values as large as 0.1 are possible .
however , we find no evidence for significant ir emission from h@xmath1 associated dust .
the mean residual infrared intensities @xmath51 from our analysis of the nep region are consistent with the values found for the other regions , and the 100 @xmath10
i ) relation for the nep is linear with small scatter over the range of @xmath26(h i ) used in our analysis ( see figures 7 and 8 of arendt et al .
excluding the nep region from our analysis would not change our derived cib results significantly . as a check of our analysis method and software , we performed fits of the form of equation ( 1 ) using data previously analyzed by lagache et al .
( 2000 ) for the q2 region .
lagache et al .
performed fits of this form for 122 positions in the region using dirbe data with interplanetary foreground subtracted , leiden - dwingeloo h i data , and preliminary wham data , all smoothed to the @xmath79 resolution of the _ cobe_/firas .
they kindly provided us with the data .
we performed fits as described above , except to be consistent with the lagache et al .
analysis , only measurement uncertainties for the dependent variable ( the infrared intensity ) were used in the calculation of @xmath57 .
table 2 gives a comparison of our results and the lagache et al .
the two analyses give the same values for the fit parameters , but the values for the fit parameter uncertainties do not agree .
our uncertainties for @xmath49 and @xmath50 are larger than those given by lagache et al . because our uncertainty calculation allows for uncertainty due to coupling between the parameters .
( the uncertainty quoted for @xmath51 by lagache et al .
is not a statistical uncertainty from the fitting , but was determined from the distribution of residual intensity values , so comparison with our uncertainty value for @xmath51 is not meaningful . )
we conclude from this test that differences between our results in 4 and those of lagache et al .
are not the result of software errors .
parameters from our fits for the five regions are listed in table 3 .
positions with h i column density greater that the cuts described in section 3.1 were excluded from the fitting . for each region and each wavelength , the first row in the table gives results from the three - component fit ( equation 1 ) and the second row gives results from the two - component fit ( equation 2 ) .
the uncertainties listed are statistical uncertainties from the 68% joint confidence region in parameter space ; the @xmath51 and @xmath53 uncertainties do not include systematic uncertainties that need to be included in determining the total uncertainty for a cib measurement .
column 7 of the table lists the number of independent wham pointings used for each fit .
sample correlation plots showing the fits to the 100 @xmath10 data for the q1 and q2 regions are shown in figure 4 , and parameters from the three - component fits for each region are plotted as a function of wavelength in figures 5 and 6 .
for all regions in all wavelength bands analyzed , we find that the mean residual intensity is nearly the same whether an h@xmath0-correlated component is included in the fitting or not ( the @xmath51 and @xmath53 values are in close agreement ) , and the quality of the fit is nearly the same in the two cases . also , we do not detect significant h@xmath0-correlated infrared emission ; the @xmath50 values are consistent with zero within the uncertainties , and the @xmath49 and @xmath80 values are nearly the same .
the lack of h@xmath0-correlated 100 @xmath10 emission is illustrated for the q1 region in figures 2 and 3 .
no correlation is seen between the h@xmath0 map in figure 2c and and the map in figure 2d of residual 100 @xmath10 emission after subtraction of h i correlated 100 @xmath10 emission .
this is also shown by the correlation plot in figure 3d .
figure 7 shows the residual intensity averaged over the five regions as a function of wavelength for each fit type ; the @xmath51 and @xmath53 values were each averaged over the regions using weighting by 1/@xmath81 .
the weighted - average residual intensities for the different fit types agree within 1@xmath2 statistical uncertainties at all wavelengths , and the agreement is within 2 nw m@xmath4 sr@xmath5 at 60 , 100 , and 240 @xmath10 . we conclude that addition of an h@xmath0-correlated component in modeling the foreground emission at high galactic latitude has negligible effect on derived cib results . to assess whether the weighted - average values for the residual emission can be identified as cib measurements , we calculated the total uncertainty for these values following the method used by hauser et al .
the total uncertainty was calculated as the quadrature sum of the statistical uncertainty , interplanetary foreground subtraction uncertainty , and dirbe detector offset uncertainty .
magnitudes of the latter two uncertainties were taken from table 6 of arendt et al .
( dirbe gain uncertainty is not included because it has the same multiplicative effect on the mean residual and its total uncertainty , and so does not affect the signal - to - noise ratio .
this uncertainty is 1014% for the dirbe bands used here . )
based on these total uncertainties , the mean residual emission is more than 3@xmath2 greater than zero at 140 @xmath10 and 240 @xmath10 for both the two - component fits and the three - component fits , and at 100 @xmath10 for the three - component fits .
the residual intensity values for the five different regions ( table 3 and figure 6 ) are consistent with isotropy at 140 @xmath10 and 240 @xmath10 ( the values are compatible within their 1@xmath2 uncertainties ) . we have not performed detailed anisotropy tests on the maps of residual intensity from our analysis , such as the tests performed by hauser et al .
( 1998 ) . however , based on the reduced chi - square values in table 3 , significant anisotropy is present in the q1 and q2 regions at 60 @xmath10 and in all five regions at 100 @xmath10
. table 4 lists the upper limits and detections of the cib at 60 , 100 , 140 , and 240 @xmath10 from hauser et al .
( 1998 ) , from our two - component and three - component fits , and from lagache et al . ( 2000 ) . for cases where the mean residual intensity does not exceed zero by greater than 3@xmath2 or
the residual intensity distribution has been shown to be anisotropic , the table lists a 2@xmath2 upper limit on the cib followed by the mean residual intensity and its 1@xmath2 uncertainty in parentheses .
our results are consistent with those of hauser et al .
our residuals have not passed a detailed anisotropy test , and in some cases our quoted uncertainties are slightly larger than those of hauser et al . , so our cib results do not supersede the hauser et al .
results .
if h@xmath0 intensity is not a good tracer of infrared emission from dust in the wim , our three - component fits may not account for emission from the wim much better than fits without an h@xmath0 correlated component do .
any emission from the wim that is not correlated with h@xmath0 or h i would contribute to the residual component , so our results would overestimate the cib .
the agreement with the hauser et al .
( 1998 ) results could be because their analysis overestimates the cib by a similar amount . in
5.4 we discuss evidence that h@xmath0 may not be a good tracer , and we estimate the possible effect of emission from the wim on the cib results from our analysis and from that of hauser et al . for each analysis
, we estimate the effect to be only about 5% at 140 and 240 @xmath10 , which is much smaller than the quoted uncertainties .
although this paper is primarily concerned with the effect of the ionized ism on cib determination , we discuss here other systematic uncertainties involving the level of the h i cut , choice of interplanetary dust model , and choice of photometric calibration .
our results for the effect of the ionized ism have no significant dependence on any of these uncertainties .
arendt et al .
( 1998 ) found that the sensitivity of the hauser et al .
( 1998 ) cib results to the h i cut is negligible compared to other sources of uncertainty .
for the lockman hole , for example , the mean 100 @xmath10 residual intensity was found to vary by less than 0.7 nw m@xmath4 sr@xmath5 when the cut was varied from 1.0 to @xmath63 cm@xmath4 .
we have investigated the sensitivity of the cib results from our three - component analysis to the h i cut used for our regions that were not included in the hauser et al .
analysis , the q1 and q2 regions .
the effect of varying the cut for these regions over the range from 2.5 to @xmath82 cm@xmath4 is small .
the derived mean residual changes by + 0.1 , -0.7 , -0.1 , and + 0.7 nw m@xmath4 sr@xmath5 at 60 , 100 , 140 , 240 @xmath10 , respectively , when the cut is changed from 3.0 to @xmath83 cm@xmath4 .
these changes are at most 25% of the uncertainties of the mean residuals given in table 4 .
the mean residual changes by -0.8 , -0.5 , -3.0 , and -1.3 nw m@xmath4 sr@xmath5 in these bands when the cut is changed from 3.0 to @xmath82 cm@xmath4 .
these changes are at most 45% of the quoted uncertainties of the mean residuals .
the effect of choice of interplanetary dust model is shown in table 5 .
this table compares mean residual intensities from hauser et al .
( 1998 ) with those obtained by wright ( 2004 ) .
the main difference between these analyses is that hauser et al .
used the interplanetary dust model of kelsall et al .
( 1998 ) and wright used the model of gorjian et al .
each of these models was obtained by fitting the time variation observed over the whole sky in each of the dirbe bands with a parameterized model of the dust cloud , but gorjian et al . added a constraint that the residual 25 @xmath10 intensity after zodiacal light subtraction be zero at high galactic latitude .
this assumed that the 25 @xmath10 cib is negligible , based on cib upper limits inferred from tev gamma ray observations of mrk 501 .
table 5 also lists the uncertainty of the zodiacal light subtraction for the hauser et al .
analysis , as estimated by kelsall et al .
this was obtained by comparing results from a series of models with different geometries for the density distribution of the dust cloud , which gave comparable quality fits to the time variation of the dirbe data .
the difference between the hauser et al .
results and the wright results is not significant relative to the uncertainty in zodiacal light subtraction , or relative to the total uncertainty of the mean residual .
hauser et al . ( 1998 ) noted the effects on cib results at 140 and 240 @xmath10 if the dirbe data are transformed to the firas photometric system .
the cib values are affected by differences in zero point and gain between the dirbe and firas photometric systems , which are not significant relative to the zero point uncertainties and gain uncertainties for the two systems .
hauser et al . found that transforming the dirbe cib results to the firas system would reduce the cib value from 25.0 to 15.0 nw m@xmath4 sr@xmath5 at 140 @xmath10 , and from 13.6 to 12.7 nw m@xmath4 sr@xmath5 at 240 @xmath10 .
the dirbe team chose to use the dirbe photometric system to report its results since comparing the two photometric systems has its own uncertainties associated with the need to integrate the dirbe map over the firas beam , and the firas spectrum over the dirbe spectral response .
the dirbe team chose not to introduce this additional uncertainty , and we made the same choice for this paper .
recently , dole et al .
( 2006 ) used _
mips observations to measure the contribution of galaxies selected at 24 @xmath10 to the cib at 70 @xmath10 and 160 @xmath10 .
they found that sources brighter than 60 @xmath9jy at 24 @xmath10 contribute 5.9 @xmath7 0.9 and 10.7 @xmath7 1.6 nw m@xmath4 sr@xmath5 to the cib at 70 @xmath10 and 160 @xmath10 , respectively . with an extrapolation of the 24 @xmath10 source counts below the 60 @xmath9jy detection threshold
, they estimated that the full population of 24 @xmath10 sources contributes 7.1 @xmath7 1.0 and 13.4 @xmath7 1.7 nw m@xmath4 sr@xmath5 at 70 and 160 @xmath10 .
they noted that these should be regarded as lower limits to the cib since there may be contributions from a diffuse background component or from sources missed by the 24 @xmath10 selection .
they estimated that these contributions account for less than @xmath84 of the far infrared cib . a diffuse component of the cib could be produced by processes such as emission from intergalactic dust or radiative decay of primordial particles .
here we compare the dole et al .
integrated galaxy brightness at 160 @xmath10 with dirbe cib measurements at 140 @xmath10 and 240 @xmath10 to estimate upper limits on emission from a diffuse component .
we scale the quoted 160 @xmath10 intensity of 13.4 @xmath7 1.7 nw m@xmath4 sr@xmath5 to 140 @xmath10 and 240 @xmath10 using the shape of the model spectral energy distribution from figure 13 of dole et al . , which is based on the lagache et al .
( 2004 ) galaxy evolution model .
this spectrum is also used to apply color corrections so the intensities can be compared with the quoted dirbe cib values , which assume a spectral shape of @xmath85 constant over the dirbe bandpass .
we obtain integrated galaxy intensity values of 13.7 @xmath7 1.7 nw m@xmath4 sr@xmath5 at 140 @xmath10 and 10.7 @xmath7 1.4 nw m@xmath4 sr@xmath5 at 240 @xmath10 .
the uncertainties here do not include any uncertainty in the shape of the adopted spectral energy distribution .
table 6 compares these values with the dirbe cib results , and with dirbe cib results transformed to the firas photometric system ( hauser et al .
the table lists values for the fractional contribution of the integrated galaxy brightness to the cib , and the difference between the cib and the integrated galaxy brightness . using the cib results on the dirbe photometric system yields 2@xmath2 upper limits for a diffuse cib component of 26 nw m@xmath4 sr@xmath5 at 140 @xmath10 and 8.5 nw m@xmath4 sr@xmath5 at 240 @xmath10 .
infrared emissivity results from our three - component fits are shown as a function of wavelength in figure 5 .
the derived values of emissivity per h atom for the neutral atomic gas phase , @xmath86(h@xmath87 ) , are comparable to previous determinations from high latitude ir h i correlation studies ( e.g. , dwek et al .
1997 , reach et al .
the values of emissivity per h@xmath88 ion for the ionized phase , @xmath86(h@xmath89 ) , were obtained from the @xmath50 values in table 3 using a conversion factor of @xmath45(h@xmath0)/@xmath26(h@xmath90 = 1.15 rayleighs/@xmath55 cm@xmath4 .
this is the mean ratio of h@xmath0 intensity to pulsar dispersion measure found by reynolds ( 1991b ) for four high latitude lines of sight toward pulsars at @xmath91 4 kpc .
it corresponds to an effective electron density , @xmath92 , of 0.08 @xmath19 for an electron temperature of 8000 k and no extinction .
the @xmath45(h@xmath0)/@xmath26(h@xmath90 ratio ranges from 0.75 to 1.9 rayleighs/@xmath55 cm@xmath4 for the lines of sight studied by reynolds , so the uncertainty in the conversion factor is large .
the value adopted here is the same as that used by lagache et al .
figure 5 shows that the derived values of @xmath86(h@xmath89 ) are consistent with zero for all regions at all wavelengths .
we have checked the dependence of derived 100 @xmath10 @xmath86(h@xmath89 ) values on the h i cut , and find that they do not change significantly when the cut is varied by @xmath7 30% . our derived values of @xmath86(h@xmath89 ) at 100 @xmath10 are compared with previous results for other regions of the sky in table 7 .
all of the results listed are from analyses similar to that used here , with either h@xmath0 intensity or centimeter wavelength radio continuum intensity used as the tracer of the ionized gas .
column 4 lists the electron density adopted in each study for calculating the conversion from h@xmath0 or radio continuum intensity to @xmath26(h@xmath88 ) .
uncertainty in this conversion may be as large as a factor of two or more for some regions , but is not included in the uncertainty listed for @xmath86(h@xmath89 ) .
the emissivity values are shown plotted as a function of electron density in figure 8 .
the 2@xmath2 upper limits for our regions are comparable to that obtained by arendt et al .
( 1998 ) for the @xmath14 region at @xmath46 .
for the other previously studied regions , 100 @xmath10 emission was detected from the ionized gas component , and the derived @xmath86(h@xmath89 ) is greater than or equal to the 100 @xmath10 @xmath86(h@xmath87 ) .
this can be explained if the dust - to - gas mass ratio is the same in the ionized and neutral atomic components , with cases of enhanced emissivity in the ionized component caused by lyman alpha heating ( for the barnard s loop region , heiles et al .
2000 ) or a local source of heating ( for the spica region , boulanger et al .
1995 , zagury , jones , and boulanger 1998 ) , or both ( for the galactic plane regions , sodroski et al .
1997 ) . with the exception of the q2 region ,
the ionized gas in these regions is not representative of the general warm ionized medium observed at high latitudes .
the electron density is an order of magnitude greater , and the ionized gas is much closer to the galactic midplane ( @xmath93 100 pc , compared to an exponential @xmath94 scale height of about 900 pc for the warm ionized medium ( reynolds 1993 ) ) . the 100 @xmath10 @xmath86(h@xmath89 ) value determined by lagache et al .
( 2000 ) for the q2 region without an h i cut is not consistent with our result for the q2 region with an h i cut applied , or with results for the other high latitude regions that sample the general warm ionized medium .
we have applied our three - component decomposition to the full q2 region ( with no h i cut ) at 100 @xmath10 at the 1 degree resolution of the wham data .
this gave @xmath86(h@xmath89 ) = 8.5 @xmath7 1.0 nw m@xmath4 sr@xmath5/@xmath55 cm@xmath4 , which is comparable to the upper limits in figure 8 , and @xmath86(h@xmath87 ) = 18.7 @xmath7 0.4 nw m@xmath4 sr@xmath5/@xmath55 cm@xmath4 .
( if an extinction correction is applied to the h@xmath0 data as described in 3.1 , the derived @xmath86(h@xmath89 ) is 5.5 @xmath7 1.0 nw m@xmath4 sr@xmath5/@xmath55 cm@xmath4 . )
the difference between these results and those of lagache et al .
is probably due to the different angular resolution used .
their analysis used data for 122 positions at 7 degree firas resolution , but the data were oversampled and there are only about 30 independent firas pointings within the region .
our analysis used data for 1762 independent wham pointings that are within 3.5 degrees ( half of the firas beam width ) of any of their 122 positions .
it is possible that some unmodeled effect , such as dust associated with molecular gas , optical depth in the 21-cm line , or variation of @xmath86(h@xmath87 ) , happens to correlate with the h@xmath0 data at firas resolution , but does not correlate as well at wham resolution .
the @xmath95 _
n_(h i ) relation for the region shows curvature , which suggests that molecular gas or 21-cm line opacity may be present .
we find that 40% of the region has temperature - corrected 100 @xmath10 intensity greater than the 3 mjy sr@xmath5 threshold for significant fractional h@xmath1 abundance ( gillmon and shull 2006 ) .
for these reasons , we consider the emissivity per h@xmath88 ion derived for the full q2 region , from either our analysis or lagache et al.s analysis , to be of questionable reliability .
our derived emissivity values for the ionized medium are statistically consistent with zero . from table 7 , our derived 2@xmath2 upper limits on 100 @xmath10 emissivity per h@xmath89 ion are 0.33 , 1.11 , 0.37 , 0.57 , and 0.30 of the 100 @xmath10 emissivity per h atom for the hqbn , lh , nep , q1 , and q2 regions , respectively .
we adopt 0.4 as a representative upper limit on @xmath96h@xmath97h@xmath98 at 100 @xmath10 for these regions .
possible explanations for this low value include a lower dust - to - gas mass ratio or a weaker radiation field in the ionized medium than in the neutral medium , a difference in grain optical properties or grain size distribution , an error in our adopted @xmath45(h@xmath0)/@xmath26(h@xmath90 conversion factor , or a shortcoming of using h@xmath0 as a tracer of infrared emission from dust in the ionized medium . in this section ,
we discuss some of these possibilities .
based on available observations and models , it appears that our low derived @xmath96h@xmath97h@xmath98 ratio is partly due to a lower dust - to - gas mass ratio in the wim and partly due to error in using h@xmath0 as a tracer , but one or more of the other factors may also contribute .
we discuss possible implications for derived cib results in 5.4 .
a lower dust - to - gas mass ratio would be expected if the ionized gas in these regions has greater @xmath94 extent and lower density than the neutral gas , as is typical for the wim in the solar vicinity ( reynolds 1991a ) . from interstellar absorption line observations , it has been inferred that abundances of heavy elements in the form of dust decrease with increasing @xmath94 , and also decrease with decreasing gas density ( e.g. , savage and sembach 1996 ) .
most of these studies have pertained to the neutral atomic medium , but from observations of al iii and s iii absorption lines , howk and savage ( 1999 ) found evidence that about 60 - 70% of aluminum atoms in the wim are in dust , compared to about 90% of al in dust in h ii regions at low @xmath94 .
this result is based on observations of two lines of sight that sample the wim up to @xmath94 distances of 690 pc and 2800 pc , and four lines of sight through low density ( @xmath99 0.2 to 4 @xmath19 ) h ii regions at @xmath93 200 pc .
howk and savage also noted that the dust phase al abundance they obtained for the low @xmath94 h ii regions is comparable to previous , somewhat uncertain determinations for the warm neutral medium at low @xmath94 . assuming that this low @xmath94 dust phase al abundance is valid for the neutral atomic medium in the regions studied here , and that 100 @xmath10 emissivity per h nucleus varies in proportion to al dust phase abundance , one would expect @xmath86(h@xmath89 ) to be about 2030% lower than @xmath86(h@xmath87 ) at 100 @xmath10.thus the howk and savage results suggest that the difference in dust abundance between the neutral and ionized components is not great enough to fully explain our derived upper limit on @xmath96h@xmath97h@xmath98 . to confirm this , dust phase al abundance determinations for additional lines of sight through the wim
would be of interest , as would calculations of the relation between al dust phase abundance and 100 @xmath10 emissivity for dust grain models .
the lower 100 @xmath10 emissivity derived for the ionized medium could also be explained if it were subject to a weaker interstellar radiation field .
heating of dust by lyman alpha is expected to be negligible at the electron density estimated for the wim ( spitzer 1978 , heiles et al . 2000 ) . also , there are no early type stars that are close enough to the lines of sight through our regions to cause significant dust heating relative to that of the general interstellar radiation field
. models of the spatial distribution of the interstellar radiation field at visual and ultraviolet wavelengths predict that the mean intensity increases with increasing @xmath94 up to about 200 pc , as light from distant stars in the galactic disk becomes less attenuated , and then decreases with further increase in @xmath94 due to geometrical dilution .
wakker and boulanger ( 1986 ) used their model of the radiation field to calculate the expected 100 @xmath10 intensity of a diffuse cloud with a mixture of silicate and graphite grains .
the intensity was calculated for different distances of the cloud along two high latitude lines of sight ( toward @xmath100 , and toward @xmath101 , @xmath102 ) .
they found that the 100 @xmath10 intensity varies by less than @xmath720% over the range in cloud height from @xmath103 0 1 kpc .
thus , it appears that the radiation field does not vary enough to explain our derived limit on @xmath96h@xmath97h@xmath98 , even if all the h i gas were located near the maximum of the radiation field and all of the h ii gas were at @xmath103 1 kpc . to change our derived upper limit on @xmath96h@xmath97h@xmath98 at 100 @xmath10 from 0.4 to 1.0 , the adopted @xmath45(h@xmath0)/@xmath104h@xmath90 conversion factor would need to be changed from 1.15 to 2.9 rayleighs/@xmath55 cm@xmath4 .
this is significantly larger than the largest value of 1.9 rayleighs/@xmath55 cm@xmath4 measured by reynolds ( 1991b ) for lines of sight to four pulsars at high @xmath94 .
values obtained by arendt et al .
( 1998 ) for an additional 5 pulsar lines of sight range from 0.8 to 1.2 rayleighs/@xmath55 cm@xmath4 .
thus , it appears that there is not an error in the adopted conversion factor value that is large enough to fully explain the low emissivity derived for the ionized medium .
independent evidence that supports our adopted conversion factor comes from interstellar absorption line studies .
our adopted value corresponds to an effective electron density of 0.08 @xmath19 .
electron density estimates from observations of absorption lines of excited c@xmath89 toward extragalactic objects and high @xmath94 stars are comparable to this . in the most extensive study to date , lehner ,
wakker , and savage ( 2004 ) presented results for 43 such lines of sight at @xmath105 .
most of the observed absorption line components are at low velocity .
for these , they find a mean density of @xmath106 @xmath19 ( 1@xmath2 dispersion ) .
for the intermediate velocity arch , they find @xmath107 @xmath19 , probably lower because the gas is at higher @xmath94 ( @xmath108 kpc ) .
the derived n@xmath109 values are averages over c@xmath89 regions in both the warm ionized medium and the warm neutral medium , but lehner et al .
used wham h@xmath0 data to estimate that at least 50% of the excited c@xmath89 column density originates in the wim for an average line of sight . our derived emissivity values and cib results are subject to error if the wham h@xmath0 data are not a good tracer of far infrared emission from the wim in the regions we study .
far infrared intensity is proportional to dust column density , whereas h@xmath0 intensity is proportional to the square of the ionized gas density integrated along the line of sight , @xmath45(h@xmath0 ) @xmath110 , where @xmath111 is electron density and @xmath112 is electron temperature ( reynolds 1992 ) . thus errors would be expected in our results if the spatial variation of h@xmath0 intensity for a region is caused more by differences in mean electron density or mean electron temperature for different lines of sight than by differences in ionized gas column density .
the approximate csc @xmath113 dependence of high latitude wham data ( haffner et al . 2003 ) provides evidence that h@xmath0 is a reasonable tracer of @xmath26(h@xmath89 ) on large angular scales , but this is nt necessarily true on the scales within the regions studied here .
haffner , reynolds , and tufte ( 1999 ) and reynolds , haffner , and tufte ( 1999 ) have found evidence that variations in h@xmath0 intensity may be largely due to variations in electron density , based on their observations of h@xmath0 , [ n ii ] @xmath8 6583 , and [ s ii ] @xmath8 6716 line intensities in the region @xmath114 , @xmath115 , and previous observations of these lines in halos of edge - on galaxies .
the [ n ii]/h@xmath0 and [ s ii]/h@xmath0 intensity ratios are observed to increase with increasing @xmath94 , while the [ s ii]/[n ii ] ratio is nearly constant .
interpreting the variations in [ n ii]/h@xmath0 as primarily due to variations in electron temperature , haffner et al . and reynolds et al .
inferred @xmath112 values ranging from 6000 k to 11000 k , with temperature increasing as h@xmath0 intensity decreases .
they showed that this anticorrelation can be explained if variations in h@xmath0 are largely due to variations in mean electron density , and a supplemental source of gas heating is present that dominates over photoionization at low density , causing temperature to increase with decreasing density .
a number of possible supplemental heating mechanisms have been proposed , and reynolds et al .
estimate the heating rate that would be needed to explain the observations for each mechanism .
if h@xmath0 is not a good tracer of ionized gas column density , our method of analysis tends to underestimate the infrared emissivity per h@xmath88 ion and overestimate the cib . to show this
, we consider a simple model in which the h@xmath0 variation in a region is partly due to variation of @xmath26(h@xmath88 ) and partly due to variation of effective electron density @xmath116 , @xmath117 where @xmath118 , @xmath119h@xmath120 , and @xmath121 are averages over all lines of sight through the region .
this assumes that electron temperature is constant , and @xmath119h@xmath122h@xmath123 .
we adopt @xmath124^{p } , \ ] ] and @xmath125^{1-p } , \ ] ] where @xmath126 is the fraction of the variation of log @xmath45(h@xmath0 ) that is caused by variation of @xmath26(h@xmath89 ) .
we assume that , averaged over large angular scales , @xmath26(h@xmath89 ) and @xmath45(h@xmath0 ) are directly proportional , @xmath127 where @xmath128 is our adopted conversion factor of 1.15 rayleighs/@xmath55 cm@xmath4 .
the infrared emissivity per h nucleus is assumed to be constant within each gas phase , so the infrared emission from each phase is proportional to its gas column density .
the infrared emission from the ionized phase is @xmath129 assuming the distributions of @xmath26(h i ) and @xmath45(h@xmath0 ) are uncorrelated , the h@xmath0 coefficient @xmath50 that would be obtained from our decomposition is the mean slope of the @xmath130
@xmath45(h@xmath0 ) relation .
if the distribution of h@xmath0 intensities is symmetric about the mean , this is given by @xmath131 and the emissivity per h@xmath89 ion derived from the analysis underestimates the actual emissivity , @xmath132 thus , our derived limit @xmath86(h@xmath89)/@xmath86(h@xmath133 may be consistent with no real difference in emissivity between the ionized and neutral phases if only a small fraction of the h@xmath0 variation in our regions is caused by variation of @xmath26(h@xmath88 ) , i.e. , if @xmath134 in the context of this simple model .
berkhuijsen , mitra , and mller ( 2006 ) have estimated @xmath135 for the high latitude diffuse ionized gas , using pulsar dispersion measure data and wham h@xmath0 data toward a sample of 157 pulsars at @xmath136 ( see their figure 7b ) .
this result suggests that our low derived @xmath86(h@xmath89)/@xmath86(h@xmath87 ) is at least partly due to error in using h@xmath0 as a tracer .
the overestimate of the cib for our simple model is given by the intercept of a linear fit to the @xmath130
@xmath45(h@xmath0 ) relation . to a good approximation , this intercept is given by @xmath137 using equations ( 6 ) , ( 7 ) , and ( 8) , we obtain @xmath138 we have used this result to estimate the possible effect of emission from the wim that is not correlated with h@xmath0 or @xmath26(h i ) on the cib results derived from our three - component fits .
we subtracted @xmath139 offsets from the derived residual intensities @xmath51 for each region assuming @xmath140 , @xmath119h@xmath120 = @xmath141h@xmath142/(1.15 rayleighs/@xmath55 cm@xmath4 ) , and @xmath143h@xmath144h@xmath145 , using values of @xmath49 in table 3 for @xmath143h@xmath145 . for each wavelength
, we then calculated the weighted average of the reduced residual intensity over the five regions , treating the @xmath146 uncertainties in @xmath126 and @xmath143h@xmath147h@xmath145 as additional independent sources of error .
the resulting average residual intensities are @xmath148 , @xmath149 , @xmath150 , and @xmath151 nw m@xmath4 sr@xmath5 at 60 , 100 , 140 , and 240 @xmath10 , respectively . at 140 and 240 @xmath10 , these results are only 6% lower than the results from our three - component fits .
equation ( 11 ) is the same as the expression for the cib overestimate from a two - component fit ( using h i without h@xmath0 ) where @xmath126 is the fraction of @xmath26(h@xmath88 ) that is correlated with @xmath26(h i ) .
thus we can use the same procedure to estimate the possible effect of emission from the wim that is not correlated with @xmath26(h i ) on the cib results of hauser et al .
we make the same assumptions as in the previous paragraph for @xmath126 , @xmath119h@xmath120 , and @xmath143h@xmath90 .
the hqbs region included in the hauser et al .
analysis is below the declination limit of the wham survey , so for this region we used data from the southern h - alpha sky survey of gaustad et al .
( 2001 ) as processed by finkbeiner ( 2003 ) .
we find that the weighted - average residual intensities from the hauser et al .
analysis would be reduced to @xmath152 , @xmath153 , and @xmath154 nw m@xmath4 sr@xmath5 at 100 , 140 , and 240 @xmath10 , respectively .
these results are only about 5% lower than those of hauser et al .
the possible effect of the ionized medium is also estimated to be small for the 125 - 2000 @xmath10 cib spectrum determined by fixsen et al .
( 1998 ) from firas observations . in one of their analyses ,
galactic emission template maps were constructed from dirbe 140 @xmath10 and 240 @xmath10 data with the hauser et al .
( 1998 ) cib and zodiacal light subtracted , and the cib spectrum was obtained by correlating firas data with these templates .
our estimate of @xmath155% for the possible error of the dirbe cib results also applies to the @xmath156 part of the cib spectrum derived from this analysis . at longer wavelengths ,
the error is expected to decrease .
this is because the spectrum of emission from the ionized medium is expected to be similar to the spectrum of emission from the h i phase , and the ratio of this spectrum to the cib spectrum decreases with increasing wavelength for @xmath157 . since all three of the fixsen et al .
analyses gave consistent results for the cib spectrum , the possible error due to the wim should be small for their average cib spectrum from the three methods .
we used wham h@xmath0 data as a tracer of far infrared emission from the warm ionized phase of the ism in an effort to determine the intensity of this emission at high galactic latitudes and to assess its effect on determination of the cosmic far infrared background .
we studied five high latitude regions , including regions previously analyzed by hauser et al .
( 1998 ) and lagache et al .
( 2000 ) . for each region
, we decomposed _ cobe_/dirbe data at 60 , 100 , 140 , and 240 @xmath10 into a sum of an h i correlated component , an h@xmath0 correlated component , and a residual component .
uncertainties in our results due to omission of an h@xmath1 correlated component are expected to be negligible , based on results of a _ fuse _ high latitude h@xmath1 absorption line survey .
we found that the intensity of the h@xmath0 correlated component is consistent with zero within the uncertainties for all regions at all wavelengths . from the mean intensities of the residual components , we derived estimates of the cib at 140 and 240 @xmath10 and upper limits to the cib at 60 and 100 @xmath10 ( table 4 ) .
we repeated the analysis without including an h@xmath0 correlated component , and the derived cib results did not change significantly .
our cib estimates and upper limits are similar to previous cib determinations for which the fir emission from the ism was traced only by h i column density .
we conclude that addition of an h@xmath0 correlated component in modeling the ism emission at high galactic latitude has negligible effect on derived cib results .
we did not perform detailed anisotropy tests on the maps of residual intensity from our analysis , so our cib results do not supersede the results of hauser et al .
( 1998 ) .
we derived 2@xmath2 upper limits to the 100 @xmath10 emissivity per h@xmath89 ion for the five regions that are typically about 40% of the emissivity per h atom for the neutral atomic medium .
available evidence suggests that this low value is partly due to a lower dust - to - gas mass ratio in the ionized medium than in the neutral atomic medium , and partly due to a shortcoming of using h@xmath0 as a tracer of fir emission , which causes our analysis to underestimate the emissivity of the ionized medium .
other possible effects that may play a role include a weaker radiation field in the ionized medium than in the neutral medium , a difference in grain optical properties or grain size distribution , or an error in our adopted @xmath45(h@xmath0)/@xmath26(h@xmath88 ) conversion factor .
( the value of 100 @xmath10 emissivity per h@xmath89 ion derived by lagache et al .
( 2000 ) is much greater than the upper limit we derived for their region .
our analysis differs from theirs in that ( 1 ) we exclude positions where @xmath26(h i ) is greater than @xmath158 cm@xmath4 , where emission from h@xmath1 associated dust or 21-cm line opacity may be significant , and ( 2 ) we analyze data at a resolution of 1@xmath159 instead of 7@xmath159 . )
h@xmath0 observations have previously been used in this kind of analysis with apparent success for ionized regions that have higher density and are at low @xmath94 . however , for the general high latitude wim , evidence from reynolds , haffner , and tufte ( 1999 ) and berkhuijsen , mitra , and mller ( 2006 ) suggests that variations in h@xmath0 are not entirely due to variations in h@xmath89 column density , but are also due to differences in mean electron density for different lines of sight .
thus , h@xmath0 may not be an accurate tracer of far infrared emission from the wim .
if h@xmath0 intensity is not a good tracer , any emission from the wim that is not correlated with h@xmath0 or h i would contribute to our derived residual component for each region , so our analysis could overestimate the cib . the agreement with the hauser et al .
( 1998 ) results could be because their analysis overestimates the cib by a similar amount .
we used wham data to estimate the possible effect on our cib results and on the hauser et al .
results , assuming that the mean h@xmath0 intensity for each region can be used to estimate its mean h@xmath89 column density . in each case
, we estimated the effect to be only about 5% at 140 and 240 @xmath10 , which is much smaller than the quoted uncertainties .
the possible effect of emission from the wim is also estimated to be small for the 125 - 2000 @xmath10 cib spectrum determined by fixsen et al .
( 1998 ) from firas observations .
we estimated upper limits on a possible diffuse component of the cib by comparing the hauser et al .
( 1998 ) cib results with the integrated galaxy brightness determined by dole et al .
( 2006 ) from _ spitzer _ source counts .
we obtained 2@xmath2 upper limits on a diffuse component of 26 nw m@xmath4 sr@xmath5 at 140 @xmath10 and 8.5 nw m@xmath4 sr@xmath5 at 240 @xmath10 .
we thank f. boulanger , j. c. howk , and g. lagache for useful discussions , and g. lagache for providing us with data from the previous study of the q2 region .
we thank the referee for helpful comments .
this paper is dedicated to the memory of our friend and colleague thomas j. sodroski , and we acknowledge his contributions in the formative stages of this work .
this research was supported by nasa astrophysical data program nra 99 - 01-adp-137 .
boulanger , f. , abergel , a. , bernard , j .-
p . , desert , f .- x .
, lagache , g. , puget , j .- l . , burton , w. b. , & hartmann , d. 1995 , in unveiling the cosmic infrared background , ed . e. dwek , ( woodbury , ny : aip ) , p. 87
brodd , s. , fixsen , d. j. , jensen , k. a. , mather , j. c. , & shafer , r. a. 1997 , _ cobe _ far infrared absolute spectrophotometer ( firas ) explanatory supplement , _ cobe _ ref .
97-c ( greenbelt , md : nasa / gsfc ) , available in electronic form from the nssdc hauser , m. g. , et al .
1997 , _ cobe _ diffuse infrared background experiment ( dirbe ) explanatory supplement , ed .
m. g. hauser , t. kelsall , d. leisawitz , and j. weiland , _ cobe _ ref .
95-a ( greenbelt , md : nasa / gsfc ) , available in electronic form from the nssdc .
heiles , c. , haffner , l. m. , & reynolds , r. j. 1999 , in new perspectives on the interstellar medium ( asp conference series vol .
168 ) , eds .
a. r. taylor , t. l. landecker , & g. joncas , ( san francisco : asp ) , 211 zagury , f. , jones , a. , & boulanger , f. 1998 , in the local bubble and beyond : proceedings of the iau colloquium no .
166 , eds .
d. breitschwerdt , m. j. freyberg , & j. truemper , ( berlin : springer - verlag ) , 385 cccc dirbe zodi subtracted & @xmath160 & & hauser et al .
( 1997 ) + mission average maps & & & + & & & + leiden dwingeloo & @xmath161 & @xmath162 km s@xmath5 & hartmann & burton ( 1997 ) + h i survey & & & + & & & + lockman hole & @xmath163 & @xmath164 km s@xmath5 & snowden et al . ( 1994 ) + h i map & & & + & & & + nep h i map & @xmath163 & @xmath165 km s@xmath5 & elvis et al .
( 1994 ) + & & & + wham h@xmath0 & @xmath166 & @xmath167 km s@xmath5 & haffner et al .
( 2003 ) + sky survey & & & + ccccc 100 & @xmath168 & @xmath169 & @xmath170 & lagache et al .
( 2000 ) + 100 & @xmath171 & @xmath172 & @xmath173 & this paper + & & & & + 140 & @xmath174 & @xmath175 & @xmath176 & lagache et al .
( 2000 ) + 140 & @xmath177 & @xmath178 & @xmath179 & this paper + & & & & + 240 & @xmath180 & @xmath181 & @xmath182 & lagache et al .
( 2000 ) + 240 & @xmath183 & @xmath184 & @xmath185 & this paper + cccccccc hqbn & 60 & 1 & @xmath186 & @xmath187 & @xmath188 & 535 & 1.25 + & & 2 & @xmath189 & & @xmath190 & 535 & 1.25 + & 100 & 1 & @xmath191 & @xmath192 & @xmath193 & 535 & 2.11 + & & 2 & @xmath194 & & @xmath195 & 535 & 2.12 + & 140 & 1 & @xmath196 & @xmath197 & @xmath198 & 535 & 1.31 + & & 2 & @xmath199 & & @xmath200 & 535 & 1.32 + & 240 & 1 & @xmath201 & @xmath202 & @xmath203 & 535 & 1.12 + & & 2 & @xmath204 & & @xmath205 & 535 & 1.12 + & & & & & & & + lh & 60 & 1 & @xmath206 & @xmath207 & @xmath208 & 215 & 0.86 + & & 2 & @xmath209 & & @xmath210 & 215 & 0.91 + & 100 & 1 & @xmath211 & @xmath212 & @xmath213 & 215 & 2.20 + & & 2 & @xmath214 & & @xmath215 & 215 & 2.27 + & 140 & 1 & @xmath216 & @xmath217 & @xmath218 & 246 & 1.34 + & & 2 & @xmath219 & & @xmath220 & 246 & 1.34 + & 240 & 1 & @xmath221 & @xmath222 & @xmath223 & 246 & 1.02 + & & 2 & @xmath224 & & @xmath225 & 246 & 1.05 + & & & & & & & + nep & 60 & 1 & @xmath226 & @xmath227 & @xmath228 & 45 & 1.12 + & & 2 & @xmath229 & & @xmath230 & 45 & 1.10 + & 100 & 1 & @xmath231 & @xmath232 & @xmath233 & 45 & 2.86 + & & 2 & @xmath234 & & @xmath235 & 45 & 2.81 + & 140 & 1 & @xmath236 & @xmath237 & @xmath238 & 54 & 1.02 + & & 2 & @xmath239 & & @xmath240 & 54 & 1.04 + & 240 & 1 & @xmath241 & @xmath242 & @xmath243 & 54 & 1.10 + & & 2 & @xmath244 & & @xmath245 & 54 & 1.22 + & & & & & & & + q1 & 60 & 1 & @xmath246 & @xmath247 & @xmath248 & 217 & 2.51 + & & 2 & @xmath249 & & @xmath250 & 217 & 2.53 + & 100 & 1 & @xmath251 & @xmath252 & @xmath253 & 225 & 4.22 + & & 2 & @xmath254 & & @xmath255 & 225 & 4.23 + & 140 & 1 & @xmath256 & @xmath257 & @xmath258 & 234 & 1.33 + & & 2 & @xmath259 & & @xmath260 & 234 & 1.34 + & 240 & 1 & @xmath261 & @xmath262 & @xmath263 & 235 & 1.42 + & & 2 & @xmath264 & & @xmath265 & 235 & 1.42 + & & & & & & & + q2 & 60 & 1 & @xmath266 & @xmath267 & @xmath268 & 499 & 2.61 + & & 2 & @xmath269 & & @xmath270 & 499 & 2.60 + & 100 & 1 & @xmath271 & @xmath272 & @xmath273 & 487 & 6.63 + & & 2 & @xmath274 & & @xmath275 & 487 & 6.63 + & 140 & 1 & @xmath276 & @xmath277 & @xmath278 & 496 & 1.21 + & & 2 & @xmath279 & & @xmath280 & 496 & 1.21 + & 240 & 1 & @xmath281 & @xmath282 & @xmath283 & 497 & 1.33 + & & 2 & @xmath284 & & @xmath285 & 497 & 1.33 + cccccc h i & @xmath286 & @xmath287 & @xmath288 & @xmath289 & hauser et al .
( 1998 ) + h i & @xmath290 & @xmath291 & @xmath292 & @xmath293 & this paper + h i & h@xmath0 & @xmath294 & @xmath295 & @xmath296 & @xmath297 & this paper + h i & h@xmath0 & & @xmath298 & @xmath299 & @xmath300 & lagache et al .
( 2000 ) + lcccc hauser et al .
( 1998 ) mean residual & 20.6 @xmath7 27 & 21.9 @xmath7 6.1 & 25.0 @xmath7 6.9 & 13.6 @xmath7 2.5 + wright ( 2004 ) mean residual & -8 @xmath7 14 & 12.5 @xmath7 5 & 22 @xmath7 7 & 13 @xmath7 2.5 + residual difference ( row 1 - row 2 ) & 28.6 & 9.4 & 3 & 0.6 + zodi subtraction uncertainty & 26.7 & 6.0 & 2.3 & 0.5 + lcc integrated galaxy brightness & 13.7 @xmath7 1.7 & 10.7 @xmath7 1.4 + dirbe cib & 25.0 @xmath7 6.9 & 13.6 @xmath7 2.5 + dirbe cib transformed to firas scale & 15.0 @xmath7 5.9 & 12.7 @xmath7 1.6 + integrated galaxy brightness / cib ( dirbe scale ) & 0.55 @xmath7 0.17 & 0.79 @xmath7 0.18 + integrated galaxy brightness / cib ( firas scale ) & 0.91 @xmath7 0.37 & 0.84 @xmath7 0.15 + cib ( dirbe scale ) - integrated galaxy brightness & @xmath301 & @xmath302 + cib ( firas scale ) - integrated galaxy brightness & @xmath303 & @xmath304 + ccccccc hqbn & 48 - 145 & 60 - 75 & 0.08 & @xmath305 & @xmath306 & this paper + lh & 130 - 163 & 46 - 65 & 0.08 & @xmath307 & @xmath308 & this paper + nep & 93 - 100 & 25 - 34 & 0.08 & @xmath309 & @xmath310 & this paper + q1 & 30- 80 & 30 - 41 & 0.08 & @xmath311 & @xmath312 & this paper + q2 ( hi cut ) & 97 - 170 & 28 - 52 & 0.08 & @xmath313 & @xmath314 & this paper + q2 ( no hi cut ) & 97 - 170 & 22 - 52 & 0.08 & @xmath315 & @xmath316 & lagache et al .
( 2000 ) + _ l _ = 144 , _
b _ = -21 & 136 - 150 & 16 - 27 & 0.09 & @xmath317 & @xmath318 & arendt et al . (
1998 ) , this paper + spica & 297 - 343 & 39 - 70 & 0.6 & @xmath319 & @xmath320 & arendt et al .
( 1998 ) , this paper + eridanus superbubble & 180 - 210 & 15 - 50 & 0.8 & @xmath321 & @xmath322 & heiles et al .
( 1999 ) + barnard s loop & 208 - 217 & 13 - 27 & 2 & @xmath323 & @xmath324 & heiles et al .
( 2000 ) + outer galactic plane & 90 - 270 & 0 - 5 & 2 & @xmath325 & @xmath326 & sodroski et al .
( 1997 ) + inner galactic plane & 270 - 90 & 0 - 5 & 10 & @xmath327 & @xmath328 & sodroski et al .
( 1997 ) + atoms cm@xmath4 .
the correlation between 100 @xmath10 intensity and h i column density ( a ) is tighter than that between 100 @xmath10 intensity and h@xmath0 intensity ( b ) or that between h@xmath0 intensity and h i column density ( c ) .
no correlation is seen between h@xmath0 intensity and residual 100 @xmath10 intensity after subtraction of the h i correlated component ( d ) .
the cross plotted in the lower right of each panel shows the typical statistical measurement uncertainty ( @xmath7 1@xmath2 ) for each quantity . ] data for the q1 and q2 regions . for each region ,
the 100 @xmath10 intercept is nearly the same for the fit using h i ( top ) and the fit using h i and h@xmath0 ( bottom ) .
the scatter about the fit line is also nearly the same for the two cases . ] | determination of the cosmic infrared background ( cib ) at far infrared wavelengths using _
cobe_/dirbe data is limited by the accuracy to which foreground interplanetary and galactic dust emission can be modeled and subtracted .
previous determinations of the far infrared cib ( e.g. , hauser et al . 1998 ) were based on the detection of residual isotropic emission in skymaps from which the emission from interplanetary dust and the neutral interstellar medium were removed . in this paper
we use the wisconsin h@xmath0 mapper ( wham ) northern sky survey as a tracer of the ionized medium to examine the effect of this foreground component on determination of the cib .
we decompose the dirbe far infrared data for five high galactic latitude regions into h i and h@xmath0correlated components and a residual component .
based on _ fuse _ h@xmath1 absorption line observations , the contribution of an h@xmath1correlated component is expected to be negligible .
we find the h@xmath0correlated component to be consistent with zero for each region , and we find that addition of an h@xmath0 correlated component in modeling the foreground emission has negligible effect on derived cib results .
our cib detections and 2@xmath2 upper limits are essentially the same as those derived by hauser et al . and are given by @xmath3nw m@xmath4 sr@xmath5 ) @xmath6 75 , @xmath6 32 , 25@xmath78 , and 13@xmath73 at @xmath8 = 60 , 100 , 140 , and 240 @xmath9 m , respectively .
our residuals have not been subjected to a detailed anisotropy test , so our cib results do not supersede those of hauser et al .
we derive upper limits on the 100 @xmath9 m emissivity of the ionized medium that are typically about 40% of the 100 @xmath9 m emissivity of the neutral atomic medium .
this low value may be caused in part by a lower dust - to - gas mass ratio in the ionized medium than in the neutral medium , and in part by a shortcoming of using h@xmath0 intensity as a tracer of far infrared emission .
if h@xmath0 is not a reliable tracer , our analysis would underestimate the emissivity of the ionized medium , and both our analysis and the hauser et al .
analysis may slightly overestimate the cib .
we estimate the possible effect for the cib to be only about 5% , which is much smaller than the quoted uncertainties . from a comparison of the hauser et al . cib results with the integrated galaxy brightness from _
spitzer _ source counts , we obtain 2@xmath2 upper limits on a possible diffuse cib component that are 26 nw m@xmath4 sr@xmath5 at 140 @xmath10 and 8.5 nw m@xmath4 sr@xmath5 at 240 @xmath10 . |
since the discovery of the kerr metric @xcite there have been many attempts to find a physically meaningful matter distribution that could serve as its source ( see @xcite and references therein ) . however , no satisfactory solution has yet been found to this problem . it is the purpose of this work , to provide an interior solution to einstein equations , satisfying all the usual physical conditions , and smoothly matched on the boundary surface of the fluid distribution , to the kerr metric . with this aim
we shall generalize a procedure , recently proposed to find sources of the weyl metrics @xcite , to the stationary case . as a particular example we shall find a source for the kerr metric which consists in an anisotropic fluid , satisfying the darmois matching conditions on the boundary surface of the matter distribution , thereby excluding the presence of thin shells , and exhibiting a well behaviour of all physical variables , for a wide range of values of the parameters of the solution .
these include values which are commonly assumed in realistic models of rotating neutron stars and white dwarfs .
the line element for a vacuum stationary and axially symmetric space time , in weyl canonical coordinates may be written as : @xmath0 where @xmath1 , @xmath2 and @xmath3 are functions of their arguments . for vacuum space
times , einstein s field equations imply for the metric functions @xmath4 @xmath5 with @xmath6 and @xmath7 notice that ( [ meq1]),([meq2 ] ) are precisely the integrability condition of ( [ meq3 ] ) , that is : given any @xmath8 and @xmath9 satisfying ( [ meq1]),([meq2 ] ) , a function @xmath10 satisfying ( [ meq3 ] ) always exists .
we can write the line element above , in erez - rosen @xcite , or standard schwarzschild type coordinates @xmath11 or in spheroidal prolate coordinates @xmath12 @xcite : @xmath13 where @xmath14 is a constant which will be identified later . in terms of the above coordinates the line element ( [ 1 ] ) may be writen as : @xmath15}dr^2 + e^{-2\left[\psi-\psi^s\right]+ 2\left[\gamma(r , y)-\gamma^s\right]}r^2 d\theta^2 + e^{-2\left[\psi-\psi^s\right ] } r^2\sin^2\theta d\phi^2 , \label{exterior}\end{aligned}\ ] ] where @xmath16 and @xmath17 are the metric functions corresponding to the schwarzschild solution , namely , @xmath18 , \label{schwfun}\ ] ] where the parameter @xmath14 is easily identified as the schwarzschild mass .
we shall now assume for the interior axially symmetric line element : @xmath19 with @xmath20 where @xmath21 , and @xmath22 and @xmath23 are functions that , on the boundary surface , equal the metric functions corresponding to the schwarzschild solution ( [ schwfun ] ) , i.e. @xmath24 and @xmath25 .
also , @xmath26 and @xmath27 , where @xmath28 is an arbitrary constant and the boundary surface of the source is defined by @xmath29 the case @xmath30 , @xmath31 , corresponds to a spherically symmetric distribution , more specifically , to the well known incompressible ( homogeneous energy density ) perfect fluid sphere , and hence the matching of ( [ interior ] ) with the schwarzschild solution implies @xmath32 .
the simple condition @xmath30 recovers , of course , the static case . it should be noticed that for simplicity we consider here only matching surfaces of the form @xmath33 , of course more general surfaces with axial symmetry could however be considered as well .
we shall now turn to the matching ( darmois ) conditions @xcite .
thus the continuity of the first and the second fundamental form across the boundary surface implies the continuity of the metric functions and the continuity of the first derivatives @xmath34 , @xmath35 , @xmath36 , producing : @xmath37 where prime denotes partial derivative with respect to @xmath38 and subscript @xmath39 indicates that the quantity is evaluated on the boundary surface .
it is important to keep in mind that we are using global coordinates @xmath40 on both sides of the boundary .
thus , our line element ( [ interior ] ) matches smoothly with any stationary exterior ( [ exterior ] ) , provided conditions ( [ matchingcond ] ) are satisfied . in the particular case when we want to match our interior with the schwarzschild exterior ( the static limit ) ,
then @xmath41 and @xmath42 , and the source is a perfect fluid if @xmath43 .
we shall now see how the field equations constrain further our possible interiors .
let us first analyse the well known case when the interior is spherically symmetric , then @xmath44 , and the physical variables are obtained from the field equations for a perfect fluid , the result is well known and reads ( in relativistic units ) @xmath45 with @xmath46 , where @xmath47 and @xmath48 denote the energy density and the isotropic pressure respectively , and for the mass function @xmath49 we have @xmath50 implying @xmath51 this model , which describes the well known incompressible perfect fluid sphere , is further restricted by the requirement that the pressure be regular and positive everywhere within the fluid distribution , which implies @xmath52 .
as it is evident from ( [ eeesf ] ) the pressure vanishes at the boundary surface .
finally , if we impose the strong energy condition @xmath53 , we should further restrict our model with the condition @xmath54 .
we shall now proceed to consider the general , non spherical case .
thus , for our line element ( [ interior ] ) we have the following non vanishing components of the energy momentum tensor : @xmath55 @xmath56 , \label{pxy}\ ] ] with @xmath57 , @xmath58 , @xmath59 denotes derivative of @xmath60 with respect to @xmath61 , and @xmath62,\nonumber\\ & & \delta \hat a= \hat a^{\prime \prime}+2\frac{\hat a^{\prime}}{r}+\frac{{\hat a}_{,\theta \theta } } { r^2}+\frac{{\hat a}_{,\theta } } { r^2}\frac{\cos \theta}{\sin \theta},\nonumber \\ & & \hat p_{xx}=-\frac{{\hat a}_{,\theta } ^2}{r^2}-\frac{\hat g^{\prime}}{r}+\hat a^{\prime 2}+\frac{{\hat g}_{,\theta } } { r^2}\frac{\cos \theta}{\sin \theta}+\nonumber\\ & + & ( 1-a)\left[2 \frac{\hat a^{\prime}}{r}\frac{\sqrt{a}}{3 \sqrt{a_{\sigma}}- \sqrt{a}}- \hat a^{\prime 2 } + \frac{\hat g^{\prime}}{r}\frac{3 \sqrt{a_{\sigma}}-2 \sqrt{a}}{3 \sqrt{a_{\sigma}}-\sqrt a}\right ] , \nonumber \\ & & \hat p_{zz}=-\frac{{\hat a}^2_{,\theta } } { r^2}-\frac{\hat g^{\prime}}{r}-\hat a^{\prime 2}-\frac{{\hat g}_{,\theta \theta } } { r^2}-\hat g^{\prime \prime}+\nonumber\\ & + & ( 1-a)\left[-2 \frac{\hat a^{\prime}}{r}\frac{\sqrt{a}}{3 \sqrt{a_{\sigma}}- \sqrt{a}}+ \hat a^{\prime 2 } + 2\frac{\hat g^{\prime}}{r}\right ] .
\label{eegeneraldet}\end{aligned}\ ] ] from the expressions above , using ( [ eegeneral])-([eegeneraldet ] ) and introducing the dimensionless parameter @xmath63 , we can now obtain the explicit expressions for the physical variables , ( with @xmath64 , @xmath65 ) : @xmath66 -\frac{{\hat g_{,\theta \theta}}}{s^2}+(1 - 2a)\frac{\hat g _
{ , s}}{s } - 2(1 - 3a ) \frac{\hat a _ { , s}}{s}+\frac{2}{s^2}\left({\hat a_{,\theta \theta}}+{\hat a_{,\theta}}\frac{\cos\theta}{\sin\theta } \right ) \right\rbrace+\nonumber \\ & + & \frac{\kappa \delta}{r_{\sigma}^4}\left\lbrace\frac 34 \left(\frac{(\omega_{,y})^2}{s^4}+\frac{a(\omega_{,s})^2}{s^2\sin^2\theta}\right)+\omega\left[\frac{\omega_{,ss}a}{2s^2 \sin^2\theta } + \frac{2\omega_{,s}\hat a_{,s } a}{s^2 \sin^2\theta}+\frac{\omega_{,yy}}{2 s^4 } + 2\frac{\omega _ { , y}\hat a_{,y}}{s^4}+\frac{(1-a)\omega_{,s}}{2 s^3 \sin^2\theta}\left ( \frac{4 \sqrt{a_{\sigma}}-3\sqrt a}{3 \sqrt{a_{\sigma}}-\sqrt a}\right ) \right ] \right\rbrace,\nonumber\\ \label{tdens}\end{aligned}\ ] ] @xmath67 , \label{t03}\ ] ] @xmath68\right\rbrace+\nonumber \\ &
-&\frac{\kappa \delta}{r_{\sigma}^4}\left\lbrace\frac 14 \left(\frac{(\omega_{,y})^2}{s^4}-\frac{a(\omega_{,s})^2}{s^2\sin^2\theta}\right)\right\rbrace , \label{t11}\end{aligned}\ ] ] @xmath69 \right\rbrace,\nonumber\\ \label{t33}\end{aligned}\ ] ] @xmath70+\frac{\kappa \delta}{r_{\sigma}^5}\left[\frac{\omega_{,\theta}\omega_{,s}}{2s^4 \sin^2\theta } \right ] .
\label{t12}\ ] ] obviously for any specific ( non spherical ) model we need to provide explicit forms for @xmath71 , @xmath72 and @xmath60 , however even at this level of generality we can assure that the junction conditions ( [ matchingcond ] ) imply @xmath73 . we shall first proceed to prove the above statement , and then we shall provide a general procedure to choose @xmath71 and @xmath72 producing physically meaningful models .
it is always possible to choose the metric functions @xmath71 and @xmath72 such that , once the junction conditions ( [ matchingcond ] ) are satisfied , the angular derivatives of such functions are continuous , i.e. : @xmath74 and @xmath75 , as well as @xmath76 .
then , using @xmath77 in ( [ eegeneral ] ) , we obtain for @xmath78 on the boundary surface : @xmath79 , \label{pxxe}\end{aligned}\ ] ] where @xmath80 , @xmath81 , and dot also denotes partial derivative with respect to @xmath82 . taking into account
the einstein s vacuum equations @xmath83 , we find the following relation for the derivatives of the metric function @xmath9 , from the einstein tensor component @xmath84 : @xmath85&= & -({\psi}_{,\theta})^2 + ( { \hat{\gamma}}_{,\theta } ) \frac{\cos\theta}{\sin\theta}+ r(r-2m)\psi^{\prime 2}-\frac{m^2}{r(r-2m)}\ -(r - m)\hat{\gamma}^{\prime},\end{aligned}\ ] ] implying the vanishing of @xmath86 . in a similar way it can be shown that @xmath87 vanishes on the boundary surface , if we take into account that the einstein vacuum equation @xmath88 produces @xmath89 and from ( [ pxy ] ) , we have the following expression for @xmath87 on the boundary @xmath90
this last condition @xmath91 , which follows from the darmois conditions , and therefore is necessary , in order to avoid the presence of shells on the boundary surface , can be obtained at once from a simple inspection of the equation ( 22 ) in @xcite .
we shall provide a general procedure to choose @xmath71 , @xmath72 and @xmath60 producing physically meaningful models . with this aim , besides the fulfillment of the junction conditions ( [ matchingcond ] ) , we shall require that all physical variables be regular within the fluid distribution and the energy density to be positive . to ensure the fulfillment of the junction conditions ( [ matchingcond ] ) , we may write without loss of generality , @xmath92 where @xmath93 , @xmath94 and @xmath95 are so far two arbitrary functions of their arguments . on the other hand , to guarantee a good behaviour of the physical variables at the center of the distribution we shall demand : @xmath96 @xmath97 where ( [ pxy ] , [ eegeneraldet ] ) have been used , and
the subscript @xmath98 indicates that the quantity is evaluated at the center of the distribution . using the conditions above in ( [ ayg ] )
we may write for @xmath99 , @xmath100 and @xmath101 @xmath102 with @xmath103 and @xmath104 , as well as @xmath105
. then we can finally write for @xmath71 and @xmath72 ( please notice a that there is a misprint in eq.(43 ) in @xcite , there it should read @xmath10 instead of @xmath8 in the two terms within the round brackets for the metric function @xmath106 ) ) @xmath107 @xmath108 @xmath109 the metric functions obtained so far , satisfy the junction conditions ( [ matchingcond ] ) and produce physical variables which are regular within the fluid distribution .
furthermore the vanishing of @xmath72 on the axis of symmetry , as required by the regularity conditions , necessary to ensure elementary flatness in the vicinity of the axis of symmetry , and in particular at the center ( see @xcite , @xcite , @xcite ) , is assured by the fact that @xmath110 and @xmath111 vanish on the axis of symmetry .
furthermore , the good behaviour of the function @xmath60 on the symmetry axis is fulfilled since @xmath112 and @xmath113 vanish when @xmath114 . finally , let us note that the energy - momentum tensor components ( [ eegeneral ] ) , ( [ pxy ] ) do not diverge on the symmetry axis because the first derivative with respect to the angular variable @xmath82 of both @xmath115 and @xmath116 vanishes on the symmetry axis since not only @xmath112 and @xmath113 vanish there , but their first derivatives with respect to @xmath82 vanish as well .
so far we have presented the general procedure to build up sources for any stationary metric , in what follows , we shall illustrate the method with the example of kerr metric .
the kerr metric in weyl coordinates is given by the following metric functions @xmath117
@xmath118 @xmath119 where @xmath120 denotes the dimensionless parameter representing the angular momentum of the source , and is related to the rotation parameter @xmath121 of kerr in its well - known boyer - lindquist representation . also , @xmath122 , which in the erez - rosen coordinates becomes @xmath123 ^ 2-m^2 j^2(1-y^2).\ ] ] as is well known , the relativistic multipole moments ( rmm ) @xcite , of the kerr solution , are easily given in terms of the rotation parameter @xmath124 ( as well as @xmath121 ) .
in fact , using the fhp method @xcite in order to calculate the rmm , these are expressed in terms of the expansion coefficients ( @xmath125 ) of the ernst potential on the axis of symmetry , but the kerr solution is the only one verifying that each rmm ( @xmath126 ) at any order @xmath127 is just equal to the corresponding coefficient @xmath125 for such order , which implies , @xmath128 therefore , the massive rmm ( even orders ) and the rotational rmm ( odd orders ) can be expressed as follows @xmath129 in particular let us remind that a first conclusion derived from these rmm ( [ rmmkerr ] ) , is that the rotation of the object leads to a negative quadrupole massive moment , @xmath130 , i.e. all the possible sources of kerr solution are oblate .
let us next , consider the line element ( [ interior ] ) with metric functions given by ( [ aygdefa ] ) and ( [ aygdef ] ) @xmath131 with @xmath132 $ ]
. for the kerr solution we may write : @xmath133 } , \label{prepsimq1}\end{aligned}\ ] ] with @xmath134 @xmath135 a straightforward calculation , using ( [ tdens])([t12 ] ) allows us to find the explicit expressions for the physical variables , these are displayed in figures ( 3)(7 ) .
however , before entering into a detailed discussion of these figures , we shall carry out some calculations with the purpose of providing some information about the `` shape '' of the source .
in particular we shall see how it is related with the rotation parameter @xmath124 of the source .
for doing so , using our interior metric , we shall calculate the proper length @xmath136 of the object along the axis @xmath137 , and the proper equatorial radius @xmath138 : @xmath139 where @xmath140 are the cylindrical coordinates associated to the erez - rosen coordinates .
obviously in the spherical case ( @xmath43 ) , both lengths are identical : @xmath141 where the fact that @xmath142 has been taken into account and where @xmath143 , @xmath144 denote the lengths corresponding to the spherical case .
it would be convenient to introduce here the concept of ellipticity ( @xmath145 ) , which in terms of @xmath136 and @xmath138 , is defined as @xmath146 .
the two extreme values of this parameter are @xmath147 , which corresponds to a spherical object , and @xmath148 for the limiting case when the source is represented by a disk . in between of these two extremes we have @xmath149 for a prolate source and @xmath150 for an oblate one . in the general
( non spherical case ) we must compare function @xmath151 with @xmath152 , since @xmath153 vanishes along the axis .
it can be seen that the sign of both functions is positive , and their relative magnitudes verify the inequality @xmath154 for all values of @xmath155 in the range @xmath156 $ ] , no matter the sign of the parameter @xmath124 , leading to the well known result that the rotation of the object always generates an oblate source ( @xmath157 ) since @xmath158 . in the figure 1 ,
one example is shown . _ functions @xmath159 and @xmath160 for values of the rotation parameter @xmath161 and @xmath162 . _ ] figure 2 shows the ellipticity @xmath145 of the source as a function of the rotation parameter @xmath124 , for different values of the parameter @xmath163 . as can be seen , the relation between @xmath145 and @xmath124 for any value of @xmath163 shows that the greater is @xmath124 , the greater is @xmath145 , and therefore the shape of the source is more oblate .
of course , for @xmath164 ( static case ) we recover the sphericity ( @xmath147 ) .
it is also observed from the figure 2 that the deformation of the source with respect to the spherical case , for any fixed value of @xmath124 , is smaller for larger values of @xmath163 ( less compact object ) .
_ relation between the ellipticity of the source @xmath145 and its rotation parameter @xmath124 for different values of @xmath163 .
_ ] let us now turn back to the physical variables of our model .
figure 3 exhibits the behaviour of the radial pressure @xmath165 for different values of @xmath124 . in it
, we observe the variation of the radial pressure with respect to the spherically symmetric case ( @xmath164 ) .
this variation is smaller for angle values close to the equator , as it is apparent for @xmath166 .
notice that the radial pressure is positive , with negative pressure gradient , and vanishes on the boundary surface .
@xmath167{prr03.eps } & \includegraphics[scale=0.21]{prr1.eps } \nonumber\\ ( a ) & ( b ) \nonumber \end{array}\ ] ] figures 4 and 5 depict the behaviour of different energy momentum components , for a specific choice of the parameters @xmath168 and @xmath163 .
@xmath167{kt00.eps } & \includegraphics[scale=0.21]{kt11.eps } \nonumber\\ ( a ) & ( b ) \nonumber \end{array}\ ] ] @xmath167{kt33.eps } & \includegraphics[scale=0.21]{kt12.eps } \nonumber\\ ( a ) & ( b ) \nonumber \end{array}\ ] ] figure 6 , shows the verification of the strong energy condition @xmath169 ( also for a specific choice of the parameters @xmath124 and @xmath163 ) . _
@xmath170 $ ] as function of @xmath171 and @xmath155 , with @xmath172 and @xmath162 . _ ] figure 7 , shows the component @xmath173 . _ @xmath174 as function of @xmath171 and @xmath155 , with @xmath172 and @xmath162 . _ ]
by extending the general procedure developed in @xcite to the stationary case , we have been able to build up a physically meaningful source for the kerr metric , satisfying the matching conditions on the boundary surface of the matter distribution . in spite of the fact that a perfect fluid source for the kerr metric might exist @xcite , our source is necessarily anisotropic in the pressure .
particular attention deserves the presence of a non vanishing @xmath175 component of the energy momentum tensor . indeed , defining as usual an energy momentum flux vector as : @xmath176 ( where @xmath177 denotes the four velocity of the fluid ) , it appears that , in the equatorial plane of our system , energy flows round in circles around the symmetry axis .
this result is a reminiscence of an effect appearing in stationary einstein
maxwell systems . indeed , in all stationary einstein
maxwell systems , there is a non vanishing component of the poynting vector describing a similar phenomenon @xcite ( of electromagnetic nature , in this latter case ) .
thus , the appearance of such a component , seems to be a distinct physical property of rotating fluids , which has been overlooked in previous studies of these sources .
we have carried out a systematic search for the range of values of @xmath163 and @xmath124 for which our models exhibit acceptable physical properties .
we have focused on the fulfillment of positive energy density ( p.e.d . ) ( @xmath178 ) , strong energy condition ( s.e.c . ) ( @xmath179 ) and positive radial pressure ( p.r.p . )
( @xmath180).the results of this search are shown in table i. it appears evident that physically meaningful sources exist for a wide range of values of the parameters .
on the other hand , the chosen range of values of the parameters , incorporates values considered in the existing literature , to describe realistic models of rotating neutron stars and white dwarfs .
let us elaborate on this issue with some detail .
the rotation of the source is determined by the parameters @xmath181 of the exterior solution .
indeed , the rotation parameter @xmath182 stands for the dimensionless angular momentum of the rotating source .
so , if we restrict ourselves to a sub - extreme kerr solution ( @xmath183 ) , then @xmath184 . in @xcite numerical models of rotating neutron stars are constructed for different eoss . for each eos
the star s angular momentum , ranges from @xmath185 to the keplerian limit @xmath186 , and the dependence of the quadrupole moment on the rotation parameter @xmath124 , is established . within the interval @xmath187 to @xmath188 solar masses for the mass @xmath14 , the parameter @xmath124 is in the interval between @xmath189 and @xmath190 . in @xcite
upper limits on the parameter @xmath124 are found for representative eoss ; uniformly rotating neutron star models with maximum mass for various equations of state are studied , and the parameter @xmath124 does not exceed the value @xmath191 . in @xcite
realistic equations of state for rapidly rotating neutron stars are explored , including a wider range of values for @xmath124 .
finally , we would like to conclude with the following comment . in some static sources it may occur that @xmath158 does not imply that the source is oblate ( @xmath157 ) ( see @xcite for a discussion on this issue ) .
however , as we have seen , this is not the case of our source .
/ p.r.p .
.fulfillment ( f ) or violation ( v ) of different criteria for good physical behaviour : positive energy density ( p.e.d . ) ( @xmath178 ) , strong energy condition ( s.e.c . ) ( @xmath179 ) and positive radial pressure ( p.r.p . ) ( @xmath180 ) .
the symbol @xmath192 over f means that although the criterion is fulfilled , nevertheless @xmath193 changes its sign in the interval @xmath194 $ ] within the source .
this table corresponds to an oblate source ( due to the rotation ) of the kerr solution for different values of @xmath124 and a sequence of different values of the parameter @xmath163 .
[ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^",options="header " , ] [ data ]
this work was partially supported by the spanish ministerio de ciencia e innovacin under research projects no .
fis2015 - 65140-p ( mineco / feder ) , and the consejera de educacin of the junta de castilla y len under the research project grupo de excelencia gr234 . 88 r. p. kerr , _ phys .
lett . _ * 11 * , 237 ( 1963 ) .
w. c. hernandez , _ phys .
rev . _ * 159 * , 1070 ( 1967 ) .
j. cohen , _ j. math .
phys . _ * 8 * , 1477 ( 1967 ) .
h. wahlquist , _ phys .
rev . _ * 172 * , 1291 ( 1968 ) .
w. c. hernndez , _ phys .
_ * 167 * , 1180 ( 1968 ) .
w. israel , _ phys .
d _ * 2 * , 641 ( 1970 ) .
a. krasinski , ann . phys . * 112 * , 22 ( 1978 ) .
s. haggag and j. marek , _ il nuov .
b _ 62 , 273 ( 1981 ) .
l. herrera and j. jimenez , _ j. math .
* 23 * , 2339 , ( 1982 ) .
s. haggag , _ il nuov .
b _ 105 , 365 ( 1990 ) .
s. drake and r. turolla , _ classical quantum gravity _ * 14 * , 1883 ( 1997 )
. t. papakostas , _ j.phys : conf
* 8 * , 13 ( 2005 ) .
s. viaggiu , _ int .
j. modern .
phys.d _ , * 15*,1441 ( 2006 ) .
s. viaggiu , _ int .
j. modern .
phys.d _ , * 19*,1783 ( 2010 ) .
e. kyriakopoulos , _ int .
j. modern .
phys.d _ , * 22*,1350051 ( 2013 ) .
m. azreg - ainou , _ eur .
j.c _ , * 74 * , 2865 ( 2014 ) .
e. kyriakopoulos , _ arxiv : 1602.04067v1_. l. herrera , a. di prisco , j. ibez and j. ospino , _ phys .
d _ * 87 * , 024014 , ( 2013 ) .
h. stephani , d. kramer , m. maccallum , c. honselaers , and e. herlt , _ exact solutions to einstein s field equations _ ( cambridge university press , cambridge , england ) , ( 2003 ) , 2nd ed .
j. carot , _ classical quantum gravity _ * 17 * , 2675 ( 2000 )
. g. t. carlson , jr . and j. l. safko , _ ann . phys .
_ * 128 * , 131 ( 1980 ) .
r. geroch , _ j. math
. phys . _ * 11 * , 2580 ( 1970 ) .
r. geroch , _ j. math
. phys . _ * 11 * , 1955 ( 1970 ) .
r. geroch , _ j. math
_ , * 12 * , 918 ( 1971 ) . | a , recently presented , general procedure to find static and axially symmetric , interior solutions to the einstein equations , is extended to the stationary case , and applied to find an interior solution for the kerr metric . the solution , which is generated by an anisotropic fluid , verifies the energy conditions for a wide range of values of the parameters , and matches smoothly to the kerr solution , thereby representing a globally regular model describing a non spherical and rotating source of gravitational field . in the spherically symmetric limit
, our model converges to the well known incompressible perfect fluid solution.the key stone of our approach is based on an ansatz allowing to define the interior metric in terms of the exterior metric functions evaluated at the boundary source .
the physical variables of the energy - momentum tensor are calculated explicitly , as well as the geometry of the source in terms of the relativistic multipole moments . |
the qualitative part of one complex variable function theory concerns with topological classification of analytical and pseudoharmonic functions as well as with foliations of their level - sets .
such kind of problems was considered by s. stoilov @xcite and g. t. whyburn @xcite who introduced notions of internal and light open maps ( respectively ) which reflect certain essential topological features of analytical mappings . at the same time foliations by level sets of harmonic function on the plane
were studied by w. kaplan @xcite .
we will say that a continuous function @xmath1 with a @xmath2-dimensional foliation @xmath3 on @xmath4 if * each leaf of @xmath3 is a connected component of some level - set @xmath5 , @xmath6 ; * for each @xmath7 there are local coordinates @xmath8 in which @xmath9 and @xmath10 .
suppose @xmath3 is a one - dimensional foliation on @xmath11 with all leaves non - compact .
w. kaplan @xcite , @xcite extending old result by e. kamke @xcite proved that then there exists a continuous function @xmath12 which agrees with @xmath3 .
moreover , one can find at most countable covering @xmath13 of @xmath11 such that * each @xmath14 consists of entire leaves of @xmath3 ; * the foliation on @xmath14 is equivalent to the foliation on the plane @xmath11 or on the half - plane by parallel lines .
in other words , @xmath11 is glued of countably many strips along open boundary intervals , see figure [ fig : strips_example ] . in @xcite w. kaplan
also shown that there exists a homeomorphism @xmath15 such that the function @xmath16 is harmonic .
this result was further extended to foliations with singularities w. boothby @xcite , @xcite , m. morse and j. jenkins @xcite , m. morse @xcite .
see also @xcite , @xcite , @xcite , @xcite , @xcite , @xcite in the last twenty years the interest to the topological classification of functions on surfaces arises due to a progress in the theory of hamiltonial dynamical systems of small degrees of freedom , see e.g. a. fomenko and a. bolsinov @xcite , a. oshemkov @xcite , v. sharko @xcite , @xcite , e. polulyakh and i. yurchuk @xcite , e. polulyakh @xcite . in the present paper we will study homotopical properties of foliations having properties ( 1 ) and ( 2 ) above on arbitrary open surfaces @xmath4 .
thus such a surface is obtained from a family of strips @xmath13 glued along some boundary intervals , and therefore such surfaces will be called .
every strip admits a natural foliation by parallel lines which gives a foliation @xmath17 on @xmath4 with all leaves non - compact . we will call this foliation .
let @xmath18 be the group of all homeomorphisms of @xmath4 which maps leaves of @xmath17 onto leaves of @xmath17 , and @xmath19 be the identity path component of @xmath18 with respect to the open compact topology
. we will prove ( theorem [ th : characterzation_of_h0 ] ) that @xmath19 is contractible .
hence the homotopy type of @xmath18 reduces to the computation of the ( or ) @xmath20 of the foliation @xmath17 . as an example we characterize stripped surfaces consisting of one strip ( theorem [ th : reducing_stripped_surfaces ] )
[ def : model_strip ] a subset @xmath21 $ ] will be called a if 1 .
@xmath22 , 2 .
the intersection @xmath23 is a ( possibly empty ) union of open finite intervals with mutually disjoint closures .
for example , @xmath24 is a model strip , while @xmath25 $ ] , @xmath26 $ ] , @xmath27 are not . for a model strip @xmath28 we will use the following notation : @xmath29 connected components of @xmath30 ( resp .
@xmath31 ) will be called ( resp . ) boundary intervals .
[ def : stripped_surface ] a is the quotient space @xmath32 where * @xmath33 is a disjoint union of model strips ; * @xmath34 is a family of pairs of boundary intervals such that @xmath35 , @xmath36 and @xmath37 for @xmath38 ; * @xmath39 , @xmath40 , is an * affine * homeomorphism preserving or reversing orientations .
thus a stripped surface is a surface obtained from a family model strips by identifying some pairs of boundary intervals via affine homeomorphisms .
it is allowed that two strips are glued along more than one pair of boundary components .
one may also glue together intervals belonging to the boundary of same strip @xmath28 , and even to the same lower or upped part of @xmath41 .
notice that if @xmath42 and @xmath43 , then there are exactly two _ affine _ homeomorphisms @xmath44 such that @xmath45 preserves orientation and @xmath46 reverses it .
namely , @xmath47 for @xmath48 .
the assumption that the gluing maps are affine is technical and not crucial , however it will be essentially used in the proof of lemma [ lm : hofzpr_is_contractible ] .
let @xmath4 be a stripped surface , defined by , and @xmath49 be the quotient map .
then @xmath4 is a non - compact two - dimensional manifold which can be nonconnected , non - orientable , and have boundary .
every connected component of @xmath50 is an interval .
also notice that a subset @xmath51 is open if and only if @xmath52 is open in @xmath53 . for each @xmath54 let @xmath55 be the composition of the inclusion of @xmath14 into @xmath56 with the quotient map @xmath57 .
we will call @xmath58 a map corresponding to @xmath14
. it will also be convenient to use the following notation : @xmath59 in particular , the image @xmath60 will be called a of @xmath4 .
notice that if @xmath58 is not an embedding , then @xmath61 and @xmath62 may intersect . on the other hand , the assumptions ( b ) and ( c ) guarantee that both restrictions @xmath63 and @xmath64 are injective .
notice that each model strip @xmath28 admits a one - dimensional foliation @xmath65 whose leaves are connected components of @xmath41 and sets @xmath66 , @xmath67 .
more generally , let @xmath4 be a stripped surface . since its strips are glued by homeomorphisms of leaves , the foliations on strips of @xmath4 yield a foliation @xmath17 on @xmath4 .
we will call this foliation .
let @xmath68 be the space of leaves endowed with the corresponding quotient topology , and @xmath69 be the quotient map .
then by definition a subset @xmath70 is open if and only if its inverse image @xmath71 is open in @xmath4 .
thus open subsets of @xmath72 can be regarded as open saturated subsets of @xmath4 .
@xmath72 is a @xmath73-space .
we should prove that every one - point set @xmath74 is closed in @xmath72 , i.e. that each leaf @xmath75 of @xmath3 is closed in @xmath4 . since @xmath76 , see , we should check that every leaf in a model strip is closed .
but the latter is evident for leaves belonging to interiors of strips @xmath77 and follows from ( 2 ) of definition [ def : model_strip ] for leaves belonging to @xmath78 .
a homeomorphism @xmath79 between stripped surfaces will be called an whenever it maps leaves of @xmath17 onto leaves of @xmath80 .
evidently , for each leaf @xmath75 we have exactly one of the following possibilities . 1
. @xmath75 belongs to @xmath81 for some @xmath82 ; in this case @xmath75 will be called .
2 . @xmath83 for some @xmath82 ; in this case @xmath75 will be called a leaf .
3 . @xmath84 for some @xmath85 and @xmath86 .
then @xmath87 for some @xmath40 .
this situation splits into the following three cases : * @xmath88 , @xmath89 , and @xmath90 ; so @xmath91 ; * @xmath92 , @xmath89 , and @xmath90 , so @xmath93 ; * @xmath94 or @xmath95 , in this case @xmath75 will be called .
our aim is to reduce the situation to the case when there is no leaves of types ( c1 ) and ( c2 ) . for this
we need the following lemma .
[ lm : reducing_strips ] let @xmath96 and @xmath97 \cup [ 1,+\infty ) \bigr)\times0 $ ] .
then there exists a homeomorphism @xmath98 fixed outside @xmath99 and preserving horizontal lines , that is @xmath100 for some continuous function @xmath101 , see figure [ fig : reducing_c3 ] .
let @xmath102 be a @xmath103-diffeomorphism given by @xmath104 .
evidently , @xmath105 .
define @xmath98 by the formula : @xmath106 , \\
\bigl(\sigma(x),\ , y\bigr ) , & y\not=0 , \
x \in [ -\frac{1}{|y| } , \frac{1}{|y| } ] , \\ \medskip \bigl(\sigma'(\tfrac{1}{|y|})\cdot ( x - \tfrac{1}{|y| } ) + \sigma(\tfrac{1}{|y|})\ , , \
y \bigr ) , & y\not=0 , \ x \in ( \tfrac{1}{|y| } , + \infty ) .
\end{cases}\ ] ] the verification that @xmath107 is indeed a homeomorphism having the corresponding properties we leave for the reader .
[ cor : reducing_infinite_strips ] let @xmath108 and @xmath109 \cup [ 1,+\infty ) \bigr)\times \{2n\}_{n\in{\mathbb{z}}}$ ] .
then there exists a homeomorphism @xmath98 fixed outside @xmath110 and preserving horizontal lines , that is @xmath100 for some continuous function @xmath101 , see figure [ fig : reducing_c3_infinite ] .
[ rem : eliminatinc_c3_leaf]notice that @xmath111 from lemma [ lm : reducing_strips ] is a stripped surface glued of two strips @xmath112 and @xmath113 so that @xmath114 is a leaf of type ( c2 ) , see figure [ fig : reducing_c3 ] . then lemma [ lm : reducing_strips ] claims that these two strips can be replaced with a unique one so that the corresponding leaf of type ( c2 ) becomes internal , i.e. having type ( a ) .
moreover , corollary [ cor : reducing_infinite_strips ] shows that even infinite sequence of strips glued in a way shown in figure [ fig : reducing_c3_infinite ] can also be replaces with one open strip .
let @xmath28 be a model strip such that @xmath115 and @xmath116 , see figure [ fig : strip_rectangle ] .
define the following two homeomorphisms @xmath117 by @xmath118 for all @xmath119 .
let also @xmath120 , @xmath121 be the quotients of @xmath28 obtained by identifying its boundary components via @xmath122 and @xmath123 respectively .
it follows from lemma [ lm : reducing_strips ] that @xmath124 is an open cylinder and @xmath125 is an open mbius band .
moreover , the leaf obtained by gluing @xmath31 with @xmath30 is of type ( c1 ) .
[ lm : reducing_c2 ] if @xmath4 has a leaf of type ( c1 ) , then it is @xmath3-homeomorphic either with @xmath124 or with @xmath125 .
suppose @xmath126 is a leaf of type ( c1 ) for some @xmath82 .
let also @xmath127 be the boundary components of @xmath128 , so @xmath129 since @xmath126 , it follows that @xmath130 but @xmath128 is an open closed subset of @xmath131 , whence from the definition of factor topology on @xmath4 it follows that @xmath132 is open closed in @xmath4 , and so it coincides with @xmath4 .
thus @xmath4 is obtained by gluing @xmath133 with @xmath134 by an affine homeomorphism @xmath135 .
if @xmath135 preserves orientation then @xmath4 is homeomorphic with @xmath124 . otherwise , @xmath4 is homeomorphic with @xmath125
. a stripped surface will be called if it has no leaves of types ( c1 ) and ( c2 ) that is every leaf of type ( c ) is of type ( c3 ) .
[ th : reducing_stripped_surfaces ] every connected stripped surface @xmath4 with countable base is @xmath3-homeomorphic either to a cylinder @xmath124 or to a mbius band @xmath125 , or to a reduced surface .
since @xmath4 has countable base , it follows that the number of strips in @xmath4 is at most countable .
we will now define a certain graph @xmath136 in the following way : * the vertices of @xmath136 are strips of @xmath4 containing leaves of types ( c1 ) or ( c2 ) ; * if there exists a leaf @xmath75 of type ( c1 ) such that @xmath126 for some @xmath137 , then we assume that @xmath75 is a loop at vertex @xmath132 ; * two such vertices @xmath132 and @xmath138 are connected by an edge in @xmath136 if and only if there exists a leaf @xmath75 of type ( c2 ) such that @xmath139 or @xmath140 .
it follows that each vertex of @xmath136 has degree either @xmath2 or @xmath141 .
therefore every connected component of @xmath136 has one of the forms ( i)-(v ) shown in figure [ fig : graph_c3 ] .
suppose @xmath136 is non - empty and let @xmath142 be a connected component of @xmath136 .
consider the cases ( i)-(v ) of figure [ fig : graph_c3 ] . *
( i ) * suppose @xmath142 is a finite non - closed path consisting of @xmath143 edges .
as noted in remark [ rem : eliminatinc_c3_leaf ] each leaf of type ( c2 ) can be regarded as an internal one after changing partition of @xmath4 by strips without changing foliation @xmath17 .
this means that each individual edge of @xmath136 with distinct ends can be reduced .
applying lemma [ lm : reducing_strips ] @xmath143 times one can completely reduce all edges of @xmath142 . *
( ii ) * suppose @xmath142 is a loop , so its edge corresponds to a leaf of type ( c1 ) .
then by lemma [ lm : reducing_c2 ] @xmath4 is @xmath3-homeomorphic either with @xmath124 or with @xmath125 . *
( iii ) * suppose @xmath142 is a cycle of @xmath143 edges .
then by arguments similar to ( i ) the situation reduces to the case ( ii ) . *
( iv ) * suppose @xmath142 is an infinite in one direction path .
then the arguments of the case ( i ) can not be applied as it requires to consider infinite sequence .
let @xmath144 be an infinite sequence of strips in @xmath4 corresponding to vertices of @xmath142 such that @xmath145 , @xmath146 is a leaf of type ( c2 ) being an edge between vertices @xmath147 and @xmath148 in @xmath142 .
interchanging @xmath149 with @xmath150 one may assume that @xmath151 notice that each @xmath152 is embedded into @xmath4 .
denote @xmath153 , so @xmath154 is omitted .
then it follows from corollary [ cor : reducing_infinite_strips ] that @xmath111 is @xmath3-homeomorphic to a model strip having only one boundary component .
@xmath155 splits @xmath4 so that @xmath156 is a connected component of @xmath157 .
it suffices to show that @xmath156 is open and closed in @xmath157 .
\1 ) first we will check the .
let @xmath158 .
\a ) if @xmath159 for some @xmath160 , then @xmath161 is an open neighborhood of @xmath162 in @xmath111 .
\b ) if @xmath163 , @xmath160 , then @xmath162 has an open neighborhood @xmath164 intersecting @xmath165 and @xmath166 only , and so @xmath167 . thus @xmath156 is open in @xmath157 .
\2 ) now let us show that .
let @xmath168 be a sequence converging to some @xmath169 .
we should prove that @xmath158 .
consider two cases .
\a ) suppose @xmath170 for some @xmath171 .
since @xmath172 is an open neighborhood of @xmath162 in @xmath157 , it follows that @xmath173 for some @xmath174 . but
@xmath175 for some @xmath160 , whence @xmath176 as well .
\b ) suppose @xmath177 for some @xmath178 .
then @xmath162 has a neighborhood @xmath164 intersecting @xmath179 and @xmath180 only .
take @xmath181 .
then @xmath182 which implies that one of these strips coincides with @xmath183 for some @xmath160 .
but @xmath184 intersects only @xmath185 and @xmath186 for @xmath187 .
hence @xmath179 and @xmath180 are contained in @xmath111 , and so @xmath188 as well .
it follows from this lemma that one can replace @xmath111 with a model strip , and so the situation reduces to the case when @xmath142 consists of a unique edge with distinct vertices . then by the case ( i ) it can be completely eliminated . *
( v ) * finally suppose that @xmath142 is an infinite in two directions path . by arguments of
( iv ) one can replace each of infinite ends of @xmath142 with a model strip , and then by ( i ) eliminate all edges of @xmath142 .
this also implies that @xmath4 is @xmath3-homeomorphic with an open model strip @xmath24 . since each strip @xmath179 may correspond to at most one connected component of @xmath136
, it follows that one can apply cases ( i)-(v ) mutually to each of the connected components of @xmath136 .
this allows to eliminate all leaves of type ( c2 ) or prove that @xmath4 is @xmath3-homeomorphic either with @xmath124 or with @xmath125 .
let @xmath4 be a stripped surface . denote by @xmath18 the group of @xmath3-homeomorphisms of @xmath4 preserving @xmath17 , that is for each @xmath189 and each leaf @xmath190 the image @xmath191 is a leaf of @xmath17 as well .
[ lm : homeo_of_model_strips ] let @xmath21 $ ] be a model strip and @xmath197 .
then @xmath198 where @xmath199\to[-1,1]$ ] is a homeomorphism , and @xmath200 is a continuous function such that for each @xmath201 the correspondence @xmath202 is a homeomorphism @xmath203 .
since @xmath204 preserves leaves of @xmath65 , i.e. the lines @xmath205 , @xmath201 , it follows that @xmath206 does not depend on @xmath162 .
moreover , as @xmath204 homeomorphically maps leaves onto leaves , the map @xmath207 is a homeomorphism @xmath208 .
finally , @xmath209 is a strictly monotone surjective continuous function , for each @xmath201 .
therefore it extends to a self - homeomorphism of @xmath210 $ ] .
[ lm : lifting_of_strip_homeo ] let @xmath211 be a strip containing leaves from @xmath195 , @xmath212 be the corresponding chart , and @xmath189 . if @xmath213 , then @xmath107 lifts to a homeomorphism @xmath214 of the model strip @xmath28 such that @xmath215 .
let @xmath212 be a strip of @xmath4 and @xmath216 . by assumption @xmath213 , whence by lemma [ lm : lifting_of_strip_homeo ] @xmath217 lifts to a self - homeomorphism @xmath214 such that @xmath218 .
moreover , by lemma [ lm : homeo_of_model_strips ] and assumption that @xmath107 preserves leaves with their orientations we have that @xmath219 where the correspondence @xmath202 , @xmath220 $ ] , is a self homeomorphism of @xmath221 preserving orientation . we will show that formulas for @xmath136 on distinct strips agree with _ affine _ gluing maps @xmath226 for all @xmath40 .
more precisely , let @xmath212 and @xmath227 be two strips of @xmath4 .
for the convenience of notation assume that @xmath228 be two boundary components glued via an affine homeomorphism @xmath229 .
let also @xmath214 and @xmath230 be the corresponding liftings of @xmath217 and @xmath231 respectively , and @xmath136 and @xmath232 the corresponding isotopies for @xmath204 and @xmath233 given by .
we have to prove that for each @xmath234 $ ] the following commutative diagram holds true : @xmath235 this diagram trivially holds for @xmath236 when @xmath237 and @xmath238 are identity maps .
moreover it also holds for @xmath239 when @xmath240 and @xmath241 , since @xmath107 agrees on @xmath242 , so @xmath243 we can assume that @xmath244 for some @xmath245 , and continuous functions @xmath246 and @xmath247 . for simplicity we will also omit the second coordinate @xmath248 .
then @xmath249 and so we get from that @xmath250 hence @xmath251 thus isotopies agree on distinct model strips , and therefore they yield a unique isotopy between @xmath107 and @xmath194 in @xmath192 .
one can easily check that such an isotopy is continuous in @xmath252 and so it yields a contraction of @xmath192 .
* @xmath254 for all @xmath171 ; * if @xmath255 , @xmath256 , @xmath171 , is a unique lifting of @xmath257 given by , then @xmath206 is increasing and @xmath258 also increasing for each fixed @xmath201 .
* @xmath107 leaves invariant each leaf @xmath259 and preserves its orientation .
\(c ) since @xmath268 and every leaf @xmath259 is a of @xmath195 , it follows that @xmath269 moreover , the restriction of @xmath270 is an isotopy between @xmath271 and @xmath272 , whence @xmath272 preserves orientation .
\(a ) it follows further , that @xmath107 also leaves invariant every connected components of @xmath273 .
but every such component is the interior of some strip , whence @xmath254 for each @xmath171 , which completes ( a ) .
\(b ) moreover , let @xmath274 be a lifting of @xmath275 .
then @xmath276 , where @xmath277 and @xmath278 are continuous in @xmath279 .
moreover , @xmath280 is an isotopy between @xmath281 and @xmath282 . hence @xmath206 is increasing .
let @xmath287 , @xmath171 , and @xmath255 be a lifting of @xmath107 , see lemma [ lm : lifting_of_strip_homeo ] .
so by ( c ) @xmath288 where then @xmath206 is increasing and @xmath289 is increasing in @xmath162 for each fixed @xmath201 . define an isotopy @xmath290 by @xmath291 then @xmath292 is fixed on @xmath293 , @xmath294 and @xmath295 preserves each leaf of @xmath296 with its orientation .
hence the family of isotopies @xmath297 yield an isotopy @xmath298 of @xmath107 to a diffeomorphism which preserves each leaf of @xmath17 with its orientation . in other words , @xmath299 , @xmath300 for all @xmath264 , and @xmath301
. one can easily check that the map @xmath302 given by @xmath303 is continuous .
moreover , if @xmath216 , then in @xmath304 , whence @xmath305 , and so @xmath306 for all @xmath264 . in other words , @xmath307 is a deformation fixed on @xmath192 , whence @xmath192 is a strong deformation retract of @xmath286 .
s. stoilov , _ lectures on topological principles of the theory of analytic functions _ , translated from the french by e. t. stekina . with a foreword by b. v .
sabat , izdat .
`` nauka '' , moscow , 1964 ( russian ) . | we study non - compact surfaces obtained by gluing strips @xmath0 with at most countably many boundary intervals along some these intervals .
every such strip possesses a foliation by parallel lines , which gives a foliation on the resulting surface .
it is proved that the identity path component of the group of homeomorphisms of that foliation is contractible . |
rrently deployed wired and wireless communication systems employ half duplex .
namely , either frequency division duplex ( fdd ) or time division duplex ( tdd ) has been used for separate transmission and reception . in fdd and tdd , transmitted signal does not interfere with received signal due to orthogonal use of time / frequency resources for transmission and reception . since two orthogonal channels are needed in half duplex systems , twice of time and/or frequency resources are required in half duplex compared to full duplex systems .
it is clear that the capacity can be doubled by simultaneous transmission and reception in the same frequency at the same time .
str or full duplex systems not only improve the physical layer capacity but also provide other important benefits in layers beyond physical layer @xcite .
for example , str can reduce or completely eliminate hidden node problem which is typical issue in csma networks such as wireless local area networks ( wlan ) . when a node receives a packet designated to it and meanwhile has a packet to transmit , by having str capability it can transmit the packet while receiving the designated packet .
this not only provides twice throughput but also enables hidden nodes to better detect active nodes in their neighborhoods . on the other hand ,
when the node has no packet to send , it can transmit a dummy signal so that any hidden node can detect the activity in its vicinity and realize that the channel is in use .
another benefit of str is significant reduction in end - to - end delay in multi - hop networks @xcite . in half duplex systems ,
each node can start transmission of a packet to the next node only when it is fully received from the previous node in network .
therefore , the end - to - end delay is equal to packet duration multiplied by the number of hops .
however , when str is employed , a node can forward a packet while receiving it , and consequently the end - to - end delay in str systems can be just a bit longer than the packet duration .
this will be a huge advantage over half duplex systems especially as the number of hops grows .
meanwhile , the forwarded packet to next node can play a role of implicit acknowledgement ( ack ) to previous node as well .
interesting application of str includes channel sensing in cognitive radio systems @xcite . in cognitive radios ,
active secondary users have to release the spectrum when primary users start their transmissions . without str capability , it is a challenge for secondary users to detect activity of primary users while they are using the spectrum for their own communications . however , str enabled secondary users can scan the activities of primary users frequently ( even as they transmit ) and stop their transmissions immediately once they detect primary users transmissions .
likewise , str makes the device discovery easier in device - to - device ( d2d ) systems .
this is due to the fact that in d2d systems , when user equipment ( ue ) has str capability , it can discover neighboring ues easily by monitoring ul signals from proximate ues without stopping its own ul transmission .
it is interesting to note that str techniques can be used for interference cancellation in co - existence of multiple radios in the same device .
multiple radios such as wlan , bluetooth , gps receiver , 2 g and 3 g cellular transceivers are put into the same device particularly a small handheld device type @xcite .
although those radios operate at different rf carriers , due to proximity of transceivers in the same device , they can still interfere with each other .
this interference can be treated as echo since the transmitted signal from a radio is arrived at other radios in the same device .
hence , by using the proposed technique in this paper , the co - existence issues in the same device can be resolved .
multiband support requires large number of switched duplexers in fdd , resulting in quite complicated rf architecture , increased cost and form factor @xcite , @xcite .
multiple input multiple output ( mimo ) and carrier aggregation techniques aggravate the situation .
however , by str techniques duplexer free systems are possible by cancelling transmitted signal appeared at receive band .
despite all the advantages mentioned , implementation of str has its own challenges .
the main difficulty in str system is echo cancellation at receive chain .
the strongest echo is introduced to the system when transmitted signal is leaked into receive chain through circulator .
this causes a large interference to the desired received signal as the echo power level is much higher than desired received signal .
for example , with 46 dbm output power at power amplifier ( pa ) , assuming 25 db isolation at circulator , the echo power will be 21 dbm .
in addition to the aforementioned echoes , echoes can be caused by impedance mismatch at antenna .
besides , transmitted signal can bounce off objects such as buildings and mountains , and return to the antenna as echoes at receive chain . hence , without echo cancellation , the received signal can not be decoded .
furthermore , in order to avoid saturation at lna and high resolution adc , echo cancellation should be performed before lna .
therefore , it is imperative to cancel echoes in analogue domain in order for str systems to be commercially deployable .
an off - the - shelf hardware @xcite is available for rf interference cancellation in analogue and can be used for echo cancellation in rf .
however , it is a narrowband interference canceller and performs interference cancellation by adjusting phase and magnitude of interference .
thus , the performance of echo cancellation in wideband signals is very limited . in @xcite
about 20 db echo suppression is reported for 5 mhz ieee 802.15.4 signals .
in addition , this narrowband interference canceller can create higher sidelobe power than in - band interference power @xcite .
an str implementation in analogue assuming separate transmit and receive antennas was reported in @xcite claiming 37 db suppression by the rf echo canceller itself .
recently , @xcite presented an open - loop technique for full duplex mimo . a hardware based experiment shows 50 db of echo suppression . in @xcite , the echo cancellation using transmit beamforming is studied .
this technique uses two transmit antennas and one receive antenna .
the additional two transmit antennas create a null toward receive antenna , resulting in reduced echo power at receive chain .
however , the transmitted signal is not omni - directionally transmitted due to the transmit beamforming .
when the intended receiver is located in the null direction , signal - to - noise power ratio ( snr ) loss is inevitable .
additionally , the transmit beamforming is narrowband beamforming .
thus , it is not suitable for wideband signal .
since this technique needs separate transmit and receive antenna set , increased form factor is unavoidable .
nonetheless , the echo suppression of 20 @xmath0 30 db by this technique is not promising .
however , with the aids from digital baseband cancellation and the narrowband noise canceller @xcite , the imperfect prototype achieves 1.84 times of throughput compared to half duplex system throughput .
a scalable design to mimo using extra antennas with special placement antennas is proposed in @xcite . in @xcite , assuming variable delay line , one - tap echo canceller is suggested .
it is shown that 45 db suppression of echo is achieved by using heuristic approach in adaptive parameter adjustment .
the throughput measurements in @xcite show 111% gain in downlink .
another prior work @xcite proposed to cancel echo before lna .
firstly , to avoid saturation at lna , separate transmit and receive antennas with large displacement are used .
various antenna orientations are investigated in @xcite .
the receiver estimates the echo channel at every subcarrier using ofdm signal . using another transmit radio chain , an echo cancelling ofdm signal
is generated based on the estimated channels of echo path and the additional radio , and added to the receive chain before lna .
the measured echo cancellation is about 31 @xmath0 32 db in @xcite and 24 db in @xcite .
this type of open - loop technique is directly sensitive to impairments as it relies on high accurate echo channel estimation . since the suppression in this case is not sufficient to realize the full duplex gain , other mechanisms such as digital domain cancellation and separate transmit / receive antenna have to be incorporated . without using additional antennas
, we propose an adaptive echo cancellation technique that can be implemented in analogue domain for wideband signals .
the proposed technique is a closed - loop technique which provides more robustness to impairments than open - loop techniques .
furthermore , the proposed technique is scalable to any mimo system . as mentioned earlier , with perfect echo cancellation , str systems
can achieve the doubled capacity especially in isolated links such as point - to - point communications and wireless backhaul .
however , in cellular systems the situation is different .
in addition to the regular co - channel interference present in half duplex systems , namely base station ( bs ) to ue and ue to bs interferences , there are two unique and serious interferences caused by system operation in full duplex mode : one is _
bs - bs _ and the other is _ ue - ue_. on one hand , due to str at bss , neighboring bss dl signals interfere with desired ul signal at home bs .
this is called _ bs - bs interference _ and is extremely severe .
firstly , unlike bs to ue channel ( downlink or uplink ) , bs - to - bs channel is closer to line - of - sight ( los ) with much smaller path loss . secondly , the transmit power and antenna gain at bs is much larger than those of ue .
hence , the interferences from neighboring bss to home bs easily dominate desired weak ul signal .
hence , without cancelling bs - bs interferences , ul communication is impossible . in this paper
we suggest a solution for bs - bs interferences and provide system level evaluations for str systems with and without bs - bs interference cancellation . on the other hand , in str systems , ul signal transmitted by a ue
creates interference to dl signals received at other ues _
i.e. _ , dl signal will be corrupted by proximate ul signals .
this is called _ ue - ue interference _ and results in loss of dl capacity .
unfortunately , with simple transceiver and omni - directional single antenna at ue , no fancy transmit and receive beamforming can be utilized in ue to handle ue - ue interference .
therefore , one option to confront ue - ue interference is to coordinate between neighboring bss such that scheduler avoids scheduling proximate ue pairs that cause serious ue - ue interference to each other .
however , in this paper we propose a non - cooperative technique which , despite its simplicity , can hugely reduce ue - ue interference and greatly improve dl capacity . unlike cellular systems ,
the aforesaid bs - bs and ue - ue interferences are absent in csma network such as wlan systems even when multiple co - channel cells are deployed .
the reason is that due to csma protocol , access points ( aps ) or terminals will hold their transmissions if they hear any transmission from other proximate aps or terminals .
in fact owing to str , the hidden node issue can be easily solved .
of course collisions still can happen between nodes due to other factors such as propagation delay and some delay in header decoding .
however , this is common problem even without employing str . in this paper
we investigate a few protocols and demonstrate how str can reduce the hidden node problem .
we show that the throughput is improved even with the assumptions of asynchronous packet arrival and some delay in decoding header or decision on channel activity .
unlike @xcite where detail issues such as variable packet length , fairness and scheduler design are addressed , we study the hidden and exposed node issues in simplified model of csma which renders comparisons with theoretical throughput expression .
the paper is organized as follows : in section ii analogue domain echo cancellation technique is described together with performance evaluations . in section iii , the application of str to csma network for reduction of hidden node problem is discussed and throughput improvements are demonstrated with multiple cell simulations .
section iv presents the solutions for bs - bs and ue - ue interferences .
finally , conclusions are provided in section v.
echo cancellation is a well - known topic studied extensively in the literature and proved technology in the field . despite the comprehensive investigations ,
the derivation of algorithms and implementations are performed in digital baseband . however , as described previously in section i , the echoes should be cancelled before lna _
i.e. _ in analogue domain .
implementing digital signal processing techniques in analogue domain is very challenging due to rf impairments and the fact that only limited mathematical operations are allowed in analogue domain . in this section ,
we derive echo cancellation techniques in analogue domain and show that our method can cancel echoes very effectively even with rf impairments .
the transmitted signal in passband can be written as @xmath1 where @xmath2 represents the real part , @xmath3 is the carrier frequency in radian / sec , and @xmath4 and @xmath5 are baseband in - phase and quadrature phase signals , respectively . without loss of generality , we assume one echo with unknown gain and delay . then , the received signal @xmath6 at receive chain can be modeled as @xmath7 where @xmath8 and @xmath9 are the unknown gain and delay of the echo , respectively , @xmath10 is the desired received signal and @xmath11 is additive white gaussian noise ( awgn ) . in this paper
we cancel the echo in passband by creating estimated echo signal using multiple replica of @xmath12 with different delays and subtracting it from the signal corrupted by echo .
for instance , the echo can be estimated by a linear combination of @xmath13 and @xmath14 where @xmath15 is the delay of @xmath16-th tap . in order to rotate the phases of @xmath13 and @xmath14
, we also need to use the hilbert transform of @xmath13 and @xmath14 .
then , the estimated echo using two taps can be written as @xmath17 where @xmath18 and @xmath19 are in - phase and quadrature component of @xmath16-th tap weight , and @xmath20 is the hilbert transform of @xmath12 .
then , the echo canceller output can be expressed as @xmath21 in order to simplify derivation of the weights @xmath18 and @xmath19 , we transform the passband model to equivalent complex baseband model .
the transmitted complex baseband signal @xmath22 is defined by @xmath23 then , from , the received signal in complex baseband is in the form of @xmath24 where @xmath25 and @xmath26 are complex baseband versions of desired received signal and awgn , respectively .
likewise , we can develop the complex baseband model of the echo canceller output from ( [ e - t0 ] ) and ( [ cancelloutput0 ] ) as follows @xmath27 after some manipulations , it is not difficult to show that @xmath28 .
for the concise representation , the echo canceller output can be written in vector form as @xmath29 where superscript @xmath30 represents transpose , @xmath31^t$ ] and @xmath32^t$ ] .
although it is not possible to implement wiener solution in analogue domain , minimum mean squared error ( mmse ) solution provides valuable insights on the performance and design parameters . minimizing the power of the echo canceller output means that the echo is cancelled
this is due to the fact that unknown desired receive signal @xmath10 and awgn can not be cancelled by linear combination of @xmath33 as @xmath33 is uncorrelated with @xmath10 and noise .
therefore , we define the cost function which minimizes the power of the echo canceller output as @xmath34 in appendix a the derivation of wiener solution in analogue domain is presented .
echo cancellation performance is shown in fig .
[ fig_mse ] where the suppression level is defined by @xmath35 the suppression is not a function of the carrier frequency and echo power as shown in appendix a. rather , it is a function of delay difference between taps and the number of them .
1 ( a ) exhibits suppression as a function of tap delay difference normalized by signal bandwidth ( b ) assuming that the echo delay is in the middle of the first and second tap delay . as it can be observed from the figure ,
smaller normalized tap delay difference @xmath36 and larger number of taps provide better echo suppression .
even with two taps , the suppression level close to 90 db can be achieved with the normalized tap delay difference equal to 0.01 which means 1 nsec of delay difference between taps in 10 mhz bandwidth systems .
this is equivalent to 100 times oversampling in digital signal processing .
it is apparent that as the tap delay difference goes to zero the suppression level goes to infinity ( _ i.e. _ no residual echo ) .
we emphasize that having a small tap delay difference is not a challenging issue .
for example , assuming the speed of electromagnetic ( em ) wave in transmission line is equal to the speed of light , @xmath37 difference in length will yield @xmath38 pico second delay difference . in fig .
[ fig_mse ] ( b ) the effect of echo delay on echo canceller performance is illustrated assuming that the normalized tap delay difference is 0.1 . in reality
the delay of echo is unknown .
however , typical value or range of values of circulator delay can be measured .
interestingly , when the tap delay coincides with the true delay of the echo , even with only two taps suppression beyond 150 db can be easily achieved .
the reason that even when @xmath9 is equal to either @xmath39 or @xmath40 the suppression is not infinity is that @xmath41 is not approaching zero .
it is simple to show that when the delay of either taps coincides with the echo delay , one tap echo canceller instead of two tap echo canceller has infinity suppression level . in two tap wiener filter , we study the impact of number of echoes under the condition of same total echo powers .
assume one echo having delay @xmath9 located in the middle of neighboring two taps @xmath42 where @xmath43 . without loss of generality assume @xmath44 .
let us add one more echo with the same power as the first echo but closer to the second tap with arbitrary phase @xmath45 and delay @xmath46 where @xmath47 and @xmath48 .
then , the received echo signal becomes @xmath49 the power of @xmath50 is @xmath51 depending on the phase @xmath52 , the new echo can reduce total echo power .
however , we consider the worst case _
i.e. _ @xmath53 .
note @xmath54 due to the assumption @xmath55 . for fair comparison ,
we normalize @xmath50 such that maximum power of @xmath50 is equal to the power of @xmath56 @xmath57 where @xmath58 .
then , @xmath56 and @xmath50 have the same power but @xmath56 has one echo in the middle of taps while @xmath50 contains two echoes : one in the middle of the taps and the other closer to either taps .
the cross correlation vector can be written as @xmath59 where @xmath60 and @xmath61 are defined in appendix a and @xmath62.\ ] ] it is not difficult to show the residual echo powers of @xmath56 and @xmath50 as following @xmath63 and , @xmath64 respectively where @xmath65 which is the residual echo power of the second echo when the first echo is not present . from fig .
[ fig_mse2echoes ] it is clear that as the second echo moves toward the second tap , the residual echo power is reduced .
notice that @xmath66 means that only one echo exists .
hence , the assumption of one echo in the middle of neighboring taps is the worst simulation condition .
this is from the fact that echo closer to either tap is easier to estimate as shown in fig .
[ fig_mse ] ( b ) .
the argument can be easily extended to arbitrary number of echoes with arbitrary echo powers located between neighboring taps . in the previous subsection ideal phase shifter
is assumed for hilbert transform . however , due to phase imbalance in phase shifter , hilbert transform of @xmath67 will not be orthogonal to @xmath67 . to overcome the phase imbalance problem we use multiple phase shifters .
as far as those phases are not identical , by a linear combination of the multiple phase shifter outputs we can rotate and scale @xmath67 even with large phase imbalances . assuming @xmath68 taps and @xmath69 phase shifters per tap , @xmath70-th phase shifter output at @xmath16-th tap can be written as @xmath71 where @xmath72 is phase shifter gain to represent amplitude imbalance , and @xmath73 , which can be modeled as @xmath74 , is phase shift including phase imbalance @xmath75
it is clear that with @xmath76 , as far as the phase imbalance is less than @xmath77 , those three phases will not be identical . by a linear combination of phase shifter outputs
, echo can be estimated as follows @xmath78 where @xmath79 is real valued weight for @xmath16-th tap and @xmath70-th phase shifter .
the complex baseband model of echo canceller output @xmath80 can be developed from ( [ cancelloutput0 ] ) and ( [ e - t ] ) as @xmath81 where @xmath82 is complex baseband version of @xmath83 .
the real and imaginary parts of @xmath82 can be written as @xmath84 and @xmath85 respectively . since implementing wiener filter in analogue domain
is quite difficult if not impossible , it is desirable to design techniques which are implementable in analogue domain . for this end
, we apply well - known steepest - descent method to the cost function .
this leads to @xmath86 where @xmath87 is step - size and superscript @xmath88 represents complex conjugate .
notice that cross correlation @xmath89 can be served as echo channel estimation at @xmath16-th tap delay of @xmath70-th phase shifter output .
since it is quite noisy estimation due to random signal @xmath12 , noise and impairments which will be discussed in next subsection , the noisy estimation is low - pass filtered by integrator with small step - size @xmath87 as shown in ( [ lms ] ) .
together with , leads to the block diagram of the adaptive echo cancellation in analogue domain shown in fig .
[ fig_blk ] .
note that the multipliers in the outputs of phase shifters are in fact variable attenuators since the magnitudes of weights can be limited to less than 1 by design .
also notice that as the step - size @xmath87 is much less than 1 , it can be implemented by a fixed attenuator .
overall , it is quite clear that the proposed structure is implementable .
unfortunately , however , @xmath90 and @xmath91 can not be obtained due to phase noise , phase / amplitude imbalances in downconverter . with unknown phase error @xmath92 and @xmath93 in in - phase and quadrature phase
respectively , and unknown gains of @xmath94 and @xmath95 in in - phase and quadrature phase respectively , following signal has to be used @xmath96 where @xmath97 represents imaginary part and @xmath98 is phase noise .
similarly , we can define @xmath99 .
in addition , although all downconverters are driven by the same oscillator , each downconverter may have different phase rotation .
hence , effectively following will be performed @xmath100 where @xmath101 is the phase difference between downconverter for @xmath91 and downconverter for @xmath82 .
it is desirable to have @xmath102 in order to avoid divergence or slow convergence .
next , the effect of phase distortion in the variable attenuator is investigated . in wideband variable
attenuator @xcite , different attenuation causes different phase rotation @xmath103 .
thus the second term in the right - hand side of is changed to @xmath104 as in @xcite , the phase distortion can be modeled as a linear function of attenuation in log scale .
for example , assuming @xmath105 and @xmath106 phase rotations at 3 db and 38 db attenuations respectively leads to @xmath107 the additional phase rotation due to the phase distortion in the attenuator can be absorbed in @xmath101 .
the phase distortion is large when the attenuation is large .
however , when the attenuation is large , its contribution to echo estimation is small .
thus , the phase distortion at high attenuation can be ignored .
the values of above mentioned impairments are unknown . hence , we do not take any attempt to compensate those impairments in our study .
simply , the outputs of downconverters which include all those impairments are used to train the weights .
although the second terms in the right hand side of and are quite different due to impairments , as far as the signs of the second terms are same , the adaptive technique can converge as explained in appendix b. this is why the proposed method exhibits immunity to impairments unlike open loop techniques .
the fixed phase shifter , combiner , splitter and coupler are quite linear components . in @xcite ,
the input third order intercept point ( iip3 ) of the variable attenuator is 50 dbm which means with 0 dbm input power non - linear distortion power can be less than -100 dbm .
hence , 0 dbm input power at the variable attenuator is maintained .
[ cols="^,^,^,^,^,^,^ " , ] however , one difficulty in solving this optimization problem is the lack of a well - known cost function that can capture gains in cell edge and cell average in both dl and ul . in general different cost functions
can be considered depending on the application and the required gain in each parameter . here , our objective is to simply maximize the sum of all gains _ i.e. _ , dl cell edge gain + ul cell edge gain + dl cell average gain + ul cell average gain .
table v - vii show gains in average and 5% cell edge capacity of str with null forming over non - str with 8 antennas without null forming .
first , consider the large cell . as discussed earlier in this case
, ue - ue interference is not a serious problem .
it can be seen from table v that more than 100% of gain in dl cell average capacity can be achieved by str with null forming in both 8 and 12 antennas and decent gain in cell edge with 12 antennas .
the gains beyond 100% in cell average come from the reduction of dl co - channel interference due to null forming .
however , since ul is less interference limited , the gain is smaller .
the cell edge capacity is not close to twice of non - str capacity since the signal power loss is unavoidable at cell edge due to null forming .
next , let us consider the case where ue - ue interference is more dominant than dl co - channel interference _ i.e. _ small cell with small @xmath108 .
as it can be observed from table vi , without addressing ue - ue interference , cell edge users experience serious loss in dl capacity .
nevertheless , by employing the proposed resource block structure and optimizing above cost function , positive gains can be achieved with str and 12 antennas by sacrificing ul capacity as shown in table vii . although we assumed that each ue uses the whole bandwidth for time division multiplexing ( tdm ) , it is possible to allocate fraction of bandwidth to each ue for frequency division multiplexing ( fdm ) .
then , ue - ue interference will be larger unless the power spectral density is maintained the same .
in addition , due to proximate ues , in the same sector higher resolution adc will be needed . due to path loss , the saturation at lna can be avoided .
we have suggested an echo cancellation technique which can be implemented in analogue domain and demonstrated sufficient suppression of echo before lna .
the technique is robust to rf impairments exhibiting outstanding performance without requiring additional antennas .
str can be employed in csma networks .
we showed that str can reduce the hidden node problem and the suggested protocols improve throughput in both single and multiple cells .
the application of str to cellular systems creates bs - bs and ue - ue interferences .
we have provided a solution for the complete cancellation of bs - bs interference . for ue - ue interference
, we studied non - cooperative method .
first , we define normalized correlation as @xmath109\ } \nonumber\\ & = [ sinc(b(\tau-\tau_1))~sinc(b(\tau-\tau_2))]\end{aligned}\ ] ] where @xmath110 is the average power of @xmath22 .
similarly , define normalized auto - correlation as @xmath111 \nonumber\\ = & \left [ { { \begin{array}{*{2}c } 1 & sinc(b(\tau_1-\tau_2 ) ) \\
sinc(b(\tau_2-\tau_1 ) ) & 1\\ \end{array } } } \right].\end{aligned}\ ] ] then , the cross correlation vector can be written as @xmath112 where @xmath61 is a diagonal matrix defined as @xmath113.\ ] ] likewise , the auto - correlation matrix can be expressed as @xmath114 the weight vector for the wiener solution is obtained by @xmath115 after manipulations , the residual echo power defined in the denominator of is in the form of @xmath116 finally , the suppression level of echo power is calculated from @xmath117
in order to give some insight on how phase noise affects the performance , without loss of generality , the following simplified model is considered .
we consider one echo and assume ideal knowledge of the delay . with one tap echo canceller and zero valued weight initially , we have however , the echo channel estimation ( [ errsig ] ) is far from @xmath122 unless it is divided by @xmath123 .
in addition , the echo channel estimation will be corrupted by noise and various impairments .
note also that random ofdm signal @xmath22 has large peak - to - average power ratio ( papr ) .
hence , in adaptive echo canceller , the echo channel estimation is low - pass filtered by an integrator with very small step - size @xmath87 as the echo channel estimation is quite noisy .
thus , the estimation error in echo channel estimation does not directly appear in the weight .
as far as the sign of @xmath124 is equal to the sign of true echo channel @xmath125 , the weight will be gradually updated toward the true echo channel until the power of @xmath91 is small enough . due to this nature , even if the echo channel estimation is not perfect and corrupted by impairments and noise , the adaptive echo canceller can converge .
@xmath126 \cos(\delta ) \nonumber \\ & & { -}\ ! \left [ \cos(\theta_h ) \left\ { \begin{array}{l } g_{x , i}g_{z , q}(\sin(\delta_{i - q}+\varepsilon_{x - z}(t))+\sin(2\theta_x(t)-\delta_{i+q}-\varepsilon_{x+z}(t)))+ \\ g_{x ,
q}g_{z , i}(\sin(\delta_{q - i}+\varepsilon_{x - z}(t))-\sin(2\theta_x(t)-\delta_{q+i}-\varepsilon_{x+z}(t)))\end{array } \right \ } \right .
\nonumber \\ & & { + } \ : \left . \left .
\sin(\theta_h ) \left\ { \begin{array}{l } g_{x , i}g_{z , q}(\cos(\delta_{i - q}+\varepsilon_{x - z}(t))+\cos(2\theta_x(t)-\delta_{i+q}-\varepsilon_{x+z}(t)))+ \\
g_{x , q}g_{z , i}(\cos(\delta_{q - i}+\varepsilon_{x - z}(t))-\cos(2\theta_x(t)-\delta_{q+i}-\varepsilon_{x+z}(t)))\end{array } \right \ } \right ] \sin(\delta ) \right
\}. \nonumber \\\end{aligned}\ ] ] @xmath127 \cos(\delta ) \right .
\nonumber \\ & & \left .
- \left [ \cos(\theta_h ) \left\ { g_{x , i}g_{z , q}\sin(\delta_{i - q})+ g_{x , q}g_{z , i}\sin(\delta_{q - i } ) \right \ } + \sin(\theta_h ) \left\ { g_{x , i}g_{z , q}\cos(\delta_{i - q})+ g_{x , q}g_{z , i}\cos(\delta_{q - i } ) \right \ } \right ] \sin(\delta ) \right \}.\end{aligned}\ ] ] then , the echo channel estimation can be written as shown in ( [ errwpn ] ) where @xmath130 is the angle of @xmath22 assuming uniformly distributed over @xmath131 to @xmath132 and @xmath133 is the phase difference between downconverters for @xmath91 and @xmath22 .
when in majority of samples the sign of @xmath134 is equal to the sign of @xmath135 , the weight will be updated to the right direction . due to random time varying phase from transmitted signal @xmath130 , phase noise @xmath136 and @xmath98 , the echo channel estimation will fluctuate .
however , by the low pass filtering , those time varying noises will be filtered . since we can assume the phase noise and the phase of transmitted signal are stationary
, the integration over time can effectively perform ensemble average .
assuming gaussian phase noise , the ensemble average over @xmath22 , @xmath136 and @xmath98 leads to ( [ eerrwpn ] ) where @xmath137 is the variance of phase noise and @xmath110 is the average power of @xmath22 .
as can be seen , the phase noise reduces the power of echo channel estimation .
hence , the performance will be more sensitive to the noise .
when all downconverters are driven by the same oscillator , the phase noises can be identical . in this scenario ,
the phase noise difference @xmath138 will be zero .
then , the scale factor @xmath139 disappears as if no phase noise is present .
yang - seok choi ( s91,m01 ) received the b.s .
degree from korea university , seoul , korea , in 1990 , the m.s.e.e .
degree from korea advanced institute of science and technology ( kaist ) , taejon , korea , in 1992 , and the ph.d .
degree from polytechnic university , brooklyn , ny , usa in 2000 , all in electrical engineering . from 1992 to 1996 , he was with samsung electronics , co. , ltd . , suwon , korea , where he developed various modems for hdtv and dbs . during 2000 summer he held a summer intern position at at&t labs - research shannon lab , florham park , nj , usa . in 2000 , he joined national semiconductor , east brunswick , nj , usa where he was involved in the development of w - cdma . during 2001 - 2002 , he was a senior technical staff member at at&t labs - research , middletown , nj , usa where he researched on mimo systems , ofdm systems and information theory . from 2002 to 2004 he had been with vivato , inc .
, spokane , wa , usa working on mimo ofdm systems , smart antenna systems , and antenna / beam selection techniques . in 2004
, he joined intel corporation , hillsboro , or , usa where he studied on broadband wireless communications systems and was a director of radio systems engineering leading standards and technology development . in 2013 , he joined intel labs where he researches future generation wireless systems .
his research interests include various aspects of wireless communications systems .
hooman shirani - mehr ( s03- m10 ) received the b.s .
degree from sharif university of technology , tehran , iran , in 2001 and m.s . and ph.d .
degrees from the university of southern california , los angeles , ca in 2006 and 2010 , respectively , all in electrical engineering .
since 2010 , he has been with intel corporation where he is currently working on 3gpp lte and lte - advanced wireless systems .
his research interests include communication theory , information theory and signal processing with applications in wireless communications . | full duplex or simultaneous transmission and reception ( str ) in the same frequency at the same time can potentially double the physical layer capacity .
however , high power transmit signal will appear at receive chain as echoes with powers much higher than the desired received signal .
therefore , in order to achieve the potential gain , it is imperative to cancel these echoes .
as these high power echoes can saturate low noise amplifier ( lna ) and also digital domain echo cancellation requires unrealistically high resolution analog - to - digital converter ( adc ) , the echoes should be cancelled or suppressed sufficiently before lna . in this paper
we present a closed - loop echo cancellation technique which can be implemented purely in analogue domain .
the advantages of our method are multiple - fold : it is robust to phase noise , does not require additional set of antennas , can be applied to wideband signals and the performance is irrelevant to radio frequency ( rf ) impairments in transmit chain .
next , we study a few protocols for str systems in carrier sense multiple access ( csma ) network and investigate mac level throughput with realistic assumptions in both single cell and multiple cells .
we show that str can reduce hidden node problem in csma network and produce gains of up to 279% in maximum throughput in such networks .
moreover , at high traffic load , the gain of str system can be tremendously large since the throughput of non - str system is close to zero at heavy traffic due to severe collisions .
finally , we investigate the application of str in cellular systems and study two new unique interferences introduced to the system due to str , namely _ bs - bs interference _ and _ ue - ue interference_. we show that these two new interferences will hugely degrade system performance if not treated appropriately .
we propose novel methods to reduce both interferences and investigate the performances in system level .
we show that bs - bs interference can be suppressed sufficiently enough to be less than thermal noise power , and with favorable ue - ue channel model , capacities close to double are observed both in downlink ( dl ) and uplink ( ul ) .
when ue - ue interference is larger than dl co - channel interferences , we propose a simple and non - cooperative " technique in order to reduce ue - ue interference .
full duplex , simultaneous transmission and reception ( str ) , echo cancellation , csma , null forming , hidden node . |
muon inelastic scattering off nuclei plays an important role in muon propagation through matter . in this process muon
may loss a significant part of its energy and can be scattered at large angle .
therefore , muon - nucleus inelastic scattering is of interest for numerous applications related to muon transport in matter , in particular , for calculations of muon intensity at large depth of matter , muon - induced hadron flux underground , background produced by atmospheric muons in underground neutrino experiments , etc .
several models [ 1 - 4 ] have been developed to describe the muon - nuclear inelastic interactions , nonetheless uncertainties of this process are much larger than for purely electromagnetic interactions .
the reason is that the bulk of this process is characterized by low squared four - momentum transfer @xmath1 .
the smallness of the @xmath1 does not allow to use of the perturbative qcd ( pqcd ) for calculation of nuclear structure function ( sf ) and phenomenological models such as the regge or general vector dominance model ( gvdm ) have to be used .
the parameterization of the nucleon sf obtained in these models , depends on free parameters which can be determined from a fit of experimental data and can be applied in the range of @xmath4 gev@xmath5 .
this range is often referred to as photo - production .
however , they fail to describe deep inelastic scattering ( dis ) data at high @xmath1 .
the pqcd ( nlo qcd ) gives a good description of the structure functions at @xmath6 gev@xmath5 .
so , a model to combine various aspects of these approaches is needed to describe the @xmath1 behavior of nucleon sfs over complete range from photo - production to dis .
the widely used approximation [ 4 ] of the muon photo - nuclear cross - section was obtained twenty years ago in the framework of gvdm .
experimental data in the region of @xmath7 gev@xmath5 and @xmath8 were used for determing the parameters . recently , precise data [ 5 ] on sf in wide ranges of @xmath1 ( @xmath9 gev@xmath5 ) and @xmath2 ( @xmath10 ) have been obtained and new nuclear effects ( anti - shadowing and emc - effect ) were observed .
the main goal of this work is the calculation of the muon - nucleus inelastic cross - section , based on the modern nucleon sf and on the present knowledge of nuclear effects .
the paper is organized as follows . in sec.2
we give the general relations and definitions used for description of neutral current charged lepton - nucleon scattering .
the procedure of the calculation of nucleon sf , using ckmt regge model ckmt [ 6 ] and mrs99 parton distribution function ( pdf ) [ 7 ] is described in sec.3 . in sec.4 , nuclear effects and their parameterization are described .
the total cross - section , the muon energy losses , and the angular distributions of scattered muons are given in sec.5 . in the conclusion
we summarize the main results of the paper .
the cross - section of neutral current charged lepton scattering off nucleon @xmath11 is given by the sum of contributions from the processes as shown in fig.1 here @xmath12 and @xmath13 are initial and final lepton four - momenta , @xmath14 is the virtual photon and z - boson momentum , @xmath15 and @xmath16 are the initial nucleon momentum and the total momentum of the final hadrons @xmath17 , respectively .
this process can be described by the four - momentum transfered @xmath1 , bjorken @xmath2 and lepton s energy loss @xmath18 ( or inelasticity @xmath19 ) defined as @xmath20 in laboratory system @xmath21 where @xmath22 and @xmath23 are nucleon and lepton masses , respectively .
the general form of the differential cross - section for scattering of charge non - polarized lepton on non - polarized nucleon , summed over the final lepton polarizations can be expressed as : @xmath24 , \ ] ] where @xmath25 @xmath26 @xmath27 and @xmath28\frac{1}{\nu } \nonumber\\ & & y_3=\big(e-\nu\big ) . \ ] ] here the term @xmath29 accounts for the z@xmath30 propagator is @xmath31 where @xmath32 is the fermi constant , @xmath33 is the fine structure constant , and @xmath34 is z - boson mass .
the constants of lepton weak coupling @xmath35 and @xmath36 are @xmath37 where @xmath38 is weinberg angle .
the functions @xmath39 are the electromagnetic ( @xmath40-exchange ) and neutral current ( @xmath41-exchange ) structure functions , respectively .
the functions @xmath42 correspond to decomposition over invariant functions of the tensor ( @xmath40-@xmath41 interference ) @xmath43 the upper sign in eq.(6 ) and eq.(7 ) corresponds to lepton scattering ( @xmath44 ) , and the lower sign is for antilepton(@xmath45 ) scattering .
the term proportional to the function @xmath46 , is due to interference between the electromagnetic scattering amplitude and the axial - vector current weak interaction amplitude .
the amplitudes have opposite @xmath47 parities , so the corresponding terms have opposite sings for lepton and antilepton scattering . at low @xmath1
the @xmath48 interference term is much smaller then the @xmath40-exchange one , but it increases linearly with @xmath1 ( eq.(9 ) ) and becomes comparable with the @xmath40-exchange term at @xmath4910@xmath50 gev@xmath5 . in terms of parton distributions in lo - approximation , the sfs can be written as @xmath51 @xmath52 @xmath53 @xmath54 @xmath55 and @xmath56 @xmath57 @xmath58 here @xmath59 and @xmath60 are parton distribution functions in the proton , @xmath61 , @xmath62 and @xmath63 are the charge , vector and axial - vector week couplings of quarks . for up - quarks ( u , c ,
t ) they are @xmath64 and for down - quarks ( d , s , b ) @xmath65 it is seen from eqs.(4 ) and ( 9 ) that the main contribution to the total cross - section is due to photo - production ( low @xmath1 process ) , however at fixed outgoing muon energy the large scattering angle corresponds to high @xmath1 @xmath66 so , for calculation of muon scattering at large angles it is necessary to know the behavior of the nucleon sfs in the wide range of @xmath67 gev@xmath5 .
at high-@xmath1 the qcd predictions for the nucleon sfs are obtained by solving the dglap evolution at nlo approximation in @xmath68 or dis schemes .
these equations yield the parton distribution functions at all values of @xmath1 provided that the pdf are given as a function of @xmath2 at some input scale @xmath69=1.2 - 5 gev@xmath5 .
the latest global fits performed by several group ( mrs99 [ 7 ] , grv98 [ 8 ] , and cteq5 [ 9 ] ) give a good description of the experimental data . at @xmath1 below @xmath70
perturbative qcd fails to describe data and phenomenological non - perturbative ( gvdm or regge ) models are required .
a considerable number of non - perturbative models have been developed [ 10]-[12 ] recently .
these models predict a correct limit of @xmath0 at @xmath1=0 and give a good description of the sfs at low and medium @xmath1 .
thus , neither the non - perturbative approaches nor pqcd can be expected to describe the @xmath1 behavior of the sfs over the complete the range from photo - production to dis . a number of models combining qcd and phenomenological approaches have been developed to describe data in the transition region of @xmath1 ( see review [ 13 ] ) . in this paper
we use the ckmt model [ 6 ] at low and moderate @xmath1 and the mrs99 fit of pdf [ 7 ] at high-@xmath1 .
the ckmt model proposes the following parameterization of the proton structure function @xmath71 @xmath72 the singlet term @xmath73 corresponds to the pomeron contribution which determines the small-@xmath2 behavior of sea quarks and gluons .
the dependence of effective intercept of the pomeron , @xmath74 on @xmath1 is parametrized as @xmath75 the @xmath76 behavior of @xmath77 is determined by the function @xmath78 the parameterization for the non - singlet term which corresponds to the secondary ( @xmath79,@xmath80 ) reggion ( valent quark ) contribution , is @xmath81 where behavior at @xmath82 is determined by the secondary reggion intercept @xmath83
. the valence quark distribution can be separated into the contributions of the u and d valence quarks by replacing @xmath84 where @xmath85 @xmath86 normalization conditions for valence quarks in proton @xmath87dx=2e^2_u \nonumber\\ & & \int\limits_0 ^ 1\frac1x\left[xd_v(x , q^2)\right]dx = e^2_d \ ] ] fix the values of parameters @xmath88 and @xmath89 at @xmath90 .
the limit of @xmath91 corresponds to interaction of real photons .
the total cross - section for real photons can be written as @xmath92 _ { q^2=0}.\ ] ]
one can see from eqs.(23),(27 ) , and ( 28 ) that @xmath93 at @xmath94 and fixed @xmath18 .
thus , the parameterization @xmath95 \ ] ] takes place in the ckmt model . in this way we find parameterizations of both @xmath96 and @xmath97 cross - section with 7 free parameters : @xmath98 , and @xmath99 . to determine the parameters we have made a joint fit of the @xmath100 data and nmc , e665 , slac , zeus , and h1 data on the proton sf @xmath0 in the region @xmath101 gev@xmath5 and @xmath102 [ 5 ] . as initial condition for the values of different parameters
, we used those obtained in the previous fit in ref.[6 ] . a global fit results in the following values of parameters ( all dimensional parameters are in gev@xmath5 ) : @xmath103=0.2513 , @xmath104=0.6186 , @xmath105=3.0292 , @xmath106=1.4817 , @xmath107=0.0988 , @xmath83=0.4056 , and @xmath99=0.12 .
the values of the parameters @xmath108 and @xmath109 , were determined from normalization conditions for valence quarks ( at @xmath69=2 gev@xmath5 ) .
the quality of the description of all experimental data is very good and @xmath110=754.8/600 , where only the statistical errors have been used .
recently , a modified version of the ckmt model used new data on @xmath96 at low @xmath1 has been published [ 14 ] .
the values of the main parameters are in a good agreement with those , obtained in the present work .
so , for calculation of @xmath0 in the entire region of @xmath1 we use the ckmt model at @xmath111 gev@xmath5 , the mrs99 pdf at @xmath112 gev@xmath5 and a linear fit between @xmath96(ckmt ) and @xmath96(mrs99 ) in the region @xmath113 gev@xmath5 .
the result of the fit of @xmath96 and @xmath114 is shown in figures 2 ( @xmath96 vs. @xmath2 for different value of @xmath1 ) and 3 ( @xmath96 vs. @xmath1 for different value of @xmath2 ) along with the experimental data [ 5 ] . the cross - section @xmath115 as a function of @xmath116 is shown in fig.4 ( the data from refs.[15],[16 ] ) .
a good description of experimental data is obtained at all @xmath2 and @xmath1 values .
it should be noted that : \a ) the recent zeus bpt97 data [ 17 ] were not included in our fit , but are in agreement with the ckmt model prediction at @xmath117 gev@xmath5 , \b ) the rise of @xmath96 at low @xmath2 and low @xmath1 is well described by the ckmt model with the slope @xmath107=0.0988 while the experimental value is [email protected] [ 17 ] , \c ) the @xmath114 values found by the zeus collaboration are the result of a phenomenologically motivated extrapolation . in figures 2 - 4
the sf @xmath96 and @xmath119 that have been obtained by bezrukov and bugaev [ 4 ] and used for calculation of the muon photo - nuclear cross - section are shown also . at @xmath120
the sfs rise slower than the present data indicate .
an the other hand in the region @xmath121 and @xmath122 gev@xmath5 the sfs are overestimated .
the ckmt parameterization gives separate contributions of valence quarks , sea quarks , and gluons .
we used this peculiarity for parameterization of the neutron sf @xmath123 , that can be extracted from the deuteron @xmath124 and the proton @xmath96 data , using the relation @xmath125 , \ ] ] and the gottfried sum rule @xmath126 in the case of su(2 ) symmetric sea , pdf of @xmath127 equals to=@xmath128 and therefore @xmath129=1/3 .
however , the nmc collaboration [ 5 ] gives @[email protected] at @xmath130 gev@xmath5 , i.e. significantly below 1/3 and shows that @xmath131 and @xmath132 at @xmath133 .
taking into account these results the singlet term of the @xmath123 has to be modified . because of the isotopical invariance of strong interaction the non - singlet term @xmath134 is @xmath135 where @xmath136 and @xmath137 are given by eq.(29 ) and eq.(30 ) .
the singlet term @xmath138 has an additional free parameter @xmath139 .
the value of this parameter obtained from fit of the @xmath124 data [ 5 ] in the region @xmath140 gev@xmath5 ( all other parameters were fixed by the fit of the @xmath96 and @xmath114 ) is @xmath141 .
the quality of the description of data is good with value of @xmath142 for the sf @xmath124 data and @xmath143 for the @xmath144 data , where only statistical errors have been used . for calculation of the sf @xmath123 in the entire region of @xmath1 we used the approximation of eq.(37 ) and eq.(38 ) at @xmath145 gev@xmath5 , the mrs99 pdf at @xmath112 gev@xmath5 and linearly fit between @xmath123(ckmt ) and @xmath123(mrs ) in the transition region @xmath146 gev@xmath5 .
the sf @xmath124 ( fig.5 ) and @xmath144 ( fig.6 ) as function of @xmath1 for different values of @xmath2 are shown with experimental data .
fig.7 shows the @xmath147 vs. @xmath2 at @xmath130 gev@xmath5 .
the calculations are in agreement with the nmc data [ 18 ] . for calculation of the cross - section of lepton - nucleon scattering
it is necessary to know the behavior of sf @xmath148 in a wide range of @xmath1 and @xmath2 also .
this sf can be expressed using the longitudinal sf @xmath149 @xmath150 then @xmath151 where @xmath152 perturbative qcd describes reasonably well the available data on the ratio @xmath153 at large values of @xmath1 and very little is known about possible extrapolations towards the region of low @xmath1 . in the limit of @xmath154
the sf @xmath149 has to vanish as @xmath155 ( for fixed @xmath18 ) and therefore @xmath156 . at @xmath157 and @xmath158
gev@xmath5 the experimental results are poor .
data show a small value of @xmath159 at moderate values of @xmath2 and a possible increase of @xmath159 as @xmath2 decreases .
the data come from experiments carried out on different targets and the differences @xmath160 are consistent with zero and do not exhibit any significant dependence on @xmath2 [ 19 ] .
data can be fitted by a parameterization @xmath159(slac98 ) [ 20 ] .
however , this fit should not be used at @xmath161 gev@xmath5 . in this region
we used the gvdm asymptotic of @xmath159 at @xmath154 given by [ 21 ] @xmath162 where @xmath163 gev is the @xmath164 meson mass .
thus , at @xmath165 gev@xmath5 the function @xmath153 is calculated as follows : @xmath166 in the region @xmath167 a linear fit between @xmath159(mrs99 , @xmath168 ) and @xmath159(slac98 , @xmath169 ) is used . at @xmath170 @xmath171 where the function @xmath172 is determined by normalization condition at @xmath70 @xmath173 and
the function @xmath174 is calculated using eq.(40 ) .
fig.8 shows the experimental values of @xmath159 as a function of @xmath1 in four ranges of @xmath2 along with the result of parameterization eq.(41)-eq.(42 ) . in the region of low @xmath170 it decreases with @xmath1 at all values @xmath2 , but
the dependence on @xmath2 is not strong ( fig.9 ) .
however , extrapolation of @xmath159 outside the kinematical range of data : namely at @xmath94 and @xmath82 , based on the presently available data , is a rather delicate problem . in figs.10 and 11
we show the results of calculations of the differential cross - section of neutral current ( nc ) @xmath175 scattering @xmath176 and @xmath177 at high @xmath1 .
the cross - section @xmath176 decreases by six orders of magnitude between @xmath178 gev@xmath5 and 4000 gev@xmath5 .
this decrease is due to the photon propagator .
the cross - section @xmath177 is shown for different @xmath1 regions . for @xmath179
gev@xmath5 the bulk of the cross - section is concentrated at small values of @xmath19 . for @xmath180 gev@xmath5
the differential cross - section is approximately constant with @xmath19 .
the predictions using the mrs99 pdf give a good description of measured cross - sections .
nc scattering at high @xmath1 is sensitive to the contribution due to the @xmath181 .
according to eqs.(4)-(9 ) the @xmath181 contribution reduces approximately by 25%(12% ) for @xmath182 ( @xmath183 ) cross - section at @xmath180 gev@xmath5 .
the sfs measured for different nuclei @xmath184 are found to differ from the sf measured on deuteron [ 24,25 ] .
the modifications are usually observed as a deviation from unity of the ratio @xmath185 , where @xmath186 and @xmath124 are the sfs per nucleon measured in a nucleus and in deuteron , respectively . neglecting nuclear effects in the deuteron
, @xmath124 can approximately stand for a isospin averaged nucleon sf , @xmath187 .
different nuclear effects are observed in different region of @xmath2 .
\(i ) shadowing at @xmath1880.1 .
the ratio @xmath189 is smaller than unity .
the experimental data cover the region @xmath19010@xmath191 and @xmath189 decreases with decreasing @xmath2 .
shadowing increases with nuclear mass @xmath184 and weakly depends on @xmath1 .
\(ii ) anti - shadowing at @xmath192 .
the nmc data have established a small ( a few percent ) but statistically significant excess over unity . within the accuracy of the data
no significant @xmath1-dependence of this effect has been found .
( iii ) emc effect at @xmath193 .
the measured ratio @xmath189 decreases as @xmath2 rises and has a minimum at @xmath194 .
the magnitude of this depletion grows slowly with nuclear mass number .
the data imply that a strong @xmath1-dependence of the @xmath189 is excluded in this region also .
\(iv ) fermi motion . at @xmath195
the ratio @xmath189 rises above unity but experimental information is rather scarce .
investigations of differences between the longitudinal - to - transverse cross - section ratio @xmath196 for different nuclei showed that @xmath197 is compatible with zero .
this implies that nuclear effects influence both sfs @xmath198 and @xmath0 in a similar way . at the moment
, there is no unique theoretical description of these effects ; it is believed that different mechanisms are responsible for them in different kinematical regions .
for example , the emc effect indicates that the averaged momentum carried by valence quarks in nuclei is reduced , relative to free nucleon .
it has been shown in [ 26 ] that the pattern of the function @xmath199 has a universal shape in the range @xmath200 and in the range of nuclei mass @xmath2014 .
namely , the ratio @xmath202 could be well approximated with phenomenological functions in different regions of @xmath2 . at @xmath203 @xmath204 where the @xmath184 dependence of @xmath205 can be approximated as @xmath206 for @xmath2074 .
the number of nucleons @xmath208 at the nucleare surface is given by the woods - saxon potential @xmath209 with values of parameters @xmath210=0.17 @xmath211 , @xmath103=0.54 fm and @xmath212 the function @xmath213 is @xmath214 where @xmath215 , @xmath216 , @xmath2170.278 , @xmath218=0.5 and @xmath219 ( @xmath220 is pion mass ) . at [email protected]
the function is given by @xmath222 with @xmath223 where @xmath224=0.129 , @xmath225=0.456 and @xmath226=0.553 .
we used eq.(49 ) for calculation of @xmath189 up to @xmath227 .
the value of @xmath228 as a function of a was obtained in the following way .
the experimental data [ 26 ] , show that in the region @xmath229 the ratio @xmath189 decreases with @xmath2 .
generally , small @xmath2 corresponds to small @xmath1 , therefore the approach of real photon interaction can be used .
so , for @xmath133 , @xmath230=@xmath231 , where @xmath232 is photon - nuclear cross - section and @xmath233 is the photon - nucleon cross section averaged over proton and neutron .
the expression for the function @xmath234 has been obtained in ref.[4 ] , using the optical nuclear model @xmath235 where the function @xmath236 is @xmath237 , \ ] ] and @xmath238 . using eq.(35 ) with the values of parameters obtained in this work , we can write the averaged photon - nucleon cross - section as follows @xmath239 in the range @xmath240 , eq.(49 ) reduces to @xmath241 then , from the asymptotic condition @xmath242 we obtain the expression for @xmath228 @xmath243^{1/m_1}. \ ] ] at @xmath244 we assumed that the function @xmath189 is constant and @xmath245 the results of approximation of the ratio @xmath189 are presented in fig.12 as a function of @xmath2 for different nuclear targets and are in a good agreement with data . taking into account the nuclear effects ,
the nuclear sfs @xmath246 and total photon - nuclear cross section can be written as : @xmath247 and @xmath248 . \ ] ] the calculated cross sections @xmath249 are shown in fig.13 as functions of the real photon energy for nuclei c , cu and pb .
we have calculated the main characteristics of inelastic muon scattering in standard rock ( @xmath250 , @xmath251 and @xmath252 g / cm@xmath50 ) .
the spectra of muon energy loss ( @xmath254 is avogadro s number ) in single interaction , @xmath255 are shown in fig.14 as functions of inelasticity @xmath19 for different muon energies .
the energy dependence of the total cross section , @xmath256 and muon energy loss , @xmath257 are shown in fig.15 .
the allowed kinematical region for the variables @xmath18 and @xmath1 is determined by the following equations : @xmath258 at @xmath259 , and @xmath260 the results from ref.[4 ] are given in these figures for comparison also .
it is necessary to note , that the cross - section and muon energy loss [ 4 ] have been calculated taking into account the shadowing effect only .
the cross - section of inelastic muon scattering obtained in the present work is larger by factor 1.2 and the muon energy losses @xmath261 is larger too by @xmath2628% at @xmath26310@xmath50 gev and by @xmath26230% at @xmath26310@xmath264 gev . as a result , the total energy losses ( the sum of bremsstrahlung , pair - production and inelastic muon scattering ) increases by @xmath2621% at @xmath26310@xmath50 gev and by @xmath2624% at @xmath26310@xmath264 gev .
these differences are mainly due to contributions of small @xmath2 and small @xmath1 where the modern sfs are large than sfs used by bezrukov and bugaev .
the probabilities @xmath265 of muon scattering in single interaction at the angles larger than @xmath266 with outgoing muon energy @xmath267 are shown in fig.16 as a function of @xmath266 for different values of @xmath268 and primary muon energies .
the results are given for @xmath269 and @xmath270 scattering .
the main peculiarities of the inelastic muon scattering are : ( i ) at fixed values of @xmath266 and @xmath271 the probability decreases very quickly with the initial muon energy @xmath272 .
for example , for @xmath273 and @xmath271=10 gev , @xmath274=6.3@xmath27510@xmath191 at @xmath272=10@xmath5gev and @xmath2763.8@xmath27510@xmath277 at @xmath272=10@xmath278gev .
\(ii ) at fixed values of @xmath266 and @xmath272 the probability increases with decreasing outgoing muon energy @xmath271 . for @xmath2792@xmath280 and @xmath272=10@xmath50gev , @xmath274=7.8@xmath27510@xmath281 at @xmath28210@xmath5gev and @xmath2763.8@xmath27510@xmath277 at @xmath28210gev .
\(iii ) at fixed values of @xmath272 and @xmath271 the mean values of @xmath2 and @xmath1 ( @xmath283 and @xmath284 ) increase with @xmath266 . for muon energies
@xmath272=10@xmath5 gev and @xmath2850.1@xmath272 the values of @xmath283 and @xmath284 increase from @xmath283=0.12 ( anti - shadowing region ) and @xmath284=0.75 gev@xmath5 at @xmath266=0.25@xmath286 up to @xmath283=0.25 ( emc - region ) and @xmath284=34 gev@xmath5 at @xmath266=6@xmath286 . for energies @xmath272=10@xmath50 gev and @xmath2850.1@xmath272 : @xmath283=0.09 ( shadowing region ) , @xmath287=28 gev@xmath5 at @xmath266=0.25@xmath286 and @xmath283=0.46 ( anti - shadowing region ) and @xmath284=925 gev@xmath5 at @xmath266=6@xmath286 .
thus , the probability of scattering at large angles is suppressed by the emc - effect .
in this work we studied inelastic muon scattering off nuclei . \1 . it is shown that the combination of the sf @xmath0 , predicted by the ckmt model at low and moderate values of @xmath1 , and the mrs99 pdf at high @xmath1 gives a good description of the data over the complete measured region from photo - production to dis .
in particular , the ckmt model well describes the rise of the @xmath96 at low @xmath2 and @xmath1 with the slope @xmath288 .
furthermore , in the framework of this model the expression for the neutron sf @xmath123 can be obtained .
the result is in a good agreement with the @xmath289 and @xmath290 data .
the mrs99 pdf well describe the differential cross - sections @xmath291 and @xmath292 , calculated taking into account not only electromagnetic current contribution but also contributions of neutral current and @xmath48 interference .
the @xmath48 interference contribution is clearly seen at high @xmath293 gev@xmath5 .
the nuclear effects modify the nucleon sfs in the entire measured region of @xmath2 and @xmath1 .
the modification depends very slowly on @xmath1 and increases with @xmath184 .
the obtained sfs have been used for calculations of the muon - nucleus scattering cross - section , muon energy losses and muon angular distributions in inelastic interaction taking into account nuclear effects and contributions of neutral current and @xmath48 interference . as a result
, the total cross - section and energy losses increase with muon energy faster then predicted in ref.[4 ] .
the scattering of high energy muons ( @xmath294 ) gev at large angles is suppressed by emc - effect and @xmath48 interference .
we are grateful to a.b .
kaidalov and f.v .
tkachov for a helpful discussions .
this work was supported by russian foundation for basic research grant number 99 - 02 - 18374 .
999 k. kobayakawa , nuovo cimento * b47 * , 156 ( 1967 ) p. ditsas , b.j .
read , and g. shaw , nucl.phys .
* b99 * , 85 ( 1975 ) p. ditsas and g. shaw , nucl.phys . *
b113 * , 246 ( 1976 ) l.b .
bezrukov and e.v .
bugaev , sov.j.nucl.phys . * 33 * , 635 ( 1981 ) http://www-spires.dur.ac.uk/hepdata/structure.html a. capella et al .
, phys.lett . *
b337 * , 358 ( 1994 ) a.d .
martin et al .
, eur.phys.j . *
c14 * , 133 ( 2000 ) m. gluck , e. reya , and a. vogt , eur.phys.j . * c5 * , 461 ( 1998 ) h.l .
lai et al . , eur.phys.j . *
c12 * , 375 ( 2000 ) d. schildknecht , acta phys . polon . *
b28 * , 2453 ( 1997 ) h. abramowicz and a. levy , hep - ph/9712415 k. abel , f. barreiro , and f.j .
yndurain , hep - ph/9610380 + k. abel and f.j .
yndurain , nucl.phys . *
b495 * , 221 ( 1997 ) a.m. cooper - sarkar , r. devenish , and a. de roeck , int j. mod . phys . *
a13 * , 3385 ( 1998 ) c. merino , a.b .
kaidalov , and d. pertermann , eur .
j. * c20 * , 301 ( 2001 ) d.o .
caldwell et al . , phys.rev .
* 40 * , 1222 ( 1978 ) m. derrick et al .
* c63 * , 391 ( 1994 ) + s. aid et al . , z.phys .
* c69 * , 27 ( 1995 ) + j. breitweg et al . , eur.phys.j .
* c7 * , 609 ( 1999 ) j. breitweg et al . , phys.lett .
* b487 * , 1 - 2 , 53 - 73 ( 2000 ) m. arneodo et al . , phys.rev .
* d50 * , r1 ( 1994 ) m. arneodo et al .
, nucl.phys . *
b487 * , 3 ( 1997 ) + l.h .
tao et al . , z.phys .
* c70 * , 83 ( 1996 ) k. abe et al.,phys.lett . *
b452 * , 194 ( 1999 ) a.d .
martin , m.g .
ryskin , and a.m. stasto , eur.phys.j . *
c7 * , 643 ( 1999 ) j. breitweg et al . , eur.phys.j .
* c11 * , 427 ( 1999 ) c. adloff et al .
, eur.phys.j . *
c13 * , 609 ( 2000 ) c. adloff et al .
, eur.phys.j . *
c19 * , 269 ( 2001 ) m. arneodo , phys.rep .
* 240 * , 301 ( 1994 ) g. piller and w. weise , phys.rep . * 330 * , 1 ( 2000 ) g.i .
smirnov , phys.lett .
* b364 * , 87 ( 1995 ) + g.i .
smirnov , eur.phys.j .
* c10 * , 239 ( 1999 ) d.o .
caldwell et al . , phys.rev . *
d7 * ( 1973 ) 1362 + d.o .
caldwell et al . , phys.rev.lett .
* 42 * ( 1979 ) 553 fig .
2a : the proton structure function @xmath96 at low @xmath1 as a function of @xmath2 .
solid lines are our calculations using the ckmt model and dashed lines calculations of bezrukov and bugaev [ 4 ] .
points are results of experiments zeus and e665 [ 5,17 ] . fig . 3 : the proton structure function @xmath96 as a function of @xmath1 at fixed values of @xmath2 .
the solid curves have been obtained by using the ckmt+mrs99 model and dashed curves are result from ref.[4 ] .
data set is due to zeus , h1 , e665 , nmc , and slac experiments [ 5 ] . for clarity
an amount @xmath295=13.6 - 0.6@xmath296 is added to @xmath96 where @xmath296=1(16 ) for the lowest ( highest ) @xmath2 value .
fig . 4 : total cross - section @xmath115 as a function of @xmath297 .
solid line is our calculations using combined ( ckmt+mrs99 ) model and dashed one is from work [ 4 ] .
experimental data are from [ 15 ] at low energies and from zeus and h1 [ 16 ] at higher energies . fig . 5 : the deutron structure function @xmath124 as a function of @xmath1 at fixed values of @xmath2 .
data set is due to bcdms , e665 , nmc , and slac experiments [ 5 ] .
for clarity an amount @xmath295=13.6 - 0.6@xmath296 is added to @xmath124 where @xmath296=1(10 ) for the lowest ( highest ) @xmath2 value .
fig . 6 : the ratio of structure functions @xmath144 as a function of @xmath1 at fixed values of @xmath2 , as compared to nmc data [ 5 ] .
for clarity an amount @xmath298 is added to @xmath144 where @xmath296=1(11 ) for the lowest ( highest ) @xmath2 value .
10 : the differential cross - section @xmath176 of neutral current proton scattering off electron ( positron ) as a function of @xmath1 .
points are experimental results of zeus [ 22 ] and h1 [ 23 ] experiments . fig . 13 : total cross - section @xmath249 for c , cu , and pb as a function of real photon energy
. results of our calculations ( solid curves ) and calculations by bezrukov and bugaeev [ 4 ] ( dashed curves ) as compered to experimental data [ 27 ] . fig . 14 : the spectra of muon energy loss due to muon inelastic scattering in standard rock as a function of @xmath19 for fixed muon energies , as compered to calculations by bezrukov and bugaev [ 4 ] .
( @xmath19 is the fraction of @xmath300 lost by muon in single interaction . )
fig . 16 : the probabilities @xmath265 of muon scattering in single interaction at angle larger than @xmath266 with energy of outgoing muon @xmath267 as a function of @xmath266 for fixed values of @xmath268 and incoming muon energies .
solid ( dashed ) lines are for positiv ( negativ ) muon charge . | it is shown that the combination of the structure functions @xmath0 predicted by the ckmt model at low and moderate values of @xmath1 and mrs99 parton distribution functions at high @xmath1 gives a good description of the data over complete measured region of @xmath2 and @xmath1 . using these structure functions
the main characteristics of the muon - nucleus inelastic scattering are calculated .
nuclear effects and contributions of the neutral current and @xmath3 interference are taken into account .
= -1 . cm * the cross - section of muon - nuclear * + + a.v.butkevich and s.p.mikheyev + + _ 60th october anniversary prospect , 7a , moscow 117312 , russia _ + |
the inf - sup constant of the divergence @xmath0 is defined for a domain @xmath1 as @xmath2 here @xmath3 denotes the space of square integrable functions with mean value zero , with norm @xmath4 and scalar product @xmath5 , and @xmath6 is the usual sobolev space , closure of the space of smooth functions of compact support in @xmath7 with respect to the @xmath8 seminorm , which for vector - valued functions is defined by @xmath9 the inf - sup condition @xmath10 is also called lbb condition @xcite or ladyzhenskaya - babuka - brezzi condition ( a term coined ca .
1980 by t. j. oden , on suggestion by j. l. lions @xcite ) , and @xmath0 is therefore also known as lbb constant @xcite .
it plays an important role in the discussion of the stability of solutions and numerical approximations of the equations of hydrodynamics .
the exact value of the constant is known only for a small class of domains , first for balls and ellipsoids ( derived from the cosserat spectrum @xcite ) , then in three dimensions for spherical shells @xcite , and finally in two dimensions for annular domains @xcite and some domains defined by simple conformal images of a disk @xcite , @xcite .
its precise value remains , however , unknown for such simple domains as a square .
finding estimates , from below and from above , for @xmath0 has therefore been an important subject for many years .
the question can be rephrased in terms of the cosserat eigenvalue problem , see relation below , and it is this problem that we will study in the present paper .
in particular , any number known to belong to the nontrivial part of the spectrum of the cosserat operator will imply an upper bound for the lbb constant .
other inequalities and eigenvalue problems are known to be related to the cosserat eigenvalue problem and the lbb condition , namely the korn and friedrichs inequalities . in two dimensions in particular ,
such relations have been investigated since the paper @xcite by friedrichs , and the equivalence between these inequalities and equations between the corresponding constants have been studied in the classical paper @xcite by horgan and payne , see @xcite for improved versions of some of their results . for the cosserat eigenvalue problem ,
the values @xmath11 and @xmath12 are eigenvalues of infinite multiplicity , and @xmath13 is an accumulation point of eigenvalues . on smooth domains ,
these are the only points in the essential spectrum , as shown by mikhlin @xcite , see @xcite for a complete proof for @xmath14 domains . on non - smooth domains ,
a non - trivial essential spectrum is present , as was already pointed out in @xcite for the equivalent friedrichs problem .
we show in theorem [ t : esspolyg ] for two - dimensional domains that a polygonal corner of opening @xmath15 contributes an interval @xmath16 $ ] to the essential spectrum of the cosserat operator .
this has been known for a while already @xcite , but the proof has not yet been published . a corollary is an upper bound for the lbb constant , see also @xcite @xmath17 for the square @xmath18 , in ( * ? ? ?
* ( 6.39 ) ) a conjecture was offered that amounts to @xmath19 , which is obviously incompatible with the upper bound @xmath20 from .
although already conjectured in @xcite , it is still unknown whether this latter inequality is an equality or whether it is strict , that is , whether for the square there exist cosserat eigenvalues below the minimum of the essential spectrum . in section [ s : rectangles ]
below , we present numerical evidence suggesting that there are no such eigenvalues .
but this is not yet proven . for three - dimensional domains with edges , the two - dimensional corner domain transversal to
the edge implies an inclusion of the corresponding interval in the essential spectrum , which therefore always contains such an interval symmetric with respect to the point @xmath13 ( see section [ ss : edges ] ) .
this is very different from the case of smooth domains , where there exists the known example of the ball that has a cosserat spectrum consisting ( apart from the trivial point @xmath12 ) of a sequence of eigenvalues @xmath21 , @xmath22 , converging to @xmath13 from below and therefore has no spectrum in the interval @xmath23@xcite . for three - dimensional bounded domains having conical boundary points with tangent cones of revolution , we present in section [ ss : cones ] numerical results showing that there are also intervals contained in the essential spectrum of the cosserat operator .
the cosserat eigenvalue problem as originally formulated by e. and f. cosserat @xcite can be written as follows : find nontrivial @xmath24 and @xmath25 such that @xmath26 this is the spectral problem of the bounded selfadjoint operator @xmath27 on @xmath28 , where @xmath29 is the inverse of the laplace operator with dirichlet conditions @xmath30 . on the orthogonal complement of the kernel of @xmath31 , which is the eigenspace for the trivial eigenvalue @xmath32 , this operator is equivalent to the operator @xmath33 the schur complement operator of the stokes system .
we define the _ cosserat constant _
@xmath34 of the domain @xmath7 as the minimum of the spectrum of the operator @xmath35 .
it is then an exercise in elementary hilbert space theory to show that there holds @xmath36 to see this , write the lbb constant using the definition of the @xmath37 norm and the fact that @xmath38 is an isometry from @xmath39 to @xmath40 : @xmath41 for numerical approximations of the cosserat eigenvalue problem , we will use the equivalent formulation as an eigenvalue problem for the stokes system : find @xmath24 , @xmath42 , @xmath25 such that @xmath43 in the following , we will mainly discuss the essential spectrum of the cosserat problem .
this is the set of @xmath25 such that the operator @xmath44 is not a fredholm operator . from the variational formulation of the operator @xmath45 for @xmath46 @xmath47 it is not hard to see that such @xmath48 have to be real , between @xmath11 and @xmath12 , and that @xmath45 is fredholm if and only if it is fredholm of index @xmath11 and if and only if it is semi - fredholm .
we will show how kondratev s classical theory @xcite applies to the cosserat operator @xmath49 when @xmath7 is a _ domain with conical points _ in @xmath50 .
this means that the boundary of @xmath7 is smooth except in a finite set @xmath51 of points @xmath52 , called the corners of @xmath7 , and in the neighborhood of each corner @xmath52 the domain @xmath7 is locally diffeomorphic to a _ regular cone _
@xmath53 , i.e. , the section @xmath54 is a smooth domain in @xmath55 . from the discussion in @xcite , we know that the system @xmath56 is elliptic at any point of @xmath7 if ( and only if ) @xmath57 , and that the dirichlet boundary condition covers @xmath56 at any smooth point of @xmath58 if , moreover , @xmath59 .
in contrast with the smooth case when @xmath60 are the only values for which @xmath49 is not fredholm , the corners of @xmath7 cause an enlargement of this exceptional set , in general . for determining the values @xmath48
such that @xmath45 in is not fredholm , we use kondratev s @xcite technique of corner localization and mellin transformation . let us choose a corner @xmath52 and write @xmath61 for the polar coordinates in the tangent cone @xmath62 .
the homogeneous lam system @xmath56 can be written as @xmath63 where the @xmath64 system @xmath65 has coefficients independent of @xmath66 .
the mellin transformation @xmath67 transforms @xmath68 into the multiplication by @xmath69 .
the _ mellin symbol _
@xmath70 at the corner @xmath52 is the operator valued function ( known as _ operator pencil _ in the russian literature ) defined as @xmath71 in kondratev s theory , the fredholm property of @xmath49 is studied in the framework of weighted sobolev spaces @xcite . in the book @xcite by kozlov - mazya - rossmann , these spaces are defined as follows : for @xmath72 and @xmath73 @xmath74 [ l : h10 ] let @xmath75 be a domain with conical points .
the space @xmath6 coincides with the subspace of @xmath76 consisting of functions with zero boundary traces .
this lemma is proved using a poincar inequality in angular variables , see ( * ? ? ?
* lemma 6.6.1 ) .
then ( * ? ? ?
* theorem 6.3.3 ) implies the following result .
[ t : mellin ] let @xmath75 be a domain with conical points .
assume that @xmath48 does not belong to @xmath77 and that for each corner @xmath52 and all complex numbers @xmath69 with real part @xmath78 , the mellin symbol @xmath79 is invertible .
then @xmath49 is fredholm .
note that we can not apply ( * ? ? ?
* theorem 3.1 ) right away because it is assumed there that the sobolev exponent @xmath80 is at least @xmath81 .
for this reason we have to use the extension performed in @xcite .
the assumption @xmath82 ensures that the boundary value problem @xmath49 is elliptic . concerning the critical abscissa @xmath78
, there is a simple rule of thumb for determining it : a corner @xmath52 being chosen together with a non - zero angular function @xmath83 and a compactly supported cut - off function @xmath84 equal to @xmath12 near @xmath52 , we find that @xmath78 is the smallest real number @xmath85 such that all functions of the form @xmath86 belong to @xmath87 , whenever @xmath88 .
let @xmath89 be a polygon , that is , a bounded lipschitz domain the boundary @xmath58 of which consists of a finite set of straight segments .
this set being chosen in a minimal way , the corners @xmath52 of @xmath7 are the ends of these segments .
each corner @xmath52 belongs to two neighboring segments and the tangent cone @xmath62 is a plane sector , the opening of which is denoted by @xmath90 .
thus the section @xmath91 can be identified with the interval @xmath92 . in two - dimension ,
the invertibility of the mellin symbol @xmath79 , which is a @xmath93 sturm - liouville system on an interval with polynomial dependence on @xmath69 , can be further reduced to the non - vanishing of a scalar holomorphic function , which we may call mellin determinant @xmath94 .
this mellin determinant is well known for the case of the dirichlet problem of the lam system of linear elasticity .
it is constructed and described in detail in the book ( * ? ? ?
* chapter 3 ) , where also references to earlier works can be found .
we can use the results of these calculations , simply by noticing that the operator @xmath45 _ is _ the lam operator , if we set @xmath95 with the poisson ratio @xmath96 .
what is non - standard here , compared to the discussion in @xcite , is first that the @xmath48 considered here correspond to the `` non - physical '' range @xmath97 , and second that we consider the question of loss of @xmath8 regularity , that is , zeros of the mellin determinant on the critical line @xmath98 . in (
* ( 3.1.22/23 ) ) , the mellin determinant for a corner of opening angle @xmath99 is given as @xmath100 the characteristic equation @xmath101 can therefore be written as the two equations @xmath102 in the following paragraph [ ss : singfun ] we present a new and straightforward way to calculate @xmath103 together with the _ singular functions _
associated with the roots @xmath69 of @xmath103 , i.e. the solutions @xmath104 of the equation @xmath105 that are homogeneous of degree @xmath69 .
[ r : sings ] these singular functions play a double role here : if @xmath106 , they do not belong to @xmath107 near the corner , which is the reason why @xmath45 is then not a fredholm operator and @xmath48 belongs to the essential spectrum , see theorems [ l : charpolyg ] and [ t : esspolyg ] .
figure [ f3.1 ] shows this case : as a function of @xmath48 and @xmath99 , the imaginary part of @xmath69 is plotted . on the other hand ,
if @xmath108 , then the singular functions @xmath104 belong to @xmath107 near the corner and they describe the corner asymptotics of any solution in @xmath39 of @xmath109 with smooth @xmath110 , in particular of cosserat eigenfunctions for such eigenvalues @xmath48 that are not in the essential spectrum : @xmath111 if @xmath112 is close to zero , the derivatives of @xmath113 ( thus the `` pressure '' part @xmath114 of the stokes solution in ) will go to infinity quickly at the corner , which poses problems for the numerical approximation of the cosserat eigenvalue problem , see the discussion in section [ s : rectangles ] . in figure [ f3.2 ] , we plot @xmath69 as function of @xmath48 and @xmath99 in this case . note that for @xmath115 , @xmath69 is either real or purely imaginary . for purely imaginary roots @xmath69 of the characteristic equation as a function of the opening @xmath116 ( abscissa ) and the cosserat spectral parameter @xmath117 $ ] ( ordinate ) . ] of the characteristic equation as a function of the opening @xmath116 ( abscissa ) and the cosserat spectral parameter @xmath117 $ ] ( ordinate ) . ]
let us choose a corner @xmath52 and drop the reference to that corner in the notation .
we assume for simplicity that @xmath52 is the origin . the symbol @xmath118 is not invertible iff it has a nonzero kernel .
we note that for @xmath119 we have the equivalence @xmath120 so we first solve @xmath121 _ without boundary conditions _ and in a second step find conditions on @xmath69 so that there exist nonzero solutions satisfying the dirichlet conditions at @xmath122 .
we know from ( * ? ? ?
* theorem 2.1 ) that for any @xmath123 the solutions of form a space of dimension 4 , which we denote by @xmath124 .
moreover ( * ? ? ?
* b(iii ) ) tells us that if @xmath125 , it is sufficient to look for each component of the homogeneous function @xmath126 in the space generated by @xmath127 , @xmath128 , @xmath129 , and @xmath130 .
here we have identified @xmath131 with the complex plane @xmath132 by the formula @xmath133 .
so the ansatz space for @xmath134 has the basis @xmath135 applying @xmath105 to these vector functions is an easy computation using complex derivatives for functions of @xmath136 and @xmath137 and writing for a scalar function @xmath138 the formulas @xmath139 and @xmath140 in this way we find that for any @xmath141 , the following four vectors @xmath142 given by @xmath143 form a basis , we could define @xmath144 by @xmath145 and @xmath146 by @xmath147 . ] for the space @xmath124 .
we look for conditions on @xmath69 so that there exists a nonzero @xmath148 which satisfies the dirichlet conditions on @xmath122 .
we note that , setting @xmath149 for any @xmath150 the function @xmath151 is zero on @xmath122 .
it is clear that @xmath152 belongs to @xmath124 iff @xmath153 and @xmath154 , i.e. @xmath155 note that this is the same equation as .
the associated singular function is @xmath156 the result of the kondratev theory for our polygon can be written as follows .
[ l : charpolyg ] let @xmath7 be a polygon in the plane .
assume that @xmath48 does not belong to @xmath77 .
then the cosserat operator @xmath45 is fredholm from @xmath157 to @xmath158 if and only if for each corner @xmath159 and @xmath160 the characteristic equations have no solution on the line @xmath106 . the if part is a consequence of theorem [ t : mellin ] . for the `` only if '' part , we could simply quote ( * ? ? ?
* remark 6.3.4 ) , but we prefer to present a rather simple explicit proof in the spirit of @xcite , namely the construction of a singular sequence approximating the non-@xmath8 corner singularity .
choose a corner @xmath52 , which we assume to sit at the origin , set @xmath161 .
let @xmath162 be such that @xmath7 coincides with the plane sector @xmath62 in the ball of center @xmath52 and radius @xmath163 .
now choose a cut - off function @xmath164 such that @xmath165 for @xmath166 and @xmath167 for @xmath168 , and for any complex @xmath69 set @xmath169 with @xmath104 the singular function defined in .
then @xmath170 on @xmath58 .
therefore @xmath171 if and only if @xmath172 .
let @xmath48 be such that the equation has a solution @xmath173 , and take @xmath174 so that @xmath175 .
define @xmath176 then @xmath177 from @xmath178 follows that @xmath179 for @xmath180 $ ] and that @xmath181 can be estimated by @xmath182 , which remains bounded as @xmath183 .
note that for @xmath183 , @xmath184 and @xmath185 remains bounded , while @xmath186 .
using the fact that @xmath38 is an isometry from @xmath39 to @xmath40 , we obtain with @xmath187 altogether , this shows that @xmath45 can not satisfy an a - priori estimate of the form @xmath188 hence @xmath45 is not fredholm and @xmath48 belongs to the essential spectrum .
we can now conclude in the case of polygonal domains .
[ t : esspolyg ] let @xmath89 be a polygon with corner angles @xmath189 , @xmath159 .
then the essential spectrum of the cosserat problem in @xmath7 is given by @xmath190\;.\ ] ] the lbb constant satisfies @xmath191 as @xmath7 is lipschitz , @xmath10 @xcite .
for @xmath192 , we have to find the @xmath48 for which has solutions @xmath69 with @xmath106 .
the function @xmath193 maps @xmath194 to @xmath195 . for @xmath196 ,
the necessary and sufficient condition is therefore @xmath197 this proves .
[ r : curvedpol ] the result of theorem [ t : esspolyg ] remains true if @xmath7 is a curved polygon , that is a bounded lipschitz domain with a piecewise @xmath198 boundary .
this follows from general perturbation techniques that are part of the kondratev theory @xcite , see also ( * ? ? ?
* ch . 6 ) .
if the corner angles have cusps , theorem [ t : esspolyg ] extends as follows : an inward cusp or crack ( @xmath199 ) does not contribute to the nontrivial essential spectrum @xcite , whereas in the presence of an outward cusp ( @xmath200 ) , the essential spectrum of the cosserat problem fills the entire interval @xmath201 $ ] @xcite .
the estimate for @xmath0 was proved by stoyan in @xcite , using the equivalence between the friedrichs constant and the lbb constant @xcite .
let @xmath202 . we consider domains with edges and corners . for the case when @xmath7 contains edges in its boundary , we prove that the essential spectrum of the cosserat problem contains an interval that depends on the opening angle of the edges . for the case when @xmath7 has a conical point with an axisymmetric tangent cone , we describe how to reduce the question to the determination of the roots of a sequence of holomorphic functions and we show the result of numerical computations .
polyhedral corners are the subject of some comments at the end of this section . in a first step ,
we assume that the boundary of @xmath7 contains a piece of a straight edge of opening @xmath15 in the sense that for some @xmath162 @xmath203 [ t : essedge ] under the hypothesis , the interval @xmath204 $ ] is contained in the essential spectrum of the cosserat problem .
let @xmath205\setminus\{\frac12\}$ ] .
we construct a singular sequence based on the one that we have used for a two - dimensional corner in section [ ss : singseq ] : like there , we set @xmath206 , and for @xmath123 we define @xmath207 as in equation and use the same cut - off function @xmath84 .
now we choose another cut - off function @xmath208 that satisfies @xmath209 if @xmath210 and @xmath211 if @xmath212
. then we define @xmath213 as @xmath214 for @xmath69 , @xmath215 , @xmath216 chosen as before , see , the vector functions @xmath217 belong to @xmath218 and @xmath219 has still the expression with @xmath220 given by @xmath221 we conclude as before that @xmath185 and @xmath222 remain bounded as @xmath183 whereas @xmath186 . as for @xmath223 , we see that @xmath224 remains bounded , so that we can again conclude that @xmath225 is a singular sequence satisfying , and that @xmath48 therefore belongs to the essential spectrum .
[ r : edge ] yet another contribution to the essential spectrum may come from the edges , this time not from the essential spectrum of the transversal cosserat problem @xmath226 as described in theorem [ t : essedge ] , but from _ eigenvalues _ of the `` edge symbol '' , which is the boundary value problem defined on the sector @xmath227 obtained by partial fourier transformation along the edge .
indeed , it is known that for an elliptic boundary value problem on a domain with edges to be fredholm , it is not sufficient that its edge symbol is fredholm ( this latter condition is satisfied as soon as the transversal mellin symbol is invertible on a certain line @xmath228 ) , but it has to be _
invertible _ for all @xmath229 ( * ? ? ?
* theorem 10.1 ) , see also @xcite . for our cosserat problem ,
if such eigenvalues exist below the essential spectrum of the transversal problem @xmath226 , then the lower bound of the essential spectrum on @xmath7 , and therefore the lbb constant , may be smaller than what is described by theorem [ t : essedge ] .
whether this actually happens is unknown , and some numerical experiments we did rather seem to indicate that it does not . [ r : curvededge ] the result of theorem [ t : essedge ] remains true for some curved edges , too , similar to what we mentioned in two dimensions in remark [ r : curvedpol ] .
in particular , if @xmath7 is a finite straight cylinder with a smooth base , then the interval corresponding to the angle @xmath230 , namely @xmath231 $ ] , will belong to the essential spectrum .
we assume now that @xmath7 is a domain in @xmath232 with conical points .
theorem [ t : mellin ] applies with the critical abscissa @xmath233 .
we can prove exactly like in dimension 2 that the sufficient condition for fredholmness is also necessary .
now the question is : does a 3d cone produce a full interval of nonzero length inside the essential spectrum ?
do we have symmetry with respect to @xmath234 like in 2d ?
we do not have a general answer to the first question , but a partial answer in the case of axisymmetric cones .
this example shows that the symmetry property with respect to @xmath234 is lost .
we learn from @xcite that semi - analytical calculation of the spectrum of the mellin symbol @xmath235 associated with an axisymmetric cone @xmath236 is made possible by the use of the boussinesq representation and separation of variables in spherical coordinates . then , relying on the equivalence , we can follow the same track as in 2d , looking for homogeneous solutions @xmath152 of the equation @xmath237 without boundary conditions in a first step , and imposing the homogeneous dirichlet conditions in a second step .
we give a rapid overview of this procedure ( see @xcite for details ) .
the boussinesq representation for solutions of the equation states that @xmath152 can be found as a linear combination of the three following particular solutions of the same equation @xmath238 where the scalar functions @xmath239 , @xmath240 , and @xmath241 are harmonic functions . here
we choose @xmath242 as the director of the cone axis .
then homogeneous @xmath152 s of degree @xmath69 are given by finding @xmath239 and @xmath240 homogeneous of degree @xmath243 , and @xmath241 of degree @xmath69 .
using spherical coordinates @xmath244 such that @xmath245 we split the 3d problem into an infinite sequence of 2d problems in @xmath246 parametrized by the azimuthal frequency @xmath247 .
the harmonic generating function can be written in separated variables as @xmath248 where @xmath249 is the associated legendre function of the first kind of order @xmath250 and degree @xmath96 . combining and
, we find for each @xmath247 three independent solutions @xmath251 , @xmath252 , and @xmath253 of degree @xmath69 for equation .
: color plot of the decimal logarithm of @xmath254 of roots @xmath69 with real part @xmath255 of characteristic equation @xmath256 as a function of the opening @xmath99 ( in degrees , abscissa ) and the parameter @xmath117 $ ] ( ordinate )
. ] denote by @xmath99 the opening of the cone @xmath236 , which means that @xmath236 is defined by the condition @xmath257 in spherical coordinates .
the exponents @xmath69 we are looking for are those for which there exists an integer @xmath258 such that the following @xmath259 matrix is singular for all @xmath260 and @xmath261 @xmath262 using the special forms and , the variables @xmath66 and @xmath261 disappear and after some obvious simplification we are left with the matrix in @xmath263 is missing in the last term of the second column . nevertheless , the numerical computations presented in these latter references where made with the correct formulas . ] : boundary of regions @xmath264 ( solid line ) and @xmath265 ( dashed line ) in the plane @xmath266 : opening @xmath99 ( in degrees , abscissa ) and @xmath117 $ ] ( ordinate ) . ]
@xmath267 \begin{minipage}{0.3\textwidth } $ ( \lambda+1 ) \cos\omega p^{-m}_{\lambda+1}(\cos\omega)$ \\
$ - ( \lambda+1-m)p^{-m}_{\lambda}(\cos\omega)$ \end{minipage } & m\cos\omega p^{-m}_{\lambda+1}(\cos\omega ) & \begin{minipage}{0.3\textwidth } $ ( \lambda+1+m)\cos\omega \,p^{-m}_{\lambda+1}(\cos\omega)$\\ $ + ( 1 - 2\sigma)\sin^2\omega \,p^{-m}_{\lambda}(\cos\omega)$ \\ $ - ( \lambda+1)\cos^2\omega \,p^{-m}_{\lambda}(\cos\omega)$ \end{minipage } \\[2ex ]
-m p^{-m}_{\lambda+1}(\cos\omega ) & \begin{minipage}{0.3\textwidth } $ ( \lambda+1-m)\cos\omega \,p^{-m}_{\lambda}(\cos\omega)$ \\ $ -(\lambda+1)\,p^{-m}_{\lambda+1}(\cos\omega)$ \end{minipage } & -m\cos\omega\,p^{-m}_{\lambda}(\cos\omega ) \end{pmatrix}\ ] ] with this matrix at hand , it is possible to compute its determinant @xmath268 and find , for any chosen @xmath258 , couples @xmath266 for which the equation @xmath269 has roots on the line @xmath233 .
we present in fig .
[ f4.1 ] the region @xmath264 of the @xmath266 plane where such roots can be found when @xmath270 .
this region is clearly non - symmetric with respect to @xmath234 .
the region @xmath265 that we have calculated for @xmath271 ( which is the same for @xmath272 ) is strictly contained in the region @xmath264 associated with @xmath270 , see fig . [ f4.2 ] .
we observe that when @xmath273 is increasing , the region @xmath274 is shrinking .
by polyhedron we understand a bounded lipschitz domain @xmath275 the boundary of which consists of a finite set of plane polygons , the faces .
this set being chosen in a minimal way , the edges @xmath276 of @xmath7 are the segments which form the boundaries of the faces , and the corners @xmath52 are the corners of the faces .
let @xmath277 and @xmath51 be the sets of edges and corners , respectively . to each edge @xmath276 is associated a transversal sector @xmath278 and its opening @xmath279 . to each corner @xmath52
is associated the tangent infinite polyhedral cone @xmath62 and its section @xmath280 .
let @xmath25 be the cosserat spectral parameter .
for each edge @xmath276 , the edge symbol @xmath281 is defined by .
for each corner @xmath52 , the corner mellin symbol @xmath282 is defined by . if @xmath82 , and if the following two conditions are satisfied @xmath283 then @xmath49 is fredholm on @xmath7 .
these conditions are also necessary and determine the essential spectrum of the operator @xmath35 .
we recall that , cf .
remark [ r : edge ] , the condition @xmath284\ ] ] is necessary for to hold but is , possibly , not sufficient .
for any rectangle @xmath89 , theorem [ t : esspolyg ] shows that the essential spectrum of the cosserat problem is the interval @xmath231 $ ] .
therefore the cosserat constant of @xmath7 satisfies @xmath285 in contrast to the essential spectrum , the discrete spectrum depends on the actual shape of the rectangle , namely of its aspect ratio ( but not on its width and height separately ) .
our convention is to characterize a rectangle by the number @xmath286 $ ] such that its aspect ratio is @xmath287:@xmath288 or @xmath289:@xmath290 .
the square has its factor @xmath291 equal to @xmath12 , and small @xmath291 corresponds to elongated rectangles ( large aspect ratio ) . to determine @xmath34 and therefore the lbb constant
, one can solve the cosserat eigenvalue problem numerically and look for the smallest eigenvalue .
a conforming discretization based on the stokes eigenvalue problem consists in choosing a pair of finite dimensional spaces @xmath292 and @xmath293 and constructing the following discrete version of the schur complement @xmath294 where @xmath295 is the stiffness matrix associated with the @xmath296 bilinear form on @xmath297 and @xmath298 is associated with the bilinear form @xmath299 on @xmath300 , @xmath301 .
the discrete cosserat eigenvalues @xmath302 , @xmath303 , are the non - zero eigenvalues of @xmath304 ordered increasingly .
two main difficulties are encountered : 1 .
the discrete pair @xmath305 constructed from finite element spaces may have a behavior of its own , somewhat independent of the continuous pair @xmath306 .
this may give rise to spurious eigenvalues .
the regularity of the eigenvectors @xmath114 of @xmath35 depends on the eigenvalue @xmath48 and gets worse as @xmath48 is closer to the essential spectrum , cf remark [ r : sings ] . in figure [ f5.0 ]
we plot the supremum of exponents @xmath307 such that @xmath114 belongs to @xmath308 : this is the minimal @xmath112 for @xmath69 solution of @xmath309 with positive real part , cf .
there exist estimates for @xmath34 from above and from below which prove that @xmath311 like @xmath312 as @xmath313 : @xmath314 the lower bound is deduced from the horgan - payne estimate @xcite and is equal to @xmath315 modulo @xmath316 . the upper bound is obtained by plugging the quasimode @xmath317 on the rectangle @xmath318 into the rayleigh quotient of the schur complement @xmath35 ( see lemma [ l : channel ] below for a more general estimate of this type ) .
it is an improvement at the order @xmath316 of the upper bound @xmath319 proven in @xcite by chizhonkov and olshanskii .
note that the upper bound in proves that for any @xmath320 , the bottom @xmath34 of the spectrum of @xmath35 is an eigenvalue . to obtain an upper bound for the cuboids @xmath322 of dimensions
@xmath3231@xmath3241 , we take @xmath325 in lemma [ l : channel ]
. we can calculate @xmath326 by fourier expansion in @xmath99 : starting from @xmath327 we find : @xmath328 hence @xmath329 and yields for our cuboids @xmath330 we observe in figure [ f5.3 ] that our computations are in good agreement with the bound . nevertheless , they are more difficult to interpret than in 2d .
the possible presence of spurious eigenvalues is not easy to distinguish from the manifestation of the essential spectrum .
moreover , at this stage , it is an open question whether the bottom of the essential spectrum comes from edges or from corners .
numerical experiments on cylinders with circular or annular sections , which are compatible with the upper bound , tend to suggest that the essential spectrum coming from the edges is restricted to @xmath335 $ ] . therefore we may conjecture that we see in figure [ f5.3 ] a manifestation of the essential spectrum coming from the corners of the cuboids .
, _ on an operator related to the convergence of uzawa s algorithm for the stokes equation .
_ , in computational science for the 21st century , m .- o .
bristeau , g. etgen , w. fitzgibbon , j. lions , j. priaux , and m. wheeler , eds .
, chichester : john wiley & sons , 1997 , pp .
242249 . height 2pt depth -1.6pt width 23pt , _ spectral problems associated with corner singularities of solutions to elliptic equations _ , mathematical surveys and monographs , 85 , american mathematical society , providence , ri , 2001 . , _
@xmath336 estimates of solutions of elliptic boundary value problems in a domain with edges _ , trans .
moscow math .
soc . , 1 ( 1980 ) , pp .
russian original in _ trudy moskov . mat .
* 37 * ( 1978 ) . | the inf - sup constant for the divergence , or lbb constant , is related to the cosserat spectrum .
it has been known for a long time that on non - smooth domains the cosserat operator has a non - trivial essential spectrum , which can be used to bound the lbb constant from above .
we prove that the essential spectrum on a plane polygon consists of an interval related to the corner angles and that on three - dimensional domains with edges , the essential spectrum contains such an interval .
we obtain some numerical evidence for the extent of the essential spectrum on domains with axisymmetric conical points by computing the roots of explicitly given holomorphic functions related to the corner mellin symbol . using finite element discretizations of the stokes problem , we present numerical results pertaining to the question of the existence of eigenvalues below the essential spectrum on rectangles and cuboids . |
general relativity is not only relativistic theory of gravitation proposed by einstein , but it is the simplest theory that is consistent with experimental data .
predictions of general relativity have been confirmed in all observations and experiments .
gravitational redshift of light is one of the predictions of general relativity and also provides evidence for the validity of the principle of equivalence .
any relativistic theory of gravitation consistent with the principle of equivalence will predict a redshift .
+ if , however , we observe on the earth the spectrum emitted by the atoms located on the sun , then , its lines appear to be shifted with respect to the lines of the same spectrum emitted on the earth .
each line with frequency @xmath0 will be shifted through the interval @xmath1 given by formula : @xmath2 where @xmath3 and @xmath4 are the potentials of the gravitational field at the points of emission and observation of the spectrum respectively ( page 269 , of landau and lifshitz [ 1 ] ) .
if we observe on the earth a spectrum emitted on the sun or the stars , then magnitude of @xmath3 is greater than magnitude of @xmath4 and it is clear that @xmath50 , i.e , the shift occurs in the direction of lower frequency . the phenomenon we have described is called the gravitational redshift .
+ adams in 1925 has claimed first about the confirmation of the predicted gravitational redshift from the measurement of the apparent radial velocity of sirius b [ 2 ] .
pound and rebka in 1959 were the first to experimentally verify the gravitational redshift from nuclear - resonance [ 3 ] .
pound and snider in 1965 had performed an improved version of the experiment of pound and rebka , to measure the effect of gravity , making use of mossbauer - effect [ 4 ] .
if the source of radiation is at a height _ h _ above an observer then the result found was ( 0.9990 @xmath6 0.0076 ) times the value of @xmath7 predicted from principle of equivalence .
snider in 1972 has measured the redshift of the solar potassium absorption line at 7699 @xmath8 by using an atomic - beam resonance - scattering technique [ 5 ] .
krisher et al . in 1993 had measured the gravitational redshift of sun [ 6 ] . +
gravitational redshift has been reported by most of the authors without consideration of rotation of a body . neglecting the rotation ,
the geometry of space time can be described using the well - known spherically symmetric schwarzschild s geometry and information on the ratio @xmath9 of a compact object can be obtained from the gravitational redshift , where m and r are mass and radius respectively .
thus the redshifted angular frequency ( @xmath10 ) and the original angular frequency ( @xmath11 ) of a photon in schwarzschild geometry are related by the relation ( page 268 , of landau and lifshitz [ 1 ] ) @xmath12 where @xmath13 is the schwarzschild radius .
p d nunez and m nowakowski in 2010 had obtained an expression for gravitational redshift factor of rotating body by using small perturbations to the schwarzschild s geometry .
two main results of this work include the derivation of a maximum angular velocity depending only on the mass of the object and a possible estimates of the radius [ 7 ] .
+ + with this background the present paper is organized as follows . in section -ii , we derive the expression for four velocity and frame dragging in kerr field . in section -iii
, we derive the expression for gravitational redshift factor ( @xmath14 ) and redshift ( z ) as the pulsar rotates .
the derived expression of gravitational redshift is applied to calculate the values of gravitational redshift for various rotational bodies which includes the sun and some millisecond pulsars . a pulsar
as it rotates , the redshifted values of spectrum shows a periodicity . in section -iv
, we derive the expression for the coefficient of latitude dependence of redshift ( @xmath15 ) .
we further discuss its variation for different values of latitude at which photon is emitted .
finally , some conclusions are made in section -v .
when rotation is taken into consideration spherical symmetry is lost and off - diagonal terms appear in the metric and the most useful form of the solution of kerr family is given in terms of t , r , @xmath16 and @xmath17 , where t , and r are boyer - lindquist coordinates running from - @xmath18 to + @xmath18 , @xmath16 and @xmath17 , are ordinary spherical coordinates in which @xmath17 is periodic with period of 2 @xmath19 and @xmath16 runs from 0 to @xmath20 .
+ covariant form of metric tensor with signature ( + , -,-,- ) is expressed as @xmath21 + non - zero components @xmath22 of kerr family are given as follows ( page 346 , of landau and lifshitz [ 1 ] and page 261 - 263 , of carroll [ 8 ] ) , @xmath23 @xmath24 @xmath25 @xmath26sin^{2}\theta}{\rho^{2}}=-[r^{2}+a^{2}+\dfrac{r_{g } r a^{2}sin^{2}\theta}{\rho^{2}}]sin^{2}\theta\ ] ] @xmath27 with @xmath28 @xmath29 where the parameters e and @xmath30 are respectively charge and rotation parameter of the source .
j is the angular momentum of the compact object or central body , which can be also written as @xmath31 .
m , i and @xmath32 are the mass , moment of inertia and angular velocity of the central body respectively . charge ( e ) is written as @xmath33 , where q is electric charge and p is magnetic charge . in case of kerr solution in 1963 both q and p vanishes [ 9 ] , for kerr - newman solution in 1965 only p vanish [ 10 ] but in case of kerr - newman - kasuya solution in 1982 both q and p are non - vanishing [ 11 ] .
+ the values of @xmath34 , @xmath35 and @xmath36 for polar and equatorial regions can be obtained from equations ( 3 ) to ( 9 ) .
+ for equatorial plane where @xmath37 , from equations ( 8) and ( 9 ) , if we consider charge e = 0 , then @xmath38 and @xmath39 .
+ from equations ( 3 ) , ( 6 ) and ( 7 ) metric elements @xmath40 , @xmath41 and @xmath35 for kerr space time at equator becomes ,
@xmath42 @xmath43 @xmath44 + at poles where @xmath45 , from equations ( 8) and ( 9 ) , if we consider charge e = 0 , then @xmath46 and @xmath47 . + from equations ( 3 ) , ( 6 ) and ( 7 ) metric elements @xmath40 , @xmath41 and @xmath35 for kerr space time at poles become , @xmath48 @xmath49 @xmath50
four - dimensional velocity ( four - velocity ) of a particle is defined as the four - vector ( page 23 , of landau and lifshitz [ 1 ] ) @xmath51 where the index i takes on the values 0,1,2,3 and @xmath52 , @xmath53 , @xmath54 , @xmath55 .
+ using equation ( 2 ) we can write @xmath56 the components of four - velocity are @xmath57 @xmath58 @xmath59 @xmath60 from equations ( 18 ) and ( 21 ) we can write @xmath61 for a sphere , the photon is emitted at a location on its surface where @xmath62 , when the sphere rotates .
+ so the expressions for four - velocity becomes , @xmath63
@xmath64 @xmath65 @xmath66 hence four - velocity of an object in kerr field can be expressed as , @xmath67 the lagrangian @xmath68 of the test particle ( * with mass @xmath69 * ) can be expressed as @xmath70 ( dot over a symbol denotes ordinary differentiation with respect to an affine parameter @xmath71 ) + from the lagrangian of the test particle , we can obtain the momentum of the test particle as , @xmath72 thus corresponding momentum in the coordinates of t , r , @xmath16 , and @xmath17 are given by @xmath73 @xmath74 @xmath75 @xmath76 since the space time of the kerr family is stationary and axially symmetric , the momenta @xmath77 and @xmath78 are conserved along the geodesics .
so we obtain two constants of motion : one is corresponding to the conservation of energy ( e ) and the other is the angular momentum ( l ) about the symmetry axis .
+ we can write from equations ( 30 ) and ( 33 ) @xmath79 and @xmath80 using equations ( 34 ) and ( 35 ) , we can obtain @xmath81 @xmath82 substituting the value of @xmath34 , @xmath36 , and @xmath35 from equations ( 3 ) , ( 6 ) and ( 7 ) in above equation ( 36 ) , @xmath83 @xmath84 alternatively following landau and lifshitz ( [ 1 ] page 351 ) , we can also find @xmath81 in the following way : + the four momentum of the test particle ( with mass @xmath69 ) is @xmath85 @xmath86 @xmath87 from above equations ( 39 ) and ( 40 ) , we can find @xmath81 as @xmath88 the above expression is identical to the expression ( 37 ) obtained earlier , where e and l are conserved energy and components of angular momentum along the axis of symmetry of the field respectively .
+ frame dragging is a general relativistic feature of all solutions to the einstein field equations associated with rotating masses . due to the influence of gravity frame dragging or dragging of inertial
frame arises in the kerr metric .
@xmath81 is termed as angular velocity of frame dragging [ 12 ] .
the equation ( 37 ) is the general expression of angular velocity of frame dragging ( @xmath81 ) in kerr field .
+ following ( cunningham and bardeen 1972 [ 13 ] ) , the photon trajectory is independent of its energy and may be described by the parameter @xmath89 , which is related to the direction cosines of a beam of radiation with respect to the @xmath17 - direction and @xmath16 - direction in the local rest frame .
+ in case of photon the energy ( e ) and linear momentum ( p ) are expressed as @xmath90 where @xmath91 is rest mass of the photon which is zero .
so above equation becomes @xmath92 + the angular momentum ( l ) about the symmetry axis may be expressed as @xmath93 we may note that the symbol @xmath68 stands for lagrangian and l for angular momentum about the symmetry axis . +
it is assumed that b@xmath94 and b @xmath95 .
here b is perpendicular distance between the axis of rotation of the central body and direction of ray of light . + considering the ray of light which is emitted radially outward from the surface of the star ( pulsar or sun ) with radius r , we can substitute , @xmath96 in above equation ( 43 ) and finally we can write @xmath97 where angle @xmath98 , is measured from the azimuthal axis , which is defined as the direction when source of light ( on the spherical surface ) is nearest to the observer .
+ @xmath99 , where t is period of rotation of star ( pulsar or sun ) .
+ substituting the value of e and l from equations ( 42 ) and ( 44 ) in the expression ( 37 ) of @xmath81 , @xmath100 @xmath101 for equatorial plane where @xmath37 and @xmath38 , the above expression ( 46 ) of @xmath81 can be written as : @xmath102 the above expression ( 47 ) is the expression of frame dragging @xmath103 on the equatorial plane . + again for any general @xmath16 and at @xmath104 , the expression for @xmath81 from equation ( 46 ) can be written as : @xmath105 substituting @xmath37 and @xmath38 , the above expression ( 48 ) of @xmath81 can be written as : @xmath106
let f be any quantity describing the field of the wave . for a plane monochromatic wave f has the form ( page 140 , of landau and lifshitz [ 1 ] ) @xmath107 where @xmath108 , and i used as superscript and subscript with indices i=0,1,2,3 .
+ k is propagation constant and @xmath109 is the wave four - vector . + we write the expression for the field in the form @xmath110 where @xmath111 , is defined as eikonal . + over small region of space and time intervals the eikonal @xmath112 can be expanded in series to terms of first order , we have @xmath113 as a result one can write ( page 141 , of landau and lifshitz [ 1 ] ) @xmath114 where @xmath109 is the wave four - vector and the components of the four - wave vector @xmath109 are related by @xmath115 or @xmath116 this equation called the eikonal equation is fundamental equation of geometrical optics .
+ using equation ( 53 ) , the components of wave four - vector are @xmath117 @xmath118 @xmath119 @xmath120 hence we can write the wave - four vector @xmath121 @xmath122 is the time component of four - wave vector and this frequency is measured in terms of the world time . + frequency measured in terms of proper time ( @xmath123 )
is defined as @xmath124 this frequency @xmath125 is different at different point of space . @xmath126 as @xmath127 is nothing but @xmath128 . + substituting the value of @xmath129 in above equation ( 62 ) from equation ( 23 ) @xmath130 this @xmath131 is the frequency measured by a distant observer in terms of proper time ( @xmath123 ) and @xmath0 is the frequency measured in terms of the world time ( t ) .
+ + * alternative method : frequency in kerr field * + + any observer measures the frequency ( @xmath131 ) of a photon following the null geodesic @xmath132 which can be calculated by the expression given as ( page 217 of carroll 2004 [ 8 ] and page 108 of straumann 1984 [ 14 ] , @xmath133 @xmath134 using equation ( 30 ) and ( 33 ) in the above equation ( 65 ) , we can write @xmath135 again using equation ( 26 ) in above equation ( 66 ) we can write : @xmath136 or @xmath137 the above expression is identical to the expression ( 62 ) obtained earlier . here
we define @xmath138 as @xmath0 , where @xmath0 can be identified as the frequency corresponding to world time as in equation ( 56 ) and equation ( 62 ) . + as a result , we can write the expression of frequency measured in terms of proper time or as observed by distant observer as : @xmath130 the above expression is identical to the expression ( 63 ) obtained earlier . thus the two approaches give us identical expression for frequency observed by distant observer . [ cols="^,^,^,^,^,^,^",options="header " , ] in general relativity , redshift ( z ) and redshift factor ( @xmath14 ) are defined as [ 15,16 ] , @xmath139 a redshift of zero corresponds to an un - shifted line , whereas @xmath140 indicates blue - shifted emission and @xmath141 red - shifted emission .
a redshift factor of unity corresponds to an un - shifted line , whereas @xmath142 indicates red - shifted emission and @xmath143 blue - shifted emission .
+ from equations ( 63 ) and ( 68 ) , we can write redshift factor ( @xmath14 ) as : @xmath144 using equation ( 68 ) , we can obtain redshift ( z ) from above equation ( 69 ) . +
if we ignore the electric and magnetic charges contribution , then using the values of metric coefficients @xmath34 , @xmath36 , and @xmath35 from equations ( 3 ) , ( 6 ) and ( 7 ) of kerr family in equation ( 69 ) and substituting the value of @xmath81 from equation ( 46 ) , we can obtain the redshift factor ( @xmath14 ) and corresponding redshift ( z ) for the values of @xmath16 from zero ( pole ) to @xmath145 ( equator ) and @xmath17 from zero to @xmath146 .
+ @xmath147 @xmath148 @xmath149 we can replace r by r which is the radius of the star . thus the redshift clearly depends on the co - ordinates @xmath16 and @xmath17 .
+ if the observer is on the axis of rotating star , we substitute @xmath150 and get the expression for redshift factor as : @xmath151 when we consider the rotation velocity of central body ( @xmath32 ) to be zero ( or spin parameter , a=0 ) , then we can obtain the corresponding gravitational redshift from a static body of same mass ( schwarzschild mass ) .
+ + now at @xmath16 = @xmath145 for equatorial plane , substituting the values of @xmath34 , @xmath36 , and @xmath35 from equations ( 10 ) , ( 11 ) and ( 12 ) and @xmath152 from equation ( 47 ) in expression ( 69 ) we get : @xmath153 @xmath154 + now as the photon is emitted from the surface of the star ( pulsar or sun ) , we can substitute the actual value for r , which is r. @xmath153 @xmath155 on simplifying @xmath156 we also express all the length scales in unit of @xmath157 , by substituting ; @xmath158 , and @xmath159 in above equation ( 74 ) .
thus after certain simplifications @xmath160 in above equation ( 75 ) @xmath161 and @xmath162 are dimensionless quantities and the redshift factor @xmath163 will show a periodic nature with respect to @xmath17 .
+ + for carrying out some numerical calculations we consider a lyman - @xmath164 line ( 1215.668 @xmath8 ) emitted from the surface of the star ( pulsar or sun ) .
+ for an asymptotic observer , the redshifted ( @xmath165 ) lyman - @xmath164 line can be calculated as , @xmath166 further for doing numerical calculations we have considered the newtonian approximation of angular momentum ( j ) , @xmath167 so rotation parameter can be written as @xmath168 where @xmath32 is angular velocity of rotating body ( pulsar or sun ) as described in section ii a. there are two possibilities for direction of spin ( @xmath169 ) of central body .
+ we have four cases for the rotation parameter ( a ) and the position ( @xmath16 ) combination on the central body at which photon is emitted ; * a = + i ve , @xmath16 = + i ve , @xmath170 * a = -ive , @xmath16 = -ive , @xmath171 * a = + i ve , @xmath16 = -ive , @xmath172 * a = -ive , @xmath16 = + i ve , @xmath173 the mass ( m ) , physical radius ( r ) , schwarzschild radius @xmath157 , rotation velocity ( @xmath32 ) , and rotation parameter ( @xmath174 ) of pulsars and sun are calculated from different sources in literature and these are listed in table - i. + + as can be seen from equation ( 75 ) each pulsar can be characterized by only two parameters @xmath161 and @xmath162 . in our case , we make table - ii , listing @xmath161 and @xmath162 values of these eight pulsars along with z@xmath175 , z@xmath176 , z@xmath177 and lyman - @xmath164 line redshifted wavelength @xmath178 values .
+ + in fig.-1 and fig.-2 , we make a plot showing variation of @xmath179 with @xmath17 for eight pulsars ( listed in table - i ) . we are making plot here for ( + , + ) or ( -,- ) , for ( a , @xmath16 ) combination .
it may be noted that @xmath17 is actually equal to @xmath180 , where t is the time measured in observers coordinate .
the nature is clearly periodic with different amplitudes at @xmath181 and @xmath182 .
thus for a pulsar as it rotates the redshifted values of spectrum shows a periodicity . from fig.-1 ,
fig.-2 and table - ii , it is clearly seen that the value of redshift ( z ) for these millisecond pulsars obtains a maximum ( primary maxima ) at @xmath183 and a second maximum ( secondary maxima with value less than the primary maxima ) at @xmath184 .
this asymmetry happens clearly due to the presence of @xmath185 and @xmath186 terms in the expression of @xmath14 as in equations ( 72 ) , ( 73 ) and ( 74 ) .
the value of redshift , at @xmath187 is actually same as that value of gravitational redshift which is generally referred ( in most of the published work ) as the gravitational redshift in kerr field .
the redshifted wavelength value @xmath188 is also given in table - ii .
this calculation have been done for lyman - @xmath164 line .
for comparison we make a similar plot ( fig.-3 ) for sun and the effect is too small .
below we define a parameter @xmath189 , showing latitude dependence of redshift . + from equation ( 76 ) , we can write @xmath190 the quantity @xmath189 thus , is defined as red - shifted wavelength ratio at any value of latitude on the central body ( star ) at which photon is emitted and at equator .
this parameter can be termed as coefficient of latitude dependence of redshift from a rotating star . + using the value of @xmath14 from equation ( 69 ) in above equation ( 79 )
, we can write @xmath191 _ { \theta=\frac{\pi}{2}}}{[\sqrt{{g_{tt}+ g_{\phi\phi}(\frac{d\phi}{c dt})^{2 } } + 2 g_{t \phi}(\frac{d\phi}{c dt } ) } ] _ { \theta}}\ ] ] here we can substitute the value of @xmath192 from equation ( 48 ) and @xmath34 , @xmath36 and @xmath35 from equations ( 3 ) , ( 6 ) and ( 7 ) for any value of @xmath16 .
we can also substitute the value of @xmath193 from equation ( 49 ) and @xmath34 , @xmath36 and @xmath35 from equations ( 10 ) , ( 11 ) and ( 12 ) for @xmath194 . + as a result we can write @xmath195 @xmath196sin^{2}\theta(\frac{\frac{r}{sin\theta } ( 1-\frac{r_{g}r}{\rho^{2}})+\frac{r_{g}ra}{\rho^{2}}}{-\frac{r_{g}r a}{\rho^{2 } } ( r sin\theta ) + ( r^{2}+a^{2}+\frac{r_{g}r a^{2 } } { \rho^{2}}sin^{2}\theta)})^{2 } } + 2 ( \frac{a sin^{2}\theta r_{g}r}{\rho^{2}})(\frac{\frac{r}{sin\theta } ( 1-\frac{r_{g}r}{\rho^{2}})+\frac{r_{g}ra}{\rho^{2}}}{-\frac{r_{g}r a}{\rho^{2 } } ( r sin\theta ) + ( r^{2}+a^{2}+\frac{r_{g}r a^{2 } } { \rho^{2}}sin^{2}\theta)})}}\ ] ] the denominator of the fraction in above equation ( 81 ) is redshift factor @xmath197 , which clearly depends on the direction of spin ( @xmath198 ) of central body and also on the position ( @xmath199 ) on the central body at which the photon is emitted . +
this also indicates that @xmath189 will also depend on the position @xmath200 and rotation parameter ( @xmath6 a ) .
+ + we obtain same results for redshift ( z ) and @xmath189 in case of ( + , + ) or ( -,- ) and similarly the same results in case of ( + , - ) or ( -,+ ) , for ( a , @xmath16 ) combination .
+ finally we have only two situations for the combination of rotation parameter ( a ) and the position ( @xmath16 ) on the central body at which photon is emitted ; ( + , + ) or ( -,- ) and ( + , - ) or ( -,+ ) , for ( a , @xmath16 ) combination .
+ + we can find @xmath189 for @xmath150 , which will give a ratio of redshifted wavelength between pole and equatorial region .
@xmath201 _ { \theta=\frac{\pi}{2}}}{[\sqrt{{g_{tt}+ g_{\phi\phi}(\frac{d\phi}{c dt})^{2 } } + 2 g_{t \phi}(\frac{d\phi}{c dt } ) } ] _ { \theta=0}}\ ] ] at poles ( @xmath202 ) , we use the values @xmath203 and @xmath204 , from equations ( 13 ) to ( 15 ) . therefore , equation ( 82 ) can be rewritten as @xmath201 _ { \theta=\frac{\pi}{2}}}{\sqrt{(1-\frac{r_{g}r}{r^{2}+a^{2}}})}\ ] ] substituting the value of @xmath193 from equation ( 49 ) and @xmath34 , @xmath36 and @xmath35 from equations ( 10 ) , ( 11 ) and ( 12 ) for @xmath205 , above equation ( 83 ) can be written as @xmath206 in fig.-4 and 5 , we show @xmath189 for eight pulsars listed in table - i and ii as a function of @xmath16 .
it is seen that the value of @xmath189 is less than one for ( + , + ) or ( -,- ) combination while greater than one for ( + , - ) or ( -,+ ) , for ( a , @xmath16 ) combination .
this clearly indicates that the gravitational redshift increases from pole to equatorial region ( maximum at equator ) . in case of
( + , + ) or ( -,- ) , for ( a , @xmath16 ) combination .
again it decreases from pole to equatorial region ( minimum at equator ) , in case of ( + , - ) or ( -,+ ) , for ( a , @xmath16 ) combination .
this shows clearly the dependence of redshift on the latitude ( @xmath207 ) .
the values of @xmath208 for these pulsars are also given in table - ii , for ( + , + ) or ( -,- ) combination .
we make fig.-6 , for sun showing @xmath209 versus @xmath16 , for the purpose of comparison and we notice that the effect is too small .
we can conclude from the present work that , 1 .
gravitational redshift is affected by rotation of the central body .
2 . gravitational redshift is a function of the @xmath17 and latitude ( @xmath207 ) .
3 . gravitational redshift will depend on the direction of spin ( @xmath198 ) of central body and also on the position ( @xmath199 ) on the central body at which the photon is emitted .
4 . when we consider the rotation velocity of central body ( @xmath32 ) to be zero , then we can obtain the corresponding gravitational redshift from a static body of same mass ( schwarzschild mass ) .
gravitational redshift shows a periodic nature with respect to @xmath17 .
the value of redshift ( z ) obtains a maximum ( primary maxima ) at , @xmath183 and a second maximum ( secondary maxima with value less than the primary maxima ) at @xmath184 .
gravitational redshift increases from pole to equatorial region ( maximum at equator ) , when rotation parameter ( a ) and the position ( @xmath16 ) on the central body at which the photon is emitted having same signs ( + , + ) or ( -,- ) , for ( a , @xmath16 ) combination .
gravitational redshift decreases from pole to equatorial region ( minimum at equator ) , when rotation parameter ( a ) and the position ( @xmath16 ) on the central body at which the photon is emitted having opposite signs ( + , - ) or ( -,+ ) , for ( a , @xmath16 ) combination .
the coefficient of latitude dependence of redshift ( @xmath15 ) , when the photon is emitted from any value of latitude on the central body and equator , clearly shows the dependence of redshift on the latitude ( @xmath207 ) .
looking at the observed redshift from a rotating object it should be possible to comment , from what latitude ( @xmath207 ) of the object , the lyman - @xmath164 line has been radiated .
we wish to thank dr .
atri deshmukhya , head department of physics , assam university , silchar , india for inspiring discussions .
a k dubey is also thankful to sanjib deb and arindwam chakraborty , department of physics , assam university , for providing help and support in programming and plots .
l. d. landau and e. m. lifshitz , _ the classical theory of fields _
( butterworth - heinemann , indian reprint 2008 ) vol .
2 , 4th revised edition , p.23 , 140 - 141 , 268 - 269 , 351 , 346 . + 2 . w. s. adams , observatory , * 48 * , 337 ( 1925 ) .
r. v. pound and g. v. jr .
rebka , phys .
lett . , * 3 * , 9 ( 1959 ) .
r. v. pound and j. l. snider , physical review , * 140 * , 3b ( 1965 ) .
j. l. snider , phys .
* 28 * , 13 ( 1972 ) .
t. p. krisher , d. morabito and j. d. anderson , phys .
, * 70 * , 2213 ( 1993 ) . + 7 .
p. d. nunez and m. nowakowski , j. astrophys .
astr . , * 31 * , 105 - 119 ( 2010 ) . + 8 .
s. carroll , _ space time and geometry _ ,
( pearson , addison wesley 2004 ) , p.217 , 261 - 263 . + 9 .
r. p. kerr , phys .
lett . * 11 * , 237 ( 1963 ) .
m. kasuya , phys .
d , * 25 * , 995 ( 1982 ) .
e. t. newman , e. couch , k. chinnapered , a. exton , a. prakash and r. torrence , j. math .
phys . , * 6 * , 918 ( 1965 ) .
p. collas and d. klein , general relativity and gravitation , * 36 * , 5 ( 2004 ) .
+ 13 . c. t. cunningham and j. m. bardeen , apj , * 183 * , 237 - 264 ( 1973 ) .
n. straumann , _ general relativity with applications to astrophysics _ , ( springer verlag 1984 ) , p.108 .
a. muller and m. wold , astronomy and astrophysics , * 457 * , 485 - 492 ( 2006 ) .
p. ji , s. qu and y. bai , general relativity and gravitation , * 40 * , 131 - 138 ( 2008 ) .
j. w. t. hessels , s. m. ransom , i. h stairs , p. c. c. freire , v.m .
kaspi , f. camilo , science , * 311 * , 1901 - 1904 , ( 2006 ) .
m. ashworth , a. g. lyne , smith , nature , * 301 * , 313 ( 1983 ) . + 19 .
b. a. jacoby , m. bailes , m. h. van kerkwijk , s. ord , a. hotan , s. r. kulkarni , s. b. anderson , apj , * 599 * , l99-l102 ( 2003 ) . + 20 .
a. g. lyne , m. burgay , m. kramer , science , * 303 * , 1153 ( 2004 ) .
i. h. stairs , z. arzoumanian , z. camilo , a. g. lyne , d. j. nice , j. h. taylor , s. e. thorsett , a. wolszczan , apj , * 505 * , 352 ( 1998 ) . | gravitational redshift is generally calculated without considering the rotation of a body . neglecting the rotation ,
the geometry of space time can be described by using the spherically symmetric schwarzschild geometry .
rotation has great effect on general relativity , which gives new challenges on gravitational redshift . when rotation is taken into consideration spherical symmetry is lost and off diagonal terms appear in the metric .
the geometry of space time can be then described by using the solutions of kerr family . in the present paper
we discuss the gravitational redshift for rotating body by using kerr metric .
the numerical calculations has been done under newtonian approximation of angular momentum .
it has been found that the value of gravitational redshift is influenced by the direction of spin of central body and also on the position ( latitude ) on the central body at which the photon is emitted .
the variation of gravitational redshift from equatorial to non - equatorial region has been calculated and its implications are discussed in detail . |
since the discovery of the cp - violation in decay of k -mesons ( christenson , cronin , fitch and turlay ( 1964 ) ) , a few attempts have been undertaken to observe experimentally this phenomenon in different processes .
however , that experiments have not been successful . at the present time ,
novel more precise experimental schemes are actively discussed : observation of the atom ( lamoreaux ( 1989 ) ) and neutron ( forte ( 1983 ) , fedorov , voronin and lavin ( 1992 ) ) electric dipole moment ; the t(time)-violating atom s ( molecule s ) spin rotation in a laser wave and the t - violating refraction of a photon in a polarized atomic or molecular gas ( baryshevsky ( 1993 , 1994 ) ) . in accordance with baryshevsky ( 1993 , 1994 )
the p(parity)- and t - violating dielectric permittivity tensor @xmath0 is given by @xmath1 where @xmath2 is the polarizability tensor of the matter , @xmath3 is the number of atoms ( molecules ) per @xmath4 , @xmath5 - the photon wave number .
the quantity @xmath6 is the tensor part of the zero - angle amplitude of elastic coherent scattering of a photon by an atom ( molecule ) @xmath7 . here
@xmath8 and @xmath9 are the polarization vectors of initial and scattered photons .
indices i=1,2,3 are refered to coordinates x , y , z , respectively , repeated indices imply summations . at zero external electric and magnetic fields
, the amplitude @xmath6 can be written as @xmath10 where @xmath11 is the p , t even ( invariant ) part of @xmath12 , @xmath13 is the p , t - violating part of @xmath12 , @xmath14 is the scalar ( vector , tensor ) p , t violating polarizability of an atom ( molecule ) , @xmath15 is the total antisymmetrical unit tensor of the rank three , @xmath16 , @xmath17 , @xmath18 is the spin density matrix of an atom ( molecule ) with the spin @xmath19 , @xmath20 .
the second rank tensor @xmath21 is given by @xmath22 ^{-1}\left\ { \hat j_i\hat j_m+\hat j_m\hat j_i-\frac 23j\left ( j+1\right ) \delta _ { im}\right\}\ ] ] in view of ( 1,2 ) , the t - violating processes affect upon the dielectric permittivity @xmath0 and , as a result , upon the refraction index only in media with polarized atoms ( molecules ) of the spin equaled or larger than 1 . this part of @xmath0 is proportional to @xmath23 .
if the atoms ( molecules ) spins are nonpolarized , only the p - violating term @xmath24 exists .
the term proportional to @xmath25 describes the p - violating rotation of a light polarization plane in metallic vapours ( barkov and zolotariov 1978 , bouchiat and pottier 1986 , khriplovich 1991 ) . as it has been shown in ( baryshevsky 1993 , 1994 ) , when an atom interacts with two coherent electromagnetic waves , the energy of this interaction depends on the t - violating scalar polarizability @xmath23 .
interaction of an atom ( molecule ) with two waves can be considered as a process of rescattering of one wave into another and vice versa .
then , as it follows from an expression for the effective interaction energy , the amplitude @xmath26 of the photon scattering by an unpolarized atom ( molecule ) at a non - zero angle is given by ( baryshevsky 1994 ) : @xmath27 + \beta _ s^t\left ( \vec n^{^{\prime } } -\vec n\right ) \left [ \vec e^{^{\prime ^{*}}}\vec e\right ] \right ) , \ ] ] where @xmath28 is the wave vector of a scattered photon , @xmath29 , @xmath30 is the scalar p , t - invariant polarizability of an atom ( molecule ) . expression
( 3 ) holds true in the absence of .external electric and magnetic fields .
it should be emphasized that expression ( 3 ) for the elastic scattering amplitude can be derived from the general principles of symmetry . indeed , there are four independent unit vectors : @xmath31 , @xmath32 , @xmath33 and @xmath9 , which completely describe geometry of the elastic scattering process .
the elastic scattering amplitude @xmath34 depends on these vectors and therewith is a scalar .
obviously , one can compose three independent scalars from these vectors : @xmath35 , @xmath36 $ ] , @xmath37 $ ] . as a result
, the scattering amplitude can be written as : @xmath38 + f_s^t\left ( \vec k^{^{\prime } } , \vec k\right ) \vec \nu _
2\left [ \vec
e^{^{\ \prime ^{*}}}\vec e\right ] , \ ] ] where @xmath39 is the p-,t- invariant scalar amplitude , @xmath40 is the p - violating scalar amplitude , and @xmath41 is the p-,t- violating scalar amplitude
. it can easily be found from ( 3,4 ) that the term proportional to @xmath42 vanishes in the case of forward scattering @xmath43 .
vice versa , in the case of back scattering @xmath44 the term proportional to @xmath45 gets equal to zero .
thus , one can conclude that the t - violating interactions manifest themselves in the processes of scattering by atoms ( molecules ) .
however , the scattering processes are usually incoherent and their cross sections are too small to hope for observation of the t - violating effect .
another situation takes place for diffraction gratings in the vicinity of the bragg resonance where the scattering process is coherent . as a result
, the intensities of scattered waves strongly increase : for instance , in the bragg ( reflection ) diffraction geometry the amplitude of the diffracted - reflected wave may reach the unity .
it gives us an opportunity to study the t - violating scattering processes ( baryshevsky 1994 ) . in the present paper
, equations describing the t - violating scattering by a diffraction grating have been obtained .
it has been shown that the photon s refraction index in a non - center - symmetrical grating depends on the t - violating amplitude @xmath41 .
it can result in a new phenomenon : the t - violating rotation of the photon polarization plane .
it has also been shown that the rotation angle gives rise sharply in the back - scattering diffraction geometry when the conditions of the photon resonance transmission are satisfied .
the phenomenon of p- , t- invariant diffraction of electromagnetic waves by diffraction gratings has been studied in detail for a very wide range of wavelengths ( see , for examples , shih - lin chang ( 1984 ) , tamir ( 1988 ) , maksimenko and slepyan ( 1997 ) ) . accoding to these articles equations of the dynamic diffraction
can be derived from the maxwell s equations if the permittivity tensor @xmath46 of a spatially periodic grating is known . to include the p , t violating processes into the diffraction theory ,
let us consider the microscopic maxwell equations : @xmath47 where @xmath48 is the electric field strength and @xmath49 is the magnetic field induction , @xmath3 and @xmath50 are the microscopic densities of the electrical charge and the current induced by an electromagnetic wave , @xmath51 is the speed of light .
the fourier transformation of these equations ( i.e. @xmath52 and so on ) leads us to the equation for @xmath53 as follows : @xmath54 , \ ] ] where @xmath55 . in linear approximation ,
the current @xmath56 is coupled with @xmath57 by the well - known dependence : @xmath58 with @xmath59 as the microscopic conductivity tensor being a sum of the conductivity tensors of the atoms ( molecules ) constituting the diffraction grating : @xmath60 here @xmath61 is the conductivity tensor of the a - type scatterers .
the summation is over all atoms ( molecules ) of the grating . in a diffraction grating ,
the tensor @xmath62 is a spatially periodic function .
it allows one to derive the expansion of @xmath63 from ( 7 ) as follows : @xmath64 where @xmath65 is the fourier transform of the conductivity tensor of a grating s elementary cell , @xmath66 is the reciprocal lattice vector of the diffraction grating . using current representation ( 9 ) , one can obtain a set of equations from ( 6 ) : @xmath67 tensor of the diffraction grating susceptibility is given by @xmath68 with @xmath69 here @xmath70 is the amplitude of coherent elastic scattering of an electromagnetic wave by a grating elementary cell from a state with the wave vector @xmath71 to a state with the wave vector @xmath28 .
the amplitude @xmath72 is obtained by summation of atomic ( molecular ) coherent elastic scatterig amplitudes over a grating s elementary cell : @xmath73 where @xmath74 is the coherent elastic scattering amplitude by an a - type atom ( molecule ) , @xmath75 is the gravity center coordinate of the a - type atom ( molecule ) , @xmath76 is the number of the atoms ( molecules ) in an elementary cell , angular brackets denote averaging over the coordinate distribution of scatterers in a grating s elementary cell .
the amplitude @xmath77 has been given by equation ( 4,3 ) . from ( 11 ) , ( 12 ) and ( 4 ) one can obtaine an expression for the susceptibility @xmath78 of the elementary cell of an optically isotropic material : @xmath79 where @xmath80 @xmath81 is the scalar p- , t- invariant susceptibility of an elementary cell , @xmath82 is the p - violating , t- invariant susceptibility of the elementary cell , and @xmath83 is the p- and t- violating susceptibility of the elementary cell , @xmath84 then , using ( 10 , 11 , 13 ) we can derive a set of equations describing the p and t violating interaction of an electromagnetic wave with a diffraction grating @xmath85 where @xmath86 assuming the interaction to be p , t invariant @xmath87 , equations ( 14 ) reduce to the conventional set of equations of dynamic diffraction theory ( shih - lin chang ( 1984 ) ) .
let us suppose , first of all , the photon @xmath88 frequency and the wave vector @xmath28 to be such that the bragg diffraction conditions @xmath89 and @xmath90 are not fulfilled exactly , and the inequality @xmath91 holds true . in this case
, the diffracted wave amplitude is much less comparing with the transmitted one : @xmath92 , and the perturbation theory can be applied for the further analysis . as a result in the first approximation of the perturbation theory one can derive from ( 10 ) that @xmath93 where @xmath94 , @xmath95 , @xmath96 .
substitution ( 15 ) into ( 10 ) results in the diffraction equations as follows @xmath97e_f(\vec k,\omega ) = 0\ ] ] which can be rewritten in more simple form @xmath98 by introducing the effective permittivity tensor @xmath99 one can see that even far away from the exact bragg conditions , where the diffracted wave amplitudes are small , a spatially periodic isotropic medium manifests the optical anisotropy being characterized by the effective permittivity tensor @xmath100 .
let a photon be incident on a grating normally to its reflection planes .
in other words , let the photon wave vector @xmath101 be antiparallel to the reciprocal lattice vector @xmath66 , i.e. @xmath102 . in this case
, the back - scattering diffraction regime can be realized for photons with wave numbers defined by the relation @xmath103 .
if , nevertheless , the inequality @xmath104 holds true , we can use set of equations ( 16 ) in which there is only one term satisfying the conditions @xmath105 and @xmath106 . let the coordinate axis @xmath107 be parallel to @xmath108 , @xmath109 . in this case
, the tensor @xmath110 has nonzero components at @xmath111 , @xmath112 = 1,2 only . as a result , set of equations ( 17 )
can be rewritten in the form as follows : @xmath113 with the permittivity tensor given by @xmath114 \varepsilon _ { ij3}\nu _ { 23}^{\vec \tau } \nonumber\end{aligned}\ ] ] in the above equations we introduced the designations : @xmath115 the term proportional @xmath116 @xmath117 describes the p - violating and t - invariant rotation of the light polarization plane about the direction @xmath118 .
this term does not depend on a structure of the diffraction grating and exists for any ordinary spatially isotropic media . unlike to that ,
the term proportional to @xmath119 is t - violating and depends on the grating s structure .
the term proportional @xmath119 looks like the term proportional to @xmath120 and is responsible for the polarization plane rotation about @xmath121 .
it is known that the phenomenon of the polarization plane rotation arises when right- and left - circularly polarized photons have different indices of refraction in a medium @xmath122 and @xmath123 , respectively .
it means that the tensor @xmath124 is diagonal for a given circular polarization and , concequently , the set of equations ( 19 ) is split into two independent equations .
really , let us write ( 19 ) in the vector notation : @xmath125 -k_0 ^ 2\alpha _
^{-1}\left [ \chi _ s\left ( \vec \tau \right ) \chi _ s^t\left ( -\vec \tau \right ) -\chi _ s\left ( -\vec \tau \right ) \chi _ s^t\left ( \vec \tau \right ) \right ] \left [ \vec e\vec \nu _ 2^\tau \right]\ ] ] and let @xmath126 be the unit polarization vector of a linearly polarized photon ; @xmath127 $ ] , @xmath128 .
then , the unit vectors corresponding to the circular polarizations are as follows @xmath129 . for the right @xmath130 , left @xmath131 circularly polarized photons , the field @xmath48 can be represented by @xmath132 . as a result
, it follows from ( 19 , 21 ) that : @xmath133 c_{\pm } = 0\ ] ] the corresponding refractive indices are obtained from ( 20 ) @xmath134\ ] ] the angle of the photon polarization plane rotation is defined by : @xmath135 where @xmath136 is the photon waylength in the medium , @xmath137 is the real part of @xmath138 .
then , the expression for @xmath139 can easily be derived from ( 24,23 ) @xmath140 l \nonumber\end{aligned}\ ] ] one can conclude , thus , that the t - violating interaction results in the phenomenon of the t - violating rotation of the photon polarization plane .
the effect manifests itself when the condition @xmath141 \neq 0 $ ] holds true .
it follows from ( 13 ) that the susceptibilities @xmath142 can be presented as @xmath143 where @xmath144 in view of ( 26,27 ) we can rewrite ( 25 ) as : @xmath145 l\ ] ] so , the t - violating rotation arises in the case of nonzero odd part of the suscetibility : @xmath146 .
such a situation is possible if an elementary cell of the diffraction grating does not posses the center of symmetry . in accordance with ( 28 ) , the angle of the t - violating rotation grows at @xmath147 .
.however , the condition @xmath148 violates at @xmath149 , where the amplitude of diffracted and transmitted waves are comparable : @xmath150 and , consequently , the perturbation theory gets unapplicable
. a rigorous dynamical diffraction theory must be applied in this case .
let the bragg condition is fulfilled for the only diffracted wave and is violated for all other possible ones .
it allows us to restrict ourselves to the two - wave approximation of the dynamical diffraction theory ( shih - lin chang ( 1984 ) ) . in that case ,
set of equations ( 14 ) reduces to two coupled equations , which for the back - scattering diffraction scheme @xmath151 take the form as follows : @xmath152 @xmath153 @xmath154 .
based on the above consideration , we can conclude that set of equations ( 29 ) can be diogonalized for the photon of a given circular polarization .
let the right - circularly polarized photon @xmath130 be incident on the diffraction grating .
the diffraction process , as it follows from ( 29 ) , results in the appearance of a back - scattered photon with the left circular polarization @xmath155 .
this is because the momentum of the back - scattered photon @xmath156 is antiparallel to the momentum @xmath28 of the incident one .
it is obvious that the left - circularly polarized photon will produce a right - circularly polarized back - scattered one .
thus , for circularly polarized photons set of vector equations ( 29 ) can be split into two independent sets of scalar equations : @xmath157 note that equations ( 30 ) are identical in form to conventional equations of the two - wave dynamical diffraction ( shih - lin chang ( 1984 ) ) : @xmath158 it allows us to write down a solution immediately , without deriving ( see , for example , ( shih - lin chang ( 1984 ) ) ) . as a result
, the amplitude of the transmitted electromagnetic wave at the output is given by @xmath159 @xmath160^{-1 } , \nonumber\end{aligned}\ ] ] where @xmath161 @xmath162 + \frac 14\{\left ( 2\chi _ s\left ( 0\right ) -\alpha \right ) ^2 - 4\left ( \chi _ { 1s}^2\left ( \vec \tau \right ) + \chi _ { 2s}^2\left ( \vec \tau \right )
\right ) \pm \nonumber\\ & & \pm 8\left [ \chi _ { 1s}\left ( \vec \tau \right ) \chi _ { 2s}^t\left ( \vec \tau \right ) -\chi _ { 2s}\left ( \vec \tau \right ) \chi _ { 1s}^t\left ( \vec \tau \right ) \right ] \}^{\frac 12 } , \end{aligned}\ ] ] @xmath163 -\frac 14\{\left ( 2\chi _ s\left ( 0\right ) -\alpha \right ) ^2 - 4\left ( \chi _ { 1s}^2\left ( \vec \tau \right ) + \chi _ { 2s}^2\left ( \vec \tau \right )
\right ) \pm \\ & & \pm 8\left [ \chi _ { 1s}\left ( \vec \tau \right ) \chi _ { 2s}^t\left ( \vec \tau \right ) -\chi _ { 2s}\left ( \vec \tau \right ) \chi _ { 1s}^t\left ( \vec \tau \right ) \right ] \}^{\frac 12 } , \end{aligned}\ ] ] @xmath164
l is the thickness of the diffraction grating @xmath165 \}^{\frac 12 } .
\nonumber\end{aligned}\ ] ] consider now the diffraction of a photon with linear polarization @xmath126 being the superposition of two opposite circular polarizations . in this case
@xmath166 and the amplitude of the transmitted wave @xmath167 can be presented by the superposition @xmath168 in the case under consideration @xmath169.as follows from ( 38 ) , this results in changing of the photon polarization at the output .
let us analyze expression ( 32 ) for the transmitted wave amplitude more attentively . according to ( 32 )
, the amplitude oscillates as a function of @xmath170 , i.e. , as a function of the wavelength , with maximums in points defined by the condition @xmath171 or @xmath172with @xmath173 as an integer .
note first of all that the condition @xmath174 dictates at @xmath175 the limit transition @xmath176 \right\ } ^{\frac 12}\rightarrow 0\ ] ] which determines the thresholds of the total bragg reflection band where there is a quickly damped inhomogeneous wave inside the diffraction grating . as a result
, the transmitted wave amplitude is small .
let now @xmath177 . as we know , the t - violating interactions are very small : @xmath178 .
it allows one to expand the square root into the teylor series truncated beyond the second term : @xmath179 where @xmath180 .
let @xmath181 , i.e. , the absorption is assumed to be sufficiently small to satisfy the condition @xmath182 which admits the consideration of the diffraction grating as an optically transparent medium with @xmath183 and @xmath184 as real functions .
let now the condition @xmath185 be fulfilled at @xmath177 .
this condition defines the resonance transmission in the grating and allows us to rewrite formula ( 40 ) in the form as follows @xmath186 l}{2\pi m}\ ] ] by substituting ( 41 ) into ( 33 , 32 ) one can express @xmath187 by @xmath188 where @xmath189 @xmath190 in view of that the second term in ( 42 ) is small and expression ( 42 ) can be written as @xmath191 where the phase terms are given by @xmath192 l\ ] ] using this equation one can find the angle of the polarization plane rotation : @xmath193 where the first term in the right - hand part defines the p - violating t - invariant rotation angle : @xmath194 and the second one corresponds to the t - violating rotation : @xmath195 , \nonumber\end{aligned}\ ] ] the @xmath196 is for @xmath197 , the @xmath198 is for @xmath199 .
the imaginary part of the t - violating polarizability @xmath200 is responsible for the t - violating circular dichroism . due to that process
, a linearly polarized photon gets a circular polarization at the diffraction grating s output .
the degree of the circular polarization of the photon is determined from the relation : @xmath201 \nonumber\end{aligned}\ ] ] it should be pointed out that the resonance transmission condition is satisfied at a given m for two different values of @xmath202 .
this is because there is a possibility to approach to the brilluan ( the total bragg reflection ) bandgap both from high and low frequencies .
the t - violating parts of the rotation angle are opposite in sign for @xmath197 and for @xmath203 .
it gives the addition opportunity to distinguish the t - violating rotation from the p - violating t - invariant rotation .
indeed , the p - violating rotation does not depend on the back bragg diffraction in the general case because the p - violating scattering amplitude equals zero for back scattering ( see ( 32 - 35 ) . in accordance with ( 47,48 ) the t - violating rotation and dichroism grow sharply in the vicinity of the resonance bragg transmission . at the first glance , one could expect for @xmath204 the dependence @xmath205 ( see ( 25 ) ) .
however , in the vicinity of resonance , the rotation angle @xmath204 turns out to be multiplied by the factor @xmath206 which provides the above mentioned growth ( for example , @xmath207 at @xmath208 , @xmath209 , @xmath210 , @xmath211 ) .
there are different types of diffraction gratings destined for use in optical and more longwave ranges .
however , it should be noted that the successful observation of the p - violating rotation has been performed by means of studying of light transmission through gas targets .
there are a lot of theoretical calculations for atoms of such gases : see , for example , ( khriplovich ( 1991 ) ) for @xmath212 , @xmath213 , @xmath214 , @xmath215 . from that point of view
, it would be preferable to use gases for studying of the t - violating phenomena of polarization plane rotation and dichroism applying the experience accumulated earlier . at the first glance
, there is a serious problem how to create such a diffraction grating in a gas .
nevertheless , the problem can be solved if we make use some well - known results of the electromagnetic theory of waveguides ( jackson ( 1962 ) , tamir ( 1988 ) , maksimenko and slepyan ( 1997 ) ) . according to this theory
, there is a correspondence between wave processes in waveguides with periodically modulated boundaries and homogeneous filling and such processes in regular waveguides filled by a periodic and , generally , anisotropic medium .
let us consider a regular waveguide constituted by two plane - parallel surfaces ; for example , two metallic mirrors ( see figure 1 ) .
let us then place a plane diffraction grating ( figure 2 ) on the surface of the mirror ( figure 3 ) and fill the waveguide by the studied gas . because of the above stated correspondence , such a system is equivalent to a regular plane waveguide filled by a gas with a spatially periodic permittivity tensor .
the permittivity modulation period is therewith equal to the grating period .
as the chosen plane grating has an asymmetric profile ( figure 2 ) , the corresponding virtual volume grating of the permittivity turns out to be noncentrosymmetrical and , thus , satisfies to the above imposed requirements for displaying of the t - violating phenomena .
let @xmath216 be the permittivity of the waveguide being considered with as a periodic function with respect to @xmath107 .
as it has been stated above , such a waveguide can be modeled by the waveguide with the effective permittivity @xmath217 which is a periodic function of @xmath107 and is independent on @xmath218 .
we can show it mathematically starting with the maxwell equations @xmath219 let an electromagnetic wave propagate in the plane regular waveguide ( fig.1(a ) ) .
in this case , the motion along the @xmath107-axis is free and , thus , it can be described by a plane wave @xmath220 .
it results in a set of equations for determining of the stationary states of the waveguide : @xmath221 where @xmath222 , @xmath202 notes the polarization state of the wave .
then , the field @xmath223 in the waveguide with the diffraction grating can be presented by the expansion as follows @xmath224 substitution of this expansion into ( 50 ) yields @xmath225 where @xmath226 , @xmath227 , @xmath228- is the period of the plane grating .
assuming the inequality @xmath229 , @xmath230 to be valid , we can restrict ourselves to a single mode approximation , which allows us to rewrite ( 51 ) as @xmath231 where @xmath232 .
equations ( 52 ) are identical with those ( see ( 10 ) ) describing the electric field in a volume diffraction grating .
so , the goal formulated above has been achieved : we have obtained a volume diffraction grating .
now , let us estimate the effect of the t - violation . to do that we must determine , in accordance with formula
( 47 ) for @xmath204 , the t - violating polarizability @xmath233 , which , in view of ( 13,3,4 ) , is proportional to the t - violating atomic polarizability @xmath234 .
the estimate carried out by ( baryshevsky ( 1993,1994 ) , khriplovich ( 1991 ) ) gives @xmath235 , where @xmath25 is the p - violating t - invariant scalar polarizability .
the polarizability @xmath236 was studied both theoretically and experimentally ( khriplovich ( 1991 ) ) .
particularly , the theory gives @xmath237 for atoms analogous to bi , tl , pb .
it yields the estimate @xmath238 for the t - violating atomic polarizability .
the polarizabelity @xmath239 causes the p - violating rotation of the polarization plane by the angle @xmath240 rad/@xmath241l for the gas density @xmath242 . as a result ,
in our case the parameter @xmath243 turns out to be @xmath244 rad/@xmath241l and can be even less by the factor h / d , where h is the corrugation amplitude of the diffraction grating while d is the distance between waveguide s mirrors . assuming this factor to be @xmath245
, we shall find @xmath246 rad/@xmath241l .
thus , the final estimate of the t - violating rotation angle @xmath204 is @xmath247 in real situation the polarizabelity of a grating @xmath248 may exceed the unity .
however , our analysis has been performed under the assumption @xmath249 . if , for example , we take @xmath250 , @xmath251 then @xmath252 rad/@xmath253l@xmath254 and , consequently , for l= 1 cm we will have the amplification by a factor of @xmath255 .
as it is seen , we have obtained the t - violating rotation angle @xmath204 of the same order of @xmath256 .
it makes possible experimental observation of the phenomenon of the t - violating polarization plane rotation .
it should be noted that the manufacturing of diffraction gratings for the range being more longwave than the visible light one may be simpler .
that is why we would like to attract attention to the possibility of studying of the t - violating polarization plane rotation in the vicinity of frequencies of atomic ( molecular ) hiperfine transitions ; for example , for ce ( the transition wavelength is @xmath257 cm ) and tl ( @xmath258 cm ) .
thus , we have shown that the phenomenon of the t - violating polarization plane rotation appears while the photon is scattered by a volume diffraction grating .
the phenomenon grows sharply in the vicinity of the resonance transmission condition . an experimental scheme based on a plane waveguide with a diffraction grating as
a mirror and gas filling has been proposed which enables real experiments on observation of the t - violating polarization plane rotation to be performed .
the rotation angle has been shown to be @xmath259 rad/@xmath241l@xmath254 , where l is the waveguide length ( thickness of the equivalent volume diffracting grating ) .
baryshevsky v.g .
1993 phys . lett a @xmath260 , 38 | the equations describing the t - violating photon scattering by a diffraction grating have been obtained .
it is shown , that the t - violating rotation of the photon polarization plane appear under diffraction in a noncenter symmetrical diffraction grating .
the rotation angle gives rise sharply , when the conditions of the photon resonance transmission are satisfied .
24.0 cm -1.5 cm 16.0 cm -0.35 cm 0.35 cm * time - violating rotation of the electromagnetic wave * * polarization plane by a diffraction grating * v.g.baryshevsky nuclear problems institute , bobruiskaya str 11 , minsk 220080 belarus . electronic address : [email protected] |
according to the prebiotic soup hypothesis @xcite , the early earth had the supply of organic compounds needed to jump - start polymerization reactions that , through gradual complexification , led to the first biochemical networks displaying some of the characteristics attributed to life , such as metabolic activity and replication .
although the road is still obscure @xcite , the situation is not all bleak . in 1953
, stanley miller simulated tentative conditions of early - earth in the laboratory to obtain amino acids from simple chemical compounds .
that same year , frank @xcite proposed that auto - catalytic polymerization could explain the emergence of biomolecular homochirality , a clear signature of terrestrial and , possibly , all life @xcite : terrestrial amino acids belonging to proteins are overwhelmingly left - handed , while sugars are right - handed .
if a bottom - up approach to the early development of life is adopted , the homochirality of life s biochemistry must have emerged dynamically , as reactions among the simplest molecular building blocks occurred with high enough yield .
alternatively , one may assume that , somehow , only monomers of a single chirality were present in the prebiotic soup : they were made that way or brought here during the intense bombardment of earth s infancy , that lasted to about 3.8 gyr ago @xcite .
we would , however , still need to understand how homochirality developed elsewhere in the cosmos and not here , and whether it developed with the same chiral bias in more than one place . here
, we consider the homochirality of life as an emergent process that took place on early - earth and , possibly , other planetary platforms . as a starting point ,
we use the reaction - network model proposed by sandars @xcite , which includes enantiometric cross - inhibition .
as shown in the interesting work of brandenburg and multamki @xcite ( bm ) , sandar s polymerization reaction network can be reduced to an effective spatio - temporal mean - field model , where the order parameter is the chiral asymmetry between left and right - handed polymers . to this , we add the effects of an external environment , showing that they can be crucial in the final determination of the net value of enantiometric excess , if any .
sandars proposed the following polymerization reactions @xcite : @xmath0 ; @xmath1 ; @xmath2 ; and , @xmath3 , where @xmath4 ( @xmath5 ) are the reaction rates for adding monomers of the same ( opposite ) chirality to a given chain . the network is completed by adding the four opposite reactions , that is , by interchanging @xmath6 , and by adding a substrate @xmath7 from which both left and right - handed monomers emerge , @xmath8 and @xmath9 , where @xmath10 determine the enzymatic enhancement of left and right - handed monomers , which is not known .
we follow sandars and choose @xmath11 $ ] , @xmath12 $ ] .
as remarked in bm , it is possible to truncate the system to @xmath13 and still maintain the essential aspects of the dynamics leading to homochiralization .
this allows us to model the reaction network as a mean - field theory exhibiting spontaneous chiral symmetry breaking @xcite .
our approach blends the work of bm with the pioneering work of kondepudi and nelson , where the reaction network was coupled to time - dependent external effects @xcite : chirality evolves spatio - temporally in contact with an environment .
the equations can be simplified by assuming that the rate of change of @xmath14 $ ] and @xmath15 $ ] is much slower than that of @xmath16 $ ] and @xmath17 $ ] .
the same for the substrate @xmath18 $ ] , so that @xmath19/dt = q-(q_l+q_r ) \simeq 0 $ ] , where @xmath20 and @xmath21 are the source terms for monomers generated from the substrate @xmath18 $ ] : @xmath22(pc_l + qc_r)$ ] , and @xmath23(pc_r + qc_l)~$ ] @xcite .
the constants @xmath24 and @xmath25 are given in terms of the fidelity of enzymatic reactions @xmath26 , an adjustable parameter .
as demonstrated by kondepudi and nelson @xcite and many others @xcite , auto - catalytic networks show a tendency to bifurcate toward homochirality for values above critical , @xmath27 .
the specific value of @xmath27 is model - dependent . under the above assumptions , and introducing the dimensionless symmetric and asymmetric variables , @xmath28 and @xmath29 , where @xmath30(2k_s / q)^{1/2}$ ] and @xmath31(2k_s / q)^{1/2}$ ]
, bm have shown that the polymerization equations reduce to @xmath32 where @xmath33 , with dimension of inverse time .
@xmath34 is a fixed point : the system will tend towards this value at time - scales of order @xmath35 . with @xmath34 ,
the equation for the chiral asymmetry has fixed points at @xmath36 , as pointed out in bm .
an enantiometric excess is only possible if @xmath37 .
we model the external environment via a stochastic spatio - temporal langevin equation , rewriting the equations above as @xmath38 where @xmath39 is the diffusion constant and @xmath40 represent white noise with zero mean and a two - point correlation function given by @xmath41 , and @xmath42 measures the strength of the external influence .
for example , in mean - field models of phase transitions , it is common to write @xmath43 , where @xmath44 is boltzmann s constant , @xmath45 is the temperature , and @xmath46 is a viscosity coefficient .
the equations can be made dimensionless by introducing @xmath47 and @xmath48 , which determine the typical spatio - temporal scales in the system .
the noise amplitude scales as @xmath49 , where @xmath50 is the number of spatial dimensions . using as nominal values for the parameters , @xmath51@xmath52s@xmath53 , @xmath54@xmath55s@xmath53 , the diffusivity of water @xmath56m@xmath57s@xmath53 , we obtain , @xmath58s)@xmath59 and @xmath60cm)@xmath61 . as in bm
[ see also @xcite ] , the concentrations are spatially - dependent quantities .
this implicitly assumes that it is possible to define an effective correlation volume within which the value of the chiral asymmetry @xmath62 is fairly homogeneous .
( we take @xmath34 . )
using well - known results from the mean - field theory of phase transitions @xcite , we can easily compute the correlation length . from the equation for @xmath62
, we obtain an effective potential @xmath63 , @xmath64~.\ ] ] for @xmath65 , @xmath63 has a typical double - well shape , with minima at the fixed points @xmath66 .
the correlation length , @xmath40 , is given by @xmath67 , where @xmath68 denotes a minimum of the potential .
for the fixed points , we get , @xmath69 . at @xmath70
the correlation length diverges , as it should for a critical point
. however , the noise parameter @xmath71 also controls the behavior of the system .
indeed , even if @xmath72 , an enantiometric excess may not develop if @xmath71 is above a critical value @xmath73 . in analogy with ferromagnets , where above a critical temperature the net magnetization is zero
, one may say that above @xmath73 the stochastic forcing due to the external environment overwhelms any local excess of @xmath74 over @xmath75 enantiometers within a domain of correlation volume @xmath76 : racemization is achieved at large scales and chiral symmetry is restored throughout space .
salam @xcite suggested that there should be a critical temperature @xmath77 above which any net homochirality is destroyed .
however , he conceded that calculating @xmath77 would be quite challenging using the electroweak theory of particle physics . here , we chose a different route which , we believe , will allow us to explore the qualitative aspects of the problem more effectively : the noise amplitude @xmath71 may represent a sudden increase in temperature and/or pressure due to a meteoritic impact or volcanic eruption , or , possibly , due to a source of circularly - polarized ultraviolet light @xcite .
the equation dictating the evolution of the enantiometric excess @xmath62 , eq .
[ aeq ] , was solved with a finite - difference method in a @xmath78 grid and a @xmath79 grid with @xmath80 and @xmath81 , and periodic boundary conditions . in 2d , this corresponds to simulating a shallow pool with linear dimensions of @xmath82 cm .
we prepared the system initially in a chirally pure phase , which we chose to be @xmath83 .
the equation is then solved for different values of the external noise , @xmath71 . as can be seen in figure [ tcritfig ] , for @xmath84 , @xmath85 ,
that is , the system becomes racemized .
@xmath86 approaches a constant for large times , indicating that the reaction network reaches equilibrium with the environment .
the results are ensemble averaged . for @xmath87 , @xmath88 .
we can describe the environmental impact on homochirality at the microscopic level by introducing the `` ginzburg criterion '' familiar of phase transitions @xcite .
consider a correlation volume with @xmath89 ( or @xmath90 ) .
what is the energy barrier ( @xmath91 ) to flip half the molecules in the volume so that @xmath85 ? if @xmath92 is the number of molecules in a correlation volume , @xmath93 , where @xmath94 is the energy to flip one molecule .
the ginzburg criterion says that this energy is @xmath95 , where @xmath96 .
comparing the two expressions we obtain , @xmath97 . from equation
[ v ] , @xmath98 .
now , @xmath99 .
( we set @xmath72 . )
using for the microscopic spatial scale @xmath100 , and that @xmath101 , we obtain @xmath102m@xmath103 . [ the energy has dimensions of ( length)@xmath104 . ] to complete the argument , we use that the critical `` environmental '' energy to restore the chiral symmetry was obtained numerically to be ( cf .
figure [ tcritfig ] ) , @xmath105m@xmath103 .
we thus obtain the ratio , @xmath106 .
it is tempting to compare this result with possible sources of homochirality .
for example , weak neutral currents are expected to induce an excess at room temperature of @xmath107 @xcite . in the language of the present work
, they would induce a tilt in the potential @xmath63 proportional to @xmath108 .
thus , within the violent environment of prebiotic earth , effects from such sources , even if cumulative , would be negligible : any accumulated excess could be easily wiped out by an external disturbance .
further work along these lines is in progress . | the development of prebiotic homochirality on early - earth or another planetary platform may be viewed as a critical phenomenon .
it is shown , in the context of spatio - temporal polymerization reaction networks , that environmental effects be them temperature surges or other external disruptions may destroy any net chirality previously produced . in order to understand the emergence of prebiotic homochirality
it is important to model the coupling of polymerization reaction networks to different planetary environments . |
graphical models of codes and the decoding algorithms associated with them are now a major focus area of research in coding theory . turbo codes , low - density parity - check ( ldpc ) codes , and expander codes are all examples of codes defined , in one way or another , on underlying graphs . a unified treatment of graphical models and the associated decoding algorithms began with the work of wiberg , loeliger and koetter @xcite,@xcite , and has since been abstracted and refined under the framework of the generalized distributive law @xcite , factor graphs @xcite , and normal realizations @xcite,@xcite .
the particular case of graphical models in which the underlying graphs are cycle - free has a long and rich history of its own , starting with the study of trellis representations of codes ; see _ e.g. _ , @xcite and the references therein .
briefly , a graphical model consists of a graph , an assignment of symbol variables to the vertices of the graph , an assignment of state variables to the edges of the graph , and a specification of local constraint codes at each vertex of the graph .
the full behavior of the model is the set of all configurations of symbol and state variables that satisfy all the local constraints .
such a model is called a realization of a code @xmath0 if the restriction of the full behavior to the set of symbol variables is precisely @xmath0 .
the realization is said to be cycle - free if the underlying graph in the model has no cycles .
a trellis representation of a code can be viewed as a cycle - free realization in which the underlying graph is a simple path .
a linear code @xmath0 has a realization on a graph @xmath1 that is not connected if and only if @xmath0 can be expressed as the direct sum of the codes that are individually realized on the connected components of @xmath1 @xcite .
thus , there is no loss of generality in just focusing , as we do , on the case of realizations on connected graphs . in this paper , we will be concerned with tree realizations cycle - free realizations in which the underlying cycle - free graph is connected , _
i.e. _ , is a tree .
it is by now well known that the sum - product algorithm on any tree realization provides an exact implementation of maximum - likelihood ( ml ) decoding @xcite,@xcite,@xcite,@xcite . a good initial estimate of the computational complexity of such an implementation is given by the constraint complexity of the realization , which is the maximum dimension of any of the local constraint codes in the realization .
now , distinct tree realizations of the same code have , in general , distinct constraint complexities .
the treewidth of a code is defined to be the least constraint complexity of any of its tree realizations .
thus , treewidth may be taken to be a measure of the ml decoding complexity of a code .
since trellis realizations are instances of tree realizations , the treewidth of a code can be no larger than the minimum constraint complexity of any of its trellis realizations . in the abstract of his paper @xcite
, forney claimed that `` the constraint complexity of a general cycle - free graph realization can be [ strictly ] less than that of any conventional trellis realization , but not by very much . '' while he substantiated the first part of his claim by means of an example , he left the `` not by very much '' part as a conjecture ( * ? ? ?
* , conjecture 2 ) .
but he also admitted that none of the arguments he gave in support of his conjecture `` is very persuasive , '' and that it is equally plausible that ( * ? ? ?
* conjecture 3 ) there exists no upper bound on the gap between the treewidth of a code and the minimum constraint complexity of any of its trellis realizations .
one of the main contributions of this paper is an example that affirms the validity of forney s conjecture 3 .
we present , in section [ complexity_section ] , a family of codes for which the difference between the minimum trellis constraint complexity and the treewidth grows logarithmically with codelength .
we conjecture that this is in fact the maximal rate of growth of this difference .
our construction of this example is based upon results from the graph theory and matroid theory literatures that connect the notions of treewidth and trellis complexity of a code to certain complexity measures defined for graphs .
this paper makes two other contributions , both relating to minimal tree realizations .
a mapping of the set of coordinates of a code @xmath0 to the vertices of a tree is called a tree decomposition .
a tree decomposition may be viewed as an assignment of symbol variables to the vertices of the tree .
it is known that given a code @xmath0 , among all tree realizations of @xmath0 that extend a given tree decomposition , there is one that minimizes the state space dimension at each vertex of the underlying tree @xcite .
this minimal tree realization , an explicit construction of which was also given in @xcite , is unique up to isomorphism .
we give two new constructions of minimal tree realizations .
the first construction involves a generalization of the idea of state merging that can be used to construct minimal trellis realizations ( * ? ? ?
* section 4 ) .
we show that any tree realization of a code can be converted to a minimal realization by a sequence of state merging transformations .
the state space and constraint code dimensions do not increase at any step of this process . from this , we obtain the fact that a minimal realization also minimizes the constraint code dimension at each vertex of the underlying tree .
our second construction of minimal tree realizations uses extensions of the code decomposition techniques that were presented in @xcite .
the main advantage of this construction is its recursive nature , which makes it suitable for mechanical implementation .
also , it is relatively straightforward to estimate the computational complexity of this construction .
we show that the complexity is polynomial in the length and dimension of the code , as well as in the size of the underlying tree , but is exponential in the state - complexity of the minimal realization , which is the maximum dimension of any state space in the realization .
the paper is organized as follows . in section [ background_section ]
, we provide the necessary background on tree realizations of linear codes .
the construction of minimal realizations by means of state merging is presented in section [ state_merging_section ] .
code decomposition techniques are developed in section [ decomp_section ] , and used in section [ new_construct_section ] to derive a recursive construction of minimal tree realizations .
proofs of some of the results from sections [ background_section][new_construct_section ] are deferred to appendices to preserve the flow of the exposition .
treewidth and related complexity measures are defined in section [ complexity_section ] , which also establishes connections between these code complexity measures and certain complexity measures defined for graphs .
these connections are used to derive the example of a code family for which the gap between minimum trellis constraint complexity and treewidth is arbitrarily large .
we also touch upon the subject of codes of bounded complexity , observing that many hard coding - theoretic problems become polynomial - time solvable when restricted to code families whose treewidth is bounded .
section [ conclusion ] contains a few concluding remarks .
our treatment of the topic of tree realizations in this section is based on the exposition of forney @xcite,@xcite ; see also @xcite .
we start by establishing some basic notation .
we take @xmath2 to be an arbitrary finite field .
given a finite index set @xmath3 , we have the vector space @xmath4 . for @xmath5 and @xmath6
, the notation @xmath7 will denote the _ projection _ @xmath8 .
also , for @xmath6 , we will find it convenient to reserve the use of @xmath9 to denote the set difference @xmath10 .
a _ linear code _ over @xmath2 , defined on the index set @xmath3 , is a subspace @xmath11
. we will only consider linear codes in this paper , so the terms `` code '' and `` linear code '' will be used interchangeably .
the dimension , over @xmath2 , of @xmath0 will be denoted by @xmath12 .
an @xmath13 $ ] code is a code of length @xmath14 and dimension @xmath15 .
if , additionally , the code has minimum distance @xmath16 , then the code is an @xmath17 $ ] code .
the dual code of @xmath0 is denoted by @xmath18 , and is defined on the same index set as @xmath0 .
let @xmath19 be a subset of the index set @xmath3 .
the _ projection _ of @xmath0 onto @xmath19 is the code @xmath20 , which is a subspace of @xmath21 .
we will use @xmath22 to denote the _ cross - section _ of @xmath0 consisting of all projections @xmath23 of codewords @xmath24 that satisfy @xmath25 . to be precise , @xmath26 . note that @xmath27 .
also , since @xmath22 is isomorphic to the kernel of the projection map @xmath28 defined by @xmath29 , we have that @xmath30 .
furthermore , projections and cross - sections are dual notions , in the sense that @xmath31 , and similarly , @xmath32 .
if @xmath33 and @xmath34 are codes over @xmath2 defined on mutually disjoint index sets @xmath35 and @xmath36 , respectively , then their _ direct sum _ is the code @xmath37 defined on the index set @xmath38 , such that @xmath39 and @xmath40 .
this definition naturally extends to multiple codes ( or subspaces ) @xmath41 , where @xmath42 is a code identifier that takes values in some set @xmath43 .
again , it must be assumed that the codes @xmath41 are defined on mutually disjoint index sets @xmath44 .
the direct sum in this situation is denoted by @xmath45 .
a tree is a connected graph without cycles . given a tree @xmath46
, we will denote its vertex and edge sets by @xmath47 and @xmath48 , respectively , or simply by @xmath49 and @xmath50 if there is no ambiguity .
vertices of degree one are called _ leaves _ , and all other vertices are called _ internal nodes_. given a @xmath51 ,
the set of edges incident with @xmath52 will be denoted by @xmath53 .
removal of an arbitrary edge @xmath54 from @xmath46 produces a disconnected graph @xmath55 , which is the disjoint union of two subtrees , which we will denote by @xmath56 and @xmath57 , of @xmath46 .
note that @xmath58 and @xmath59 form a partition of @xmath47 .
let @xmath0 be a code over @xmath2 , defined on the index set @xmath3 .
to each @xmath60 , we associate a _ symbol variable _ @xmath61 , which is allowed to take values in @xmath2 .
a _ tree decomposition _ of @xmath3 is a pair @xmath62 , where @xmath46 is a tree ( _ i.e. _ , a connected , cycle - free graph ) and @xmath63 is a mapping from @xmath3 to the vertex set of @xmath46 .
pictorially , a tree decomposition @xmath62 is depicted as a tree with an additional feature : at each vertex @xmath52 such that @xmath64 is non - empty , we attach special `` half - edges '' , one for each index in @xmath64 ; see figure [ tree_decomp ] . at this point
, we introduce some notation that we will consistently use in the rest of the paper . given a tree decomposition @xmath62 of an index set @xmath3 , and an edge @xmath65
, we define @xmath66 and @xmath67
. thus , @xmath68 and @xmath69 are the subsets of @xmath3 that get mapped by @xmath70 to vertices in @xmath56 and @xmath57 , respectively .
clearly , @xmath68 and @xmath69 form a partition of @xmath3 .
recall that @xmath53 , @xmath51 , denotes the set of edges incident with @xmath52 in @xmath46 .
consider a tuple of the form @xmath71 , where * @xmath62 is a tree decomposition of @xmath3 ; * for each @xmath65 , @xmath72 is a vector space over @xmath2 called a _ state space _ ; * for each @xmath51 , @xmath73 is a subspace of @xmath74 , called a _ local constraint code _ , or simply , a _
local constraint_. such a tuple will be called a _
tree model_. the elements of any state space @xmath72 are called _
states_. the index sets of the state spaces @xmath72 , @xmath65 , are taken to be mutually disjoint , and are also taken to be disjoint from the index set @xmath3 corresponding to the symbol variables . finally , to each @xmath65 , we associate a _ state variable _ @xmath75 that takes values in the corresponding state space @xmath72 .
a _ global configuration _ of a tree model as above is an assignment of values to each of the symbol and state variables . in other words
, it is a vector of the form @xmath76 .
a global configuration is said to be _
valid _ if it satisfies all the local constraints .
thus , @xmath76 is a valid global configuration if for each @xmath51 , @xmath77 .
the set of all valid global configurations of a tree model is called the _ full behavior _ of the model . note
that the full behavior is a subspace @xmath78 . as usual , @xmath79 denotes the projection of @xmath80 onto the index set @xmath3 .
if @xmath81 , then the model @xmath71 is called a _ ( linear ) tree realization _ of @xmath0 . a tree realization @xmath71 of @xmath0 is said to _ extend _ ( or be an extension of ) the tree decomposition @xmath62 of the index set of @xmath0 .
any tree decomposition of the index set of a code can always be extended to a tree realization of the code , as explained in the following example .
let @xmath0 be a code defined on index set @xmath3 , and let @xmath62 be a tree decomposition of @xmath3 .
pick an arbitrary @xmath51 , and define @xmath82 .
now , consider the set , @xmath53 , of edges incident with @xmath52 .
removal of any @xmath83 produces the two subtrees @xmath56 and @xmath57 .
we specify @xmath56 to be the subtree that does _ not _ contain the vertex @xmath52 , and as usual , @xmath66 . for each @xmath83 ,
the state space @xmath72 is taken to be a copy of @xmath84 .
the remaining state spaces and local constraints are chosen so that , for each @xmath83 , the symbol variables indexed by @xmath68 simply get relayed ( unchanged ) to the state variable @xmath75 ; see figure [ trivial_ext ] .
it should be clear that the resulting tree model is a tree realization of the code @xmath0 .
this will be called a _
trivial extension _ of @xmath62 .
we will present constructions of non - trivial extensions of tree decompositions a little later .
[ trivial_example ] a _ simple path _ is a tree with exactly two leaves ( the end - points of the path ) , in which all internal nodes have degree two ; see figure [ path_fig ] .
let @xmath0 be a code defined on index set @xmath3 , and let @xmath62 be a tree decomposition of @xmath3 , in which @xmath46 is a simple path , and @xmath70 is a surjective map @xmath85 .
any tree realization of @xmath0 that extends @xmath62 is called a _ trellis realization _ of @xmath0 .
when @xmath70 is a bijection , then any trellis realization extending @xmath62 is called a _ conventional _ trellis realization . when @xmath70 is not a bijection ( but still a surjection ) , a trellis realization that extends @xmath62
is called a _
sectionalized _ trellis realization . in trellis terminology ,
the local constraint codes in a trellis realization are called _
branch spaces_. the theory of trellis realizations is well established ; we refer the reader to @xcite for an excellent survey of this theory .
[ trellis_example ] let @xmath80 be the full behavior of a tree model @xmath71
. we will find it useful to define certain projections of @xmath80 , other than @xmath86 for @xmath6 .
let @xmath87 be a global configuration in @xmath80 . at any given @xmath51 ,
the _ local configuration _ of @xmath88 at @xmath52 is defined as @xmath89 the set of all local configurations of @xmath80 at @xmath52 is then defined as @xmath90 . by definition , @xmath91 .
similarly , for @xmath92 , and @xmath88 as above , we define the projections @xmath93 and @xmath94 .
clearly , @xmath95 is a subspace of @xmath96 .
if @xmath97 consists of a single edge @xmath54 , then we simply denote the corresponding projections by @xmath98 and @xmath99 .
the following elementary property of the projections @xmath100 will be useful later ; a proof for it is given in appendix [ sec2_lemmas_app ] .
let @xmath80 be the full behavior of some tree realization of a code @xmath0 , defined on the index set @xmath3 , that extends the tree decomposition @xmath62 .
suppose that @xmath101 and @xmath65 are such that @xmath102 .
then , @xmath103 .
[ b|e_lemma ] a tree model ( or realization ) @xmath104 , with full behavior @xmath80 , is said to be _ essential _ if @xmath105 for all @xmath65 .
this definition actually implies something more . if the tree model @xmath106 , with full behavior @xmath80 , is essential , then @xmath107 for all @xmath51 .
[ ess_lemma ] a proof of the lemma can be found in appendix [ sec2_lemmas_app ] .
+ an arbitrary tree model can always be `` essentialized '' . to see this ,
let @xmath108 be a tree model with full behavior @xmath80 .
recall that @xmath109 is a subspace of @xmath72 , and @xmath110 is a subspace of @xmath73 .
define the _ essentialization _ of @xmath111 to be the tree model @xmath112 .
it is readily verified that @xmath113 has the same full behavior as @xmath111 .
+ given a code @xmath0 and a tree decomposition @xmath62 of its index set @xmath3 , there exists an essential tree realization , @xmath114 , of @xmath0 with the following property @xcite,@xcite : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ if @xmath71 is a tree realization of @xmath0 that extends @xmath62 , then for all @xmath65 , @xmath115 . _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ this _ minimal _ tree realization , which we henceforth denote by @xmath116 , is unique up to isomorphism . more precisely , if @xmath117 is also a tree realization of @xmath0 with the above property ( except that @xmath118 is replaced by @xmath119 ) , then @xmath120 for each @xmath65 , and @xmath121 for each @xmath51
. we will not distinguish between isomorphic tree realizations .
we outline a construction , due to forney @xcite , of @xmath116 . for any edge @xmath65 , the sets @xmath68 and @xmath69 form a partition of the index set @xmath3 .
e^ * = / ( _ j(e )
_ |j(e ) ) , [ se*_def ] and let _
e^ * : / ( _ j(e ) _
|j(e ) ) [ se*_def ] be the canonical projection map .
in other words , for @xmath24 , @xmath122 is the coset @xmath123 .
now , let @xmath80 be the vector space consisting of all global configurations @xmath124 corresponding to codewords @xmath24 , where @xmath125 .
it is worth noting that @xmath81 , and furthermore , @xmath126 , since @xmath127 implies that @xmath128 .
we can now define for each @xmath51 , the local constraint c_v^ * = |_v = \{(|_^-1(v ) , ( _ e^ * ( ) , ee(v ) ) ) : c}. [ cv*_def ] the minimal realization @xmath116 is the tuple @xmath114 .
it may be verified that @xmath80 is the full behavior of @xmath129 , so that @xmath129 is indeed an essential tree realization of @xmath0 . from the definition of @xmath118 in ( [ se*_def ] ) ,
it is clear that for each @xmath65 , ( _ e^ * ) = ( ) - ( _ j(e ) ) - ( _ |j(e ) ) .
[ dimse * ] it is useful to point out that @xmath130 may also be expressed as ( _ e^ * ) = ( |_j(e ) ) + ( |_|j(e ) ) - ( ) , [ dimse*_alt ] a consequence of the fact that for any @xmath6 , @xmath30 .
thus , by the uniqueness of minimal tree realizations , if @xmath131 is a tree realization of @xmath0 with the property that for all @xmath65 , @xmath132 equals one of the expressions in ( [ dimse * ] ) or ( [ dimse*_alt ] ) , then @xmath133 is in fact @xmath129 .
forney @xcite also derived an expression for the dimension of the local constraints @xmath134 .
consider any @xmath51 .
for each @xmath83 , we specify @xmath56 to be the component of @xmath55 that does _ not _ contain @xmath52 . as usual , @xmath66 .
then ( * ? ? ?
* theorem 1 ) , ( c_v^ * ) = ( ) - _ e e(v ) ( _ j(e ) ) .
[ dimcv * ] forney gave the following bound for @xmath135 ( * ? ? ?
* theorem 5 ) : for any @xmath83 , @xmath136 , where @xmath137 denotes the length of the code @xmath134 .
the upper bound can be improved slightly . in the minimal tree realization @xmath129
, we have , for @xmath51 and @xmath83 , @xmath138 [ dimcv*_bnd ] the upper bound may be proved as follows .
since @xmath30 for any @xmath6 , we may write ( [ dimcv * ] ) as @xmath139 now , let @xmath83 be fixed .
we have @xmath140 however , as can be seen from figure [ vpersp_fig ] , @xmath141 is the disjoint union of @xmath64 and the sets @xmath142 , @xmath143
. therefore , @xmath144 and hence , @xmath145 the lemma now follows from ( [ dimse*_alt ] ) .
as already mentioned , among all tree realizations of @xmath0 extending @xmath62 , the minimal realization @xmath129 minimizes state space dimension at each edge of the tree @xmath46 .
it is natural to ask whether @xmath129 also minimizes local constraint code dimension at each vertex of @xmath46 .
we will show in the next section that @xmath129 does in fact have the following property : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ if @xmath71 is a tree realization of @xmath0 that extends @xmath62 , then for all @xmath51 , @xmath146 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we will deduce this fact from an alternative construction of @xmath129 that we present next .
the construction we describe in this section takes an arbitrary tree realization @xmath111 that extends the tree decomposition @xmath62 for example , the trivial extension given in example [ trivial_example ] and via a sequence of transformations , converts @xmath111 to @xmath129 .
these transformations constitute a natural generalization of the state - merging process in the context of minimal trellis realizations ; see , for example , ( * ? ? ?
* section 4 )
. it would be useful to keep this special case in mind while going through the details of the description that follows .
let @xmath147 be an essential instead ; see theorem [ merge_theorem ] . ]
tree realization of a code @xmath0 with index set @xmath3 , and let @xmath80 be the full behavior of @xmath111 . as @xmath111 is essential , we have that @xmath105 for all @xmath65 ( by definition ) , and @xmath148 for all @xmath51 ( by lemma [ ess_lemma ] ) .
pick an arbitrary edge @xmath149 , and for ease of notation , set @xmath150 and @xmath151 .
let @xmath152 be the subspace of @xmath153 defined by @xmath154 we will define a new tree model @xmath155 , such that states in the same coset of @xmath152 in @xmath153 are represented by a single `` merged '' state in @xmath156 .
let @xmath157 be the mapping defined by @xmath158 define @xmath159 .
it is clear from the definitions that @xmath160 , and that @xmath161 .
consider now the tree model @xmath155 , where @xmath162 for each @xmath65 , and @xmath163 for each @xmath51 .
note that @xmath164 , and for @xmath165 , we have @xmath166 .
all states in @xmath153 belonging to the same coset of @xmath152 , say , @xmath167 , are mapped to ( or merged into ) the single state @xmath168 in @xmath156
. further note that if @xmath52 is not one of the two vertices incident with @xmath169 , then @xmath170 . at the two vertices that are incident with @xmath169 ,
the local constraints are appropriately modified to take into account the state - merging at edge @xmath169 .
in any case , we have ( |_e ) = ( ||_e ) ( |_e ) = ( _ e ) , e e , [ dimse_ineq ] and ( |c_v ) = ( ||_v ) ( |_v ) = ( c_v ) , v v. [ dimcv_ineq ] + we claim that @xmath171 is an essential tree realization of @xmath0 . to prove this claim , we must show that @xmath172 for all @xmath65 , and that @xmath173 , where @xmath174 denotes the full behavior of @xmath171 .
note that we do _ not _ claim that @xmath175 ; indeed , this may not be true .
it is easy to see that the inclusion @xmath176 holds . indeed , since @xmath177 , it is evident that any @xmath178 satisfies all the local constraints of @xmath171 , and hence is in @xmath174
. therefore , @xmath179 , and in particular , @xmath180 .
the reverse inclusion , @xmath181 , follows from part ( a ) of the lemma below .
let @xmath182 be a global configuration in @xmath174 .
then , * @xmath183 ; and * @xmath184 if and only if @xmath185 . [ gbar_lemma1 ] we defer the proof of the lemma to appendix [ gbar_lemmas_app ] . lemma [ gbar_lemma1](a ) shows that @xmath181 , thus proving that @xmath171 is a tree realization of @xmath0 .
it remains to show that @xmath111 is essential , _
i.e. _ , that @xmath172 for all @xmath65 .
this is shown by the following simple argument .
we have already seen that @xmath186 , and hence , @xmath187 for all @xmath65 . on the other hand , at any @xmath65 , @xmath188 is , by definition , a subspace of @xmath162 .
hence , @xmath171 is essential , thus proving our original claim , which we record in the following proposition .
the tree model @xmath171 is an essential tree realization of @xmath0 .
[ gbar_prop ] let us call the process described above of obtaining @xmath171 from @xmath111 as the _ state - merging process at edge @xmath169_. we use the notation @xmath189 to denote this transformation .
our goal now is to show that , starting from an essential tree realization , if we apply the state - merging process at each edge of the underlying tree , then we always end up with a minimal realization .
a proof of this assertion requires the following technical lemma , whose proof we also defer to appendix [ gbar_lemmas_app ] .
suppose that there exists @xmath190 such that the full behavior , @xmath80 , of @xmath111 satisfies the following property : for @xmath101 , we have @xmath191 if and only if @xmath192 .
then , for any @xmath193 , we also have @xmath194 if and only if @xmath195 .
[ gbar_lemma2 ] we are now in a position to prove the main result of this section , which provides a construction of @xmath129 via state merging .
let @xmath111 be a tree realization of @xmath0 that extends the tree decomposition @xmath62 .
let @xmath196 be a listing of the edges of @xmath46 .
set @xmath197 , and for @xmath198 , set @xmath199 .
then , @xmath200 is the minimal tree realization @xmath129 .
[ merge_theorem ]
let @xmath80 denote the full behavior of @xmath111 ( and hence , also of @xmath113 ) , and for @xmath198 , let @xmath201 denote the full behavior of @xmath202 . by proposition [ gbar_prop ] , each @xmath202 is an essential tree realization of @xmath0 . by lemma [ gbar_lemma1](b ) , for any @xmath203
, we have @xmath204 if and only if @xmath205 .
furthermore , by lemma [ gbar_lemma2 ] , for any @xmath206 , @xmath207 satisfies the following property : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ for any @xmath208 , we have @xmath204 if and only if @xmath205 . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ in particular , @xmath209 satisfies the following property for @xmath198 : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ for any @xmath210 , we have @xmath204 if and only if @xmath205 .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ let us call the above property ( p ) .
property ( p ) has two important consequences .
firstly , it implies that if @xmath210 is such that @xmath211 , then @xmath212 for all @xmath65 .
this means that the projection @xmath213 defined by @xmath214 is in fact an isomorphism . for the second consequence of ( p ) ,
consider , for any @xmath65 , the homomorphism @xmath215 defined by @xmath216 .
this map is well - defined since @xmath217 is an isomorphism .
property ( p ) is equivalent to the assertion that , for any @xmath65 , the kernel of @xmath218 is precisely @xmath219 .
therefore , @xmath220 .
thus , for each @xmath65 , state space @xmath221 is isomorphic to @xmath118 defined in ( [ se*_def ] ) , and the map @xmath218 is the canonical projection map @xmath222 given by ( [ se*_def ] ) .
it easily follows that for each @xmath51 , @xmath223 is isomorphic to @xmath134 defined in ( [ cv*_def ] ) .
hence , @xmath224 is the minimal realization @xmath129 .
observe that at each step of the procedure outlined in theorem [ merge_theorem ] , the dimensions of the state spaces and the local constraints do not increase . to make this precise , given tree models @xmath225 and @xmath226 , let us say that @xmath227
if @xmath228 for all @xmath65 , and @xmath229 for all @xmath51 . then , for @xmath111 and @xmath202 , @xmath230 , as in the statement of theorem [ merge_theorem ] , we have by virtue of ( [ dimse_ineq ] ) and ( [ dimcv_ineq ] ) , @xmath231 thus , we have that if @xmath111 is any tree realization of @xmath0 that extends the tree decomposition @xmath62 , then @xmath232 .
we record this strong property of minimal realizations as a corollary to theorem [ merge_theorem ] .
let @xmath62 be a tree decomposition of the index set of a code @xmath0 , and let @xmath233 be the corresponding minimal tree realization of @xmath0 .
then , for any tree realization , @xmath71 , of @xmath0 that extends @xmath62 , we have @xmath115 for all @xmath65 , and @xmath146 for all @xmath51 . [ dimcv_cor ]
the procedure outlined in theorem [ merge_theorem ] does not translate to an efficient algorithm for the construction of @xmath129 .
this is because the state - merging procedure that creates @xmath202 from @xmath234 requires knowledge of the full behavior of @xmath234 , which may not be easily determined .
so , as a practical method for constructing @xmath129 , given @xmath0 and @xmath62 , we propose a novel construction that relies upon the code decomposition techniques of the next section .
in previous work @xcite , it was demonstrated that techniques from the decomposition theory of matroids @xcite,@xcite could be put to good use in a coding - theoretic setting .
the decomposition theory in that work was presented in the context of binary linear codes . as we will now show
, the basic elements of that theory can be easily extended to cover the case of nonbinary codes as well .
the object of this exercise is not just to create a more general code decomposition theory , but as we will see in the next section , this decomposition theory ties in very nicely with the theory of tree realizations .
let @xmath33 and @xmath34 be linear codes over the finite field @xmath235 , defined on the index sets @xmath35 and @xmath36 , respectively .
let @xmath236 denote the symmetric difference , @xmath237 , of the index sets .
we will construct a code @xmath238 with @xmath236 as its index set . for @xmath239 and @xmath240 , let @xmath241 be defined by @xmath242 setting @xmath243 , we see that @xmath244 has @xmath38 as its index set .
we take @xmath238 to be the cross - section @xmath245 .
note that when @xmath246 , we have @xmath247 . for @xmath248 ,
let @xmath249 and @xmath250 denote the projection @xmath251 and the cross - section @xmath252 , respectively .
the codes @xmath249 and @xmath250 , for @xmath248 , all have @xmath253 as their index set .
the dimension of @xmath238 can be expressed in terms of the codes @xmath254 , @xmath249 and @xmath250 , @xmath255 , as stated in the following lemma . for codes
@xmath256 , we have @xmath257 where @xmath258 .
[ dim_sum_prop ] for a code @xmath0 , and a subset @xmath19 of its index set , the kernel of the projection map @xmath28 is isomorphic to @xmath22 , and hence , @xmath30 .
thus , taking @xmath259 , and @xmath260 , we find that @xmath261 since @xmath262 .
so , we must show that @xmath263 .
let @xmath264 be a copy of @xmath34 defined on an index set that is disjoint from @xmath35 .
for each @xmath265 , denote by @xmath266 its copy in @xmath264 .
consider the homomorphism @xmath267 defined by @xmath268 .
note that @xmath269 iff @xmath270 and @xmath271 .
equivalently , @xmath269 iff @xmath272 , @xmath273 , and @xmath274 .
it follows that the kernel of @xmath275 is isomorphic to @xmath276 which is simply @xmath277 .
hence , @xmath278 , as desired .
we will restrict our attention to a particular instance of the @xmath238 construction , in which we require that the codes @xmath249 and @xmath250 , @xmath248 , take on a specific form .
we need to introduce some notation first . for each positive integer @xmath279 ,
set @xmath280 , and fix an @xmath281 matrix , which we denote by @xmath282 , over @xmath283 , with the property that each pair of columns of @xmath282 is linearly independent over @xmath283 .
note that @xmath282 is a parity - check matrix for an @xmath284 $ ] hamming code over @xmath283 ( cf .
let @xmath285 denote the dual of this hamming code , _
i.e. _ , @xmath285 is the @xmath286 $ ] code over @xmath283 _ generated _ by @xmath282 .
the code @xmath285 is sometimes referred to as a _ simplex code_. we take a moment to record an important property of the matrix @xmath282 that we will use later .
the column vectors of @xmath282 form a maximal subset of @xmath287 with the property that each pair of vectors from the subset is linearly independent over @xmath283 .
this is due to the fact that the number of distinct one - dimensional subspaces of @xmath287 is precisely @xmath288 .
therefore , any ( column ) vector in @xmath287 is a scalar multiple of some column of @xmath282 . given an @xmath289 , suppose that the codes @xmath33 and @xmath34 , defined on the index sets @xmath35 and @xmath36 , respectively , are such that @xmath290 , and for @xmath248 , we have @xmath291 and @xmath292 . in such a case , @xmath238 is called the _ @xmath279-sum _ of @xmath33 and @xmath34 , and is denoted by @xmath293 .
it is convenient to extend this definition to the case of @xmath294 as well : when @xmath295 , the @xmath296-sum @xmath297 is defined to be the direct sum @xmath298 .
consider the case of codes defined over the binary field @xmath299 .
note that @xmath300 .
suppose that @xmath301 , and that the coordinates of @xmath33 are @xmath34 are ordered so that the index common to @xmath35 and @xmath36 corresponds to the last coordinate of @xmath33 and the first coordinate of @xmath34 .
the conditions necessary for the 1-sum @xmath302 to be defined can then be stated as * @xmath303 is not a codeword of @xmath33 , and the last coordinate of @xmath33 is not identically zero ; * @xmath304 is not a codeword of @xmath34 , and the first coordinate of @xmath34 is not identically zero .
the composite code @xmath238 resulting from @xmath305 that satisfy ( p1 ) , ( p2 ) above was studied in @xcite , where it was actually called a `` 2-sum '' .
we would also like to point out that the specialization of our @xmath279-sum operation to the case @xmath306 was called `` @xmath307-sum '' in @xcite .- sum operations defined in @xcite imposed additional conditions on the lengths of the codes involved in the sum , which we have dropped here . ] to add to the confusion , there was in fact an operation called `` 3-sum '' defined in @xcite , but that , in a certain sense , dualizes the 2-sum operation we have given in this paper . for @xmath289 , note that if @xmath249 and @xmath250 ( @xmath248 ) are in the form needed to define an @xmath279-sum , then @xmath308 , and @xmath309 .
therefore , as a corollary to proposition [ dim_sum_prop ] , we have the following result ( which also applies trivially to the @xmath294 case ) . for @xmath310 ,
if @xmath33 , @xmath34 are such that @xmath293 can be defined , then @xmath311 [ dim_rsum_cor ] an elementary property of direct sums ( _ i.e. _ , 0-sums ) is that a code @xmath0 is expressible as a direct sum of smaller codes if and only if there exists a partition @xmath312 of the index set of @xmath0 such that @xmath313 .
this property extends beautifully to @xmath279-sums in general .
let @xmath0 be a linear code over @xmath283 , defined on the index set @xmath3 , and let @xmath279 be a positive integer .
then , the following statements are equivalent .
* @xmath314 for some codes @xmath33 , @xmath34 .
* there exists a partition @xmath312 of @xmath3 , with @xmath315 , such that @xmath316 [ rsum_theorem ] : see appendix [ rsum_app ] .
+ : we give here a complete proof of this direction of the theorem , as it gives an explicit construction of codes @xmath33 , @xmath34 such that @xmath314 , given a partition @xmath312 as in ( b ) .
the proof generalizes ideas from similar constructions presented in @xcite .
let @xmath312 be a partition of @xmath3 such that @xmath317 .
set @xmath318 and @xmath319 , and let @xmath320 be a @xmath321 generator matrix for @xmath0 . without loss of generality , we may assume that the columns of @xmath320 are ordered so that the first @xmath322 columns are indexed by the elements of @xmath19 , and the rest by the elements of @xmath323 .
in the following exposition , we will often permute the columns of @xmath320 to bring the matrix into some desired form .
whenever this is the case , it will be tacitly assumed that column indices migrate with the columns .
let @xmath324 and @xmath325 denote the restrictions of @xmath320 to the columns indexed by the elements of @xmath19 and @xmath323 , respectively ; thus , @xmath326 $ ] .
let @xmath327 and @xmath328 ; by our assumption on @xmath312 , we have we have @xmath329 .
bring @xmath320 into reduced row - echelon form ( rref ) over @xmath283 .
permuting within the columns of @xmath324 and within those of @xmath330 if necessary , @xmath331 may be assumed to be of the form @xmath332 , \label{rref_eq}\ ] ] where @xmath333 , for @xmath334 , denotes the @xmath335 identity matrix , @xmath43 is a @xmath336 matrix , @xmath337 is a @xmath338 matrix , @xmath339 is a @xmath340 matrix , and the @xmath341 s denote all - zeros matrices of appropriate sizes .
the fact that the submatrix @xmath342 $ ] must have rank equal to @xmath328 implies that @xmath337 must have rank @xmath343 .
hence , @xmath337 has @xmath279 linearly independent rows , call them @xmath344 , which form a basis of the row - space of @xmath337 . permuting the first @xmath345 rows of @xmath346 if necessary
, we may assume that @xmath344 constitute the first @xmath279 rows of @xmath337 .
( permuting these rows of @xmath346 will also permute the rows of the @xmath347 matrix , but the effects of this can be negated by appropriately permuting the first @xmath345 columns of @xmath346 . )
any row of @xmath337 is uniquely expressible as a linear combination ( over @xmath283 ) of @xmath348 . in particular , for @xmath349 , the @xmath350th row of @xmath337 can be uniquely expressed as @xmath351 for some @xmath352 .
let us denote by @xmath353 , the rows of the @xmath281 generator matrix , @xmath282 , of the code @xmath285 .
let @xmath354 be the @xmath355 matrix such that for @xmath349 , the @xmath350th row of @xmath354 equals @xmath356 , where the @xmath357 s are such that the @xmath350th row of @xmath337 is @xmath351 .
thus , the row - space of @xmath354 is the span of @xmath358 , _
i.e. _ , it is the code @xmath285 . to the columns of @xmath354 , we assign indices from some set @xmath359 disjoint from @xmath3 .
now , define the @xmath360 matrix g_1 = , [ g1_def ] allowing the submatrix @xmath361 $ ] to retain its column indices from @xmath346 . also , define the @xmath362 matrix g_2 = , [ g2_def ] again allowing the submatrix @xmath342 $ ] to retain its column indices from @xmath346 .
thus , the index set of the columns of @xmath363 is @xmath364 , while that of the columns of @xmath365 is @xmath366 .
finally , for @xmath248 , let @xmath254 denote the code over @xmath283 generated by @xmath367 .
the following facts about @xmath33 and @xmath34 may be verified : * @xmath368 , @xmath248 .
* @xmath293 can be defined , so that by corollary [ dim_rsum_cor ] , @xmath369 . *
all rows of @xmath346 are in @xmath293 .
since @xmath346 generates the same code as @xmath320 ( recall that column indices get permuted along with columns ) , we see that @xmath293 contains all the codewords of @xmath0 .
we leave the details of the routine verification of the above facts to the reader .
it only remains to point out that facts ( ii ) and ( iii ) above show that @xmath370 , thus completing the proof of the implication ( b ) @xmath371 ( a ) .
the procedure described in the above proof can be formalized into an algorithm that takes as input a @xmath321 generator matrix @xmath320 ( over @xmath283 ) for @xmath0 , and a partition @xmath312 of the index set of @xmath0 , and produces as output generator matrices of two codes @xmath33 and @xmath34 ( and their associated index sets ) such that @xmath314 , where @xmath372 .
the run - time complexity of this procedure is determined by the following : * an rref computation to find @xmath346 as in ( [ rref_eq ] ) ; this can be carried out in @xmath373 time , which is the run - time complexity of bringing a @xmath321 matrix to reduced row - echelon form via elementary row operations ; * the computations required to identify a basis ( @xmath348 ) of the row - space of the matrix @xmath337 , and correspondingly the coefficients @xmath357 ; this could be done by computing the rref of @xmath337 , which would also take @xmath373 time ; * the computations needed to determine the @xmath355 matrix @xmath354 ; each row of the matrix requires @xmath374 computations , and there are @xmath375 rows , so the computation of @xmath354 takes @xmath376 time .
therefore , the entire procedure can be carried out in @xmath377 time .
it is worth noting that the run - time complexity of the procedure is polynomial in @xmath14 , @xmath15 and @xmath378 , but exponential in @xmath279 .
the above procedure for determining an @xmath279-rum decomposition of a given code forms the basis of a new construction of minimal tree realizations that we present in this section .
the key observation behind this construction is that if a code @xmath0 has a partition @xmath312 of its index set such that @xmath379 , then @xmath0 has an essential tree realization of the form depicted in figure [ rsum_fig ] .
the tree in the figure consists of a single edge @xmath380 , the state space @xmath72 is the code @xmath285 , and the local constraint codes at the two vertices are the codes @xmath33 and @xmath34 such that @xmath370 . in fact
, this is the minimal realization @xmath129 , for the tree @xmath46 consisting of the single edge @xmath380 , and the index map @xmath70 such that @xmath381 and @xmath382 .
this is simply because @xmath383 , so by virtue of ( [ dimse*_alt ] ) , @xmath72 has the same dimension as the state space @xmath384 in the minimal realization @xmath129 .
so , by the uniqueness of minimal tree realizations , the tree realization depicted in figure [ rsum_fig ] is @xmath129 . to summarize ,
if @xmath0 is a code defined on the index set @xmath3 , and @xmath62 is a tree decomposition of @xmath3 such that @xmath46 consists of the single edge @xmath385 , then we may construct @xmath129 as follows .
set @xmath386 and @xmath387 , and compute @xmath372 . assign an index set @xmath388 that is disjoint from @xmath3 to the code @xmath285 .
use the procedure in the proof of theorem [ rsum_theorem ] to determine codes @xmath33 and @xmath34 , defined on the respective index sets @xmath389 and @xmath390 , such that @xmath314 . for @xmath255 , assign @xmath254 to be the local constraint code at vertex @xmath391 , and assign @xmath285 to be the state space at edge @xmath54 .
the resulting tree model @xmath392 is the minimal tree realization @xmath129 . before describing how the construction may be extended to the case of trees with more than one edge
, we deal with the trivial case of trees without any edges .
if @xmath46 is a tree consisting of a single vertex @xmath52 , and no edges , then given any code @xmath0 defined on some index set @xmath3 , there is only one way of realizing @xmath0 on @xmath46 .
this is the realization @xmath393 , where @xmath70 is the unique mapping @xmath394 , and @xmath73 is the code @xmath0 itself .
of course , this is also the minimal realization @xmath129 . at this point , we know how to construct @xmath129 , for any code @xmath0 , and any tree decomposition @xmath62 such that @xmath46 has at most one edge .
from this , we can recursively construct @xmath129 for any @xmath0 and any @xmath62 , as we now describe .
suppose that we know how to construct @xmath129 for any @xmath0 , and any @xmath62 such that @xmath46 has at most @xmath395 edges , for some integer @xmath396 .
let @xmath0 be a code defined on the index set @xmath3 , and let @xmath62 be a tree decomposition such that @xmath397 . pick any @xmath398 , and as usual , let @xmath399 and @xmath400 be the two components of @xmath401 .
we will assume that @xmath402 and @xmath403 .
let @xmath404 and @xmath405 .
compute r = ( |_j ( ) ) + ( | _ ( ) ) - ( ) , [ r_def ] which determines the code @xmath285 .
assign @xmath285 an index set @xmath388 that is disjoint from @xmath3 . use the procedure in the proof of theorem [ rsum_theorem ] to determine codes @xmath33 and @xmath34 , defined on the respective index sets @xmath406 and @xmath407 , such that @xmath408 .
now , define the index maps @xmath409 and @xmath410 as follows ( see figure [ rec_constr_fig ] ) : @xmath411 thus , @xmath412 and @xmath413 are tree decompositions of the index sets of @xmath33 and @xmath34 , respectively . as neither @xmath414 nor @xmath415 contains the edge @xmath169 , we have @xmath416 and @xmath417 .
therefore , by our assumption , we know how to construct @xmath418 and @xmath419 .
let @xmath420 finally , set @xmath421 , where _ e = _ e^(1 ) , & e e(t _ ) + _ r , & e = + _ e^(2 ) , & e e ( _ ) , [ se_def ] and c_v = c_v^(1 ) , & v v(t _ ) + c_v^(2 ) , & v v ( _ ) .
[ cv_def ] figure [ gamma_star ] contains a depiction of @xmath422 .
it is easy to see that @xmath422 is a tree realization of @xmath0 .
indeed , @xmath418 is a realization of @xmath33 , and @xmath419 is a realization of @xmath34 , and hence ( as should be clear from figure [ gamma_star ] ) , @xmath422 is a realization of @xmath370 .
it is not immediately obvious that @xmath422 is actually @xmath129 , but this is in fact true , as stated in the following proposition , a proof of which is given in appendix [ rec_constr_app ] .
@xmath422 is the minimal tree realization @xmath129 .
[ rec_constr_prop ] in summary , we have the following recursive procedure for constructing @xmath129 , given a code @xmath0 and a tree decomposition @xmath62 .
+ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ * procedure * ` min_realzn`(@xmath423 ) + _ input _ : a @xmath321 generator matrix for a code @xmath0 , and a tree decomposition @xmath62 of the index set of @xmath0 .
+ _ output _ : a specification of the state spaces and the local constraints in the minimal realization @xmath129 . +
* step m1*. if @xmath46 consists of a single vertex , then return @xmath424 . +
* step m2*. if @xmath46 contains at least one edge , then choose an @xmath425 .
let @xmath426 be the vertex of @xmath399 incident with @xmath169 , and let @xmath427 be the vertex of @xmath400 incident with @xmath169 .
+ ( m2.1 ) compute @xmath428 .
+ ( m2.2 ) determine @xmath285 , and assign it an index set @xmath388 disjoint from @xmath429 .
+ ( m2.3 ) determine codes @xmath33 and @xmath34 , with index sets @xmath406 and @xmath407 , respectively , such that @xmath408 .
+ ( m2.4 ) determine the index maps @xmath430 and @xmath431 as in ( [ omega1_def ] ) and ( [ omega2_def ] ) .
+ * step m3*. determine @xmath418 by calling ` min_realzn`(@xmath432 ) ; determine @xmath419 by calling ` min_realzn`(@xmath433 ) .
we may assume that @xmath418 and @xmath419 are in the form given in ( [ minc1_form ] ) and ( [ minc2_form ] ) .
+ * step m4*. return @xmath434 , where @xmath72 and @xmath73 are as defined in ( [ se_def ] ) and ( [ cv_def ] ) .
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ a simplified version of the above procedure may be obtained by choosing , in step m2 , the edge @xmath169 to be an edge incident with a leaf of @xmath46 .
then , one of the two components of @xmath435 , say , @xmath400 , consists of a single vertex , so that the call to ` min_realzn`(@xmath436 ) may be avoided , as it would simply return @xmath437 .
we will use this modification of the procedure to give an estimate of its run - time complexity .
let @xmath14 denote the length of @xmath0 , let @xmath319 , and let @xmath438 .
also , define r _
= _ e e ( |_j(e ) ) + ( |_(e ) ) - ( ) .
[ rmax_eq ] observe that , as a result of the modification suggested above , in the determination of @xmath129 , the procedure ` min_realzn ` gets called @xmath439 times , once for each edge @xmath65 .
the run - time complexity of any particular run of ` min_realzn ` is determined by the computations in step m2 . in the @xmath350th run ,
the procedure acts upon some code @xmath440 of length @xmath441 and dimension @xmath442 , and in step m2 , it computes an @xmath443 , a code @xmath444 with index set @xmath445 , and a code @xmath446 . via lemma
[ appd_lemma ] , we have that @xmath447 .
we bound @xmath442 and @xmath441 as follows .
note that @xmath446 is the code @xmath448 that the @xmath449th run of the procedure takes as input .
thus , we have @xmath450 , and @xmath451 . since @xmath452 , @xmath453 , and @xmath454
, we have , for @xmath455 , @xmath456 and @xmath457 .
now , by the estimate given in section [ decomp_section ] of the run - time complexity of the @xmath279-sum decomposition procedure , we see that the @xmath350th run of step m2 of ` min_realzn ` takes @xmath458 time .
hence the overall run - time complexity of ` min_realzn ` may be estimated to be @xmath459 .
this expression can be simplified by observing that @xmath460 \\ & = & ( k^2 + r_{\max}q^{r_{\max } } ) \ , \left[n|e| + ( 1/2)|e|(|e|-1)q^{r_{\max}}\right].\end{aligned}\ ] ] it follows that ` min_realzn ` runs in @xmath461 time .
note that this is polynomial in @xmath14 , @xmath15 , @xmath378 and @xmath439 , but exponential in @xmath462 .
as observed in @xcite , any graphical realization of a code specifies an associated decoding algorithm , namely , the sum - product algorithm . the sum - product algorithm specified by a tree realization
, @xmath463 , of a code @xmath0 provides an exact implementation of ml decoding for @xmath0 . a reasonable initial estimate of the computational complexity of the sum - product algorithm on @xmath111 is provided by the _ constraint complexity _ of @xmath111 , which is defined as @xmath464 .
as implied by corollary [ dimcv_cor ] , given a tree decomposition @xmath62 of the index set of @xmath0 , the minimal realization @xmath129 has the least constraint complexity among all tree realizations of @xmath0 that extend @xmath62 .
let @xmath465 denote the constraint complexity of @xmath129 .
note that , by ( [ dimcv * ] ) , ( ; t , ) = _ v v ( ( ) - _ e e(v ) ( _ j(e ) ) ) .
[ k_ct_omega ] thus , @xmath465 is a measure of the complexity of implementing ml decoding for @xmath0 as a sum - product algorithm on @xmath129 .
let us now define the _ treewidth _ of the code @xmath0 to be ( ) = _ ( t , ) ( ; t , ) , [ treewidth_def_0 ] where the minimum is taken over all tree decompositions @xmath62 of the index set of @xmath0 .
the treewidth of a code is an indicator of how small the computational complexity of an ml decoding algorithm for @xmath0 can be .
the notion of treewidth ( _ i.e. , _ minimal constraint complexity ) of a code was first introduced by forney @xcite . a related notion , called minimal tree complexity ,
was defined and studied by halford and chugg @xcite .
treewidth , as defined in ( [ treewidth_def_0 ] ) , is an upper bound on the minimal tree complexity measure of halford and chugg .
a tree is called _ cubic _ if all its internal nodes have degree 3 .
forney @xcite showed that the minimum in ( [ treewidth_def_0 ] ) is always achieved by a tree decomposition @xmath62 in which @xmath46 is a cubic tree , and @xmath70 is a bijection may be taken to be such that @xmath46 is a cubic tree and @xmath70 is a surjective map onto the leaves of @xmath46 .
however , the symbol - splitting argument in section v.f of his paper actually implies that @xmath70 in the minimizing tree decomposition may be taken to be one - to - one as well . ] between the index set of @xmath0 and the set of leaves of @xmath46 .
let @xmath466 denote the set of all tree decompositions @xmath62 in which @xmath46 is cubic and @xmath70 maps the index set of @xmath0 bijectively onto the set of leaves of @xmath46 .
we may then re - write ( [ treewidth_def_0 ] ) as ( ) = _ ( t , ) ( ) ( ; t , ) .
[ treewidth_def ] an alternate measure of code complexity may be obtained from the notion of _ state complexity _ of a tree realization @xmath111 , which is the largest dimension of a state space in @xmath111 .
thus , by virtue of ( [ dimse * ] ) and ( [ dimse*_alt ] ) , the state complexity of a minimal realization @xmath129 is given by @xmath467 we then define , in analogy with ( [ treewidth_def ] ) , ( ) = _ ( t , ) ( ) ( ; t , ) .
[ branchwidth_def ] note that the minimum in the above definition is taken over tree decompositions in @xmath466 only .
it must be emphasized that @xmath468 , as defined in ( [ branchwidth_def ] ) , need _ not _ be the same as the least @xmath469 over all tree decompositions @xmath62 of the index set of @xmath0 .
a notion analogous to @xmath468 is known as branchwidth in the matroid theory literature ; see _ e.g. _ , @xcite . in keeping with that nomenclature ,
we will call @xmath468 the _ branchwidth _ of the code @xmath0 .
branchwidth and treewidth are very closely related , as shown by the following result , which can be obtained in a straightforward manner from the bounds in lemma [ dimcv*_bnd ] . given a code @xmath0 , if @xmath470 , then @xmath471 hence , @xmath472 .
[ bw_tw_bnds ] the notions of state and constraint complexity have been studied extensively in the context of conventional trellis realizations of a code ; see _ e.g. _ , @xcite .
recall from example [ trellis_example ] that a conventional trellis realization of a code is a tree realization that extends a tree decomposition @xmath62 in which @xmath46 is a simple path and @xmath70 is a bijection between the index set of @xmath0 and the vertices of @xmath46 .
this special case of a tree decomposition is referred to as a _ path decomposition_. specifically , a path decomposition of a code @xmath0 defined on the index set @xmath3 is a pair @xmath62 , where @xmath46 is a simple path on @xmath473 vertices , and @xmath85 is a bijection .
let @xmath474 denote the set of all path decompositions of @xmath0 .
we then define _
( ) = _ ( t , ) ( ) ( ; t , ) [ k_trellis_def ] and _ ( ) = _ ( t , ) ( ) ( ; t , ) . [ s_trellis_def ] it is well - known , and indeed readily follows from lemma [ dimcv*_bnd ] , that @xmath475 .
it is clear from ( [ treewidth_def_0 ] ) and ( [ k_trellis_def ] ) that @xmath476 .
forney @xcite asked the question of whether @xmath477 could be significantly smaller than @xmath478 .
he conjectured that either @xmath479 for all codes @xmath0 , or @xmath480 is unbounded .
we show here that it is in fact the latter that is true .
to do so , we need to introduce some new concepts .
in their fundamental work on graph minors @xcite , robertson and seymour introduced two notions of complexity of graphs , namely , treewidth and pathwidth .
these notions have proved to be invaluable tools with many applications in graph theory and theoretical computer science .
an overview of such applications can be found , for example , in @xcite .
we will define the notions of treewidth and pathwidth of a graph in this subsection , and subsequently , relate them to the complexity measures @xmath477 and @xmath478 defined above for codes .
let @xmath1 be a graph with vertex set @xmath481 and edge set @xmath482 .
the graph may contain self - loops and parallel edges .
a _ tree decomposition _ of @xmath1 is a pair @xmath483 , where @xmath46 is a tree , and @xmath484 is a mapping that satisfies the following : * @xmath485 ; * for each pair of adjacent vertices @xmath486 , we have @xmath487 for some @xmath488 ; and * for each pair of vertices @xmath489 , if @xmath490 is any vertex on the unique path between @xmath491 and @xmath492 , then @xmath493 .
it may be helpful to point out that ( t3 ) above is equivalent to the following : * for each @xmath494 , the subgraph of @xmath46 induced by @xmath495 is a ( connected ) subtree of @xmath46 . a reader familiar with the notion of `` junction trees '' ( see _ e.g. _ , @xcite ) will recognize a tree decomposition of @xmath1 to be a junction tree .
the _ width _ of a tree decomposition @xmath483 as above is defined to be @xmath496 .
the _ treewidth _ of @xmath1 , which we denote by @xmath497 , is the minimum among the widths of all its tree decompositions .
note that if @xmath1 has at least one edge , then , because of ( t2 ) , any tree decomposition of @xmath1 must have width at least one .
thus , for any graph @xmath1 with @xmath498 , we have @xmath499 . for any tree @xmath46 with at least two vertices , we have @xmath500 .
this can be seen as follows .
fix a vertex @xmath501 .
define a mapping @xmath502 as follows : @xmath503 , and for @xmath504 , @xmath505 , where @xmath506 is the first edge on the unique path from @xmath491 to @xmath279 .
it is easily verified that @xmath483 is a tree decomposition of @xmath46 .
since this tree decomposition has width one , it follows that @xmath500 .
[ treewidth_example1 ] if @xmath483 is a tree decomposition in which @xmath46 is a simple path , then @xmath483 is called a _
path decomposition_. the minimum among the widths of all the path decompositions of @xmath1 is called the _ pathwidth _ of @xmath1 , which we denote by @xmath507 .
it is evident that @xmath508 .
analogous to the situation of example [ treewidth_example1 ] , a simple path has pathwidth one .
however , trees may have arbitrarily large pathwidth .
the following example is due to robertson and seymour @xcite .
let @xmath509 be the complete bipartite graph @xmath510 .
for @xmath511 , we inductively define @xmath512 by taking a copy of @xmath513 , and to each leaf @xmath52 of this graph , adding two new vertices adjacent to @xmath52 .
figure [ y_trees ] shows the trees @xmath509 ,
@xmath514 and @xmath515 .
the pathwidth of @xmath512 , @xmath516 , is @xmath517 @xcite .
[ treewidth_example2 ] thus , for trees @xmath46 , the difference @xmath518 can be arbitrarily large .
we will use this fact to construct codes @xmath0 for which @xmath480 is arbitrarily large .
we remark that the problem of determining the treewidth or pathwidth of a graph is known to be np - hard @xcite,@xcite .
as we will see a little later , this implies that the problem of determining the treewidth of a code , or its trellis counterpart , is also np - hard .
let @xmath2 be an arbitrary finite field . to any given graph @xmath1 , we will associate a code @xmath519 $ ] over @xmath2 as follows .
let @xmath520 be any directed graph obtained by arbitrarily assigning orientations to the edges of @xmath1 , and let @xmath521 be the vertex - edge incidence matrix of @xmath520 .
this is the @xmath522 matrix whose rows and columns are indexed by the vertices and directed edges , respectively , of @xmath520 , and whose @xmath523th entry , @xmath524 , is determined as follows : @xmath525 the code @xmath519 $ ] is defined to be the linear code over @xmath2 generated by the matrix @xmath521 . when @xmath2 is the binary field , the code @xmath519 $ ] is the _ cut - set code _ of @xmath1 , _
i.e. _ , the dual of the cycle code of @xmath1 @xcite . the following fundamental result that relates the treewidths of the graph @xmath1 and the code @xmath519 $ ] is due to hlinn and whittle @xcite .
if @xmath1 is a graph with at least one edge , then @xmath526)$ ] .
[ treewidth_theorem ] since determining the treewidth of a graph is np - hard , it immediately follows from the above theorem that the problem of determining the treewidth of a code ( over any fixed finite field ) is also np - hard .
we remark that the problem of determining the branchwidth of a code is also np - hard .
this follows from a result @xcite that relates the branchwidth of the code @xmath519 $ ] to the branchwidth of the graph @xmath1 , the latter being a notion we have not defined in this paper .
unfortunately , it is not true that @xmath527)$ ] . as an example , consider the code @xmath528 $ ] over the binary field , for an arbitrary tree @xmath46 .
it is not hard to see that @xmath528 = \{0,1\}^{|e(t)|}$ ] which , being the direct sum of multiple copies of @xmath529 , has @xmath530 ) = 1 $ ] .
but as we have already noted , trees can have arbitrarily large pathwidth .
we get around this problem by means of a suitable transformation of graphs . given a graph @xmath1 ,
let @xmath531 be a graph defined on the same vertex set as @xmath1 , having the following properties ( see figure [ gbar ] ) : * @xmath531 is loopless ; * a pair of distinct vertices is adjacent in @xmath531 iff it is adjacent in @xmath1 ; and * in @xmath531 , there are exactly two edges between each pair of adjacent vertices .
define @xmath532 to be the graph obtained by adding an extra vertex , @xmath491 , to @xmath531 , along with a pair of parallel edges from @xmath491 to each @xmath533 ( see figure [ gbar ] ) .
it is easy to see that @xmath532 is constructible directly from @xmath1 in @xmath534 time .
the following result was used in @xcite to show that the problem of determining @xmath535 for an arbitrary code @xmath0 ( over any fixed finite field ) is np - hard .
if @xmath532 is the graph constructed from a given graph @xmath1 as described above , then @xmath536 ) = \k_\path(\cg ) + 1 $ ] .
[ pathwidth_theorem ] since @xmath535 is always within one of @xmath478 , the above theorem implies that _ ( ) + 1 _ ( ) _ ( ) + 2 .
[ k_trellis_bnd ] while this falls short of establishing the np - hardness of computing @xmath478 for an arbitrary code @xmath0 , it is certainly enough to provide us with the desired example of codes @xmath0 for which @xmath480 is arbitrarily large .
we just need to make one more observation : @xmath537 .
the proof of this fact , which is along the lines of the proof of lemma 3.5 in @xcite , is left to the reader as a straightforward exercise .
we can now prove the following corollary to theorems [ treewidth_theorem ] and [ pathwidth_theorem ] .
over any finite field @xmath2 , there exists a family of codes @xmath254 , @xmath538 , such that @xmath539 [ pathwidth_cor ] let @xmath512 , @xmath540 , be the family of trees defined in example [ treewidth_example2 ] .
define @xmath541 $ ] , where @xmath542 refers to the graph obtained from @xmath512 by the transformation depicted in figure [ gbar ] .
note that @xmath543 , since @xmath544 , as shown in example [ treewidth_example1 ] .
thus , on the one hand , from theorem [ treewidth_theorem ] , we have @xmath545 . and on the other hand , from ( [ k_trellis_bnd ] ) and example [ treewidth_example2 ]
, we have @xmath546 . using standard facts known about the incidence matrix @xmath521 for a graph @xmath1 ( see , for example , ( * ? ? ?
* chapter 5 ) ) , it may be verified that the codes @xmath254 , @xmath516 , constructed in the above proof are @xmath547 $ ] codes , where @xmath548 note that @xmath549 grows as @xmath550 .
we conjecture that this is in fact the maximal rate of growth of the difference @xmath549 for any code family @xmath254 . if @xmath254 , @xmath516 , is any sequence of codes over @xmath2 , then @xmath551 where @xmath441 denotes the length of the code @xmath254 .
[ conj1 ] the codes @xmath254 constructed in the proof of corollary [ pathwidth_cor ] all have treewidth equal to two .
issues related to families of codes whose treewidth is bounded by a constant are discussed next .
many np - hard combinatorial problems on graphs are known to be solvable in polynomial ( often , linear ) time when restricted to graphs of bounded treewidth @xcite,@xcite . in this subsection
, we will see that the same general principle applies to problems pertaining to codes as well .
let @xmath235 be a fixed finite field .
given an integer @xmath552 , denote by @xmath553 ( resp .
@xmath554 ) the family of all codes over @xmath283 of treewidth ( resp .
branchwidth ) at most @xmath555 .
thus , a family @xmath556 of codes over @xmath283 is said to have bounded treewidth ( resp .
branchwidth ) if @xmath557 ( resp .
@xmath558 ) for some integer @xmath555 . note that by proposition [ bw_tw_bnds ] , @xmath559 , and so , a code family @xmath556 has bounded treewidth if and only if it has bounded branchwidth . a fundamental result of coding theory @xcite states that the problem of ml decoding is np - hard for an arbitrary family of codes
. however , we will now show that this problem becomes solvable in linear time for any code family of bounded treewidth .
so , consider a code family @xmath557 , where @xmath555 is a _ fixed _
integer , and pick an arbitrary @xmath560 .
let @xmath14 denote the length of @xmath0 . by definition
, @xmath0 has a minimal realization @xmath129 with constraint complexity at most @xmath555 .
moreover , by ( [ treewidth_def ] ) , @xmath62 can be chosen to be in @xmath466 , _
i.e. _ , it may be chosen so that @xmath46 is a cubic tree , and @xmath70 maps the index set of @xmath0 bijectively onto the leaves of @xmath46 .
in particular , the number of leaves of @xmath46 equals the cardinality , @xmath14 , of the index set of @xmath0 .
now , recall that ml decoding of @xmath0 may be implemented as a sum - product algorithm on any tree realization of @xmath0 , and in particular , on @xmath129 .
the computational complexity of the sum - product algorithm on @xmath129 is determined by the computations that take place at the internal nodes of @xmath46 . by an estimate of forney (
* theorem 5.2 ) , the number of computations at the internal node @xmath561 is of the order of @xmath562 , where @xmath563 is the degree of @xmath52 in @xmath46 . since @xmath46 is cubic , @xmath564 , and since the constraint complexity of @xmath129 is at most @xmath555 , we have @xmath565 . hence ,
the number of computations performed by the sum - product algorithm at any internal node of @xmath46 is bounded by @xmath566 , which is a constant .
now , @xmath46 is a cubic tree on @xmath14 leaves , so it has at most @xmath567 internal nodes .
it follows that the computational complexity of the sum - product algorithm on @xmath129 is @xmath568 , the constant in the @xmath569-notation being proportional to @xmath566 .
thus , there is a linear - time implementation of ml decoding for any @xmath560 .
a question that naturally arises in this context is that of how hard it is to explicitly determine the minimal tree realization required for linear - time implementation of ml decoding . note that this is not exactly a decoding complexity issue , since the determination of a suitable @xmath129 may be done `` off - line '' for each @xmath560 .
an explicit determination of @xmath129 involves finding a tree decomposition @xmath470 such that @xmath570 , and the specification of the state spaces and the local constraint codes of @xmath129 . given a @xmath470 satisfying @xmath570 ,
the state spaces and local constraint codes of @xmath129 may be determined by the ` min_realzn ` procedure of section [ new_construct_section ] .
an estimate of the computational complexity of this procedure was given in that section , in terms of the length and dimension of @xmath0 , the number of edges in @xmath46 , and the quantity @xmath462 defined in ( [ rmax_eq ] ) .
comparing ( [ rmax_eq ] ) with ( [ s_ct_omega ] ) , we see that @xmath462 is simply @xmath469 , which by proposition [ bw_tw_bnds ] , is bounded from above by @xmath555 .
the number of edges of @xmath46 is @xmath571 , and @xmath47 consists of @xmath14 leaves and at most @xmath567 internal nodes , @xmath14 being the length of @xmath0 .
therefore , by the estimate of the computational complexity of ` min_realzn ` given in section [ new_construct_section ] , for an @xmath13 $ ] code @xmath560 , and a @xmath470 such that @xmath570 , the minimal realization @xmath129 may be constructed in @xmath572 time . note that @xmath555 appears in the exponent of the constant implicit in the @xmath569-notation .
this leaves us with the problem of finding , for a given code @xmath573 , a tree decomposition @xmath470 such that @xmath570 .
unfortunately , there appears to be no efficient algorithm known for solving this problem .
however , reasonably good algorithms do exist for solving a closely related problem : given a code @xmath574 , find a tree decomposition @xmath470 such that @xmath575 .
several polynomial - time algorithms for solving this problem are given in @xcite , the most efficient of these being an algorithm that runs in @xmath576 time - notation depends exponentially on @xmath555 .
] , @xmath14 being the length of @xmath0 .
now , by proposition [ bw_tw_bnds ] , any @xmath573 is also in @xmath554 , and furthermore , @xmath577 .
therefore , the algorithms of @xcite find , for a given code @xmath573 , a tree decomposition @xmath470 such that @xmath578 .
this is sufficient for our purposes , as the computational complexity of the sum - product algorithm on the resulting @xmath129 would still be @xmath568 , except that the constant in the @xmath569-notation would now be proportional to @xmath579 .
while code families of bounded treewidth have the desirable property of having linear decoding complexity , it is very likely that they do not have good error - correcting properties .
we give an argument to support the plausibility of this statement .
recall from coding theory that a code family @xmath556 is called _ asymptotically good _ if there exists a sequence of @xmath547 $ ] codes @xmath580 , with @xmath581 , such that @xmath582 and @xmath583 are both strictly positive .
the code family @xmath254 , @xmath516 , from the proof of corollary [ pathwidth_cor ] has bounded treewidth , but is not asymptotically good : @xmath584 , but @xmath585 , as @xmath586 .
it is known that if a code family @xmath556 has bounded trellis complexity , _
i.e. _ , if there exists an integer @xmath555 such that @xmath587 for all @xmath560 , then @xmath556 is not asymptotically good .
this is a consequence of the following bound for an @xmath17 $ ] code @xmath0 @xcite : _ ( ) ( d-1 ) .
[ s_trellis_bnd ] since @xmath588 , the above is also a lower bound on @xmath478 . now , suppose that @xmath254 , @xmath516 , is a sequence of @xmath547 $ ] codes of bounded treewidth , so that there exists some @xmath552 such that for all @xmath350 , @xmath589 .
hence , from ( [ s_trellis_bnd ] ) , we have @xmath590 . therefore
, assuming the validity of conjecture [ conj1 ] , we have @xmath591 since @xmath555 is a constant , this implies that @xmath592 from which we infer that at least one of @xmath582 and @xmath583 must be zero .
observe that the final conclusion of the above argument can also be reached if we only assume that @xmath593 grows at most logarithmically with @xmath441 .
we formalize this as a conjecture .
let @xmath254 , @xmath516 , be any sequence of @xmath547 $ ] codes such that @xmath594 .
then , @xmath582 and @xmath583 can not both be strictly positive .
in particular , for any @xmath552 , the code family @xmath553 is not asymptotically good .
[ conj2 ] we wrap up our discussion on complexity measures for codes by elaborating on a comment we made at the beginning of this subsection , in which we implied that hard coding - theoretic problems often become polynomial - time solvable when restricted to codes of bounded complexity .
we saw earlier several examples of algorithms that , given a code @xmath573 , solve some problem in time polynomial in the length of @xmath0 . in each of these cases ,
the computational complexity of the algorithm displayed an exponential dependence on the parameter @xmath555 .
but since @xmath555 was a fixed constant , this exponential dependence could be absorbed into the constant hidden in the `` big-@xmath569 '' estimate of the complexity .
thus , fixing the parameter @xmath555 allowed a potentially intractable coding - theoretic problem to become tractable . problems that may be hard in general , but which become solvable in polynomial time when one of the parameters of the problem is fixed , are called _ fixed - parameter tractable_. we noted previously that the problems of computing the treewidth and branchwidth of a code are np - hard .
it should come as no surprise that these problems are in fact fixed - parameter tractable .
hlinn @xcite gives an @xmath576 algorithm that , for a fixed integer @xmath555 , determines whether or not a given length-@xmath14 code is in @xmath554 . from this
, one can also prove the existence of an @xmath576 algorithm for deciding membership of a given length-@xmath14 code in @xmath553 @xcite .
perhaps the most significant problem that remains open in the context of minimal tree realizations of codes is the resolution of conjecture [ conj2 ] , which proposes that codes of bounded treewidth can not be asymptotically good . it may be possible to resolve this by deriving a lower bound on treewidth along the lines of the bound in ( [ s_trellis_bnd ] ) .
another possibility is an inductive approach using code decompositions .
however , an open problem of far greater significance is the development of a general theory of minimal realizations of codes on graphs with cycles . at present
, such a theory only exists for the case of realizations of codes on graphs consisting of a single cycle , _
i.e. _ , tail - biting trellis realizations @xcite .
this simplest case of graphs with cycles is already more difficult to study than the cycle - free case for example , there can be several non - equivalent definitions of minimality in the context of tail - biting trellis realizations .
the challenge posed by graphs with more complex cycle structures can only be greater .
_ proof of lemma [ b|e_lemma]_. consider an arbitrary @xmath65 .
an arbitrary global configuration @xmath88 may be written in the form @xmath595 . now , suppose that @xmath88 is such that @xmath102 , _
i.e. _ , @xmath596 .
observe that the global configurations @xmath597 also satisfy all local constraints ( since @xmath598 for each @xmath51 ) , and hence are in @xmath80 .
therefore , @xmath599 , and so by definition of @xmath600 , we have @xmath601 . similarly , @xmath602 , so that @xmath603 . hence , @xmath604 .
+ _ proof of lemma [ ess_lemma]_. for any tree model ( essential or not ) , we have , by definition , @xmath91 for all @xmath51 .
so we need only show the reverse inclusion in the case when @xmath605 is an essential tree model .
pick an arbitrary @xmath51 .
let @xmath606 be the edges of @xmath46 incident with @xmath49 .
for @xmath607 , let @xmath608 denote the component of @xmath609 that does not include @xmath52 .
set @xmath610 , and @xmath611 .
we will write an arbitrary configuration @xmath101 as @xmath612 consider any @xmath613 , where @xmath614 , and @xmath615 for @xmath616 .
as the tree model @xmath111 is essential , we have @xmath617 for all @xmath350 .
in particular , @xmath618 , so that there exists @xmath619 such that @xmath620 .
as @xmath621 is in @xmath80 , its `` sub - configuration '' @xmath622 satisfies the local constraints of @xmath111 at all vertices in @xmath623 .
hence , @xmath624 satisfies the local constraints of @xmath111 at all vertices in @xmath625 .
now , @xmath52 is the only vertex of @xmath46 that is not in @xmath625 .
but , by construction , @xmath626 , and so , @xmath182 also satisfies the local constraint at @xmath52 .
thus , @xmath182 satisfies all local constraints of @xmath111 , so that @xmath627 .
hence , @xmath628 is in @xmath110 , which proves the lemma .
_ proof of lemma [ gbar_lemma1]_. for simplicity of notation , let @xmath97 denote the edge set of the subtree @xmath399 , and let @xmath629 denote that of the subtree @xmath630 . note that @xmath631 . throughout this proof
, we will write an arbitrary global configuration @xmath88 , belonging to @xmath80 or @xmath174 , in the form @xmath632 . consider any @xmath633 .
let @xmath634 and @xmath279 be the two vertices incident with the edge @xmath169 in @xmath46 .
we assume that @xmath635 and @xmath636 , as depicted in figure [ lemma1_pf_fig ] .
we write the local configuration @xmath637 as @xmath638 , and @xmath639 as @xmath640 .
suppose first that @xmath641 ; note that the zero element of @xmath642 ( @xmath643 ) is @xmath152 . by definition of @xmath171 , @xmath644 .
hence , there exists @xmath645 , for some @xmath646 .
now , @xmath647 ( being a `` sub - configuration '' of @xmath648 ) satisfies the local constraints of @xmath171 at all vertices in @xmath649 .
but these local constraints are of the form @xmath650 which , for @xmath651 , is identical to @xmath652 .
therefore , the sub - configuration @xmath647 satisfies the local constraints of @xmath111 at all vertices in @xmath649 .
it follows that @xmath653 satisfies the local constraints of @xmath111 at all vertices in @xmath654 , including @xmath634 . by a similar argument
, there exists a @xmath655 such that @xmath656 satisfies the local constraints of @xmath111 at all vertices in @xmath657 .
now , by definition of @xmath152 , there exist @xmath88 and @xmath658 in @xmath80 , such that @xmath659 , @xmath660 , and @xmath661 .
note , in particular , that the sub - configuration @xmath662 of @xmath88 satisfies the local constraints of @xmath111 at all vertices in @xmath657 .
therefore , the global configuration @xmath663 satisfies the local constraints of @xmath111 at all vertices in @xmath46 , and hence is in the full behavior , @xmath80 , of @xmath111 .
a similar argument shows that @xmath664 is also in @xmath80 .
as @xmath80 is a vector space , it must also contain @xmath665 and @xmath666 since @xmath111 is a tree realization of @xmath0 , we have @xmath81 . in particular , @xmath667 , and similarly , @xmath668 . hence , @xmath669 and @xmath670 .
however , @xmath88 and @xmath658 were chosen so that @xmath671 and @xmath672
. thus , we also have @xmath673 and @xmath674 .
this finally yields @xmath675 , thus proving one direction of part ( b ) of the lemma
. we will next show that if @xmath676 , then @xmath677 but @xmath678 .
this will prove both part ( a ) and the reverse direction of part ( b ) .
so , suppose that @xmath679 .
thus , @xmath680 is some coset of @xmath152 in @xmath681 , but is not @xmath152 itself .
pick some @xmath682 . as @xmath683
, there exists some @xmath101 such that @xmath684 .
observe that @xmath685 , but since @xmath686 , @xmath687 .
define @xmath688 , so that @xmath689 .
furthermore , @xmath690 , and @xmath691 ( @xmath692 ) is in @xmath0 but not in @xmath693 .
we have already noted ( prior to the statement of lemma [ gbar_lemma1 ] ) that @xmath179 .
therefore , @xmath694 , and since @xmath174 is a vector space , @xmath695 .
however , @xmath696 , and as we showed above , this implies that @xmath697 . since @xmath691 is in @xmath0 but not in @xmath693 , we find that @xmath698 , but @xmath699 .
the proof of the lemma is now complete .
+ _ proof of lemma [ gbar_lemma2]_. as @xmath171 is a tree realization of @xmath0 , lemma [ b|e_lemma ] shows that for any @xmath193 , we have @xmath194 only if @xmath195 .
thus , we need only prove the converse .
suppose that @xmath193 is such that @xmath700 , but @xmath701 .
now , @xmath702 , the last equality being a consequence of the fact that @xmath703 . therefore , there exists a @xmath101 such that @xmath704 .
note that , by the hypothesis of the lemma , @xmath705 .
set @xmath688 , so that @xmath706 .
observe that @xmath707 , and since @xmath703 , we also have @xmath708 .
thus , @xmath709 and @xmath710 .
but now , we have @xmath695 , with @xmath711 , and @xmath712 .
this contradiction of lemma [ b|e_lemma ] proves that there exists no @xmath193 such that @xmath700 , but @xmath701 .
_ proof of _ ( a ) _
@xmath371 _ ( b ) _ in theorem [ rsum_theorem]_. let @xmath314 for codes @xmath33 and @xmath34 defined on the index sets @xmath35 and @xmath36 , respectively . by definition , @xmath713 .
set @xmath714 and @xmath715 , so that @xmath312 forms a partition of @xmath3 . in what follows
, words defined on the index set @xmath35 will be written in the form @xmath716 ; words defined on the index set @xmath36 will be written in the form @xmath717 ; words defined on the index set @xmath3 will be written as @xmath718 ; and finally , words on the index set @xmath38 will be written as @xmath719 .
we begin by proving that @xmath720 .
this is accomplished by a two - step argument : we first show that @xmath721 , and then we show that @xmath722 . if @xmath723 , then there exists some @xmath724 , @xmath265 such that @xmath725 . however , @xmath726 as @xmath19 lies outside @xmath253 .
hence , @xmath727 .
conversely , suppose that @xmath728 . then , there exists @xmath729 such that @xmath730 .
since @xmath731 , there exists @xmath732 .
now , @xmath733 , and hence @xmath734 .
thus , @xmath723 , which completes the proof of the fact that @xmath721 . now , to show that @xmath735 , let us consider the projection map @xmath736 defined by @xmath737 .
this map is a homomorphism , with kernel isomorphic to @xmath738 , which is @xmath739 by definition .
hence , @xmath217 is in fact an isomorphism , which proves that @xmath735 .
we have thus shown that @xmath720 .
a similar argument yields the fact that @xmath740 .
hence , @xmath741 by corollary [ dim_rsum_cor ] .
it remains to show that @xmath315 .
note that since @xmath379 , and @xmath742 , we must have @xmath743 .
therefore , @xmath744 . by a similar argument
, we also have @xmath745 .
the proof of proposition [ rec_constr_prop ] requires the following lemma , which presents a property of the codes @xmath33 and @xmath34 obtained via the @xmath279-sum decomposition procedure of section [ decomp_section ] .
let @xmath0 be a code defined on the index set @xmath3 , and let @xmath312 be a partition of @xmath3 , with @xmath379 .
suppose that @xmath33 and @xmath34 are the codes , defined on the respective index sets @xmath35 and @xmath36 , that are obtained by the procedure described in the proof of theorem [ rsum_theorem ] .
then , for any @xmath746 , and any @xmath747 , we have @xmath748 [ appd_lemma ] we use notation from the proof of the ( b ) @xmath371 ( a ) direction of theorem [ rsum_theorem ] .
thus , @xmath0 , @xmath33 and @xmath34 are generated by the matrices @xmath346 , @xmath363 and @xmath365 given by ( [ rref_eq ] ) , ( [ g1_def ] ) and ( [ g2_def ] ) , respectively , which we reproduce here for the sake of convenience . @xmath749 , \\
g_1 & = & \left[\begin{array}{ccc } i_{k_1 } & a & x \\ \end{array } \right ] , \\ g_2 & = & \left[\begin{array}{ccc } x & \o & b \\
\o & i_{k - k_1 } & c \end{array}\right].\end{aligned}\ ] ] for any matrix @xmath750 , given a subset @xmath751 of the column indices of @xmath750 , we will denote by @xmath752 the restriction of @xmath750 to the columns indexed by @xmath751 .
thus , @xmath753,\ \ \ \ { g_1|}_j = \left [ \begin{array}{cc } i_{k_1 } & a \end{array } \right],\ ] ] @xmath754.\ ] ] our proof of the lemma uses only elementary linear algebra . we prove ( [ appd_lemma_eq1 ] ) first .
consider any @xmath746 .
it is clear that @xmath755 $ ] , and therefore , we have @xmath756 .
next , note that @xmath757 , from which we have @xmath758\right).\ ] ] now , observe that by performing column operations on @xmath759 , we can bring it into the form @xmath760.\ ] ] hence , @xmath761\right ) & = & \rank\left(\left[\ { \bar{g}|}_{j - j_1 } \ \ \
w\ \right]\right ) \\ & = & \rank(i_{k - k_1 } ) + \rank\left(\left[\ { g_1|}_{j - j_1 } \ \ \
b\ \right]\right ) \\ & = & k - k_1 + \rank\left(\left[\ { g_1|}_{j - j_1 } \ \ \ b\ \right]\right).\end{aligned}\ ] ] at this point , we have @xmath762\right),\ ] ] which upon re - arrangement yields @xmath763\right ) - \dim(\cc_1).\ ] ] thus , ( [ appd_lemma_eq1 ] ) would be proved if we could establish that @xmath764\right)$ ] .
now , @xmath765 , and hence , @xmath766\right).\ ] ] thus , we have to show that @xmath767\right ) = \rank\left(\left[\ { g_1|}_{j - j_1 } \ \ \ x\ \right]\right ) .
$ ] we will prove that the matrices @xmath337 and @xmath354 have identical column - spaces . clearly , the desired result then follows .
recall that for @xmath768 , the @xmath350th row of @xmath337 can be uniquely expressed as a linear combination , @xmath351 , of its first @xmath279 rows @xmath348 .
furthermore , the @xmath350th row of @xmath354 equals @xmath356 for the same @xmath357 s , where @xmath358 are the rows of the generator matrix , @xmath282 , of the code @xmath285 .
in particular , the first @xmath279 rows of @xmath354 constitute the matrix @xmath282 .
denote by @xmath769 the submatrix of @xmath337 comprised by its first @xmath279 rows .
now , it was pointed out in section [ decomp_section ] ( a little after the proof of proposition [ dim_sum_prop ] ) that any column vector in @xmath287 is a scalar multiple of some column of @xmath282 .
therefore , any column of @xmath769 is a scalar multiple of some column of @xmath282 .
but because of the way @xmath354 was constructed , this implies that any column of @xmath337 is a scalar multiple of some column of @xmath354 .
thus , the column - space of @xmath337 is a subspace of the column - space of @xmath354 .
however , we also have @xmath770 , and so , the column - spaces of the two matrices are in fact identical .
this proves that @xmath771\right ) = \rank\left(\left[\ { g_1|}_{j - j_1 } \ \
\ x\ \right]\right)$ ] , and ( [ appd_lemma_eq1 ] ) follows .
+ to show ( [ appd_lemma_eq2 ] ) , consider any @xmath747 .
arguments similar to the ones above establish that ( |_j_2 ) = ( _ 2|_j_2 ) ( |_i - j_2 ) = k_1 + ( |_- j_2 ) .
[ appd_eq1 ] now , consider @xmath772 .
noting that @xmath773 , we see that the matrix @xmath774 has the form @xmath775,\ ] ] with @xmath776 , for some @xmath777 .
since the columns of @xmath337 are contained in the column - space of @xmath354 , we can perform column operations on @xmath774 to bring it into the form @xmath778.\ ] ] hence , @xmath779 ) \notag \\ & = & ( k_1+k_2-k ) + \rank\left({\left[i_{k - k_1}\ \ \ c\right]\bigm|}_{\oj - j_2}\right ) .
\label{appd_eq2}\end{aligned}\ ] ] some trivial manipulations of ( [ appd_eq1 ] ) and ( [ appd_eq2 ] ) yield ( [ appd_lemma_eq2 ] ) , which proves the lemma .
_ proof of proposition [ rec_constr_prop]_. recall that @xmath421 , where @xmath72 and @xmath780 are as defined in ( [ se_def ] ) and ( [ cv_def ] ) . to show that @xmath422 is the minimal realization @xmath129 , it is enough to show that for all @xmath781
, @xmath782 equals the expression in ( [ dimse*_alt ] ) , _ i.e. _ , ( _ e ) = ( |_j(e ) ) + ( |_|j(e ) ) - ( ) .
[ appd_eq3 ] note that this is true when @xmath783 , since @xmath784 , and from ( [ r_def ] ) , we have @xmath428 .
we must therefore show that ( [ appd_eq3 ] ) holds for @xmath785 .
we will prove this for @xmath786 ; the proof for @xmath787 is similar .
so , consider any @xmath786 .
one of the two components , @xmath56 and @xmath788 , of @xmath789 is contained in @xmath399 .
without loss of generality , we may assume that it is @xmath56 that is a subtree of @xmath399 , as depicted in figure [ subtree ] .
hence , @xmath790 .
now , by ( [ se_def ] ) , @xmath791 , the latter being the state space associated with @xmath54 in @xmath418 .
therefore , by ( [ dimse*_alt ] ) , @xmath792 but , by lemma [ appd_lemma ] , the above expression is equal to the expression on the right - hand side of ( [ appd_eq3 ] ) .
hence , ( [ appd_eq3 ] ) holds for any @xmath786 , and the proposition follows .
bodlaender , `` dynamic programming on graphs of bounded treewidth , '' _ proc .
15th international colloquium on automata , languages and programming _ , vol .
317 , lecture notes in computer science , springer - verlag , pp .
105118 , 1988 . | a tree decomposition of the coordinates of a code is a mapping from the coordinate set to the set of vertices of a tree .
a tree decomposition can be extended to a tree realization , _ i.e. _ , a cycle - free realization of the code on the underlying tree , by specifying a state space at each edge of the tree , and a local constraint code at each vertex of the tree .
the constraint complexity of a tree realization is the maximum dimension of any of its local constraint codes .
a measure of the complexity of maximum - likelihood decoding for a code is its treewidth , which is the least constraint complexity of any of its tree realizations .
it is known that among all tree realizations of a code that extends a given tree decomposition , there exists a unique minimal realization that minimizes the state space dimension at each vertex of the underlying tree . in this paper , we give two new constructions of these minimal realizations . as a by - product of the first construction , a generalization of the state - merging procedure for trellis realizations , we obtain the fact that the minimal tree realization also minimizes the local constraint code dimension at each vertex of the underlying tree .
the second construction relies on certain code decomposition techniques that we develop .
we further observe that the treewidth of a code is related to a measure of graph complexity , also called treewidth .
we exploit this connection to resolve a conjecture of forney s regarding the gap between the minimum trellis constraint complexity and the treewidth of a code .
we present a family of codes for which this gap can be arbitrarily large . |
this work was supported by the eu - fp7 within the eu - india project athena .
supercomputer time was provided by the vienna scientific cluster . c. f. acknowledges grant support from the cas ( fellowship for young international scientists ) and the nsfc ( grand number : 51050110444 ) .
we are grateful to b .- j .
kennedy for sharing with us the unpublished low temperature structural data .
the value of the mixing parameter @xmath15=0.1 is reduced with respect to the standard value ( 0.25 ) since self - interaction error are considered to be smaller in 4@xmath3 transition metal oxides . se ref . | the evolution of the magnetic ordering temperature of the 4@xmath0 perovskites @xmath1tco@xmath2 ( @xmath1=ca , sr , ba ) and its relation with its electronic and structural properties has been studied by means of hybrid density functional theory and monte carlo simulations . when compared to the most widely studied 3@xmath3 perovskites the large spatial extent of the @xmath4 shells and their relatively strong hybridization with oxygen weaken the tendency to form jahn - teller like orbital ordering .
this strengthens the superexchange interaction .
the resulting insulating g - type antiferromagnetic ground state is characterized by large superexchange coupling constants ( 26 - 35 mev ) and ne@xmath5l temperatures ( 750 - 1200 k ) .
these monotonically increase as a function of the @xmath1 ionic radius due to the progressive enhancement of the volume and the associated decrease of the cooperative rotation of the tco@xmath6 octahedra .
intermediately located between manganese and rhenium , technetium ( 5@xmath74@xmath8 ) shares with its isovalent neighbors the intriguing possibility to form oxides with a complex structural , electronic and magnetic phase diagram .
however , the rare occurrence of natural tc ( tc is essentially an artificial product of fission reactions ) and the related radioactive risks ( tc is the lightest radioactive element , whose most abundant isotope , @xmath9tc , decays with a half - life of 10@xmath10 years ) , have made investigations of tc - based oxides very sparse @xcite . overcoming these difficulties , avdeev _ et al . _
@xcite and rodriguez _ et al . _
@xcite have recently reported the successful synthesis of the first ever fabricated tc - based perovskites , namely catco@xmath2 and srtco@xmath2 .
furthermore they have shown that these compounds display the anomalously high ne@xmath5l temperatures ( @xmath11 ) of 800 k for catco@xmath2 and 1000 k for srtco@xmath2 , by far the highest among materials not incorporating 3@xmath3 transition metals .
these results are surprising and challenge our understanding of the magnetic interaction in perovskites . in particular they pose three fundamental questions : i ) is the origin of such a large magnetic ordering temperature related to the strong tc 4@xmath3-o @xmath12 hybridization ? , ii ) what is the role played by the structural degrees of freedom ? and
iii ) is the enhancement of @xmath11 upon ca@xmath13sr substitution related to the observed increase in unit cell volume and the corresponding modification of the internal atomic positions @xcite ? in the present letter , by using a combination of hybrid density functional theory @xcite and monte carlo ( mc ) simulations @xcite , we address these issues at the microscopic level through a systematic study of the series @xmath1tco@xmath2 ( @xmath1=ca , sr , ba ) .
our aim here is twofold .
first we wish to interpret and understand the experimental findings for catco@xmath2 and srtco@xmath2 , which can be only captured by beyond - local density functional theory ( dft ) , due to the incorrect treatment of the residual exchange - correlation effects still present in 4@xmath3 compounds @xcite .
secondly we anticipate the experiments predicting the properties of the technetiates series end member batco@xmath2 .
we have employed the heyd , scuseria and ernzerhof ( hse)@xcite scheme as implemented in the vasp code @xcite , using a 4@xmath144@xmath144 k - points set , a cutoff energy of 300 ev , a mixing parameter of @xmath15=0.1@xcite and standard structural optimization conditions@xcite .
structure of the distorted perovskite @xmath1tco@xmath2 .
the tc atoms [ medium ( light - blue ) spheres ] are surrounded by o@xmath6 octahedran delimited by full ( red ) lines .
the @xmath1 atoms are represented by big ( dark green ) spheres . both the conventional ( @xmath16 ) and cartesian ( @xmath17 ) coordinate systems are shown . in
( a ) a schematic representation ( dashed lines ) of the distorted tc - o subnetwork indicating the tc - o1-tc and tc - o2-tc superexchange paths is given , where o1 and o2 are the apical and planar oxygen atom , respectively . ]
@xmath1tco@xmath2 compounds adopt the same distorted @xmath18 perovskite structure as lamno@xmath2 @xcite , characterized by a gdfeo@xmath2-like ( gfo ) tilting of the tco@xmath6 octahedra caused by the @xmath1-o and @xmath1-tc covalencies , and a concomitant jahn - teller ( jt ) tc - o bond length disproportionation , with long ( l ) and short ( s ) tc - o2 in - plane distances and medium ( m ) tc - o1 vertical ones ( see fig . [ fig : fig1 ] and table [ tab:1 ] ) .
when compared to lamno@xmath2 , the experimental values of the jt modes @xmath19 and @xmath20 in catco@xmath2 and srtco@xmath2 ( see table [ tab:1 ] ) are significantly smaller due to the intrinsically different electronic configuration of the transition metal ions . in lamno@xmath2
the mn@xmath21 ions 3@xmath3 states are in a @xmath22 configuration and a strong jt effect is required to release the degeneracy of the @xmath23 orbitals and to form an orbitally ordered state @xcite .
in contrast in @xmath1tco@xmath2 the tc@xmath24 ions ( 4@xmath0 : @xmath25 ) have completely filled ( empty ) @xmath26(@xmath23 ) manifolds , similarly to mn@xmath24 ions in camno@xmath2@xcite , a configuration which inhibits the occurence of electronic and structural jt instabilities , and leads to an insulating g - type antiferromagnetic ( afm - g ) ground state@xcite . however , at variance with @xmath27 transition metal ( tm ) ions in perovskites , where @xmath0 configurations unavoidably lead to fully polarized @xmath28= 3@xmath29 moments on tm ( e.g. camno@xmath2 ) , here due to the much larger extension of the @xmath4 electrons and in turns a much stronger hybridizations with o @xmath12 states , a sizable amount of charge fills the @xmath4 minority states as well , with a consequential reduction of magnetic moment and an overtly evident intra - atomic hund s rule violation . .
comparison between the low temperature ( 3 k ) @xcite experimental and calculated structural parameters for the g - type oriented @xmath1tco@xmath2 family .
the corresponding cooperative jahn - teller local modes @xmath19 and @xmath20 are also reported . in the definition @xmath30 , @xmath31 and @xmath28
indicate the three inequivalent long ( tc - o2@xmath32 ) , short ( tc - o2@xmath33 ) and medium ( tc - o1 ) tc - o distances , respectively .
room temperature experimental data for catco@xmath2 and srtco@xmath2 are from refs . and , respectively .
no experimental data are available for batco@xmath2 . as a reference we report here the relevant experimental structural data for the prototypical jt / gfo compound lamno@xmath2 : q@xmath34=0.398 , q@xmath2=-0.142 , @xmath35=154.3 , @xmath36=156.7 , mn - o2@xmath32=2.184 , mn - o2@xmath33=1.903 and mn - o@xmath37=1.957 @xcite . [
cols="<,^,^,^,^,^ " , ] [ tab:2 ] by mapping the hse total energies for different magnetic configurations onto a heisenberg hamiltonian ( for unitary spins ) we have evaluated the nearest - neighbor ( nn , @xmath38 ) and next - nearest - neighbor ( nnn , @xmath39 ) magnetic exchange parameters as a function of the ionic radius @xmath40@xcite . the results are summarized in fig .
[ tab:2 ] .
the nn superexchange ( se ) tc - o - tc path , as illustrated in fig .
[ fig : fig1 ] , connects nn tc atoms via planar ( tc - o2-tc ) and apical ( tc - o1-tc ) oxygen atoms . because of the almost identical tc - o distances and @xmath41 angles ( see table [ tab:1 ] ) we have found that the in - plane ( @xmath42 ) and out - of - plane ( @xmath43 ) values of @xmath38 are equal , both in sign ( thus establishing the afm - g ordering ) and magnitude ( within 10@xmath44 mev ) . therefore we have finally adopted an effective hamiltonian including an isotropic @xmath38 magnetic interaction .
in addition to the ferromagnetic ( fm ) alignment three additional afm configurations have been considered in fitting the @xmath45 s : afm - a [ ( 100 ) planes of like spins alternating along the @xmath46 axis ] , afm - c ( antiferromagnetic order in the @xmath47 plane with a ferromagnetic order along @xmath48 ) and afm - g ( all spins are antiferromagnetically coupled to their nearest neighbors ) . at variance with what found with conventional local dft functionals @xcite , hse provides very stable magnetic solutions for every spin ordering with almost identical values of the magnetic moment of tc ( @xmath49 2 @xmath50 , see table [ tab:2 ] ) .
furthermore the afm - g phase turns out to be the most stable ( see table [ tab:2 ] ) .
the heisenberg hamiltonian is then used to evaluate @xmath11 via mc calculations @xcite .
note that in doing that we have rescaled the @xmath45 s by a factor @xmath51 ( s=3/2 ) , which accounts for quantum fluctuations@xcite .
the results are summarized in table [ tab:2 ] .
the calculated @xmath11 for catco@xmath2 ( 750 k ) and srtco@xmath2 ( 1135 k ) are in very good agreement with experiments and the @xmath11 predicted for batco@xmath2 ( 1218 k ) , if experimentally confirmed , would represent the highest magnetic ordering temperature for any system without 3@xmath3 states .
tco@xmath2 series in the g - type afm ground state decomposed over majority and minority o @xmath12 and tc t@xmath52 and e@xmath53 states .
the calculated band gaps are : 1.4 ev ( catco@xmath2 ) , 1.48 ev ( srtco@xmath2 ) and 1.57 ev ( batco@xmath2 ) . ]
these unexpectedly large magnetic ordering temperatures can be understood in the framework of anderson s theory of se interactions @xcite , which links the strength of the se coupling constant with the actual hybridization between the metal and the mediating atom , and van vleck s theory of antiferromagnetism @xcite , which connects the strength of the se interaction with the magnetic ordering temperature .
the calculated inter - atomic nn and nnn coupling constants listed in table [ tab:2 ] show that the antiferromagnetic nn @xmath38 is the dominating parameter ( @xmath54 -30 mev ) and that it is almost two orders of magnitude larger than @xmath39 ( @xmath54 -0.5 mev ) .
the calculated density of states ( dos ) displayed in fig .
[ fig : fig2 ] shows that these huge @xmath38 value arises from the strong covalency between the tc @xmath26 and o @xmath12 orbitals evolving along the wide 4@xmath3 @xmath26 manifold , in particular for the topmost valence states spreading from -1.5/2 ev to the fermi level ( @xmath55 ) .
the increasing bandwidth ( @xmath56 ) of this group of hybridized bands observed when going from catco@xmath2 ( @xmath57 ev ) to srtco@xmath2 and batco@xmath2 ( @xmath58 ev ) associated with the enhanced @xmath26-@xmath12 hybridization in the 3 ev wide @xmath26 band around -5.5 ev explains the larger @xmath38 and the corresponding larger @xmath11 for srtco@xmath2 and batco@xmath2 . at this point
a fundamental question naturally arises : _ why the tc - o hybridization increases along the @xmath1tco@xmath2 series when @xmath40 get larger ? _ and @xmath38 ) and structural [ volume and q=@xmath59(q@xmath37+q@xmath34 ) ] quantities on the @xmath41 angle ( average between @xmath60 and @xmath61 ) in the @xmath1tco@xmath2 series . ] in order to answer to this question we analyse the coupling between the structural changes , caused by the substitution of ca with bigger isovalent atoms ( sr and ba ) , and the electronic and magnetic properties , as exemplified in fig .
[ fig : fig3 ] .
we recall , that the most significant effects on the crystal structure caused by increasing @xmath40 are ( i ) the volume enhancement , ( ii ) the quenching of the jt distortions q@xmath34 and q@xmath2 and ( iii ) the decrease of the cooperative rotation of the tco@xmath6 octahedra represented by the @xmath41 bond angles .
the angles @xmath60 and @xmath61 are crucial quantities to explain the evolution of the se interactions and of the magnetic ordering temperature .
in fact , the monotonic increase of the @xmath41 bond angles ( in brief the average between @xmath60 and @xmath61 , see fig .
[ fig : fig3 ] ) leads to the progressive rectification of the nn superexchange paths .
this generates , in a tight - binding framework , an enhanced tc-@xmath26/o-@xmath12 hybridization , as confirmed by the dos ( see fig . [
fig : fig2 ] ) . as displayed in fig .
[ fig : fig3 ] , @xmath11 steeply increases from catco@xmath2 ( 750 k ) to srtco@xmath2 ( 1135 k ) as a consequence of the observed larger change in @xmath41 , which goes from 151@xmath62 to 167@xmath62 . when moving from srtco@xmath2 to batco@xmath2 ( 1218 k )
the rise of @xmath11 is weaker due to a smaller change of @xmath41 ( from 167@xmath62 to 179@xmath62 ) and to a further reduction and sign change in @xmath39 . in analogy with the rmno3 perovskites ,
the increase of @xmath11 for larger tc - o - tc angles correlates with a progressive reduction of the jt dis- tortions ( i.e. a decrease of the associated structural ordering tempera- ture).@xcite in conclusion we remark that our findings for the 4@xmath3 @xmath1tco@xmath2 series are consistent with the goodenough - kanamori rules @xcite and follow the general trend observed in 3@xmath3 @xmath1mno@xmath2 , namely that @xmath11 increases by increasing the @xmath41 angles .
the remarkably different @xmath11 s in manganites ( @xmath63 150 k ) and technetiates ( @xmath64 750 k ) can be explained with the reduced spatial extension of 3@xmath3 shell , which suppresses the @xmath3-@xmath12 hybridization in manganites and thereby inhibits strong se interactions .
we also emphasize , that the correct description of this complex class of materials can be only achieved within a beyond - local functional method such as hse .
hse is capable of capturing the delicate balance between the se mechanism and both , the hund s coupling ( unlike local functionals , hse provides well defined magnetic solutions for different spin configurations ) and the intra - atomic coulomb repulsion ( though substantially reduced with respect to 3@xmath3 perovskites , on site coulomb interaction survives in 4@xmath3 technetiates and contributes to the formation of a rather large gap which local functionals seriously underestimates ) . |
over the last several years the torsional oscillator ( to ) technique has become a popular method @xcite to study solid @xmath0he .
the measurements , which involve monitoring the resonant frequency _ f _ of a high quality factor torsion pendulum filled with @xmath0he , have revealed a `` transition '' below an onset temperature _ t@xmath2 _ of @xmath1200 mk .
the results of several control experiments @xcite indicate that the sudden increase in the frequency is due to a superfluid - like decoupling of a fraction of the @xmath0he mass .
the apparent nonclassical rotational inertia fraction ( ncrif ) , proportional to the frequency shift @xmath3__f _ _ , is independent @xcite of the maximum oscillation speed _ _ v__@xmath4 of the to provided that it is smaller than a critical value . above this critical velocity _ v@xmath5 _ the magnitude of ncrif is attenuated . the value of _
v@xmath5 _ corresponds to several quanta of circulation @xmath6 , suggesting that the important excitations in the system are vortices .
anderson has proposed @xcite a vortex liquid model that qualitatively captures a number of experimental results . in multiple solid @xmath0he samples slowly grown within an annulus of radius , _ r _ = 5 mm , and width @xcite , _ t _ = 0.95 mm , a value of _ v@xmath5 _
= 10 @xmath7m@xmath8s@xmath9 was found @xcite ( corresponding to @xmath6 = 2@xmath10__rv__@xmath4 @xmath11 3__h__/_m _ , where _ h _ is planck s constant and _
m _ is the bare @xmath0he mass ) .
ncrif decreased almost linearly with ln[__v__@xmath4 ] as the speed was increased further .
this attenuation , according to ref . , is attributable to the nonlinear susceptibility of a vortex liquid phase , which consists of an entanglement of many thermally activated vortices @xcite . the ability of the vortices to move counter to the time - dependent superflow ( relative to the cell s oscillation ) results in the screening of supercurrents . as the temperature is lowered the motion and number of vortices
is reduced so that ncrif becomes finite .
the accompanying dissipation @xcite that peaks in the middle of the transition is due to a matching of _ f _ and the optimal rate at which vortices respond to changes in the velocity field .
residual @xmath13he atoms may @xcite cling to vortices , thereby slowing vortex motion and ultimately enhancing the onset temperature @xcite .
one prediction of anderson s model @xcite is that _ t@xmath2 _ , and perhaps ncrif , for a particular sample will decrease as the measurement frequency is lowered .
performed @xcite measurements in a double oscillator that operated in either an antisymmetric ( high _ f _ ) or a symmetric ( low _ f _ ) mode and found that for the same @xmath0he sample , _
t@xmath2 _ @xmath11 240 mk at 1173 hz and _ t@xmath2 _ @xmath11 160 mk at 496 hz . however , ncrif became independent of frequency below @xmath135 mk .
in addition , irreversible changes in ncrif below @xmath145 mk occurred upon variation of the oscillation speed between 10 @xmath7m@xmath8s@xmath9 and 800 @xmath7m@xmath8s@xmath9 .
nonsuperfluid explanations of the to experiments have also been proposed , in which some sort of freezing process @xcite or glass transition @xcite takes place .
although the former was only discussed qualitatively , the pinning of dislocation lines within solid @xmath0he is one possible realization of such a `` freezing '' transition .
the mechanism for dislocation pinning is likely related to interactions with isotopic impurities .
it is known that dislocations @xcite and @xmath13he atoms @xcite are mobile in solid @xmath0he below @xmath11 k , but the strong coupling between them in the low temperature limit can immobilize them both @xcite . in a to study @xcite of many solid @xmath0he samples we found that the dependence of _ t@xmath2 _ on the @xmath13he concentration _ _
x__@xmath14 is quantitatively consistent with this impurity - pinning model .
recently day and beamish observed @xcite a dramatic increase ( between 5% and 20% ) in the shear modulus _ _ c__@xmath15 of solid @xmath0he ( _ _ x__@xmath14 @xmath11 0.3 ppm ) below @xmath1250 mk , with the temperature dependence of _ _ c__@xmath15 strongly resembling that of _ f _ in to experiments .
the authors also reported hysteresis in _ _ c__@xmath15 upon adjusting the stress amplitude of the measurement at _ t _ @xmath11 20 mk , similar to what is seen in ref . .
the increase in the shear modulus is very likely due to the stiffening of the dislocation network via the immobilization of individual lines .
it was suggested by the authors of ref . that an increase in _
_ c__@xmath15 will result in a stronger coupling between the @xmath0he sample and its container and therefore mimic an enhanced mass loading of a to , causing _
f _ to decrease .
however , any enhancement to the overall rigidity of the system actually leads to a higher resonant frequency .
we have carried out a finite element method ( fem ) calculation @xcite employing realistic parameters of the to and solid helium in order to confirm this result .
the positive correlation between _ _ c__@xmath15 and _ f _ , their resemblances in the temperature dependence , and the hysteresis common to both , all suggest that either the frequency shift is a direct consequence of the enhanced modulus and can be explained without invoking supersolidity or that the microscopic mechanism responsible for ncri also affects the elastic properties of the solid ( e.g. , a recent suggestion is given in ref . ) . the results presented below appear to favor the latter .
temperature dependence of @xmath3__f _ _ for a bc sample before and after annealing , and one ct sample .
for all samples _ _
x__@xmath14 @xmath11 1 ppb .
data was obtained with the beryllium - copper to from ref . .
freezing ( _ t@xmath16 _ ) and annealing ( _ t@xmath17 _ ) temperatures are given for each sample .
all data was taken during cooling scans in which the temperature was successively lowered in controlled steps .
for any one sample the measured frequency at _
t _ @xmath18 _
t@xmath2 _ is identical for all values of _ _ v__@xmath4 .
for easy comparison we have set @xmath3__f _ _ = 0 at _ t _ = 175 mk by subtracting 1071.007 hz , 1071.011 hz , 1071.035 hz , and 1071.898 hz from the data of bc-12/10/06 , bc-12/10/06 post - annealing , ct-010507 , and the empty to .
thus , the effective mass loading for each of these samples is 891 mhz , 887 mhz , and 863 mhz . ]
[ sec : data ] we present our experimental observations in a general manner by simply describing the dependencies of the observed frequency shifts and dissipation signals on temperature ( see figs .
[ fig : one ] and [ fig : two ] ) , oscillation speed @xcite ( see fig . [ fig : three ] ) , and growth method .
two different interpretations of the data are given in secs .
[ sec : ncri ] and [ sec : c44 ] . in the former we attempt to describe the experiments in the context of a supersolid that is permeated by vortices .
since vortex lines interact with velocity fields one of the important parameters in this model is _ _ v__@xmath4 . in sec .
[ sec : c44 ] we discuss the consequences of attributing the entire frequency shift to an increase in the shear modulus of solid @xmath0he .
it is assumed in this picture that the rigidity of the solid is strongly affected by the motion of dislocation lines under the applied stress @xmath19 . in anticipation of these two interpretations
, we have simultaneously plotted the data in fig .
[ fig : three ] as a function of _ _ v__@xmath4 and @xmath19 .
in addition , in both figs .
[ fig : four ] and [ fig : five ] we have plotted the frequency shift in terms of ncrif and @xmath3__c__@xmath15/__c__@xmath15 .
temperature dependence of @xmath3__f@xmath20 _ _ and @xmath3__q@xmath20__@xmath9 for the same samples presented in fig .
[ fig : one ] , obtained by subtracting the temperature dependence of the empty cell . the symbols for bc-12/10/06 , pre- and post - annealing , are identical to those in fig .
[ fig : one ] . ]
a recent study @xcite of samples grown at constant temperature and pressure ( ct / cp ) at a single point on the solid - liquid coexistence curve found the onset of the frequency shift to be sharp and reproducible , which is consistent with the expected formation of single crystals .
in contrast , wide variations in @xmath3__f _ _ and _ t@xmath2 _ were observed when employing the blocked capillary ( bc ) method , which typically yields polycrystalline samples or highly strained crystals due to the pressure drop during freezing . in this paper
we examine in detail the different thermal histories of @xmath3__f _ _ in bc and ct / cp samples that follow from changes in oscillation speed above and below _ t@xmath2_. we have measured the temperature dependence of @xmath3__f _ _ in 30 samples , with the bulk of our work being carried out on `` isotopically pure '' ( _ _ x__@xmath14 @xmath11 1 ppb of @xmath13he ) samples at 1 @xmath7m@xmath8s@xmath9 @xmath21 _ _ v__@xmath4 @xmath21 100 @xmath7m@xmath8s@xmath9 . to overlap with ref .
we also studied commercially pure ( _ _ x__@xmath14 @xmath11 0.3 ppm ) samples up to speeds of 880 @xmath7m@xmath8s@xmath9 .
figure [ fig : one ] shows the temperature dependence of @xmath3__f _ _ for a bc sample before and after annealing , compared with that for a sample grown at ct .
we find that the temperature dependence of @xmath3__f _
_ for each sample is reproducible in warming and cooling scans at low oscillation speeds , i.e. , on the order of one micron per second .
although the differences between bc and ct / cp samples are apparent at low speeds , the temperature dependence obtained at _ _
v__@xmath4 @xmath18 50 @xmath7m@xmath8s@xmath9 is very similar in all samples studied ( including 0.3 ppm samples ) . after subtracting the temperature dependence of the empty cell
we obtain the frequency shift and dissipation due to solid @xmath0he , which we respectively denote by @xmath3__f@xmath20 _ _ and @xmath3__q@xmath20__@xmath9 .
these results are plotted in fig .
[ fig : two ] .
there is considerable dissipation in rapidly grown bc samples even at the lowest temperatures .
after annealing sample bc-12/10/06 the high temperature tail of @xmath3__f@xmath20 _ _ ( and thus _ t@xmath2
_ ) , the low temperature limiting value of @xmath3__f@xmath20 _ _ , and the magnitude of @xmath3__q@xmath20__@xmath9 are all reduced .
thus it appears that annealing affects the temperature dependence of @xmath3__f@xmath20 _ _ much in the same way as driving the to at high speed ( see fig . [ fig : two ] ) .
one difference , however , is that the results from annealing are permanent . in fig .
[ fig : three ] we have plotted the low temperature value of @xmath3__f@xmath20 _ _ measured at different _
_ v__@xmath4 , which again demonstrates that the data converge at high speeds .
one interesting difference from previous to studies @xcite is the lack of saturation in @xmath3__f@xmath20 _ _ even well below 10 @xmath7m@xmath8s@xmath9 .
this is most evident in two bc samples ( 01/25/07 and 12/10/06 ) grown from the normal fluid phase , where @xmath3__f@xmath20 _ _ actually doubles when _ _ v__@xmath4 is decreased from 10 @xmath7m@xmath8s@xmath9 to 1 @xmath7m@xmath8s@xmath9 . in samples assumed to be of higher quality , we find _
@xmath11 3 @xmath7m@xmath8s@xmath9 ( 3.5 @xmath7m@xmath8s@xmath9 corresponds to @xmath6 = _ h_/_m _ ) .
dependence of the low temperature value of @xmath3__f@xmath20 _ _ on oscillation speed and applied stress .
the inertial stress is approximated by @xmath19 @xmath11 @xmath22__fv__@xmath4/2 , for a given helium density @xmath23 .
values of _
f@xmath20 _ were extracted from cooling scans that began well above _
t@xmath2_. samples are labeled according to growth method and date .
the data are plotted versus a linear abscissa in the inset . ] the curves in fig .
[ fig : three ] were compiled from cooling scans that began well above the onset temperature . using this procedure we found that @xmath3__f@xmath20 _ _ is diminished at higher _ _
v__@xmath4 for all _ t _ @xmath21 _ t@xmath2 _ in a completely reproducible manner , whereas measurements taken during warming are history - dependent when obtained at speeds greater than a few microns per second .
as we will show below , varying the oscillation speed below the onset temperature can result in irreversible changes in @xmath3__f@xmath20__. careful examination of how the oscillation speed affects @xmath3__f@xmath20 _ _ reveals metastability at the lowest temperatures , as depicted in figs .
[ fig : four ] and [ fig : five ] ( see secs . [
sec : ncri ] and [ sec : c44 ] ) .
using the following protocol , the thermal history of @xmath3__f@xmath20 _ _ was investigated in crystals grown under different conditions .
first , the temperature dependence of @xmath3__f@xmath20 _ _ was measured while cooling at low speed ( _ _ v__@xmath4 @xmath21 3 @xmath7m@xmath8s@xmath9 ) . at _ t _ @xmath11 20 mk _ _ v__@xmath4 was slowly increased .
final velocities were usually @xmath120 @xmath7m@xmath8s@xmath9 , but in some cases much greater ( e.g. , several hundred microns per second ) .
multiple thermal cycles were then performed in succession , indicated by the arrows in fig .
[ fig : four ] .
complete equilibration was purposely avoided while cycling between 30 mk and 60 mk in order to observe multiple metastable states .
the measurement culminated with a high velocity cooling trace that began well above _
t@xmath2_. figures [ fig : four ] and [ fig : five ] show that for one oscillation speed a number of different values of @xmath3__f@xmath20 _ _ ( @xmath24 ncrif and @xmath3__c__@xmath15/__c__@xmath15 ) can be `` frozen in '' below a characteristic temperature , _ t _ * @xmath11 30 mk .
this is true regardless of the sample growth procedure .
however , there are significant differences in the time evolution ( following temperature steps ) and the velocity dependence of @xmath3__f@xmath20 _ _ between samples formed by the bc method and crystals grown at ct / cp .
similar behavior is seen in 0.3 ppm samples , but with _
t _ * @xmath11 45 mk .
for all samples , there is no observable decay from any of the metastable states of @xmath3__f@xmath20 _ _ on a timescale of days as long as the temperature is maintained below _
t_*. further discussion of the hysteresis is deferred to secs .
[ sec : ncri ] and [ sec : c44 ] below .
as stated in sec . [ sec : intro ] , the to experiments to date are qualitatively consistent with many features of anderson s vortex liquid model . the present work ,
however , demonstrates that below 60 mk the system does not act like a free vortex liquid as originally proposed .
in fact , below 30 mk it exhibits severe pinning . with this picture in mind
we can explain the temperature dependence of ncrif ( @xmath24 @xmath3__f@xmath20 _ _ ) in fig .
[ fig : two ] , the velocity dependence in fig .
[ fig : three ] , and the history dependence in figs .
[ fig : four ] and [ fig : five ] .
first we consider the temperature dependence . as the system is cooled through the transition some number of free vortices
are produced , which is proportional to _
_ v__@xmath4 .
t _ @xmath25 _
t@xmath2 _ the mobility of the existing vortices diminishes and it is unfavorable to create additional vortex lines .
the inability of vortices to screen supercurrents leads to a sizeable ncrif in the low temperature limit . when the system is driven at high speeds , not only
is ncrif smaller but the high temperature tail of ncri is reduced .
this suggests that vortex pinning is very weak at _ t _ @xmath26 60 mk .
it is also likely that vortices are more easily excited as _ t@xmath2 _ is approached . the velocity dependence in fig .
[ fig : three ] reflects the distribution in the strength of individual vortex pinning sites at low temperature .
annealing might effectively remove the weakest pinning centers so that the low temperature limiting value of ncrif is only decreased at small _ _ v__@xmath4 . at greater speeds these weak regions are irrelevant since there is a plethora of free vortices , hence the convergence in figs .
[ fig : two ] and [ fig : three ] at _ _ v__@xmath4 @xmath18 50 @xmath7m@xmath8s@xmath9 . in the context of superconductivity or magnetism ( or superfluidity ) , the system is prepared differently if ( velocity ) field - cooling or zero field - cooling are employed .
the hysteresis shown in figs .
[ fig : four ] and [ fig : five ] is a consequence of what the authors of ref .
have called vortex glass behavior .
if the state of the system is prepared at low _
_ v__@xmath4 there is only a `` small '' number of vortices frozen into the solid .
_ v__@xmath4 is gradually increased while _
t _ @xmath21
_ t _ * , ncrif becomes either unstable ( see fig . [
fig : four ] ) or metastable ( see fig .
[ fig : five ] ) . in the former case
, ncrif decays until it reaches a metastable state . for _ t _ @xmath18 _ t _ * ,
ncrif is reduced by the enhanced mobility of vortices .
metastability of @xmath3__f@xmath20 _ _ ( plotted as ncrif and @xmath3__c__@xmath15/__c__@xmath15 ) for a bc sample .
the values of @xmath3__c__@xmath15/__c__@xmath15 were calculated based on the assumption that the solid sample was isotropic , i.e. , highly polycrystalline @xcite due to bc method that was employed .
even larger values would be obtained if the sample was assumed to be a single crystal .
decay of the signal occurs for _ t _
@xmath18 30 mk .
cycling above and below _ t _ * = 30 mk results in successively smaller values until equilibrium ( i.e. , that obtained when cooling from above _ t@xmath2 _ ) is reached .
the speeds of _ _
v__@xmath4 = 1.3 @xmath7m@xmath8s@xmath9 and 21 @xmath7m@xmath8s@xmath9 correspond to @xmath19 = 2.0 mpa and 32 mpa . ]
metastability of ncrif and @xmath3__c__@xmath15/__c__@xmath15 for a cp sample .
the sample was assumed to be a single crystal with its c - axis oriented horizontally .
if we instead assume that the c - axis is oriented vertically ( or that the sample is isotropic ) the low temperature limiting value of @xmath3__c__@xmath15/__c__@xmath15 would be 60% ( 30% ) .
labels a through i denote changes in the decay rate at several different temperatures .
the speeds of _ _
v__@xmath4 = 2.2 @xmath7m@xmath8s@xmath9 and 22 @xmath7m@xmath8s@xmath9 correspond to @xmath19 = 3.4 mpa and 34 mpa . ]
the low temperature metastability seems to depend on the difference between the initial ncrif prepared at low speed and the `` equilibrium '' value obtained upon cooling at the higher speed .
for instance , the difference is ncrif = 0.9% for bc-12/10/06 ( see fig .
[ fig : four ] ) and ncrif = 0.08% for cp-11/21/06 ( see fig . [
fig : five ] ) , resulting in a much more stable value for the latter .
the measured value for bc-12/10/06 immediately drops by about 40% of its magnitude following the increase in _ _
v__@xmath4 , while there is no change for cp-11/21/06 .
the behavior of cp-11/21/06 is similar to that in ref . , which was compared with the meissner effect ( i.e. , a robust ncrif due to the inability of vortices to enter into the solid ) .
we note that the metastability up to 800 @xmath7m@xmath8s@xmath9 in ref .
may be related to the small ncrif @xmath11 0.1% of the sample . in comparison , upon raising _ _
v__@xmath4 to 880 @xmath7m@xmath8s@xmath9 we observe , for a 0.3 ppm sample with ncrif @xmath18 1% ( @xmath3__f@xmath20 _ _ @xmath18 8 mhz ) , an abrupt drop in the signal that is followed by a gradual decay .
in fact , in this extreme limit of _ _
v__@xmath4 , the decay rate is essentially insensitive to temperature steps ( warming and cooling ) between 20 mk and 45 mk . only upon warming above 60
mk does the decay rate increase further . in 1 ppb samples ,
the stability is reduced as the temperature is raised above 30 mk .
fragments of data from scans of cp-11/21/06 are displayed in fig .
[ fig : six ] . the data at any _
t _ @xmath21 60 mk can not be fit satisfactorily using a simple mathematical formula .
for example , the decay at 40 mk in fig .
[ fig : six](a ) ( between points a and b ) begins immediately , but slows down abruptly after only 1 h. in fig .
[ fig : six](b ) ,
ncrif barely changes in the first hour at the same temperature ( data prior to point g ) .
irregular rates are even more apparent at 50 mk .
the decay slows down dramatically when ncrif crosses ( e.g. , points e and h ) from the region above the low velocity cooling trace to the region below it .
this is most obvious in the fig .
[ fig : six](b ) , where the sample was warmed to 50 mk rapidly to result in a larger initial ncrif .
the very different decay rates in these two regions , and the identical temperature dependence in warming and cooling scans , both suggest that there is physical significance to the reproducible data obtained in the low velocity limit .
similar behavior could not be observed in bc samples because of higher decay rates , i.e. , it was impossible to exceed the low velocity trace upon warming ( see fig .
[ fig : four ] ) .
interestingly , the decay in cp-11/21/06 at 60 mk ( data following point i ) is different from that at 40 mk and 50 mk in that it can be well fit to an exponential form , which yields a time constant of 2.25 h. the smooth decay across the low velocity trace may indicate that the small tail of ncrif for _ t _ @xmath26 60 mk does not denote the same `` boundary '' as seen at lower temperatures .
this observation is complemented by the enhanced decay rate at 60 mk that we find in 0.3 ppm samples as well as the _
_ x__@xmath14-independent specific heat peak that has been observed @xcite in the same vicinity , and suggests that there is some inherent transition between the `` free '' and `` pinned '' regimes .
segments of raw data used to compile the `` cycling '' and `` warming '' curves in fig .
[ fig : five ] .
decay rates above ( d to e and g to h ) the low velocity trace are greater than those below ( e to f and h to i ) .
the ordinate axis is given in terms of ncrif for easy comparison with fig .
[ fig : five ] . ]
.[tab : table1]approximate stress amplitudes corresponding to _ v@xmath5 _ in various cells .
values are estimated by ignoring the effects from the top and bottom surfaces of the cell and equating two different expressions for the torque : _ _ i__@xmath27 = _ _ ra__@xmath19 . here
, @xmath27 is the angular acceleration and _ a _ is the area at the rim of the cell .
the stress exerted in an annular geometry differs from that in a cylindrical one by the approximate factor , ( 2__tf__)@xmath28/(_rf_)@xmath29 . [
cols="^,^,^,^",options="header " , ] another intriguing result is the nearly equal spacing between each metastable ncrif among the four samples that we studied in detail ( e.g. , see figs .
[ fig : four ] and [ fig : five ] ) .
it is clear that the system prefers to possess specific values of ncrif in the low temperature limit .
for example , counter to the slow decay upon warming a solid sample , as the temperature is lowered we find ncrif to `` jump '' up to a steady value ( see the data in fig . [ fig : six ] following point f ) .
if we naively estimate the frequency shift due to the presence of a single vortex line using the expression @xcite , _ i@xmath30
_ = @xmath22__r__@xmath0ncrif / ln[_r_/__a__@xmath4 ] , we calculate a shift corresponding to 25% of the observed ncrif for a vortex core radius @xcite of _ _
a__@xmath4 = 0.1 nm .
this value is of the same order of magnitude to what we observe ( e.g. , from fig .
[ fig : five ] we get 0.04/0.33 = 11% for cp-11/21/07 ) .
although our experimental results can be qualitatively interpreted within the framework of anderson s vortex liquid model , very few quantitative comparisons are possible @xcite .
one of the present experimental results that is difficult to understand in the vortex picture is depicted in fig .
[ fig : three ] . in two samples
the observed frequency shift continues to increase when _
_ v__@xmath4 is reduced to 1 @xmath7m@xmath8s@xmath9 , such that even the peak to peak amplitude of 2 @xmath7m@xmath8s@xmath9 is less than that which corresponds to a single quantum of circulation ( @xmath13.5 @xmath7m@xmath8s@xmath9 ) .
due to the similarities discussed in sec .
[ sec : intro ] between the observed frequency shifts and the results of ref .
, we are compelled to interpret the data from sec .
[ sec : data ] in terms of the response of the dislocation network to oscillating stress fields imposed on the solid . here
we report our preliminary estimates of changes in the shear modulus of solid @xmath0he . by calculating @xcite the resonant frequency of the torsion mode for different _ _ c__@xmath15 values we extracted the percentage change in the modulus that corresponds to a small frequency shift of the to .
the change is given by @xmath3__c__@xmath15/__c__@xmath15 = ( _ _ d__ln_f_/__d__ln__c__@xmath15)@xmath9(@xmath3__f@xmath20__/_f _ ) , where the derivative term comes from the fem calculation and frequency shift comes directly from our to measurements .
the results of our calculations are shown in figs .
[ fig : four ] and [ fig : five ] .
since the low temperature change in the resonant frequency is very small ( parts per million ) , it scales linearly with the change in the shear modulus . as a result the inferred temperature dependence of @xmath3__c__@xmath15/__c__@xmath15 is identical to @xmath3__f@xmath20__.
it is therefore natural that the qualitative features of the data in ref .
be very similar to that from to studies . however , the magnitude of the relative change in _ _
c__@xmath15 appears to be different . in order to account for the observed @xmath3__f@xmath20 _ _ in to measurements the required @xmath3__c__@xmath15/__c__@xmath15 values among the samples studied are between two and five times larger than those seen in ref . .
moreover , the changes are greater than the theoretical upper limit of 30% that can result from the pinning of dislocation lines @xcite .
we have carried out similar calculations for some of the to s in the literature @xcite and found that even larger values of @xmath3__c__@xmath15/__c__@xmath15 are necessary to account for the observed @xmath3__f@xmath20__. for instance , typical ncrif values in refs .
all translate into an increase in _
_ c__@xmath15 by approximately a factor of 10 . for the double oscillator from ref .
_ c__@xmath15 to increase by factors of 5 and 10 in the anti - symmetric and symmetric modes , respectively .
this implies that the effect is very dependent on frequency , contrary to the findings of day and beamish .
furthermore , it is difficult to apply the dislocation pinning picture to the experiments in porous vycor glass @xcite , in which the dimensions of helium ( i.e. , the pore size ) are smaller than typical dislocation loop lengths in bulk crystals @xcite .
this , and the fact that the crystalline phase @xcite of the confined @xmath0he is body - centered - cubic rather than hexagonal - close - packed , would lead one to expect very different behavior in the response to torsional oscillations , counter to what is observed . day and beamish found the measured shear modulus to be a function of stress amplitude . below a critical stress @xmath19__@xmath5 _ _ the magnitude of _ _ c__@xmath15 saturates at a maximum value , as does @xmath3__f@xmath20 _ _ in some samples ( see fig .
[ fig : three ] ) .
this critical value , @xmath19__@xmath5 _ _ = 300 mpa , is smaller than the estimated breakaway stress of 4 pa for a @xmath13he-@xmath13he separation of 5 @xmath7 m ( a typical loop length @xcite ) along a dislocation , but closely matches the stress at which @xmath3__f@xmath20 _ _ tends to zero in fig .
[ fig : three ] .
such strong nonlinearity between the drive and response , also seen in other low frequency acoustic measurements @xcite , occurs at `` critical '' speeds that are 100 to 1000 times less ( e.g. , @xmath150 nm@xmath8s@xmath9 in ref . )
than in to studies @xcite .
it was concluded in ref .
that @xmath19__@xmath5 _ _ is a more fundamental quantity than _
v@xmath5 _ due to it being independent of the measurement frequency .
however , this is inconsistent with ref . and
implies that the two types of measurements , although clearly related on some level , are distinct .
we have studied a number of solid @xmath0he samples grown within a torsional oscillator at constant temperature and pressure , as well as with the blocked capillary method , and found that the resonant frequency of the system possesses a very strong thermal history .
the magnitude of the apparent ncrif measured in the low temperature limit is reproducible when obtained upon cooling the sample from temperatures well above the transition .
modulation of the oscillation speed below the onset temperature reveals the existence of many metastable states .
we find that these properties are qualitatively consistent with two different interpretations : the response of vortices to velocity fields and/or the response of dislocation lines within the solid to oscillating stress fields .
the mobility of either of these entities becomes very limited with decreasing temperature , resulting in hysteretic behavior below 60 mk .
extreme pinning takes place below 30 mk ( 1 ppb ) or 45 mk ( 0.3 ppm ) , such that their motion is essentially frozen out . at present there
remain discrepancies between the experiments and either of the two interpretations . in the vortex liquid picture ,
the most notable problem is the very low critical velocity found in this study . in regard to the elastic properties of solid @xmath0he
, we find that the majority of the frequency shifts observed are appreciably larger than upper limit expected from the pinning of dislocations .
+ we thank p. w. anderson , j. r. beamish , w. f. brinkman , d. a. huse , j. jain , h. kojima , x. lin , l. pollet , n. v. prokofev , a. s. c. rittner , j. toner , and j. t. west for their input .
we are grateful to j. a. lipa for providing us with isotopically pure @xmath0he .
funding was provided by the usa nsf under grants dmr-0207071 and dmr-0706339 .
e. kim and m. h. w. chan , nature ( london ) * 427 * , 225 ( 2004 ) .
e. kim and m. h. w. , science * 305 * , 1941 ( 2004 ) .
e. kim and m. h. w. chan , j. low temp . phys . * 138 * , 859 ( 2005 ) .
e. kim and m. h. w. chan , phys .
lett . * 97 * , 115302 ( 2006 ) .
a. s. c. rittner and j. d. reppy , phys .
* 97 * , 165301 ( 2006 ) .
a. s. c. rittner and j. d. reppy , phys .
lett . * 98 * , 175302 ( 2007 ) .
y. aoki , j. c. graves , and h. kojima , phys .
lett . * 99 * , 015301 ( 2007 ) .
a. penzev , y. yasuta , and m. kubota , j. low temp . phys . * 148 * , 677 ( 2007 )
. m. kondo , s. takada , y. shibayama , and k. shirahama , j. low temp .
148 * , 695 ( 2007 ) .
a. c. clark , j. t. west , and m. h. w. chan , phys .
99 * , 135302 ( 2007 ) .
e. kim _ et al .
_ , to be published in phys .
( 2008 ) ; arxiv:0710.3370v1 ( 2007 ) .
a. penzev , y. yasuta , and m. kubota , arxiv:0711.0212v2 ( 2007 ) .
p. w. anderson , nature phys .
* 3 * , 160 ( 2007 ) .
we note an error in refs . and , i.e. , _ t _ = 0.95 mm rather than 0.6 mm .
@xmath13he atoms are attracted to vortices in superfluid liquid @xmath0he due to the lower local density in the vortex core .
for this reason it has been suggested that vortices in solid @xmath0he are nucleated between lattice sites in regions of lower density .
for instance , see w. m. saslow , phys .
b * 71 * , 092502 ( 2005 ) . repulsion of @xmath13he atoms from the vortex core is expected if the local density is higher than the average value .
this might in fact be applicable to solid @xmath0he if the supersolid phase is due to a finite concentration of vacancies .
for example , see c. wu , h. chen , j. hu , and s. c. zhang , phys .
a * 69 * , 043609 ( 2004 ) .
z. nussinov , a. v. balatsky , m. j. graf , and s. a. trugman , phys .
b * 76 * , 014530 ( 2007 ) .
a. v. balatsky , m. j. graf , z. nussinov , and s. a. trugman , phys .
b * 75 * , 094201 ( 2007 ) .
we note that the particulars of the theory are not quite reconciled by the observed frequency dependence of @xmath3__f__. for example , based on the glass parameters obtained from the symmetric mode of oscillation in ref . , the predicted frequency shift of the antisymmetric mode is approximately 10 times smaller than the experimental value .
r. wanner , i. iwasa , and s. wales , solid state commun .
* 18 * , 853 ( 1976 ) .
lin , x. , a. c. clark , m. h. w. chan , nature ( london ) * 449 * , 1025 ( 2007 ) .
m. g. richards , j. pope , and a. widom , phys . rev .
* 29 * , 708 ( 1972 ) ; v. n. grigorev _ et al .
_ , j. low temp
* 13 * , 65 ( 1973 ) ; a. r. allen , m. g. richards , and j. schratter , _ ibid . _
* 47 * , 2890 ( 1982 ) .
i. iwasa and h. suzuki , j. phys .
jpn . * 49 * , 1722 ( 1980 ) .
m. a. paalanen , d. j. bishop , and h. w. dail , phys .
46 * , 664 ( 1981 ) . j. day and j. r. beamish , arxiv:0709.4666v1 ( 2007 ) .
an eigenfrequency analysis of the to was carried out using the structural mechanics module of comsol multiphysics .
the entire to ( the base , torsion rod , and torsion bob both with and without @xmath0he ) , was drawn in a computer - aided design ( cad ) environment , from which a mesh of finite , tetrahedral elements was generated . to perform calculations in a convenient period of time the number of elements
was typically limited to less than 30,000 .
however , meshes with higher densities were also tested in several cases to ensure the accuracy of the resonant frequency . based only on the dimensions of the to and the elastic properties of its constituents we found _ f
_ to be in excellent agreement with the experimental value ( i.e. , within 1% ) .
similarly , high accuracy was obtained for the calculated mass loadings of a number of samples ( within 2% ) .
further details of our presently ongoing fem study will be published elsewhere .
j. p. bouchand and g. biroli , arxiv:0710.3087v1 ( 2007 ) .
we discuss the `` velocity dependence '' in a general sense , in that it is simply proportional to the drive level input to the system ( as is amplitude , acceleration , stress , etc . ) .
j. e. vos _
_ , physica * 37 * , 51 ( 1967 ) . g. b. hess , phys . rev .
* 161 * , 189 ( 1967 ) . due to the logarithmic dependence ,
the magnitude of the vortex angular momentum is fairly insenstive to _ _ a__@xmath4 ( e.g. , taking _ _
a__@xmath4 = 100 nm would reduce the calculated frequency shift by less than a factor of 2 ) .
a very recent attempt at an in depth analysis has been made in ref . . if we plot our data in the same manner as fig . 3 in ref .
, we observe very different behavior . at present
we do not understand this discrepancy .
however , it is interesting to note that for such plots we see qualitative differences between ct / cp and bc samples .
we note that paalanen _
@xcite previously reported a 40% change in the shear modulus , but this included high temperature data near the melting point where there are likely other stiffening processes to account for ( e.g. , the presence or absence of thermally activated vacancies ) .
calculations were also done for a to now being used by rittner and reppy .
the construction is similar to that used in ref . , but with a slightly simpler design that is easier to accurately model .
i. iwasa , k. araki , and h. suzuki , j. phys .
. jpn . * 46 * , 1119 ( 1979 ) .
d. wallacher , d. , m. rheinstaedter , t. hansen , and k. knorr , j. low temp . phys .
* 138 * , 1013 ( 2005 ) .
mukharsky , o. avenel , and e. varoquaux , j. low temp . phys . *
148 * , 689 ( 2007 ) . | we have studied the thermal history of the resonant frequency of a torsional oscillator containing solid @xmath0he .
we find that the magnitude of the frequency shift that occurs below @xmath1100 mk is multivalued in the low temperature limit , with the exact value depending on how the state is prepared .
this result can be qualitatively explained in terms of the motion and pinning of quantized vortices within the sample .
several aspects of the data are also consistent with the response of dislocation lines to oscillating stress fields imposed on the solid . |
1 the binder s cumulant @xmath35 ( see eq.([binder ] ) ) vs. temperature @xmath83 for different sizes @xmath33 . at the critical temperature @xmath35 does not depend on @xmath33 for large @xmath33 .
3 the binder s cumulant ( see eq.([binder ] ) ) as a function of @xmath84 for two values of the parameter that controls the disorder @xmath20 . in fig.3a )
the data for @xmath73 are reported having choosen @xmath85 and @xmath8 . in fig.3b ) the data for @xmath50 are reported having choosen @xmath86 and @xmath87 .
temperature is in units of @xmath88 .
4 the scaled helicity modulus @xmath89 as a function of @xmath90 for two values of the parameter that controls the disorder @xmath20 . in fig.4a ) the data for @xmath73 and @xmath91 ( from below to the top )
are reported .
the best fit is given by the full curve that corresponds to eq.([scaling ] ) with @xmath92 and @xmath78 . in fig.4b )
the data for @xmath93 and @xmath94 ( from below to the top ) are reported .
the best fit is given by the full curve that corresponds to eq.([scaling ] ) with @xmath95 and @xmath78 . | the critical behaviour of the fully frustrated xy model in presence of weak positional disorder is studied in a square lattice by monte carlo methods . the critical exponent associated to the divergence of the chiral correlation length
is found to be @xmath0 already at very small values of disorder .
furthermore the helicity modulus jump is found larger than the universal value expected in the xy model .
the study of the fully frustrated xy ( ffxy ) model@xcite in a planar lattices has attracted much attention since it provides a simple model for many physical systems .
in particular it is related to josephson - junction arrays and superconducting wire networks in a magnetic field @xcite for which recent experiments show interesting properties@xcite . in these systems
a perpendicular magnetic field @xmath1 penetrates the array inducing a quenched vortex density .
when the field is such that each plaquette of area @xmath2 in the planar lattice traps half a vortex , e.g. the ratio of the magnetic field to the flux unit @xmath3 , the system can be described by the ffxy model . as first noted by villain@xcite the ffxy model posses , besides the continous @xmath4 symmetry , a discrete @xmath5 symmetry which is not present in the unfrutrated xy model .
the existence of two different symmetries leads to the possibility of two critical temperatures that have been investigated by both analytical methods@xcite and monte carlo ( mc ) simulations@xcite .
however , in spite of accurate mc simulations , the scenario is not fully understood .
in fact , there are numerical evidences supporting both the existence of two very close critical temperatures ( @xmath6 ) with critical behavior typical of ising and thouless - kosterlitz - berezinski ( tkb ) transitions respectively@xcite and the existence of a single transition with novel critical behavior@xcite . in the latter case
it has been claimed that the transition associated to the @xmath5 symmetry is characterized by a correlation length critical exponent @xmath7 lower than that expected in the ising case .
recent experiments on nb wire networks@xcite in an external magnetic field have shown that the critical exponent @xmath7 which can be extracted by the scaling of current - voltage ( i - v ) characteristics , is in agreament neither with a ising - like exponent ( @xmath8 ) nor with the value suggested in ref.[@xcite ] ( @xmath9 ) .
on the contrary , the scaling of the experimental data suggests a much larger value . the aim of the present work is to investigate if the effect of weak quenched disorder in the ffxy model is able to change the critical behaviour of the system and to explain the unexpected critical behaviour .
the hamiltonian of the ffxy model is @xmath10 where @xmath11 is the phase in the site @xmath12 and @xmath13 is the line integral of the vector potential along the line ( @xmath14 ) and @xmath15 is the flux quantum . in order to simulate positional disorder
we consider a uniform distribution of the @xmath16 whose average value @xmath17 is equal to @xmath18 every second column and zero otherwise .
the disorder configuration is thus specified by the probability distribution @xmath19 the parameter @xmath20 controls the degree of randomness and is related to the variance according to @xmath21 . of course , the presence of positional disorder effects the behaviour of the model .
in fact it has been shown by renormalization group techniques@xcite and monte carlo simulations@xcite that the critical temperatures associated to both @xmath5 and @xmath4 simmetries decrease with increasing disorder .
it has been also suggested for the @xmath4 transition a reentrant behaviour , even if recent work on the disordered unfrustrated xy model seems to exclude this possibility@xcite .
however , the critical behaviour that characterize the ising - like and the tkb transitions have not been studied to our knowledge . in particular , the study of the critical exponent @xmath7 associated to the divergence of the coherence lenght of the ising - like transitions is interesting in view of its possible connection with the @xmath22 characteristics reported by ling@xcite .
in fact the experimental data i - v , or equivalently the data j - e , scale very well according to the relation@xcite @xmath23 where the dynamical critical exponent @xmath24 is found to be about 2 and @xmath25 . this type of scaling is consistent with the reasonable assumption that the relevant excitations that contribute to the critical behaviour of the i - v characteristic are domain - wall . in order to obtain an estimation of @xmath7 we have calculated the binder s cumulant of the staggered magnetization @xmath26 @xmath27 where @xmath26 , associated to the chirality of each plaquette , is defined as @xmath28 in eq .
( [ m ] ) the chirality @xmath29 is the rotation of the current around a plaquette @xmath30 where @xmath31 is the phase difference at the edges of the plaquette .
binder s cumulant allows an accurate estimation of the critical temperature associated to the ordering of the plaquette chirality ( ising - like transition ) .
in fact @xmath32 does not depend on lattice size @xmath33 for large systems and , then , @xmath34 can be identified without making any assumption on the critical exponents .
once a satisfactory estimation of @xmath34 is obtained the critical exponent @xmath7 is estimated through a data collapsing with the parameter @xmath7 only .
the binder s cumulant @xmath35 has been calculated by monte carlo ( mc ) simulations using a standard metropolis dynamics . for each disorder configuration the mean value
@xmath36 has been obtained averaging over @xmath37 mc sweeps after discarding the first @xmath38 for equilibrating the system .
then the mean values are further averaged over a number of disorder configurations that range from @xmath39 to @xmath40 depending on lattice size and amount of disorder present in the system . in fig .
1 ) it is shown how the critical temperature @xmath41 can be extracted by using the mc estimations of @xmath35 for different lattice sizes . in the case reported in the figure ,
if we exclude the smallest lattice size , @xmath42 , all the data cross about @xmath43 . from this type of analysis
it is , then , possible to follow @xmath41 as a function of the disorder ( fig.2 ) ) .
the results confirm that @xmath41 decrease with increasing disorder going from @xmath44 for the case without disorder ( @xmath45 ) to @xmath46 for the case with largest disorder studied ( @xmath47 ) .
the results are in qualitative agreement with those reported by choi et al.@xcite . for larger values of the disorder preliminar results
seems to suggest a continous collaps of @xmath41 . in figs .
3a ) and 3b ) we show binder s cumulant @xmath35 as a function of @xmath48 for two values of @xmath20 . by using the estimated values of @xmath41 at @xmath49 and @xmath50 we see that the best data collapsing is obtained for @xmath51 in the case of lower disorder and for @xmath52 in the other case .
even if the numerical errors do not allow a very accurate estimation of @xmath7 it is very clear that already for very weak disorder ( @xmath53 ) the critical exponent is quite different from that expected in the ising model ( @xmath8 ) and from that suggested by granato et al.@xcite in the case without disorder .
instead the value obtained is very close to that extracted from the @xmath22 characteristics@xcite .
it is also interesting to note that a very similar value is obtained in the 2d xy spin glass that has a critical temperature @xmath54.@xcite the same analysis for @xmath45 and @xmath55 gives clearly @xmath56 and @xmath0 , respectively . on the other hand for the intermediate value @xmath57
the data are not able to clearly exclude either @xmath56 or @xmath0 ; larger lattices and more mc sweeps are needed to reach a satisfactory estimation of @xmath7 .
these results indicate the existence of a threshold disorder above which @xmath7 changes from 1 to 1.7 . the value of the threshold can be , then , estimated to be @xmath58 but it must be understood as an upper limit . as in the ffxy model
we expect the system to present a jump in the helicity modulus at temperature lower or equal to @xmath41 . in order to study this possibility
we have calculated the helicity modulus that for the system studied takes the form @xmath59 where @xmath60 and @xmath61 are the unitary vectors in the directions of the link @xmath62 and @xmath63 , respectively .
the symbol @xmath64 indicates the average over the spin configurations .
the elicity modulus @xmath65 has to be averaged further over the disorder configurations to obtain the mean value @xmath66 .
following the analysis developed for the xy model@xcite and extended to the ffxy model@xcite we can estimate the critical temperature @xmath67 by using the following ansatz for the size dependence of @xmath66 @xmath68 where @xmath69 corresponds to the universal jump expected at the critical temperature in a ktb transition .
this critical scaling , being based on the mapping between a neutral coulomb gas and a xy model , is valid only if the transition is of ktb type .
therefore eq .
( [ scaling ] ) can be used as a test for the existence of a standard ktb transition . in figs .
4a ) and 4b ) we show the results of this analysis for two different values of the disorder . in both cases
it is not possible to obtain a satisfactory fit of the data by using eq .
( [ scaling ] ) with @xmath70 .
however , a more accurate data analysis shows that a satisfying fitting of the data can be obtained assuming a helicity modulus jump larger than the universal value .
from such an analysis we get @xmath71 and @xmath72 for @xmath73 and @xmath74 and @xmath75 for @xmath76 in both cases the fitting parameter @xmath77takes the same value found in the ffxy model without disorder , @xmath78,@xcite and @xmath67 is slightly smaller than @xmath41 even if the estimated numerical error does not rule out @xmath79 in the case of lowest disorder ( @xmath73).@xcite this analysis shows that the only way to fit the data with eq.([scaling ] ) is to assume a larger helicity modulus jump and this is an indication that the transition associated to this jump is not a ktb transition .
summarizing we have calculated , by mc simulations , the critical temperature and the critical exponent @xmath7 associated to the chiral degree of freedom in a disordered ffxy model .
unexpected we find that the critical exponent @xmath80 already for very weak disorder becomes equal to the value found in the xy spin glass , i.e. @xmath81 correspondently , the helicity modulus jump becomes larger than the universal value expected in ktb transition .
finally , we note that the value obtained for @xmath7 is in agreement with recent results in josephson wire networks.@xcite j. villain , j. phys .
c * 10 * , 4793 ( 1977 ) t.c .
halsey , j. phys .
c * 18 * , 2437 ( 1985 ) for a review see : physica b * 222 * , 253 ( 1996 ) x.s .
ling , et al .
phys .
rev .
lett . * 76 * , 2989 ( 1996 ) e. granato and m.p .
nightingale , phys .
rev .
b * 48 * , 7438 ( 1993 ) s. teitel and c. jayaprakash , phys . rev .
b * 27 * , 598 ( 1983 ) ; phys . rev
. lett . * 51 * , 199 ( 1983 ) jooyoung lee , j.m .
kosterlitz and e. granato , phys .
rev .
b * 43 * , 11531 ( 1991 ) g. ramirez - santiago and j.v .
jos , phys .
rev .
b * 49 * , 9567 ( 1994 ) p. olsson , phys . rev . lett . *
75 * , 2758 ( 1995 ) g. grest , phys .
rev .
b * 39 * , 9267 ( 1989 ) p. olsson , phys . rev .
b * 55 * , 3585 ( 1997 ) e. granato and j.m .
kosterlitz , phys .
rev .
lett . * 62 * , 823 ( 1989 ) ; phys .
rev .
b * 33 * , 6533 ( 1986 ) v. cataudella and m. nicodemi , physica a * 233 * , 293 ( 1996 ) m.y . choi , j.s .
chung , and d. stroud , phys .
rev .
b * 35 , * 1669 ( 1987 ) stefan scheidl phys . rev .
b * 55 * , 457 ( 1997 ) ; m.c .
cha and h.a .
fertig , phys .
rev .
lett .
74 , 4867(1995 ) d.s .
fisher , m.p.a .
fisher and d. huse , phys .
rev .
b*43 * , 130 ( 1991 ) h.s .
bokil and a.p .
young , j. phys .
a * 29 * , l89 ( 1996 ) h. kawamura and m. tanemura , phys .
rev .
b * 36 * , 7177 ( 1987 ) p. ray and m.a . moore , phys . rev .
b * 45 * , 5361 ( 1992 ) h. weber and p. minnhagen , phys .
rev .
b * 37 * , 5986 ( 1988 ) the estimated values of @xmath67 and @xmath82 show that increasing the degree of disorder in the system the difference betwwen the two critical temperatures increases .
this is in agreement with the results reported by granato and kosterlitz @xcite |
recently we identified the two most unusual aspects of our solar system ( compared to all the observed exoplanetary systems ) to be , first , the lack of super - earths ( planets with a mass in the range @xmath1 , e.g. * ? ? ?
* ) and , secondly , the lack of planets inside of mercury s orbit @xcite .
more than half of the observed sun - like stars in the solar neighborhood have one or more super earth planets on low eccentricity orbits with periods of days to months @xcite .
this is in contrast to our solar system which is depleted in mass in this region . to examine potential reasons for this difference , in this work we consider formation processes for super earths and investigate the conditions in the solar nebula that could have affected the outcome for our solar system .
the timeline for the formation of the planets in our solar system is thought to have been as follows .
the planetesimals formed within a few million years of the birth of the sun @xcite .
the giant planets formed quickly allowing for the accretion of material from the gas disk ( e.g. * ? ? ? * ) .
however , the terrestrial planets formed long after the gas disk was dispersed , on a timescale of around @xmath2 million years ( e.g. * ? ? ? * ) .
there are no super earths in our solar system so understanding their formation on the basis of solar system data alone is more difficult . the observed close in super
earths exhibit a wide range of densities ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) suggesting that there may be several different mechanisms for their formation ( see also figure 5 in * ? ? ?
* ) . in fig . [ density ]
we show the planet mass and semi major axis of observed exoplanets that have a density measurement ( the data are taken from exoplanets.org , * ? ? ?
the largest points show the low density planets ( planets with densities similar to the giant planets in our solar system , density @xmath3 ) , the medium size points have density in the range @xmath4 and the small points show the high density planets ( with density greater than @xmath5 , similar to the terrestrial planets in our solar system ) .
there appears to be little correlation between the density and the planet mass .
there is a slight correlation between the density and the semi major axis in that there are fewer low density planets close to the star .
the compositions of extrasolar super - earths suggest that at least some of them have substantial gaseous atmospheres .
this is at odds with the timescale for terrestrial planet formation and therefore ( at least some ) super - earths likely formed while the gas disk was still present .
the composition of the planets is dependent on their formation location .
planets that form outside of the water snow line ( the radial location from the star where the temperature is low enough for water to become solid , e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) will be water
rich while those that form close to their star will likely be more rocky .
we expect that planets that form farther out in the disk may be less dense than those that form close to the star .
however , accretion of gaseous material from the disk may significantly reduce the final average density of the planet and thus the time of planet formation is also important .
that have a density measurement .
the large ( blue ) points denote planets with a low density @xmath6 , the medium size ( magenta ) points denote those with a density in the range @xmath7 and the small ( red ) points denote those with a high density , @xmath8 . for reference , in our solar system the terrestrial planets have average densities @xmath9 and the giant planets have average densities @xmath10 .
data are from exoplanets.org.,width=317 ] there are two general orbital locations suggested for the formation of super - earths .
either they formed _ in situ _ with no significant migration , or else they formed similarly to giant planets , outside the snow line in the protoplanetary disk and then migrated inwards .
giant planets with short orbital periods are thought to have formed farther out and migrated inwards to their current locations ( e.g. * ? ? ? * ) .
however , hot jupiters are only found around about 1% of stars
( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
typically their orbits are highly misaligned , suggesting that they were driven to these locations through secular perturbations or planet planet scattering ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
on the other hand , super earths could have an occurrence rate as high as 50% or more @xcite .
thus , the hot jupiters probably migrated to their observed location while super earths , that are far more abundant , may be more likely to have formed _ in situ_. we discuss both possibilities here . @xcite constructed a minimum - mass extrasolar nebula ( mmen ) from observations of super earth exoplanets with orbital periods @xmath11days .
this is a circumstellar disk of solar composition that allows _ in situ _ formation of close in super earths .
they found a surface density for the gas disk of @xmath12 where @xmath13 is the radial distance from the central star .
this is somewhat higher than the minimum mass solar nebular ( mmsn ) , required to form the planets in our solar system , which is given by @xmath14 @xcite .
however , the mmsn is not necessarily applicable inside of mercury s orbit at @xmath15 .
@xcite used n body simulations to form super - earth planets _ in situ_. they found it was possible if the amount of rocky material interior to @xmath16 is about @xmath17 .
they suggested that this would require significant radial migration of solid material before the end stages of planet formation .
the minimum mass solar nebular @xcite has only @xmath18 in solids interior to @xmath16 , although the planet formation process is not likely to be completely efficient .
the ratio of dust to gas is often quoted to be 0.01 in the ism . however , in accretion disks observations indicate that it may be much higher ( e.g. * ? ? ?
the mass in the gas disk in this region would need to be in the range @xmath19 .
an advantage of the _ in situ _ formation model is that it can explain several properties of the observed planet inclination and eccentricity distributions as well as the orbital spacing @xcite .
the alternative super earth formation theory suggests that they form farther out in the disk where there is more solid material and then migrate inwards @xcite . in this scenario ,
planets grow through earth size embryo embryo collisions .
smooth migration means that the planets form in chains of mean motion resonances .
however , multiplanet systems observed with kepler generally do not show planets locked into resonances ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) and thus the resonances must be broken after planet formation ( e.g. * ? ? ?
* ; * ? ? ?
@xcite suggested that hot super earths and giant planet cores form in the same way .
they both migrate inwards but the giant planet cores become massive enough for the direction of the migration to reverse .
the super earths pile up at the inner edge of the disk . in this work we present a gas disk model that allows for the formation via both mechanisms in both locations , depending upon the disk properties . in sections [ turb ] and
[ dead ] we construct numerical models of protoplanetary disks without and with a dead zone ( a region of low turbulence ) .
this will allow us to draw some conclusions about the formation mechanisms for super earths .
in section [ solarsystem ] we discuss the implications for our own solar system and we discuss and summarize our results in sections [ discussion ] and [ conc ] .
in this section we consider the evolution of a fully turbulent protoplanetary gas disk .
the disk is in keplerian rotation with angular velocity @xmath20 around a central mass @xmath21 at radial distance @xmath13 .
turbulence , driven by the magneto rotational instability ( mri ) , drives a kinematic turbulent viscosity @xmath22 where @xmath23 is the @xcite viscosity parameter , the sound speed is @xmath24 with mid plane temperature @xmath25 , @xmath26 is the gas constant and @xmath27 is the mean molecular weight .
the evolution of the surface density , @xmath28 , is governed by @xmath29 , \label{sigma}\ ] ] ( e.g. * ? ? ?
* ; * ? ? ?
* ) where @xmath30 in this fully mri turbulent disk .
the temperature evolves according to the simplified energy equation @xmath31 @xcite .
the disk specific heat for temperatures around @xmath32 is @xmath33 .
the disk is heated by viscous dissipation according to @xmath34 we assume that each annulus of the disk radiates as a black body and so the local cooling is @xmath35 where @xmath36 is the temperature at the surface of the disk and @xmath37 is the stefan - boltzmann constant . assuming energy balance , the mid plane temperature and the surface temperature are related through @xmath38 where the optical depth is @xmath39 we use the simplified opacity of @xcite @xmath40 that is valid in the inner parts of the disk . at early times in the protostellar accretion history the infall accretion rate on to the disk
is expected to be around @xmath41 , ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . for temperatures around @xmath42 ,
the infall accretion rate is of the order of @xmath43 .
we first run a disk model with a constant infall accretion rate of @xmath44 until it reaches a steady state .
this is the initial disk setup that we use for all of the simulations
. the infall accretion rate on to the disk evolves with time as @xmath45 where the initial accretion rate is @xmath46 and @xmath47 ( see * ? ? ?
* ; * ? ? ?
* ) . as a function of time ( lower line ) and total disk mass up to a radius of @xmath48 ( upper line ) .
the initial disk surface density is a steady state fully turbulent disk with an infall accretion rate of @xmath49 .
the infall accretion rate decreases exponentially in time according to equation ( [ mdot ] ) .
the dashed lines show the mass in @xmath50 for the mmsn ( lower ) and the mmen ( upper ) .
, width=264 ] we integrate the disk evolution equations ( [ sigma ] ) and ( [ temp ] ) for a disk that extends from @xmath51 to @xmath52 around a solar mass star .
the grid contains @xmath53 points distributed uniformly in @xmath54 .
material is added to the disk at a radius of @xmath55 . at the inner edge of the disk
a zero torque boundary condition allows the inward flow of gas out of the grid and toward the central star .
the flow is prevented from leaving the outer boundary with a zero radial velocity outer boundary condition .
we consider a fully turbulent disk model that has a viscosity parameter @xmath56 everywhere in the disk ( e.g. * ? ? ?
figure [ fig1 ] shows the mass of the disk inside of @xmath57 ( lower line ) and the total disk mass up to @xmath58 ( upper line ) as a function of time .
the material in @xmath50 is potentially available for the formation of super - earth planets in these inner regions .
note that the amount of mass in the mmsn in this region is @xmath59 .
it seems a little surprising that the fully turbulent disk model is never above this value .
this is partially because the mmsn is not applicable inside of mercury s orbit .
the mmsn has a very steep dependence on radius and thus predicts a rather high mass in the inner regions of the disk .
similarly , the mmen also predicts a very high mass in the inner regions in order to be able to form the observed exoplanets . we can conclude that the fully turbulent disk model probably can not form the planets that are observed close to their host stars _ in situ_. while _ in situ _ formation of super
earths seems to be ruled out in the case of a fully turbulent disk , formation of super earths farther out in the disk followed by inward migration remains a possibility in the fully turbulent disk model .
there are two types of planetary migrations depending on the mass of the planet and the properties of the disk . in type
i migration the planet is not massive enough to open a gap in the disk and the surface density of the disk remains largely unperturbed by the presence of the planet .
this type of migration is not dependent on the viscosity of the disk . in a fully turbulent disk ,
earths are not large enough to open a gap in the disk and so they migrate via type i migration .
the timescale for this migration is given approximately by @xmath60 @xcite , where @xmath61 is the mass of the planet , @xmath62 is the orbital radius of the planet , @xmath21 is the mass of the star and @xmath63 is the disk aspect ratio .
we do note that this timescale is very sensitive to disk parameters and the direction of type i migration is determined by the mass of the planet .
equation ( [ type1 ] ) shows that the inward migration timescale is around a few @xmath64 and so the migration can comfortably take place within the lifetime of the disk .
the difficulty is rather how to stop the migration ( e.g. * ? ? ?
. there has been much discussion about planet traps that change the direction of type i migration ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these can occur at transitions such as snow lines , dead zone boundaries and heat transitions .
the radial location of such a trap may move slowly in time , thus transporting the planet on a much longer timescale .
thus , formation farther out followed by inward migration is possible in principle in the fully turbulent disk model .
it is generally accepted that protoplanetary disks are not fully turbulent , but rather that they contain a region of low turbulence , a dead zone ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
we therefore solve the accretion disk equations that include a dead zone region .
we define @xmath65 to be the surface density in the mri active layers and @xmath66 is the surface density in the dead zone layer . if there is no dead zone at a particular radius , then @xmath67 and @xmath68 there .
we consider two prescriptions for determining the surface density of the dead zone : 1 .
we assume that the disk surface layers are ionized by external sources ( cosmic rays or x - rays from the central star , e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) to a maximum surface density depth of @xmath69 on the upper and lower disk surfaces , where @xmath70 is constant ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
cosmic rays are thought to ionize about @xmath71 while x - rays are about an order of magnitude smaller .
the disk surface layers always contain mri turbulence .
furthermore , if the temperature is greater than a critical value , @xmath72 , then the disk is thermally ionized allowing the mri to operate throughout the vertical extent .
the value of @xmath72 is not well determined and so we consider two values of @xmath73 and @xmath74 ( e.g. * ? ? ?
* ; * ? ? ?
thus , the disk is mri active if either @xmath75 or if @xmath76 and otherwise , there is a dead zone layer at the mid plane .
the dead zone surface density is determined via a critical magnetic reynolds number ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the disk is `` dead '' if @xmath77 , where @xmath78 , where @xmath79 is the disk scale height and the ohmic resistivity is @xmath80 @xcite and @xmath81 is the electron fraction .
we use the analytic approximations for the active layer surface density shown in equations ( 27 ) and ( 28 ) in @xcite that include thermal ionization , cosmic ray ionization and the effects of recombination .
because the dead zone acts like a plug in the accretion flow , material accumulates in this region and may become massive enough to become self gravitating .
viscosity may be generated by the mri and by self
gravity @xcite .
the viscosity driven by the mri in the magnetic layers is @xmath82 where we take @xmath83 and @xmath84 is the sound speed in the magnetic layer with temperature @xmath85 .
the dead zone is assumed to have zero turbulence , unless it becomes self gravitating .
however , the inclusion of a small amount of turbulence in the dead zone ( e.g. @xmath86 ) does not affect the behaviour of the disk significantly @xcite .
as material builds up , the disk becomes self - gravitating if the toomre parameter @xmath87 , where @xmath88 and the sound speed at the disk mid - plane is given by @xmath89 .
we approximate the temperature of the self - gravitating region as the mid plane temperature , @xmath90 .
gravity drives a viscosity that is approximated by @xmath91 \label{nug}\ ] ] for @xmath92 and zero otherwise ( e.g. * ? ? ?
* ; * ? ? ?
* ) and we take @xmath93
. the surface density and temperature evolution for a disk with a dead zone is given by equations ( [ sigma ] ) and ( [ temp ] ) but here with @xmath94 . the magnetic layer temperature is related to the surface temperature with @xmath95 where the optical depth to the magnetic region is @xmath96 the optical depth within the dead zone layer is @xmath97 and @xmath98 the cooling function is found to be @xmath99 @xcite . although the disk is not in thermal equilibrium , we apply this cooling function to equation ( [ temp ] ) and so we do not attempt to treat the cooling during viscosity transitions consistently .
the initial disk setup is the steady state disk with an infall accretion rate of @xmath46 .
the infall accretion rate decreases exponentially in time according to equation ( [ mdot ] ) .
thus , the total amount of mass that is accreted on to the disk is exactly the same in these disk models as in the fully turbulent disk model described in the previous section .
the disk with a dead zone is not in a quasi steady state as the infall accretion rate decreases , as the fully turbulent disk model is .
instead , material building up becomes self gravitating . the extra heating by self
gravity can lead to the mri being triggered within this region , if the infall accretion rate is sufficiently high .
when this happens , the disk becomes mri active throughout , leading to a large amount of material accreting on to the star in a short time interval @xcite .
this is an accretion outburst that is thought to be the explanation for the observed fu orionis type outbursts ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
these outbursts occur during the initial disk evolution but at later times when the infall accretion rate has dropped there may still be a dead zone , but there are no further outbursts .
planets that survive must form after the last outburst otherwise they will likely be swept on to the central star during the outburst .
we first consider the possibility of _ in situ _ super earth formation by examining the amount of material in the inner parts of the disk available for planet formation .
we consider a dead zone defined by model 1 with @xmath100 and @xmath101 .
[ fig2 ] shows the evolution of the amount of material at @xmath50 .
the sharp increases in the amount of material are when an outburst is triggered and a large amount of material flows through the region .
we suggest that if super
earths are to form and survive within such a disk , they must form after the last accretion outburst . for this model ,
the dead zone is accreted and the disk becomes fully turbulent at a time of around @xmath102 .
the dead zone does not last the entire disk lifetime because the active layer surface density is relatively high . as material drains slowly from the dead zone
, the disk soon has @xmath103 or @xmath104 everywhere in the disk , and it is fully turbulent . at this time
, the amount of material inside of @xmath16 decreases very rapidly . consequently , with these dead zone model parameters it is unlikely that there would be enough material , for sufficient time , for super earths to form in the inner parts of the disk . except the disk has a dead zone defined by model 1 with @xmath105 and @xmath100.,width=317 ] fig .
[ fig3 ] shows a disk with a dead zone defined by model 1 with parameters @xmath100 and @xmath106 . with the lower active layer surface density
, the dead zone persists for the entire disk lifetime .
in the later stages of the disk lifetime there is sufficient material to form super earths _ in situ_. in this model , there is a final accretion outburst at a time of around @xmath107 .
however , observationally the lifetime of protoplanetary disks is only a few myr ( e.g. * ? ? ? * ; * ? ? ?
therefore , the disk will mostly be photoevaporated before this time ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
photoevaporation of the outer parts of the disk will cut off the supply of material to the inner parts .
the final accretion outburst shown in the model will most likely not take place and super earths that form while the disk mass is high will survive . for this disk model ,
in which the dead zone lasts the entire lifetime of the disk , the total mass of the disk is somewhat high , around @xmath108 at @xmath109 .
observations of disk masses are typically derived from measuring the amount of dust . assuming a gas to dust ratio of 100 , observed masses are in the range @xmath110 to @xmath108 ( e.g. * ? ? ?
thus , the disk masses predicted by this model are on the high side . however , in a disk with a quiescent dead zone , the dust will concentrate to the midplane and therefore observations may underestimate the mass of the gas disk .
the process of photoevaporation depends upon the accretion flow changing from being dominated by viscous torques , to being dominated by the wind mass loss .
this transition occurs once the accretion flow through the disk reaches a critical value which is typically in the range of @xmath111 depending upon the the dominant photon flux , x
rays , euv or fuv ( e.g. * ? ? ?
* ) . in our model ,
most of the disk mass ( 97% ) is in the dead zone at the end of the simulation and so the accretion flow rate through the disk is small .
while the total disk mass may be high , the viscous accretion rate is low , and therefore photoevaporation should be able to efficiently clear the disk . except the disk has a dead zone defined by model 1 with @xmath112 and @xmath113.,width=317 ] fig .
[ fig4 ] shows a disk with a dead zone defined by model 1 with parameters @xmath114 and @xmath106 .
the higher critical temperature allows more time for super earth formation at earlier times .
again , we suggest that the final outburst shown will not take place because the disk will be photoevaporated before this time . except the disk has a dead zone defined by model 1 with @xmath112 and @xmath115.,width=317 ] fig .
[ fig5 ] shows a disk with a dead zone determined by model 2 .
the critical magnetic reynolds number is @xmath116 . for this model , the active layer surface density changes with radius .
the inner parts of the disk are almost entirely non turbulent and a significant amount of material builds up there .
there is more than enough material for super earths to form _ in situ_. after a time of about @xmath102 the disk has more material than the mmen in @xmath50 for the rest of the disk lifetime . in our disk model ,
the inner edge of the dead zone is determined by the radial distance at which the temperature of the disk drops below that required for thermal ionisation .
it does not change in time because the temperature profile is not significantly affected by a dead zone ( see for example * ? ? ? * ) unless it becomes self gravitating . in conclusion
, we have shown that the disk mass inside of @xmath16 may be several times that of the mmen for the later stages of the disk life and thus formation of super earths in this region is possible , depending on the dead zone parameters .
we should note that we have described only a representative sample of the results obtained from a series of simulations .
from all the simulations we find that the dead zone persists long enough for the formation of super earths provided that @xmath117 . while the value of the active layer surface density is still somewhat uncertain , this value may be representative of protoplanetary disks .
still , since different protoplanetary disks may have different values , super earths may form _ in situ _ in some systems but not in others .
except the disk has a dead zone defined by model 2 with@xmath118 ( bottom right).,width=317 ] formation of super earths farther out in the disk with subsequent inward migration may be more difficult in a disk with a dead zone . while the rate of type i migration is not affected by viscosity ( or the presence of a dead zone ) , the low viscosity means that the planet more likely migrates by type ii migration ( e.g. * ? ? ?
* ; * ? ? ?
traditionally it was thought that there are two gap opening criteria that need to be satisfied in order for a planet to open a gap in the disk .
the first is the viscous gap opening criterion @xmath119 @xcite .
@xcite suggested that a second criterion , the thermal criterion , must also be satisfied for a planet to open a gap .
that is , the hill radius of the planet must be larger than the disk scale height . however , more recently , it has been found that this criterion is too strong ( e.g. * ? ? ?
@xcite find that due to the non linear wave steepening , a low mass planet can open a gap in contradiction to the thermal criterion . in this work
we assume that the gap opening criterion is determined by the viscous torque on the disk balancing the tidal torque from the planet .
rearranging equation ( [ mcrit ] ) , we find that a super earth can open a gap in the disk provided that the viscosity parameter satisfies @xmath120 that is , the dead zone does not have to be entirely turbulence
free for a super earth to open a gap .
type ii migration typically proceeds on the viscous timescale for the disk @xmath121 ( e.g. * ? ? ?
* ) . for typical parameters
this is given by @xmath122 this can be longer than the lifetime of the disk for @xmath123 .
if a super earth forms inside a dead zone , it may be trapped there in the dead region , while migrating only very slowly by type ii migration .
the disk becomes depleted in time because the infall accretion rate declines while material continues to accrete on to the star through the active surface layers .
concomitantly , the dead zone size gradually decreases . in the end stages of the disk s life ,
the disk becomes fully turbulent down to the mid plane ( this occurs when @xmath124 everywhere that @xmath125 ) and the inner parts of the disk accrete on to the star .
when the disk becomes turbulent , the planet may no longer be able to hold a gap open , and type i migration may occur .
we discuss this possibility further in the next section .
the lack of super earths in our solar system sets us somewhat apart from observed exoplanetary systems .
there are two possible explanations for this dearth of super earths .
either conditions in the solar nebula did not allow for the formation of super earths , or else , they did form but were subsequently somehow removed .
given that the orbits of the planets in the solar system are coplanar and not very eccentric , planet
planet scattering does not seem to be a likely ejection mechanism .
thus , if super earths formed , they most likely fell into the sun .
if super earths form outside of the snow line and migrate inwards through a gas disk , they affect the composition of the terrestrial planets , if they migrate slowly enough @xcite . therefore , if they formed in our solar system , they would have had to migrate quickly , on a timescale of about @xmath126 .
if the timescale is longer than this , then a super earth shepherds rocky material interior to its orbit and depletes the terrestrial planet - forming zone .
terrestrial planets that form may be volatile rich and are more likely to be water
worlds that are not very earth like . consequently , the most probable formation site in our solar system is in the inner regions , inside of mercury s orbit .
the lack of any objects in this region may indeed suggest that super
earths formed close to our sun , clearing the region of debris , but subsequently they fell into the sun . the mechanism that pushed the super earths into the sun
could be simply migration through the gas disk . at the end of the gas disk s lifetime
, photoevaporation removes the remaining gas on a short timescale ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. however , this process only operates outside of the gravitational radius from the star , typically around @xmath127 ( e.g. * ? ? ? * ) .
the inner parts of the disk that we are interested in accrete on to the central star on a viscous timescale .
the maximum total surface density present in the disk during this process is the critical active layer surface density , @xmath128 .
if the critical active layer s surface density is small , then the planets will not move far during this process . however , a large critical surface density may be sufficient for a super earth to migrate into the sun .
we therefore find that there is a delicate balance between the need for a sufficiently large surface density in the active layer for the planets to migrate into the star , but also a small enough active layer surface density to allow the planets to form _ in situ _ in the first place . for the super earth to migrate into the sun
, it must do so on a timescale shorter than the viscous timescale ( the timescale for the disk to accrete ) .
equating equations ( [ type1 ] ) and ( [ tvisc ] ) we find that the minimum surface density in the disk for the planet to migrate in to the sun is @xmath129 now , if @xmath130 then we expect that super
earths that form _ in situ _ will migrate into the sun at the end of the disk lifetime . on the other hand , if @xmath131 then there may be some type i migration , but not enough to allow the super earth to be accreted .
note that equation ( [ sigmin ] ) is very sensitive to the disk aspect ratio .
for example , if the disk aspect ratio is decreased by a factor of two , down to @xmath132 , then we find that @xmath133 .
we speculate that in our solar system , super
earths formed in the inner parts of a relatively cool disk , close to the dead zone inner boundary .
there was sufficient time for them to migrate through the disk to be accreted on to the sun . while this outcome is less likely because of the cool conditions required , it is definitely not impossible and should happen in other planetary systems also .
we note that an alternative mechanism for pushing the super
earths into the sun is the grand tack @xcite . in this model
, jupiter migrates inwards to @xmath134 before it gets locked into resonance with saturn and then they both move outwards to their current locations . during this process
the innermost super - earths get shepherded in to the sun .
our disk model which includes a dead zone also provides an alternative explanation for the cleared inner regions of our solar system and the lack of super earths .
furthermore , the small masses of mercury and mars can be explained if the terrestrial planets form from a narrow annulus of rocky debris in the orbital range @xmath135 @xcite .
our model can explain the inner truncation radius for this annulus as being where the super
earths cleared the material .
the mmen is criticized because of the unusually high amount of material required for the _ in situ _ formation of super earths .
however , sub - mm observations of disks measure the properties of the outer disk while the inner parts of the disk are not very well constrained . furthermore
, @xcite suggest that _ in situ _ formation can not operate unless type i migration is suppressed in the region inside of @xmath16 . a disk model with a dead zone ( such as the one proposed in this work ) not only provides sufficient material , but it also suppresses the rate of migration in that region .
@xcite argue that super
earths could not have formed _ in situ_. they use observations of systems that contain three or more planets to construct an mmen and find that the surface density profile @xmath136 has @xmath137 .
they suggest that because viscous accretion disk models have difficulty in reproducing such extreme profiles that the super earths could not have predominantly formed _ in situ_. these conclusions are based on a steady state disk model with the temperature profile dominated by stellar irradiation ( e.g. * ? ? ?
* ) . however , we suggest that the extreme values could be a result of a discontinuous surface density distribution brought about by the presence of a dead zone . in the dead zone material
builds up and the surface density in the innermost mri active parts of the disk may be significantly lower than that farther out in the dead zone .
if this is the case , then one can not smooth the profile from three planets out to a continuous surface density distribution .
consequently , the most important factor is the amount of material in the inner regions available for planet formation , rather than its distribution .
because planets that form _ in situ _ are expected to require an unusually large amount of solids in the inner parts of the disk , the prediction is that the metallicity of the host star must be higher in such cases .
@xcite found that close - in super - earths are more likely to be found around metal - rich stars .
similarly , the widely separated super - earths are more often around metal poor stars . as metal - rich stars have more solid material closer to the star within the disk , this appears to favour the _ in situ _ formation .
however , there may be be two populations of super earths because there is a distinct transition in orbital period observed , rather than a smooth transition .
@xcite proposed a mechanism to form the close in super earths at the pressure trap at the inner edge of a dead zone , known as inside out planet formation .
once a planet forms there , the inner edge of the dead zone moves out allowing planet formation further out @xcite .
this model requires a high rate of supply of pebbles to the inner disk .
we suggest that a dead zone model can provide all the required material for planet formation without the need for accumulation .
however , the inside out planet formation mechanism potentially increases the range of possible dead zone parameters that are able to form the super earths _ in situ_. in this work we have considered only the formation of super earth planets during the gas disk lifetime .
it is entirely possible that some super earth planets may form after the gas disk has dissipated and we expect these to have a high density , similar to the terrestrial planets in our solar system .
however , this is not the only mechanism to form high density super earths .
a super earth that forms in a dead zone with a small active layer will not accrete much material once it has carved out a gap
. the only accretion on to the planet is through the active layer .
thus , planets that form within a dead zone are likely to be less gaseous and more dense than planets that form farther out in a fully turbulent disk and migrate inwards .
this may explain the weak dependence of density on semi major axis observed in fig . 1 . however , the gaseous atmosphere of a close in super earth could be stripped from the planet , by tidal evolution or evaporation , leaving only the solid core behind ( e.g. * ? ? ?
* ; * ? ? ?
* ) . consequently ,
as we can not tell the difference between these two mechanisms in the observed exoplanets , it is difficult to make firm conclusions as to which super earths form by which mechanism
. nevertheless , the existence of these different mechanisms may explain the range in the densities .
there are two possible formation locations for observed close in super earths in exoplanetary systems : either _ in situ _ , or farther away from the star followed by migration to their observed location .
we find that a disk that contains a dead zone ( a region of low turbulence ) may have sufficient material for the planets to form _ in situ _ although it depends upon the dead zone parameters . in order for the dead zone to last long enough for super earths to form , the active layer surface density must be sufficiently low , @xmath138 .
if the active layer surface density is too large , the disk becomes fully turbulent before there is sufficient time to form the super earths .
migration of super earths through a dead zone is very slow and thus formation farther out in a disk with a dead zone is more difficult .
we find that a fully turbulent protoplanetary disk model does not have sufficient material in the inner parts of the disk for _ in situ _ formation of super earths . in this case , the only possible formation mechanism involves migration from farther out in the disk . the fast rate of migration in a fully turbulent disk lends itself to this scenario .
we suggest that the observed large range in super earth compositions may be the result of these two very different formation locations .
the lack of super earths in our solar system is somewhat puzzling given that more than half of observed exoplanetary systems contain one .
however , the fact that there is nothing inside of mercury s orbit may not be a coincidence .
_ in situ _ formation of super earths in that region could have cleared the solid material .
the super earths would have had to subsequently fall into the sun .
this is possible if the active layer surface density is sufficiently large that during the final accretion process there was enough material in the disk for the planets to migrate into the sun . in order to satisfy both this constraint , and the constraint that the dead zone must last throughout the disk lifetime ,
requires a sufficiently cool disk during the final accretion process .
the level of fine tuning required is certainly possible , but we do nt expect it to happen in all systems and this can explain why the solar system is somewhat special in its lack of super earths .
we thank an anonymous referee for useful comments . this research has made use of the exoplanet orbit database and the exoplanet data explorer at exoplanets.org . | we first consider how the level of turbulence in a protoplanetary disk affects the formation locations for the observed close in super earths in exosolar systems .
we find that a protoplanetary disk that includes a dead zone ( a region of low turbulence ) has substantially more material in the inner parts of the disk , possibly allowing for _ in situ _ formation .
for the dead zone to last the entire lifetime of the disk requires the active layer surface density to be sufficiently small , @xmath0 .
migration through a dead zone may be very slow and thus super earth formation followed by migration towards the star through the dead zone is less likely . for fully turbulent disks , there is not enough material for _ in situ _ formation .
however , in this case , super earths can form farther out in the disk and migrate inwards on a reasonable timescale .
we suggest that both of these formation mechanisms operate in different planetary systems .
this can help to explain the observed large range in densities of super earths because the formation location determines the composition .
furthermore , we speculate that super earths could have formed in the inner parts of our solar system and cleared the material in the region inside of mercury s orbit .
the super earths could migrate through the gas disk and fall into the sun if the disk was sufficiently cool during the final gas disk accretion process . while it is definitely possible to meet all of these requirements , we do nt expect them to occur in all systems , which may explain why the solar system is somewhat special in its lack of super earths .
[ firstpage ] |
ordinal categorical data appear frequently in the biomedical research literature , for example , in the analysis of @xmath0 independent binary random variables related to an increasing ordered categorical variable .
it is well - known that for such data it is not possible to use the classical test - statistics such as chi - square or likelihood ratio with chi - squared asymptotic distribution , but there exist appropriate order - restricted test - statistics with chi - squared - bar asymptotic distribution . to illustrate this problem a modification of an example given in silvapulle and sen ( 2005 )
is considered in this introductory section .
2.8pt [ c]||clc||crr||crc|crc|| & & & + & @xmath1 & & & @xmath2 & & & @xmath3 & & & @xmath4 & + & @xmath5 & & & @xmath6 & & & @xmath7 & & & @xmath8 & + & @xmath9 & & & @xmath10 & & & @xmath11 & & & @xmath12 & + & @xmath13 & & & @xmath14 & & & @xmath5 & & & @xmath15 & + @xmath16 table [ tt1 ] contains a subset of data from a prospective study of maternal drinking and congenital malformations .
women completed a questionnaire , early in their pregnancy , concerning alcohol use in the first trimester ; complete data and details are available elsewhere ( graubard and korn , 1987 ) .
specifically , women were asked what was the amount of alcohol taken during the first three months of their pregnancy and four categories of drink doses are considered ( @xmath17 ) , no alcohol consumption ( @xmath18 ) , average number of alcoholic drinks per day less than one but greater than zero ( @xmath19 ) , one or more and less than three alcoholic drinks per day ( @xmath20 ) and three or more alcoholic drinks per day ( @xmath21 ) . in terms of a binary random variable with @xmath22 individuals in total ( see the second column in table [ tt1 ] ) with independent behavior with respect to having congenital malformations , the individuals not having congenital malformations are considered to be unsuccessful ( @xmath23 , see the last column in table [ tt1 ] ) and successful otherwise ( @xmath24 , see the third column in table [ tt1 ] ) .
let @xmath25 be the probability of a success associated with the @xmath26-th alcohol dose .
let us consider some statistical inference questions that may arise in this example and in similar ones with binomial probabilities . 1 .
is there any evidence of maternal alcohol consumption being related to malformation of sex organ ? to answer this question , the null and alternative hypotheses may be formulated as @xmath27 respectively . however , this formulation is unlikely to be appropriate because the main issue of interest is the possible increase in the probability of malformation as alcohol consumption increases .
2 . is there any evidence that an increase in maternal alcohol consumption is associated with an increase in the probability of malformation ? .
this question , as it stands , is quite broad to give a precise formulation of the null and the alternative hypotheses .
one possibility is to formulate the problem in the following way,@xmath28 consider an experiment with @xmath0 increasing ordinal categories for a variable @xmath29 .
suppose that @xmath22 prefixed individuals are assigned to the @xmath26-th category and @xmath30 .
the individuals are followed over time for the development of an event of interest @xmath31 and the events related to the individuals are independent .
let @xmath32 be the random variable that represents the number of individuals related to successful events ( @xmath33 ) out of the total assigned to the @xmath26-th category , @xmath22 , @xmath34 if we denote by @xmath35 the probability of a success associated with the @xmath26-th category , we have that @xmath32 is a binomial random variable with parameters @xmath22 and @xmath25 , @xmath36 .
let @xmath37 denote the number of unsuccessful events associated with the @xmath26-th category , i.e. @xmath38 , then the contingency table of a realization of @xmath39 , @xmath36 , is in the last two columns of the following table@xmath40{||c||cc||}\hline\hline $ n_{1}$ & $ n_{11}$ & $ n_{12}=n_{1}-n_{11}$\\ $ \vdots$ & $ \vdots$ & $ \vdots$\\ $ n_{i}$ & $ n_{i1}$ & $ n_{i2}=n_{i}-n_{i1}$\\ $ \vdots$ & $ \vdots$ & $ \vdots$\\ $ n_{i}$ & $ n_{i1}$ & $ n_{i2}=n_{i}-n_{i1}$\\\hline\hline \end{tabular } \ .\ ] ] our purpose in this paper is to propose new order - restricted test statistics , wald - type and phi - divergence based test - statistics for testing@xmath41 the classical likelihood ratio test statistic will appear as a particular case of phi - divergence based test - statistics .
a log - linear formulation of ( [ j1 ] ) is proposed in section [ sec1b ] , fundamental for defining the wald type test - statistics . in section [ sec2 ]
the families of phi - divergence test statistics are presented .
section [ sec3 ] is devoted to solve the problem presented in this section [ sec1 ] for a illustrative example .
an extensive simulation study is carried out in section [ sec4 ] .
reparametrizating the initial problem through log - linear modeling , the formulation of the null hypothesis is strongly simplified since all the interaction parameters are zero under the null hypothesis and this is appealing , in special , to create wald type test - statistics .
let@xmath42 be the probability vector of the following saturated log - linear model@xmath43 with@xmath44 being the identifiability constraints,@xmath45 the unknown parameters vector and @xmath46 with@xmath47@xmath48 the redundant parameters , obtained through @xmath49 taking into account @xmath50 , @xmath36 . in terms of the log - linear formulation , ( [ j1a ] )
is equivalent to@xmath51 notice that @xmath52 is a nuisance parameter since it does not interfere in ( [ j1b ] ) .
in particular , under the null hypothesis of @xmath53 , the value of the nuisance parameter is @xmath54 $ ] , and thus it contains all the information about the homogeneous probability vector . in matrix notation
, we can express the vector of parameters of the log - linear model in terms of the following logistic regression@xmath55 where @xmath56 @xmath57 is the the identity matrix of order @xmath58 , @xmath59 is the @xmath58-vector of ones and @xmath60 is the @xmath58-vector of zeros .
since a saturated model has been considered , @xmath61 is a full rank matrix and thus we can consider@xmath62 and on the other hand ( [ j1b ] ) in matrix notation is given by@xmath63 with @xmath64 , and @xmath65 is a square matrix of order @xmath66 with @xmath1-s in the main diagonal and @xmath67-s in the upper superdiagonal .
we shall consider three parameter spaces for @xmath49@xmath68 i.e. while @xmath69 is the restricted parameter space , @xmath70 is unrestricted , and @xmath71 becomes the parameter space under the null hypothesis .
it is well known that , the fisher information matrix for @xmath72 in the logistic regression is given by @xmath73 where @xmath74 , @xmath36 .
the following result provides the explicit expression of the fisher information matrix under the null hypothesis given in ( [ j1a ] ) or ( [ j1b ] ) .
[ th0]for @xmath75 in the model ( [ loglineq ] ) or ( [ logistic ] ) , the fisher information matrix is given by@xmath76 it is immediate by plugging @xmath77 , @xmath36 , to ( [ fim1 ] ) . if @xmath78 , @xmath79 and @xmath80 represent the maximum likelihood estimator ( mle ) of @xmath49 focussed on the parameter spaces , @xmath71 , @xmath81 , @xmath69 respectively , according to silvapulle and sen ( 2005 , pages 154 and 166 ) we can consider three wald - type test - statistics,@xmath82 where@xmath83 these test - statistics , have according to proposition 4.4.1 in silvapulle and sen ( 2005 ) , the same asymptotic distribution as the likelihood ratio test - statistic . under the null hypothesis given in ( [ j1a ] ) or ( [ j1b ] ) , the expression of @xmath84 , given in ( [ w0 ] ) , is as follows@xmath85 where @xmath86 , @xmath87,@xmath88 from theorem [ th0 ] the @xmath89 block structure of the fisher information matrix is@xmath90 and@xmath91 with @xmath92 , @xmath93 .
therefore,@xmath96 there is an explicit formula for the mles of @xmath49 under the null hypothesis,@xmath97 for the calculation of @xmath98 or @xmath99 , it is much easier to calculate first the corresponding mle for the probability vector , @xmath100 or @xmath101 , and plugging it to ( [ loglin ] ) .
there is an explicit formula for@xmath102 and for calculating @xmath103 the following pava algorithm can be used .
[ order restricted estimation of probabilities][algorp]the mle of @xmath104 under the restriction of @xmath105 , @xmath106 , is calculated in the following way : ` step 1 : do ` @xmath107 ` , where ` @xmath108 is ( [ pibar])`.``step 2 : while not ` @xmath109 ` ` @xmath110 ` do``for ` @xmath111`if ` @xmath112 ` do ` @xmath113 ` and ` @xmath114 .
the classical order - restricted likelihood ratio test for testing ( [ j1a ] ) is given by@xmath115 ( see for instance mancuso et al ( 2001 ) ) .
the kullback - leibler divergence measure between two @xmath116-dimensional probability vectors @xmath117 and @xmath118 , is given by @xmath119 it is an easy exercise to verify that @xmath120 where@xmath121 the classical order - restricted chi - square test statistic for testing ( [ j1a ] ) , known as bartholomew s test - statistic , is given by @xmath122 which can be written as @xmath123 where @xmath124 is the pearson divergence measure defined by@xmath125 details about this test - statistic can be found in fleiss et al .
( 2003 , section 9.3 ) .
more general than the kullback - leibler divergence and pearson divergence measures are @xmath126-divergence measures , defined as @xmath127 where @xmath128 is a convex function such that @xmath129 , @xmath130 , @xmath131 , @xmath132 , for @xmath133 . for more details about @xmath126-divergence measures
see pardo ( 2006 ) . based on @xmath126-divergence measures we shall consider in this paper two families of order - restricted @xmath126-divergence test statistics valid for testing ( [ j1a ] ) or ( [ j1b ] ) .
the first one generalizes the order - restricted likelihood ratio test given in ( [ g ] ) in the sense that we replace on it the kullback - leibler divergence measure by a phi - divergence measure and its expression is @xmath134 for @xmath135 , we get the likelihood ratio test .
the second one generalizes the order - restricted pearson test statistic given in ( [ j3 ] ) in the sense that we replace on it the pearson divergence measure by a phi - divergence measure and its expression is @xmath136 for @xmath137 , we get the pearson test - statistics .
the following theorem provides the link between the both test - statistics , @xmath138 and @xmath139 , and wald - type test - statistics .
[ th1]for testing ( [ j1a ] ) or ( [ j1b ] ) , the asymptotic distribution of @xmath140 is common and is given by@xmath141 where@xmath142 and @xmath143 is the set of weights such that @xmath144 and its values are given in theorem [ prop1 ] .
let @xmath78 be the @xmath0-dimensional vector given in ( [ thetahat ] ) .
the second order taylor expansion of function @xmath145 about @xmath78 is@xmath146 where@xmath147 and @xmath148 let @xmath149 , where @xmath108 is ( [ pibar ] ) .
in particular , for @xmath150 we have@xmath151 and for @xmath152@xmath153 and then taking into account that @xmath154,@xmath155 in a similar way it is obtained@xmath156 and then taking into account that @xmath154,@xmath157 according to proposition 4.4.1 in silvapulle and sen ( 2005)@xmath158 which means that the asymptotic distribution of @xmath140 is common .
such a distribution can be established from the likelihood ratio test - statistic used for the problem formulated in ( 6.13 ) of silvapulle and sen ( 2005)@xmath159 where@xmath160 with@xmath161 using the partitioned structure of @xmath162 and @xmath163 , ( [ a ] ) is obtained .
the following result is based on the third way for computation of weights given in page 79 of silvapulle and sen ( 2005 ) .
[ prop1]the set of weights @xmath143 of the asymptotic distribution given in theorem [ th1 ] is computed as follows@xmath164 @xmath165 , @xmath166 and @xmath167 such that @xmath26 components are strictly positive and @xmath168 components are null . in particular , for @xmath169 ( [ w ] ) has the explicit expressions given in ( 3.24 ) , ( 3.25 ) and ( 3.26 ) of silvapulle and sen ( 2005 ) . for computing ( [ w ] )
is useful to know the following explicit expression @xmath170 , where @xmath171 is an upper triangular matrix of @xmath1-s , @xmath172 , and to simulate the probability according to the following algorithm . for simulation @xmath173
are needed rather than @xmath174 and @xmath175 respectively .
[ estimation of weights][algorw]the estimators of the weights given in theorem [ prop1 ] , @xmath176 , are obtained by monte carlo in the following way : ` step 1 : for ` @xmath177 ` , do ` @xmath178`.``step 2 : repeat the following steps ` @xmath179 ` ( say ` @xmath180 ` ) times:``step 2.1 : generate an observation , ` @xmath181 ` , from ` @xmath182 ` .
e.g. , thenag fortran library subroutines g05cbf , g05eaf , and g05ezf can be useful.step 2.2 : compute ` @xmath183 .
` e.g. , the imsl fortran library subroutine dqprog can be useful.step 2.3 : count ` @xmath184 ` , the number of strictly positive components contained in ` @xmath185 ` , and do ` @xmath186.`step 3 : do ` @xmath187 for @xmath177 .
in this section the data set of the introduction ( table [ tt1 ] ) is analyzed .
the sample , a realization of @xmath188 , is summarized in the following vector@xmath189 the estimated vectors of interest are@xmath190 and@xmath191 for the asymptotic distribution the weighs can be calculated though@xmath192 calculating the correlation coefficients@xmath193 the partial correlation coefficients@xmath194 and evaluating the following expressions@xmath195 if we take @xmath196 , where for each @xmath197 , the power divergence family is obtained@xmath198 it is also possible to cover the real line for @xmath199 , by defining @xmath200 , for @xmath201 . it is well known that @xmath202 and @xmath203 .
this is very interesting since this means that the power divergence based family of test - statistics contain as special cases @xmath204 and @xmath205 .
finally , the expressions of the test - statistics are summarized in table [ test ] .
it can be seen that the null hypothesis can not be rejected for @xmath84 , @xmath206 , @xmath207 , @xmath208 , @xmath209 , @xmath210 , @xmath211 , @xmath212 , @xmath213 and should be rejected for @xmath214 , @xmath215 , @xmath216 , @xmath217 , @xmath218 , @xmath219 . even though the sample size seems to be large enough ,
this is a case of small values of @xmath220 , @xmath221 , @xmath222 , @xmath223 which is known to have not reliable behavior in the values calculated for the @xmath224-values in order to make decisions . in the simulation study
we shall study such a case and according to the results the rejection of the null hypothesis is supported since with @xmath225 are obtained the most realiable test - statistics . as conclussion , an increase in maternal alcohol consumption is associated with an increase in the probability of malformation
. 2.8pt @xmath226{c}\begin{tabular } [ c]{ccccccc}\hline\hline test - statistic & $ ] = -1.5@xmath227=-1@xmath227=-0.5@xmath227=0@xmath227=@[email protected]@[email protected]@[email protected]@xmath2278.4895@xmath230p@xmath231(t_)@[email protected]@[email protected]@[email protected]@[email protected]@[email protected]@[email protected]@xmath230p@xmath231(s_)@[email protected]@[email protected]@[email protected]@xmath234{cccccc}\hline\hline $ ] ( , ) @xmath2292.5979@xmath227h(,)@xmath2272.6363@xmath227d(,,)@xmath2272.6462@xmath230p@xmath231(w(,))@xmath2270.1686@xmath227p@xmath231(h(,))@xmath2270.1653@xmath229p@xmath231(d(,,))@xmath2320.1645@xmath235
for testing ( [ j1 ] ) , by considering @xmath17 binomial random variables , the following scenarios will be considered : * scenario * a * ( small / big proportions ) : @xmath236 , @xmath237 , @xmath238 , @xmath239 . *
* scenario * a-0 * : @xmath240 . * * scenario * a-1 * : @xmath241 , @xmath242 . * * scenario * a-2 * : @xmath241 , @xmath243 , @xmath244 . * * scenario * a-3 * : @xmath241 , @xmath243 , @xmath245 , @xmath246 . *
scenario * b * ( small / big proportions ) : @xmath247 , @xmath248 , @xmath249 , @xmath250 . * * scenario * b-0 * : @xmath240 . * * scenario * b-1 * : @xmath241 , @xmath242 . * * scenario * b-2 * : @xmath241 , @xmath243 , @xmath244 . * * scenario * b-3 * : @xmath241 , @xmath243 , @xmath245 , @xmath246 .
* scenario * c * ( small / big proportions ) : @xmath251 , @xmath252 , @xmath253 , @xmath254 . * * scenario * c-0 * : @xmath240 . * * scenario * c-1 * : @xmath241 , @xmath242 . * * scenario * c-2 * : @xmath241 , @xmath243 , @xmath244 . * * scenario * c-3 * : @xmath241 , @xmath243 , @xmath245 , @xmath246 . *
scenario * d * ( intermediate proportions ) : @xmath236 , @xmath237 , @xmath238 , @xmath239 . * * scenario * d-0 * : @xmath255 . * * scenario * d-1 * : @xmath256 , @xmath257 . * * scenario * d-2 * : @xmath256 , @xmath258 , @xmath259 . * * scenario * d-3 * : @xmath256 , @xmath258 , @xmath260 , @xmath261 . *
scenario * e * ( intermediate proportions ) : @xmath247 , @xmath248 , @xmath249 , @xmath250 . * * scenario * e-0 * : @xmath255 . * * scenario * e-1 * : @xmath256 , @xmath257 . * * scenario * e-2 * : @xmath256 , @xmath258 , @xmath259 . * * scenario * e-3 * : @xmath256 , @xmath258 , @xmath260 , @xmath261 . *
scenario * f * ( intermediate proportions ) : @xmath251 , @xmath252 , @xmath253 , @xmath254 . * * scenario * f-0 * : @xmath255 . * * scenario * f-1 * : @xmath256 , @xmath257 . * * scenario * f-2 * : @xmath256 , @xmath258 , @xmath259 . * * scenario * f-3 * : @xmath256 , @xmath258 , @xmath260 , @xmath261 .
the simulation experiment is performed with @xmath262 replications and in each of them , apart from the wald type test - statistics @xmath263 , @xmath264 , @xmath265 , @xmath266 , all the power divergence test statistics , @xmath267 , @xmath266 , associated with the interval @xmath268 are considered . from the p - values it is possible to calculate the proportion of replications rejected according with the nominal size @xmath269 , i.e.@xmath270 where @xmath271 represents the indicator function .
the scenarios ending in 0 represent that the null hypothesis is true , and are useful to obtain the simulated significance levels , @xmath272 , @xmath273 , with different sample sizes and kinds of test - statistics .
the scenarios ending in either 1 , 2 or 3 represent that the null hypothesis is false and are useful to obtain the simulated powers , @xmath274 , with different alternatives , sample sizes and types of test - statistics . for calculating both ,
@xmath272 and @xmath274 , ( [ prop ] ) is applied , each one in the corresponding scenario . in figures
[ fig1 ] and [ fig2 ] , @xmath272 and @xmath274 for all the aforementioned test - statistics are plotted in different scenarios .
the curves represent either @xmath275 or @xmath276 power divergence test - statistics , located respectively on left or right of the panel of plots .
the asterisk , square and circle symbols , represent @xmath84 , @xmath206 and @xmath207 respectively and all of them are repeated on the left as well as on the right in order to make easier their comparison with @xmath275 or @xmath277 respectively . the black color lines and symbols , representing @xmath272 , are useful to select the test - statistics closed to nominal level according to the criterion given by dale ( 1986 ) , a reasonable exact significance level should verify@xmath278 for @xmath279 , being closed to nominal level the exact significance levels verifying the inequality with @xmath280 and fairly closed to nominal level the ones with @xmath281 . in this study
only the test statistics satisfying the condition with @xmath280 are considered , and the corresponding upper and lower bounds appear plotted with two horizontal lines , having in the middle the line associated with the nominal level , @xmath269 . among the test - statistics with simulated significance levels closed to the nominal level , the test - statistics with higher powers should be selected but since in general high powers correspond to high significance levels , this choice is not straightforward . for this reason , based on @xmath282 or @xmath283 as baseline , the efficiencies relative to the likelihood ratio test ( @xmath284 ) @xmath285 or the bartholomew s test ( @xmath286 ) @xmath287 are considered . in table
[ table1 ] , the efficiency of @xmath288 is compared with respect to @xmath289 , and then if @xmath290 , since @xmath289 is better than @xmath288 , the plot of the efficiencies in figures [ fig3 ] and [ fig4 ] will be only focussed on ( [ ro ] ) .
similarly , if @xmath291 , since @xmath289 is worse than @xmath288 , the plot of the efficiencies in figures [ fig3 ] and [ fig4 ] will be only focussed on ( [ ro * ] ) .
[ c]ccccccc & sc a & sc b & sc c & sc d & sc e & sc f + 1 & @xmath290 & @xmath290 & @xmath290 & @xmath290 & @xmath290 & @xmath290 + 2 & @xmath290 & @xmath291 & @xmath290 & @xmath290 & @xmath290 & @xmath290 + 3 & @xmath290 & @xmath291 & @xmath290 & @xmath290 & @xmath290 & @xmath290 + in view of the plots , it is possible to propose test - statistics with better performance in comparison with @xmath204 and @xmath205 . from figures
[ fig1 ] and [ fig3 ] , the so - called cressie - read test - statistic , @xmath215 , can be recommended for small / big proportions either for small or moderate sample sizes . on the other hand , @xmath84 , @xmath206 and the test - statistic based on the hellinger distance , @xmath294 ,
can be recommended for intermediate proportions and moderate sample sizes , however for small sample sizes the likelihood ratio test - statistic still remains being the best one .
silvapulle , m. j. and sen . , p. k. ( 2005 ) .
_ constrained statistical inference . inequality , order , and shape restrictions_. wiley series in probability and statistics .
wiley - interscience ( john wiley & sons ) . | in this paper new test statistics are introduced and studied for the important problem of testing hypothesis that involves inequality constraint on proportions when the sample comes from independent binomial random variables : wald type and phi - divergence based test - statistics . as a particular case of phi - divergence
based test - statistics , the classical likelihood ratio test is considered .
an illustrative example is given and the performance of all of them for small and moderate sample sizes is analyzed in an extensive simulation study .
= 1 _ _ keywords and phrases _ _ * * : * * wald - type statistics , phi - divergence statistics , inequality constrains , loglinear model , logistic regression . |
the eigenvalue problem for @xmath1 hermitian octonionic matrices , henceforth referred to as _ jordan matrices _ , contains some surprises .
notable among these is that , whereas each jordan matrix satisfies its characteristic equation , its real eigenvalues do not .
as shown in @xcite , each jordan matrix admits six real eigenvalues , rather than three .
however , the eigenvalues divide naturally into two families of three , and the corresponding families of eigenvectors do have the expected properties , such as orthonormality , provided that these properties are properly formulated . due to the nonassociativity of the octonions
, most jordan matrices also appear to admit eigenvalues which are not real ; several examples were discussed in @xcite .
however , to our knowledge there is no known algorithm for finding the non - real eigenvalues of such matrices , nor is it clear how many there are . in this paper , which is an extension of @xcite
, we take a different approach . rather than attempt to find the non - real eigenvalues and corresponding eigenvectors of a given jordan matrix
, we instead find the jordan matrices which admit a given eigenvector / eigenvalue pair .
specifically , for any vector @xmath2 which is not quaternionic , we use the associator of the elements of @xmath3 as the eigenvalue , and find all jordan matrices for which @xmath3 is an eigenvector with that eigenvalue , which is nonzero by assumption .
we show below that a necessary condition for such matrices to exist is that @xmath4 , and that if this condition is satisfied there is a 6-parameter family of such matrices .
we begin in section [ octonions ] by reviewing the octonions and their properties , and then briefly summarize some known examples @xcite of @xmath1 hermitian octonionic matrices with imaginary eigenvalues in section [ examples ] . in section [ results ] , we present our new results , which we then summarize in section [ conclusions ] , where we also propose some further conjectures .
= 3.5 in we use the standard basis @xmath5 for the quaternions @xmath6 , and we construct the octonions @xmath7 via the cayley - dickson process as @xmath8 .
the resulting multiplication table is neatly summarized by the oriented fano geometry shown in figure [ omult ] . as is well - known , the octonions are neither commutative nor associative .
writing the components of an octonion @xmath9 as @xmath10 we have @xmath11 and @xmath12 can be assumed without loss of generality to take the form @xmath13 by suitable choice of basis ; we refer to @xmath14 as _ generic octonions_. choosing @xmath15 the associator of @xmath3 is by definition the associator of its elements , that is @xmath16 = [ x , y , z ] = ( xy)z - x(yz ) = 2 x_2 y_3 z_8 \,k\ell\ ] ] with the last equality holding for generic octonions .
we seek solutions of the ( right ) eigenvalue problem @xmath17 where @xmath18 ; in what follows we will consider only the case @xmath19 \ne 0 $ ] , noting that @xmath20 $ ] is pure imaginary .
any jordan matrix can be written in the form @xmath21 with @xmath22 and @xmath23 .
then takes the form @xmath24 as shown below , admits solutions only if @xmath4 , in which case there is a family of jordan matrices @xmath25 which satisfy it .
little is known about solutions of with non - real eigenvalues .
although the problem of finding real eigenvalues for @xmath1 hermitian octonionic matrices has been completely solved @xcite , no such solution exists for finding imaginary eigenvalues @xcite .
in fact , we know of only a handful of explicit examples of families of @xmath1 hermitian octonionic matrices admitting imaginary eigenvalues , such as those given in @xcite , which are reproduced below .
note in each case that @xmath26 can always be chosen so that the eigenvalue has no real part .
furthermore , in the first example , the eigenvectors have no real part , the associator of each eigenvector is a ( possibly zero ) multiple of @xmath27 , the imaginary direction of the eigenvalue , and this last property also holds for the elements of @xmath28 .
however , the first two properties fail to hold in the second example , while the last property fails in the third .
[ [ example-1 ] ] example 1 + + + + + + + + + the matrix @xmath29 with @xmath30 and @xmath31 has , among others , the eigenvalues and eigenvectors , @xmath32 where @xmath33 [ [ example-2 ] ] example 2 + + + + + + + + + the matrix @xmath34 has , among others , the eigenvectors and eigenvalues , @xmath35 [ [ example-3 ] ] example 3 + + + + + + + + + the matrix @xmath36 admits the eigenvector @xmath37 with eigenvalue @xmath38
as already noted , for each example in the previous section , @xmath26 can be chosen so that a given eigenvalue is purely imaginary . more generally ,
the real part of an eigenvalue can be changed by adding a suitable multiple of the identity matrix to the original matrix .
more formally , we have the following result : the hermitian matrices having a given eigenvector can be divided into families which differ only by ( real ) multiples of the identity matrix .
the imaginary part of the corresponding eigenvalue is the same for each member of such a family , and each family contains a unique member such that the real part of the corresponding eigenvalue vanishes .
[ nonreal ] if @xmath3 is an eigenvector of @xmath25 with eigenvalue @xmath39 , then @xmath3 is also an eigenvector of @xmath40 for any @xmath41 , with eigenvalue @xmath42 . in particular
, @xmath3 is an eigenvector of @xmath43 , with eigenvalue @xmath44 .
note that this technique can in general be used to eliminate the real part of only one eigenvalue at a time . nonetheless , any eigenvector with a non - real eigenvalue is also an eigenvector of a closely related matrix with a purely imaginary eigenvalue .
this suggests the following strategy for trying to find eigenvectors with non - real eigenvalues : rather than trying to find eigenvector / eigenvalue pairs @xmath3 and @xmath39 satisfying for given @xmath25 , with @xmath39 non - real , we will instead seek to categorize the matrices which admit such eigenvalues .
lemma [ nonreal ] now tells us that we can assume @xmath45)=0 without loss of generality , at least so long as we consider only a single eigenvector .
we will therefore attempt to find the matrices @xmath25 which admit a given vector @xmath3 as an eigenvector , with given eigenvalue @xmath39 satisfying @xmath45)=0 .
motivated by the first example , we will further assume that @xmath39 is a real multiple of @xmath20 $ ] , and we will consider only the case where @xmath20\ne0 $ ] . finally , by rescaling @xmath25 , the constant of proportionality can be assumed to be @xmath46 .
thus , we assume that @xmath47 \ne 0\ ] ] since @xmath20\ne0 $ ] by assumption , none of @xmath14 can be zero .
in particular , @xmath48 , and it is straightforward to solve for @xmath49 and for @xmath50 , yielding @xmath51 inserting these expressions into reduces to the form @xmath52 _ if @xmath20\ne0 $ ] , then @xmath53 , that is , @xmath54 . _ multiply both sides of on the left by @xmath55 .
the lhs of the resulting expression must be real , since the rhs is , but direct computation shows that the coefficient of @xmath56 on the left is @xmath57 .
since each factor except for @xmath58 is nonzero by assumption , @xmath58 must be zero .
if @xmath20\ne0 $ ] , then there are no solutions to unless @xmath4 . direct computation , as follows .
inserting @xmath54 into , the @xmath59-component yields @xmath60 . since each factor except for @xmath61 is nonzero by assumption , @xmath62 . in a separate computation
, the @xmath63-component of can be solved for @xmath64 , yielding @xmath65 inserting the result into the @xmath66-component of , along with @xmath67 , results in @xmath68 , which forces @xmath69 .
finally , the @xmath70- and @xmath71-components of can be solved for @xmath72 and @xmath73 , yielding @xmath74 and the result inserted into the @xmath63-component ( along with @xmath75 and the above expression for @xmath64 ) , resulting in @xmath76 , which forces @xmath77 .
thus , @xmath4 . _
if @xmath20\ne0 $ ] and @xmath4 , then there is a 6-parameter family of solutions to _ .
[ result ] inserting the above expressions for @xmath72 , @xmath73 and @xmath64 , as well as the condition @xmath4 , into results in a single nonzero component , which can be solved for @xmath78 , yielding @xmath79 as with the equations solved above for @xmath72 , @xmath73 and @xmath64 , the relevant coefficients are nonzero under the stated assumptions , so that the given solutions always exist .
we have thus constructed @xmath25 explicitly , with @xmath80 , @xmath81 , @xmath82 , @xmath26 , @xmath83 , and @xmath84 as free parameters .
we have created a method for finding a hermitian matrix @xmath85 which has an eigenvalue relationship with an imaginary vector @xmath2 , with the associator of @xmath3 , assumed to be nonzero , playing the role of the eigenvalue @xmath39 . for our method to be successful , rather than fixing the matrix
, we must begin by fixing the vector @xmath3 , thus fixing @xmath39 as well .
the question we must ask is if the resulting eigenvalue / eigenvector system from our construction method represents a variation of one of the existing three family examples presented by dray , janesky and manogue @xcite . the last two eigenvectors in example 1 satisfy the conditions of theorem [ result ] ( vanishing real part and non - vanishing associator ) , and @xmath26 and @xmath86 can be chosen so that the eigenvalue is precisely the vector associator , as required by our hypotheses .
it is straightforward to verify that the corresponding matrix @xmath28 is indeed contained in the 6-parameter family constructed in theorem [ result ] .
lemma [ nonreal ] can now be used , along with some obvious renormalization , to construct the matrices @xmath28 for any values of @xmath26 and @xmath86 . in this sense ,
example 1 is contained within our solution method , although our method generates many more solutions but can not yet handle the first eigenvector shown , whose associator vanishes . each of the eigenvectors in examples 2 and 3 , however , either has a non - zero real part or a vanishing associator , so that our results do not apply to these cases .
note that in example 2 , although the imaginary part of the eigenvalues is indeed in the direction of @xmath87 $ ] ( the associator of the elements of @xmath88 ) , namely @xmath27 , none of the given eigenvectors has an associator in this direction .
furthermore , in example 3 , the imaginary part of the eigenvalue no longer points in the direction of the matrix associator @xmath89 $ ] .
these examples therefore make clear that the results in this paper represent only the tip of the iceberg ; our assumptions are too restrictive . by understanding more about the problems encountered in trying to find a characteristic eigenvalue equation for @xmath1 hermitian matrices over the octonions ,
we hope our work will aid in the discovery of a method for finding the imaginary eigenvalues ( if any ) , and their corresponding eigenvectors , for any given @xmath1 hermitian octonionic matrix . | we show that any @xmath0-component octonionic vector which is purely imaginary , but not quaternionic , is an eigenvector of a 6-parameter family of hermitian octonionic matrices , with imaginary eigenvalue equal to the associator of its elements . |
in his remarkable proof of the kneser conjecture @xcite , lovsz gave a lower bound for the chromatic number of a graph using equivariant algebraic topology .
this was essentially done via a functor ( the edge complex , defined below ) from the category of graphs and graph morphisms to the category @xmath8@xmath9 ( topological spaces equipped with an action of the group @xmath8 and @xmath8maps , i.e. , continuous functions commuting with the given actions ) . in the case of the edge complex functor , the group @xmath8 is of order two , and the actions are free .
the edge complex of the complete graph @xmath3 is ( equivariantly homeomorphic to ) the sphere @xmath10 with the antipodal action , and the punchline is provided by the borsuk - ulam theorem . in @xcite ,
csorba observed that , for @xmath11 , @xmath12 , and @xmath13 , the @xmath1cycle complex ( defined below ) of the complete graph @xmath3 , denoted @xmath0 , is equivariantly homeomorphic to the stiefel manifold @xmath5 of ( ordered ) orthonormal 2-frames in the euclidean space @xmath7 , and he conjectured that this was true for all @xmath14 .
the non - equivariant version of csorba s conjecture was proven by c. schultz in @xcite , who also proved that @xmath0 is equivariantly homotopy equivalent to @xmath5 . in this note
we give a proof of the equivariant csorba conjecture : @xmath0 is equivariantly homeomorphic to @xmath5 with respect to the actions described below .
this provides a rather mysterious combinatorial model for the stiefel manifolds @xmath5 which are fundamental topological spaces .
the edge complex and the ( @xmath1)cycle complex are special cases of the lovsz graph multimorphism complex , which we proceed to define precisely now .
a * graph morphism * @xmath15 is a function from @xmath16 to @xmath17 ( where @xmath16 denotes the vertex set of the graph @xmath18 ) so that the image of an edge of @xmath18 is an edge of @xmath19 .
( in particular , a graph morphism from @xmath18 to the complete graph @xmath3 is just an admissible coloring of the vertices of @xmath18 . )
a * multimorphism * from @xmath18 to @xmath19 is a relation @xmath20 such that : 1 .
@xmath21 is non - empty for all @xmath22 , and 2 .
any function @xmath23 is a graph morphism . in other words
, @xmath24 is a multimorphism when there is a function @xmath23 and any function @xmath23 is a graph morphism .
the lovsz * multimorphism complex * @xmath25 @xcite is a bifunctor ( contravariant in the first variable , covariant in the second ) assigning a regular cell complex to the ( simple ) graphs @xmath26 .
any regular cell complex is determined by its face poset ( @xcite p200 , @xcite ch 3 1 ) , and the face poset of @xmath25 is the set of multimorphisms from @xmath18 to @xmath19 ordered by inclusion .
geometrically , the cells of the complex @xmath25 are products of simplices ; the cell corresponding to the multimorphism @xmath24 : @xmath27 where @xmath28 is the full simplex on the vertex set @xmath29 .
@xmath25 is the union of all these cells indexed by the multimorphisms .
we will identify each cell with the multimorphism @xmath24 indexing it and refer to it as such .
the vertices ( @xmath30cells ) of @xmath25 are the graph morphisms from @xmath18 to @xmath19 , the vertices of a given cell @xmath24 are obtained by choosing any @xmath31 from @xmath32 for each @xmath33 .
the composition of two multimorphisms is also a multimorphism , hence @xmath25 is ( bi-)functorial for multimorphisms ( not just for morphisms , this extended functoriality was put to very good use in @xcite ) .
that is , a multimorphism @xmath34 induces a ( cellular ) map @xmath35 , and a multimorphism @xmath36 induces a ( cellular ) map @xmath37 . if @xmath8 and @xmath38 are groups acting on the graphs @xmath18 and @xmath19 respectively , then there is an action of @xmath39 on @xmath40 inducing an action on the relations , i.e. , subsets of @xmath40 , which restricts to an action on @xmath25 . in particular
, a graph @xmath18 equipped with an action of a group @xmath8 defines the functor @xmath41 from the category of graphs and graph ( multi-)morphisms to @xmath8@xmath9 , the equivariant category of @xmath8spaces and @xmath8maps .
lovsz s proof of the kneser conjecture @xcite essentially employs the * edge complex * functor @xmath42 with @xmath8 being the automorphism group of ( the edge ) @xmath43 , the complete graph on @xmath6 vertices .
if we denote the vertices of @xmath43 with @xmath44 and @xmath45 , then the nontrivial element of @xmath8 is the involution on @xmath46 switching @xmath47 and @xmath48 , the subsets ( of @xmath17 ) assigned to the two vertices of @xmath43 .
this action is free because @xmath47 and @xmath48 are distinct since they are nonempty and disjoint . in particular
, @xmath49 is a poset consisting of pairs @xmath50 of nonempty disjoint subsets of @xmath51 , ordered by component - wise inclusion , where @xmath52 and @xmath53 .
@xmath49 can be geometrically realized as the @xmath54sphere of radius two with respect to the @xmath55 norm ( inducing the taxicab metric ) in the hyperplane orthogonal to the diagonal vector @xmath56 in the euclidean space @xmath57 , where @xmath47 and @xmath48 are the sets of coordinates with positive and negative values respectively .
switching @xmath47 and @xmath48 yields the antipodal action .
interestingly , long before the definition of @xmath58 , the underlying spaces of @xmath59 figured prominently in two unrelated applications of equivariant algebraic topology ( neither one involving chromatic numbers or graphs ) : in alon s elegant necklace splitting theorem ( with @xmath60 prime ) @xcite and the solution of the prime power case of the topological tverberg problem ( @xcite and @xcite ) , which was conjectured by brny , shlossman , and szcs in @xcite .
the lovsz conjecture ( proven by babson and kozlov @xcite , see also @xcite ) is about the ( odd ) * * cycle complex**es @xmath61 where @xmath62 is the @xmath60gon .
@xmath61 also has an involution induced by a reflection ( of the @xmath60gon ) .
when @xmath60 is odd , any such reflection flips an edge and hence induces a free action on @xmath61 as above .
the lovsz conjecture reduces to a computation involving the equivariant cohomology of @xmath63 .
a graph multimorphism @xmath24 in @xmath64 is also determined by specifying @xmath65 for each @xmath66 .
clearly , for any @xmath24 , each @xmath67 is independent ( i.e. , no two elements are adjacent in @xmath18 ) .
the * independence complex * @xmath68 of a graph @xmath18 is the simplicial complex with vertex set @xmath16 and simplices being independent subsets of @xmath16 .
equivalently , @xmath68 is the flag complex ( the largest simplicial complex whose @xmath69skeleton is the given graph ) of the edge complement of @xmath18 ( the graph with the same vertices as @xmath18 but with the complementary edge set ) .
any flag complex is the independence complex of the edge complement of its @xmath69skeleton . when @xmath70 , the only condition on @xmath67 is being in @xmath68 , so @xmath71 consists of @xmath24 such that : 1 .
@xmath32 is non - empty for all @xmath22 , and 2 .
@xmath72 is independent for @xmath73 .
we can identify @xmath74 with the face poset of the join ( over @xmath73 ) of the links of the @xmath72 in @xmath68 . since @xmath75 is the face poset of the boundary of a product ( over the vertices @xmath76 ) of the simplices on @xmath32 , @xmath71 is a ( closed ) manifold if @xmath68 is a pl ( piecewise - linear ) sphere .
in fact , the converse is also true @xcite .
the complex @xmath63 is a manifold only when @xmath77 since this is the only time when @xmath78 is a sphere : @xmath79 is 3 points , @xmath80 is the disjoint union of two edges , @xmath81 is a pentagon , and @xmath78 with @xmath82 has maximal simplices of different dimensions . on the topological side , the orthogonal group @xmath83 acts on the stiefel manifold @xmath84 ( with the grassmannian @xmath85 of @xmath6planes in @xmath86space as the quotient ) .
the group @xmath83 is the semi - direct product of rotations @xmath87 with a subgroup generated by an arbitrary reflection .
two natural involutions on @xmath5 are ( i ) @xmath88 and ( ii ) @xmath89 . since any two reflections are conjugate via a rotation , these give equivalent actions .
an explicit map @xmath90 interchanging the actions ( i ) and ( ii ) is @xmath91 .
schultz @xcite used the action ( i ) on the stiefel manifold .
below , we will use the ( equivalent ) action ( ii ) .
the corresponding involution on @xmath0 is induced by any reflection of the pentagon @xmath2 ( they all give equivalent actions ) .
it is convenient to work with a smaller model : @xmath92 , in which two multimorphisms are considered the same if their values differ only on the independent set @xmath93 of vertices .
thus , @xmath92 is the subcomplex of @xmath94 consisting of the cells @xmath24 that can be extended to @xmath18 .
the projection from @xmath64 to @xmath92 is a homotopy equivalence @xcite since the fibers are contractible .
it turns out that @xmath71 is homeomorphic ( but not via the aforementioned projection ) to @xmath95 when @xmath68 is a pl sphere @xcite .
for @xmath92 to inherit the @xmath8action ( induced from an action on @xmath18 ) , the set @xmath93 needs to be ( setwise ) @xmath8invariant .
when @xmath8 is the group of order two generated by the reflection on @xmath2 switching the vertex @xmath96 with the vertex @xmath97 for @xmath98 , there are two @xmath8invariant independent subsets : @xmath99 and @xmath100 . in @xcite schultz
uses @xmath101 ; we use @xmath102 .
we need @xmath103 and @xmath0 to be equivariantly homeomorphic which we get from an equivariant version of a lemma of schultz @xcite . while @xmath0 ( or @xmath104 , which is equivariantly homeomorphic to it ) is the star of this story , the hero is @xmath105 , where @xmath106 is the path with four edges from @xmath69 to @xmath1 ( the subgraph of @xmath2 missing the edge @xmath107 ) .
our story also features @xmath108 , which is the disjoint union of two copies of @xmath43 , namely the edges @xmath109 and @xmath110 .
the graph morphism @xmath111 sending @xmath69 and @xmath1 to @xmath44 and @xmath6 and @xmath112 to @xmath45 induces the diagonal embedding of ( the sphere ) @xmath49 in @xmath113 .
we also have the restriction map @xmath114 induced from the inclusion of @xmath106 in @xmath2 and the inclusion of @xmath105 into @xmath115 .
all these maps are equivariant with respect to the involutions induced from the aforementioned @xmath116 .
( the action on @xmath49 is trivial since it maps homeomorphically to the diagonal in @xmath115 , the fixed point set of the involution . )
note that @xmath105 is not homeomorphic to @xmath117 : we will show that @xmath105 is a @xmath118dimensional manifold with boundary while @xmath117 is not a manifold , having maximal cells of every dimension from @xmath119 to @xmath120 .
( recall that @xmath121 . )
the face posets of these lovsz complexes above are actually very easy to describe concretely ; all are made up of ( ordered ) quadruples of nonempty subsets @xmath122 of @xmath123 satisfying further conditions .
for any cell @xmath24 , @xmath122 are @xmath124 , and @xmath125 respectively . in @xmath126
we have @xmath127 .
in @xmath105 we also have that @xmath128 , and @xmath104 has the further restriction that @xmath129 .
the diagonal @xmath49 has @xmath130 and @xmath131 ( in addition to @xmath132 ) .
we will denote these cells as @xmath133 to remind us of their locations in terms of the vertices of the pentagon @xmath2 .
such an array should not be thought of as a @xmath134 matrix .
on the topological side we have the involution on @xmath135 switching the two coordinates , with the diagonal @xmath10 as the fixed set .
an equivariant regular neighborhood of the diagonal is @xmath136 , which is a manifold with boundary @xmath137 .
we prove in section 5 that @xmath105 is a manifold with boundary @xmath104 and also an equivariant regular neighborhood of the diagonal in @xmath138 .
from the equivariant pl theory of collapsing , shelling , and regular neighborhoods developed in section 4 , our main result follows : the regular cell complex @xmath105 is a pl manifold with boundary @xmath104 and is equivariantly homeomorphic ( with respect to the involution described above ) to @xmath139 , where the involution on @xmath140 interchanges @xmath141 with @xmath142 .
the stiefel manifold @xmath143 is therefore equivariantly homeomorphic to @xmath104 , which is equivariantly homeomorphic to @xmath0 .
the structure of the rest of the paper is as follows : we establish notation and give basic definitions and prove that @xmath0 and @xmath104 are equivariantly homeomorphic in section 2 .
the small amount of discrete morse theory that we use is in section 3 .
section 4 is devoted to the stating the equivariant versions of facts we need from piecewise - linear topology given in @xcite .
finally , in section 5 we specialize to our setup and prove the equivariant csorba conjecture .
all simplicial complexes and posets we consider are finite .
an ( abstract ) * simplicial complex * @xmath144 on a vertex set @xmath145 is a collection of ( finite ) subsets of @xmath145 such that if @xmath146 , then @xmath147 .
the dimension of the simplex @xmath148 is one less than its cardinality , and the * @xmath14skeleton * of the simplicial complex @xmath144 ( denoted @xmath149 ) is the sub - simplicial complex made up of all simplices of @xmath144 with dimension at most @xmath14 .
however , we will abuse the notation and use @xmath150 to also mean @xmath151 .
the nonempty simplices @xmath147 are also called the faces of @xmath144 .
the * geometric realization * of @xmath144 is the topological space @xmath152 where @xmath153 is the standard basis vector of @xmath154 corresponding to @xmath155 .
the underlying topological space of a face @xmath148 of @xmath144 is @xmath156 note that we are using vertical bars to denote both the cardinality of a finite set and the geometric realization of a simplex or a simplicial complex ; which one is meant should be clear from the context . if @xmath148 is a simplex in a simplicial complex @xmath144 , its * link * is defined @xmath157 .
if @xmath144 and @xmath158 are two simplicial complexes , we define their * join * @xmath159 .
a subcomplex @xmath158 of a simplicial complex @xmath144 is called * full * if it satisfies the property that if @xmath147 and @xmath160 , then @xmath161 .
a * regular cell structure * on a ( compact hausdorff ) topological space @xmath162 is a ( finite ) collection @xmath163 of subspaces ( called cells or faces ) , each homeomorphic to a ( closed ) disk of dimension @xmath164 for some @xmath164 , such that ( 1 ) @xmath162 is the disjoint union of the relative interiors of its cells ( called the open cells and denoted by int @xmath165 ) , and ( 2 ) the boundary of each cell is a union of ( lower dimensional ) cells .
a topological space with a regular cell structure is a * regular cell complex*. for any simplicial complex @xmath144 , the collection @xmath166 gives a regular cell structure on @xmath167 .
for any poset @xmath168 , its * order complex * @xmath169 is the simplicial complex whose simplices are chains @xmath170 , with @xmath171 .
the faces of a regular cell complex under inclusion form a poset whose order complex is its barycentric subdivision ( @xcite p200 , @xcite ch 3 1 ) .
note that this notation is consistent with @xmath28 being the full simplex on a vertex set @xmath29 if the set @xmath29 is endowed with any total order . also note that an abstract simplicial complex @xmath144 is itself a poset and includes the empty simplex ( unlike its face poset ) , so the order complex @xmath172 is the cone of the barycentric subdivision of @xmath144 .
a * @xmath8simplicial complex * , where @xmath8 is a ( finite ) group , is a simplicial complex @xmath144 equipped with a permutation action of @xmath8 on the vertex set @xmath145 so that the induced action on the subsets of @xmath145 sends simplices to simplices .
similarly , in a @xmath8cell complex , the group action permutes the cells .
for any cell ( or simplex ) @xmath165 , @xmath173 is called its * stabilizer*. it will be convenient for us to define a * @xmath8regular cell complex * to be a topological space with a @xmath8action and a regular cell structure with the group @xmath8 permuting the cells so that every closed cell is @xmath174homeomorphic to a cone on its boundary with the apex fixed by @xmath174 .
the geometric realization of a @xmath8simplicial complex @xmath144 is a @xmath8regular cell complex with cells @xmath166 since the stabilizer of a simplex always fixes its centroid .
similarly , if the cells of a regular cell complex are convex with @xmath8 acting affinely on each one , the complex is @xmath8regular .
if the stabilizer of each cell ( or simplex ) fixes the cell pointwise , the complex is called * admissible*. if @xmath8 acts on a poset @xmath168 ( preserving the order ) , then @xmath169 is an admissible @xmath8simplicial complex .
just as in the non - equivariant case ( @xcite p200 , @xcite ch 3 1 ) , the face poset of a @xmath8regular cell complex determines its @xmath8homeomorphism type : [ gcellcx ] if x is a g - regular cell complex with face poset f , then @xmath162 is @xmath8homeomorphic to @xmath175 .
the proof below is very similar to the non - equivariant case and is mainly given for completeness .
the proof is by induction on the number of cells in @xmath162 .
there is nothing to prove if x consists of a single @xmath30cell .
now , choose a maximal cell @xmath165 and define @xmath176 with the induced @xmath8cell structure .
by the induction hypothesis , @xmath177 . also , @xmath178 .
then we take the cone of @xmath179 with its apex being a point @xmath180 fixed by @xmath174 . extending this coning equivariantly to @xmath181
gives the homeomorphism from @xmath175 to @xmath162 .
we are specifically interested in the @xmath8equivariant homeomorphism type of some lovsz multimorphism complexes @xmath71 , for a graph @xmath18 with a group of symmetries @xmath8 .
the following equivariant version of lemma 3.5 in @xcite enables us , in some cases , to work with the smaller and more convenient @xmath8regular cell complexes @xmath182 , where @xmath93 is a @xmath8invariant independent subset of vertices of @xmath18 .
[ schultz2 ] let @xmath18 be a graph with a g - action , n @xmath183 1 , and @xmath93 a @xmath8invariant , independent subset of the vertex set @xmath16 .
for all @xmath22 , define @xmath184 if there is a @xmath8homeomorphism @xmath185 such that @xmath186 for all @xmath22 , then @xmath71 is @xmath8homeomorphic to @xmath95 .
following @xcite , we consider the equivariant poset embedding @xmath187 given by @xmath188 .
then @xmath189 can naturally be identified with those relations @xmath190 that are multimorphisms from the induced subgraph on the vertices with @xmath32 nonempty to the complete graph @xmath3 .
the additional condition that no @xmath32 can be empty yields the following description : @xmath191 all the @xmath192 satisfy the condition that if @xmath193 and @xmath194 , then @xmath195 .
therefore , taking the order complex commutes with unions , and we obtain that @xmath196 we use a similar argument for @xmath95 .
the image of its embedding in @xmath197 has the additional condition that for each vertex in @xmath93 , there is some element of @xmath198 that is not related to any of its neighbors in @xmath18 .
we have that , for @xmath199 , @xmath200 satisfies the same condition as @xmath192 above . for @xmath201 , @xmath200 also satisfies the condition that if @xmath202 and @xmath203 , then @xmath204 .
hence , @xmath205 thus , using the @xmath8homeomorphism from the hypothesis on each coordinate in the product , we obtain that @xmath206 .
in our case , we choose @xmath102 as our independent set in @xmath2 .
the existence of a @xmath207homeomorphism @xmath208 satisfying the necessary conditions is easy to see from figure [ fig1 ] .
thus , we have @xmath0 is equivariantly homeomorphic to @xmath209 . and @xmath210 [ fig1 ] ]
we will use robin forman s discrete morse theory to find collapsings of simplicial complexes .
more thorough discussions of the subject can be found in @xcite and @xcite .
let @xmath144 be a finite ( abstract ) simplicial complex .
we use the notation @xmath211 if @xmath212 and dim @xmath213 . by a vector ,
we mean a pair @xmath214 , where @xmath215 is thought of as the head of the vector and @xmath148 the tail . a * discrete vector field * on @xmath144
is defined to be a collection of vectors @xmath216 such that each simplex @xmath217 belongs to at most one element of @xmath145 , either as a head or a tail of a vector . given a discrete vector field @xmath145 on @xmath144 , we have the notion of a * path * , which is a sequence of simplices in k of the form : @xmath218 where @xmath219 , @xmath220 and @xmath221 .
we say a path as above has length @xmath222 . by a * cycle * we mean a path as above with @xmath223 .
a * morse matching * ( or a * discrete gradient field * ) is a discrete vector field @xmath145 with no cycles .
the simplices which are unpaired in @xmath145 are called * critical*. an equivalent concept to a morse matching is a height function on @xmath144 . a * height * ( or * morse * )
* function * is a map @xmath224 satisfying @xmath225 , @xmath226 given a height function @xmath227 , the corresponding morse matching is the collection of pairs @xmath228 for which @xmath229 .
conversely , it is not difficult to construct a height function inducing a given morse matching ( @xcite , @xcite ) .
this height function is clearly not unique .
in fact , we may adjust it to be one - to - one and to take values in @xmath230 ( without changing the morse matching ) .
the following is the key lemma from discrete morse theory we will use in this paper .
a more general version for cell complexes can be found in @xcite , theorem 11.13 .
[ morse2 ] let k be a finite simplicial complex with a morse matching whose critical simplices form a subcomplex l. then k collapses simplicially to l. let @xmath231 be a one - to - one height function corresponding to the given morse matching . define a new height function @xmath232 as follows : for @xmath161 , set @xmath233 . for @xmath234 , set @xmath235 . under this new height function ,
all of the simplices in @xmath158 remain critical , and the relative heights of all the simplices outside of @xmath158 are unchanged , preserving their pairings
. therefore , @xmath236 corresponds to the same morse matching as @xmath227 , and @xmath236 is one - to - one on @xmath237 .
now , for @xmath238 define @xmath239 note that @xmath240 . for @xmath241 , either @xmath242 or @xmath243 where @xmath244 and @xmath245 . in the latter case ,
@xmath246 collapses to @xmath247 along the free face @xmath248 . since @xmath249
, @xmath144 collapses to @xmath158 via a sequence of these collapsings .
in this section , we discuss the equivariant theory of @xmath8regular neighborhoods with suitable hypotheses , similar to rourke and sanderson s non - equivariant version @xcite . throughout ,
we assume @xmath8 is a finite group , all simplicial complexes are finite , and maps between polyhedra are piecewise - linear .
also , when there is a @xmath8action on a space @xmath162 , the product @xmath250 is assumed to have the @xmath8action @xmath251 . a simplicial complex @xmath144 * polyhedrally triangulates * a space @xmath252 if the @xmath30skeleton @xmath150 is identified with a subset of @xmath162 so that the canonical piecewise - linear function from @xmath167 to @xmath57 mapping @xmath253 of each simplex @xmath148 to the convex hull of its vertices in @xmath162 is a homeomorphism onto @xmath162 .
then @xmath162 is called a * polyhedron * , and two polyhedra are equivalent if there is a piecewise - linear homeomorphism between them . if @xmath144 is a @xmath8complex , the ambient vector space @xmath57 into which the polyhedron @xmath162 is @xmath8embedded is the underlying space of a ( real ) linear representation of the group @xmath8 , and the piecewise - linear homeomorphism from @xmath167 to @xmath162 is a @xmath8map , then we have a * polyhedral @xmath8triangulation * of a * @xmath8polyhedron * @xmath162 .
we implicitly identify each simplex @xmath148 of @xmath144 with the image of its geometric realization in @xmath162 .
note that one abstract ( @xmath8)simplicial complex may triangulate @xmath162 in multiple ways by having different choices for some of the vertices in @xmath162 , in which case we will give the simplicial complex different names .
we will only consider polyhedral triangulations .
let @xmath254 be polyhedra and @xmath158 be a subcomplex of @xmath144 such that @xmath255 triangulates @xmath256 where @xmath140 is a neighborhood of @xmath257 in @xmath162 .
define the * simplicial neighborhood * of @xmath158 in @xmath144 , @xmath258 . also , define @xmath259 .
the * derived subdivision * @xmath260 of @xmath144 near @xmath158 is ( combinatorially ) the simplicial complex with vertex set @xmath261 , and the simplices are of the form @xmath262 where @xmath161 or @xmath263 with @xmath264 and @xmath265 .
geometrically , @xmath260 is realized by selecting the new vertex @xmath266 in the interior of each simplex @xmath215 of @xmath237 that intersects a simplex of @xmath158 and then , in ascending order of dimension , replacing each @xmath215 with the cone ( with apex @xmath266 ) on its boundary ( which has already been subdivided in the previous steps ) . now suppose @xmath158 is full in @xmath144 ( i.e. , if a set of vertices in @xmath158 forms a simplex in @xmath144 , then they form a simplex in @xmath158 ) and @xmath267 in @xmath162
. then @xmath268 is called a * regular neighborhood * of @xmath257 in @xmath162 .
if @xmath144 and @xmath158 are both admissible @xmath8complexes and , when defining @xmath260 , the set of new vertices @xmath269 is chosen to be @xmath8invariant , we say @xmath270 is a * @xmath8regular neighborhood*. [ reg1 ] if k is a @xmath8simplicial complex , @xmath271 $ ] has an admissible @xmath8triangulation with no new vertices in @xmath272 .
let @xmath273 be the face poset of @xmath144 .
then the order complex @xmath274 of the product poset @xmath275 satisfies @xmath276 .
hence @xmath274 gives the desired @xmath8triangulation .
let @xmath277 be a collection of simplicial complexes , each one triangulating the polyhedron @xmath162 . for @xmath278 ,
let @xmath279 .
define an equivalence relation on @xmath280 by @xmath281 .
we define a poset @xmath282 with the partial order @xmath283 \leq [ \psi ] \leftrightarrow |\phi| \subseteq |\psi|$ ] .
we will identify an equivalence class @xmath283 $ ] with the geometric realization @xmath284 of its representatives .
[ commonsubdivision ] given a collection @xmath285 of simplicial complexes that triangulate a polyhedron x , then @xmath286 is a regular cell structure on x , so that @xmath287 is a triangulation of x and a common subdivision of @xmath288 .
we show first that the open cells of @xmath286 are disjoint . in this discussion , let @xmath289 and @xmath290 .
note that if @xmath291 , we have that @xmath292 .
thus , the intersection of any two closed cells is a closed cell .
if @xmath283 < [ \psi]$ ] , then , without changing @xmath293 $ ] or @xmath283 $ ] , we may replace each @xmath294 with its minimal face containing @xmath295 and each @xmath296 with @xmath297 .
then @xmath298 is always nonempty , and for some @xmath96 we have @xmath299 .
therefore we have that @xmath300 , implying that the intersection of any two distinct closed cells must occur on the boundary of at least one of them , so any two distinct open cells are disjoint .
each @xmath295 is a ( nonempty ) compact , convex polytope , yielding that @xmath301 for some @xmath302 .
furthermore , @xmath303 < [ \psi ] } |\phi|$ ] : if @xmath304 , there is a unique @xmath305 such that @xmath306 , giving some @xmath24 with @xmath307 and @xmath283 \leq [ \psi]$ ] .
however @xmath283 < [ \psi]$ ] because @xmath308 is not in @xmath309 , so there is an @xmath96 with @xmath308 not in @xmath310 , i.e. , @xmath311 .
conversely , if @xmath312 < [ \psi ] } as above .
this proves that @xmath286 is a regular cell structure on @xmath162 .
therefore , we have that @xmath313 .
[ reg2 ] if @xmath162 is a @xmath8polyhedron and @xmath144 is a nonequivariant triangulation of @xmath162 , then there is an admissible @xmath8triangulation of @xmath162 which is a subdivision of k. for each @xmath314 , @xmath315 is another triangulation of @xmath162 because @xmath8 is acting linearly on the ambient representation space .
let @xmath316 . using the notation from the proof of [ commonsubdivision
] , each cell @xmath283 $ ] is given by a map @xmath317 .
then @xmath318 . for @xmath319
, define @xmath320 .
this induces an order - preserving @xmath8action on @xmath286 because , for any @xmath24 and any @xmath227 in @xmath8 , @xmath321 .
lastly , @xmath286 is a @xmath8regular cell complex because the average of the vertices of any cell is fixed by the cell s stabilizer .
the result follows by lemma [ gcellcx ] .
[ reg3 ] let ( k , l ) be an admissible @xmath8triangulation of ( x , y ) with l full in k and let @xmath322 be a @xmath8subdivision of ( k , l )
. then @xmath323derived subdivisions @xmath260 and @xmath324 of k and @xmath325 near l and @xmath326 such that @xmath327 .
we follow the proof of the non - equivariant version , lemma 3.7 in @xcite , and clarify some details with the combinatorial definition of derived subdivisions .
we define a map @xmath328 $ ] by setting @xmath329 if @xmath330 and @xmath331 if @xmath332 gives a map on the vertices of @xmath144 which can be linearly extended on simplices
. choose @xmath333 small enough so that no vertex of @xmath334 is contained in @xmath335 $ ] .
then choose @xmath8derived subdivisions @xmath260 and @xmath324 of @xmath144 and @xmath325 near @xmath158 and @xmath326 respectively with all the new vertices @xmath266 lying in @xmath336 .
we can choose these vertices equivariantly because @xmath15 is @xmath8invariant and @xmath144 is admissible .
now we will show that @xmath337 = the map @xmath15 takes values of @xmath30 or @xmath333 on all of the vertices of @xmath338 and @xmath339 , so both of these neighborhoods are contained in @xmath335 $ ] .
now let @xmath308 be a point in @xmath340 \subset |k^{\prime}| = @xmath341 of @xmath260 ( respectively @xmath324 ) as in the combinatorial definition of a derived subdivision .
then @xmath342 where @xmath343 and @xmath344 .
suppose @xmath148 is not in @xmath158 ( resp .
@xmath326 ) . then @xmath345 ( resp .
@xmath346 ) for @xmath347 .
meanwhile , @xmath348 for @xmath349 .
we have then , in both cases , that @xmath350 .
this is a contradiction , and @xmath148 must be in @xmath158 ( resp .
@xmath326 ) , yielding that @xmath308 is in @xmath351 ( resp .
@xmath352 ) .
[ links ] let @xmath353 be a vertex of a @xmath8simplicial complex @xmath144 . if @xmath260 is a derived @xmath8subdivision of @xmath144 near @xmath353 , then @xmath354 is @xmath355-homeomorphic to @xmath356 .
assume that the derived vertices are chosen in @xmath357 for some @xmath358 .
then a point in @xmath359 is of the form @xmath360 where @xmath361 .
mapping this point to @xmath362 gives the desired homeomorphism .
[ reg4 ] if @xmath270 and @xmath363 are @xmath8regular neighborhoods of y in x , then there exists a @xmath8homeomorphism @xmath364 that maps @xmath270 to @xmath363 and is the identity on y. the proof mirrors that of the non - equivariant version , theorem 3.8 in @xcite .
we have @xmath365 for @xmath366 as above , where @xmath367 are @xmath8triangulations of a neighborhood of @xmath257 in @xmath162 . as in @xcite , @xmath325 and @xmath43
may be subdivided and extended to triangulations of @xmath368 . applying [ commonsubdivision ]
, there is a triangulation @xmath369 of @xmath370 containing subdivisions of @xmath325 and @xmath43 and inducing @xmath371 , a common subdivision of @xmath326 and @xmath372 . by [ reg2 ]
, @xmath369 may be assumed to be a @xmath8triangulation . by [ reg3 ] ,
for @xmath366 , we can find @xmath8derived subdivisions @xmath373 and @xmath374 of @xmath367 and @xmath369 near @xmath375 and @xmath371 respectively such that @xmath376 .
therefore we have that @xmath377 where each @xmath8homeomorphism is induced by a @xmath8isomorphism of @xmath8derived subdivisions .
such an isomorphism is given by changing the placement of each derived vertex @xmath266 within the interior of a simplex @xmath215 touching @xmath257 and fixing the placement of every other vertex in the subdivision , so the corresponding @xmath8homeomorphism is the identity on @xmath257 itself .
thus , the same is true of the composition of these @xmath8homeomorphisms .
let @xmath254 be @xmath8polyhedra .
a * @xmath8collar * on @xmath257 in @xmath162 is a pl @xmath8embedding @xmath378 such that @xmath379 and @xmath380 is an open neighborhood of @xmath257 in @xmath162 .
suppose that for every @xmath381 there are ( closed , polyhedral ) neighborhoods @xmath382 and @xmath145 of @xmath383 in @xmath162 and @xmath257 respectively with @xmath384 , such that for any @xmath314 , @xmath385 implies that @xmath386 and @xmath387 .
suppose further that @xmath388 with @xmath389 on @xmath145 .
then we say that @xmath257 is * locally @xmath8collarable * in @xmath162 and we have that @xmath390 .
local @xmath8collarability is equivalent to @xmath8collarability .
[ reg5 ] suppose @xmath254 is locally @xmath8collarable . then there is a @xmath8collar on y in x. the proof is essentially the same as that of the non - equivariant version , theorem 2.25 in @xcite , substituting local @xmath8collars for local collars . construct a new @xmath8polyhedron @xmath391 $ ] by attaching a @xmath8collar to @xmath257 outside of @xmath162 , identifying @xmath254 with @xmath392 .
we will construct a @xmath8homeomorphism between @xmath162 and @xmath393 which carries @xmath257 to @xmath394 .
then the preimage of @xmath395 $ ] will be a @xmath8collar on @xmath257 in @xmath162 . for each @xmath381 ,
let @xmath396 be a local @xmath8collar at @xmath383 .
using compactness , cover @xmath257 with the interiors of finitely many @xmath397 .
then for each @xmath96 , we will define a @xmath8homeomorphism @xmath398 which maps the interior of @xmath399 into the interior of @xmath400 $ ] and is the identity outside of @xmath401 $ ] . to do this ,
we consider a @xmath8triangulation @xmath402 of @xmath403 . taking the product of this triangulation with an interval
defines a regular @xmath8cell structure on @xmath401 $ ] .
we now equivariantly subdivide this cell complex by its order complex : we must choose a new vertex @xmath404 in the interior of each cell @xmath165 .
first , equivariantly select a point @xmath405 in the interior of each simplex @xmath406 .
for the cell @xmath407 , choose @xmath408 .
likewise , for the cell @xmath409 , choose @xmath410 . finally , for the cell @xmath411 $ ] , choose @xmath412 .
define a different subdivision simply by moving @xmath404 to @xmath413 when @xmath165 intersects the relative interior of @xmath414 $ ] and making no change in the placement for @xmath404 otherwise .
then the @xmath8homeomorphism @xmath415 is given by mapping the first subdivision to the second , since they are both realizations of the same order complex .
note that @xmath415 is the identity except on the relative interior of @xmath401 $ ] where @xmath416 with @xmath417 .
thus , @xmath415 maps all of the relative interior of @xmath418 into @xmath419 .
[ fig - gcollar ] ] now define @xmath227 to be the composition @xmath420 . since the interiors of the local @xmath8collars cover all of @xmath257 , @xmath227 maps all of @xmath392 into @xmath421 .
let @xmath144 @xmath8triangulate @xmath257 , and thus also @xmath422 .
then we consider @xmath423 $ ] .
it has a regular cellular @xmath8structure with a face poset isomorphic to that of @xmath424 $ ] .
the cells in the former come in three types : simplices of @xmath144 triangulating @xmath392 ( which correspond to the same in the latter complex ) , simplices of @xmath144 triangulating @xmath422 ( which correspond to simplices of @xmath144 triangulating @xmath394 ) , and the intersection of @xmath425 with cells @xmath426 $ ] for @xmath147 ( which correspond to the cells @xmath426 $ ] ) .
therefore , @xmath423 $ ] is @xmath8homeomorphic to @xmath395 $ ] , fixing @xmath392 .
we extend this homeomorphism by the identity to the rest of @xmath162 to get the desired @xmath8homeomorphism @xmath427 carrying @xmath257 to @xmath394 .
[ reg7 ] if @xmath254 is locally @xmath8collarable then a @xmath8regular neighborhood of y in x is a @xmath8collar .
we have by [ reg5 ] that @xmath257 is @xmath8collarable .
let @xmath428 be an admissible @xmath8triangulation of @xmath257 .
let @xmath144 be the @xmath8triangulation of the @xmath8collar @xmath429 as in [ reg1 ] .
now choose a @xmath8derived subdivision @xmath260 of @xmath144 near @xmath158 such that all of the new vertices have second coordinate @xmath430 .
then @xmath431 \approx_g y \times i$ ] .
the result now follows from [ reg4 ] .
we will also make use of the notion of bicollarability .
we say @xmath254 is * @xmath8bicollarable * in @xmath162 if there exists a @xmath8embedding of @xmath432 $ ] into @xmath162 with @xmath433 for all @xmath434 and @xmath435 mapping to an open neighborhood of @xmath257 in @xmath162 .
[ regnbhd - bicollar ] if @xmath436 is a @xmath8regular neighborhood of @xmath257 in @xmath162 , then @xmath437 is @xmath8bicollarable in @xmath162 . by [ reg4 ] , it suffices to consider the case @xmath438 $ ] for some @xmath358 .
that is , the derived vertices @xmath269 of @xmath260 were chosen in @xmath336 .
let @xmath439 .
equivariantly , choose alternate derived vertices @xmath440 and @xmath441 in @xmath442 and @xmath443 respectively , giving derived @xmath8subdivisions @xmath324 and @xmath444 of @xmath144 near @xmath158 .
then there are the natural homeomorphisms @xmath445 given by sending each @xmath266 to @xmath446 .
we now define a @xmath8bicollar @xmath447 \rightarrow_g cl ( |n_{k_2^{\prime}}(l)| \setminus |n_{k_1^{\prime}}(l)|)$ ] by setting @xmath448 note that this @xmath8bicollarability can alternatively be expressed as @xmath437 being @xmath8collarable in both @xmath140 and in @xmath449 . for the remainder of the section
, we wish to consider @xmath8regular neighborhoods within manifolds . for that purpose
, we need to define a @xmath8manifold .
we will consider @xmath8polyhedra and @xmath8complexes that are manifolds and that have particularly well - behaved @xmath8actions .
first , consider an orthogonal representation @xmath450 .
we denote by @xmath451 and @xmath452 the unit sphere and disk respectively in the corresponding representation space .
further , denote by @xmath453 the hemisphere with last coordinate nonnegative and similarly for @xmath454 .
these have unique piecewise - linear structures coming from their smooth structures @xcite .
we now inductively define a * combinatorial @xmath8sphere*. @xmath455 with a @xmath8action is a combinatorial @xmath30dimensional @xmath8sphere .
an admissible simplicial @xmath8complex @xmath144 with @xmath167 ( pl ) @xmath8homeomorphic to @xmath451 for some @xmath456 is an @xmath14dimensional combinatorial @xmath8sphere if for every @xmath457 , @xmath458 is an @xmath86dimensional combinatorial @xmath355sphere , itself @xmath355homeomorphic to @xmath459 , where @xmath460 is the orthogonal complement in @xmath461 of the trivial representation @xmath462 .
similarly , we may define a * combinatorial @xmath8hemisphere * by substituting @xmath453 and allowing links of vertices to be @xmath14dimensional @xmath8spheres or @xmath8hemispheres in the above definition .
finally , a * combinatorial @xmath8disk * is simply the cone on a @xmath8sphere or @xmath8hemisphere with a @xmath8fixed point .
an admissible simplicial @xmath8complex @xmath144 is an @xmath14dimensional * combinatorial @xmath8manifold * if for every @xmath457 , @xmath458 is an @xmath86dimensional combinatorial @xmath355sphere or hemisphere .
a @xmath8polyhedron @xmath463 is an @xmath14dimensional * ( pl ) @xmath8manifold * ( with boundary ) if its admissible @xmath8triangulations are @xmath14dimensional combinatorial @xmath8manifolds .
( if it is true for one , it is true for all . )
the boundary @xmath464 is easily shown to be a @xmath8submanifold . [ submanifold ] let @xmath463 be an @xmath14dimensional @xmath8manifold and @xmath465 be an @xmath14dimensional @xmath8invariant submanifold with @xmath466 @xmath8bicollarable in @xmath463 .
then @xmath465 is a @xmath8manifold .
let @xmath144 be a @xmath8triangulation of @xmath463 with subcomplexes @xmath467 triangulating @xmath468 .
we need to show that the link of any vertex @xmath469 is a @xmath355sphere or hemisphere .
we consider the case @xmath470 .
the link of any other vertex of @xmath325 is the same in both @xmath144 and @xmath325 .
since @xmath471 is @xmath8bicollarable , we may consider a closed @xmath355invariant neighborhood @xmath472 = u_v$ ] with @xmath473 identified with @xmath308 for all @xmath474 and @xmath475 \subset m_1 $ ] .
triangulate @xmath476 in the following way : first triangulate @xmath477 $ ] .
add to it by coning @xmath478 and @xmath479 with @xmath480 and @xmath481 respectively .
let @xmath482 be this triangulation of @xmath477 \cup |\operatorname{st}_l(v)| \times \{-1,1\}$ ] .
then @xmath483 triangulates @xmath472 $ ] .
now subdivide @xmath144 so that it contains a subdivision of @xmath483 as a subcomplex .
choose a derived @xmath355subdivision @xmath260 near @xmath353 .
from the construction in the proof of [ reg3 ] , we may assume @xmath484 is an @xmath333neighborhood of @xmath353 in @xmath485 .
thus , by [ links ] , @xmath477 \cup |\operatorname{st}_l(v)| \times \{-1,1\}$ ] is @xmath355homeomorphic to @xmath486 , which we know is an @xmath86@xmath355sphere or hemisphere since @xmath463 is a @xmath8manifold .
consider the point @xmath487 .
its stabilizer in @xmath355 is all of @xmath355 .
hence , @xmath488 is an @xmath54@xmath355sphere or hemisphere .
but @xmath489 is @xmath355homeomorphic to @xmath490 , giving us that @xmath491 is the suspension of @xmath490 by @xmath31 and @xmath480 , and thus a @xmath355sphere or hemisphere . in conclusion , @xmath492 \cup |\operatorname{st}_l(v)| \times \{1\}$ ] , which is @xmath355homeomorphic to @xmath493 , is a @xmath355hemisphere .
( note that when @xmath494 is a @xmath355hemisphere , @xmath492 \cup |\operatorname{st}_l(v)| \times \{1\}$ ] is actually a quadrant of a @xmath355sphere where the two nonnegative coordinates give trivial subrepresentations .
this is easily seen to be @xmath355homeomorphic to a @xmath355hemisphere . )
now assume for the remainder of the section that @xmath495 is a polyhedron and @xmath463 is an @xmath14@xmath8manifold .
the following is a direct result of [ submanifold ] , [ regnbhd - bicollar ] , and the non - equivariant proposition 3.10 from @xcite [ reg8 ] a @xmath8regular neighborhood n of y in m is a @xmath8manifold with boundary .
if @xmath496 int m , @xmath497 where l and @xmath260 are as in the definition of regular neighborhood .
we omit the proofs of the following two results since they are exactly the same as the non - equivariant versions ( theorem 3.11 and corollary 3.18 ) in @xcite , only utilizing @xmath8bicollarability in place of collarability of boundaries of manifolds .
[ reg9 ] let y be a @xmath8polyhedron in int @xmath498 and n be a @xmath8neighborhood of y in int @xmath498 .
then n is @xmath8regular
if and only if * n is an @xmath14manifold with boundary @xmath499 @xmath8bicollarable in @xmath463 , * there are admissible @xmath8triangulations ( k , l , j ) of ( n , y , @xmath499 ) with l full in k , @xmath500 and @xmath501 [ reg10 ] suppose @xmath502 are two @xmath8regular neighborhoods of y in int m. then cl(@xmath503 ) is a @xmath8collar of @xmath504 .
let @xmath254 be polyhedra such that @xmath505 and @xmath506 where @xmath507 and @xmath508 are disks of dimensions @xmath509 and @xmath510 respectively and there is a pl homeomorphism @xmath511 with @xmath508 mapping by the identity to @xmath512 .
then we say there is an ( @xmath60dimensional ) * elementary collapse * of @xmath162 onto @xmath257 .
if there is a sequence @xmath513 of polyhedra such that there is an elementary collapse of @xmath514 onto @xmath515 , we say @xmath162 * collapses * to @xmath257 , or @xmath516 .
in particular , if @xmath517 and @xmath518 with @xmath144 collapsing simplicially to @xmath158 , then @xmath162 collapses to @xmath257 . note that the dimensions of the individual elementary collapses in a sequence are allowed to vary .
we say that a collapse is @xmath60dimensional if every elementary step has dimension @xmath519 .
consider now the case that @xmath8 is a finite group and @xmath162 and @xmath257 are @xmath8polyhedra with @xmath520 where @xmath507 ( as well as @xmath521 for each @xmath314 ) is a disk as in the definition of an elementary collapse .
suppose we also have that : ( 1 ) @xmath522 implies that @xmath523 and ( 2 ) there exists a point @xmath524 fixed by the stabilizer @xmath525 such that @xmath508 is @xmath38homeomorphic to a cone with apex @xmath526 on some @xmath38complex , and ( 3 ) @xmath507 is @xmath38homeomorphic to @xmath527 , then we say there is an * elementary @xmath8collapse * from @xmath162 to @xmath257 .
( note then that any point @xmath528 $ ] will have stabilizer @xmath529 . )
a sequence of these is called a @xmath8collapse , denoted @xmath530 .
a simplicial collapse from admissible @xmath144 to @xmath158 gives rise to a @xmath8collapse if and only if whenever @xmath228 is in its corresponding morse matching , so too is @xmath531 .
an elementary collapse from an n - manifold @xmath463 to another n - manifold @xmath465 where @xmath532 and @xmath533 and @xmath534 is called an * elementary shelling*. a sequence of such collapses is a * shelling*. note that every elementary step must be @xmath14dimensional .
the equivariant version of shelling requires some additional conditions .
let @xmath535 be @xmath14manifolds with an elementary @xmath8collapse from @xmath536 to @xmath465 such that : ( 1 ) @xmath537 lies in a @xmath8collarable subpolyhedron @xmath538 , ( 2 ) under the @xmath539triangulation @xmath540 of @xmath541 , if @xmath542 , then @xmath543 , and ( 3 ) @xmath544 implies that , in the @xmath8collar on @xmath545 , @xmath546 .
if these three extra conditions are satisfied , this elementary @xmath8collapse is called an * elementary @xmath8shelling*. a sequence of these is called a * @xmath8shelling*. [ reg11 ] if m @xmath8shells to @xmath465 , then there is a @xmath8homeomorphism @xmath547 which is the identity outside an arbitrary neighborhood of @xmath548 . as in the corresponding proof of lemma 3.25 in @xcite , we need only to consider the case of an elementary @xmath8shelling .
let @xmath536 give the elementary @xmath8shelling .
denote @xmath539 by @xmath38 .
let @xmath540 be the @xmath38triangulation of @xmath541 from the definition of elementary @xmath8collapse .
choose a @xmath8collar on @xmath549 in @xmath465 within the given neighborhood of @xmath548 .
we may consider the disk @xmath550 $ ] with @xmath551 $ ] , @xmath552 \subset m_1 $ ] , and @xmath553 .
then if @xmath554 , we have that @xmath550 $ ] may only intersect @xmath555 $ ] in @xmath556 $ ] .
we will define an @xmath38homeomorphism from @xmath550 $ ] to @xmath557 which is the identity on @xmath556 \cup |k| \times \{1\}$ ] ; such a homeomorphism can then be extended , first equivariantly to all of @xmath558)$ ] and then by the identity to the rest of @xmath463 .
this last extension is possible because either @xmath559 or @xmath560 . ]
let @xmath260 be a derived @xmath38subdivision of @xmath144 near @xmath526 .
we have that @xmath561 is @xmath38homeomorphic to @xmath562 because they are both @xmath38homeomorphic to @xmath167 with an @xmath38collar attached outside to @xmath563 .
therefore , we have an @xmath38homeomorphism from @xmath564 $ ] to @xmath565 $ ] .
coning the two polyhedra with @xmath566 and @xmath567 gives the desired @xmath38homeomorphism from @xmath568 to @xmath557 .
[ reg12 ] suppose @xmath569 are @xmath8polyhedra in a @xmath8manifold @xmath463 .
if @xmath530 , then a @xmath8regular neighborhood of @xmath162 @xmath8shells to a @xmath8regular neighborhood of @xmath257 in @xmath463 .
we follow the proof of the non - equivariant version , theorem 3.26 in @xcite , checking that the conditions of @xmath8shelling are satisfied .
the proof uses induction on the dimension of the collapse from @xmath162 to @xmath257 .
suppose that the theorem holds when the @xmath8collapse is @xmath570dimensional .
we now consider the case where there is an @xmath60dimensional elementary @xmath8collapse from @xmath162 to @xmath257 .
let @xmath520 , with @xmath571 where @xmath572 . for simplicity
, we will from now on denote the subgroup @xmath573 by @xmath38 .
let @xmath144 be an admissible @xmath8triangulation of @xmath463 with full subcomplexes @xmath574 triangulating @xmath257 and @xmath162 respectively .
denote by @xmath482 the subcomplex triangulating @xmath575 , and by @xmath576 , the resulting @xmath8triangulation of its @xmath8orbit , @xmath577 .
finally , let @xmath526 be the apex in the @xmath573cone structure of @xmath508 .
note then that @xmath578 is fixed pointwise by @xmath38 , and any point @xmath579 with @xmath580 has stabilizer exactly @xmath38 .
let @xmath581 . as in @xcite , by breaking up the collapsing into smaller steps
, we may assume that there are no vertices of @xmath150 in @xmath582 .
now we choose a derived @xmath8subdivision @xmath260 of @xmath144 near @xmath583 .
choose the derived vertices for simplices in @xmath584 in @xmath585 ensuring that @xmath308 is one of them , and denote by @xmath586 the new triangulation of @xmath162 .
then @xmath587 gives a @xmath8regular neighborhood of @xmath162 , which is the union of @xmath588 and @xmath589 , @xmath8regular neighborhoods of @xmath257 and @xmath577 respectively .
there is an @xmath570dimensional @xmath38collapse from @xmath491 to @xmath590 , so the induction hypothesis and [ reg11 ] together imply that @xmath591 is an @xmath14dimensional @xmath38disk .
let @xmath592 .
as an @xmath38disk with @xmath308 an @xmath38fixed point , @xmath557 is seen to be @xmath38homeomorphic to @xmath593 . ] we will show that if @xmath594 , the two subcomplexes must be disjoint .
for such a @xmath595 , suppose there exists a vertex @xmath596 .
( note that it must be a derived vertex since @xmath597 and @xmath482 are themselves disjoint . ) then @xmath598 contains vertices @xmath362 and @xmath31 of @xmath597 and @xmath482 respectively .
thus , @xmath599 since @xmath326 is a full subcomplex of @xmath144 , but @xmath600 is not contained in @xmath257 and it is not contained in @xmath601 since a simplex in @xmath507 is may only contain vertices from @xmath372 and @xmath482 , not @xmath597 .
this contradicts @xmath520 , so @xmath602 must be empty . since we have shown that @xmath603 implies @xmath604 , it must be true that @xmath605 .
we next prove that @xmath606 is an @xmath86disk @xmath607 which is @xmath38homeomorphic to @xmath608 , giving that @xmath557 is @xmath38homeomorphic to @xmath609 as required .
to see this , we show that @xmath607 is an @xmath38regular neighborhood of @xmath610 in the @xmath86@xmath38manifold @xmath611 , so that we may again invoke the induction hypothesis for @xmath570dimensional collapses and [ shelling - homeo ] ( since @xmath610 @xmath38collapses to @xmath308 and an @xmath38regular neighborhood of @xmath308 is the desired star of @xmath308 ) .
let @xmath168 be the subcomplex of @xmath260 triangulating @xmath610 and let @xmath612 for brevity .
the claim then is that @xmath613 .
let @xmath614 , we easily see that @xmath148 can not intersect @xmath615 or @xmath616 and must consist only of derived vertices of the form @xmath617
. then there must exist @xmath618 and @xmath619 such that @xmath620 and @xmath621 are both simplices of @xmath260 .
this implies that there exists @xmath622 for some @xmath217 containing both @xmath362 and @xmath31 .
but then @xmath623 due to the fullness of @xmath326 .
thus , @xmath624 is in @xmath168 and can be added to @xmath148 , so @xmath625 .
hence , we have @xmath626 . for the other inclusion ,
if @xmath148 is in @xmath627 , it means that there is a @xmath628 such that @xmath629 is in @xmath630 for some @xmath215 which contains vertices from both @xmath372 and @xmath482 .
we note again that @xmath148 consists only of derived vertices since it is in @xmath612 , so let @xmath248 be the minimal face such that @xmath631 . then @xmath632 , so we have that @xmath633 .
since @xmath248 was subdivided , it must contain some vertex @xmath618 .
therefore , @xmath362 may be added to @xmath148 to get a simplex of @xmath260 intersecting @xmath615 , i.e. , @xmath634 , and it is already in @xmath635 .
this proves that @xmath636 .
this finishes the proof that @xmath607 is an @xmath38regular neighborhood of @xmath610 in @xmath637 and therefore @xmath38homeomorphic to the @xmath86disk @xmath638 as explained .
observe that @xmath639 , which is @xmath8collarable in @xmath640 by [ regnbhd - bicollar ] .
there is one remaining condition to check for this to be a @xmath8shelling .
write @xmath641 .
then we must verify that @xmath642 implies that within the @xmath8collar on @xmath643 in @xmath640 , @xmath644 .
it suffices for us to show that every simplex of @xmath645 lies on @xmath464 because the @xmath8collar is given by moving derived vertices around with simplices of @xmath144 .
thus , if a simplex @xmath148 consisting only of derived vertices lies on @xmath464 , then @xmath646 lies on @xmath464 , and hence also on @xmath647 . since @xmath607 is an @xmath38regular neighborhood of @xmath648 , if @xmath642 , then @xmath649 .
thus , @xmath650 , forcing it to also belong to @xmath647 . likewise , any simplex of @xmath651 containing @xmath308 lies in @xmath652 but not in @xmath653 , forcing @xmath148 to be in @xmath464 and therefore @xmath654 .
hence , we have proven the final condition that this constitutes a @xmath8shelling of a @xmath8regular neighborhood of @xmath162 to a @xmath8regular neighborhood of @xmath257 .
[ reg13 ] if y @xmath8collapses to a point , then any @xmath8regular neighborhood of y ( in a @xmath8manifold ) is a @xmath8disk .
[ reg14 ] a collapsible @xmath8manifold is a @xmath8disk .
[ reg15 ] if @xmath655 and @xmath530 then a @xmath8regular neighborhood of x in m is a @xmath8regular neighborhood of y in m. let @xmath270 and @xmath363 be @xmath8regular neighborhoods of @xmath162 and @xmath257 respectively in @xmath463 . by [ reg12 ] , @xmath270 @xmath8shells to @xmath363 , and
so by [ reg11 ] , they are @xmath8homeomorphic fixing @xmath257 .
the homeomorphism carries any @xmath8triangulation of @xmath363 to a @xmath8triangulation of @xmath270 .
therefore , by [ reg9 ] , @xmath270 is a @xmath8regular neighborhood of @xmath257 in @xmath463 .
our final goal for this section is the following result that enables us to recognize @xmath8regular neighborhoods from @xmath8collapses .
[ reg16 ] let n be a @xmath8neighborhood of y in int m. then n is @xmath8regular if and only if * n is an @xmath14manifold with @xmath499 @xmath8bicollarable in @xmath463 , * n @xmath656 y. the proof follows exactly the non - equivariant case ( corollary 3.30 in @xcite ) , but we include it here nonetheless because this theorem is a key tool in proving our main result .
for the first implication , suppose @xmath657 is regular with @xmath144 an admissible simplicial neighborhood of @xmath158 ; then ( i ) follows from [ reg8 ] and [ regnbhd - bicollar ] .
recall the map @xmath658 from the proof of [ reg3 ] . choose @xmath358 and choose a @xmath8derived @xmath260 of @xmath144 near @xmath158 with all of the new vertices lying in @xmath336 for a given @xmath358 to obtain @xmath659 $ ] , a @xmath8regular neighborhood of @xmath257 in @xmath463 .
we have that @xmath660 , and we have a @xmath8cell structure on @xmath270 whose cells are obtained by intersecting the interior simplices of @xmath144 with @xmath661 , @xmath336 , or @xmath335 $ ] .
we may collapse , along with its orbit , each cell @xmath662 $ ] , @xmath663 from its face @xmath664 in order of decreasing dimension . that this is a @xmath8collapse follows from the admissibility of @xmath144 .
for the other implication , suppose we have @xmath140 satisfying conditions ( i ) and ( ii ) .
let @xmath665 $ ] be a @xmath8bicollar with @xmath666 .
then let @xmath667)$ ] , which constitutes a @xmath8regular neighborhood of @xmath140 in @xmath463 because we can triangulate it to be a simplicial neighborhood .
therefore , by [ reg15 ] , since @xmath668 , @xmath270 is also a @xmath8regular neighborhood of @xmath257 .
but we can define a @xmath8homeomorphism on @xmath669 fixing @xmath670 and carrying @xmath671 to @xmath672 .
we can extend this by the identity to all of @xmath463 , mapping @xmath270 to @xmath140 , showing that the latter is also a @xmath8regular neighborhood of @xmath257 in @xmath463 .
in this section , @xmath8 is the group @xmath673 unless otherwise noted . working inside @xmath115 ( @xmath8homeomorphic to the @xmath8manifold @xmath674 )
, we show that @xmath675 is a @xmath8regular neighborhood of the diagonal @xmath49 using the collapsing criterion , i.e. , we show that it is a manifold of the correct dimension and that it ( simplicially ) @xmath8collapses to the diagonal .
recall from section 1 how we represent elements of the posets in question as arrays whose entries @xmath122 are nonempty subsets of @xmath51 .
we reiterate that @xmath463 , @xmath144 , and @xmath158 are the face posets of the @xmath8regular cell complexes @xmath126 , @xmath675 , and @xmath680 respectively , and @xmath681 that of the diagonal @xmath49 . by passing to order complexes , we obtain that @xmath682 and @xmath683 are full @xmath8subcomplexes of @xmath172 , which is a full @xmath8subcomplex of @xmath684 , and they are all admissible .
our goal is to show that @xmath685 is a @xmath8regular neighborhood of @xmath686 whose boundary is @xmath687 .
we show that the link of an element of @xmath144 is a sphere or a disk of dimension @xmath689 . for any @xmath690 , @xmath691 . for any @xmath692
, we obtain an element of its lower link by deleting proper subsets from each of @xmath29 , @xmath693 , @xmath669 , and @xmath694 , at least one of which is nonempty . therefore , @xmath695 is isomorphic to the face poset of @xmath696 , yielding that @xmath697 is a combinatorial sphere of dimension @xmath698 .
( recall that , if @xmath29 is an unordered set , @xmath28 is the full simplex having @xmath29 as its vertex set , whereas , if @xmath168 is a poset , @xmath169 is its order complex . )
when @xmath699 , we show that @xmath700 is a sphere of dimension @xmath701 , yielding that @xmath702 is a sphere of dimension @xmath703 . for any @xmath704 , we have that @xmath705 so that @xmath706 .
thus , to obtain an element of the upper link of @xmath24 , any element of @xmath707 can be added to either @xmath29 or @xmath693 , but not to both , and similarly for elements of @xmath708 . as a consequence
, we have that @xmath709 is isomorphic to the face poset of @xmath710 where @xmath711 , and therefore @xmath700 is a sphere of dimension @xmath701 as claimed . in the case
where @xmath712 , we claim that @xmath700 is a disk of dimension @xmath701 , meaning that @xmath702 is a disk of dimension @xmath703 .
this will finish the proof that @xmath172 is a manifold and @xmath683 is its boundary . to see that @xmath709 is the face poset of a subcomplex of a join of spheres , we consider the various types of elements that we can add to one or more of @xmath29 , @xmath693 , @xmath669 , and @xmath694 to obtain a larger element of @xmath144 . 1
. an element of @xmath713 can not be added anywhere ( while remaining in @xmath463 ) .
the same is true for elements of @xmath714 and @xmath715 .
thus , these elements contribute nothing to the upper link .
an element of @xmath716 can be added to @xmath669 or to @xmath694 ; doing so will give us something in @xmath144 , since the element was already in @xmath717 .
thus , each of these elements contributes a copy of @xmath455 to the join of spheres .
similarly , each element of @xmath718 contributes a copy of @xmath455 to the join .
an element of @xmath719 can be added to @xmath669 or to @xmath694 , contributing a copy of @xmath720 to the join with @xmath721 indicating that the element was added to @xmath669 and @xmath722 indicating @xmath694 .
similarly , an element of @xmath723 can be added to @xmath29 @xmath724 or to @xmath693 @xmath725 .
adding elements of type @xmath726 to @xmath693 or @xmath694 could produce something not in @xmath144 .
an element of @xmath727 can be added to @xmath29 or @xmath693 ( but not both ) and , at the same time , to @xmath669 or @xmath694 ( but not both ) .
this contributes a copy of @xmath728 ( treated as a single coordinate ) to the join of spheres with the @xmath721 s corresponding to @xmath29 and @xmath669 and the @xmath722 s corresponding to @xmath693 and @xmath694 . as with type @xmath726 , adding this type of element to @xmath693 or @xmath694 could yield something not in @xmath144 . to ensure that we remain in @xmath144
, there must be an element of @xmath729 which is not added to @xmath730 . in terms of coordinates , this means there must be at least one coordinate corresponding to type @xmath726 or @xmath112 above that has no @xmath722 s . before proceeding ,
we define @xmath731 to be the subcomplex whose simplices have at least one coordinate from the join with no @xmath722 s .
( note that , as before , each copy of @xmath732 is regarded as a single coordinate . )
we will prove a lemma ( [ main2 ] ) stating that @xmath733 is a disk of dimension @xmath734 . assuming that result , since @xmath709 is isomorphic to the face poset of @xmath735 where @xmath736 , @xmath737 , and @xmath738 , we have that @xmath700 is a disk of dimension @xmath739 as we had claimed
. we proceed by induction on the dimension , @xmath741 . in the initial case
, @xmath742 has a single @xmath455 coordinate which must be @xmath721 , so it is a single point , i.e. a disk of dimension @xmath30 . to prove @xmath733 is a disk , we will show that it is a @xmath743manifold , show it collapses to a vertex , and then apply corollary [ reg14 ] .
there are four types of vertices whose links we need to consider : 1 .
@xmath721 coming from one of the @xmath744 @xmath745 coordinates has as its link @xmath746 , a @xmath747disk by induction .
@xmath722 coming from one of the @xmath745 coordinates has as its link @xmath748 , a @xmath747disk .
@xmath721 coming from one of the @xmath749 @xmath673 coordinates has as its link @xmath750 , a @xmath747sphere .
@xmath722 coming from one of the @xmath673 coordinates has as its link @xmath751 , a @xmath747disk .
now we will define a matching on @xmath733 .
first , we order the coordinates . in each @xmath745
coordinate , we also choose one of the two copies of @xmath673 to be distinguished .
associate each simplex in @xmath733 with the simplex obtained by inserting or removing @xmath721 to or from the first coordinate lacking a @xmath722 ( in the distinguished copy of @xmath673 in the case the first such coordinate is @xmath745 ) . doing this does not change
which coordinate is the first without a @xmath722 , so the pairing is well - defined .
every simplex is paired ( @xmath752 is paired with the vertex with a @xmath721 in the first coordinate and nothing in any other coordinate ) , so if there are no cycles in this matching , @xmath733 .
suppose there were a cycle .
it would have to be of the form : @xmath753 where each @xmath296 is paired with @xmath294 .
also , for @xmath754 , @xmath296 must be @xmath755 minus a vertex @xmath756 . therefore , since this is a cycle , there must be a @xmath757 such that @xmath758 .
for this to be possible , @xmath756 must be a @xmath721 .
thus , all of the simplices in the cycle must have all the same @xmath722 coordinates , but if that is the case , the vertex to be added in any @xmath759 pair is always the same , and @xmath756 must be the same for every @xmath96 .
this is a contradiction .
therefore , there are no cycles , and we have a morse matching with a single critical simplex . [ main3 ] let @xmath8 be any group and @xmath168 be a finite poset , @xmath760 an order - preserving poset map such that for any @xmath761 , @xmath762 ( or @xmath763 ) .
define @xmath630 to be the set of fixed points of @xmath227 .
then @xmath764 collapses simplicially to @xmath765 . in the case
that h is a g poset map , p @xmath656 q. we prove it for the case that @xmath762 , the other case being almost identical . since @xmath168 is finite , we may choose @xmath140 large enough so that for all @xmath308 in @xmath168 , we have @xmath766 .
now let @xmath767 be a chain @xmath768 .
if @xmath769 such that @xmath770 , let @xmath744 be the largest such @xmath96 . then we may insert @xmath771 into the chain immediately following @xmath772 because @xmath773 if @xmath774 .
associate to @xmath148 the chain obtained by inserting @xmath771 or by deleting it in the case @xmath775 .
since it is an element of @xmath630 being inserted or deleted , the selection of @xmath744 is not affected , and @xmath772 uniquely determines the other chain in the pair . therefore , this matching is well - defined .
also , this matching is equivariant if @xmath227 is a @xmath8map .
suppose there is a cycle @xmath753 we have for @xmath754 that @xmath776 for some @xmath777 .
then there must be some pair @xmath778 , so @xmath779 for all @xmath96 .
thus every simplex in the cycle has all the same elements of @xmath780 , so @xmath781 that is the greatest such element in every simplex .
hence @xmath782 for all @xmath757 , and @xmath783 for all @xmath96 .
this is a contradiction because the same element is being added and deleted in consecutive steps .
therefore , we have a morse matching whose critical simplices are exactly the elements of @xmath765 , a subcomplex of @xmath169 .
thus , @xmath169 @xmath8collapses to @xmath765 .
the collapsing will occur in three steps .
define @xmath784 @xmath785 first , we collapse @xmath172 to @xmath786 .
let @xmath148 be a chain of the form @xmath787 in @xmath172 where @xmath788 .
if @xmath789 , we want to pair @xmath148 with another chain for which that is also true .
find the last @xmath744 such that @xmath790 . then compare @xmath791 and @xmath792 to @xmath793 and @xmath794 . if @xmath795 and @xmath796 , pair @xmath148 with the chain obtained by adding to ( or deleting from ) the end of @xmath148 the element @xmath797
otherwise , find the first @xmath798 where @xmath799 or @xmath800 . now
pair @xmath148 with the chain obtained by inserting ( or removing if it equals @xmath801 ) @xmath802 before @xmath803 .
nowhere are we inserting or deleting elements with @xmath804 , so the selection of @xmath744 is not affected . in the second case , we are inserting or deleting an element with @xmath805 and @xmath806 , so the selection of @xmath749 is not affected .
therefore , the matching is well - defined .
the critical simplices are exactly those where @xmath807 , forming @xmath786 , a subcomplex . therefore ,
if there are no cycles , we have a collapsing from @xmath172 to @xmath786 . also , the pairings are chosen equivariantly , so we will have a @xmath8collapse .
suppose we have a cycle @xmath753 again , for @xmath754 , @xmath296 is obtained from @xmath755 by deleting an element @xmath808 , so there must be a pair @xmath809 coming from our matching .
therefore , @xmath810 for all @xmath96 , which means that all the simplices in our cycle have all of the same elements with @xmath129 .
thus , they all have the same @xmath811 , so @xmath791 and @xmath792 are fixed and we know that every @xmath808 has them as its second column . as a result , the elements after @xmath811 that have @xmath812 or @xmath813 are not changing as we move through the cycle , implying that @xmath808 is the same for all @xmath96 .
this is a contradiction , so our matching has no cycles .
this proves that @xmath172 @xmath8collapses to @xmath786 .
the next two collapsings are proved by lemma [ main3 ] .
for the first , we define @xmath814 by @xmath815 .
this is an order - preserving @xmath8poset map , and @xmath816 .
the fixed point set of @xmath817 is exactly @xmath43 , so lemma [ main3 ] implies that @xmath786 @xmath8collapses to @xmath818 . for the second collapsing
, we now define @xmath819 by @xmath820 .
this is an order - preserving @xmath8poset map , @xmath821 , and the fixed point set is @xmath681 .
therefore , the same lemma implies that @xmath822 @xmath8collapses to @xmath682 .
hence , @xmath172 @xmath8collapses to @xmath682 .
@xmath8 acts freely outside of @xmath686 , so @xmath823 is @xmath8bicollarable in @xmath824 .
now the theorem follows immediately from [ reg16 ] ( the collapsing criterion for @xmath8regular neighborhoods ) and propositions [ main1 ] and [ main4 ] .
the regular cell complex @xmath105 is a pl manifold with boundary @xmath104 and is equivariantly homeomorphic ( with respect to the involution described above ) to @xmath139 , where the involution on @xmath140 interchanges @xmath141 with @xmath142 .
the stiefel manifold @xmath143 is therefore equivariantly homeomorphic to @xmath104 , which is equivariantly homeomorphic to @xmath0 .
it follows from [ gcellcx ] that we have @xmath825 with the subcomplex @xmath826 . because @xmath685 and @xmath140 are both @xmath8regular neighborhoods of the diagonal , they are equivariantly homeomorphic by [ reg4 ] .
the stiefel manifold @xmath5 has a natural action of the orthogonal group @xmath83 ( with the grassmannian as the quotient ) .
the equivariant homeomorphism above is with respect to a single reflection of @xmath83 .
the multimorphism complex @xmath0 does not have a combinatorial @xmath83 action ; however , there is the induced action of the dihedral group @xmath827 ( a subgroup of @xmath83 ) which is the group of symmetries of the cycle @xmath2 .
it seems natural to ask : unfortunately , neither of the smaller models @xmath104 or @xmath828 is @xmath827invariant , so it seems that one needs to work with @xmath0 which does not have an obvious @xmath827equivariant embedding into @xmath674 .
also , a good ( equivariant ) combinatorial candidate for @xmath140 is missing , which is the obstacle to applying the methodology above to answer this question positively .
* s illman * , _ existence and uniqueness of equivariant triangulations of smooth proper @xmath8manifolds with some applications to equivariant whitehead torsion , _ j. reine angew .
math . , 524 ( 2000 )
, 129 - 183 | we prove csorba s conjecture that the lovsz complex @xmath0 of graph multimorphisms from the @xmath1cycle @xmath2 to the complete graph @xmath3 is @xmath4equivariantly homeomorphic to the stiefel manifold , @xmath5 , the space of ( ordered ) orthonormal @xmath6frames in @xmath7 .
the equivariant piecewise - linear topology that we need is developed along the way .
stiefel manifolds , pl topology , group action , graph morphism complexes 57s25 , 57q15 , 05e18 , 57q91 , 57q40 , 05e45 , 57s17 |
experimental and theoretical studies of the transport , electronic and optical properties of the armchair graphene nanoribbon ( gnr ) have attracted much attention in recent years .
one of the reason for this is that the gnrs used as the interconnects in graphene - based nanoelectronic and as the basic elements in the logic transistors could provide ultrahigh carrier mobility between the unbounded gapless 2d graphene monolayers .
however , the opened band gap in gnr reduces considerably the mobility of the carriers @xcite .
additional inevitable difficulties come from the fact that in contrast to gapless graphene monolayers , in which the bound impurity states are forbidden @xcite in gapped graphene @xcite and quasi 1d gnr @xcite the bound impurity states can be realized .
the binding energies @xmath0 of the impurity electrons in the gnr of width @xmath1 reach the considerable amount of the order of @xmath2 .
in addition , it was shown @xcite that impurity electrons , possessing energies close to the breit - wigner meta - stable resonances , contribute negligibly little to the conductance . clearly
, the impurity centres suppress strongly the mobility of the gnrs .
of course the binding effect of the impurity centres could be reduced by the technologically involved procedure of the improvent of the sample standard @xcite .
nevertheless the less elaborate mechanism of the liberation of the carriers captured by the impurities is a much more immediate demand at the present time .
the process of the ionization by an external electric field can be used as an instrument for the release of the blocked carriers . besides , the quasi-1d structures , in particular the bulk semiconductors subject to strong magnetic fields @xcite , quantum wires ( qwrs ) @xcite and armchair gnrs @xcite are favorable for the formation of both the strictly discrete and meta - stable ( fano resonances ) @xcite impurity and exciton states adjacent to the ground and excited size - quantized energy levels , respectively . the latter are caused by the confinement effect associated with the magnetic field in bulk semiconductors and the boundaries of the qwrs and gnrs .
the nature of the fano resonances comes from the inter subband coupling between the discrete and low - lying continuous rydberg states . with emergence of the electric field only one channel of the ionization
is opened for the ground series of the rydberg discrete states , while the excited meta - stable states decay into two channels : the autoionization channel , open due to the inter - subband fano coupling , and the channel of the electric field ionization , related to the under - barrier tunneling ( fig .
the interaction of these two channels of the ionization is of immediate interest .
( color online ) .
the combined potential @xmath3 , formed by the impurity potential @xmath4 ( [ e : pot ] ) and electric field @xmath5 adjacent to the ground and first excited size - quantized energy levels @xmath6 ( [ e : ener ] ) with @xmath7 , respectively .
the continuous and discrete rydberg states are labeled by the indexes @xmath8 and @xmath9 ( ground state ) and @xmath10 ( first excited state ) , respectively.,width=325 ] it is clear that the problem of the impurity states in the armchair gnr in the presence of a longitudinal electric field directed parallel to the gnr axis is important on account of two aspects : ( i ) its considerable interest in the context of basic research , and ( ii ) possible nano - electronic applications .
narrow gnrs of several nanometers width are the best candidates among the other 1d structures for fundamental studies .
the binding energies of the impurity electron in gnr exceed those in the corresponding semiconductor structures by a factor of about @xmath11 that in particular manifests itself in the strong electric fields , providing the complete ionization of the impurity states in gnr .
gnrs seem to be a unique structure in which both channels of the auto- and electric field ionization are opened .
the process of electric field ionization transforms the strictly bound and the fano resonant states into states of transporting carriers that in turn improve the conductance properties of the gnrs and of the nano - electronic devices into which these ribbons are incorporated .
the finite lifetime of the quasi - discrete impurity states associated with the two - channel ionization should be taken into account in the practical use of the gnrs exposed to the electric field .
there are two comments in order .
first , the theoretical approaches to this problem are mostly based on numerical calculations ( density functional theory and bethe - salpiter equation @xcite , nonorthogonal tight - binding model @xcite , tight - binding scheme and hartree - fock approximation @xcite ) requiring significant computational efforts . only a few recent works
elaborate on analytical methods . in ref .
@xcite the bound and quasi - discrete impurity states in the armchair gnr have been studied by solving the dirac equation for a massless neutrino .
ratnikov and silin @xcite empirically extended to the gnr the model earlier developed for the semiconductor qwr @xcite , and calculated the excitonic energy levels by the variational method and their red shift induced by the electric field .
second , to our knowledge analytical results based on the dirac equation adequately describing the impurity electrons in gnrs subject to external electric field are not present in the literature .
thus an analytical approach to the problem of the impurity states in biased armchair gnr is desirable .
particularly it renders the basic physics transparent and governs the electronic , optical and transport properties of the graphene based devices . in the present work we develop an analytical approach to the problem of the impurity state in the narrow armchair gnr in the presence of an external electric field directed parallel to the ribbon axis .
the coulomb impurity attraction is taken to be much weaker than the influence of the ribbon confinement and much stronger than the effect of the electric field .
the impurity centre can be positioned anywhere within the gnr .
the 2d dirac equation for the massless neutrino subject to the coulomb and external uniform electric field is solved in the adiabatic approximation .
this approximation implies the transverse motion of the electron governed by the ribbon confinement to be much faster than the longitudinal motion controlled by the impurity and external electric field .
our approach is based on the matching of the wave functions in the intermediate regions .
the latter separates the impurity interaction from the electric field interaction dominated regimes . in the approximation of the isolated size - quantized subbands
the binding energies and widths of the quasi - discrete states as a function of the ribbon width , position of the impurity and the electric field are calculated in explicit form . also , the phases of the functions of the continuous spectrum specifying the s - matrix are derived . in the double - subband approximation
the total widths of the first excited rydberg series of impurity states , associated with the ionization effect of the electric field and inter - subband fano coupling are calculated .
also the capturing of the electron by the impurity potential for the lifetime determined by the electric field is explored .
numerical estimates made for realistic gnr show that for narrow ribbons the impurity states in the presence of a weak electric field remain quite stable which is to be proven experimentally , while significantly strong field could unlock the captured electrons .
the aim of this work is to clarify the ionization mechanism of the release of the strictly bound and quasi - bound impurity electrons yielding the increase of the mobility of the carriers in the gnr .
this work is organized as follows . in section 2 the general analytical approach
is described . the complex quasi - discrete energy levels dictating the binding energies and energy widths caused by the electric field along with the phases of the wave functions of the continuous states
are calculated in the single - subband approximation in section 3 .
the combined effect of the autoionization of the fano resonant states and their ionization by the electric field is under consideration in section 4 . in section 5
we discuss the obtained results and estimate the expected experimental values .
section 6 contains the conclusions .
we consider a ribbon of width @xmath12 placed in the @xmath13 plane and bounded by the lines @xmath14 the impurity centre of charge @xmath15 is shifted from the mid - point of the ribbon @xmath16 by the distance @xmath17 the equation describing the impurity electron at a position @xmath18 subject to the external uniform electric field @xmath19 possesses the form of a dirac equation @xmath20 where the hamiltonian @xmath21 relevant to the inequivalent dirac points + @xmath22 is the graphene lattice constant ) is given by @xcite @xmath23 + \left[v(\vec{\rho } ) - efy\right]\hat{{\rm i}};\ ] ] containing the pauli matrixes @xmath24 , the graphene parameter @xmath25 , the unit matrix @xmath26 and the 2d coulomb impurity potential @xmath27 here @xmath28 is the effective dielectric constant determined by the static dielectric constant @xmath29 of the substrate @xcite and by the parameter @xmath30 the envelope wave function four - vector @xmath31 @xmath32 consists of the wave functions @xmath33 describing the electron states in the sublattices @xmath34 and @xmath35 of graphene in the vicinity of the dirac points @xmath36 , respectively .
the boundary conditions require the total wave function to vanish at both edges for each sublattice @xcite @xmath37 by solving eq .
( [ e : basic ] ) the components @xmath38 of the total wave vector @xmath39 subject to the boundary conditions ( [ e : bound ] ) can be found .
following the procedure presented in details in ref .
@xcite we expand the wave functions @xmath38 in a series @xmath40 in which @xmath41 \right ) \right ] \right \}\end{aligned}\ ] ] are the components of the orthonormal @xmath42-vector wave function relevant to the transverse confined @xmath43-motion of the free electron with the size - quantized energies @xmath44.\ ] ] below for estimates we take the gnrs of the family @xmath45 providing along with @xmath46 the semiconductor - like gapped structure , leaving aside @xmath47 , corresponding to the metallic - like gapless ribbon .
this leads to the set of the equations for the coefficients @xmath48 @xmath49 @xmath50,\ ] ] with eq .
( [ e : coulomb ] ) for the potential @xmath51 . at @xmath52 @xmath53~;\,s=0,1,2,\ldots ; \ ] ] below we solve the set ( [ e : set ] ) in the adiabatic approximation .
the longitudinal @xmath54-motion , governed by the quasi - coulomb potentials @xmath55 slightly perturbed by the electric field @xmath5 , is assumed to be much slower than the transverse @xmath43-motion affected by the boundaries of the narrow ribbon .
the relevant parameters are the strength of the impurity potential scaled to that of the graphene @xmath56 , the impurity bohr radius @xmath57 , the quantum number of the bound impurity state @xmath58 and the dimensionless electric field @xmath59 , which is the external electric field @xmath5 scaled to the impurity electric field @xmath60 .
they are defined by @xmath61 the other parameters @xmath62 and @xmath63 - are the first and second quasi - classical turning points calculated from @xmath64 , where @xmath65^{1/2};\ ] ] is the quasi - classical momentum .
further we impose the conditions @xmath66 meaning the narrowness of the ribbon @xmath67 ( at any rate for the low excited size - quantized @xmath42 subbands ) i.e. the smallness of the impurity effect comparatively to that of the confinement , and @xmath68 providing the weakness of the external electric field @xmath5 relatively to the impurity electric field in the state with quantum number @xmath58 . under these conditions the relationships @xmath69 are valid .
at the first stage we neglect the coupling between the states associated with the subbands of different @xmath70 it follows from eq .
( [ e : onoff ] ) , that in the narrow ribbon of small width @xmath12 the diagonal potentials @xmath71 dominate the off - diagonal terms which allows in turn allows to take @xmath72 and then to decompose the set ( [ e : set ] ) into independent equations with the potentials @xmath73 @xmath74 the set ( [ e : set ] ) for @xmath75 is solved by matching in the intermediate regions the two - vectors @xmath76 valid in the inner region @xmath77 , coulomb region @xmath78 and in the `` electric '' region @xmath79 @xcite . in the inner and coulomb regions the impurity electric field @xmath60 considerably exceeds the external uniform field @xmath5 , while in the `` electric '' region the potentials @xmath4 can be treated as a small perturbation to the effect of the field @xmath5 . _ inner region
_ + + in this region an iteration procedure is employed .
the subsequent integration of the set ( [ e : set ] ) , in which we keep only diagonal potentials @xmath4 ( [ e : limit ] ) and take arbitrary constants for the trial functions @xmath81 , gives for the even states @xmath82 in the intermediate region @xmath83 @xcite @xmath84 where @xmath85 and @xmath86 is an arbitrary constant phase .
+ + _ coulomb region _
+ + in this region the wave two - vector @xmath87 can be written in the form @xmath88 where @xmath89 are the vectors increasing and decreasing , respectively at @xmath90 , and where @xmath91 are the corresponding arbitrary constants .
the components determining the vector @xmath92 , have been calculated in ref .
@xcite in terms of the exact solutions to eqs .
( [ e : set ] ) at @xmath93 @xmath94,\ ] ] where @xmath95 and where @xmath96 is the whittaker function having the asymptotics @xmath97 @xcite
. the function @xmath98 can be obtained from eq .
( [ e : outer ] ) by replacing @xmath99 .
the wave functions @xmath100 , corresponding to the vector @xmath101 , are derived from the functions @xmath102 , respectively by replacing @xmath103 is the whittaker function having the asymptotics @xmath104 @xcite . at @xmath105 @xmath106 @xmath107 where @xmath108 with @xmath109 in eq .
( [ e : theta ] ) @xmath110 is the euler constant and @xmath111 is the @xmath112-function . at @xmath113
@xmath114 @xmath115;\\ v_{nc(+)}^{(2)}(y)=\frac{q}{2\eta^2 } \left [ \frac{{\rm e}^{-\rm i\pi\eta}}{\gamma ( \eta)}\tau^{\eta}{\rm e}^{-\frac{\tau}{2}}- \frac{1}{\gamma ( -\eta)}\tau^{-\eta}{\rm e}^{+\frac{\tau}{2 } } \right ] ; \end{array } \right \}\end{aligned}\ ] ] _ `` electric region '' _ + + the problem of the relativistic electron in the presence of a uniform electric field has been studied initiatively by sauter @xcite . using the original notations @xmath116 the set ( [ e : set ] ) for the functions @xmath117 reads
@xmath118 v_{n\mbox{el}}^{(1)}(\xi)=0 ~;\\ v_{n\mbox{el}}^{(2)}(\xi ) = \frac{1 } { k_0 + \xi + \frac{q}{\xi-\xi_0}}v_{n\mbox{el}}^{(1)'}(\xi)~ ; \end{array } \right \}\end{aligned}\ ] ] using the relationships @xmath119 induced by the conditions ( [ e : weakimp ] ) and ( [ e : weakel ] ) and setting @xmath120~ , x=(2k_0)^{\frac{1}{3}}\sqrt{\frac{ef}{p}}(y - y_2);\quad x\ll k_0^{\frac{4}{3}};\end{aligned}\ ] ] we obtain from eq .
( [ e : setel ] ) @xmath121 where @xmath122 eq . ( [ e : gfunct ] ) is solved by the method of a comparison equation @xcite successfully employed in ref . @xcite in which the impurity and exciton in a biased quantum wire have been studied
. the key point of this method is the replacements of the coefficient @xmath123 and the function @xmath124 by others which transform eq .
( [ e : gfunct ] ) into an exactly analytically solvable comparison equation ( see refs . @xcite and @xcite for details ) . the solutions to eq .
( [ e : gfunct ] ) @xmath125 are written in terms of the airy functions @xmath126 and @xmath127 @xcite @xmath128^{\frac{1}{6}}}{g(x)^{\frac{1}{4 } } } ai\left[\left(\frac{3}{2}s(x ) \right)^{\frac{2}{3 } } \right]~;\quad \varphi_2 ( x ) = \frac{\left[\frac{3}{2}s(x ) \right]^{\frac{1}{6}}}{g(x)^{\frac{1}{4 } } } bi\left[\left(\frac{3}{2}s(x ) \right)^{\frac{2}{3 } } \right]~;\ ] ] where @xmath129 at @xmath130 resulting in @xmath131 , the asymptotic expansions for @xmath132 @xcite in eqs .
( [ e : airy ] ) give for the functions @xmath133 ( [ e : subst ] ) and @xmath134 ( [ e : setel ] ) @xmath135~;~ v_{n\mbox{el}}^{(1)(+)}(y ) = \pi^{-\frac{1}{2}}x_2^{-\frac{1}{4}}\frac{1}{2}\exp \left [ - s(y ) \right]~ ; \nonumber \\
v_{n\mbox{el}}^{(2)(\mp)}(y ) & = & \mp \frac{q^2}{4\eta^2}v_{n\mbox{el}}^{(1)(\mp)}(y)~ ; \quad x_2 = ( 2k_0)^{\frac{1}{3}}\sqrt{\frac{ef}{p}}y_{2}~;\end{aligned}\ ] ] where @xmath136 the components @xmath137 in eqs . (
[ e : elcomp ] ) determine the two - vector @xmath138 in the region @xmath139
@xmath140,\ ] ] where @xmath141 is an arbitrary constant .
note , that in the region @xmath142 the vector state ( [ e : genel ] ) with eqs .
( [ e : setel ] ) , ( [ e : subst ] ) , ( [ e : airy ] ) , possesses the asymptotics of the outgoing wave @xmath143 \right\}.\ ] ] on equating in the intermediate region @xmath83 the two - vectors @xmath144 and @xmath145 ( [ e : coulvec ] ) with the components ( [ e : iter ] ) and ( [ e : outer1 ] ) , ( [ e : outer2 ] ) for the vectors @xmath144 and @xmath145 , respectively , we obtain @xmath146 with @xmath147.\ ] ] taking in eq.([e : theta ] ) @xmath148 , the function @xmath149 reads in an explicit form @xmath150 + \ln z + \psi\left ( 1 + \frac{q}{z}\right ) + x_n~ ; \\
x_n & = & \ln \left\{\frac{|n - \tilde{\sigma}|\pi } { 2}\sqrt{1-s^2}\exp\left [ \frac{s}{2}\ln\frac{1+s}{1-s } \right ] \right\ } + 2c - 1~ ; \nonumber\end{aligned}\ ] ] in eq . (
[ e : y ] ) @xmath151 and @xmath152 is the logarithmic derivative of the @xmath153-function . in an effort to make the further results more readable and transparent
, we utilize the logarithmic approximation @xmath154 , which transforms eq .
( [ e : y ] ) into @xmath155 for the @xmath156 function and for its derivative we obtain @xmath157 a comparison in the other intermediate region @xmath158 the coulomb vector @xmath145 ( [ e : coulvec ] ) , ( [ e : outer1 ] ) , ( [ e : outer2 ] ) and `` electric '' vector @xmath138 ( [ e : genel ] ) , ( [ e : elcomp ] ) yield @xmath159 where @xmath160 on solving the set ( [ e : eq1 ] ) , ( [ e : eqel ] ) by the determinantal method we arrive at the equation for the complex quantum numbers @xmath58 and complex energies @xmath161 @xmath162\phi_{\eta}^2 + { \rm i } = 0.\ ] ] the quantum numbers @xmath163 of the strictly discrete states related to the zero electric field @xmath164 can be found from equation @xmath165 with eq .
( [ e : y1 ] ) for the @xmath156 function . on expanding this function in eq .
( [ e : compleq ] ) in the vicinity of the quantum numbers @xmath166 and taking into account the derivative ( [ e : der ] ) we calculate the complex quantum numbers @xmath167 which in turn determine the quasi - discrete energy levels @xmath168 where the energy width @xmath169 replacing the vector ( [ e : genel ] ) by the `` electric '' vector @xmath138 @xmath170,\ ] ] where @xmath141 and @xmath171 are the arbitrary constant and phase , respectively ,
we obtain @xmath172\gamma(1-\eta)\phi_{\eta}^2 \right\}^{-1}\ ] ] the results of this subsection summarized in eqs .
( [ e : elen ] ) , ( [ e : width ] ) , and ( [ e : cot ] ) are valid under the conditions ( [ e : weakimp ] ) as well as @xmath173 and ( [ e : weakel ] ) .
_ inner region _ + + as above the wave functions , corresponding to the inner region are given by eqs .
( [ e : iter ] ) .
+ + _ coulomb region _ + + in the region @xmath175 , where @xmath176 , the two - vector @xmath145 reads @xmath177 where @xmath178 is an arbitrary phase . the arbitrary constants analogous to those in eqs .
( [ e : genel ] ) and ( [ e : stand ] ) do not contribute to the results of this paragraph and are therefore omitted .
the components of the vectors @xmath179 were calculated in ref .
@xcite in terms of the exact solutions to eqs .
( [ e : set ] ) at @xmath180 and @xmath181 . in particular @xmath182,\ ] ] where @xmath183 the function @xmath184 can be obtained from eq .
( [ e : outcont ] ) by replacing @xmath185 by @xmath186 and @xmath56 by @xmath187 . for @xmath188
the components @xmath189 of the vector @xmath145 ( [ e : coulcont ] ) become @xmath190 while the function @xmath191 can be obtained from eq .
( [ e : outcont1 ] ) by replacing @xmath192 and @xmath193 by @xmath194 . in eq .
( [ e : outcont1 ] ) @xmath195,~ c_n = \frac{\pi}{2}\left(1 + \coth \frac{\pi q}{\varphi_{n}}\right).\ ] ] in the region @xmath196 the vector @xmath145 ( [ e : coulcont ] ) , ( [ e : outcont ] ) is determined by the components @xmath197 and @xmath191 , calculated from eq .
( [ e : outcont2 ] ) by replacing @xmath198 by @xmath199 and @xmath200 by @xmath201 with @xmath202 _ `` electric '' region _
+ + at the same time the `` electric '' two - vector @xmath203 where @xmath204 is an arbitrary phase , is written in terms of the two - vectors @xmath205 calculated analogously to the vectors @xmath206 ( [ e : elcomp ] ) incorporated into eq .
( [ e : genel ] ) . as a result
the components @xmath117 of the vector @xmath138 ( [ e : electr ] ) in the region @xmath207 become @xmath208;\quad v_{n\mbox{el}}^{(2)}= \tan\frac{\varphi_{n}}{2 } \cos \left[\tilde{s}(y ) + \frac{\pi}{4 } + \vartheta \right],\ ] ] where @xmath209 on matching the wave - vectors @xmath144 ( [ e : iter ] ) and @xmath87 ( [ e : coulcont ] ) in the intermediate region @xmath210 we obtain for the phase @xmath178
@xmath211 where @xmath212.\ ] ] a comparison in the other intermediate region @xmath213 the wave vectors @xmath87 ( [ e : outcont2 ] ) and @xmath138 ( [ e : electr1 ] ) yield @xmath214 equation ( [ e : cotcont ] ) allows to calculate the phase @xmath178 as a function of the energy @xmath161 . as expected setting @xmath215 and matching the functions ( [ e : coulcont ] ) , ( [ e : outcont ] ) taken at @xmath216 with the iteration functions ( [ e : iter ] ) and then with the `` electric '' functions ( [ e : electr1 ] ) at @xmath217 , we obtain the equation ( [ e : cot ] ) for the @xmath218 . employing the equation @xmath219 ,
determining the poles of the @xmath220-matrix @xmath221 @xcite , we arrive at eq .
( [ e : compleq ] ) for the quasi - discrete energy levels .
note , that the wave - vector @xmath138 ( [ e : electr ] ) has at @xmath222 the asymptotic form of the standing wave with the components @xmath223 ; \nonumber\\ v_{n\mbox{el}}^{(2)}(y)&= & \tan\frac{\varphi}{2 } \cos \left[\frac{2^{2/3}}{3 } \left(\frac{f}{f_0}\right)^{1/2 } \left(\frac{y}{a_0}\right)^{3/2 } + \frac{\pi}{4 } + \vartheta \right ] ; \end{aligned}\ ] ] the main result ( [ e : cotcont ] ) of this subsection is valid under the conditions @xmath224 and @xmath225 .
in this section we consider the coupling between the continuous states branching from the ground size - quantized energy level @xmath226 and discrete states adjacent to the energy level @xmath227 , having the common energies @xmath228 .
the corresponding four - fold set can be derived from the set ( [ e : set ] ) limited by @xmath229 .
+ + _ continuous states _ @xmath230 + + in the inner region @xmath231 the above
described iteration procedure leads to the components @xmath232 of the vector @xmath233 @xcite @xmath234 in this set @xmath235 is given by eq .
( [ e : iter ] ) , @xmath236 and @xmath237 are the corresponding arbitrary constants and phases , respectively . the parameter @xmath238 \nonumber\\ & + & \sin\alpha_0 \left [ { \rm { si}}\left ( \frac{\pi}{2 } + \alpha_0\right)+ { \rm { si}}\left ( \frac{\pi}{2 } -\alpha_0\right ) \right],~\alpha_0=\frac{\pi x_0}{d}.\end{aligned}\ ] ] consisting of the integral sine @xmath239 @xcite , describes the coupling induced by the potentials @xmath240 ( [ e : pot ] ) . in this region
the components @xmath241 of the coulomb vector @xmath242 can be calculated from eq .
( [ e : outcont1 ] ) for @xmath230 . in the `` electric '' region @xmath243 the components @xmath241 of the coulomb vector @xmath242 coincide with those presented in eq .
( [ e : outcont2 ] ) , while the wave functions @xmath244 relevant to the `` electric '' vector @xmath245 are given by eq .
( [ e : electr1 ] ) . matching in the inner region the wave vectors @xmath233 ( [ e : set0 ] ) with @xmath246 and @xmath242 ( [ e : outcont1 ] ) we arrive at @xmath247 - r_1\gamma_{01}\cot \omega = 0,\ ] ] where @xmath248 is defined in eq .
( [ e : t ] ) . on equating in the `` electric '' region the coulomb vector @xmath242 ( [ e : outcont2 ] ) and `` electric '' vector @xmath245 ( [ e : electr1 ] ) the relationship ( [ e : link ] ) between the phases @xmath178 and @xmath204 of the coulomb and `` electric '' wave - vectors , respectively , is obtained .
+ + _ discrete states _
@xmath249 + + in the inner region @xmath250 the components @xmath251 of the wave - vector @xmath252 are obtained from the wave functions @xmath232 ( [ e : set0 ] ) , respectively by replacing @xmath253 and @xmath254 .
the coulomb wave - vector @xmath255 is defined by the components @xmath256 ( [ e : outer1 ] ) and @xmath257 ( [ e : outer2 ] ) of the wave - vectors @xmath258 in eq .
( [ e : coulvec ] )
. in the `` electric '' region @xmath259 the corresponding wave functions @xmath256 and @xmath257 have the form ( [ e : outer3 ] ) and ( [ e : outer4 ] ) , respectively .
the `` electric '' wave vector @xmath260\ ] ] formed by the vectors @xmath261 , having the components ( [ e : elcomp ] ) for @xmath249 , gives for the @xmath262 @xmath263 + \frac{1}{2}\cos\vartheta\exp \left[- s(y ) \right ] \right)~;\\ v_{1 \mbox{el}}^{(2)}(y)=c_1 \pi^{-\frac{1}{2}}x_2^{-\frac{1}{4 } } \frac{1}{2 } \frac{q}{\eta } \left(-\sin\vartheta\exp \left[s(y ) \right]+ \frac{1}{2}\cos\vartheta\exp \left[- s(y ) \right ] \right)~ ; \end{array } \right \}\end{aligned}\ ] ] with eq .
( [ e : momas ] ) for @xmath264 . on equating in the inner region the wave vectors @xmath252 and @xmath255 calculated from eqs . ( [ e : set0 ] ) and ( [ e : coulvec ] ) , ( [ e : outer1 ] ) , ( [ e : outer2 ] ) , respectively ,
we obtain @xmath265 where @xmath149 is given by eq .
( [ e : y1 ] ) . a comparison in the `` electric '' region for the coulombic @xmath255 ( [ e : coulvec ] ) , ( [ e : outer3 ] ) , ( [ e : outer4 ] ) and `` electric '' @xmath266 ( [ e : elreal ] ) wave vectors leads to the set @xmath267 with eq .
( [ e : phi ] ) for @xmath268 .
the total set of eqs .
( [ e : setcoupl ] ) , ( [ e : eqel1 ] ) and ( [ e : eqel2 ] ) for the coefficients @xmath269 and @xmath270 being solved by the determinantal method gives @xmath271 \left[y(\eta)-\frac{\gamma_{01}^2}{t(k)+ \frac{c_0}{\cot\omega}}\right ] -1 = 0,\ ] ] where @xmath272 $ ] ( [ e : y1 ] ) , @xmath273 $ ] ( [ e : t ] ) , @xmath274 $ ] ( [ e : outcont1 ] ) are introduced above .
the phases @xmath178 and @xmath204 are linked by eq .
( [ e : link ] ) . by solving eq .
( [ e : cotcoupl ] ) the phases @xmath178 and @xmath204 as a function of the energy @xmath161 can be found in principle . as expected the general eq .
( [ e : cotcoupl ] ) describes the limiting cases studied above for negligibly small coupling @xmath275 or electric field @xmath276 .
equation @xmath277 with @xmath278 calculated from eq .
( [ e : cotcoupl ] ) at @xmath279 coincides with eq .
( [ e : compleq ] ) derived in the approximation of the isolated subbands . equation @xmath280 with @xmath281 taken from eq .
( [ e : cotcoupl ] ) at @xmath282 transforms into that describing the fano resonances in the double - subband approximation @xcite .
the complete energy width @xmath283 caused by both mechanisms of ionization can be derived by setting @xmath277 in eqs .
( [ e : link ] ) and ( [ e : cotcoupl ] ) and then expanding @xmath149 ( [ e : y1 ] ) in the vicinity of the quantum numbers @xmath284 of the strictly discrete states for which @xmath285 . the complex quantum numbers @xmath286 calculated from eq .
( [ e : cotcoupl ] ) determine the quasi - discrete energy states @xmath287 including the complete energy width @xmath288 where the width @xmath289 induced by the electric field is given by eq .
( [ e : width ] ) . for the width @xmath290 of the fano resonances we obtain @xmath291 where the quantum defects @xmath292 n=0,1,2 , can be calculated from eq .
@xmath165 using eq .
( [ e : y1 ] ) at @xmath249 for @xmath149 .
this point is suitable to demonstrate one of the possible applications of the obtained results . since the breit - wigner resonant scattering on the quasi - discrete state caused by the inter - subband coupling has been considered in ref .
@xcite below we focus on the effect of the resonant capturing of the electron induced by the electric field .
the electron density within the coulomb well is determined by the coefficient @xmath293 in the wave vector ( [ e : coulvec ] ) , growing towards the impurity centre .
the electron density @xmath294 related to the ground size - quantized energy level @xmath226 can be obtained from eqs .
( [ e : eqel1 ] ) , ( [ e : eqel2 ] ) , ( [ e : cotcoupl ] ) at @xmath279 and @xmath230 . using the function @xmath295 derived from eq .
( [ e : compleq ] ) and then expanded in a series in the vicinity of the resonant energy level @xmath296 and the coefficient @xmath141 providing the unit flux density of the waves in eq .
( [ e : stand ] ) , the electron density reads @xmath297 where @xmath298 in eq .
( [ e : dens ] ) @xmath299 is the probability of the resonant capturing of the ingoing electron within the impurity region for a lifetime @xmath300 before being transformed into an outgoing wave .
( [ e : dens ] ) is completely in line with the results obtained for the @xmath301function @xcite and 3d rectangular @xcite potential barriers .
_ single subband approximation _ + + we define the binding energy of the electron @xmath302 of the impurity electron in the @xmath303 quasi - discrete state associated with the @xmath42 subband as the difference between the size - quantized energy @xmath304 ( [ e : ener ] ) of the free electron and the real part of the the energy of the impurity electron @xmath305 given by eqs .
( [ e : elen ] ) , and ( [ e : encompl ] ) , yielding @xmath306 .
the dependencies of the binding energy on the ribbon width @xmath12 and the displacement of the impurity centre @xmath307 from the mid - point of the ribbon @xmath16 were discussed in detail in ref .
here we only mention that the binding energy decreases with increasing ribbon width @xmath308 and with shifting the impurity from the ribbon centre towards the boundaries .
note , that we ignore the small effect of the electric fields on the binding energy . in order to calculate the corrections @xmath309 and @xmath310 to the non - relativistic energy @xmath311 ( [ e : elen ] ) and the wave function @xmath312 ( [ e : outer ] ) , respectively , caused by the electric field
@xmath313 , we trivially solved the equation @xmath314 by setting @xmath315 and @xmath316 to find @xmath317 the obtained red shift of the energy level @xmath311 coincides completely with that calculated by ratnikov and silin @xcite by the dalgarno - lewis perturbation theory method @xcite .
for the gnr of width @xmath318 placed on the sapphire substrate @xmath319 and exposed to the electric field @xmath320 the relative shift of the binding energy of the ground impurity state @xmath321 is negligibly small .
( color online ) .
the width @xmath322 ( [ e : width ] ) of the ground impurity state @xmath323 in the gnr placed on the @xmath324 substrate @xmath325 as a function of the electric field @xmath5 and of the graphene width @xmath12 for the different impurity positions @xmath326 with @xmath327.,width=325 ] the main effect of the electric field is the ionization of the impurity states which are accompanied by the emergence of the energy widths @xmath328 .
it follows from eq .
( [ e : width ] ) that with increasing ribbon width @xmath12 and strength of the electric fields @xmath5 the width of the quasi - discrete impurity states increases .
however , the greater the shift of the impurity centre @xmath307 from the mid - point @xmath16 is the wider the impurity state becomes .
this means that in contrast to quasi-1d semiconductor structures ( qwr , bulk material subject to a magnetic field ) in which the ionization of the impurity centre is reached only by the increasing electric field , in the gnr the mechanism of the dimensional ionization can be realized .
the electric field could be kept constant , while the widening of the ribbon and the displacement of the impurity would lead to the ionization .
note , that the dimensional ionization is more efficient as compared to the electric ionization , because the argument of the exponent function in eq .
( [ e : phi ] ) changes @xmath329 with changing the ribbon width @xmath12 , and @xmath330 with changing the electric field @xmath5 .
the width @xmath322 of the ground impurity state adjacent to the ground size - quantized level @xmath230 as a function of the ribbon width @xmath12 and electric field @xmath5 for the different impurity positions @xmath307 is depicted in fig.2 .
iso - width lines @xmath331 providing the width @xmath332 and given in fig.3 evidently follow from eq .
( [ e : width ] ) .
( color online ) .
the iso - width curves @xmath333 calculated from eq .
( [ e : width ] ) for the ground impurity state @xmath334 in the gnr placed on a @xmath324 substrate @xmath325
. the impurity is positioned at the mid - point @xmath335 of the gnr of width @xmath12 ; @xmath5 is the strength of the electric field .
the energy widths are taken to be @xmath336.,width=325 ] fig.4 demonstrates the iso - width surfaces when all parameters @xmath337 are changed . in the gnrs
the effects of both parameters @xmath5 and @xmath12 are governed by the exponential factor ( [ e : phi ] ) in contrast to the semiconductor qwr in which the dependence on the radius of the qwr @xmath338 has the less pronounced logarithmic character @xcite .
equations ( [ e : width ] ) and ( [ e : phi ] ) show that the impurity electron becomes practically unbound if the electric field @xmath5 exceeds the critical value @xmath339 .
thus , the ground state @xmath340 is less sensitive to the ionizing effect of the electric field and remains stable up to the significantly greater electric fields @xmath341 than those @xmath342 , destroying the excited states @xmath343 .
( color online ) .
the iso - width surfaces @xmath344 calculated from eq .
( [ e : width ] ) for the ground impurity state @xmath334 in the gnr of width @xmath12 placed on the @xmath324 substrate @xmath325 and exposed to the electric field @xmath5 . the impurity centre is positioned at the coordinate @xmath326 .
the energy widths are taken to be @xmath345.,width=325 ] _ double subband approximation _ + + the double subband @xmath7 approximation describes the combined effect of the electric field ionization and inter - subband autoionization . since the influence of the electric field was discussed just above here we briefly remind the reader of the sequences of the inter - subband interaction . at @xmath181 the rydberg series of the strictly discrete energy levels adjacent to the excited size - quantized energies @xmath346 transform into the quasi - discrete levels ( fano resonances ) of widths proportional to @xmath347 and increasing both with decreasing the ribbon width @xmath12 and with the displacement of the impurity centre from the mid - point of the ribbon @xmath16 @xcite .
the combined effect of the both types of the ionization reflected in eqs .
( [ e : link ] ) and ( [ e : cotcoupl ] ) leads to the summation of the widths @xmath328 and @xmath348 , associated with the electric filed and fano mechanisms , respectively ( [ e : sum ] ) .
the energy widths @xmath349 and @xmath350 as a function of the width @xmath12 for the different strengths @xmath5 of the electric field and for the impurity positioned at @xmath351 are presented in fig.5 .
( color online ) .
the dependence of the widths @xmath349 ( [ e : width ] ) and @xmath350 ( [ e : fano ] ) of the ground state @xmath9 of the first excited size - quantized rydberg series @xmath249 , caused by the electric field @xmath5 and the inter - subband coupling , respectively , and the binding energy @xmath352 on the width @xmath12 of the gnr placed on the sapphire substrate @xmath353 .
the impurity centre is positioned at the coordinate @xmath354 .
the electric fields are taken to be @xmath355 .
the corresponding critical ribbon widths @xmath356 provides the balance between the widths @xmath349 and @xmath350.,width=325 ] clearly , @xmath328 and @xmath348 change with changing @xmath12 in the opposite way . as a result in narrow gnrs the widening effect of the fano coupling
exceeds that of the electric field , but with increasing the ribbon width both effects come into balance and then the electric field ionization dominates the autoionization .
the greater is the electric field @xmath5 the less the critical width @xmath356 becomes , providing the equality between the both widths . the parameters @xmath356 and @xmath5 obey the relationship @xmath357 , following from the condition @xmath358 , in which the widths @xmath328 and @xmath348 are given by eqs .
( [ e : width ] ) and ( [ e : fano ] ) , respectively for @xmath359 .
since the fano coupling does not contribute to the most interesting ground impurity series @xmath230 and does not manifests itself in not significantly narrow gnrs exposed to sufficiently strong electric fields , we clarify below the mechanism of the resonant capturing of the electron by the impurity centre in the presence of the electric fields .
( [ e : eq1 ] ) , ( [ e : eqel ] ) show that the ratio @xmath360 for the energy of the ingoing electron @xmath161 apart from the resonant value @xmath361 ( arbitrary @xmath58 and @xmath362 ) reads @xmath363 .
the ingoing wave then almost entirely reflects from the barrier . however , under the condition @xmath364 implying @xmath365 , eqs .
( [ e : eq1 ] ) , ( [ e : eqel ] ) and ( [ e : compleq ] ) result in @xmath366 , @xmath367 for the @xmath368 given by eq .
( [ e : dens ] ) . in case of the exact resonance @xmath369 the electron density reaches a maximum @xmath370 while for the energy deviations considerably exceeding the resonant width @xmath371 the electron density reduces relatively @xmath372 by a factor @xmath373 .
note , that the probability of the resonant capturing is very sensitive to the accuracy of the resonant energy @xmath361 . neglecting the deviation of the potential @xmath4 from the coulomb form @xmath374 ( [ e : pot ] ) at small distances @xmath375 for which @xmath376 @xcite and setting @xmath377 , induce the energy shift @xmath378 .
this shift significantly exceeding the resonant width @xmath379 results in @xmath380 . in conclusion of this paragraph note that the specific problems of the electron scattering on the impurity centres in gnrs having the signs of the resonant and potential scattering require special consideration .
+ + _ estimates of the expected experimental values _
+ + in an effort to render our results close to an experimental setup , we present below the estimates of the expected values for the gnrs of the family corresponding to @xmath381 placed on @xmath382 and sapphire @xmath383 substrates @xcite . since the @xmath324 material is not the best candidate to be described by the theory implying @xmath384 the general equation @xmath385 for the energy @xmath161 has to be taken to calculate the binding energy @xmath386 , width @xmath387 , electric field @xmath342 and other parameters .
being derived from this equation and from eq .
@xmath165 ( [ e : y ] ) the binding energy of the ground rydberg state @xmath9 of the ground size - quantized series @xmath230 for the impurity positioned at the mid - point @xmath388 of the gnr of width @xmath318 reads @xmath389 . for the critical electric field
@xmath342 providing the complete depletion of the @xmath58 impurity level and estimated from the condition @xmath390 we obtain @xmath391 .
the less bound first excited impurity level @xmath392 can be ionized by the electric field @xmath393 .
the above mentioned condition related to the fields @xmath342 is suitable to introduce the parameter of stability @xmath394 of the @xmath303 impurity state associated with the @xmath42 subband relatively to the ionization effect of the electric field @xmath5 @xmath395}{z^{2}+1 } ; \nonumber \end{aligned}\ ] ] where @xmath396 is the root of the equation @xmath397 ( see eq.([e : y ] ) ) . under the condition
@xmath398 the @xmath399 state is practically ionized , while in the opposite case @xmath400 the given state can be treated as relatively stable .
for the possibly employed substrates , namely @xmath324 @xmath325 , sapphire @xmath319 and @xmath401 @xmath402 the corresponding parameters calculated for the impurity positioned at the ribbon mid - point @xmath403 read @xmath404 , respectively . in order to estimate the combined effect of the electric field and fano - ionization on
the impurity states adjacent to the @xmath249 subband we are forced to avoid the ribbon placed on the @xmath324 substrate and address a sapphire substrate .
the point is that the condition of the adiabatic approximation @xmath405 with eq .
( [ e : par ] ) for @xmath57 being written strictly looks like @xmath406 , which , as pointed above , transforms into eq .
( [ e : weakimp ] ) for the low excited @xmath42 subbands . for the chosen @xmath407
the ground subband @xmath230 provides for the @xmath324 substrate @xmath408 , while for the subband @xmath249 this parameter is already @xmath409 that makes the adiabatic approximation for this subband for the @xmath324 substrate to be inappropriate .
it follows from fig .
5 that the resonant fano width consists to a considerable part of the binding energy of the ground state @xmath410 .
the possible reasons for this are first the parameter @xmath353 being close to the threshold of the adiabatic approximation @xmath411 and second the ground state @xmath412 is more sensitive to the fano - coupling @xmath413 than the excited states @xmath414 .
the excited states @xmath415 are expected to be significantly narrower than the ground state @xmath416 . in the presence of relatively weak electric field
@xmath417 the lifetime @xmath418 of the state @xmath419 in the ribbon of width @xmath318 is determined only by the fano width @xmath420 resulting in @xmath421 .
however the lifetime of the first excited state @xmath422 in the same ribbon @xmath423 is of the order of @xmath424 .
thus even in the absence of electric field the resonant fano widths of the impurity states should be taken into account in the study of the electronic and transport processes in gnrs .
recently gong _ et.al . _ @xcite reported that the analogous line - defect - induced fano resonant states in the conduction band of the armchair gnr impede the electron transport in this region . with increasing electric field @xmath5 and decreasing the ribbon width @xmath12
the contribution of the electric field to the complete width becomes more pronounced .
the critical width @xmath356 at which the electric field @xmath425 and the fano coupling contribute equally to the energy width @xmath426 are @xmath427 .
the dependence @xmath428 is valid to a high accuracy .
it should be noted that at the critical ribbon widths @xmath356 the ground impurity state @xmath9 in the ribbon located on the chosen specific substrate @xmath353 seems to be completely ionized . at the same time
the substrates with the greater dielectric constant @xmath29 ( the less @xmath56 ) provide significantly stable impurity states especially those having the quantum numbers @xmath429 . a comparison of our results with those obtained numerically based on density functional theory @xcite and on the tight binding approximation @xcite , @xcite
, @xcite demonstrates that the dirac equation approach employed in this paper quite adequately describes the electronic structure and the impurity and exciton states in the gnrs .
the exciton characteristics can be obtained from the corresponding impurity ones by replacing @xmath430 by @xmath431 and @xmath56 by @xmath432 .
the dependencies @xmath433 of the effective electron mass @xmath434 @xcite , the energy gap @xmath435 ( [ e : ener ] ) and the binding energy @xmath436 ( [ e : elen ] ) , ( [ e : encompl ] ) on the ribbon width @xmath12 are qualitatively in line with those presented in all above mentioned refs .
moreover , the energy gaps @xmath437 reveal a quantitative good agreement .
thus , the energy gaps @xmath438 and @xmath439 calculated from ( [ e : ener ] ) for @xmath318 and @xmath440 , respectively are close to the values @xmath441 @xcite and @xmath442 @xcite presented for the corresponding widths .
a greater discrepancy is found for the masses @xmath443 of the electron in the ribbon of width @xmath444 scaled to the mass of the free electron @xmath445 and @xmath446 @xcite .
though , the dependence @xmath433 of the binding energy on the ribbon width @xmath12 correlates completely with that obtained numerically @xcite , @xcite , @xcite the different environments prevent us from a detailed quantitative comparison .
this is because our data are calculated for the effective dielectric constant @xmath447 ( [ e : coulomb ] ) resulting in @xmath384 , while others for the gnrs or suspended @xmath448 @xcite , @xcite or placed on the @xmath324 substrate , with unspecified dielectric constant @xmath449 ( see eqs .
( 10 ) and ( 11 ) in paper @xcite ) inducing @xmath450 .
we therefore conclude that the presented analytical results well correlate with those obtained by the numerical approaches in the literature . along with the estimates of the expected experimental values
this could be extended to further studies of the wide range of the gnr structures and their applications in nanoelectronics .
in summary , we have developed an analytical approach to the problem of the impurity electron in a narrow armchair gnr exposed to the external electric field directed parallel to the graphene axis .
the effect of the strong confinement is taken to be much greater than the influence of the impurity coulomb electric field , which in turn considerably exceeds the external field . in the approximation of the isolated size - quantized subbands
we have calculated the complex energy levels of the quasi - discrete impurity rydberg states and phases of the wave functions of the continuous spectrum .
the complex energies determine the binding energies and widths of the quasi - stationary states , while the phases ( and s - matrix ) allow to study the various scattering problems .
the explicit form of the obtained results makes it possible to trace the dependence of the listed above values on all the parameters of the structure , namely , on the ribbon width , position of the impurity centre , and the electric field .
in particular it was found that the gnr is the structure in which the mechanism of the dimensional ionization occurs : the impurity centre can be ionized by increasing the ribbon width . in the approximation of the ground and
first excited size - quantized subbands the complete widths of the first excited rydberg series caused by the combined effect of the electric fields and the fano resonant inter - subband coupling have been calculated .
estimates of the expected experimental values for realistic gngs show that there are two aspects of the effect of the electric field .
weak field provides the resonant capturing of the electrons by the impurity centres for a significantly long lifetime , and remain quasi - discrete impurity states available for the experimental in particular optical study . relatively strong field releases the bound electrons to activate the transport properties of the gnrs .
the authors are grateful to d.b .
turchinovich for useful discussions and a. zampetaki for significant technical assistance . | we present an analytical investigation of the quasi - coulomb impurity states in a narrow gapped armchair graphene nanoribbon ( gnr ) in the presence of a uniform external electric field directed parallel to the ribbon axis .
the effect of the ribbon confinement is taken to be much greater than that of the impurity electric field , which in turn considerably exceeds the external electric field . under these conditions
we employ the adiabatic approximation assuming that the motion parallel ( `` slow '' ) and perpendicular ( `` fast '' ) to the ribbon axis are separated adiabatically . in the approximation of the isolated size - quantized subbands induced by the `` fast '' motion
the complex energies of the impurity electron are calculated in explicit form .
the real and imaginary parts of these energies determine the binding energy and width of the quasi - discrete state , respectively .
the energy width increases with increasing the electric field and ribbon width .
the latter forms the background of the mechanism of dimensional ionization .
the s - matrix - the basic tool of study of the transport problems can be trivially derived from the phases of the wave functions of the continuous spectrum presented in explicit form . in the double - subband approximation
we calculate the complete widths of the impurity states caused by the combined effect of the electric field and the fano resonant coupling between the impurity states of the discrete and continuous spectra associated with the ground and first excited size - quantized subbands .
our analytical results are shown to be in agreement with those obtained by other theoretical approaches .
estimates of the expected experimental values for the typically employed gnrs show that for weak electric field the impurity quasi - discrete states remain sufficiently stable to be observed in corresponding experiment , while relatively strong field unlock the captured electrons to further restore their contribution to the transport . |
it is well known that the spiral arms of galaxies are density waves .
this means that the spiral arms are not always composed of the same matter but they only represent the maxima of the density along every circle around the center .
the stars passing through the spiral arms stay longer close to them , thus producing an increase of the local density .
the linear theory of spiral density waves assumes that the potential v , the density @xmath0 ( or the surface density @xmath1 ) and the distribution function _ f _ ( phase space density ) have small deviations from their axisymmetric values ( @xmath2,@xmath3,@xmath4 ) .
this linear theory was initiated by lindblad @xmath5 but it was developed in its modern form by lin and shu @xmath6 , kalnajs @xmath7 , lynden - bell and kalnajs @xmath8 toomre @xmath9,and others .
however the deviations near the main resonances of a galaxy ( inner and outer lindblad resonances and corotation ) are large , thus near these resonances a nonlinear theory is necessary ( contopoulos @xmath10 ) . in both the linear and the nonlinear theory ,
whenever the perturbation is relatively small , chaos is unimportant . although some chaotic orbits appear near all the unstable periodic orbits , their proportion and their effects are small .
this is the case of most normal spirals , where the density perturbations are of order @xmath11 of the axisymmetric background . on the other hand in barred galaxies
the perturbations are large , of the order of @xmath12 . in such cases
the chaotic orbits play an important role in the dynamics of the galaxies .
the main chaotic effects in a galaxy appear near corotation .
it is well known now that chaos is generated by the overlapping of resonances ( contopoulos @xmath13 , rosenbluth et al .
@xmath13 , chirikov @xmath14 ) . in the region near corotation there are many resonances between the angular velocities of the stars @xmath15 ( in the frame rotating with the angular velocity of the pattern @xmath16 ) and the epicyclic frequence @xmath17 .
the ratio of these frequencies is @xmath18 these resonances are congested in a relatively small interval of jacobi constants @xmath19 around the jacobi constant @xmath20 at corotation
. some important resonances in an extended region around corotation are @xmath21 ( inside corotation ) , @xmath22 ( at corotation ) and @xmath23,@xmath24 ( outside corotation ) .
chaos produced by the interactions of resonances , extends around the envelope of the bar and along the spiral arms beyond the end of the bar .
the first example of a chaotic orbit that fills an envelope of the bar and the inner parts of the spiral structure was provided by kaufmann and contopoulos ( 1996 ) ( fig.1 ) .
however the most surprising result came from n - body simulations , which indicated that most orbits along the spiral arms beyond the ends of the bar are chaotic ( voglis et al .
@xmath25 ) . thus started a systematic study of the chaotic spiral arms outside corotation in strong barred galaxies ( voglis and stavropoulos 2005 , voglis et al .
@xmath26 , romero - gomez et al .
@xmath27,@xmath28 , tsoutsis et al .
@xmath29 , athanassoula et al.@xmath30 , * 2010 * , harsoula et al.@xmath31 , @xmath32 ) . in the present paper
we present the most recent results of this study .
although the spiral arms beyond the ends of the bars are composed of chaotic orbits , nevertheless these spiral arms are density waves , i.e. the density maxima are populated by different stars at every time .
the density of the stars is maximum in areas where their velocities are minimum .
this happens mainly near the apocentres and the pericentres of the orbits . and
these apocentres and pericentres appear close to the asymptotic manifolds of the main unstable periodic orbits around corotation .
the most important families of unstable periodic orbits in the corotation region are the families @xmath33,@xmath34 around the unstable lagrangian points @xmath35 and @xmath36 . in a simple model of a barred galaxy the hamiltonian near corotation @xmath37 ( contopoulos @xmath38 )
is given by the relation @xmath39 where @xmath40 is the jacobi constant , @xmath41 the epicyclic action , @xmath42 is the azimuthal action , ( where @xmath43 , @xmath44 are the angular momenta of the star and of corotation ) @xmath45 is the azimuth of the star and @xmath46 , @xmath47 , @xmath48 , @xmath49 , @xmath50 , @xmath51 are constants .
the action @xmath41 is also constant , because the conjugate angle * @xmath52 * , does not appear in the hamiltonian .
the orbits @xmath33 , @xmath34 are represented by the unstable equilibrium points of the system , namely when @xmath53 therefore @xmath54 and @xmath55 the orbits close to the lagrangian points @xmath35 , @xmath36 have @xmath56 and @xmath57 , ( while the orbits close to the lagrangian points @xmath58,@xmath59 have @xmath60 , and @xmath61 ) , in the lowest approximation the orbits @xmath33 , @xmath34 start with @xmath56 , or @xmath62 and @xmath63 hence @xmath64 therefore these orbits exist whenever @xmath65 . after finding @xmath41 we can find @xmath66 from eq.(5 ) .
we have ( contopoulos @xmath67 ) @xmath68 where @xmath69 is the angular velocity , and @xmath17 is the epicyclic frequency .
the subscript @xmath70 denotes the values at corotation and the accents mean derivatives with respect to the radius r. thus eq.(5 ) can be written @xmath71 we consider now a simple model of the form @xmath72 where @xmath73 is the potential and @xmath74 is the force as a function of @xmath75 .
then we have @xmath76 and @xmath77 thus eq.(9 ) gives @xmath78 and if @xmath79 we find that @xmath66 is negative . in particular in a keplerian model @xmath80 and the relation ( 13 ) becomes @xmath81 therefore @xmath82 .
the initial point of the orbit @xmath33 ( with @xmath83 and increasing @xmath84 ) is @xmath85 inside the corotation distance with @xmath86 the orbit @xmath33 is described counterclockwise ( while the galaxy is rotating clockwise . )
thus the orbit @xmath33 intersects the @xmath83 axis once more with @xmath87 and decreasing @xmath84 ( fig .
@xmath34 on the configuration plane . ]
the asymptotic manifolds from the orbits @xmath33 , @xmath34 on the configuration plane are shown in fig.2 .
there are two unstable manifolds from @xmath33 namely @xmath88 along a trailing spiral and @xmath89 along the leading edge of the bar , and two stable manifolds , @xmath90 ( along a leading direction ) and s ( along the trailing edge of the bar ) .
these manifolds contain successive apocentres ( @xmath91=0 ) ) of the asymptotic orbits i.e. orbits starting on this manifold .
similar manifolds emanate from @xmath85 , @xmath92 , corresponding to the pericentres of the asymptotic orbits . the orbits starting close to @xmath33 but not exactly on the asymptotic manifold approach this point along the stable directions @xmath93 and @xmath90 and then deviate along the unstable directions @xmath88 and @xmath89 .
after a longer time these orbits form trailing spiral arms and the envelope of the bar .
the forms of the asymptotic curves emanating from @xmath33 and @xmath34 are shown in fig .
we see that the curve u from @xmath33 reaches the neighborhood of @xmath34 , making oscillations around the asymptotic curve @xmath94 of larger and larger amplitude as it approaches @xmath34 , coming closer and closer to the asymptotic curves @xmath95 and @xmath96 from @xmath34 , which are symmetric to @xmath88 and @xmath89 with respect to the center of the galaxy .
thus the matter that starts close to @xmath33 and moves along the asymptotic curve @xmath88 , after reaching the neighborhood of @xmath34 moves very close to the asymptotic curves @xmath95 and @xmath96 and approaches again the neighborhood of @xmath33 . in a similar way matter moves along the asymptotic curves @xmath89 , @xmath95 and @xmath96 .
thus we have circulation of the material along the spiral arms that lasts for a substantial fraction of the hubble time , until the spiral arms fade away . and @xmath34 .
] the individual orbits along the asymptotic curve @xmath88 or close to it , are of the form of fig.4 .
the orbits start by making some rotations close to the periodic orbit @xmath33 and they deviate away from @xmath33 , reaching the neighborhood of @xmath34
. then they proceed either close to the spiral @xmath95 ( fig.4a ) , or close to the envelope of the bar @xmath96 ( fig .
the unstable asymptotic curves of the various unstable periodic orbits for the same jacobi constant can not cross themselves or each other .
thus they are obliged to follow nearly parallel paths .
the main unstable orbits inside corotation are the families @xmath97 , @xmath98 and @xmath99 ( inner lindblad ) , while close to corotation the most important families are @xmath100 and @xmath34 whose asymptotic manifolds are shown in fig
on the other hand , all the manifolds of the unstable periodic orbits corresponding to the same jacobi constant , are approximately parallel and contribute to the formation of spiral arms outside corotation ( fig .
this is the phenomenon of coalescence that was described by tsoutsis et al.(2008 ) .
( around the lagrangian point @xmath35 ) approach the lagrangian point @xmath36 and then deviate ( a ) along @xmath95 , or ( b ) along @xmath96 . ] and @xmath34 in the configuration space ( b ) the `` coalescence '' of the invariant asymptotic manifolds from the @xmath33 , @xmath34 , @xmath101 , @xmath23 and @xmath102 families . ,
title="fig:",width=234 ] and @xmath34 in the configuration space ( b ) the `` coalescence '' of the invariant asymptotic manifolds from the @xmath33 , @xmath34 , @xmath101 , @xmath23 and @xmath102 families .
, title="fig:",width=207 ] if we consider also similar figures for other values of the jacobi constant ( in which case overlapping of various curves is permitted ) we see the formation of thick spiral arms .
in fact the observed spiral arms in n - body simulations of barred galaxies are very close to the general form of the spiral arms produced by the superposition of the various asymptotic curves .
however orbits close to the particular unstable periodic orbits support particular parts of the spiral arms .
thus it is of interest to study the orbits close to every resonance .
the planar orbits in a time independent rotating model have a fixed jacobi constant ( that we call `` energy in the rotating frame '' ) @xmath103 where u is the velocity in the rotating frame , @xmath104 is the potential in polar coordinates , @xmath43 is the angular momentum and @xmath16 is the angular velocity of the system .
if @xmath19 is larger that the energy @xmath105 at the lagrangian points @xmath106 , the orbits inside and outside corotation can communicate . if , however , @xmath107 the orbits inside corotation can not get outside and those outside corotation can not get inside .
we consider a particular n - body system that was studied by voglis and stavropoulos ( 2005 ) and voglis et al .
( 2006a ) simulating a barred spiral galaxy and we separate the orbits in particular intervals @xmath108 .
we consider first the n - body orbits that have energies in the interval @xmath109 . in this energy level
there are almost no regular orbits at all ( see fig . 3 of harsoula et al .
2011 ) . for these energies the orbits can move both inside and outside corotation .
we integrate the orbits with initial conditions outside corotation for @xmath110 ( half - mass crossing times ) and find the distribution of their @xmath111 values ( fig .
the time interval @xmath112 is about one third of the hubble time . during that time most resonant
orbits are concentrated near the resonances @xmath23 ( outer lindblad ) , @xmath24 and @xmath113 .
only a few orbits have moved inside corotation @xmath114 . however after about one and a half hubble time ( fig .
6b ) the above resonances have fewer stars .
some stars have escaped , but a substantial proportion of stars have reached the inner resonances @xmath98 and @xmath99 ( inner lindblad ) .
these stars are trapped near these resonances for very long times before escaping again outwards . on the other hand stars starting inside corotation
remain close to the resonances @xmath98 and @xmath99 for more than @xmath115 hubble times before escaping outside the galaxy ( without being trapped by the outer resonances ) .
it must be pointed out here that the 2-body relaxation time for galaxies is of the order of @xmath116 hubble times , a time exceedingly longer that the diffusion time of the chaotic orbits in our n - body model .
then we consider the stars outside corotation for an energy interval @xmath117 , where again only chaotic orbits exist outside corotation . in this case
the orbits with initial conditions outside corotation can not enter inside corotation . in figs .
7a , b the distribution of the @xmath111-values of these stars is given , for the time intervals @xmath118 ( fig.@xmath119 ) and @xmath120 ( fig.@xmath121 ) . in this case
there are no @xmath122 orbits , but we notice some important resonances like @xmath23 ( outer lindblad ) , @xmath123 and @xmath124 .
after one and a half hubble times ( fig.@xmath121 ) the number of resonant stars has decreased and many stars have escaped beyond the ends of the galaxy . nevertheless there is still an appreciable number of stars close to these resonances .
this is due to the stickiness phenomenon that lasts for very long times , before the stars escape from the galaxy . comparing figs .
6 and 7 we notice that in the case of chaotic orbits that are restricted in the area outside corotation , stickiness to resonances lasts for longer times in smaller values of jacobi constants than in greater ones .
thus in general , the diffusion of chaotic orbits supporting the outer parts of the spiral arms outwards is more slow than the diffusion of orbits supporting more inner parts of the spiral structure .
an example is given in fig .
8 where the percentage of orbits starting close to the @xmath125 unstable periodic orbits that stay located inside a radius @xmath126 ( which in fact confines the bound part of the galaxy , @xmath127 being the half mass radius of the system ) is plotted as a function of time in @xmath128 , for four different energy levels ( jacobi constants ) , namely for @xmath129 ( black solid curve ) , @xmath130 ( black dotted curve ) , @xmath131 ( gray curve ) and @xmath132 ( black dashed curve ) .
after 1 hubble time ( @xmath133 ) , @xmath134 of the orbits with @xmath129 is still located inside @xmath135 , while @xmath136 of the orbits with @xmath130 , @xmath137 with @xmath131 and only @xmath138 of the orbits with @xmath132 is still located inside @xmath126 .
the functions @xmath139n versus @xmath140 for the various energy levels are approximated by exponentials of the form @xmath141 where @xmath142 and @xmath143 take the values given in table 1 .
+ .the values of @xmath142 and @xmath143 in eq .
( 17 ) [ cols="^,^,^",options="header " , ] the values of @xmath142 are close to @xmath144 , while @xmath143 can be given by the approximate formula @xmath145 with @xmath146 x @xmath147 and @xmath148 x @xmath149 . therefore the diffusion is much faster for larger jacobi constants . on the other hand for smaller jacobi constants
the diffusion is slower , and most of the stars remain close to the outer parts of the spiral arms for more than a hubble time . in the case of chaotic orbits with initial conditions inside corotation or close to it , the diffusion happens quickly for an initial time interval corresponding to @xmath150 1/3 of the hubble time ( during which the spiral structure survives ) , while later on it is very slow ( see fig .
24 of harsoula et al . 2011 ) .
we therefore conclude that the outer parts of the spiral structure of our n - body model survive for longer times than the inner parts of the spiral structure .
, where the areas inside and outside corotation can communicate ( b ) @xmath151 , where the areas inside and outside corotation can not communicate ( c ) @xmath152 where again the areas inside and outside corotation can not communicate and ( d ) the isodensities of particles belonging to all previous three levels .
the spiral structure is apparent here . ] in previous papers ( harsoula and kalapotharakos 2009 , harsoula et al . 2011 ) we have emphasized the role of stickiness of chaotic orbits along the unstable asymptotic manifolds of the unstable periodic orbits , in supporting the structure of the spiral arms . in what follows , we investigate the role of the 2-d asymptotic orbits , i.e. orbits having initial conditions on the unstable manifolds of the various unstable periodic orbits , in supporting the spiral structure of the model .
( or @xmath153 ) family having initial conditions inside corotation and jacobi constant @xmath154 .
superimposed is the orbit @xmath153 plotted in black .
( b ) the density distribution of 2-d orbits having initial conditions on a grid close and around the unstable periodic orbit @xmath99 ( in color ) .
superimposed , in black , are the isodensities of the real n - body particles belonging to the same energy level .
the corresponding time of integration of the orbits is @xmath150 one and a half hubble time . ]
orbits in different energy levels support different parts of the spiral structure .
this is obvious in fig.9 where the isodensities of the n - body particles are plotted on the configuration plane of rotation , belonging to three different energy levels .
more precisely , particles having values of jacobi constant @xmath155 correspond to the envelope of the bar and the innermost parts of the spiral arms .
particles with values of jacobi constant @xmath156 correspond to parts of the spiral arms that extend further beyond and finally particles with values of jacobi constant @xmath157 correspond to the outermost part of the spiral arms . in fig .
9d we plot the isodensities of particles belonging to all previous energy levels of figs .
9a , b , c .
the spiral structure of the galaxy is apparent . in harsoula
2011 we studied the density distribution of @xmath158 orbits starting close to the various resonances .
however we find similar results if we consider the projections of the various orbits on the plane of symmetry ( @xmath159 ) of the galaxy ( having the bar along the z - axis ) . in fig .
10 an example of the density distribution of 2-d asymptotic orbits of an unstable periodic orbit is plotted ( in color ) for a jacobi constant @xmath154 , where the areas inside and outside corotation can communicate .
more precisely in fig.10a we plot ( in color ) 10000 asymptotic orbits of the @xmath99 ( or @xmath153 ) family , having initial condition inside corotation , together with the unstable periodic orbit ( in black ) .
the corresponding time of integration of the orbits is @xmath150 one and a half hubble time . for an initial time interval equal to @xmath150 1/5 of the hubble time ,
the chaotic orbits stay close to the periodic orbit , following its shape , but later on they are diffused outwards modulating the inner parts of the spiral structure . if we take the same number of orbits , with initial conditions not on the unstable manifold , but on a grid close and around the unstable periodic orbit having small deviations from it , on the ( @xmath160 ) surface of section(fig .
10b ) , we find that the distribution of the orbits ( in color ) follows the inner parts of the spiral pattern , derived from the isodensities of the real n - body particles of the corresponding energy level ( in black ) .
this is an example of stickiness of chaotic orbits along the unstable asymptotic manifolds of the unstable periodic orbits .
in fact the stickiness of chaotic orbits delays their diffusion outwards and is responsible for the survival of the spiral structure of the galaxy for more that 10 rotations of the bar ( see harsoula et al .
2011 ) .
similar results are found for the @xmath161 orbits near @xmath106 at the end of the bar , for the @xmath24 orbits outside corotation and for the @xmath162 orbits around @xmath58 and @xmath59 . and
@xmath23 for various jacobi constants , superimposed with the isodensities of the real n - body particles belonging to the corresponding energy levels ( black curves ) . ] finally , in fig .
11 we present the density distribution of the sticky chaotic orbits near the resonances @xmath163 superimposed with the isodensities of the real n - body particles belonging to the corresponding energy levels .
we conclude that by using a sample of sticky chaotic orbits around a number of unstable periodic orbits inside and outside corotation in different energy levels , we are able to reproduce quite well the outer envelope of the bar and the spiral structure of the galaxy .
the main conclusions of our paper are the following : \1 ) stickiness of chaotic orbits close to the unstable asymptotic manifolds of various periodic orbits delays the diffusion of these orbits outwards and therefore modulates the shape of the spiral structure of the galaxy for more than 10 rotations of the bar , corresponding to 1/3 of the hubble time .
\2 ) chaotic orbits that are limited outside corotation modulate the outer parts of the spiral structure for smaller values of jacobi constant while orbits with greater values of jacobi constant modulate the inner parts of the spiral structure .
moreover , in our n - body model , stickiness to resonances for smaller values of jacobi constants lasts for longer times than stickiness for greater values of jacobi constants .
\3 ) asymptotic orbits ( having initial conditions on the unstable asymptotic curve of an unstable periodic orbit ) stay located close to the periodic orbit for an initial interval of time , following the shape of this specific orbit , before diffusing from it and supporting the spiral structure .
chaotic orbits having initial conditions inside corotation modulate the envelope of the bar and the innermost spiral structure during a time interval of fast diffusion ( @xmath164 of the hubble time ) and then they are diffused outwards with much slower rates .
\4 ) using a sample of sticky chaotic orbits close to a number of unstable periodic orbits inside and outside corotation , in different energy levels , we are able to reproduce quite well the outer envelope of the bar and the spiral structure of the galaxy . 14 athanassoula , e. , romero - gomez , m. and masdemont , j.j . :
rings and spirals in barred galaxies - i. building blocks , mon . not .
astron . soc .
, * 394 * , 67 - 81 ( 2009a ) athanassoula , e. , romero - gomez , m. , bosma , a. and masdemont , j.j . : rings and spirals in barred galaxies - ii . ring and spiral morphology .
mon . not .
, * 400 * , 1706 - 1720 , ( 2009b ) athanassoula , e. , romero - gomez , m. , bosma , a. and masdemont , j.j . : rings and spirals in barred galaxies - iii .
further comparisons and links to observations . mon . not .
soc . , * 407 * , 1433 - 1448 , ( 2010 ) contopoulos , g. : resonance effects in spiral galaxies .
astrophys.j . , * 160 * , 113 - 133 , ( 1970 ) contopoulos , g. : the particle resonance in spiral galaxies .
nonlinear effects .
astrophys.j . ,
* 181 * , 657 - 684 ( 1973 ) contopoulos , g. : inner lindblad resonance in galaxies - nonlinear theory .
i. astrophys.j .
, * 201 * , 566 - 584 ( 1975 ) contopoulos , g . : periodic orbits near the particle resonance in galaxies .
astrophys . ,
* 64 * , 323 - 332 ( 1978 ) harsoula , m. and kalapotharakos , c. : orbital structure in n - body models of barred - spiral galaxies .
mon . not .
, * 394 * , 1605 - 1619 , ( 2009 ) harsoula , m. , kalapotharakos , c. and contopoulos , g. : asymptotic orbits in barred spiral galaxies . mon . not .
, * 411 * , 1111 - 1126 ( 2011 ) kaufmann d.e . , and
contopoulos g. : self - consistent models of barred spiral galaxies .
astron.astrophys . ,
309 , 381 ( 1996 ) .
romero - gomez , m. , athanassoula , e.,masdemont , j.j . and
garcia - gomez , c. : the origin of rr1 ring structures in barred galaxies .
astron.astrophys . , * 453 * , 39 - 45 ( 2006 ) romero - gomez , m. , athanassoula , e. , masdemont , j.j . and garcia - gomez , c. : the formation of spiral arms and rings in barred galaxies .
astron.astrophys . ,
* 472 * , 63 - 75 ( 2007 ) tsoutsis , p. , efthymiopoulos , ch . and voglis , n. : the coalescence of invariant manifolds and the spiral structure of barred galaxies .
mon . not .
roy.astron.soc . , * 387 * , 1264 - 1280 ( 2008 ) voglis , n . and
stavropoulos , i .
: can chaotic motion be responsible for the formation of spiral arms ? in solomos ,
n.(ed ) `` recent advances in astronomy and astrophysics '' aip conf .
proceedings * 848 * , 647 - 659 ( 2006 ) voglis , n. , stavropoulos , i. and kalapotharakos , c. : chaotic motion and spiral structure in self - consistent models of rotating galaxies .
mon . not .
soc . , * 372 * , 901 - 922 ( 2006b ) voglis , n. , tsoutsis , p. and efthymiopoulos , ch .
: invariant manifolds , phase correlations of chaotic orbits and the spiral structure of galaxies . mon . not .
soc . , * 373 * , 280 - 294 ( 2006c ) | we study the role of asymptotic curves in supporting the spiral structure of a n - body model simulating a barred spiral galaxy .
chaotic orbits with initial conditions on the unstable asymptotic curves of the main unstable periodic orbits follow the shape of the periodic orbits for an initial interval of time and then they are diffused outwards supporting the spiral structure of the galaxy .
chaotic orbits having small deviations from the unstable periodic orbits , stay close and along the corresponding unstable asymptotic manifolds , supporting the spiral structure for more than 10 rotations of the bar .
chaotic orbits of different jacobi constants support different parts of the spiral structure .
we also study the diffusion rate of chaotic orbits outwards and find that chaotic orbits that support the outer parts of the galaxy are diffused outwards more slowly than the orbits supporting the inner parts of the spiral structure . |
organic superconductors of the bedt family exhibit fascinating phenomena due to the interplay between strong electronic correlations and large magnetic frustration.@xcite for instance , their rich phase diagram contains @xmath9-wave superconducting , antiferromagnetic@xcite and possibly quantum spin liquid states at absolute zero.@xcite at finite temperature , a mott metal - insulator transition has clearly been identified.@xcite recently , a moderately one - dimensional organic compound has been synthesized.@xcite this 2d organic mott insulator , @xmath0-h@xmath1(cat - edt - ttf)@xmath2 , does not exhibit magnetic order at very low temperature ( @xmath10 mk ) , which makes it a serious candidate for a quantum spin liquid .
microscopically , the simplest model describing this organic compound is the two - dimensional single - band hubbard model on an anisotropic triangular lattice.@xcite this model is the same as that often used for bedt organic compounds.@xcite in order to focus our attention on @xmath0-h@xmath1(cat - edt - ttf)@xmath2 , we study the phase diagram of this model in the moderately one - dimensional ( m1d ) regime characterized by a ratio between the hopping terms @xmath4 and @xmath11 , defined in fig.[fig.sym_a ] , larger than unity .
since we are interested in the mott transition and the possibility of a spin liquid , we use cellular dynamical mean - field theory ( cdmft),@xcite a cluster extension of dynamical mean - field theory ( dmft ) that can treat both the metallic and the insulating phases , the mott transition between them as well as magnetic phases.@xcite before we proceed , we briefly recall related studies , setting aside superconductivity that we do not consider here .
previous work focused mostly on the frustrated regime @xmath12 ( square lattice to triangular lattice ) since these anisotropy values , usually obtained from ab initio calculations,@xcite corresponded to all known bedt organic compounds .
theoretical investigations that were concerned with the mott transition in the interaction - frustration ( @xmath13 ) phase diagram used methods that included path - integral renormalization group @xcite , variational methods , @xcite exact diagonalization,@xcite variational cluster approximation , @xcite cdmft @xcite and dual - fermion @xcite approaches .
although there are quantitative discrepancies between the different methods , metallic , insulating , non - magnetic , and antiferromagnetic phases were found , generally in good agreement with experiments .
@xcite a more detailed comparison between experiment and some of the above theoretical calculations appears in ref .. cdmft was one of the most successful approaches.@xcite more recently , the m1d regime ( @xmath14 ) has been investigated:@xcite for @xmath15 , the hubbard model exhibits a spiral order and possibly a spin liquid phase .
different theoretical lattice approaches seem to agree with the presence of a spin liquid and a collinear magnetic state with an associated ordering vector @xmath16 for @xmath17,@xcite even though they do not completely agree on the precise form of the phase diagram , unlike for the case @xmath18 . in the strong - interaction limit , _
i.e _ for @xmath19 , where the hubbard model in the insulating phase reduces to the heisenberg model with exchange interactions @xmath20 and @xmath21 to second order in perturbation theory@xcite and , at higher order , to models with more complicated spin interactions , such as the two distinct ring exchange couplings @xmath22 and @xmath23.@xcite the heisenberg model corresponding to the m1d regime , _ i.e
_ @xmath24 has been extensively studied using different methods such as linear spin - wave,@xcite coupled cluster method,@xcite variational monte - carlo@xcite or density matrix renormalization group.@xcite these methods show that a spiral state is present for @xmath25 but they give different magnetic phases in the m1d regime , _
e.g _ a spiral phase , a collinear magnetic phase or a spin liquid state .
more sophisticated hamiltonians including the ring exchange coupling @xmath26 in the anisotropic triangular lattice give a rich phase diagram where the presence of a nel state , a spin liquid state or a spiral phase depends on the relative strength between @xmath27 and @xmath28.@xcite this paper is organized as follows : the hubbard model and the cellular dynamical mean - field theory ( cdmft ) on a plaquette with an exact diagonalization ( ed ) impurity solver are detailed in sec.[sec.model&method ] . in sec.[sec.results_normal ] , results for the normal state , showing a first order mott metal - insulator transition are presented .
magnetic states are explored in sec.[sec.results_magnet ] .
our results are summarized in the phase diagram of fig.[fig.entire_phase_diagram ] .
finally , the choice of cluster is motivated in sec.[sec_cluster ] , where we show , by comparing results for the magnetic phases with other methods , that a cluster sharing symmetries with the lattice is essential for a reliable cdmft calculation in this regime .
we conclude in sec.[sec.conclusion ] .
we focus on the physics embodied in the hubbard model on the anisotropic triangular lattice @xmath29 all quantities are expressed in natural units ( @xmath30 and @xmath31 ) . here , @xmath32 are the hopping amplitudes between sites @xmath33 and @xmath34 and can take two different values , @xmath11 and @xmath4 , illustrated in fig.[fig.sym ] .
the fourier transform of the hopping amplitudes @xmath32 determines the anisotropic bare dispersion @xmath35 -2t'\cos(k_x + k_y)$ ]
. the on - site coulomb repulsion is @xmath36 and @xmath37 is the chemical potential set so that the system is half - filled . for that filling , the signs of @xmath11 and @xmath4 do not modify the phase diagram .
layered organic compounds are usually half - filled , but doped compounds @xcite have been investigated experimentally @xcite and theoretically.@xcite we focus on the m1d regime @xmath6 using cdmft,@xcite a cluster extension of dynamical mean - field theory ( dmft).@xcite cdmft approximates the infinite lattice as a finite size cluster self - consistently coupled to a bath of non - interacting electrons , thus taking into account dynamical correlations as well as spatial correlations up to the size of the cluster.@xcite cdmft , like dmft , maps the system into an anderson impurity problem , which is then solved self - consistently . in this paper , the quantum impurity problem
is solved using the exact diagonalization ( ed ) method @xcite at zero temperature .
this method is restricted to a small number @xmath38 of bath sites .
while the hamiltonian of the quantum impurity problem is coded exactly , the ground state and the green functions of interest are found in a quasi - exact way with the lanczos algorithm.@xcite exact diagonalization is robust in the presence of frustration , unlike quantum monte - carlo methods which suffer from the infamous sign problem.@xcite moreover , it directly computes dynamical quantities as a function of real frequencies .
to summarize , the assumption inherent to cluster approaches is that the essential physics of the system originates from short - ranged correlations , which should be the case in strongly - correlated magnetically - frustrated organic compounds .
we solve the following cluster - bath hamiltonian : @xmath39 where the indices @xmath40 label the sites within the cluster whereas @xmath41 label the bath sites .
the second quantized operators @xmath42 and @xmath43 annihilate electrons on the cluster and in the bath , respectively .
@xmath32 are the hopping matrix elements within the cluster , @xmath44 are the bath energies , and @xmath45 are the bath - cluster hybridization matrix elements . besides , in order to allow antiferromagnetism to appear , @xmath44 and @xmath45 explicitly carry a spin variable @xmath46 .
a complication of the ed method is that the cdmft self - consistency condition can not be satisfied exactly because of the finite number of bath sites .
this condition is rather satisfied approximately by minimizing a distance function . for further information on the matter ,
see refs . , and .
we use the same distance function parameters as in the last two references , namely a frequency cut - off at @xmath47 and a `` fictitious '' inverse temperature @xmath48 . to check that there are no artifacts associated with the finite bath
, we checked our results for the mott transition using cdmft with a continuous - time quantum monte - carlo ( ctqmc ) impurity solver @xcite at finite but low temperature ( @xmath49 ) .
for both clusters that we describe below , we found agreement with our cdmft plus ed solver for the values of @xmath7 that we tested ( 0.4 and 1.5 ) . as illustrated in fig.[cluster_cartoon ] , we use clusters of @xmath50 sites coupled to @xmath51 bath sites .
although the calculation is for 2@xmath522 clusters , we expect to capture the main physics of the lattice since studies using cdmft have confirmed that results on a 2@xmath522 cluster are quantitatively similar to those on larger clusters , at least at high temperature.@xcite all physical results presented in the next sections are extracted from the symmetry - preserving ( sp ) cluster of fig.[fig.sym_b ] , whose parametrization within the model eq . is detailed in fig.[fig_sp ] . for large values of @xmath7 , one might argue that the quasi one - dimensional character of the lattice must be present in the cluster . in order to shed light on this question , the quasi - one dimensional ( q1d ) cluster of fig.[fig.sym_b ] ,
whose parametrization within the model eq . is detailed in fig.[fig_q1d ] , has also been investigated .
our results , presented in the following sections , will show that the physics extracted from this second cluster geometry does not compare well with other methods , leading us to conclude that the sp cluster is a better representation of the infinite lattice in the m1d regime .
first , let us focus on the mott transition in the normal state .
magnetic states are forbidden if one suppresses the spin dependence of the bath parameters .
the phase diagram was obtained by changing @xmath53 in steps of @xmath54 and varying @xmath55 in much smaller steps .
therefore , the values of @xmath53 quoted for phase boundaries in the following have an uncertainty of order @xmath56 .
we find that as long as @xmath57 , the double occupancy displays hysteresis bounded by jumps at @xmath58 and @xmath59 as @xmath36 decreases or increases , respectively ( @xmath60 ) .
these jumps are the signature of the usual mott transition . for the sp cluster at @xmath61 ,
the hysteresis region is still present but is bounded by a jump only when @xmath36 decreases .
as @xmath36 increases , a mere change of slope occurs , as shown in fig.[d ] . even without a jump in the double - occupancy ,
the mott transition can be observed by studying the low - frequency behavior of the local density of states @xmath62 , as presented in fig.[fig.dos ] for @xmath63 . for @xmath64 ,
slightly smaller than the upper critical ratio @xmath65 at @xmath63 , the local density of states exhibits a low - frequency metallic behavior .
when @xmath55 is increased only by a tiny fraction to @xmath66 , the opening of an insulating gap in the local density of states signals the mott transition .
the mott transition of the q1d cluster , presented in fig.[d_ng ] , does not even feature a jump or some hysteresis in the double occupancy .
the local density of states , however , indicates a mott transition ( not shown ) .
+ the critical ratios @xmath65 presented in fig.[fig.comparaison ] as a function of @xmath7 illustrate one of the main differences between the two cluster geometries considered in this paper .
indeed , for @xmath67 , @xmath65 first decreases before increasing for the q1d cluster whereas it only increases monotonically for the sp cluster .
surprisingly , this discrepancy between the two cluster geometries does not hold for @xmath68 since both geometries yield the same value of @xmath65 even if one could assume that the q1d cluster should be more appropriate only in the m1d regime .
an acceptable explanation for this phenomenon has not been found yet , but the following section will give arguments that lead us to believe that the results for the sp cluster capture the correct physics .
within cdmft , one can only look for commensurate magnetic orders on the cluster .
therefore , this restriction does not allow us to explore all possible magnetic phases nor to distinguish between a spin liquid and an incommensurate magnetic order ( in the sense of a magnetic order whose unit cell does not perfectly fit or repeat within the cluster ) .
hence , we can only rule out a spin liquid by demonstrating that a magnetic phase exists , but we can not prove that a spin liquid state will occur since we can not explore all possible magnetic states .
in other words , not finding one of the allowed magnetic states of our cluster in a mott insulating phase is a necessary , but not sufficient , condition for a spin liquid .
a spin - liquid state could occur only in a non - magnetic insulating state ( nmi state ) . for reasons that will become clear below
, we present our final phase diagram , including magnetic order , only for the sp cluster .
one can check from fig.[fig.phase_diagram ] that for @xmath57 , we find the same results as in ref . , namely a transition from a metal to a nel state for @xmath69 , followed by a nmi state that starts right above the mott transition for @xmath70 and then undergoes a nel transition at larger @xmath55 if @xmath71 .
we did not investigate @xmath72 .
previous studies indicate that a spiral order or a spin liquid could be present in this area of the phase diagram,@xcite corresponding to the nmi state of fig.[fig.phase_diagram ] . for @xmath73 , a first - order transition between a metal and a collinear magnetic insulating state ( cms ) takes place for @xmath74 or @xmath75 upon increasing @xmath55 , before the mott transition , as shown in fig.[fig.phase_diagram ] , and survives at larger values of @xmath55 .
the presence of this phase is not surprising since different studies predict the appearance of this magnetic phase in the m1d regime using lattice models @xcite or spin models.@xcite at this magnetic transition , we observe a jump in the double - occupancy and a gap opening in the spectral function .
some hysteresis can be seen in the double - occupancy , but while the upper critical interaction @xmath76 can always be detected for any value of @xmath7 , the lower critical interaction @xmath77 can not always be found due to some numerical instability ( hence the absence of bars for the red triangles of fig.[fig.phase_diagram ] ) .
fig.[fig.phase_diagram_alt ] shows the same phase diagram as fig.[fig.phase_diagram ] using @xmath4 instead of @xmath11 as the unit of energy to allow an easier comparison with the results of ref .. it is interesting to compare the phase diagram of fig.[fig.phase_diagram ] with the one presented in ref .
for the half - filled square lattice with nearest - neighbor hopping ( more suited to the study of cuprates ) .
this model is different and the method is the variational cluster approximation ( vca),@xcite but it presents the same collinear magnetic phase with ordering wave vector @xmath74 or @xmath75 , for @xmath7 larger than @xmath78 .
we saw that the results obtained with the q1d cluster differ from those for the sp cluster .
we have checked that the assumption that the mott transition occurs when @xmath36 is of the order of the bandwidth is not sufficient to choose the appropriate cluster , although in this context the non - monotonic dependence on @xmath7 of the mott metal - insulator transition is suggestive of the inadequacy of the q1d cluster . however , based on the symmetries satisfied by the sp cluster , as illustrated in fig.[fig.sym ] , it should capture the correct physics .
this is confirmed by the fact that the predictions for the magnetic state obtained with the sp cluster agree with the results of other methods that are available for comparisons . with the q1d cluster
there is no commensurable magnetic state at all in the m1d regime while other methods find stable magnetic long - range order .
first , we searched for collinear magnetic state @xmath74 using the restricted hartree - fock approximation on a 18@xmath5218 cluster .
this method allows one , in principle , to map the phase diagram for a large but finite number of magnetic states and to study larger clusters than in cdmft . here , we used hartree - fock just to confirm our magnetic phase diagram in the m1d regime for the sp cluster .
we thus allowed only two magnetic states : the nel order and the collinear magnetic state .
even though cdmft and hartree - fock methods can not be compared quantitatively , a qualitative agreement is found : for @xmath79 , a first - order metal to antiferromagnetic ( @xmath80 , @xmath80 ) insulator transition takes place at a finite interaction @xmath81 , whereas for @xmath17 , a first order metal to collinear magnetic insulator transition is found .
although the same magnetic states and the same order of transition are found in the same range of @xmath7 as in the cdmft plus ed solver method , the critical hartree - fock interaction @xmath81 has a lower value , around @xmath82 , mainly due to the fact that the hartree - fock method is purely a mean - field theory that neglects the fluctuations that renormalize down the value of the interaction .
kanamori - brckner screening is an example of renormalization mechanism that is neglected in hartree - fock . finally , the critical interaction for collinear magnetism found with the sp cluster exhibits qualitatively the same dependency on frustration as in the phase diagram of ref . obtained by variational methods .
the phase diagram in fig.[fig.tocchio ] presents our results with the same axis as in ref . to ease the comparison .
even though it is not a rigorous proof , the fact that the appearance of a collinear magnetic state for @xmath83 is supported by three different numerical methods gives solid arguments in favor of its presence in this region of the phase diagram .
the lack of collinear magnetism with the q1d cluster is an additional argument , beyond symmetry , to reject that cluster .
in the moderately one - dimensional regime of the hubbard model on the anisotropic triangular lattice , a symmetry - preserving cluster should be preferred to a quasi one - dimensional cluster geometry for calculations with cluster dynamical mean - field theory .
the symmetry - preserving cluster gives magnetic phases in agreement with other methods . with the symmetry - preserving cluster
, we obtained the phase diagram using cdmft with an exact diagonalization solver in the moderately one - dimensional regime .
there is a line of first - order mott transition where the critical @xmath55 monotonically increases with @xmath7 .
we also found a first - order metal - to - collinear magnetic state transition that occurs for @xmath61 and does not allow any spin liquid state to appear in this regime . for @xmath84 , no sign of magnetic states covering the metal - insulator transition has been found .
a spin liquid or a magnetic order which is not commensurate with our cluster , such as a spiral order , might however appear in this region
. our results at @xmath85 are particularly relevant for experiment since they are supposed to describe the organic compound @xmath0-h@xmath1(cat - edt - ttf)@xmath2 that seems to be a good candidate for a spin liquid state.@xcite at @xmath85 , the collinear magnetic state appears at a critical interaction strength that is slightly lower than the mott critical interaction in the normal state .
this would suggest that there can not be a spin liquid state , _
i.e. _ a non - magnetic insulating state , since the magnetic phase covers the mott metal - insulator transition .
however , the uncertainty on the value of @xmath7 and the fact that the magnetic and mott transitions are so close to each other does not allow us to strictly exclude a spin liquid state for @xmath85 .
had @xmath7 been much larger , this possibility would have been excluded .
one should also recall that other physical effects may need to be taken into account to model the real material , such as multi - band effects , near - neighbor repulsion etc .
we are grateful to r. valenti , s. verret and especially m. gringas , for numerous suggestions and discussions .
we also thank p. smon for the ctqmc impurity solver .
this work has been supported by the natural sciences and engineering research council of canada ( nserc ) under grant rgpin-2014 - 04584 , and by the research chair in the theory of quantum materials at universit de sherbrooke ( a .- m.s .
simulations were performed on computers provided by canadian foundation for innovation , the ministre de lducation des loisirs et du sport ( qubec ) , calcul qubec , and compute canada . | spin - liquid behavior was recently suggested experimentally in the moderately one - dimensional organic compound @xmath0-h@xmath1(cat - edt - ttf)@xmath2 .
this compound can be modeled by the one - band hubbard model on the anisotropic triangular lattice with @xmath3 , where @xmath4 is the minority hopping .
it thus becomes important to extend previous studies , that were performed in the range @xmath5 , to find out whether there is a regime where mott insulating behavior can be found without long - range magnetic order . to this end , we study the above model in the range @xmath6 using cluster dynamical mean - field theory ( cdmft ) .
we argue that it is important to choose a symmetry - preserving cluster rather than a quasi one - dimensional cluster .
we find that , upon increasing @xmath7 beyond @xmath8 , the mott transition at zero - temperature is replaced by a first - order transition separating a metallic state from a collinear magnetic insulating state .
nevertheless , at the physically relevant value @xmath3 , the transitions toward the magnetic and the mott insulating phases are very close .
the phase diagram obtained in this study can provide a working basis for moderately one - dimensional compounds on the anisotropic triangular lattice . |
let @xmath1 be a topological dynamical system ( for short tds ) , where @xmath2 is a compact metric space with compatible metric @xmath3 and @xmath4 a continuous @xmath5-action on @xmath6 .
the set @xmath7 denotes the compact convex set of all borel probability measures , and @xmath8 the compact convex set of @xmath9invariant borel probability measures .
we denote by @xmath10 the set of all non - negative integers .
entropy is one of the most widely used notions in the characterization of the complexity of topological dynamical systems . among the notions of entropy
, there are two classical ones which are topological entropy and measure - theoretical entropy . in 1958
kolmogorov @xcite introduced the definition of measure - theoretical entropy for an invariant measure and in 1965 adler _
et al _ @xcite defined topological entropy .
according to goodwyn @xcite , goodman @xcite and misiurewicz @xcite s work , a classical variational principle is found .
the topological variational principle establishes that topological entropy is supremum over all @xmath11 of the measure - theoretical entropy .
in1973 bowen @xcite introduced the topological entropy @xmath12 for any set @xmath13 in a tds @xmath14 resembling hausdorff dimension .
he also proved the remarkable result that @xmath15 for ergodic measure @xmath16 , where @xmath17 denotes the set of _ generic _ points of @xmath16 , and @xmath18 is the measure - theoretical entropy .
bowen s topological entropy plays a key role in topological dynamical systems and dimension theory , see pesin @xcite . in 1983
brin and katok @xcite gave a topological version of the shannon - mcmillan - breiman theorem with a local decomposition of the measure - theoretical entropy .
recently , feng and huang @xcite gave a certain variational relation between bowen s topological entropy and measure - theoretic entropy for arbitrary non - invariant compact set .
i.e. @xmath19 where @xmath13 is any non - empty compact subset of @xmath14 , @xmath20 is the measure - theoretical lower entropy of borel probability measure @xmath16 ( see @xcite ) .
the name * slow entropy * was introduced into dynamical systems by katok and thouvenot@xcite , hochman@xcite for @xmath21-actions . in this paper , we use the carath@xmath22odory dimension structure to study the slow entropy . we define a new topological slow entropy @xmath23 of any subset @xmath24 for higher dimension @xmath0actions like hausdorff dimension inspired by bowen@xcite s definition of topological entropy , but it varies or converges more slowly than bowen s .
we also give a modification of brin an katok s definition of measure - theoretical lower entropy , i.e. measure - theoretical slow entropy @xmath25 for higher dimension @xmath0actions .
we prove that if bowen entropy is positive , the topological slow entropy must be infinite .
we prove a variational principle for slow entropies : @xmath26 where @xmath13 is any non - empty compact subset of @xmath6 .
the paper is organized as follows . in sect .
2 we give the definition of topological slow entropy for @xmath0actions in the form of hausdorff dimension , topological slow entropy using open covers , and some properties . in sect .
3 some examples in a symbolic dynamical system are given . in sect .
4 measure - theoretical slow entropy for @xmath0actions is presented .we prove a variational principle , the topological slow entropy is supermum over all borel probability measure of the measure - theoretical slow entropy .
in this section , we give definitions and some related properties of two slow entropies of subsets in a topological dynamical system : topological slow entropy for @xmath0actions in the form of dimension and using open covers for @xmath27action .
firstly , we introduce the notion of @xmath5-action found in @xcite for convenience .
let @xmath1 be a tds .
a family of continuous transformations @xmath28 is called a continuous @xmath5-action , with @xmath29 or @xmath30,@xmath31 , if @xmath4 satisfies @xmath32,and @xmath33 is the identity map . for @xmath34 and @xmath35 , we set @xmath36 . for @xmath37 , let @xmath38 and @xmath39,where @xmath40 denotes the cardinality of the set @xmath41 .
secondly , the definition of bowen ball for @xmath0actions is as follow : for @xmath42 , we denote by @xmath43 the open bowen ball of radius @xmath44 in the metric @xmath45 around @xmath46 i.e. @xmath47 @xmath48 we now give the definition of slow entropy with @xmath49 or @xmath30-actions .
let @xmath2 be a compact metric space and @xmath4 be a continuous @xmath5-action on @xmath6 with @xmath29 or @xmath30 , @xmath31 .
for @xmath50 .
define @xmath51 where the infimum is taken over all finite or countable families @xmath52 such that @xmath53 and @xmath54 .
clearly , @xmath55 does not decrease as @xmath56 increases and @xmath57 decreases , hence the following limit exists : @xmath58 then we can easily see that there exists the critical value @xmath59 such that @xmath60 finally we set @xmath61 which is called the _
topological slow entropy _ for @xmath62 with respect to @xmath4 .
this quantity @xmath63 is defined in way which resembles the hausdorff dimension , also satisfies most properties like bowen entropy@xcite for @xmath10 action . for convenience
, we use @xmath64 , @xmath65 instead of @xmath55 , @xmath66 without any confusion .
( i)if @xmath67 , then @xmath68 ( ii)if @xmath69 , then @xmath70 \(iii ) @xmath71 ; if @xmath72 is a countable set , then @xmath73 we will only prove proposition ( i ) , others proof are omitted , because we can get them from the definition without difficulty .
suppose a finite or countable family @xmath74 , and @xmath75 corresponding to @xmath76 in the situation of @xmath67 .
suppose @xmath77 we may suppose @xmath78 as well , then @xmath79 where @xmath80 denotes the integral part of a real number @xmath81 .
obviously , @xmath82 which implies @xmath83 . then @xmath84 taking the infimum of both sides , and we get @xmath85 by letting @xmath86 , then @xmath87 , and then @xmath88 letting @xmath89 , we have @xmath90 .
next , we will prove the reverse side . because @xmath76 is uniformly continuous , for @xmath91 such that @xmath92
we suppose bowen family @xmath93 , @xmath94 corresponding to @xmath95 , and @xmath96 , then @xmath97 in fact , for any arbitrary @xmath98 , there exists some @xmath99 such that @xmath100 , i.e. @xmath101 so @xmath102 then @xmath103 , and then @xmath104 .
we notice that any finite or countable family @xmath105 covering @xmath62 contains @xmath106 .
so we get @xmath107 and @xmath108 letting @xmath86 , hence @xmath109 . and noticing @xmath110 arbitrary , we get @xmath111 .
we point out @xmath112 bowen topological entropy defined by using bowen balls . for details , see[18 , page 74 ] .
\(i ) for @xmath10-action , @xmath72 , @xmath113 \(ii ) for @xmath10-action , if @xmath114 , then @xmath115 .
\(i ) is obvious . we only prove ( ii ) .
we suppose @xmath116 . from the definition of bowen entropy , for arbitrary @xmath117
, there exists @xmath118 , for arbitrary @xmath119 , we have @xmath120 which implies @xmath121 . because @xmath122 increases as @xmath56 ,
so @xmath123 then for any subfamily @xmath124 of @xmath62 , @xmath125 . for arbitrary real number @xmath126 , because @xmath127 , i.e. for arbitrary @xmath128 , there is some @xmath129 , for @xmath130 , @xmath131 . hence @xmath132 . and thus @xmath133 that is @xmath134 as @xmath135 .
letting @xmath86 , we have @xmath136 so @xmath137 because @xmath57 and @xmath13 are arbitrary , hence @xmath138 .
thus we complete the proof .
we now give a equivalent definition of @xmath139 for the situation @xmath67 , which comes from bowen @xcite .
we set @xmath14 be a tds , @xmath140 a finite open cover of @xmath6 .
@xmath141 , let @xmath142 for @xmath143 , we define @xmath144 then for each @xmath145 , set @xmath146 we write @xmath147 for the case @xmath148 by convention .
when @xmath149 , @xmath150 increases , therefore we could define @xmath151 we notice that if @xmath152 , then @xmath153 . we define _ bowen topological slow entropy _
@xmath154 of @xmath155 with respect to @xmath76 as follows : @xmath156 and _ bowen topological slow entropy _ of @xmath62 as follows : @xmath157 @xmath158 it can be similarly proved using techniques in @xcite .
we set @xmath14 be a tds , @xmath140 a finite open cover of @xmath6 .
set @xmath159 to be the minimal cardinality of sub - families @xmath160 with @xmath161 , where @xmath162 .
and we write @xmath163 .
obviously , @xmath164 .
let @xmath165 @xmath166 increases with respect to @xmath155 .
we define the topological slow entropy of @xmath62 by @xmath167 * remark * : from the definition of topological slow entropy and topological entropy using open covers , there exists a relation : @xmath168 noticing @xmath169 as @xmath170 .
therefore , if @xmath171 , then @xmath138 ; if @xmath172 , then @xmath173 .
@xmath174 we only prove @xmath175 .
suppose @xmath176 such that @xmath177 .
obviously , @xmath178 , @xmath179 and we get @xmath180 so @xmath181 and if @xmath182 , then @xmath183 .
therefore , @xmath184
we take examples with a continuous @xmath185-action to explain the new topological slow entropy in a symbolic dynamical system with a special metric .
suppose a finite alphabet @xmath186 , where @xmath187 .
@xmath188 suppose @xmath189 , let @xmath190 endow @xmath191 with the metric @xmath192 for @xmath193 , then @xmath194 is a compatible metric for @xmath195 .
@xmath191 is the onesided full shift of @xmath196 symbols , i.e. for @xmath197 , we define the shift action @xmath198 as @xmath199 and then @xmath200 is a continuous @xmath5 -action on @xmath195 . for any subset @xmath201 , @xmath202 , where @xmath203 is denoted the hausdorff dimension in @xmath2(see@xcite ) .
in fact , for @xmath204 we set a cylinder set as @xmath205 it is not hard to see the @xmath206 hausdorff outer measure of @xmath62 can be @xmath207 and where the infimum is taken over all finite or countable family @xmath208 , which covers @xmath62 with @xmath209 .
let @xmath44 be sufficiently small and choose @xmath210 such that @xmath211 .
so it also follows from the choice of the metric @xmath194 that @xmath212 for all @xmath213 and @xmath214 . comparing the two definitions
, we get the result . for any real number @xmath215
, there exists a compact subset @xmath216 , such that @xmath217 . from mattila@xcite , we know the fact : if the hausdorff dimension of a set @xmath218 is @xmath219 , then for any @xmath220 , there exists a compact subset @xmath221 such that @xmath222 . from theorem 3.1 ,
we get @xmath223 , this means @xmath224 . in the symbolic system @xmath225 , for any non - empty subset @xmath226 , @xmath227 , where @xmath228 is denoted the upper box dimension of @xmath62 ( see @xcite ) .
suppose @xmath226 , for @xmath229 , let @xmath230 be the smallest number of @xmath231ball needed to cover @xmath62 : @xmath232 suppose @xmath233 be a generator , @xmath234 , @xmath235 and every @xmath236 is clopen set for @xmath237 .
suppose @xmath238 for @xmath239 , for any @xmath240 , bowen ball @xmath241 with metric @xmath194 be the cylinder set as follows : @xmath242 noticing @xmath243 , the union of elements of @xmath244 covers @xmath62 , then @xmath245 similarly , @xmath246 , and then @xmath247 from ( 3.1),(3.2 ) , therefore , @xmath248 taking @xmath249 , @xmath250 and supper limit , we get @xmath251 if @xmath2 is a compact metric space , @xmath72 , @xmath252 is a @xmath231cover of @xmath62 , if @xmath253 for @xmath44 and @xmath254 , where @xmath255 denotes the diameter of @xmath256 .
if @xmath2 and @xmath257 are both compact metric spaces , @xmath258 is a map , and if there exists @xmath259 , so that @xmath260 , @xmath261 for @xmath262 , @xmath263 is a bi - lipschitz map i.e. @xmath264 then for any @xmath72 , we have @xmath265 .
suppose @xmath266 is a @xmath267cover of @xmath62 , since @xmath268 is a bi - lipschitz map for aibitray @xmath269 , so @xmath270 then @xmath271 is a @xmath272cover of @xmath273 . for @xmath274 , @xmath275 which implies that @xmath276 taking @xmath277 , we get @xmath278 .
if @xmath279 , then @xmath280 , that is @xmath281 for all @xmath279 , so @xmath282 .
noticing the bi - lipschitz mapping @xmath283 has a inverse mapping @xmath284 and using the above result , we have @xmath285 . in the symbolic system @xmath225 , for arbitray @xmath286 , there exists a @xmath287 subset @xmath221 , @xmath288 , such that @xmath289 .
from proposition 3.2 and lemma 3.1 , for any @xmath286 , there is a compact subset @xmath226 , @xmath290 .
we set @xmath291 , then @xmath221 is a @xmath287 subset , @xmath288 , and @xmath292 , which implies @xmath293 .
we firstly introduce the measure - theoretic slow entropy . the notion of weighted topological slow entropy
is presented , which is important to prove the variational principle . * remark * : from the definition of measure - theoretic slow entropy , it is easy to know @xmath305 because @xmath306 as @xmath170 . if @xmath25 be finite , then @xmath307 ; if @xmath308 , @xmath25 must be infinite .
it s not difficult to prove that there exists a critical value of parameter @xmath219 , which we will denote by @xmath318 , such that @xmath319 we call @xmath318 the _ weighted topological slow entropy _ of @xmath62 with respect to @xmath4 .
@xcite let @xmath2 be a compact metric space and @xmath329 be a family open of ( or closed ) balls in @xmath6 .
then there exists a finite or countable subfamily @xmath330 of pairwise disjoint balls in @xmath331 such that @xmath332 _ proof of proposition4.2 .
_ we follow the argument in @xcite for the condition of @xmath5-actions .
let @xmath333 , set @xmath334 in the definition of weighted topological entropy , we have @xmath335 for @xmath336 .
we denote @xmath344 and @xmath345 for @xmath346 we write @xmath347 for @xmath348 .
obviously we may assume@xmath349 for @xmath350 . for @xmath286 ,
set @xmath351 and @xmath352 we divide the proof of ( 4.2 ) into the following three steps .
step 1 . for each @xmath353 , and @xmath286 , there exists a finite set @xmath354 such that the ball @xmath355 are pairwise disjoint , @xmath356 , and @xmath357 we will use the method of federer @xcite , also mattila @xcite for @xmath358actions .
since @xmath359 is finite , by approximating the @xmath360 s from above , we may assume that each @xmath360 is a positive rational , and then by multiplying with a common denominator we may assume that each @xmath360 is a positive integer .
let @xmath361 be the least integer with @xmath362 .
denote @xmath363 , and define @xmath364 , by @xmath365 .
since @xmath366 , for @xmath350 , so @xmath367 is well defined .
we define by introduction integer - valued functions @xmath368 on @xmath331 and sub - families @xmath369 of @xmath331 starting with @xmath370 . using lemma 4.1 repeatedly , we define inductively for @xmath371 , disjoint subfamilies @xmath372 of @xmath331 such that @xmath373 @xmath374 and the functions @xmath375 such that @xmath376 this is possible for @xmath377 , @xmath378 whence every @xmath379 belongs to some ball @xmath380 with @xmath381 .
thus @xmath382 choose @xmath383 so that @xmath384 is the smallest .
then @xmath385 so @xmath386 is desired .
assume @xmath388 , otherwise ( 4.3 ) is obvious .
since @xmath389 , @xmath390 for large enough @xmath196 .
let @xmath391 be the sets constructed in step 1 .
then @xmath392 for large enough @xmath196 .
set @xmath393 .
note that the family of all non - empty compact subsets of @xmath6 is compact with respect to hausdorff distance(federer[5,2.10.21 ] ) .
it follows that there is a subsequence @xmath394 of natural numbers and a non - empty compact set @xmath395 such that @xmath396 converges to @xmath397 in the hausdorff distance as @xmath398 .
since any two points in @xmath399 have a distance ( with respect to @xmath400 ) not less than @xmath57 , so do the points in @xmath397 . thus @xmath397 is a finite set , moreover , @xmath401 when @xmath402 is large enough .
hence @xmath403 when @xmath402 is large enough , and thus @xmath404 . by the way ,
since @xmath405 when @xmath402 is large enough , we have @xmath406 therefore @xmath407 step 3 . for any @xmath408 , we have @xmath409 which implies ( 4.2 ) .
in fact , fix @xmath408
. then @xmath410 .
thus by proposition 4.1(i ) and ( 4.3 ) , we get @xmath411 this completes the proof of the proposition .
clearly @xmath416 .
we define a function @xmath417 on the space @xmath418 of continuous real - valued functions on @xmath6 by @xmath419 let @xmath420 denote the constant function @xmath421 .
it is easy to verify that by the haha - banach theorem , we can extend the linear functional @xmath430 , from the subspace of constant functions to a linear functional @xmath431 satisfying @xmath432 if @xmath423 with @xmath433 , then @xmath434 and so @xmath435 .
hence we can use the riesz representation theorem to find a borel probability measure @xmath16 on @xmath6 such that @xmath436 for @xmath423 .
next , we prove @xmath414 . for any compact set @xmath437 , by urysohn lemma there exists @xmath423 such that @xmath438 for @xmath439 and @xmath440 for @xmath441
. then @xmath442 and thus @xmath443 .
hence @xmath444 .
this shows @xmath445 that is @xmath414 . in the end , we prove @xmath446 for @xmath447 .
in fact , for any compact set @xmath448 , by urysohn lemma again , there is @xmath423 , such that @xmath449 for @xmath450 and @xmath451 for @xmath452
. then @xmath453 .
since @xmath454 , and @xmath455 , we get @xmath456 and hence @xmath457 .
therefore , we have @xmath458 .
it follows that @xmath459 this completes the proof of the lemma . * remark * : there is a related slow entropy distribution principle . using techniques in@xcite
, we have : for any borel set @xmath460 and borel probability measure @xmath16 on @xmath221 , if @xmath461 for all @xmath439 , then @xmath462 ; if @xmath463 for all @xmath464 , then @xmath465 firstly , we prove @xmath468 , for any @xmath469 .
we set @xmath470 for @xmath471 .
it s easy to see that @xmath472 is nonnegative and increases as @xmath57 decreases . by the monotone convergence theorem
, we get @xmath473 thus to show @xmath468 , we only to show @xmath474 for any @xmath44 .
now we fix @xmath475 , set @xmath476 , then exist a borel set @xmath477 such that @xmath478 let @xmath479 ba a finite or countable family such that @xmath480 , and @xmath481 .
we may as well assume that for each @xmath269 , @xmath482 , and select @xmath483 .
then by ( 4.4),we have @xmath484 so , we get @xmath485 therefore , @xmath486 . letting @xmath487 , we get @xmath474 .
thus @xmath468 .
we next prove @xmath488 .
we may as well assume @xmath489 , otherwise the conclusion is obvious .
by proposition 4.2 , @xmath490 .
suppose @xmath491 , then there exists@xmath492 such that @xmath493 .
by lemma 4.2 , there exists @xmath494 , such that @xmath495 for any @xmath496 . and
then @xmath497 for each @xmath498 . therefore , @xmath499 the proof is completed .
goodwyn , comparing topological entropy with measure - theoretic entropy , _ proc .
* 94 * ( 1972 ) 366 - 388 .
m.hochman , slow entropy and differentiable models for infinite - measure preserving @xmath21 actions , to appear in _ ergod .
dynam . sys . _ , 2011 .
kolmogorov , a new metric invariant of transient dynamical systems and automorphisms of lebesgue spaces , _ dokl .
* 119 * ( 1958 ) 861 - 864 .
a.katok and j.p.thouvenot , slow entropy type invanriants and smooth realization of commuting measure - preserving transformations , _ ann .
h.poincare probab .
, _ * 33(3 ) * ( 1997 ) 323 - 338 .
p.mattila , geometry of sets and measures in euclidean spaces .
cambridge university press , 1995 .
m.misiurevicz , topological conditional entropy , _ studia math . _
* 2(55 ) * ( 1976 ) 175 - 200 .
j.ma and z.wen , a billingsley type theorem for bowen entropy , _ c.r.math.acad.sci.paris _ * 346 * ( 2008 ) , no.9 - 10 , 503 - 507 . | this paper defines and discusses the dimension notion of topological slow entropy of any subset for @xmath0actions .
also , the notion of measure - theoretic slow entropy for @xmath0actions is presented , which is modified from brin and katok @xcite .
relations between bowen topological entropy [ 3,17 ] and topological slow entropy are studied in this paper , and several examples of the topological slow entropy in a symbolic system are given . specifically , a variational principle is proved . |
the study of dynamics and pattern formation in nonlinear spatially extended dynamical systems continues to attract considerable attention @xcite , with potential solution behaviors ranging from attracting steady states to complex spatiotemporal dynamics including spatiotemporal chaos ( stc ) , which is characterized by temporally chaotic dynamics with decaying spatial correlations and a finite density of positive lyapunov exponents @xcite . in the search for general principles ,
much effort has gone into the derivation and study of simplified model equations to highlight and clarify pattern - forming properties of more complicated `` full '' systems .
such paradigmatic models , such as the ginzburg - landau equation for the slow evolution of modulations of an underlying pattern , have been remarkably fruitful in predicting the leading - order scaling behavior and stability of patterns , by comparison with numerical or experimental results .
the identification of relevant scaling laws forms an integral part of the asymptotic procedures used to derive such reduced model systems .
the successful application of such methods rests on the expectation that the leading - order truncation obtained subject to a consistent scaling assumption indeed captures the dominant scaling of the full problem , in the sense that the perturbative effects of neglected terms are `` small '' and occur at the scaling implied by the dominant balance . in this paper , we present an example which appears to challenge this expectation . it is known from various examples that the stability of finite - wavelength pattern or wave phenomena may be complicated through interaction with a mean flow ; for instance , in convection with free - slip boundary conditions @xcite , in some reaction - diffusion systems @xcite and elsewhere , such coupling can lead to unusual instability mechanisms of the underlying pattern or flow , whose analysis may display mixing of scales in the perturbation parameter and require higher - order terms in the perturbative expansions for a complete understanding @xcite . in the present work we study a model for one - dimensional pattern formation in the presence of continuous symmetry , for which the interaction of modes with two well - separated length scales leads to spatiotemporal chaos .
focussing in this case on characterizing the long - term chaotic regime rather than on its onset via the initial stability problem , for the statistics of the long - wave mode we find a scaling discrepancy between predictions of leading - order multiple - scale analysis and computations of the full system .
unexpectedly , though , the agreement in scaling is restored by including next - order corrections in the amplitude equations .
we investigate the nikolaevskiy partial differential equation ( pde ) , given in canonical ( derivative ) form as @xmath0 u\ ] ] ( where @xmath1 ) , which was originally proposed as a model for longitudinal seismic waves in viscoelastic media @xcite , and has more recently been obtained in the context of phase dynamics in reaction - diffusion systems @xcite and finite - wavelength transverse instabilities of traveling fronts @xcite .
in fact , the nikolaevskiy equation has a continuous symmetry in the form it is galilean - invariant , and preserves the spatial mean and appears to be a paradigmatic model for short - wave pattern formation with such symmetry @xcite .
considering individual fourier modes of the form @xmath2 , the linear dispersion relation about the trivial solution @xmath3 is @xmath4 $ ] . since for @xmath5 ,
all initial conditions of decay to a spatially homogeneous solution , in the following we consider only @xmath6 and write @xmath7 ; as usual the @xmath8 width of the unstable band and @xmath9 maximum growth rate suggest natural slow length and time scales @xmath10 , @xmath11 . for @xmath12 , the nikolaevskiy equation
has a one - parameter ( up to translation ) family of stationary periodic solutions , or rolls , @xmath13 for @xmath14 , which may be found by weakly nonlinear analysis to be @xcite @xmath15 previous studies of have revealed some surprises . due to the presence of a symmetry - induced long - wave ( goldstone ) mode
@xmath16 interacting with the weakly unstable pattern mode @xmath17 , all roll equilibria @xmath13 are unstable for all @xmath18 @xcite ; instead , there is a direct supercritical transition from spatial uniformity to spatiotemporal chaos @xcite .
similar behavior occurs , for instance , in electroconvection in liquid crystals @xcite , and has been termed `` soft - mode turbulence '' @xcite . by deriving leading - order modulation equations using multiple - scale analysis , matthews and
cox @xcite furthermore showed that in the chaotic state , @xmath17 ( and hence @xmath19 ) has a rather unusual @xmath20 scaling . in the present work , through extensive numerical studies
we show that nikolaevskiy stc exhibits yet more unexpected scaling : amplitudes and correlation times of the large - scale mode @xmath16 have an anomalous dependence on @xmath21 , seemingly inconsistent with leading - order multiple - scale analysis , though recovered with the addition of next - order terms to the amplitude equations
. in particular , in the chaotic regime the mean - square amplitude of @xmath16 appears to scale as @xmath22 , rather than the predicted @xmath23 .
we numerically computed ( mean - zero ) solutions of the nikolaevskiy equation on an @xmath24-periodic domain in the chaotic regime , over a wider @xmath25-range than in earlier studies : @xmath26 .
the system size @xmath24 was chosen so that the number of fourier modes in the unstable band of the dispersion relation @xmath27 , about @xmath28 , sufficed to capture the large - system limit ; to minimize system size effects across our computations , we fixed the domain length in the slow space variable @xmath10 , at @xmath29 with @xmath30=1 and 2 ( about 12 and 25 unstable modes , respectively ) , and checked our results with @xmath30=4 .
we used a pseudo - spectral method in space with @xmath31 fourier modes for @xmath32=11 to 16 , chosen so that the maximum wave number retained was well within the strongly decaying regime of @xmath27 : @xmath33 .
our time integration was performed using a fourth - order runge - kutta exponential time differencing ( etdrk4 ) scheme @xcite , with a typical time step of @xmath34 , fixed in the slow time variable @xmath11 ; after a transient , we computed statistics over a time interval of length @xmath35 .
the ( time - averaged ) power spectrum @xmath36 of fig .
[ fig : nikps ] captures the dominant features of the numerically observed chaotic dynamics : as pointed out by tanaka @xcite , for `` large '' @xmath25 the stc appears qualitatively similar to that of the related well - known kuramoto - sivashinsky equation @xcite ; fourier modes with @xmath37 are ( at most ) weakly damped , and the power spectrum reveals no significant scale separation for @xmath38 ( dashed curves ) .
however , for sufficiently `` small '' @xmath25 , we observe the distinct gap between the @xmath39 and @xmath40 energetic modes which characterizes `` nikolaevskiy chaos '' @xcite .
note that for @xmath41 , the subdominant peaks at @xmath42 represent higher harmonics slaved to the fourier modes @xmath40 via the quadratic nonlinearity ; we remark that exponential time differencing integration schemes @xcite are particularly well - suited to capturing this slaving behavior and improving the accuracy at small scales .
the strong scale separation of the spatiotemporally chaotic dynamics in this small-@xmath25 regime is readily apparent in a representation of a typical numerically computed time series on the multiple - scale chaotic attractor of the nikolaevskiy pde , as in fig .
[ fig : nikxt ] , or of a snapshot of a typical solution as seen in fig .
[ fig : nikux](a ) : solutions on the attractor are characterized by long - wave modulations , aperiodic in space and time , of small - scale rolls of wavelength @xmath43 ( corresponding to the linearly unstable modes of @xmath27 ) . of a solution @xmath44 ( blue ) of the nikolaevskiy equation with @xmath45 and domain length @xmath46 .
the figure also shows the reconstructed mean mode @xmath16 ( black ) and envelope @xmath47 ( black dashed line ) .
( b ) instantaneous spectrum @xmath48 of the solution in ( a ) , showing the wavenumber ranges for the fourier filters.,width=307 ] this motivates a natural decomposition of the solution @xmath49 of the nikolaevskiy pde : with @xmath10 , @xmath11 , we have @xmath50 where @xmath17 and @xmath16 are slowly varying envelopes of the dynamically significant fourier modes near @xmath51 and @xmath52 , respectively .
we have implemented fourier filters to extract @xmath16 and @xmath17 at each time step from the fourier space numerical solution @xmath53 of the full pde ; see fig . [
fig : nikux](b ) : for some @xmath54 , we identify the modes with @xmath55 as being `` large - scale '' , while those with @xmath56 belong to the patterned mode and its higher harmonics . in view of the strong suppression of the linearly damped modes near @xmath57 for @xmath58 ( see fig .
[ fig : nikps ] ) , we may safely choose @xmath59 ; we have verified that our results are insensitive to the choice of @xmath60 .
thus we obtain @xmath16 from @xmath61 , and @xmath62 and hence @xmath17 from @xmath63 . in fig .
[ fig : nikux](a ) we have also shown the mean mode @xmath16 and the reconstructed envelope @xmath47 .
in the light of the decomposition for solutions of , one may seek a modulation theory for the slow evolution of the envelopes , coupling the amplitude @xmath17 of the finite - wavelength spatial pattern mode with the long - wave mode @xmath16 induced by the continuous symmetry @xcite . in this context
it is of particular interest to determine the exponents @xmath64 and @xmath65 describing the scaling of @xmath17 and @xmath16 as @xmath66 .
thus one postulates the ansatz @xmath67 where @xmath68 and @xmath69 are assumed @xmath70 , and @xmath69 has mean zero . the @xmath8 amplitude of the known roll solutions @xmath13 suggests the apparently natural choice @xmath64=1 ( see for instance @xcite ) , familiar from other pattern - forming systems @xcite , which we will denote the ginzburg - landau ( gl ) scaling .
it turns out , though , as discussed in @xcite , that amplitude equations for derived using @xmath64=1 in are asymptotically inconsistent for any choice of @xmath65 if @xmath71 . instead ,
as shown by matthews and cox @xcite , the only asymptotically consistent amplitude equations are obtained by using the ansatz with @xmath64=3/2 , @xmath65=2 ; we shall call this the mc scaling .
standard multiple - scale techniques then yield the leading - order amplitude equations proposed to describe stc in the nikolaevskiy equation : the matthews - cox ( mc ) equations @xcite are @xmath72 an immediate prediction of with the mc exponents , that the amplitude of solutions @xmath49 of ( computed as the time - averaged root - mean - square ( rms ) amplitude @xmath73 , where @xmath74 and @xmath75 denote time and space averages , respectively ) should scale as @xmath76 , was verified in @xcite ; see @xcite for an analogous result in two space dimensions .
we have performed the multiple - scale analysis of to the next order in @xmath25 , using the ansatz with the mc scaling @xmath64=3/2 , @xmath65=2 , to obtain the asymptotically consistent @xmath8 amplitude equations @xmath77 \nonumber \\ & & - 2 i { \varepsilon}\left [ a_x + 6 a_{xxx } \right ] , \label{eq : mce1 } \\
f_t & = & f_{xx } - |a|^2_x - { \varepsilon}f f_x .
\label{eq : mce2 } \end{aligned}\ ] ] ( the linear @xmath8 terms in correspond to next - order corrections to the exact dispersion relation @xmath27 ) . observe that unlike the @xmath70 equations
, these @xmath8 equations also capture the roll solutions @xmath13 of the gl regime , which have size @xmath78 in this scaling .
one would expect , though , that in the chaotic mc regime ( in which @xmath79 , @xmath80 ) , the @xmath8 correction terms in would lead to small @xmath8 perturbations to the solutions and their scaling .
following previous studies @xcite , we set out to verify that @xmath17 and @xmath16 indeed scale for small @xmath25 as predicted by the mc exponents @xmath64=3/2 , @xmath65=2 . with a view to isolating and thereby computing the individual scaling exponents directly , tanaka @xcite numerically integrated the pde on a domain of length @xmath24=512 for @xmath81 $ ] , and computed the @xmath82 scaling of the fourier coefficients nearest in his system to @xmath83=0 and @xmath83=1 ; these fourier modes were used as proxies for @xmath16 and @xmath17 to argue that the results were consistent with mc scaling .
however , the numerical evidence of ( * ? ? ?
5 ) does not appear sufficiently well - averaged to establish the asserted @xmath9 large - scale scaling convincingly .
our implementation of fourier filters to extract the full slowly - varying envelopes @xmath16 and @xmath17 from the computed solution @xmath49 of the nikolaevskiy pde provides a more powerful method to estimate the critical exponents , via the @xmath25-dependence of the time - averaged @xmath82 magnitudes of @xmath16 , @xmath17 . in fig .
[ fig : u1u0rms ] we plot @xmath84 and @xmath85 computed as functions of @xmath25 for @xmath86 $ ] . , and anomalous scaling of the large - scale mode @xmath16 : + ( a ) scaling of averaged rms amplitudes of @xmath87 from and @xmath88 from
; we find @xmath89 for @xmath90 .
the ( black ) dashed line indicates the mc scaling prediction @xmath91 .
+ ( b ) scaling of averaged rms amplitudes of @xmath16 from and @xmath92 from ; we find @xmath93 for @xmath94 ( black dashed reference line ) . the ( blue ) dotted line indicates the mc prediction @xmath95 .
+ symbols for this and the next figures ( hardly distinguishable here ) : ( magenta and red ) triangles : @xmath96 ; ( blue ) circles and ( black ) diamonds : @xmath97 .
@xmath98 , @xmath99 : nikolaevskiy pde , @xmath100 , @xmath30=1 and 2 ; @xmath101 , @xmath102 : @xmath8 amplitude equations , @xmath103 , @xmath30=1 and 2.,width=307 ] as seen in fig .
[ fig : u1u0rms](a ) , we obtain @xmath104 for @xmath105 ; that is , we find in agreement with @xcite that the pattern mode @xmath17 indeed satisfies the mc scaling @xmath106=@xmath64=3/2 @xcite .
the computed scaling of the reconstructed large - scale mode @xmath16 is rather more surprising : as shown in fig .
[ fig : u1u0rms](b ) , over a wide range @xmath86 $ ] our data is consistent with @xmath107 with the anomalous exponent @xmath108 , contrary to the prediction @xmath109=@xmath65=2 from multiple - scale analysis . in the alternative representation of our data in fig .
[ fig : frms ] the deviation from the mc prediction is more readily apparent : @xmath110 is clearly not @xmath25-independent , as confirmed also by some larger - domain computations performed as an additional check . from and @xmath69 from . this is the data of fig .
[ fig : u1u0rms](b ) rescaled vertically by @xmath111 to emphasize the anomalous scaling : according to the prediction of the matthews - cox scaling ansatz , which for the large - scale mode is @xmath112 , @xmath113 , these values are expected to be @xmath25-independent .
+ symbols are as in fig .
[ fig : u1u0rms ] .
the solid black squares are obtained from additional computations of the nikolaevskiy pde for larger @xmath114 ( @xmath30=4).,width=307 ] the leading - order amplitude equations , being @xmath25-independent , are manifestly unable to capture this unexpected scaling behavior .
however , for comparison with the full nikolaevskiy pde , we have integrated the @xmath8 amplitude equations for @xmath68 and @xmath69 as functions of the slow variables @xmath115 $ ] and @xmath116 , for the same @xmath25- and @xmath117-values as before .
the functions @xmath49 reconstructed from @xmath68 and @xmath69 using are qualitatively similar to computed solutions of ; for @xmath118 trajectories appear chaotic and are not observed to settle down .
we extracted amplitude scaling behavior for by computing the time - averaged @xmath82 amplitudes of @xmath88 and @xmath92 ; as shown in fig .
[ fig : u1u0rms ] , the results agree remarkably well with those obtained for @xmath87 and @xmath16 from the full pde .
equivalently , the statistics of the pattern mode @xmath68 are , unsurprisingly , asymptotically @xmath25-independent ; but for the large - scale mode , we find that approximately @xmath119 , or , as shown in fig . [ fig : frms ] , @xmath120 : whereas one might have anticipated a small ( @xmath8 ) effect of the added @xmath8 terms in the amplitude equations
, in fact they are able to capture the ( asymptotically large ) corrections to mc scaling .
we observe in fig .
[ fig : frms ] a possible weak dependence of @xmath121 ( and of @xmath110 ) on the system size @xmath122 , which may merit further study ; but that for each @xmath117 we have the same scaling with @xmath25 ( with some possible finite - size effects appearing at the smaller domain size for @xmath123 ) .
the deviation for @xmath124 , where the separation of scales is weak and the `` perturbation '' parameter is relatively large , is unsurprising .
the corrections to scaling appear not only in the rms amplitudes : while the multiple - scale ansatz introduces only one slow time scale @xmath11 , a glance at time series of fourier coefficients @xmath53 computed from for @xmath83@xmath1250 and @xmath83@xmath1251 , or at the evolution of @xmath16 and @xmath87 , suggests that for small @xmath25 , at large scales the dynamics are much slower than for the pattern mode . to begin to quantify this observation of potentially distinct time scales , for solutions of we considered time series @xmath126 and @xmath127 of the @xmath82 mode amplitudes depending on the slow time variable @xmath116 , and computed ( full - width at half - maximum ) autocorrelation times @xmath128 , @xmath129 as functions of @xmath25 . again for comparison , we similarly computed correlation times @xmath130 , @xmath131 from time series of @xmath82 values of @xmath69 , @xmath132 computed via . our results are shown in fig . [
fig : tau1tau0 ] ; again , the @xmath8 amplitude equations closely match the behavior of the full pde . from , and
@xmath132 from ; we find @xmath133 , decreasing slightly for larger @xmath117 .
( b ) scaling of correlation times for the rms values of @xmath16 from and @xmath69 from . symbols are as in fig .
[ fig : u1u0rms].,width=307 ] the asymptotic validity of the leading - order modulation equations
implies that these correlation times should be ( asymptotically ) @xmath25-independent .
indeed , fig .
[ fig : tau1tau0](a ) shows these predictions again satisfied for the pattern mode @xmath17 : we find @xmath133 , seemingly independent of @xmath25 ( though with , again , an apparent weak dependence on domain size @xmath117 ) .
however , as seen in fig .
[ fig : tau1tau0](b ) the long - wave mode @xmath16 is substantially slower : while our statistics are not yet well - converged ( particularly for small @xmath25 ) , our data appear to indicate the presence of another time scale diverging as @xmath66 , and scaling possibly as @xmath134 , where our results suggest @xmath135 may be near 4/3 .
the description that emerges for the nikolaevskiy pde with sufficiently small @xmath25 is of spatiotemporally chaotic dynamics with strong scale separation and anomalous long - wave behavior .
the solution space features two distinct scaling regimes : the @xmath8 gl regime , which contains the roll solutions @xmath13 together with their stable manifolds , is unstable to spatially varying perturbations , which lead to collapse to a smaller , apparently attracting , set in phase space , the @xmath136 mc regime of `` nikolaevskiy chaos '' .
our observed anomalous scaling of the long - wave mode @xmath16 implies that the leading - order amplitude equations do not fully describe the dynamics on the attractor of nikolaevskiy stc , at least for the parameter range @xmath137 $ ] we investigated .
this conclusion is supported by recent observations of non - extensive behavior in the matthews - cox equations @xcite .
the higher - order modulation equations , however , do appear to capture the corrections to amplitude and time scaling : the function @xmath138 , evolving according to the coupled equations
, scales in the same way as @xmath139 , where @xmath16 is the large - scale mode extracted from solutions of by fourier filtering .
further investigation of is needed to clarify the mechanism by which the addition of @xmath8 terms to the mc equations modifies the scaling .
this work was partially supported through an nserc grant to r.w .
we would like to thank david muraki for extensive valuable discussions and comments concerning this work .
r.w . would like to acknowledge helpful conversations with participants at the ` pattern formation in large domains ' program at the newton institute for mathematical sciences , cambridge .
we also thank youngsuk lee for useful discussions , and the irmacs centre at simon fraser university for its support . | the nikolaevskiy model for pattern formation with continuous symmetry exhibits spatiotemporal chaos with strong scale separation .
extensive numerical investigations of the chaotic attractor reveal unexpected scaling behavior of the long - wave modes .
surprisingly , the computed amplitude and correlation time scalings are found to differ from the values obtained by asymptotically consistent multiple - scale analysis .
however , when higher - order corrections are added to the leading - order theory of matthews and cox , the anomalous scaling is recovered . |
measuring the quark condensate remains the main concern for physicists of non perturbative quantum chromodynamics .
the purest process allowing a direct measurement of this parameter is @xmath4 scattering .
information concerning the latter can be obtained from the rescattering of two pions in the final state of pionium @xcite , @xmath5 @xcite , or @xmath2 @xcite decays .
let @xmath6 be the phase of a two - pion state of angular momentum @xmath7 and isospin @xmath8 and consider the charged @xmath2 decay process @xmath9 where the lepton @xmath10 is either a muon @xmath11 or an electron @xmath12 , and @xmath13 stands for the corresponding neutrino . in the isospin limit , the decay amplitude @xmath14 for process ( [ eq : process ] ) can be parameterized in terms of three vectorial ( @xmath15 , @xmath16 , and @xmath17 ) and one anomalous ( @xmath18 ) form factors : @xmath19 \right .
\nonumber \\ & & \left .
-\dfrac{1}{m_{k^{\pm}}^3}\ , \epsilon^{\mu\nu\rho\sigma}(p_{\ell}+p_{\nu})_{\nu } ( p_1+p_2)_{\rho}(p_1-p_2)_{\sigma}h\right\rbrace \,,\end{aligned}\ ] ] where @xmath20 denotes the cabibbo - kobayashi - maskawa flavor - mixing matrix element and @xmath21 is the so - called fermi coupling constant .
note that form factors are made dimensionless by inserting the normalizations , @xmath22 and @xmath23 .
the fact that we have used the _ charged _ kaon mass is a purely conventional matter and corresponds to the choice of defining the isospin limit in terms of charged masses .
form factors are analytic functions of three independent lorentz invariants , @xmath24 and the angle @xmath25 formed by @xmath26 , in the dipion rest frame , and the line of flight of the dipion as defined in the kaon rest frame @xcite . in the following
, we will be interested only in two form factors , @xmath15 and @xmath16 , and consider the partial wave expansion , @xmath27 where a convenient parametrization of @xmath28 , @xmath29 , @xmath30 , and @xmath31 in the experimentally relevant region has been proposed in ref .
@xcite .
the currently running na48 experiment aims at measuring form factors for @xmath2 decay of the charged kaon with an accuracy better than the one offered by previous measurement @xcite .
the outgoing data on form factors contain , besides a strong interaction contribution , a contribution coming from the electroweak interaction .
the latter breaks isospin symmetry and is expected to be sizable near the @xmath4 production threshold @xcite . in order to extract @xmath4 scattering parameters from the na48 measurement
, the isospin breaking correction to form factors should therefore be under control . in this direction
, we recently published analytic expressions for @xmath15 and @xmath16 form factors calculated at one - loop level in the framework of chiral perturbation theory based on the effective lagrangian including mesons , photons , and leptons @xcite . in the present work , we will use the method proposed in ref .
@xcite to split analytically the isospin limit and isospin breaking part in form factors , allowing a first evaluation of isospin breaking effects in charged @xmath2 decays .
we shall start things off by the general form of the decay amplitude for process ( [ eq : process ] ) as dictated by lorentz covariance , @xmath32 \gamma_{\mu } \right .
\nonumber \\ & & + \dfrac{i}{m_{k^{\pm}}^3}\ , \epsilon^{\mu\nu\rho\sigma}(p_{\ell}+p_{\nu})_{\nu } ( p_1+p_2)_{\rho}(p_1-p_2)_{\sigma}h \nonumber \\ & & \left .
+ \frac{1}{2m_{k^{\pm}}^2}\,[\gamma_{\mu},\gamma_{\nu}]\,p_1^{\mu}p_2^{\nu}\,t\right\rbrace v(\boldsymbol{p}_l)\ , .
\nonumber\end{aligned}\ ] ] the quantities @xmath0 , @xmath1 , @xmath33 , and @xmath34 , will be called the _ corrected _ @xmath2 form factors since their isospin limits are nothing else than the @xmath2 form factors , @xmath15 , @xmath16 , @xmath17 , and @xmath18 , respectively .
the tensorial form factor @xmath35 is purely isospin breaking and has been calculated at leading chiral order in ref .
the corrected form factors as well as the tensorial one are analytic functions of five independent lorentz invariants , @xmath36 , @xmath37 , @xmath25 , @xmath38 , and @xmath39 .
@xmath38 is the angle formed by @xmath40 , in the dilepton rest frame , and the line of flight of the dilepton as defined in the kaon rest frame .
@xmath39 is the angle between the normals to the planes defined in the kaon rest frame by the pion pair and the lepton pair , respectively .
let us denote by @xmath41 and @xmath42 the next - to - leading order corrections to the @xmath15 and @xmath16 form factors , respectively , @xmath43 the analytic expressions for @xmath41 and @xmath42 were given in @xcite .
we shall distinguish between _ photonic _ and _ non photonic _ contributions to @xmath44 and @xmath42 .
the photonic contribution comes from those feynman diagrams with a virtual photon exchanged between two meson legs or one meson leg and a pure strong vertex . obviously , this contribution is proportional to @xmath45 , where @xmath12 is the electric charge , and depends in general on the five independent kinematical variables , @xmath36 , @xmath37 , @xmath25 , @xmath38 , and @xmath39 through lorentz invariants like @xmath46 , say .
the non photonic contribution comes from diagrams having similar topology as the ones in the pure strong theory with isospin breaking allowed in propagators and vertices .
this contribution generates isospin breaking terms proportional to the rate of @xmath47 to @xmath48 breaking , @xmath49 and to mass square difference between charged and neutral mesons , @xmath50 or equivalently , @xmath51 , @xmath52 , and @xmath53 .
the kinematical dependence is on three lorentz invariants , @xmath54 , @xmath55 , and @xmath56 which represent respectively the dipion mass square , the exchange energy between the kaon and the neutral pion , and that between the kaon and the charged pion . in terms of independent kinematical variables ,
the preceding scalars are functions of @xmath36 , @xmath37 , and @xmath57 .
it has been noted in ref .
@xcite that for @xmath58 the photonic contribution neither depends on @xmath38 nor on @xmath39 and , consequently , it can be written as @xmath59 where @xmath60 and @xmath61 are analytic functions of @xmath36 .
note that , to the order we are working , that is , to leading order in isospin breaking , the power counting scheme we use dictates the following on - shell conditions to be used in the argument of @xmath60 and @xmath61 , @xmath62 with respect to the nonphotonic contribution , it depends on @xmath36 , @xmath55 , @xmath56 and masses through one- and two - point functions . in order to split strong and electromagnetic interactions in one - point functions we use the formula @xmath63 \delta_p\,,\ ] ] where @xmath64 denotes a pion , @xmath65 , or a kaon , @xmath66 , and @xmath67 the difference , @xmath68 concerning the splitting in two - point functions @xmath69
, we have to expand exchange energies in powers of the fine structure constant @xmath70 and @xmath53 .
we then inject the obtained expansion in the expression of @xmath71 where @xmath72 , @xmath73 , and @xmath74 , are leading order in isospin breaking , @xmath75 the final step consists on expanding two - point functions to first order in @xmath72 , @xmath73 , and @xmath74 , @xmath76 \delta_0 \nonumber \\ & & + \dfrac{1}{32\pi^2p_1 ^ 2}\left [ \ln\left ( \dfrac{m_0 ^ 2}{m_1 ^ 2}\right ) -(p_1 ^ 2-m_1 ^ 2+m_0 ^ 2)\tau ( p_1 ^ 2,m_0 ^ 2,m_1 ^ 2)\right ] \delta_1 \nonumber \\ & & -\dfrac{1}{32\pi^2p_1
^ 4}\left\lbrace 2p_1 ^ 2+(m_1 ^ 2-m_0 ^ 2)\ln\left ( \dfrac{m_0 ^ 2}{m_1 ^ 2}\right ) \right .
\nonumber \\ & & + \left [ ( m_1 ^ 2-m_0 ^ 2)^2-p_1 ^ 2(m_1 ^ 2+m_0 ^ 2)\right ] \tau ( p_1 ^ 2,m_0 ^ 2,m_1 ^ 2)\,\bigg\}\,\delta\,,\end{aligned}\ ] ] with @xmath77 a generic integral defined by , @xmath78 putting all this together , form factors for @xmath2 decay of the charged kaon can be written in the following compact form which shows explicitly the splitting between strong and electromagnetic interactions , @xmath79 \ , , \;\ ; x\,=\,f\,,\ ; g\,,\end{aligned}\ ] ] where , @xmath80 are analytic functions of @xmath36 .
if one makes the following substitutions , @xmath81 then , equations ( [ eq : principal_one ] ) and ( [ eq : principal_two ] ) read , @xmath82 the aim of the present work is to determine the @xmath83 functions corresponding to @xmath0 and @xmath1 form factors for @xmath2 decay of the charged kaon .
we have @xmath84 \nonumber\\ & & + \frac{2}{f_0 ^ 2}\left [ 16(s_{\pi}-2m_{\pi^{\pm}}^2)l_1 + 4(m_{k^{\pm}}^2-m_{\ell}^2+s_{\pi})l_2 \right . \nonumber\\ & & \left .
+ ( m_{k^{\pm}}^2 - 8m_{\pi^{\pm}}^2-m_{\ell}^2 + 5s_{\pi})l_3 - 2(2m_{k^{\pm}}^2 - 7m_{\pi^{\pm}}^2)l_4+m_{\ell}^2l_9\right ] \nonumber\\ & & -\frac{1}{8f_0 ^ 2}\left [ 5-\frac{2}{t_{\pi}}\left ( m_{k^{\pm}}^2 - 2m_{\pi^{\pm}}^2\right)+\frac{2}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right)^2\right ] a(m_{\pi^{\pm}}^2 ) \nonumber\\ & & + \frac{1}{8f_0 ^ 2}\left [ 2-\frac{1}{t_{\pi}}\left ( 8m_{k^{\pm}}^2 - 5m_{\pi^{\pm}}^2 - 3m_{\eta}^2\right ) \right . \nonumber\\ & & \left .
-\frac{2}{t_{\pi}^2}\left ( 2m_{k^{\pm}}^2-m_{\pi^{\pm}}^2-m_{\eta}^2\right ) ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2)\right ] a(m_{k^{\pm}}^2 ) \nonumber\\ & & + \frac{1}{8f_0 ^ 2}\left [ 3+\frac{1}{t_{\pi}}\left ( 2m_{k^{\pm}}^2 + 3m_{\pi^{\pm}}^2 - 3m_{\eta}^2\right ) \right .
\nonumber\\ & & \left .
-\frac{2}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right ) ( m_{\eta}^2-m_{k^{\pm}}^2)\right ] a(m_{\eta}^2 ) \nonumber\\ & & + \frac{1}{12f_0 ^ 2}\left [ 6(2s_{\pi}-m_{\pi^{\pm}}^2)b(s_{\pi},m_{\pi^{\pm}}^2,m_{\pi^{\pm}}^2 ) \right . \nonumber\\ & & \left .
+ 6m_{\pi^{\pm}}^2b(s_{\pi},m_{\eta}^2,m_{\eta}^2)+9s_{\pi}b(s_{\pi},m_{k^{\pm}}^2,m_{k^{\pm}}^2)\right ] \nonumber\\ & & + \frac{1}{4f_0 ^ 2}\left [ 3m_{k^{\pm}}^2+m_{\pi^{\pm}}^2 - 4t_{\pi } \right .
\nonumber\\ & & \left .
+ \frac{2}{t_{\pi}}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right)^2+\frac{1}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right)^3\right ] b(t_{\pi},m_{\pi^{\pm}}^2,m_{k^{\pm}}^2 ) \nonumber\\ & & -\frac{1}{8f_0 ^ 2}\left [ 2m_{k^{\pm}}^2-m_{\pi^{\pm}}^2 + 3m_{\eta}^2-\frac{1}{t_{\pi}}\left ( 4m_{k^{\pm}}^2+m_{\pi^{\pm}}^2 - 5m_{\eta}^2\right ) m_{k^{\pm}}^2 \right .
\nonumber\\ & & \left .
+ \frac{3}{t_{\pi}}\left ( m_{\pi^{\pm}}^2-m_{\eta}^2\right ) m_{\eta}^2-\frac{2}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right ) ( m_{\eta}^2-m_{k^{\pm}}^2)^2\right ] b(t_{\pi},m_{\eta}^2,m_{k^{\pm}}^2)\ , , \\
u_s^g & = & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 12m_{k^{\pm}}^2 + 21m_{\pi^{\pm}}^2 + 3m_{\eta}^2 \right . \nonumber\\ & & \left .
-4s_{\pi}-2t_{\pi}-\frac{3}{t_{\pi}}\left ( 2m_{k^{\pm}}^2+m_{\pi^{\pm}}^2+m_{\eta}^2\right ) ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2)\right ] \nonumber\\ & & -\frac{2}{f_0 ^ 2}\left [ ( m_{k^{\pm}}^2-m_{\ell}^2+s_{\pi})l_3 + 2(m_{\pi^{\pm}}^2 + 2m_{k^{\pm}}^2)l_4-m_{\ell}^2l_9\right ] \nonumber\\ & & -\frac{1}{24f_0 ^ 2}\left [ 5-\frac{6}{t_{\pi}}\,m_{\pi^{\pm}}^2-\frac{6}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right)^2\right ] a(m_{\pi^{\pm}}^2 ) \nonumber\\ & & + \frac{1}{24f_0 ^ 2}\left [ 2+\frac{3}{t_{\pi}}\left ( 4m_{k^{\pm}}^2 - 3m_{\pi^{\pm}}^2-m_{\eta}^2\right ) \right .
\nonumber\\ & & \left .
+ \frac{6}{t_{\pi}^2}\left ( 2m_{k^{\pm}}^2-m_{\pi^{\pm}}^2-m_{\eta}^2\right ) ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2)\right ] a(m_{k^{\pm}}^2 ) \nonumber\\ & & + \frac{1}{8f_0 ^ 2}\left [ 1+\frac{1}{t_{\pi}}\left ( m_{\eta}^2 - 3m_{\pi^{\pm}}^2\right ) + \frac{2}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right ) ( m_{\eta}^2-m_{k^{\pm}}^2)\right ] a(m_{\eta}^2 ) \nonumber\\ & & + \frac{1}{12f_0 ^ 2}\left [ 2(s_{\pi}-4m_{\pi^{\pm}}^2)b(s_{\pi},m_{\pi^{\pm}}^2,m_{\pi^{\pm}}^2)+(s_{\pi}-4m_{k^{\pm}}^2)b(s_{\pi},m_{k^{\pm}}^2,m_{k^{\pm}}^2)\right ] \nonumber\\ & & -\frac{1}{4f_0 ^ 2}\left [ m_{k^{\pm}}^2-m_{\pi^{\pm}}^2-t_{\pi } \right .
\nonumber\\ & & \left .
+ \frac{1}{t_{\pi}}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right)^2+\frac{1}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right)^3\right ] b(t_{\pi},m_{\pi^{\pm}}^2,m_{k^{\pm}}^2 ) \nonumber\\ & & -\frac{1}{8f_0 ^ 2}\left [ 2m_{k^{\pm}}^2+m_{\pi^{\pm}}^2+m_{\eta}^2 - 2t_{\pi } \right .
\nonumber\\ & & + \frac{1}{t_{\pi}}\left ( 2m_{k^{\pm}}^2+m_{\pi^{\pm}}^2 - 3m_{\eta}^2\right ) m_{k^{\pm}}^2+\frac{1}{t_{\pi}}\left ( 2m_{k^{\pm}}^2 - 3m_{\pi^{\pm}}^2+m_{\eta}^2\right ) m_{\eta}^2 \nonumber\\ & & \left .
+ \frac{2}{t_{\pi}^2}\left ( m_{\pi^{\pm}}^2-m_{k^{\pm}}^2\right ) ( m_{\eta}^2-m_{k^{\pm}}^2)^2\right ] b(t_{\pi},m_{\eta}^2,m_{k^{\pm}}^2)\,.\end{aligned}\ ] ] in the preceding expressions , we used the notation @xmath85 note that in the isospin breaking correction , the same expression holds for @xmath86 with the replacement @xmath87 .
the correction due to @xmath88 reads : @xmath89\frac{1}{16\pi^2 } \right . \nonumber\\ & & + \left [ -3+\frac{2}{t_{\pi}}\,m_{\pi}^2+\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ] a(m_{\pi}^2 ) -\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_{\eta}^2\right ) ( m_{\pi}^2-m_k^2)a(m_k^2 ) \nonumber\\ & & + \left [ 3-\frac{2}{t_{\pi}}\,m_{\pi}^2-\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)\right ] a(m_{\eta}^2)+12(s_{\pi}-m_{\pi}^2)b(s_{\pi},m_{\pi}^2,m_{\pi}^2 ) \nonumber\\ & & -4m_{\pi}^2b(s_{\pi},m_{\eta}^2,m_{\eta}^2)-4(3s_{\pi}-4m_{\pi}^2)b(s_{\pi},m_{\pi}^2,m_{\eta}^2 ) \nonumber\\ & & + 2\left [ m_k^2 + 5m_{\pi}^2 - 3t_{\pi}-\frac{1}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right)^2-\frac{1}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^3\right ] b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \left [ -2(5m_k^2+m_{\pi}^2 - 3t_{\pi})+\frac{1}{t_{\pi}}\left ( 3m_{\eta}^2 - 5m_k^2\right ) ( m_{\pi}^2-m_k^2 ) \right .
\nonumber\\ & & \left.\left .
+ \frac{1}{t_{\pi}}\left ( 3m_{\eta}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)+\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)^2\right ] b(t_{\pi},m_{\eta}^2,m_k^2)\right\}\ , , \\
u_{\epsilon}^g & = & -\frac{1}{4f_0 ^ 2}\left\ { ( m_{\pi}^2-m_{\eta}^2)\left [ -1+\frac{1}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right)\right]\frac{1}{16\pi^2 } \right . \nonumber\\ & & + \left [ 1+\frac{2}{t_{\pi}}\,m_k^2+\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ] a(m_{\pi}^2 ) \nonumber\\ & & -\frac{1}{t_{\pi}}\left [ 4m_k^2 - 3m_{\pi}^2-m_{\eta}^2+\frac{2}{t_{\pi}}\left ( m_{\pi}^2-m_{\eta}^2\right ) ( m_{\pi}^2-m_k^2)\right ] a(m_k^2 ) \nonumber\\ & & + \left [ -1+\frac{1}{t_{\pi}}\left ( 2m_k^2 - 3m_{\pi}^2-m_{\eta}^2\right ) -\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)\right ] a(m_{\eta}^2 ) \nonumber\\ & & -2\left [ m_k^2 - 3m_{\pi}^2 + 2t_{\pi}+\frac{1}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^3\right ] b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \left [ -2(3m_k^2+m_{\pi}^2 - 2m_{\eta}^2 - 2t_{\pi})-\frac{1}{t_{\pi}}\left ( 4m_k^2 - 3m_{\eta}^2\right ) ( m_{\pi}^2-m_k^2 ) \right . \nonumber\\ & & \left.\left . + \frac{1}{t_{\pi}}\left ( 4m_k^2+m_{\eta}^2\right ) ( m_{\eta}^2-m_k^2)+\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)^2\right ] b(t_{\pi},m_{\eta}^2,m_k^2)\right\}\,.\end{aligned}\ ] ] that due to @xmath90 is equal to : @xmath91 -\frac{24}{f_0 ^ 2}\,l_4 \nonumber\\ & & + \frac{1}{24f_0 ^ 2}\left [ 9+\frac{2}{t_{\pi}}\left ( 13m_{\pi}^2-m_k^2\right ) -\frac{2}{t_{\pi}^2}\left ( m_k^2 - 2m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)\right]\frac{a(m_{\pi}^2)}{m_{\pi}^2 } \nonumber\\ & & -\frac{1}{24f_0 ^ 2}\,\frac{1}{t_{\pi}}\left [ 9+\frac{2}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right)\right ] a(m_k^2)-\frac{3}{8f_0 ^ 2}\,\frac{1}{t_{\pi}}\,a(m_{\eta}^2 ) \nonumber\\ & & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 6-\frac{1}{t_{\pi}}\left ( 5m_k^2+m_{\pi}^2\right ) \right . \nonumber\\ & & \left .
+ \frac{2}{t_{\pi}^2}\left ( m_k^2 - 2m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)-\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^3\right]\ln\left(\frac{m_{\pi}^2}{m_k^2}\right ) \nonumber\\ & & -\frac{1}{384\pi^2f_0 ^ 2}\left [ m_k^2 - 7m_{\pi}^2 + 6t_{\pi}+\frac{1}{t_{\pi}}\left ( 7m_k^2 - 3m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2 ) \right .
\nonumber\\ & & \left .
-\frac{1}{t_{\pi}^2}\left ( m_k^2 - 3m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)^2+\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^4\right]\tau ( t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{1}{32\pi^2f_0 ^ 2}\left ( s_{\pi}-m_{\pi}^2\right)\tau ( s_{\pi},m_{\pi}^2,m_{\pi}^2)+\frac{5}{2f_0 ^ 2}\,b(s_{\pi},m_{\pi}^2,m_{\pi}^2 ) \nonumber\\ & & -\frac{1}{2f_0 ^ 2}\left [ b(s_{\pi},m_{\eta}^2,m_{\eta}^2)-4b(s_{\pi},m_k^2,m_k^2)\right ] \nonumber\\ & & + \frac{1}{12f_0 ^ 2}\left [ 3+\frac{1}{t_{\pi}}\left ( 10m_k^2 - 13m_{\pi}^2\right ) -\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ]
b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & -\frac{3}{8f_0 ^ 2}\left [ 1-\frac{1}{t_{\pi}}\left ( m_{\eta}^2-m_k^2\right)\right ] b(t_{\pi},m_{\eta}^2,m_k^2)\ , , \\
u_{\pi}^g & = & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 10-\frac{1}{t_{\pi}}\left ( m_k^2 + 5m_{\pi}^2\right ) -\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ] + \frac{8}{f_0 ^ 2}\,l_4 \nonumber\\ & & + \frac{1}{24f_0 ^ 2}\left [ 11-\frac{22}{t_{\pi}}\,m_{\pi}^2+\frac{2}{t_{\pi}^2}\left ( m_k^2 - 2m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)\right]\frac{a(m_{\pi}^2)}{m_{\pi}^2 } \nonumber\\ & & + \frac{1}{24f_0 ^ 2}\,\frac{1}{t_{\pi}}\left [ 7+\frac{2}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right)\right ] a(m_k^2)+\frac{3}{8f_0 ^
2}\,\frac{1}{t_{\pi}}\,a(m_{\eta}^2 ) \nonumber\\ & & + \frac{1}{384\pi^2f_0 ^ 2}\left [ 7-\frac{1}{t_{\pi}}\left ( 7m_k^2 + 3m_{\pi}^2\right ) \right .
\nonumber\\ & & \left .
+ \frac{1}{t_{\pi}^2}\left ( m_k^2 - 3m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)-\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^3\right]\ln\left(\frac{m_{\pi}^2}{m_k^2}\right ) \nonumber\\ & & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 10m_{\pi}^2 - 7t_{\pi}-\frac{8}{t_{\pi}}\,m_k^2(m_{\pi}^2-m_k^2 ) \right .
\nonumber\\ & & \left .
-\frac{2}{t_{\pi}^2}\,m_{\pi}^2(m_{\pi}^2-m_k^2)^2-\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^4\right]\tau ( t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{1}{12f_0 ^ 2}\left [ 23+\frac{1}{t_{\pi}}\left ( 11m_{\pi}^2 - 8m_k^2\right ) + \frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ]
b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{3}{8f_0 ^ 2}\left [ 1-\frac{1}{t_{\pi}}\left ( m_{\eta}^2-m_k^2\right)\right ] b(t_{\pi},m_{\eta}^2,m_k^2)\,.\end{aligned}\ ] ] the one due to @xmath92 is given by : @xmath93 \nonumber\\ & & + \frac{4}{f_0 ^ 2}\,l_4-\frac{1}{12f_0 ^ 2}\,\frac{1}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right ) a(m_{\pi}^2 ) \nonumber\\ & & -\frac{1}{24f_0 ^ 2}\left [ 6-\frac{1}{t_{\pi}}\left ( 22 m_k^2 - 7m_{\pi}^2 - 9m_{\eta}^2\right ) \right .
\nonumber\\ & & \left .
-\frac{2}{t_{\pi}^2}\left ( 8m_k^2-m_{\pi}^2 - 3m_{\eta}^2\right ) ( m_{\pi}^2-m_k^2)\right]\frac{a(m_k^2)}{m_k^2 } \nonumber\\ & & -\frac{1}{4f_0 ^ 2}\,\frac{1}{t_{\pi}}\left [
1+\frac{1}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right)\right ] a(m_{\eta}^2 ) \nonumber\\ & & + \frac{1}{384\pi^2f_0 ^ 2}\left [ 6-\frac{1}{t_{\pi}}\left ( 5m_k^2+m_{\pi}^2\right ) \right .
\nonumber\\ & & \left .
+ \frac{2}{t_{\pi}^2}\left ( m_k^2 - 2m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)-\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^3\right]\ln\left(\frac{m_{\pi}^2}{m_k^2}\right ) \nonumber\\ & & + \frac{1}{256\pi^2f_0 ^ 2}\,\frac{1}{t_{\pi}}\left [ 2m_k^2-m_{\pi}^2 + 3m_{\eta}^2+\frac{1}{t_{\pi}}\left ( 5m_{\eta}^2-m_{\pi}^2 - 4m_k^2\right ) m_k^2 \right .
\nonumber\\ & & \left .
+ \frac{3}{t_{\pi}}\left ( m_{\pi}^2-m_{\eta}^2\right ) m_{\eta}^2-\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)^2\right]\ln\left(\frac{m_{\eta}^2}{m_k^2}\right ) \nonumber\\ & & + \frac{1}{64\pi^2f_0 ^ 2}\,s_{\pi}\tau ( s_{\pi},m_k^2,m_k^2 ) \nonumber\\ & & + \frac{1}{384\pi^2f_0 ^ 2}\left [ 11m_k^2 - 5m_{\pi}^2 - 6t_{\pi}+\frac{1}{t_{\pi}}\left ( 3m_k^2 + 5m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2 ) \right .
\nonumber\\ & & \left .
-\frac{1}{t_{\pi}^2}\left ( 3m_k^2 - 5m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)^2+\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^4\right]\tau ( t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & -\frac{1}{256\pi^2f_0 ^ 2}\left [ 2m_k^2-m_{\pi}^2 + 3m_{\eta}^2-\frac{6}{t_{\pi}}\left ( m_{\eta}^2-m_k^2\right)^2-\frac{1}{t_{\pi}^2}\left ( 6m_k^2-m_{\pi}^2 - 5m_{\eta}^2\right ) m_k^2 \right .
\nonumber\\ & & \left .
+ \frac{1}{t_{\pi}^2}\left ( 2m_k^2+m_{\pi}^2 - 3m_{\eta}^2\right ) m_{\eta}^2-\frac{2}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)^3\right]\tau ( t_{\pi},m_{\eta}^2,m_k^2 ) \nonumber\\ & & -\frac{1}{12f_0 ^ 2}\left [ 3+\frac{1}{t_{\pi}}\left ( 2m_k^2 - 5m_{\pi}^2\right ) -\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ] b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{1}{8f_0 ^ 2}\left [ 4-\frac{1}{t_{\pi}}\left ( 6m_k^2+m_{\pi}^2 - 5m_{\eta}^2\right ) + \frac{4}{t_{\pi}^2}\left
( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)\right ] b(t_{\pi},m_{\eta}^2,m_k^2)\ , , \\
u_k^g & = & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 14-\frac{3}{t_{\pi}}\left ( 4m_k^2 - 3m_{\pi}^2-m_{\eta}^2\right)\right ] + \frac{4}{f_0 ^ 2}\,l_4 \nonumber\\ & & -\frac{1}{12f_0 ^ 2}\,\frac{1}{t_{\pi}}\left [ 1-\frac{1}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right)\right ] a(m_{\pi}^2)+\frac{1}{4f_0 ^ 2}\,\frac{1}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right )
a(m_{\eta}^2 ) \nonumber\\ & & + \frac{1}{24f_0 ^ 2}\left [ 10-\frac{1}{t_{\pi}}\left ( 6m_k^2 - 5m_{\pi}^2 - 3m_{\eta}^2\right ) \right .
\nonumber\\ & & \left .
-\frac{2}{t_{\pi}^2}\left ( 8m_k^2-m_{\pi}^2 - 3m_{\eta}^2\right ) ( m_{\pi}^2-m_k^2)\right]\frac{a(m_k^2)}{m_k^2 } \nonumber\\ & & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 7-\frac{1}{t_{\pi}}\left ( 7m_k^2 + 3m_{\pi}^2\right ) \right .
\nonumber\\ & & \left .
+ \frac{1}{t_{\pi}^2}\left ( m_k^2 - 3m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)-\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^3\right]\ln\left(\frac{m_{\pi}^2}{m_k^2}\right ) \nonumber\\ & & -\frac{1}{256\pi^2f_0 ^ 2}\left [ 2-\frac{1}{t_{\pi}}\left ( 2m_k^2+m_{\pi}^2+m_{\eta}^2\right ) -\frac{1}{t_{\pi}^2}\left ( 2m_k^2+m_{\pi}^2 - 3m_{\eta}^2\right ) m_k^2 \right .
\nonumber\\ & & \left .
-\frac{1}{t_{\pi}^2}\left ( 2m_k^2 - 3m_{\pi}^2+m_{\eta}^2\right ) m_{\eta}^2-\frac{2}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)^2\right]\ln\left(\frac{m_{\eta}^2}{m_k^2}\right ) \nonumber\\ & & -\frac{1}{192\pi^2f_0 ^ 2}\left ( s_{\pi}-4m_k^2\right)\tau ( s_{\pi},m_k^2,m_k^2 ) \nonumber\\ & & -\frac{1}{384\pi^2f_0 ^ 2}\left [ 14m_k^2 - 4m_{\pi}^2 - 7t_{\pi}+\frac{6}{t_{\pi}}\left ( m_{\pi}^2+m_k^2\right ) ( m_{\pi}^2-m_k^2 ) \right .
\nonumber\\ & & \left .
-\frac{2}{t_{\pi}^2}\left ( m_k^2 - 2m_{\pi}^2\right ) ( m_{\pi}^2-m_k^2)^2+\frac{1}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right)^4\right]\tau ( t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & -\frac{1}{256\pi^2f_0 ^ 2}\left [ 4m_k^2+m_{\pi}^2-m_{\eta}^2 - 2t_{\pi}-\frac{2}{t_{\pi}}\left ( m_{\pi}^2-m_{\eta}^2\right ) m_{\eta}^2 \right .
\nonumber\\ & & -\frac{1}{t_{\pi}^2}\left ( m_{\pi}^2-m_{\eta}^2\right ) ( m_{\eta}^2-m_k^2)m_{\eta}^2+\frac{1}{t_{\pi}^2}\left ( 4m_k^2-m_{\pi}^2 - 3m_{\eta}^2\right ) ( m_{\eta}^2-m_k^2)m_k^2 \nonumber\\ & & \left .
+ \frac{2}{t_{\pi}^3}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)^3\right]\tau ( t_{\pi},m_{\eta}^2,m_k^2)-\frac{1}{3f_0 ^ 2}\,b(s_{\pi},m_k^2,m_k^2 ) \nonumber\\ & & + \frac{1}{12f_0 ^ 2}\left [ 5-\frac{3}{t_{\pi}}\,m_{\pi}^2-\frac{2}{t_{\pi}^2}\left ( m_{\pi}^2-m_k^2\right)^2\right ] b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{1}{8f_0 ^ 2}\,\frac{1}{t_{\pi}}\left [ 2m_k^2+m_{\pi}^2-m_{\eta}^2-\frac{4}{t_{\pi}}\left ( m_{\pi}^2-m_k^2\right ) ( m_{\eta}^2-m_k^2)\right ] b(t_{\pi},m_{\eta}^2,m_k^2)\,.\end{aligned}\ ] ]
we have @xmath94 \nonumber\\ & & -\frac{1}{18}\left ( 24k_1 - 264k_2 - 16k_5 - 88k_6 - 36k_{12}+120x_1 + 9x_6\right ) \nonumber\\ & & -\frac{1}{2}\,\frac{a(m_{\ell}^2)}{m_{\ell}^2}+2\,\frac{a(m_{\pi}^2)}{m_{\pi}^2}+\frac{a(m_k^2)}{m_k^2}-\frac{1}{2t_{\pi}}\left [ a(m_{\pi}^2)-a(m_k^2)\right ] \nonumber\\ & & -\frac{m_{\ell}^2}{2}\left\{\frac{3}{t_{\pi}}-\frac{1}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -\frac{m_{\ell}^2}{t_{\pi}}\left ( m_k^2 + 5m_{\pi}^2 - 3m_{\ell}^2\right)\right.\right .
\nonumber\\ & & \left.\left .
+ \frac{m_{\pi}^2}{t_{\pi}}\left ( m_k^2 + 2m_{\pi}^2\right ) + m_k^2 - 12m_{\pi}^2 - 5m_{\ell}^2 + 2t_{\pi}\right]\right\ } b(0,m_{\ell}^2,m_k^2 ) \nonumber\\ & & + \frac{m_{\ell}^2}{2}\,\frac{m_k^2 - 2m_{\pi}^2+m_{\ell}^2 - 2t_{\pi}}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ 1+\frac{12m_{\ell}^2m_{\pi}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\right ]
b(m_{\ell}^2,0,m_k^2 ) \nonumber\\ & & + \left [ 2+\frac{m_{\ell}^2}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left ( m_k^2 - 3m_{\ell}^2 + 3s_{\pi}\right)\right .
\nonumber\\ & & \left .
+ \frac{3m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left ( m_{\pi}^2+m_{\ell}^2-t_{\pi}\right)\right ] b(m_{\ell}^2,0,m_{\ell}^2 ) \nonumber\\ & & + \left\ { 2-\frac{8m_{\pi}^2(m_{\pi}^2-t_{\pi})}{\lambda ( t_{\pi},m_{\pi}^2,m_k^2)}-\frac{12m_{\ell}^2m_{\pi}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\right .
\nonumber\\ & & + \frac{3m_{\ell}^2m_{\pi}^2}{\lambda^2(t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -m_{\pi}^2(m_k^2 - 4m_{\pi}^2)-m_{\ell}^2(m_k^2 + 3m_{\pi}^2-m_{\ell}^2 ) \right . \nonumber\\ & & \left.\left .
+ ( m_k^2 - 4m_{\pi}^2-m_{\ell}^2)t_{\pi}\right]\right\ } b(m_{\pi}^2,0,m_{\pi}^2 ) \nonumber\\ & & + \left [ 1-\frac{2m_k^2}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left ( m_k^2 - 2m_{\ell}^2-s_{\pi}\right)\right .
\nonumber\\ & & \left .
-\frac{4m_k^2}{\lambda ( t_{\pi},m_{\pi}^2,m_k^2)}\left ( m_k^2 - 3m_{\pi}^2-t_{\pi}\right)\right ] b(m_k^2,0,m_k^2 ) \nonumber\\ & & + \left\ { -2+\frac{1}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left [ 2m_k^2(m_k^2-s_{\pi } ) \right.\right .
\nonumber\\ & & \left.\left .
-m_{\ell}^2(5m_k^2 - 3m_{\ell}^2 + 3s_{\pi})\right]\right\ } b(s_{\pi},m_{\ell}^2,m_k^2 ) -b(s_{\pi},m_{\pi}^2,m_{\pi}^2 ) \nonumber\\ & & -\frac{1}{2}\left\ { 9+\frac{1}{t_{\pi}}\left ( m_k^2-m_{\pi}^2 - 3m_{\ell}^2\right ) \right . \nonumber\\ & & + \frac{8}{\lambda ( t_{\pi},m_{\pi}^2,m_k^2)}\left [ ( m_k^2 - 2m_{\pi}^2)(m_{\pi}^2-m_k^2)+(m_k^2 + 2m_{\pi}^2)t_{\pi}\right ] \nonumber\\ & & + \frac{m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ 2(m_k^2 - 13m_{\pi}^2 - 2m_{\ell}^2)\right .
\nonumber\\ & & \left .
-\frac{m_{\ell}^2}{t_{\pi}}\left ( m_k^2 + 5m_{\pi}^2 - 3m_{\ell}^2\right ) + \frac{m_{\pi}^2}{t_{\pi}}\left ( m_k^2 + 2m_{\pi}^2\right)\right ] \nonumber\\ & & + \frac{6m_{\ell}^2m_{\pi}^2}{\lambda^2(t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ m_{\ell}^2(m_k^2 - 7m_{\pi}^2 + 3m_{\ell}^2)\right . \nonumber\\ & & \left.\left .
-m_{\pi}^2(m_k^2 - 4m_{\pi}^2)+(m_k^2 - 4m_{\pi}^2 - 5m_{\ell}^2)t_{\pi}\right]\right\ } b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{3m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left ( m_{\pi}^2-m_{\ell}^2+t_{\pi}\right ) b(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) \nonumber\\ & & + 2(m_k^2+m_{\ell}^2-s_{\pi})c(m_{\ell}^2,s_{\pi},m_k^2,m_{\gamma}^2,m_{\ell}^2,m_k^2 ) \nonumber\\ & & + 2(2m_{\pi}^2-s_{\pi})c(m_{\pi}^2,s_{\pi},m_{\pi}^2,m_{\gamma}^2,m_{\pi}^2,m_{\pi}^2 ) \nonumber\\ & & + \frac{2m_{\ell}^2m_k^2s_{\pi}}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\,c(m_k^2,s_{\pi},m_{\ell}^2,0,m_k^2,m_k^2 ) \nonumber\\ & & + m_{\ell}^2\left\ { 1+\frac{1}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left [ m_{\ell}^2(2m_k^2-m_{\ell}^2+s_{\pi})-m_k^2(m_k^2-s_{\pi})\right]\right .
\nonumber\\ & & + \frac{1}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -m_{\ell}^2(m_k^2-m_{\pi}^2-m_{\ell}^2 ) \right .
\nonumber\\ & & \left.\left .
-m_{\pi}^2m_k^2+(m_k^2-m_{\ell}^2)t_{\pi}\right]\right\ } c(m_{\ell}^2,0,m_{\ell}^2,0,m_{\ell}^2,m_k^2 ) \nonumber\\ & & -m_{\ell}^2\left\ { 1-\frac{1}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left [ m_k^2(m_k^2 - 3s_{\pi } ) \right.\right .
\nonumber\\ & & \left.\left .
-m_{\ell}^2(2m_k^2-m_{\ell}^2+s_{\pi})\right]\right\ } c(s_{\pi},s_{\pi},0,m_{\ell}^2,m_k^2,m_k^2 ) \nonumber\\ & & + \frac{m_{\ell}^2}{2}\left\ { 1+\frac{3}{t_{\pi}}\left ( m_{\ell}^2-m_{\pi}^2\right ) + \frac{1}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -m_{\pi}^2(3m_k^2 + 8m_{\pi}^2 ) \right.\right . \nonumber\\ & & + m_{\ell}^2(m_k^2 + 7m_{\pi}^2 + 3m_{\ell}^2)+\frac{m_{\ell}^4}{t_{\pi}}\left ( m_k^2 + 8m_{\pi}^2 - 3m_{\ell}^2\right ) \nonumber\\ & & -2(m_k^2 - 3m_{\pi}^2)t_{\pi}-\frac{m_{\ell}^2m_{\pi}^2}{t_{\pi}}\left ( 2m_k^2 + 7m_{\pi}^2\right ) \nonumber\\ & & \left.\left .
+ \frac{m_{\pi}^4}{t_{\pi}}\left ( m_k^2 + 2m_{\pi}^2\right)\right]\right\ } c(t_{\pi},t_{\pi},0,m_{\ell}^2,m_{\pi}^2,m_k^2 ) \nonumber\\ & & -\frac{m_{\pi}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left\ { 2(m_k^2 - 3m_{\ell}^2)(m_{\ell}^2-m_k^2 ) \right .
\nonumber\\ & & -\frac{3m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ m_{\ell}^4(2m_k^2 + 3m_{\pi}^2-m_{\ell}^2)-2m_{\ell}^2m_{\pi}^2m_k^2 \right .
\nonumber\\ & & -m_k^4(m_{\pi}^2+m_{\ell}^2)+4m_{\pi}^4(m_k^2-m_{\ell}^2)-m_{\ell}^2(2m_k^2 - 4m_{\pi}^2-m_{\ell}^2)t_{\pi } \nonumber\\ & & \left.\left .
+ m_k^2(m_k^2 - 4m_{\pi}^2)t_{\pi}\right]\right\ } c(m_{\pi}^2,t_{\pi},m_{\ell}^2,0,m_{\pi}^2,m_k^2)\ , , \\
u_{e^2}^g & = & -\frac{1}{32\pi^2}\left [ 9 + 2\ln\left(\frac{m_{\gamma}^2}{m_{\ell}^2}\right ) + 4\ln\left(\frac{m_{\gamma}^2}{m_{\pi}^2}\right ) + 2\ln\left(\frac{m_{\gamma}^2}{m_k^2}\right)\right ] \nonumber\\ & & -\frac{1}{18}\left ( 24k_1 + 24k_2 - 144k_3 - 72k_4 + 32k_5 - 40k_6 - 36k_{12}-24x_1 + 9x_6\right ) \nonumber\\ & & -\frac{1}{2}\,\frac{a(m_{\ell}^2)}{m_{\ell}^2}+2\,\frac{a(m_{\pi}^2)}{m_{\pi}^2}+\frac{a(m_k^2)}{m_k^2}+\frac{1}{2t_{\pi}}\left [ a(m_{\pi}^2)-a(m_k^2)\right ] \nonumber\\ & & + \frac{1}{2}\left\ { 3+\frac{3m_{\ell}^2}{t_{\pi}}+\frac{m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ \frac{m_{\ell}^2}{t_{\pi}}\left ( m_k^2 + 5m_{\pi}^2 - 3m_{\ell}^2\right)\right.\right .
\nonumber\\ & & \left.\left .
-\frac{m_{\pi}^2}{t_{\pi}}\left ( m_k^2 + 2m_{\pi}^2\right ) -(m_k^2 - 12m_{\pi}^2 - 5m_{\ell}^2 + 2t_{\pi})\right]\right\ } b(0,m_{\ell}^2,m_k^2 ) \nonumber\\ & & -\frac{1}{2}\left\ { 3-\frac{m_{\ell}^2}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left ( m_k^2+m_{\ell}^2-s_{\pi}\right ) \right .
\nonumber\\ & & \left .
+ \frac{m_{\ell}^2(m_k^2 - 2m_{\pi}^2+m_{\ell}^2 - 2t_{\pi})}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ 1+\frac{12m_{\ell}^2m_{\pi}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\right]\right\ }
b(m_{\ell}^2,0,m_k^2 ) \nonumber\\ & & + \left [ 2+\frac{2m_{\ell}^2}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left ( m_k^2-m_{\ell}^2+s_{\pi}\right)\right .
\nonumber\\ & & \left .
+ \frac{3m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left ( m_{\pi}^2+m_{\ell}^2-t_{\pi}\right)\right ] b(m_{\ell}^2,0,m_{\ell}^2 ) \nonumber\\ & & + \left\ { -4 + 2\left(\frac{8m_{\pi}^2 - 3s_{\pi}}{4m_{\pi}^2-s_{\pi}}\right)+\frac{8m_{\pi}^2m_k^2}{\lambda ( t_{\pi},m_{\pi}^2,m_k^2 ) } \right .
\nonumber\\ & & + \frac{3m_{\ell}^2m_{\pi}^2}{\lambda^2(t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left[m_{\pi}^2(m_k^2 - 4m_{\pi}^2)+m_{\ell}^2(m_k^2 + 3m_{\pi}^2-m_{\ell}^2)\right .
\nonumber\\ & & \left.\left .
-(m_k^2 - 4m_{\pi}^2-m_{\ell}^2)t_{\pi}\right]\right\ } b(m_{\pi}^2,0,m_{\pi}^2 ) \nonumber\\ & & + \left [ 1-\frac{m_k^2}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left ( 2m_k^2-m_{\ell}^2 - 2s_{\pi}\right)\right .
\nonumber\\ & & \left .
-\frac{4m_k^2}{\lambda ( t_{\pi},m_{\pi}^2,m_k^2)}\left ( m_k^2+m_{\pi}^2-t_{\pi}\right)\right ] b(m_k^2,0,m_k^2 ) \nonumber\\ & & -2\left\ { 1+\frac{1}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\left [ m_{\ell}^2(2m_k^2-m_{\ell}^2)-m_k^4+(m_k^2+m_{\ell}^2)s_{\pi}\right]\right\ } b(s_{\pi},m_{\ell}^2,m_k^2 ) \nonumber\\ & & -4\left(\frac{2m_{\pi}^2-s_{\pi}}{4m_{\pi}^2-s_{\pi}}\right ) b(s_{\pi},m_{\pi}^2,m_{\pi}^2 ) + \frac{m_{\ell}^2(m_k^2-m_{\ell}^2+s_{\pi})}{2\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\,b(s_{\pi},m_k^2,m_k^2 ) \nonumber\\ & & -\frac{1}{2}\left\ { 3-\frac{1}{t_{\pi}}\left ( m_k^2-m_{\pi}^2 - 3m_{\ell}^2\right ) -\frac{8m_k^2}{\lambda ( t_{\pi},m_{\pi}^2,m_k^2)}\left ( m_k^2-m_{\pi}^2-t_{\pi}\right)\right .
\nonumber\\ & & + \frac{m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -2(m_k^2 - 13m_{\pi}^2 - 2m_{\ell}^2)\right .
\nonumber\\ & & \left .
+ \frac{m_{\ell}^2}{t_{\pi}}\left ( m_k^2 + 5m_{\pi}^2 - 3m_{\ell}^2\right ) -\frac{m_{\pi}^2}{t_{\pi}}\left ( m_k^2 + 2m_{\pi}^2\right)\right ] \nonumber\\ & & -\frac{2m_{\ell}^2m_{\pi}^2}{\lambda^2(t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ 3m_{\ell}^2(m_k^2 - 7m_{\pi}^2 + 3m_{\ell}^2)\right .
\nonumber\\ & & \left.\left .
-3m_{\pi}^2(m_k^2 - 4m_{\pi}^2)+3(m_k^2 - 4m_{\pi}^2 - 5m_{\ell}^2)t_{\pi}\right]\right\ } b(t_{\pi},m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{3m_{\ell}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left ( m_{\pi}^2-m_{\ell}^2+t_{\pi}\right )
b(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) \nonumber\\ & & + 2(m_k^2+m_{\ell}^2-s_{\pi})c(m_{\ell}^2,s_{\pi},m_k^2,m_{\gamma}^2,m_{\ell}^2,m_k^2 ) \nonumber\\ & & + 2(2m_{\pi}^2-s_{\pi})c(m_{\pi}^2,s_{\pi},m_{\pi}^2,m_{\gamma}^2,m_{\pi}^2,m_{\pi}^2 ) \nonumber\\ & & -\frac{m_{\ell}^2m_k^2}{\lambda ( s_{\pi},m_{\ell}^2,m_k^2)}\,(m_k^2-m_{\ell}^2)c(m_k^2,s_{\pi},m_{\ell}^2,0,m_k^2,m_k^2 ) \nonumber\\ & & -m_{\ell}^2\left\ { 2+\frac{1}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ m_{\ell}^2(m_k^2 + 7m_{\pi}^2-m_{\ell}^2 ) \right.\right .
\nonumber\\ & & \left.\left .
+ m_{\pi}^2m_k^2-(m_k^2-m_{\ell}^2)t_{\pi}\right]\right\ } c(m_{\ell}^2,0,m_{\ell}^2,0,m_{\ell}^2,m_k^2 ) \nonumber\\ & & -\frac{m_{\ell}^2}{2}\,c(s_{\pi},s_{\pi},0,m_{\ell}^2,m_k^2,m_k^2 ) \nonumber\\ & & + \frac{m_{\ell}^2}{2}\left\ { 3-\frac{3}{t_{\pi}}\left ( m_{\ell}^2-m_{\pi}^2\right ) + \frac{1}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -m_{\pi}^2m_k^2\right.\right .
\nonumber\\ & & + m_{\ell}^2(3m_k^2 + 5m_{\pi}^2 - 7m_{\ell}^2)-\frac{m_{\ell}^4}{t_{\pi}}\left ( m_k^2 + 8m_{\pi}^2 - 3m_{\ell}^2\right ) \nonumber\\ & & -2(m_k^2-m_{\pi}^2 - 2m_{\ell}^2)t_{\pi}+\frac{m_{\ell}^2m_{\pi}^2}{t_{\pi}}\left ( 2m_k^2 + 7m_{\pi}^2\right ) \nonumber\\ & & \left.\left .
-\frac{m_{\pi}^4}{t_{\pi}}\left ( m_k^2 + 2m_{\pi}^2\right)\right]\right\ } c(t_{\pi},t_{\pi},0,m_{\ell}^2,m_{\pi}^2,m_k^2 ) \nonumber\\ & & + \frac{m_{\ell}^2m_{\pi}^2}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left\ { 4(2m_k^2+m_{\pi}^2-m_{\ell}^2-t_{\pi } ) \right .
\nonumber\\ & & + \frac{3}{\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}\left [ -m_{\ell}^4(2m_k^2 + 3m_{\pi}^2-m_{\ell}^2 ) \right .
\nonumber\\ & & + 2m_{\ell}^2m_{\pi}^2(m_k^2 + 2m_{\pi}^2)+m_k^4(m_{\pi}^2+m_{\ell}^2 ) \nonumber\\ & & -4m_{\pi}^4m_k^2+m_{\ell}^2(2m_k^2 - 4m_{\pi}^2-m_{\ell}^2)t_{\pi } \nonumber\\ & & \left.\left .
-m_k^2(m_k^2 - 4m_{\pi}^2)t_{\pi}\right]\right\ } c(m_{\pi}^2,t_{\pi},m_{\ell}^2,0,m_{\pi}^2,m_k^2)\,.\end{aligned}\ ] ]
we shall proceed to the numerical evaluation of isospin breaking corrections . to this end
, we must handle all types of singularities encountered in our expressions .
these are of three types in general : ultraviolet , infrared , and coulomb . although our expressions are ultraviolet finite , they are infrared divergent .
we showed in ref .
@xcite that the latter singularity is canceled by the emission of a real soft photon at the level of differential decay rate .
since we are interested in measuring form factors , a subtraction of infrared divergence at this level is needed .
there are infinitely many choices to do so .
we shall choose the simplest _ minimal subtraction scheme _ consisting on dropping out , from the expression of form factors , @xmath95 terms only .
finally , coulomb interaction between charged particles induces singularities due to a photon exchange between : 1 .
the kaon and a pion .
this occurs at @xmath96 or @xmath97 hence , the singularity is situated outside the _ allowed kinematical region _
, @xmath98 from the left .
2 . the two pions .
this occurs at @xmath99 the former value represents a pseudo - threshold and is situated outside the allowed kinematical region from the left .
the latter value is a normal threshold and is situated at the lower bound of the allowed kinematical region .
the corresponding singularity is of great experimental importance for the present work and we will study it further in the following .
3 . the kaon and the lepton . this occurs at @xmath100 the pseudo - threshold is situated at the upper bound of the allowed kinematical region . the normal threshold outside the latter from the right .
4 . a pion and the lepton . this occurs at @xmath101 or @xmath102 hence , the singularity is situated outside the allowed kinematical region from the right .
let us return to the coulomb interaction between the two pions and shift the value of @xmath36 from @xmath103 by an infinitesimal positive amount @xmath104 we then expand our expressions in powers of @xmath105 .
the coulomb singularity shows up then as poles in the @xmath105-plane . in order to obtain finite ( regularized ) results ,
we simply remove these poles allowing the numerical evaluation of form factors .
we shall use the following numerical values equal to the rho mass . ] for the various parameters @xcite : 1 .
the fine structure constant , @xmath106 corresponding to the classical electron charge @xmath107 ; 2 .
the masses of the charged leptons , @xmath108 3 .
the masses of the light mesons , @xmath109 4 .
the quark masses and condensates , @xmath110 5 .
the low - energy constants in the strong sector , @xmath111 6 .
the low - energy constants in the electromagnetic sector , @xmath112 with an error of @xmath113 assigned to each of them ; 7 .
the low - energy constants in the leptonic sector , @xmath114 8 .
the coupling of axial currents to the vacuum , @xmath115 9 .
the charged pion decay constant and electromagnetic mass , @xmath116 using the preceding input parameters , we drew ( fig . [ fig:1 ] ) the curve of the variation of one - loop level correction to the real part for @xmath0 form factor as function of @xmath36 . in fig .
[ fig:2 ] , we drew the isospin breaking correction to the same quantity and compared the contributions of orders @xmath117 and @xmath118 .
radiative correction to the real part of the first term in the partial wave expansion for @xmath0 form factor under the assumptions @xmath119 .
the plain curve represents the one - loop correction in the absence of isospin breaking .
the dashed curve gives the isospin breaking correction of order @xmath120 .
the infrared divergence has been removed applying a minimal subtraction scheme .
error bands come exclusively from the uncertainty in the determination of low - energy constants and have been developed in quadrature . ]
isospin breaking correction to the real part of the first term in the partial wave expansion for @xmath0 form factor under the assumptions @xmath119 .
the infrared divergence has been removed applying a minimal subtraction scheme .
error bands come exclusively from the uncertainty in the determination of low - energy constants and have been developed in quadrature . ]
the same has been done for the real part of @xmath1 form factor in figs .
[ fig:3 ] and [ fig:4 ] .
radiative correction to the real part of the first term in the partial wave expansion for @xmath1 form factor under the assumptions @xmath119 .
the plain curve represents the one - loop correction in the absence of isospin breaking .
the dashed curve gives the isospin breaking correction of order @xmath120 .
the infrared divergence has been removed applying a minimal subtraction scheme .
error bands come exclusively from the uncertainty in the determination of low - energy constants and have been developed in quadrature . ]
isospin breaking correction to the real part of the first term in the partial wave expansion for @xmath1 form factor under the assumptions @xmath119 .
the infrared divergence has been removed applying a minimal subtraction scheme .
error bands come exclusively from the uncertainty in the determination of low - energy constants and have been developed in quadrature . ] after removing infrared singularity and coulomb poles , the na48 experiment should measure what we will call _ subtracted form factors_. the corresponding _ modules _ are found to be @xmath121\ln ( m_{\gamma}^2 ) \nonumber\\ & & + \frac{e^2}{8\pi^2}\,\frac{m_k^2+m_{\ell}^2-s_{\pi}}{\sqrt{(m_k+m_{\ell})^2-s_{\pi}}\sqrt{(m_k - m_{\ell})^2-s_{\pi}}}\,\ln ( \sigma_{\ell k})\ln ( m_{\gamma}^2)\,,\end{aligned}\ ] ] where @xmath122 is a logical function equal to @xmath123 if @xmath124 is true and to @xmath125 if @xmath124 is false . the @xmath126-wave iso - scalar and @xmath64-wave iso - vector @xmath4 phase shifts are given by @xmath127 respectively .
the imaginary part of the first term in the partial wave expansion of form factors reads : @xmath128\right\ } \nonumber\\ & & + \frac{3\alpha}{4}\left ( 1+\frac{m_{\pi}^2-m_{\ell}^2}{t_{\pi}}\right)\frac{m_{\ell}^2}{\sqrt{t_{\pi}-(m_{\ell}+m_{\pi})^2}\sqrt{t_{\pi}-(m_{\ell}-m_{\pi})^2}}\ , , \\ \im\,u^g(s_{\pi } ) & = & \delta_1 ^ 1(s_{\pi } ) \nonumber\\ & & -\frac{\alpha}{\sigma_{\pi}}\left ( 1-\frac{2m_{\pi}^2}{s_{\pi}}\right)\left [ 1-\ln ( \sigma_{\pi})+\frac{1}{2}\,\ln\left(\frac{m_{\gamma}^2}{s_{\pi}}\right)\right ] \nonumber\\ & & + \frac{3\alpha}{4}\left ( 1+\frac{m_{\pi}^2-m_{\ell}^2}{t_{\pi}}\right)\frac{m_{\ell}^2}{\sqrt{t_{\pi}-(m_{\ell}+m_{\pi})^2}\sqrt{t_{\pi}-(m_{\ell}-m_{\pi})^2}}\,.\end{aligned}\ ] ] if the infrared divergence is removed using a minimal subtraction scheme the imaginary part for @xmath0 form factor takes the shape of fig .
[ fig:5 ] . in fig .
[ fig:6 ] , we compared the size of each contribution to the isospin breaking part of the same quantity . finally , the imaginary part for @xmath1 form factor is sketched in fig .
[ fig:7 ] .
note that the isospin breaking part is purely of order @xmath117 . the imaginary part ( in radians ) of the first term in the partial wave expansion for @xmath0 form factor under the assumptions @xmath129 .
the infrared divergence has been removed applying a minimal subtraction scheme .
the plain curve represents @xmath130 .
the dashed one includes isospin breaking effects . ]
isospin breaking correction to the imaginary part ( in radians ) of the first term in the partial wave expansion for @xmath0 form factor under the assumptions @xmath131 .
the infrared divergence has been removed applying a minimal subtraction scheme . ]
the imaginary part ( in radians ) of the first term in the partial wave expansion for @xmath1 form factor under the assumptions @xmath129 .
the infrared divergence has been removed applying a minimal subtraction scheme .
the plain curve represents @xmath132 .
the dashed one includes isospin breaking effects . ] for experimental purposes , we define the _ subtracted phase shifts _ as : @xmath133\textrm{if}\left ( s_{\pi}=4m_{\pi}^2\right ) \nonumber\\ & & + \frac{\alpha}{2}\left ( 1-\frac{2m_{\pi}^2}{s_{\pi}}\right)\frac{1}{\sigma_{\pi}}\,\ln ( m_{\gamma}^2)\ , , \\ \delta_p(s_{\pi } ) & \doteq & \im\,u^g(s_{\pi } ) \nonumber\\ & & + \frac{\alpha}{2}\,\frac{m_{\pi}}{\varrho}\left [ 1-\ln ( 2\varrho ) \right]\textrm{if}\left ( s_{\pi}=4m_{\pi}^2\right ) \nonumber\\ & & + \frac{\alpha}{2}\left ( 1-\frac{2m_{\pi}^2}{s_{\pi}}\right)\frac{1}{\sigma_{\pi}}\,\ln ( m_{\gamma}^2)\,.\end{aligned}\ ] ]
in this work we made the splitting between strong and electromagnetic interactions in @xmath2 decay of the charged kaon , @xmath3 .
our expressions were evaluated at the production threshold for the lepton pair , @xmath134 .
thanks to this assumption , a partial wave expansion of form factors with exactly the same structure as in the pure strong theory was possible .
the imaginary part of such an expansion involves the @xmath126-wave isoscalar and @xmath64-wave isovector @xmath4 phase shifts , @xmath130 and @xmath132 , respectively .
these can be related to @xmath4 scattering lengths via roy equations . in their turn ,
scattering lengths are sensitive to the way chiral symmetry is spontaneously broken .
consequently , a theoretical study of the process in question including all possible contributions is imperative .
we gave here the first analytic and numerical evaluation of the isospin breaking contribution .
this would allow the extraction of @xmath130 and @xmath132 from the experimental measurement of form factors .
our results can be summarized as follows : * isospin breaking affects modules of form factors only by the effect of coulomb interaction between charged particles .
the one between the two pions is of great importance and induces a singularity at @xmath135 .
we gave the residue of the pole in the present work .
* the effect of isospin breaking on the imaginary parts of form factors is considerable if the infrared divergence is removed using a minimal subtraction scheme .
we gave here all analytical expressions for the imaginary part including the finite part , the infrared divergent part , the singular part with the residue of the pole .
our results are of great utility for the interpretation of the outgoing data from the upgraded na48 experiment at cern .
we use dimensional regularization and adopt the @xmath136 subtraction scheme @xmath137\,,\ ] ] where @xmath138 is space - time dimension and @xmath139 the euler constant . all the technical material necessary for the calculation of one - loop integrals
is given in the appendix of ref .
@xcite .
we need the following two - point functions : @xmath144 \ , , \\
\re\,b(s_{\pi},m_{\pi}^2,m_{\pi}^2 ) & = & \frac{a(m_{\pi}^2)}{m_{\pi}^2}+\frac{1}{16\pi^2}\left [ 1-\sigma_{\pi}\ln\left(\frac{1+\sigma_{\pi}}{1-\sigma_{\pi}}\right)\right]\ , , \\ \im\,b(s_{\pi},m_{\pi}^2,m_{\pi}^2 ) & = & \frac{\sigma_{\pi}}{16\pi}\ , , \\
b(s_{\pi},m_k^2,m_k^2 ) & = & \frac{a(m_k^2)}{m_k^2}+\frac{1}{16\pi^2}-\frac{1}{8\pi^2}\left(\frac{4m_k^2}{s_{\pi}}-1\right)^{1/2}\arctan\left(\frac{4m_k^2}{s_{\pi}}-1\right)^{-1/2}\ , , \\
b(s_{\pi},m_{\eta}^2,m_{\eta}^2 ) & = & \frac{a(m_{\eta}^2)}{m_{\eta}^2}+\frac{1}{16\pi^2}-\frac{1}{8\pi^2}\left(\frac{4m_{\eta}^2}{s_{\pi}}-1\right)^{1/2}\arctan\left(\frac{4m_{\eta}^2}{s_{\pi}}-1\right)^{-1/2}\ , , \\
b(s_{\pi},m_{\ell}^2,m_k^2 ) & = & \frac{1}{2}\,\frac{a(m_{\ell}^2)}{m_{\ell}^2}+\frac{1}{2}\,\frac{a(m_k^2)}{m_k^2}+\frac{1}{16\pi^2}\left [ 1-\frac{1}{2s_{\pi}}\,(m_{\ell}^2-m_k^2)\ln\left(\frac{m_{\ell}^2}{m_k^2}\right)\right ] \nonumber\\ & & + \frac{1}{16\pi^2s_{\pi}}\,\sqrt{(m_{\ell}+m_k)^2-s_{\pi}}\sqrt{(m_{\ell}-m_k)^2-s_{\pi } } \nonumber\\ & & \times\ln\frac{\sqrt{(m_{\ell}+m_k)^2-s_{\pi}}+\sqrt{(m_{\ell}-m_k)^2-s_{\pi}}}{\sqrt{(m_{\ell}+m_k)^2-s_{\pi}}-\sqrt{(m_{\ell}-m_k)^2-s_{\pi}}}\ , , \\ \re\,b(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) & = & \frac{1}{2}\,\frac{a(m_{\ell}^2)}{m_{\ell}^2}+\frac{1}{2}\,\frac{a(m_{\pi}^2)}{m_{\pi}^2}+\frac{1}{16\pi^2}\left [ 1-\frac{1}{2t_{\pi}}\,(m_{\ell}^2-m_{\pi}^2)\ln\left(\frac{m_{\ell}^2}{m_{\pi}^2}\right)\right ] \nonumber\\ & & -\frac{1}{16\pi^2t_{\pi}}\,\sqrt{t_{\pi}-(m_{\ell}+m_{\pi})^2}\sqrt{t_{\pi}-(m_{\ell}-m_{\pi})^2 } \nonumber\\ & & \times\ln\frac{\sqrt{t_{\pi}-(m_{\ell}-m_{\pi})^2}+\sqrt{t_{\pi}-(m_{\ell}+m_{\pi})^2}}{\sqrt{t_{\pi}-(m_{\ell}-m_{\pi})^2}-\sqrt{t_{\pi}-(m_{\ell}+m_{\pi})^2}}\ , , \\
\im\,b(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) & = & \frac{1}{16\pi t_{\pi}}\,\sqrt{t_{\pi}-(m_{\ell}+m_{\pi})^2}\sqrt{t_{\pi}-(m_{\ell}-m_{\pi})^2}\ , , \\
b(t_{\pi},m_{\pi}^2,m_k^2 ) & = & \frac{1}{2}\,\frac{a(m_{\pi}^2)}{m_{\pi}^2}+\frac{1}{2}\,\frac{a(m_k^2)}{m_k^2}+\frac{1}{16\pi^2}\left [ 1-\frac{1}{2t_{\pi}}\,(m_{\pi}^2-m_k^2)\ln\left(\frac{m_{\pi}^2}{m_k^2}\right)\right ] \nonumber\\ & & + \frac{1}{16\pi^2t_{\pi}}\,\sqrt{(m_{\pi}+m_k)^2-t_{\pi}}\sqrt{(m_{\pi}-m_k)^2-t_{\pi } } \nonumber\\ & & \times\ln\frac{\sqrt{(m_{\pi}+m_k)^2-t_{\pi}}+\sqrt{(m_{\pi}-m_k)^2-t_{\pi}}}{\sqrt{(m_{\pi}+m_k)^2-t_{\pi}}-\sqrt{(m_{\pi}-m_k)^2-t_{\pi}}}\ , , \\
b(t_{\pi},m_{\eta}^2,m_k^2 ) & = & \frac{1}{2}\,\frac{a(m_{\eta}^2)}{m_{\eta}^2}+\frac{1}{2}\,\frac{a(m_k^2)}{m_k^2}+\frac{1}{16\pi^2}\left [ 1-\frac{1}{2t_{\pi}}\,(m_{\eta}^2-m_k^2)\ln\left(\frac{m_{\eta}^2}{m_k^2}\right)\right ] \nonumber\\ & & -\frac{1}{8\pi^2t_{\pi}}\,\sqrt{(m_{\eta}+m_k)^2-t_{\pi}}\sqrt{t_{\pi}-(m_{\eta}-m_k)^2 } \nonumber\\ & & \times\arctan\frac{\sqrt{t_{\pi}-(m_{\eta}-m_k)^2}}{\sqrt{(m_{\eta}+m_k)^2-t_{\pi}}}\,,\end{aligned}\ ] ] for the following integral , we shall distinguish between two cases .
* the lepton is an electron : @xmath145 \nonumber\\ & & + \textrm{if}\left ( 4m_{\pi}^2<s_{\pi}<(m_{\eta}-m_{\pi})^2\right ) \nonumber\\ & & \times\frac{1}{16\pi^2s_{\pi}}\,\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}\sqrt{(m_{\eta}-m_{\pi})^2-s_{\pi } } \nonumber\\ & & \times\ln\frac{\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}+\sqrt{(m_{\eta}-m_{\pi})^2-s_{\pi}}}{\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}-\sqrt{(m_{\eta}-m_{\pi})^2-s_{\pi } } } \nonumber\\ & & -\textrm{if}\left ( ( m_{\eta}-m_{\pi})^2<s_{\pi}<(m_k - m_e)^2\right ) \nonumber\\ & & \times\frac{1}{8\pi^2s_{\pi}}\,\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}\sqrt{s_{\pi}-(m_{\eta}-m_{\pi})^2 } \nonumber\\ & & \times\arctan\frac{\sqrt{s_{\pi}-(m_{\eta}-m_{\pi})^2}}{\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}}\ , , \end{aligned}\ ] ] * the lepton is a muon : @xmath146 \nonumber\\ & & + \frac{1}{16\pi^2s_{\pi}}\,\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}\sqrt{(m_{\eta}-m_{\pi})^2-s_{\pi } } \nonumber\\ & & \times\ln\frac{\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}+\sqrt{(m_{\eta}-m_{\pi})^2-s_{\pi}}}{\sqrt{(m_{\eta}+m_{\pi})^2-s_{\pi}}-\sqrt{(m_{\eta}-m_{\pi})^2-s_{\pi } } } \ , .
\end{aligned}\ ] ] these are scalar three - point functions whose definition and expressions were given in the appendix of ref .
@xcite . in
what follows , we sketch some of the particular cases that we need for the numerical evaluation of isospin breaking corrections : @xmath149\ , , \\
c(s_{\pi},s_{\pi},0,m_{\ell}^2,m_k^2,m_k^2 ) & = & \frac{1}{m_k^2-m_{\ell}^2}\left [ b(s_{\pi},m_k^2,m_k^2)-b(s_{\pi},m_{\ell}^2,m_k^2)\right ] \ , , \\
\re\,c(m_{\pi}^2,s_{\pi},m_{\pi}^2,m_{\gamma}^2,m_{\pi}^2,m_{\pi}^2 ) & = & -\frac{1}{32\pi^2s_{\pi}\sigma_{\pi}}\left\ { 4\textrm{li}_2\left(\frac{1-\sigma_{\pi}}{1+\sigma_{\pi}}\right ) + \frac{4\pi^2}{3 } \right .
\nonumber\\ & & \left .
+ \left [ 4\ln ( \sigma_{\pi})-2\ln\left(\frac{m_{\gamma}^2}{s_{\pi}}\right ) + \ln\left(\frac{1-\sigma_{\pi}}{1+\sigma_{\pi}}\right)\right]\ln\left(\frac{1-\sigma_{\pi}}{1+\sigma_{\pi}}\right)\right\ } \ , , \\
\im\,c(m_{\pi}^2,s_{\pi},m_{\pi}^2,m_{\gamma}^2,m_{\pi}^2,m_{\pi}^2 ) & = & -\frac{1}{16\pi s_{\pi}\sigma_{\pi}}\left [ 2\ln ( \sigma_{\pi})-\ln\left(\frac{m_{\gamma}^2}{s_{\pi}}\right)\right]\ , ,
\\ c(m_{\ell}^2,s_{\pi},m_k^2,m_{\gamma}^2,m_{\ell}^2,m_k^2 ) & = & \frac{1}{16\pi^2}\,\frac{1}{m_km_{\ell}}\,\frac{\sigma_{\ell k}}{1-\sigma_{\ell k}^2 } \nonumber\\ & & \times\left\{\left [ 2\ln ( 1-\sigma_{\ell k}^2)-\frac{1}{2}\,\ln ( \sigma_{\ell k})-\ln\left(\frac{m_{\gamma}^2}{m_km_{\ell}}\right)\right]\ln ( \sigma_{\ell k } ) \right.\nonumber\\ & & -\frac{\pi^2}{6}+\frac{1}{2}\,\ln^2\left(\frac{m_{\ell}}{m_k}\right ) + \textrm{li}_2(\sigma_{\ell k}^2)\nonumber\\ & & \left .
+ \textrm{li}_2\left ( 1-\frac{m_{\ell}}{m_k}\,\sigma_{\ell k}\right ) + \textrm{li}_2\left ( 1-\frac{m_k}{m_{\ell}}\,\sigma_{\ell k}\right)\right\}\ , , \\
c(m_k^2,s_{\pi},m_{\ell}^2,0,m_k^2,m_k^2 ) & = & \frac{1}{16\pi^2}\,\frac{1}{m_km_{\ell}}\,\frac{\sigma_{\ell k}}{1-\sigma_{\ell k}^2 } \nonumber\\ & & \times\left[\ln\left(\frac{m_k^2}{m_k^2-m_{\ell}^2}\right ) + \ln\left(\frac{m_km_{\ell}}{m_k^2-m_{\ell}^2}\right ) \right . \nonumber\\ & & -\ln^2(\sigma_{kk})-\frac{1}{2}\,\ln^2\left(\frac{m_{\ell}}{m_k}\right ) -\frac{1}{2}\,\ln^2(\sigma_{\ell k } ) \nonumber\\ & & + \textrm{li}_2\left ( 1-\frac{m_{\ell}}{m_k}\,\sigma_{\ell k}\right ) + \textrm{li}_2\left ( 1-\frac{m_k}{m_{\ell}}\,\sigma_{\ell k}\right ) \nonumber\\ & & -\textrm{li}_2\left ( 1-\frac{m_k}{m_{\ell}}\,\frac{\sigma_{\ell k}}{\sigma_{kk}}\right ) -\textrm{li}_2\left ( 1-\frac{m_k}{m_{\ell}}\,\sigma_{\ell k}\sigma_{kk}\right ) \nonumber\\ & & \left . -\textrm{li}_2\left ( 1-\frac{m_{\ell}}{m_k}\,\frac{\sigma_{\ell k}}{\sigma_{kk}}\right ) -\textrm{li}_2\left
( 1-\frac{m_{\ell}}{m_k}\,\sigma_{\ell k}\sigma_{kk}\right)\right\}\ , , \\
c(t_{\pi},t_{\pi},0,m_{\ell}^2,m_{\pi}^2,m_k^2 ) & = & \frac{1}{32\pi^2t_{\pi}}\,\frac{1}{m_k^2-m_{\ell}^2}\left\ { ( m_k^2-m_{\pi}^2+t_{\pi})\ln\left(\frac{m_{\ell}^2}{m_k^2}\right ) \right . \nonumber\\ & & + x_0\ln\frac{m_k^2-m_{\pi}^2+t_{\pi}+x_0}{m_k^2-m_{\pi}^2+t_{\pi}-x_0}-x_1\ln\frac{m_k^2-m_{\pi}^2+t_{\pi}+x_1}{m_k^2-m_{\pi}^2+t_{\pi}-x_1 } \nonumber\\ & & -x_0\ln\frac{(x_0+m_k^2-m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}{(x_0-m_k^2+m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & + x_1\ln\frac{(x_1+m_k^2-m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}{(x_1-m_k^2+m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & -(m_k^2-m_{\ell}^2)\ln\frac{(x_0+m_k^2-m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}{(x_1+m_k^2-m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & -(m_k^2-m_{\ell}^2)\ln\frac{(x_0-m_k^2+m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2)}{(x_1-m_k^2+m_{\ell}^2)^2-\lambda ( t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & -\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)\ln\frac{m_k^2-m_{\ell}^2+x_0+\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)}{m_k^2-m_{\ell}^2+x_0-\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & -\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)\ln\frac{m_k^2-m_{\ell}^2-x_0+\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)}{m_k^2-m_{\ell}^2-x_0-\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & + \lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)\ln\frac{m_k^2-m_{\ell}^2-x_1+\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)}{m_k^2-m_{\ell}^2-x_1-\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2 ) } \nonumber\\ & & \left .
+ \lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)\ln\frac{m_k^2-m_{\ell}^2+x_1+\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)}{m_k^2-m_{\ell}^2+x_1-\lambda^{1/2}(t_{\pi},m_{\ell}^2,m_{\pi}^2)}\right\}\ , , \\
c(m_{\pi}^2,t_{\pi},m_{\ell}^2,0,m_{\pi}^2,m_k^2 ) & = & \frac{1}{16\pi^2}\,\frac{1}{m_{\ell}m_{\pi}}\,\frac{\sigma_{\ell\pi}}{1-\sigma_{\ell\pi}^2 } \nonumber\\ & & \times\left\{\ln ( -\sigma_{\ell\pi})\left[\ln\left(\frac{m_{\ell}m_k}{m_k^2-m_{\ell}^2}\right)+\ln\left(\frac{m_{\pi}m_k}{m_k^2-m_{\ell}^2}\right)\right ] -\frac{\pi^2}{6 } \right . \nonumber\\ & & + \frac{1}{2}\,\ln^2\left(\frac{m_{\ell}}{m_{\pi}}\right ) -\ln^2\left(\frac{m_{\ell}}{m_k}\right ) -\frac{1}{2}\,\ln^2(-\sigma_{\ell\pi})-\ln^2(\sigma_{\pi k } ) \nonumber\\ & & -\frac{1}{2}\,\ln^2\left ( 1-\frac{m_{\ell}}{m_{\pi}}\,\sigma_{\ell\pi}\right)-\frac{1}{2}\,\ln^2\left ( 1-\frac{m_{\pi}}{m_{\ell}}\,\sigma_{\ell\pi}\right ) \nonumber\\ & & + \frac{1}{2}\,\ln^2\left ( 1-\frac{m_{\ell}}{m_k}\,\frac{\sigma_{\ell\pi}}{\sigma_{\pi k}}\right ) + \frac{1}{2}\,\ln^2\left(1 -\frac{m_k}{m_{\ell}}\,\frac{\sigma_{\ell\pi}}{\sigma_{\pi k}}\right ) \nonumber\\ & & + \frac{1}{2}\,\ln^2\left ( 1-\frac{m_{\ell}}{m_k}\,\sigma_{\ell\pi}\sigma_{\pi k}\right ) + \frac{1}{2}\,\ln^2\left ( 1-\frac{m_k}{m_{\ell}}\,\sigma_{\ell\pi}\sigma_{\pi k}\right ) \nonumber\\ & & -\textrm{li}_2\left(\frac{m_{\ell}}{m_{\ell}-m_{\pi}\sigma_{\ell\pi}}\right ) -\textrm{li}_2\left(\frac{m_{\pi}}{m_{\pi}-m_{\ell}\sigma_{\ell\pi}}\right ) \nonumber\\ & & + \textrm{li}_2\left(\frac{m_{\ell}}{m_{\ell}-m_k\sigma_{\ell\pi}\sigma_{\pi k}}\right ) + \textrm{li}_2\left(\frac{m_k}{m_k - m_{\ell}\sigma_{\ell\pi}\sigma_{\pi k}}\right ) \nonumber\\ & & \left .
+ \textrm{li}_2\left(\frac{m_{\ell}\sigma_{\pi k}}{m_{\ell}\sigma_{\pi k}-m_k\sigma_{\ell\pi}}\right ) + \textrm{li}_2\left(\frac{m_k\sigma_{\pi k}}{m_k\sigma_{\pi k}-m_{\ell}\sigma_{\ell\pi}}\right)\right\}\,,\end{aligned}\ ] ] | we evaluate the size of isospin breaking corrections to form factors @xmath0 and @xmath1 of the @xmath2 decay process @xmath3 which is actually measured by the extended na48 setup at cern .
we found that , keeping apart the effect of coulomb interaction , isospin breaking does not affect modules .
this is due to the cancelation between corrections of electromagnetic origin and those generated by the difference between up and down quark masses . on the other hand ,
electromagnetism affects considerably phases if the infrared divergence is dropped out using a minimal subtraction scheme .
consequently , the greatest care must be taken in the extraction of @xmath4 phase shifts from experiment . |
simulating the data that a space instrument like corot will provide might look presomptuous .
indeed , it is certainly , when comparing to previous comparable instruments like iphir or golf .
these two examples show that the nominal behaviour of the instrument is not always reached , but this does not prevent this instrument to provide very interesting data .
however , despite some technical problems , iphir and golf yielded a wealth of scientific results .
thus , what is the interest of simulating corot data ? how close to reality these simualtions will get ? this might not be the most important fact as the preparation of these simulations will help us to prepare the analysis of real data and to be ready in case of unexpected technical behaviour of the instrument perturbating the data , or unexpected physical behaviour of the targets of the instrument .
a consequence of that is that the simulation tool must include technical and physical aspects , making the task even more difficult .
these aspects cover : photon noise , p modes excitation , granulation signal , stellar activity signal , orbital perturbations , stellar rotation ... the software presented here is freely available at : + www.lesia.obspm.fr/@xmath0corotswg/simulightcurve.html
the photon noise is certainly the easiest component of the noise to simulate ... as far as it has the expected behaviour : a true white noise with a variance depending on the photon counts .
the corot specifications impose a level of photon noise of @xmath1 ppm in the amplitude spectrum of a star of magitude @xmath2 for an observing duration of 5 days . the stellar flux for a given magnitude @xmath3 being related to the stellar magnitude by : @xmath4 and knowing that the level of noise varies as the square root of the flux , the level of noise is related to the star magnitude by : @xmath5 as indicated above , this simple relation can model the photon noise but not some other photon counting perturbations as slow ( periods of hours or more ) trends in photon countings which might contributes to the low frequency spectrum of the noise .
the solar - like oscillations are stochastically excited and simulated here following the recipe of @xcite , recalled below .
each solar - like oscillation is a superposition of a large number of excited and damped proper modes .
each solar - like oscillation can then be decomposed as : @xmath6 \ , exp[-\eta ( t - t_j ) ] h(t - t_j ) + { \rm c.c } \label{simu : eq1}\ ] ] where @xmath7 is the amplitude at which the mode @xmath8 with proper frequency @xmath9 is excited by turbulent convection , @xmath10 at which it is excited , @xmath11 is the ( linear ) mode damping rate , @xmath12 is the heaviside function and `` c.c .
'' means complex conjugate .
the fourier transform of eq.[simu : eq1 ] yields the spectrum : @xmath13 whith @xmath14 , @xmath15 and @xmath16 is a complex number proportional to @xmath17 $ ] .
as the excitations are random , @xmath10 is random and hence @xmath16 has a random phase .
as the excitation are very numerous , according to the central limit princip the real and imaginary parts of the complex number @xmath18 are distributed according to a normal statitics .
hence the spectrum of the oscillations , @xmath19 , can be simulated by generating the real and imaginary parts of the complex number @xmath18 according to a normal statitics .
an inverse fourier transform is next applied in order to simulate the oscillations in the time domain .
this is the principe of the @xcite s recipe .
the question is now that is the constraints on the mean and variance of @xmath18 . from eq.[simu :
eq2 ] we deduce the power spectrum @xmath20 of the oscillation : @xmath21 the mean mode profile @xmath22 is obtained by averaging eq.[simu : eq3 ] over a larger number of realizations : @xmath23 where @xmath24 , which is by definition the variance of the complex number @xmath18 since @xmath25 .
we see from eq.[simu : eq4 ] that the mean profile has a lorentzian shape as it is in first approximation observed . according to the parseval - plancherel relation ,
@xmath12 is related to the mean square of the mode intrinsic amplitude in terms of luminosity fluctuations , @xmath26 , as ( see eg .
@xcite ) : @xmath27 where @xmath28 is a visibilty factors , which depends on the degree @xmath29 .
once the mode intrinsic amplitude @xmath30 , the mode line width @xmath31 and @xmath28 are given , we have then a constraint on the variance of @xmath18 .
note that the derivation of eqs.[simu : eq2]-[simu : eq5 ] assumes that the the mode life - time ( @xmath32 ) is much shorter than the duration of the observation ( @xmath33 ) , that is @xmath34 .
for life - times much longer than @xmath33 , different expressions are derived but the principe of the simulation remains the same . for sake of simplicity , we use the adiabatic assumption formulated by @xcite to deduce the maximum of @xmath30 from the maximum of the root mean square mode velocity @xmath35 according to : @xmath36 where @xmath37 is the effective temperature and the symbol @xmath38 refers to quantities related to the sun .
we take for the sun the _ rms values _ @xmath39 ppm ( see table 2 in @xcite and more recently @xcite ) and @xmath40 27 cm / s according to @xcite s seismic observations . in turn
the root mean square of the mode intrinsic velocity , @xmath41 , is related to the rate @xmath42 at which energy is injected into the mode by turbulent convection and the mode damping rate @xmath43 as : @xmath44 where @xmath45 is the mode mass , @xmath46 the mode inertia , @xmath47 the mode radial displacement and @xmath48 the height above the photosphere where oscillations are measured ( we consider @xmath49 ) . the way the quantities involved in eq.[simu : eq6 ] are model is explained in the next two sects .
theoretical mode damping rates are obtained from the tables calculated by @xcite .
these calculations rely on the non - local and time - dependent formulation of the convection of @xcite ( for more details , see @xcite ) .
the computation of the excitation rates @xmath42 is performed according to the model of stochastic excitation of @xcite .
the calculations assume - as in @xcite - a lorenzian function for modelling the convective eddy time - correlations .
furthermore , the characteritic wavenumber @xmath50 involved in the theory is assumed to be constant according to the simplification proposed by @xcite .
its value is related to the value @xmath51 inferred in @xcite from a 3d simulation of the sun as : @xmath52 where @xmath53 is the mixing - length evaluated at the layer the convective velocity is maximum .
the solar model we consider is calculated with the cesam code ( @xcite ) and appropriate input physics , described in details in @xcite .
in particular , convection is modelled according to the classical mixing - length theory ( * ? ? ? * hereafter mlt ) with a mixing - length @xmath54 where @xmath55 is the pressure scale height and @xmath56 is the mixing - length parameter .
in contrast with @xcite , the atmosphere is restored from the eddington classical gray atmosphere and microscopic diffusion is included according to the simplified formalism of @xcite .
the calibration of the solar model , in luminosity and radius for an age of @xmath57 gyr , fixes the initial helium content @xmath58 , metallicity @xmath59 and the mlt parameter @xmath60 .
the oscillation eigenfunctions and hence the mode masses , @xmath61 , in eq .
[ simu : eq6 ] are calculated with the adiabatic adipls pulsation code ( @xcite ) .
granulation can be considered as noise if the aim of the observation is the stellar oscillations , but it carries some information about the physics of the star , and very valuable information as the convection is a very badly described phenomenon in stellar modelling .
thus , we call it a @xmath62 and not a noise . however , granulation can be described only with a statistical approach . moreover ,
its contribution in the fourier domain is not independent of frequency : it will contribute more at low frequency .
a common description is to consider that signal @xmath63 as a random signal with some memory : then , its autocorrelation function is : @xmath64 where @xmath65 is the amplitude of this signal ( to be related to its variance ) and @xmath66 a characteristic time . knowing that the fourier transform of the autocorrelation function is the squared modulus of the function @xmath63 , one has : @xmath67 thus , the fourier spectrum of the granulation is modelled as a lorentzian which , if it is summed over all frequencies , yields the variance of the signal and then its relation with @xmath65 : @xmath68 so , knowing the intensity standard deviation due to granulation and its characteristic time , it is possible to model its fourier spectrum ( and considering its phase is random , it is possible to model the corresponding time series by an inverse fourier transform ) .
this is in fact the approach developped by @xcite to model the solar noise spectrum , including granulation , mesogranulation and supergranulation .
it can be used to describe any noise with memory ( some electronic noise for example ) . in the case of granulation ,
the required parameters are estimated from theory of convection .
the eddy size @xmath69 is assumed to be equal to @xmath70 where @xmath71 is the mixing - length parameter and @xmath55 the pressure scale height .
the eddy overturn time is related to the eddy convective velocity , @xmath72 , by @xmath73 and is considered as the characteristic time of granulation @xmath66 .
the number of number of eddies seen at the star surface , @xmath74 , is @xmath75 where @xmath76 is the star radius .
this relation ignores of course that the medium is higly anisotropic .
according to the mixing - length theory ( see eg .
@xcite ) , the eddies contrast ( border / center of the granule ) , @xmath77 , can be related to the difference between the temperature gradient of the eddy , @xmath78 and that of the surrounding medium , @xmath79 , according to the relation : @xmath80 in turn , eq .
[ dl_l_eddy ] can be reduced to : @xmath81 where @xmath82 is the convection efficiency and @xmath83 the ratio between the convective flux and the total flux of energy .
this relation is finally calibrated in order to match the solar constraints , and the intensity standard deviation for the whole observed disc is @xmath84 all the quantities are obtained on 1d stellar models computed as explained in sect .
[ sect : solar - like - exc ] .
the stellar magnetic activity will induce intensity variation in time , mainly due to the presence of starspots crossing the observed disc because of the rotation .
however , some other sources of intensity variations are expected ( flares for example ) .
a first approach is to consider the intensity variations as described by the same way than granulation . knowing the standard deviation and the characteristic time of intensity variations allows to build the fourier spectrum of these variations as a lorentzian .
this approach has been used for example by @xcite .
the difficulty in the case of magnetic activity is that there is no theoretical description of the phenomenon .
thus , the parameters describing it are empirically derived .
+ the characteristic time is taken as the period of rotation of the star , or , if the rotation is slow , the instrinsic lifetime of a spot .
the latter is arbitrarily chosen as the one in the solar case , computed by @xcite .
the rotation period is computed from an empirical law involving the age and the @xmath85 color index of the star described in the same reference .
+ the standard deviation of intensity variations due to magnetic activity is also derived from an empirical law .
however , this law involves the rossby number @xmath86 ( ratio of rotation period @xmath87 and the overturn convective time at the base of the convection zone @xmath88 ) .
this number can be derived empirically @xcite , but in the present simulations , we derive @xmath88 from models
. then we use empirical laws to estimate @xmath89 .
another approach is to simulate in the intensity time series the influence of individual spots , to estimate their number and contrast ( as for example @xcite ) .
the expected result in the fourier spectrum should be similar to the first approach , but this detailed simulation should allow for rotation measurements .
this approach will be included in a further version of our simulation software . , showing that the p modes are easily detected , as well as granulation and activity , width=359 ]
a first example of the output of this software is shown in fig.[comp_sun ] : a simulation of a sun with a magnitude @xmath90 ( in order to have a very weak phoon noise ) is compared with a spectrum from the loi photometer ( @xcite ) onboard soho .
the agreement between simulation and ovservation is good , except maybe at very low frequency : the activity component of the signal is overestimated in the simulation ( this is explained as the sun does not fit very well the empirical law used for activity estimation ) . + the following example is a sun with a magnitude @xmath91 , showing that both p modes and activity should be visible with corot for such a star ( fig.[sun_m6 ] ) .
another example is shown in fig.[star1.5_m8 ] for a star of 1.5 solar mass and an age of 2.4 gyr : again , p modes and activity are detectable . for a star with a mass @xmath92 and a magnitude @xmath93 , showing also p modes , granulation and activity signatures , width=359 ]
this simulator software will continue to evolve with time . as indicated above ,
intensity modulation due to starspots will be included , as well as other stellar or instrumental signals , as for example instrumental perturbations due to orbital vraiations .
moreover , this effort of simulation will not end with the delivery of first data but will be continued after that .
the comparison with real data will allow to check for the validity of physical hypothesis used to simulate the different signals of astrophysics origin in the data .
this shoud bring a great amount of information on our knowmedge of these often not well known phenomena , which stellar simulation is often derived from the solar case . in parallel , the simulation of instrumental components of the signal will be improved to help the interpretation of real data .
all these reasons justify in our opinion the need for the simulation tool presented here . ,
c. , de ridder , j. , mazumdar , a. , et al .
2004 , in proceedings of the soho 14 / gong 2004 workshop ( esa sp-559 ) .
`` helio- and asteroseismology : towards a golden future '' .
12 - 16 july , 2004 .
new haven , connecticut , usa . editor : d. danesy . | in order to prepare the analysis of corot data , it has been decided to build a simple tool to simulate the expected light curves .
this simulation tools takes into account both instrumental constraints and astrophysical inputs for the corot targets .
for example , granulation and magnetic activity signatures are simulated , as well as p modes , with the expected photon noise . however , the simulations rely sometimes on simple approach of these phenomenons , as the main goal of this tool is to prepare the analysis in the case of corot data and not to perform the most realistic simulations of the different phenomenons . |
top quark is the heaviest elementary particle known to - date discovered at fermilab in 1995 @xcite . during the current run ii period of the tevatron accelerator ,
the growing statistics permits to study in detail further kinematic properties of the top pair production as well as recently observed electroweak production of individual top quarks .
we preset the latest results on the @xmath0 cross sections measurements , their combination and interpretation in terms of limits on new physics .
we summarise the current state of the observation of the single top quark production and the measurements and searches based on the @xmath1 vertex properties .
as the top quark was first observed produced in @xmath0 pairs via the strong interaction , the pair production is the obvious first handle to test perturbative quantum chromodynamics ( pqcd ) while the decaying system properties can be used as an electroweak laboratory to search for new physics . at tevatron
the production proceeds mostly via @xmath2 annihilation ( 85% ) accompanied by the @xmath3 fusion ( 15% ) .
the final states of the @xmath0 system are determined by the leptonic , hadronic or mixed decays modes of the @xmath4-bosons .
the dilepton ( @xmath5 ) final state is easy to identify , has low background , but also low branching fraction ; the @xmath6jet channels ( with prompt @xmath7 or cascade @xmath8 ) have larger branching fraction , more background , but have only one escaping neutrino ; the hadronic channel has the largest branching ratio , but suffers from overwhelming background from qcd multijet production .
dedicated @xmath9jets analyses are challenged by the @xmath10 lepton identification , but provide important additional information and offer windows on testing new models with prominent couplings to @xmath10 .
the cross - section analysis in the @xmath6jets was divided into several bins in terms of the jet multiplicity , with @xmath11 + 1,2 jets serving as a control sample to tune selection and data / simulation agreement , and the 3 and @xmath12 jet bins ( with larger signal fraction ) for the measurement itself .
we required at least one jet identified as a @xmath13-jet using a neural network ( nn ) @xmath13-tagger , one isolated lepton and large missing transverse energy .
additionally the leading jet @xmath14 was required to be @xmath15gev .
most of standard model backgrounds were modelled using alpgen+pythia parton shower , including theory nlo / lo scale factors .
the largest physics background comes from the @xmath16jets events , other backgrounds include diboson and single top production .
as there is no reliable model of the multijet production beyond several final state hard partons , the multijet background was taken from a special data with inverted cut on the lepton isolation .
the signal fraction grows with the number of the @xmath13-tagged jets ( see figure [ fig : ljjets ] , left ) . a kinematic likelihood discriminant based on several topological and kinematic variables ( see figure [ fig : ljjets ] , right )
was employed to fit the @xmath0 fraction and the cross - section .
cross - section measurements in @xmath6jets , @xmath5 and @xmath17 channels were combined using a total of 14 orthogonal channels , the combined cross - section was fitted using a likelihood function with poisson terms for signal and background yields .
systematic uncertainties are included using gaussian probability density function .
the combined result is @xmath18 , assuming @xmath19 @xcite .
the systematic uncertainties dominate the error on this result .
we also provide the ratios of cross - sections between the channels @xmath20 , where most of systematics cancel , leading to results on @xmath21 and @xmath22 which are consistent with the sm value of 1 and are shown in figure [ fig : xsect_combi_d0_cdf ] .
in addition , cross - section ratios between individual channels can be used to set upper limits on the @xmath23 branching ratio , as for @xmath24 one would observe an enhancement of the @xmath10 channel ( using @xmath25 ) .
leptophobic @xmath26 would lead to an enhancement of the @xmath6jets channel ( using @xmath27 .
we extracted the expected and observed upper limits as shown in figure [ fig : ttbar_h_limits ] using pseudo - experiments derived from the data distributions .
the current tevatron @xmath28 combined measurement is precise to @xmath29gev @xcite , triggering discussions on the definition and interpretations of the top mass which is actually measured .
a complementary information on @xmath28 with different sensitivity to theoretical and experimental uncertainties can be extracted from the @xmath30 dependence .
assuming different @xmath31 , the measured @xmath0 cross - section can be compared to nlo and approximate nnlo ( nll ) calculations , where @xmath28 definition is usually the pole mass .
the extracted values of @xmath28 range from @xmath32 to @xmath33gev depending on the theory used , with errors of @xmath34gev , being in good agreement with the world - average of @xmath35gev @xcite as depicted in figure [ fig : mass_from_xsect ] .
the electroweak ( single ) top quark production had been one of the remaining sm processes to be observed until recently @xcite , with the evidences for single top production reported by the tevatron collider experiments in 2007 @xcite and 2008 @xcite . indeed , this process has a much larger background and smaller cross - section compared to the @xmath0 production . at the same time ,
observation of the single top production is rewarding in providing a direct access to the @xmath1 vertex properties and another window to new physics .
two basic production modes can be distinguished at lo , denoted the @xmath36 and @xmath37-channels . the basic experimental requirements to measure the single top cross - section include the selection of events with one isolated lepton and at least two jets with at least one with @xmath38gev .
in addition , the missing transverse energy ( @xmath39-signature ) is required to be @xmath40gev in the two - jet ( @xmath41 ) bin .
one or two of the jets is required to originate from @xmath13-hadrons using the secondary vertex information incorporated into a neural net @xmath13-tagger . selecting 4,519 @xmath13-tagged events
, we expect @xmath42 single top events with signal acceptances of 3.7% ( 2.5% ) for @xmath43 ( @xmath44 ) .
as the expected signal fraction of 39% is in fact smaller than the background uncertainty , a simple counting experiment method is impossible and multivariate techniques are needed to extract the signal .
three analysis methods named boosted decision trees ( bdt ) , bayesian neural - network ( bnn ) , matrix element ( me ) giving consistent results were employed and used to produce a combined discriminant ( see figure [ fig : d0_singletop_2 ] ) . the result based on @xmath45 data collected by the d detector
is shown in table [ table : singletop_individual ] .
being 5060% correlated , combining the individual results leads to an improved significance .
the combined cross - section measurement is then @xmath46 one can also search new phenomena in the same final state , e.g. for charged higgs bosons decaying to top and bottom quarks in @xmath49 collisions @xcite .
limits on the charged higgs mass a function of @xmath50 were set based on a discriminating variable of the jet - jet-@xmath4 invariant mass ( figure [ fig : d0_singletop_4 ] ) .
single top production also provides a direct access to the properties of the @xmath1 vertex , allowing a search for anomalous top quark couplings @xcite .
the sm lagrangian allows only the @xmath51 vertex structure , but a more general form can be probed : @xmath52 where sm corresponds to the case of @xmath53 , @xmath54 . using the @xmath4 helicity measurement in the @xmath55jet channel ( which constrains form factors ratios ) for probability density priors , the extracted bayesian limits on the vector and axial vector form factors are @xmath56 , @xmath57 and @xmath58 assuming @xmath53 . | we report on measurements of the @xmath0 production cross section at a center - of - mass energy of 1.96 tev at the d experiment during run ii of the fermilab tevatron collider .
we use candidate events in lepton+jets and dilepton final states . in the most sensitive channel ( lepton+jets channel ) , a neural network algorithm that uses lifetime information to identify b - quark jets is used to distinguish signal from background processes .
we also present measurements of single top quark production at d using several multivariate techniques to separate signal from background . |
positive and negative feedback loops are frequently - occurring motifs in gene transcription regulatory networks and signaling pathways @xcite .
the components of a feedback loop are genes , proteins and other molecules which are connected by regulatory interactions .
depending on the components and their interactions , feedback loops have distinct roles in diverse regulatory systems .
a regulatory interaction is positive ( negative ) if an increase in the amount or activity of one component increases ( decreases ) the amount or activity of its interaction partner .
a feedback loop is positive ( negative ) if the number of repressing interactions is zero or even ( odd ) .
a large number of experiments and theoretical studies elucidate the major functional characteristics of feedback loops with simple structure @xcite .
positive feedback in a gene transcription regulatory network ( gtrn ) tends to enhance protein levels whereas negative feedback favours homeostasis , i.e. , maintenance of proteins at a desired level .
the simplest feedback loop has only one component which is thus self - regulating .
for such a motif in a gtrn , a protein promotes / represses its own production via autoactivation / autorepression of the expression of its gene .
a positive feedback loop with two components and two regulatory interactions is of two types : double negative and double positive . again , considering a gtrn , the protein products of the two genes in a double negative feedback loop repress each other s synthesis .
the construction of a synthetic circuit , the genetic toggle , is based on this motif @xcite .
the double positive feedback loop is defined by two genes , the protein products of which promote each other s synthesis .
there are several examples of two - component positive feedback loops in natural cellular networks @xcite , a prominent example being the cell division cycle , the regulatory network of which contains both double positive and double negative feedback loops @xcite . in this case
, the loops control enzymatic activity .
the double negative feedback loop , because of its more common occurrence , has been extensively studied in contrast to the double positive feedback loop .
the next stage of complexity in feedback loops involves linked positive and negative feedback loops @xcite .
the key variables in the dynamics of feedback loops are the concentrations of the component molecules . in the case of a gtrn
, these may be the protein concentrations . in a deterministic description ,
the time evolution of the concentrations is determined by solving a set of coupled differential equations , the number of equations being equal to the number of variables . in reality , the biochemical events associated with gene expression are probabilistic in nature and this is reflected in the presence of fluctuations ( termed noise ) around mean protein levels @xcite .
a stochastic description of time evolution is thus more appropriate .
a single positive feedback loop has a tendency to amplify noise , also the time taken to reach the steady state protein level is longer than that in the case of an unregulated gene @xcite .
interlinking of two positive feedback loops with slow and fast dynamics results in a switch with rapid activation and slow deactivation times and a marked resistance to noise in the upstream signaling pathway @xcite .
addition of a single negative feedback loop leads to rapid deactivation in the absence of the signal which activates the switch @xcite .
the combination of positive and negative feedback loops may give rise to excitability with transient activation of protein levels .
recent experiments suggest that competence development in b. subtilis is achieved via excitability @xcite . in this paper
, we study the functional characteristics of a two - gene double positive feedback loop coupled with autorepression of the expression of one of the genes .
the major motivation for studying this specific motif is its presence in the gtrn controlling the pancreatic @xmath0-cell function @xcite .
the hormone insulin is a small protein that is synthesized in the @xmath0-cells and secreted when an increase in the blood glucose level is sensed .
glucose metabolism releases energy needed by cells to do useful work .
insulin is necessary to metabolize glucose and thereby control its level in the blood .
diabetes occurs due to an excessive accumulation of glucose in the blood brought about by an insufficient production of or reduced sensitivity to insulin .
the core of the @xmath0-cell transcriptional network consists of a double positive feedback loop in which the transcription factors @xmath2 and @xmath3 , belonging to the nuclear hepatocyte family , activate each other s synthesis .
there is also some evidence that @xmath3 autorepresses its own synthesis @xcite .
mutations in the transcription factors @xmath2 and @xmath3 give rise to a type of diabetes known as maturity - onset diabetes of the young ( mody ) which has an early onset with age less than usually @xmath4 years .
there are six different forms of mody of which mody 1 and mody 3 are caused by mutations in the genes @xmath5 and @xmath6 respectively @xcite .
the structure of the regulatory network , of which the two genes @xmath6 and @xmath5 are integral components , is not fully known .
a partial structure of the complex network is shown in @xcite involving the genes @xmath7 @xmath8 @xmath9 @xmath10 @xmath11 @xmath12 @xmath13 and @xmath14 the genes collectively control the transcription of a number of important genes involved in glucose metabolism in the @xmath15cell .
these include the glucose transporter @xmath16 ( @xmath17 gene , the glucokinase gene encoding the glycolytic enzyme glucokinase which acts as glucose sensor and also the insulin gene .
odom et al . @xcite
combined chromatin immunoprecipitation assays with promoter microarrays to gain insight on the regulatory circuits formed by @xmath6 and @xmath18 both the proteins are found to control the activity of a large number of target genes in the @xmath15cell .
this recent finding as well as earlier experiments @xcite indicate that the @xmath6 and @xmath5 genes play a prominent role in the pancreatic @xmath15cell function .
mutations in the genes give rise to mody resulting in the impairment of glucose - stimulated insulin secretion .
several experiments @xcite provide clues on the possible molecular origins of mody .
the cross - regulatory interactions between @xmath2 and @xmath3 are switched on as pancreatic @xmath0-cells receive the signals to differentiate .
the double positive feedback loop has the potential for bistability , i.e. , two stable steady states .
the two states are a basal state in which the two genes have low activity and an activated state which corresponds to high protein levels .
the states are analogous to the off and on states of a switch .
normal functioning of the pancreatic @xmath0-cells requires the two - gene feedback loop to be in the on state .
the circuit operation is , however , vulnerable to decreased gene dosage caused by mutations ( in a diploid organism each gene has two identical copies ) .
genetic disorders , termed haploinsufficiency , are known to occur due to reduced gene dosage resulting in decreased protein levels @xcite .
gene expression noise increases the probability that a protein level falls below a threshold value so that the protein amount is insufficient for meaningful activity .
the loss of vital protein functions is responsible for the occurrence of genetic disorders .
mody , brought about by reduced gene dosage , is thus an example of haploinsufficiency @xcite .
we construct a mathematical model to study the dynamics of the core circuit consisting of a double positive feedback loop coupled with autorepression of the @xmath5 gene .
we use both deterministic and stochastic approaches to identify the functional features of the motif and discuss their possible relevance in the occurrence of mody
the circuit diagram of the motif to be studied is shown in the figure 1 .
@xmath19 and @xmath20 represent the genes @xmath6 and @xmath5 respectively .
the arrow sign denotes activation by the appropriate protein product and the hammerhead sign denotes repression .
the chemical kinetic schemes corresponding to the expression of genes @xmath19 and @xmath20 are shown in figures 2(a ) and 2(b ) .
the protein products of @xmath19 and @xmath20 are denoted by @xmath21 and @xmath22 .
we assume that the regulation of gene expression is mediated by the protein dimers @xmath23 and @xmath24 , @xmath25 and @xmath26 being the binding constants of dimerization . for each gene , there are two rates of protein synthesis : a basal rate ( rate constants @xmath27 and @xmath28 ) and an activated rate ( rate constants @xmath29 and @xmath30 ) . in the second case ,
protein synthesis occurs in the activated state of the gene ( @xmath31 and @xmath32 ) attained via the binding of protein dimers @xmath23 and @xmath24 to the genes @xmath19 and @xmath20 respectively .
the associated binding constants are @xmath33 and @xmath34 .
the rate constants for protein degradation are @xmath35 and @xmath36 with @xmath37 denoting the degradation product .
dimer degradation is not taken into account as its rate is few - fold lower than the degradation rate of protein monomers .
for the gene @xmath20 , there is an extra biochemical event representing autorepression .
the dimers @xmath23 and @xmath24 bind the promoter region of the gene gs competitively , i.e. , the binding of one type of dimer excludes the binding of the other type . when the dimer @xmath23 binds at @xmath20 , there is complete repression
. the binding constant is denoted by @xmath38 .
the protein concentrations @xmath22 and @xmath21 are the dynamical variables in the system .
the time scale of binding events , in general , is much faster than that of protein synthesis and degradation .
the bound complexes thus reach the steady state at an earlier time point . taking this into account ,
the differential rate equations describing the time evolution of the protein concentrations @xmath22 and @xmath21 are : @xmath39 @xmath40 with @xmath41 @xmath42 there are two conservation equations for the total concentrations @xmath43 and @xmath44 of the genes @xmath20 and @xmath19 .
@xmath45 @xmath46 after an appropriate change in variables @xmath47 the differential rate equations ( 1 ) and ( 2 ) are transformed into @xmath48 @xmath49 the different parameters are given by @xmath50 the variable @xmath51 is dimensionless whereas the variables @xmath52 and @xmath53 have the dimensions of concentration expressed in units of [ nm ] .
the parameters @xmath54 and @xmath55 are dimensionless while the parameters @xmath56 , @xmath57 and @xmath0 are expressed in units of @xmath58^{2}}$ ] .
the dimensions of @xmath59 and @xmath60 are in units of @xmath61 $ ] with one gene copy corresponding to approximately @xmath62 $ ] . from now on ,
the units will not be explicitly mentioned .
table 1 displays all the parameters and rescaled parameters of the two - gene model as well as their meanings and defining formulae .
we use the software package xppaut @xcite to probe the dynamics of the double positive feedback loop and the effect of autorepression of the @xmath22 proteins on the dynamics .
we focus on how the steady state value of u ( rescaled concentration of @xmath22 proteins ) changes as a function of the different parameters in equations ( 8) and ( 9 ) . in the steady state ,
the rates of change @xmath63 and @xmath64 are zero .
figure 3(a ) shows a plot of @xmath52 versus @xmath54 when the autorepression strength given by @xmath56 is zero .
the other parameters have values @xmath43
= @xmath65@xmath66 , @xmath67 , @xmath68 , @xmath69 and @xmath70 .
the plot shows that a region of bistability separates two region of monostability .
the two stable states in the bistable region correspond to low and and high values of @xmath52 . in this region and at a specific value of @xmath54
, the choice between the stable steady states is history - dependent , i.e. , depends on initial conditions @xcite .
if the value of @xmath54 is initially low , the system ends up in the low @xmath52 state .
as @xmath54 increases , the system enters the region of bistability but continues to be in the low expression state till a bifurcation point is reached . at this point ,
a discontinuous jump to the high @xmath52 state occurs and the system becomes monostable .
bistability is accompanied by hysteresis , i.e. , the value of @xmath54 at which the switch from the low to the high expression state occurs is greater than the value of @xmath54 ( the lower bifurcation point ) at which the reverse transition takes place .
the two stable branches are separated by a branch of unstable steady states ( dash - dotted line ) which are not experimentally accessible .
there are now several known systems in which bistability and hysteresis have been observed experimentally @xcite .
figure 3(b ) shows the plot of @xmath52 versus @xmath54 for the same parameter values as in figure 3(a ) except that the copy number of the @xmath20 gene is reduced from two to one , i.e. , @xmath43 has the value @xmath71 . a comparison of figures 3(a ) and
( b ) shows that with reduced copy number the extent of the region of bistability in considerably increased .
the same conclusion is reached when the steady state values of @xmath52 are plotted versus the parameter @xmath72 the region of bistability is lower in extent when the parameter @xmath73 a measure of the autorepression strength , is increased from zero .
the value of @xmath56 is changed by modifying the value of @xmath38 ( equation ( 10 ) ) , the binding constant for repressor binding at the @xmath20 gene .
figure 4 shows the phase portrait corresponding to equations ( 8) and ( 9 ) with the parameter values @xmath74
@xmath75 @xmath76 @xmath77 and @xmath78 the system is bistable for the parameter values quoted .
the nullclines , obtained by putting @xmath79 @xmath80 intersect at three points , the fixed points of the dynamics .
the lower and upper fixed points are stable whereas the intermediate fixed point is unstable , in fact , a saddle node @xcite .
the stable manifold of the saddle node divides the @xmath81phase space into two basins of attraction .
trajectories starting in the lower ( upper ) basin of attraction end up at the lower ( upper ) stable fixed point as shown in figure 4 .
a trajectory initiated on the stable manifold stays on it and ends at the saddle node .
a typical trajectory asymptotically approaches the unstable manifold as @xmath82 a trajectory is obtained by plotting the values of @xmath52 and @xmath53 at different time points , determined by solving equations ( 8) and ( 9 ) .
the arrow direction on a trajectory denotes increasing time .
> p2.5in > p3 in * parameter / rescaled parameter * & * * meaning / defining formula * * + @xmath83 & rate constants for basal protein synthesis + @xmath84 & rate constants for activated protein synthesis + @xmath85 & binding constants of protein dimers @xmath23 and @xmath24 + @xmath86 & binding constants for the binding of protein dimers @xmath23 and @xmath24 at the genes @xmath19and @xmath20 + @xmath87 & rate constants for protein degradation + @xmath38 & binding constant for repressor dimer binding at gene @xmath88 @xmath38 thus denotes the strength of repression + @xmath54 & @xmath89 , ratio of activated and basal rate constants for synthesis of s proteins + @xmath55 & @xmath90 ratio of activated and basal rate constants for synthesis of @xmath21 proteins + @xmath56 & @xmath91 with @xmath92 and @xmath26 kept fixed , @xmath56 can be varied by changing @xmath38 thus providing a measure of repression strength + @xmath57 & @xmath93 + @xmath0 & @xmath94 + @xmath95 & abbreviations defined in equations ( 3 ) and ( 4 ) + figures 5(a ) shows the plot of @xmath55 versus @xmath54 exhibiting regions of monostability and bistability .
the parameter values are the same as before with @xmath70 and @xmath96 .
the regions of bistability , enclosed within the red and black curves , correspond to @xmath97 and @xmath98 respectively .
the difference in the locations of the two loops in the logarithmic plots clearly shows that the bistable region is of greater extent when the gene copy number is reduced from two to one .
the region of bistability is decreased in extent when autorepression is taken into account ( figure 5(b ) with @xmath99 figure 6 shows the @xmath100 plot with the regions of bistability falling within the red ( @xmath97 ) and black ( @xmath98 ) curves respectively .
the value of @xmath56 is changed by varying @xmath38 ( equation 10 ) with @xmath101 .
a major advantage of combining a double positive feedback loop operating between two genes with autorepression of the expression of one of the genes lies in dosage compensation @xcite .
this relates to the fact that the fall in steady state protein levels , brought about by a reduction in the gene copy number , is less when autorepression is included , compared to the case when there is no autorepression .
a measure of dosage compensation is provided by the quantity @xmath102 , termed percentage gain , defined as @xmath103 where @xmath104 denotes the steady state concentration of @xmath22 proteins when the copy number of the @xmath20 gene , @xmath43 , is one .
the parameter @xmath56 is a measure of the repression strength .
@xmath102 is calculated by keeping the mean level of @xmath22 proteins to be the same in the two cases @xmath96 and @xmath105 when @xmath98 .
this is achieved by adjusting the binding constant @xmath33 contained in the parameter @xmath57 in equation ( 10 ) .
the other parameter values are @xmath106 , @xmath68 , @xmath69 and @xmath77 .
figure 7 shows the plot of @xmath102 versus @xmath56 for the parameter values mentioned .
as @xmath56 increases from zero , there is initially a sharp increase in the value of g followed by a slower growth which ultimately leads to a near - saturation of g values .
the results obtained in the deterministic approach provide insight on the advantages of autorepression in the non - occurrence of the genetic disorder mody .
the normal functioning of pancreatic @xmath0-cells requires the @xmath2 and @xmath3 protein levels to be high , i.e. , the two - gene system should be in the on state .
the genesis of mody lies in a substantial fraction of the @xmath0-cells being in the off state .
this is brought about by mutations in the @xmath6 and @xmath5 genes giving rise to a fall in the steady state protein levels . in terms of the two - gene model studied by us ,
the monostable high state , in which the levels of the p and s proteins are both high , represents the on state of normal @xmath0-cells .
the system may enter a region of bistability , in which both the on and off states are possible , due to the loss of a gene copy brought about by mutations .
we will show in the next section that fluctuations in the protein levels are responsible for transitions between the on and off states . in the deterministic scenario ,
the major advantages of the autorepressive feedback loop appear to be dosage compensation ( figure 7 ) as well as a lesser possibility of the system being in the bistable region due to a reduction in gene copy number .
the continuance of the system in the monostable high state ensures the normal functioning of cells .
similar conclusions are reached if the gene copy number @xmath44 is reduced from two to one .
there is , however , an asymmetry in the @xmath22 and @xmath21 protein levels as the expression of the gene @xmath19 is not autorepressed .
consider the two - gene network to be originally in the monostable high state . in the deterministic formalism
, the system continues to be in the high , i.e. , on state even if it enters the region of bistability due to the loss of a gene copy .
this is due to history dependence , since the system is initially in the on state it continues to be in the on state in the bistable region .
the protein levels corresponding to the on state are , however , lower in magnitude in the bistable region . in the pancreatic @xmath0-cells ,
the occurrence of mody is possible only when a sizable fraction of cells is in the off state .
the on@xmath1off and off@xmath1on transitions can be understood only when stochasticity in gene expression is taken into account .
we now give a simple physical picture of the origin of stochastic transitions @xcite . in the case under consideration ,
the dynamical variables are the protein concentrations @xmath52 and @xmath53 . in the case of deterministic time evolution ,
trajectories starting in individual basins of attraction stay confined to the specific basins with no possibility of a trajectory crossing from one basin to another . in the stochastic approach ,
the trajectories are no longer deterministic as the dynamical variables @xmath107 and @xmath108 are fluctuating . in the deterministic case , given the initial state defined by @xmath109 , the trajectory is fixed in the @xmath110-phase space . in the stochastic case ,
different trajectories are generated in repeated trials . a transient fluctuation , if sufficiently strong , switches the system dynamics from one basin to
the other brought about by the excursion of the trajectory across the boundary separating the two basins of attraction . in terms of the pancreatic @xmath0-cells ,
the switch to the off state hampers the normal functioning of the cells . for proper regulatory functions as transcription factors , the @xmath2 and @xmath3 protein levels are high with an optimal value as excessive protein amounts
are known to be harmful rather than beneficial @xcite . in this context
, it is pertinent to undertake a comparison of the functional characteristics of two - gene network models with and without the autorepressive loop and with the mean protein levels kept at the same high values in the two cases .
the last condition ensures the normal functioning of the cells in both the cases . in section 2 , we have identified certain advantages of the autorepressive loop as regards the system dynamics in a deterministic framework .
our goal is now to identify the desirable features of the model incorporating both a double positive feedback loop and an autorepressive loop taking the stochastic aspects of the dynamics into consideration .
this is done with the help of a detailed computer simulation based on the gillespie algorithm @xcite .
the algorithm enables one to keep track of the stochastic time evolution of the system .
the different biochemical reactions considered in the simulation are depicted in figures 2(a ) and 2(b ) .
the reactions are sixteen in number and are given by @xmath111 @xmath112 @xmath113 @xmath114 @xmath115 @xmath116 @xmath117 @xmath118 @xmath119 @xmath120 @xmath121 @xmath122 @xmath123 @xmath124 @xmath125 @xmath126 the different symbols are as explained in section 2 .
the stochastic rate constants , associated with the reactions , are c(i ) , i=1, ...
,16 , in the appropriate units .
the results of the simulation are shown in figures 8 - 9 .
figures 8(a)-(d ) show the distribution of @xmath20 protein levels , @xmath127 , in an ensemble of 4500 cells for repressor strengths @xmath128 and @xmath129 respectively after a simulation time of @xmath130 time units .
the gene copy numbers are @xmath131 and @xmath97 so that the system is in the region of bistability .
the values of the stochastic rate constants are @xmath132 @xmath133@xmath134,@xmath135,@xmath136,@xmath137 , @xmath138@xmath139 .
the value of @xmath56 is changed by varying the stochastic rate constant @xmath140 .
the value of the rate constant @xmath141 is changed to keep the mean protein levels to be the same when @xmath142 for all values of @xmath56 .
the value @xmath96 implies that only the double positive feedback loop contributes to the dynamics .
the distribution @xmath127 is found to be bimodal , i.e. , has two distinct peaks corresponding to the off and on states .
in all the cases , the cells are in the on state at time @xmath143 one finds that the fraction of cells in the off state decreases as the value of @xmath56 increases .
in fact , when @xmath144 the number of cells which are in the off state is larger than that in the on state . since initially
all the cells are in the on state , a large number of on @xmath1 off state transitions occur during the simulation time . for @xmath144
the reverse transition is , however , much rarer .
the role of the autorepressive loop thus appears to be to reduce the number of stochastic transitions from the on to the off state .
this makes the occurrence of mody , brought about by a sizeable fraction of the cell population existing in the off state , less probable .
there are two distinct time scales over which protein fluctuations occur .
the probability distribution @xmath127 versus @xmath52 has a two - peaked structure .
fluctuations on a short time scale confine the @xmath52 values to lie predominantly within individual peaks .
the long time scale corresponds to the time at which large fluctuations occur bringing about transitions between states belonging to different peaks .
the `` escape time '' is often very large and a quantitative measure is provided by the mean first passage time @xmath51 @xcite . in the present case ,
the values of @xmath145 and @xmath146 are quite large for different values of @xmath147 the maximum simulation time @xmath148 is @xmath149 time units for all values of @xmath147 for @xmath144 @xmath145 is around @xmath150 time units whereas @xmath146 is even larger . as @xmath56 increases , @xmath145 increases whereas @xmath146 decreases . for @xmath151 @xmath145
is as large as @xmath152 because of large escape times , the probability distribution @xmath127 versus @xmath52 is metastable on a large time scale @xcite . over shorter periods of time ,
the shape of the distribution remains more or less invariant .
the plots in figure 9 are obtained for an ensemble of 4500 cells .
for gene copy numbers @xmath131 and @xmath98 , the mean protein levels are adjusted to be the same irrespective of the values of @xmath56 .
the parameter values are so chosen that the system is in the monostable high region .
on reduction of @xmath43 to @xmath71 ( one copy of the gs gene ) , the system enters the region of bistability and is in the on state at time @xmath153 . after a period @xmath154 time units of stochastic time evolution ,
the percentage of cells in the off state is determined .
the red curve shows this percentage as a function of the repression strength @xmath56 .
the drop in the percentage of cells in the off state is found to be exponential .
the black curve shows the percentage of cells in the on state after @xmath154 time units , with all the cells being initially in the off state .
one finds that with increasing @xmath56 , the fraction of cells in the on state becomes larger .
the autorepressive loop has the effect of making the on state more stable and the off state more unstable .
this feature enhances the probability of the nonoccurrence of mody as there are infrequent transitions from the on to the off state . on the other hand
, the system has a lesser probability of remaining stuck in the off state compared to the case when there is no autorepressive loop .
in this paper , we have studied the functional characteristics of a motif consisting of a double positive feedback loop operating between two genes and a negative feedback loop in which the protein product of one gene represses its own synthesis .
the motif appears in the gene regulatory network controlling the pancreatic @xmath0-cell function @xcite .
loss of a gene copy due to mutations has been shown @xcite to be responsible for abnormal @xmath0-cell function resulting in mody .
we have studied the effect of reduced gene copy number on the dynamics of the model describing the two - gene motif . in a deterministic formalism based on differential rate equations , we identified regions of bistability in appropriate parameter regions .
the stable steady states , designated as the off and on states , correspond respectively to low and high protein levels .
the normal @xmath0-cells are expected to be in the monostable on state .
the occurrence of mody is brought about by a fraction of @xmath0-cells being in the off state .
the on@xmath1off switch can occur only in the bistable region .
negative feedback reduces the extent of the bistable region making it less likely that the cellular state falls in this part of the phase diagram .
the region of bistability , however , increases in size on reduction of the gene copy number making the on@xmath1off transitions more probable .
negative feedback also produces a mechanism of dosage compensation ( figure 7 ) .
the results hold true for a wide range of parameter values .
since switching to the off state is detrimental , one would have thought , from an evolutionary point of view , that two genes which are constitutively on would be more appropriate . in reality , the genes @xmath6 and @xmath5 form a positive feedback loop .
cross - regulation between the two genes is established when the pancreatic cells receive signals to differentiate @xcite .
the positive feedback loop provides a stable mechanism of gene expression since the two genes reset each other s activity to the functional state under physiological perturbations .
this serves to self - perpetuate the activity of the two genes and their targets in the pancreatic @xmath15cells .
normal functioning of these cells requires both the protein levels to be high .
the system of two genes that are constitutively on are less robust under physiological perturbations since there is no resetting mechanism by which both the genes are in the functional on state .
the theoretical suggestions of bistability due to the existence of a positive feedback loop @xcite , backed up by the results of our mathematical model , should be tested in actual experiments .
the on@xmath1off switch is brought about by protein fluctuations the origin of which lies in stochastic gene expression .
our major finding is that negative feedback makes the on@xmath1off transitions less probable and the off@xmath1on transitions more probable .
thus the function of the negative feedback appears to be to protect the normal @xmath0-cell function since the cell is more likely to be in the on state in this case .
the asymmetric response to fluctuations prevents switching off and facilitates switching on of the high expression state . in the deterministic scenario
, one finds that the difference between the on state and the unstable steady state protein level increases as the autorepression strength is increased whereas the difference between the unstable steady state and off state protein levels decreases on increasing the autorepression strength .
this may explain the asymmetry in the on@xmath1off and off@xmath1on switches when stochasticity is taken into account . for moderate strengths of autorepression ,
the system is locked in the on state for extremely long times . in our simulations
, we did not encounter on@xmath1off switches for very long trajectories ( @xmath155 seconds ) with @xmath156 .
this translates into lifetimes measured in years and explains the delayed onset of the diabetic phenotype @xcite .
the phenotype generally appears after several years indicating that the activation of the on@xmath1off switch is rare .
the average age at which mody is manifest could thus be dictated by the probability that a sufficient number of @xmath0-cells is locked in the off state .
we have considered the simplest form of negative autoregulation in our two - gene model .
there are recent suggestion that negative autoregulation of the hnf-4@xmath57 gene in the pancreatic @xmath0-cells may be more complex @xcite .
also , the number of transcription factor binding sites of the two genes is not known with certainty .
cooperative binding at multiple sites is expected to promote the stability of the gene expression states .
our two - gene motif constitutes a minimal model which seeks to explains the desirable features of combining a double positive feedback loop with an autorepressive loop vis - - vis the normal functional activity of @xmath0-cells . the insight gained from the model study
is expected to provide a basis for the investigation of more complex cases .
the two - gene network model .
the protein products of the @xmath19 and @xmath20 genes activate each other s synthesis .
there is also an autorepressive loop in which the proteins of the @xmath20 gene repress their own synthesis .
fig3 . @xmath52 versus @xmath54 curves showing bistability and hysteresis .
the solid ( dashed ) lines represent stable ( unstable ) steady states for gene copy numbers ( a ) @xmath131 , @xmath98 and ( b ) @xmath131 , @xmath97 .
the parameter @xmath56 , a measure of the autorepression strength , is zero .
phase portrait described by equations ( 8) and ( 9 ) .
the dark and light solid lines represent the nullclines intersecting at three fixed points .
the stable and unstable fixed points are denoted by solid and empty circles respectively .
the stable manifold divides the phase space into two basins of attraction .
some typical trajectories are shown with arrow directions denoting increasing time .
plots of @xmath55 versus @xmath54 showing regions of monostability and bistability when the parameter @xmath73 a measure of the autorepression strength , is zero ( a ) and 0.005 ( b ) .
the regions of bistability are enclosed within the red and black curves with gene copy numbers @xmath131 , @xmath97 and @xmath131 , @xmath98 respectively .
distribution of steady state @xmath20 protein levels , @xmath127 , in an ensemble of @xmath157 cells for repressor strengths ( a ) @xmath158 , ( b ) @xmath159 , ( c ) @xmath160 and ( d ) @xmath161 respectively . fig9 . for gene copy numbers @xmath162 and after a time interval @xmath154 time units of stochastic time evolution , the percentage of cells in an ensemble of @xmath157 cells in the off state ( red curve ) versus the repression strength @xmath56 with all the cells being in the on state at time @xmath153 .
the black curve shows the percentage of cells in the on state versus @xmath56 with all the cells being in the off state at @xmath153 . | we study the functional characteristics of a two - gene motif consisting of a double positive feedback loop and an autoregulatory negative feedback loop .
the motif appears in the gene regulatory network controlling the functional activity of pancreatic @xmath0-cells .
the model exhibits bistability and hysteresis in appropriate parameter regions .
the two stable steady states correspond to low ( off state ) and high ( on state ) protein levels respectively . using a deterministic approach
, we show that the region of bistability increases in extent when the copy number of one of the genes is reduced from two to one .
the negative feedback loop has the effect of reducing the size of the bistable region .
loss of a gene copy , brought about by mutations , hampers the normal functioning of the @xmath0-cells giving rise to the genetic disorder , maturity - onset diabetes of the young ( mody ) .
the diabetic phenotype makes its appearance when a sizable fraction of the @xmath0-cells is in the off state . using stochastic simulation techniques ,
we show that , on reduction of the gene copy number , there is a transition from the monostable on to the on state in the bistable region of the parameter space .
fluctuations in the protein levels , arising due to the stochastic nature of gene expression , can give rise to transitions between the on and off states .
we show that as the strength of autorepression increases , the on@xmath1off state transitions become less probable whereas the reverse transitions are more probable .
the implications of the results in the context of the occurrence of mody are pointed out .. p.a.c.s .
nos . : 87.18cf , 87.18tt , 87.18vf department of physics , bose institute , 93/1 , a. p. c road , kolkata-700009 , india |
over the last two decades , the striking optical properties of noble metal nanoparticles have raised considerable interest . among the variety of morphologies obtainable by chemical synthesis , `` nanorods '' ( elongated cylindrical or , more commonly , spherocylindrical particles ) are some of the most studied , due to their strong longitudinal resonance , which can be tuned by varying their anisotropy .
silver and gold nanorods were thoroughly studied by uv - vis - ir spectroscopy .
although numerical simulation techniques are nowadays quite accessible , see e.g. refs .
, these particles are often modeled as ellipsoids @xcite , since this shape has the great advantage of being amenable to analytical treatment , as briefly discussed in the following ; see @xcite for a detailed and up - to - date discussion .
the scattering problem was solved exactly for a sphere @xcite and for an arbitrary spheroid @xcite , but the expressions are cumbersome . in the rayleigh ( or electrostatic )
limit , analytical approximations can be derived for ellipsoids @xcite .
they are valid for particles much smaller than the wavelength , a condition often fulfilled in practice .
more elaborate but still tractable approximations exist , for spheres @xcite and spheroids @xcite beyond the rayleigh regime .
unfortunately , the currently prevailing opinion @xcite is that the ellipsoid model does not accurately describe more realistic shapes , in particular with respect to the longitudinal plasmon resonance ( that associated to the longest dimension ) .
the authors reach this conclusion by comparing the position of the longitudinal plasmon peak ( lpp ) for cylinders , spherocylinders , and prolate spheroids with the same length @xmath0 and transverse diameter @xmath1 and hence the same aspect ratio @xmath2 .
however , the literature provides no justification for using @xmath3 as the comparison parameter .
we consider the relevant criterion for identifying the ellipsoid corresponding to a given particle .
in contrast with the established procedure discussed above , we argue that the relevant quantities are the various moments of the mass distribution , leading to an effective aspect ratio @xmath4 .
the spectra of particles with different shapes : cylinder , spherocylinder and ellipsoid and the same @xmath4 agree much better than those for an equal aspect ratio @xmath3 .
the most general description of a mass distribution , for a body or a system of particles ) , is in terms of its various moments @xcite .
the first of these are , in increasing tensorial order , @xmath5 ( the total mass of the system ) , @xmath6 ( with @xmath7 the center - of - mass vector ) and the inertia tensor : @xmath8\ ] ] where @xmath9 is the volume of the body , @xmath10 is the local mass density , @xmath11 is the unit tensor and @xmath12 denotes the outer product .
@xmath13 is widely used in mechanics , but is also relevant for the interaction of particles with radiation : for instance , the gyration radius @xmath14 defined as @xmath15 is extensively used in small - angle scattering techniques @xcite . this strategy is also very similar to the traditional way of describing a charge distribution by its total charge , dipole , quadrupole and higher multipolar moments . in this framework ,
the ellipsoid equivalent to a given body is the one with the same distribution moments .
geometrically , an ellipsoid is completely defined by three parameters ( e.g. the semi - axes ) chosen such that the three eigenvalues of @xmath13 are the same as those of the initial body .
the particle position @xmath7 does not influence the absorbance ; as to the mass , we will return to it in appendix [ sec : appmass ] .
three nanoparticles ( from top to bottom : cylinder , spherocylinder and prolate ellipsoid ) with the same total length @xmath16 and diameter @xmath17 , and thus the same aspect ratio @xmath18 . for the top two shapes
, the superimposed contour represents the equivalent ellipsoid ( see text).,scaledwidth=45.0% ] we studied three types of particles : cylinders , spherocylinders and prolate ellipsoids , described by their aspect ratio @xmath19 ( exemplified in figure [ fig : geom ] for @xmath20 . ) and considered homogeneous : @xmath21 . for ellipsoids , one can also write @xmath22 ( the ratio of the two semi - axes ) .
all these shapes have azimuthal symmetry around their length , taken as parallel to @xmath23 , so that only @xmath24 and @xmath25 need to be determined .
@xmath26 @xmath27 \right . \\ & + \left
. \dfrac{4}{3 } \left [ \dfrac{83}{320 } + \left ( r - 1 + \dfrac{3}{8}\right ) ^2 \right ] \right \rbrace\\ i_z & = \rho \pi r^5 \left [ ( r-1 ) + \dfrac{8}{15}\right ] \end{array } \right.\ , \text{spherocylinder}\\ \label{eq : isphcyl}\ ] ] @xmath28 [ eq : iall ] where @xmath29 and @xmath30 are the masses of the cylinder and ellipsoid , respectively . for the spherocylinder ,
the moments of inertia are not easily expressed as a function of the total mass of the object and are therefore given in terms of the geometrical parameters and of the mass density @xmath31 .
for a given cylinder or spherocylinder we determine @xmath32 and @xmath33 as a function of @xmath34 and @xmath3 from or and identify them with , yielding the semi - axes of the equivalent ellipsoid @xmath35 and @xmath36 or , equivalently , @xmath37 and @xmath38 .
it is easily checked that @xmath4 does not depend on the radius @xmath34 : @xmath39 where @xmath40 .
we performed numerical simulations of the extinction cross - section @xmath41 of anisotropic gold nanoparticles using the discrete dipole approximation ( dda ) code ddscat 7.3 @xcite with the filtered coupled dipole method @xcite .
the refractive index of bulk gold @xmath42 is that given by johnson & christy @xcite , with no corrections for boundary dissipation . the ambient medium is water , described by a constant refractive index @xmath43 .
the particles are discretized using 90 dipoles along the diameter and @xmath44 along the length , where @xmath3 goes from 1 to 8 .
the particle radius is 5 nm .
the electric field is parallel to the long axis of the particles ( along @xmath23 ) in order to probe the lpp . in the transverse configuration ( electric field perpendicular to @xmath23 )
, the spectrum changes very little with the aspect ratio @xcite . for randomly oriented particles , e.g. in colloidal solution ,
the overall spectrum is a superposition of the longitudinal and transverse components , so that a satisfactory description of the longitudinal spectrum ensures that the total spectrum is also correctly described . to extend the results to larger particle sizes ( 10 and 20 nm radii ) we performed additional simulations using the boundary element method ( bem ) @xcite as implemented by the matlab package mnpbem @xcite .
the two methods yield very similar results , as shown in appendix [ sec : ddabem ] .
we fitted the extinction spectrum with the sum of a lorentzian peak and a cubic background to determine the position of the lpp , plotted in figure [ fig : simul ] for the three shapes , as a function of @xmath3 ( left panel ) and @xmath45 ( right panel ) .
as reference , we also added the lpp position for ellipsoids in the rayleigh limit , using the gans formula ( ref . , eq .
( 5.32 ) ) . as @xmath3 increases , so does the difference in lpp position between different shapes . when plotted against @xmath45 , however , the lpp values are much closer together and the variation is non - monotonic .
the tendency also holds for larger particles , as shown in figure [ fig : r10r20 ] for radii of 10 and 20 nm . for clarity ,
only the differences with respect to the rayleigh limit are shown , corresponding to the bottom of figure [ fig : simul ] .
the peak positions for the spherocylinder and ellipsoid are remarkably close together , while those of the cylinder exhibit a red shift which is increasingly pronounced with the aspect ratio and the radius .
this is probably due to field concentration at the sharp edges of the cylindrical particles , a localized effect that can not be captured by our simplified model .
we also checked our model against the experimental data measured for gold nanorods with varying aspect ratio in ref . .
the mean aspect ratio @xmath3 and its standard deviation @xmath46 are determined via tem and are shown in figure [ fig : ar ] as grey bars versus the sample code .
the uv - vis - ir spectroscopy curves are fitted with an ellipsoid model , yielding the distribution of _ effective _ aspect ratio .
its mean @xmath4 and standard deviation @xmath47 are shown as solid diamonds with error bars . if our model is correct , the @xmath3 values ( open dots ) obtained from the experimental @xmath4 distribution by inverting the middle relation in eq .
are the true geometrical ones and should coincide with the tem results . for most samples
this is indeed the case , confirming the improvement .
one should however keep in mind that tem only samples a very small fraction of the particles ( those deposited on grids and , among these , only those visible in the images ) in contrast with uv - vis - ir spectroscopy , which averages over all particles contained in a few milliliters of solution .
the latter technique should therefore be much more representative of the complete particle distribution .
determined by transmission electron microscopy ( tem ) ( grey bars ) and effective aspect ratio @xmath4 measured by uv - vis - ir spectroscopy ( solid diamonds ) . the geometric aspect ratio for spherocylinders @xmath3 ( open dots )
is then extracted from @xmath4 by inverting eq . .
the error bars ( uv - vis - ir ) and bar height ( tem ) indicate the mean @xmath48 standard deviation : @xmath49 or @xmath50 .
based on figure 5 of ref .
.[fig : ar],scaledwidth=45.0% ]
for ellipsoids in the rayleigh approximation , the lpp is completely described by a geometrical factor @xmath51 .
one can then define an equivalent ellipsoid of a particle by fitting its simulated spectrum with the mie - gans formula and finding the corresponding @xmath51 @xcite .
our approach here is completely different : we find the equivalent ellipsoid of the particle by applying very general principles to its mass distribution and without any consideration of its electrostatic or optical properties .
we conclude that the lpp positions of particles with the same moments of inertia ( defined in ) are significantly closer than for particles with the same aspect ratio @xmath3 . in conclusion ,
when the mass distribution is properly taken into account , the effective ellipsoid approximation describes fairly well the optical properties of elongated gold nanoparticles , in particular those of spherocylinders .
we therefore expect the results can be generalized to other rounded particles ( e.g. dumbbells @xcite ) but probably not to those exhibiting edges or tips ( e.g. prisms @xcite ) , where the field concentration is significant
. it would also be interesting to study the influence of faceting , observed in certain nanorods@xcite , on their optical spectra . combining this correction with an approximate analytical relation for the response of ellipsoids @xcite and accounting for the polydispersity in aspect ratio @xcite should yield a quantitatively accurate model for the experimental extinction spectra .
the three moments of inertia @xmath52 and @xmath33 completely define the equivalent ellipsoid for a given particle , including its volume @xmath53 and hence its mass @xmath54 . in figure [
fig : volume ] we show the ratio of @xmath53 to the volume @xmath9 of the target particle , for the cylinder and spherocylinder and for aspect ratios from 1 to 8 . for reference , we also show the same ratio for ellipsoids having the same aspect ratio @xmath3 as the target particle .
ratio of the volume @xmath53 of the equivalent ellipsoid to the volume @xmath9 of the target particle ( cylinder or spherocylinder ) , based on identifying the moments of inertia ( dotted line ) or by using the same aspect ratio ( solid line ) , compared to the exact value of 1 ( dashed line).,scaledwidth=45.0% ] clearly , the volumes are much closer when the corresponding ellipsoid is chosen based on the moments of inertia ( the discrepancy is below 3 % ) than based on the aspect ratio ( where the discrepancy can reach 33 % ) .
this result is noteworthy on two counts : first , because it comforts our choice of the moments of inertia as relevant parameters and second because the volume must be correctly described for practical applications , e.g. when estimating the particle concentration in solution from the extinction spectrum .
the dda method is more time - consuming , so we only used it for the smaller particles ( 5 nm in radius ) and employed the bem technique for the larger objects . to make sure that the results are compatible
, we also ran the bem simulations for the small particles , with the results shown in figure [ fig : ddabem ] . | we compare the light extinction spectra of elongated gold nanoparticles with different shapes ( cylinder , spherocylinder and ellipsoid ) and sizes of 10 to 100 nm . we argue that the equivalence of the various moments of mass distribution is the natural comparison criterion rather than the length - to - diameter ( aspect ) ratio generally used in the literature and that it leads to better spectral correspondence between the various shapes . |
let @xmath0 be @xmath1-dimensional minkowski space with the standard metric @xmath2 where @xmath3 is a point in @xmath0 .
let @xmath4 be a surface in @xmath0 given by @xmath5 , and @xmath6 be a function in a domain @xmath7 .
we shall assume that @xmath6 is @xmath8-smooth everywhere in @xmath9 except for a set @xmath10 consisting of isolated points only . a surface @xmath4 is _ space - like _ if the induced from @xmath11 metric is the positive definite metric .
this is equivalent to the following inequality @xmath12 where the lower index denotes a partial derivative with respect to the corresponding variable : @xmath13 , and @xmath14 .
a space - like surface @xmath4 is called _ maximal _ if the following equality holds @xmath15 everywhere in @xmath16 .
then it is well - known that ( [ first ] ) is equivalent to the vanishing of the mean curvature of @xmath4 ( with respect to its embedding in the minkowski space @xmath11 ) .
cheng and yau in @xcite proved that for every entire maximal surface @xmath4 ( i.e. the surface to be defined over the whole @xmath17 ) satisfies the bernstein property . in other words ,
an entire solution @xmath18 to ( [ second])-([first ] ) is always an affine function . on the other hand , the study of _ almost - entire _ solutions to ( [ second])-([first ] ) , i.e. solutions that they are of @xmath8 outside of a non - empty descrete set @xmath19 , is of great interest
really , there are a lot of connections between this theory and the modern physics ( see @xcite ) .
the first break - through in this direction was due to ecker @xcite , who established that the rotationally symmetric maximal surfaces ( ` maximal catenoids ' ) @xmath20 are only almost - entire solutions with @xmath19 consisting of the origin .
moreover , in the same paper it was proved that every isolated singular point of a maximal ( not necessarily entire ) surface behaves as a light cone .
namely , let @xmath21 be an isolated singular point of a solution @xmath6 which is defined at a neighborhood of @xmath22 then @xmath6 can be extended continuously at @xmath22 and @xmath23 an important characteristic of a singular point is the flux of the solution which is defined as @xmath24 where the integral is taken over a closed surface @xmath25 which encloses @xmath22 and contains no singular points , and @xmath26 denotes the outward unit normal to @xmath27 .
it follows from ( [ first ] ) that integral ( [ fourth ] ) is independent on a choice of @xmath27 .
it was shown in @xcite that @xmath28 is a lorentzian invariant of @xmath4 .
it is not hard to prove that a point @xmath29 is an essential singularity of the solution @xmath6 if and only if @xmath30 @xcite . in recent papers @xcite ,
@xcite the asymptotic behaviour and the existence questions for solutions to ( [ second])-([first ] ) were studied . in
particularly , it was shown in @xcite that under some natural geometrical assumptions on the finite set @xmath19 there exists a unique almost - entire solution @xmath6 with the prescribed fluxes @xmath31 , @xmath32 , where @xmath19 is the singular set of the solution .
on the other hand , there are no explicit examples except for the mentioned above `` maximal catenoids '' ( [ third ] ) even in the two - dimensional case . in this paper
we construct a one - parametric family of periodic almost - entire maximal surfaces whose singular set is discrete ( consists of isolated points ) and located on the fixed line .
these examples are based on the following general assertion which completely characterizes all maximal surfaces with harmonic level - sets .
let @xmath33 satisfy ( [ second])-([first ] ) and let @xmath34 , where @xmath35 is a harmonic function .
then @xmath35 is a real part of the holomorphic function @xmath36 , where @xmath37 is one of the following \(i ) @xmath38 ; \(ii ) @xmath39 ; \(iii ) @xmath40 . + here @xmath41 and @xmath42 , @xmath43 , @xmath44 . the cases ( i ) and ( ii ) lead us to the well - known examples : the plane , the rotational surface , the helicoid and sherk s maximal surfaces .
in the case ( iii ) the surface is space - like only if @xmath45 , moreover in that case it has infinitely many isolated singularity points . the explicit expression and analysis we shall give in the remained part of the paper . in fact , the examples of one - periodic maximal surfaces constructed in this paper are a small part of a bigger family of double- and one - periodic maximal two - dimensional surfaces which we treat in the forthcoming paper @xcite .
let @xmath46 be a @xmath8-function such that @xmath47 and @xmath48 then the graph of @xmath46 is a maximal surface in @xmath49 .
now we consider the function @xmath46 such that @xmath50 , where @xmath51 and @xmath52 are some twice - differentiable functions .
then ( [ ur1 ] ) can be brought to the following form : @xmath53 where @xmath54 , @xmath55 , @xmath56 .
let @xmath57 , @xmath37 be holomorphic functions , @xmath41 .
then the following identities take place @xmath58 @xmath59 here @xmath60 denotes the conjugate to @xmath61 function .
[ lemm1 ] the cauchi - riemann conditions imply @xmath62 @xmath63 we prove the validity of the first equality only : @xmath64 the second equality can be proved by the same way , hence the lemma is proved completely .
let @xmath65 , @xmath41 , where @xmath66 is a holomorphic function in the domain @xmath67 .
in order to find the coefficients @xmath19 , @xmath68 and @xmath69 of equation ( [ ur2 ] ) , from ( [ koshi ] ) we notice @xmath70 @xmath71 then @xmath72 and the equation ( [ ur2 ] ) becomes @xmath73 setting @xmath74 we find that @xmath75 the term @xmath76 in the equation ( [ chast1 ] ) depends only on @xmath65 if and only if @xmath77 where @xmath78 is a real constant .
[ lemm2 ] let @xmath79 .
we show that the condition of the functional dependence @xmath80 is equivalent to @xmath81 , @xmath82 . indeed , by virtue of ( [ koshi ] ) @xmath83 , @xmath84 and ( [ l1 ] )
, we have @xmath85 then @xmath86 @xmath87 @xmath88 and , finally , @xmath89 simplifying the last expression yields @xmath90 the latter identity holds in a non - empty domain @xmath67 , hence by the uniqueness theorem for analitic functions , there exists a real constant @xmath78 such that @xmath81 . the lemma is proved .
now we consider the differential equation ( [ mmm ] ) with a holomorphic in some domain @xmath67 function @xmath37 . one can easily show that the set of solutions of this equation makes up the following functional family : ( a ) @xmath38 ; ( b ) @xmath39 ; ( c ) @xmath91 , @xmath42 , @xmath43 , @xmath92 .
now we consider the last case , when @xmath93 . here
we have @xmath94 and , hence , @xmath95 without loss of generality , we may assume that the constant in the right hand side of the last equality is identically zero .
then @xmath96 and @xmath97 on the other hand , @xmath98 then the equation ( [ chast1 ] ) takes the form @xmath99 by solving the ordinary differential equation we arrive at @xmath100 and @xmath101 to find the admissible values of the parameter @xmath102
which correspond to the space - like examples , we check when the inequality @xmath103 holds . for this purpose
we write @xmath104 and by using a new variable @xmath105 , we obtain @xmath106 we denote @xmath113 and assume without loss of generality that @xmath114
. then @xmath115 @xmath116 where @xmath117 and @xmath118 .
let us introduce @xmath119 .
then @xmath120 i.e. by means of the jacobi elliptic sinus , we find @xmath121 let @xmath124 and @xmath125 .
then the surface @xmath126 given implicitly by @xmath127 is a maximal surface in @xmath128 .
moreover , this surface is a graph of a real analytic function everywhere except for the set consisting of singular points @xmath129 we observe that for different values of @xmath130 , the surfaces @xmath126 are lorentz non - isometric .
one can also see that @xmath126 is located in the parallel slab @xmath131 , where @xmath132 is the complete elliptic integral of the first kind ( the least positive solution of equation @xmath133 ) .
moreover , one can show that the flux @xmath134 at the singular point @xmath135 is equal to @xmath136 | in the present paper we study two - dimensional maximal surfaces with harmonic level - sets . as a corollary
we obtain a new class of one - periodic maximal surfaces . |
the seyfert 1 galaxy ngc 7469 is one of a handful of bright , nearby active galactic nuclei ( agn ) that has been intensively studied at multiple wavelengths and monitored extensively in the x - ray , uv , and optical bands ( @xcite ; @xcite ; @xcite ; @xcite ) .
asca x - ray spectra ( @xcite ; @xcite ) were modeled with absorption edges of o vii and o viii indicative of a warm absorber , and _
hubble space telescope _
faint object spectrograph observations show associated uv absorption lines in , and .
such x - ray and uv absorption is common in seyfert 1s , occurring in tandem in roughly half the population ( @xcite ; @xcite ; @xcite ) .
the x - ray and uv absorbing gas components thus appear to be related , but it is not yet clear whether this is the consequence of a more general phenomenon linking the gas visible at both wavelengths , or if the absorption arises in the same gas with identical physical conditions in both the x - ray and the uv . at high resolution ,
both the x - ray and the uv absorption is kinematically complex ( @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) . generally only some of the uv absorption lines have velocities and physical conditions similar to those seen in the x - ray . the locations of both the x - ray and the uv absorbing media are also highly uncertain .
if the outflowing gas originates in the accretion disk , then it may be relatively close to the central engine ( e.g. , @xcite ; @xcite ) and interior to the broad emission - line region .
on the other hand , if the absorbing gas is a thermal wind driven off the obscuring torus ( @xcite ; @xcite ) , then it will lie at distances of roughly 1 pc , be external to the broad - line region , but probably interior to the narrow - line region .
studies of the x - ray absorption variability ( ngc 3516 ; @xcite ) show that at least some of the gas does lie at distances approaching 1 pc .
variability of uv absorption in ngc 4151 suggests distances of tens of parsecs for the absorbing gas ( @xcite ) , and even greater distances for the associated uv absorption seen in some quasars ( @xcite ; @xcite ) . using the _ far ultraviolet spectroscopic explorer ( fuse ) _ , we obtained high - resolution uv spectra below 1200 covering the o vi @xmath1 resonance doublet and the high - order lyman lines .
the o vi doublet provides a key link to the x - ray band since high energy transitions from the same ion can be viewed with _
xmm - newton_. this permits us to directly compare the kinematics and the column densities measured in both wavelength regions as a crucial test for establishing the link between x - ray and uv absorbing gas in seyferts .
_ xmm - newton _ observations of ngc 7469 are discussed in a companion paper ( @xcite ) .
observes the far - ultraviolet wavelength range from 9121187 using four independent optical channels . in each channel
a primary mirror gathers light for a rowland - circle spectrograph .
two - dimensional photon - counting detectors record the dispersed spectra .
two of the optical trains use lif coatings on the optics to cover the 9901187 wavelength range .
the other two channels cover shorter wavelengths down to 912 using sic - coated optics .
see @xcite for a full description of and its in - flight performance .
we observed ngc 7469 with on 1999 december 6 through the 30-square low - resolution apertures .
data were obtained on 22 consecutive orbits for a total integration time of 37,803 s. unfortunately the sic channels were not properly aligned ; the only detectable flux was in the lif channels .
the time - tagged data were processed using v1.8.7 of the calibration pipeline . @xcite
describe the standard data processing steps .
we added an additional customized step to normalize and subtract a flat background image from each detector segment .
the resulting extracted spectra were merged into a linearized spectrum with 0.05 bins .
while this rebinning process introduces some correlated error , since each of these bins holds @xmath7 original pixels , the resulting bins are more than 80% statistically independent .
poisson errors and data quality flags are propagated through the data reduction process along with the science data . to firmly establish the zero - point of our wavelength scale , we compare the positions of low - ionization galactic absorption lines from species such as ar i , fe ii , o i and @xmath8 to the galactic 21 cm velocity as measured by murphy et al .
( unpublished ) .
this requires a shift of @xmath9 to be applied to our wavelengths to place them in a heliocentric reference frame .
we estimate that the flux scale is accurate to @xmath1010% , and that wavelengths are accurate to @xmath1015 .
the full merged spectrum is shown in the top panel of fig .
[ h4204f1.ps ] .
numerous galactic absorption features obscure much of the spectrum , so we have constructed a simple model of the interstellar absorption to remove them from the spectrum .
we assume a galactic column of @xmath11 at a heliocentric velocity of @xmath12 with a doppler width of 10 .
heavy elements are included at solar abundances following the depletion pattern typical of warm gas toward @xmath13 oph ( @xcite ) .
matching the molecular hydrogen absorption requires one component with gas at 50 k and a @xmath8 column density of @xmath14 , and a second component with a temperature of 300 k and a total column density of @xmath15 .
we show the spectrum of ngc 7469 corrected for this foreground absorption in the lower panel of fig .
[ h4204f1.ps ] .
strong , broad o vi emission dominates the spectrum .
weaker broad emission lines of @xmath16977 , @xmath16991 , and he ii @xmath17 are also visible , as well as several unidentified emission features . the low broad bump on the red wing of o vi is also present in the spectrum of mrk 509 ( @xcite ) as well as in the spectra of low - redshift quasars observed using the _ hubble space telescope _
( @xcite ) . based on observations of a large sample of low - redshift agn , kriss et al .
( 2003 ) have identified this as blended emission due to s iv @xmath18 . at the high resolution of ,
absorption features intrinsic to ngc 7469 are also discernible .
[ h4204f2.ps ] shows the and o vi region of the spectrum in 0.05 bins . here
one can see two absorption systems due to the o vi @xmath191032,1038 resonance doublet in the blue wing of the broad o vi emission line .
the absorption associated with these systems is weak , if present at all , and severely blended with foreground galactic absorption .
similarly , no or @xmath16977 absorption is visible at shorter wavelengths .
to measure the emission and absorption line properties of ngc 7469 , we model the spectrum using the iraf task specfit ( @xcite ) .
we use a reddened power law in @xmath20 to describe the continuum emission . for extinction
, we use a @xcite curve with @xmath21 and e(b@xmath22v)=0.069 ( @xcite ) . to model the continuum
, we use regions of the spectrum free of both emission and absorption lines covering the full observed wavelength range . the best - fit , extinction - corrected continuum has the form @xmath23 .
to model the o vi emission and absorption , we fix the continuum parameters as determined above and restrict our subsequent fits to the 1033.01060.3 wavelength range . as one can see from the models shown in fig .
[ h4204f2.ps ] , the o vi emission line profile requires both broad and narrow components .
we model these with gaussian line profiles .
the narrow component is only cleanly visible near the peak of the red component of the o vi doublet ; strong foreground galactic @xmath8 absorption obscures the corresponding feature near the blue peak .
although we can only clearly see part of the red portion of the narrow doublet , our baseline model assumes that they are optically thin with a 2:1 intensity ratio .
( tests using other assumptions are described in our discussion of the absorption line fits below . )
the broad component is also treated as a doublet with two blended gaussians having a 2:1 intensity ratio . for each narrow and broad doublet , we require the widths to be identical and the wavelengths to have the same ratio as their laboratory values .
the @xmath24 , @xmath25 , s iv @xmath18 , and he ii @xmath17 emission lines are all modeled as single gaussian emission lines .
the two components of the s iv doublet have identical widths , a fixed flux ratio of 1:1 , and wavelengths fixed at the ratio of their laboratory value . in addition , we include a single gaussian emission line at the location of the unidentified emission feature at 1031 .
table [ tbl - elines ] gives the best - fit parameters for these emission lines .
lcccccc line & @xmath26 & flux@xmath27 & velocity@xmath28 & fwhm + & ( ) & & @xmath29 & @xmath29 + c iii & 977.02 & @xmath30 & @xmath31 & @xmath32 + n iii & 991.58 & @xmath33 & @xmath34 & @xmath32 + unknown & 1014.86 & @xmath35 & @xmath36 & @xmath37 + o vi broad & 1031.93 & @xmath38 & @xmath39 & @xmath40 + o vi broad & 1037.62 & @xmath41 & @xmath39 & @xmath40 + o vi narrow & 1031.93 & @xmath42 & @xmath43 & @xmath44 + o vi narrow & 1037.62 & @xmath45 & @xmath43 & @xmath44 + s iv & 1062.66 & @xmath46 & @xmath47 & @xmath48 + s iv & 1072.97 & @xmath46 & @xmath47 & @xmath48 + heii & 1085.15 & @xmath49 & @xmath50@xmath51 & @xmath52 + observed flux in units of @xmath53 .
velocity is relative to a systemic redshift of @xmath54 @xcite .
we model absorption lines in the ngc 7469 spectrum using voigt profiles . for the intrinsic absorbers
, we permit the absorption to partially cover the source .
that is , for a covering fraction @xmath55 , the transmittance at a given wavelength , @xmath56 , has the form @xmath57 , where @xmath58 is the optical depth at that wavelength .
the intrinsic absorption lines are clustered into two complexes that we designate as # 1 ( at a relative systemic velocity of @xmath59 ) and # 2 ( at velocity @xmath60 ) .
each of these complexes is modeled with four separate blended absorption lines .
the o vi lines in each component are treated as doublets with their wavelengths fixed at the ratio of their laboratory values , their optical depths fixed at a 2:1 ratio , and their velocity widths forced to be the same .
for the corresponding lines , we fix their wavelengths at the ratio of their laboratory values to those of the o vi lines and force them to have the same velocity widths .
covering fractions vary freely within each complex .
the individual o vi lines within a complex are forced to have the same velocity widths and the same covering fractions .
( this assumption is relaxed below in an alternative model for the absorption . ) since absorption is not even unambiguously detectable , we fix the relative optical depths of the lines within each complex to be the same as the ratios observed for o vi . however , we permit the overall optical depth for each complex to vary freely
. table [ tbl - alines ] gives the best - fit parameters for this model of the absorption lines .
parameters in the table with error bars of zero had their values linked to another parameter in the fit . lccccccc feature & comp & @xmath61 & @xmath62 & @xmath63 & @xmath64 & fwhm & @xmath55 + & # & ( ) & ( ) & @xmath65 & ( @xmath66 ) & ( @xmath66 ) & + ly@xmath67 & 2a & 1025.72 & @xmath680.05 & @xmath69 & @xmath70 & @xmath71 & @xmath72 + ly@xmath67 & 2a & 1025.72 & & @xmath73 & @xmath74 & @xmath71 & @xmath72 + ly@xmath67 & 2a & 1025.72 & & @xmath75 & @xmath76 & @xmath71 & @xmath72 + ly@xmath67 & 2a & 1025.72 & & @xmath77 & @xmath78 & @xmath71 & @xmath72 + ly@xmath67 & 1a & 1025.72 & @xmath680.03 & @xmath79 & @xmath50@xmath80 & @xmath81 & @xmath82 + ly@xmath67 & 1b & 1025.72 & & @xmath83 & @xmath50@xmath84 & @xmath81 & @xmath82 + ly@xmath67 & 1c & 1025.72 & & @xmath85 & @xmath50@xmath86 & @xmath81 & @xmath82 + ly@xmath67 & 1d & 1025.72 & & @xmath87 & @xmath50@xmath88 & @xmath81 & @xmath82 + ovi & 2a & 1031.93 & @xmath89 & @xmath90 & @xmath91 & @xmath71 & @xmath72 + ovi & 2b & 1031.93 & & @xmath92 & @xmath93 & @xmath71 & @xmath72 + ovi & 2c & 1031.93 & & @xmath94 & @xmath95 & @xmath71 & @xmath72 + ovi & 2d & 1031.93 & & @xmath96 & @xmath97 & @xmath71 & @xmath72 + ovi & 1a & 1031.93 & @xmath98 & @xmath99 & @xmath50@xmath100 & @xmath81 & @xmath82 + ovi & 1b & 1031.93 & & @xmath101 & @xmath50@xmath84 & @xmath81 & @xmath82 + ovi & 1c & 1031.93 & & @xmath102 & @xmath50@xmath103 & @xmath81 & @xmath82 + ovi & 1d & 1031.93 & & @xmath104 & @xmath50@xmath105 & @xmath81 & @xmath82 + equivalent widths are integrated over all components of each spectral feature .
upper limits are @xmath106 .
velocity is relative to a systemic redshift of @xmath54 @xcite .
since much of the peak of the o vi emission - line profile is obscured by foreground galactic absorption , we have explored the effects that this uncertainty in the true emission - line profile might have on our characterization of the absorption lines .
we consider a wide range of models : a. this is our baseline model , described in detail above .
the narrow emission lines have line ratios fixed at an optically thin value of 2:1 , and we assume that this narrow line emission is obscured at the same covering fraction as the continuum and broad emission line .
individual o vi lines within a complex have the same velocity widths and covering fractions .
b. in this model the narrow emission lines are treated the same as in a " , but the velocity widths and the covering fractions of individual o vilines within a complex are all permitted to vary freely . in these fits ,
the strongest , deepest absorption line in each complex has parameters typical of those found in our baseline model .
the weaker adjacent lines retain similar widths in these fits , but their best - fit covering fractions do change .
however , the individual covering fractions are not well constrained .
their error bars are typically @xmath107% , and the new values lie within one error bar of the best - fit value in the baseline model . c. this variation has the narrow emission line fluxes fixed at an optically thick ratio of 1:1 .
the emission is obscured with the same covering fraction as the continuum and broad emission line .
as in the baseline model , individual o vi lines within a complex have the same velocity widths and covering fractions . in a pattern that repeats with each of the subsequent variations below , the best - fit o vi column densities in this model are slightly lower , but not more than twice the values of the error bars shown in table [ tbl - alines ] .
while this model significantly changes the overall emission - line profile , since the o vi absorption is so highly blueshifted , little of the intrinsic absorption falls in the region of the line profile with the most dramatic changes .
thus the overall optical depth of the absorption lines is roughly the same as in the baseline model .
d. we also considered a model with no narrow emission - line component .
as in the baseline model , individual o vi lines within a complex have the same velocity widths and covering fractions .
this model provides a significantly worse fit than the baseline , but the measured o vi column densities and covering fractions are the same as the baseline model to within the error bars in table [ tbl - alines ] .
e. this model has narrow - line emission with line ratios fixed at 2:1 , but here we assume that the intrinsic absorption does not obscure the narrow - line emission at all .
as in the baseline model , individual o vi lines within a complex have the same velocity widths and covering fractions . again the o vi column densities are lower , but they are not significantly different from those of the baseline model .
f. in this model we also assume that the intrinsic absorption does not absorb the narrow - line emission , but we fix that emission at a ratio of 1:1 . as in the baseline model , individual o vi lines within a complex have the same velocity widths and covering fractions .
again we find no significant variation from the results of our baseline model .
the overall properties of these various emission - line profile models are summarized in table [ tbl - models1 ] . for the o vi components 1 and 2 we list the total column density from the fit , the velocity of the strongest feature in each component , the best - fit widths of the absorption lines , and the best - fit covering fractions . the comparative summary for component # 1
is listed first in the table , followed by the summary for component # 2 .
l c c c c c model / component & @xmath108 & @xmath109 & @xmath110 & fwhm & @xmath55 + & & @xmath111 & ( @xmath66 ) & ( @xmath66 ) & + a/1 & 604.0/344 & @xmath112 & @xmath50@xmath113 & @xmath81 & @xmath82 + b/1 & 596.9/332 & @xmath114 & @xmath50@xmath115 & @xmath116 & @xmath117 + c/1 & 613.1/344 & @xmath118 & @xmath50@xmath113 & @xmath119 & @xmath120 + d/1 & 747.5/344 & @xmath121 & @xmath50@xmath122 & @xmath119 & @xmath123 + e/1 & 622.1/344 & @xmath124 & @xmath50@xmath125 & @xmath81 & @xmath126 + f/1 & 616.1/344 & @xmath127 & @xmath50@xmath113 & @xmath119 & @xmath128 + a/2 & 604.0/344 & @xmath129 & @xmath130 & @xmath131 & @xmath72 + b/2 & 596.9/332 & @xmath132 & @xmath133 & @xmath134 & @xmath72 + c/2 & 613.1/344 & @xmath135 & @xmath130 & @xmath131 & @xmath136 + d/2 & 747.5/344 & @xmath137 & @xmath130 & @xmath131 & @xmath138 + e/2 & 622.1/344 & @xmath139 & @xmath140 & @xmath141 & @xmath142 + f/2 & 616.1/344 & @xmath143 & @xmath140 & @xmath144 & @xmath142 + velocity is relative to a systemic redshift of @xmath54 @xcite .
[ h4204f2.ps ] graphically illustrates our best fit to the o viand region of the spectrum . using our model for the underlying emission and the foreground galactic absorption ,
we divide this into the data to produce a normalized spectrum that is corrected for the effects of foreground galactic absorption .
this normalized spectrum is shown in velocity space surrounding the two absorption line complexes in o vi and in fig .
[ h4204f3.ps ] .
since the relative oscillator strengths of the blue and red lines in the o vi doublet have roughly a 2:1 ratio , one would expect the red line to have a depth that is half that of the blue line in optically thin gas , and an equal depth in optically thick gas .
from the normalized spectrum it is readily apparent that the equal depths of the red and blue lines in component # 1 imply that the absorption is heavily saturated .
however , since the line cores are not black at their centers , we also can see that the absorbing gas does not fully cover the underlying emission .
the covering fraction is only @xmath5% .
our fits may underestimate the total column density in this absorption complex .
the lack of visible damping wings on the line profile permits us to set an upper limit of @xmath145 on the o vi column of this component .
this upper limit is reflected in the error bar that we assign to the strongest line in component # 1 in table [ tbl - alines ] .
the relative depths of the red and blue lines in component # 2 are not quite equal , but they also are not at an optically thin ratio .
this component is approaching saturation , but is not yet optically thick , and so its total column density is better determined .
as typical for seyfert 1 galaxies viewed at high spectral resolution in the far - ultraviolet , we see complex , multiple absorption lines from highly ionized o vi . although absorption in the spectrum is weak or absent , there are absorption lines in the fos spectrum of ngc 7469 obtained in 1996 at systemic velocities of @xmath146 and @xmath147 ( @xcite ) , very close to the velocities of components # 1 and # 2 , so there is some minimal neutral hydrogen column .
the high o vi to column - density ratios for both components imply very highly ionized gas consistent with those typical of x - ray warm absorbers and similar to the highest ionization component in mrk 509 ( @xcite ) .
however , only component # 1 has velocities comparable to those of higher ionization states of o vii and o viii ( @xmath148 ) as seen in the _ xmm - newton _ grating spectrum of ngc 7469 ( @xcite ) .
the upper limit on the o vi column density measured in the _ xmm - newton _ grating spectrum of @xmath149 is consistent with the column densities measured in the spectrum .
l c c c c c c c @xmath150 & @xmath151 & @xmath152 & @xmath153 & @xmath154 & @xmath155 & @xmath156 & @xmath157 + & @xmath158 & @xmath158 & @xmath158 & @xmath158 + 6.0@xmath27 & 20.55 & @xmath159 & @xmath160 & @xmath161 & @xmath162 & @xmath163 & @xmath164 + & observed : & @xmath165 & @xmath166 & @xmath167 & @xmath168 & & + 0.2@xmath28 & 18.58 &
@xmath169 & @xmath170 & @xmath171 & @xmath172 & @xmath173 & @xmath174 + & observed : & @xmath175 & @xmath176 & @xmath177 & @xmath178 & &
+ 0.08@xmath179 & 18.58 & @xmath180 & @xmath181 & @xmath182 & @xmath183 & @xmath184 & @xmath185 + this model is our best match to component # 1 .
this model is our best match to component # 2 .
this model corresponds to component # 2 at the epoch of the 1996 fos observations . for a doppler width of 100 ,
the blue line in the doublet would have an equivalent width of 0.20 . for a doppler width of 100 , the blue line in the doublet would have an equivalent width of 0.16 . to test whether the same gas is responsible for both the uv and the x - ray absorption in component # 1
, we computed photoionization models similar to those used by krolik & kriss ( 1995 , 2001 ) .
these models cover a grid in total column density ranging from @xmath186 to @xmath187 , and in ionization parameter from @xmath188 to 10.0 , where @xmath150 is the ratio of ionizing photons in the lyman continuum to the local electron density .
for the illuminating spectrum in our models , we used two different spectral energy distributions .
the first , sed1 , is the same as that described by kriss et al .
( 2000a ) , which is based on the simultaneous iue , fos , and rxte observations of ngc 7469 performed in 1996 .
we show sed1 as a solid line in fig .
[ h4204f4.ps ] .
the second spectral energy distribution , sed2 , is based on the current spectrum presented in this paper and the _ xmm - newton _ spectrum from blustin et al .
( 2003 ) . at long wavelengths we assume @xmath189 , as in sed1 . at 2500 ,
this breaks to @xmath190 to match the spectrum . at x - ray energies , we use the energy index of 0.7 from the _ xmm - newton _ spectrum to span the 0.5 kev to 100 kev . at higher energies we use a steeper index of 2.0 to prevent divergence in the total ionizing flux . from 50 ev to 0.5 kev , the steep @xmath191 is a reasonable match to the soft excess seen in the _ xmm - newton _ data .
the and _ xmm - newton _ data were not obtained simultaneously , and so we normalize the x - ray continuum relative to the uv continuum by choosing @xmath192 , the same value used by kriss et al .
( 2000a ) based on the 1996 simultaneous observations .
sed2 is the dashed line shown in fig .
[ h4204f4.ps ] .
for comparison , we show the absorption - corrected uv and x - ray fluxes at 2120 and 2 kev , respectively , as observed with _
xmm - newton _
( @xcite ) .
these data have @xmath193 , and they are consistent with sed2 . after examining the results of our modeling
, we find that none of the models using sed1 achieves sufficiently high ratios of o vi , o vii , and o viii relative to to match the values observed in our spectra .
for sed2 , we do find a solution that is compatible with our observations .
this model has u = 6.0 and log @xmath194 = 20.55 .
the predicted column densities for ions of interest are summarized in table [ tbl - models2 ] , where we compare them to the observed column densities .
note that and are virtually invisible in this model , and , indeed , no absorption from these ions was seen by kriss et al .
( 2000a ) in the 1996 fos spectrum at the velocity of component # 1 . for component # 2 ,
the required column density and ionization level is substantially lower .
we obtain a reasonable match to the observed o vi and column densities for @xmath195 and @xmath194 = 18.58 , as also shown in table [ tbl - models2 ] .
the predicted column densities for and in this model are also rather low , although easily detectable .
we note that the uv continuum in the current observation is a factor of @xmath196 brighter than that observed with the fos in 1996 .
if we assume that any variations since then are simply due to changes in flux with a constant column density of absorbing gas , then we would have expected the gas in 1996 to be responding to an ionization parameter of @xmath197 .
this would then predict and column densities of @xmath198 and @xmath199 , respectively , with corresponding equivalent widths of the blue components of each doublet ( for optically thin gas ) of 0.2 and 0.16 .
while these are still lower than the observed values in the fos spectrum , ( @xmath200 and @xmath201 for and , respectively ) given the uncertainties in the spectral energy distribution and other potentially variable factors , we find the agreement quite good .
the covering fraction and line depth of component # 1 is consistent with its absorbing only the continuum flux and none of the emission - line flux in ngc 7469 .
its high ionization parameter , and its likely identification with the same gas responsible for the x - ray absorption suggests that it is much higher ionization than component # 2 . taken together , these points imply that component # 1 lies very close to the central engine , possible interior to the broad emission line region .
this would favor its association with an accretion disk wind , rather than a thermally driven wind emanating from the obscuring torus .
in constrast , the high covering fraction of component # 2 clearly places it exterior to both the blr and continuum regions as would be expected for the thermally driven winds of krolik & kriss ( 1995 ; 2001 ) .
overall , in this one object we may be seeing absorption from both a torus wind and a disk wind .
our high - resolution far - ultraviolet spectrum of ngc 7469 obtained with shows broad emission lines of , , o vi , and he ii , as well as possible emission from s iv @xmath18 .
intrinsic absorption in the o vi @xmath1 resonance doublet arises in two distinct kinematic components at systemic velocities of @xmath2 ( component # 1 ) and @xmath3 ( component # 2 ) .
both components are very highly ionized with no significant absorption detected at either velocity .
component # 2 , although highly ionized , has a lower total column density than component # 1 , and it is consistent with having no associated x - ray absorption .
it covers more than 90% of both the continuum and broad - line emission .
component # 1 at @xmath2 is the best match in velocity to the highly ionized x - ray absorbing gas detected in the _ xmm - newton _ grating spectrum of ngc 7469 ( @xcite ) at a blueshift of @xmath202 .
photoionization models for component # 1 show that for a total column density of @xmath203 and an ionization parameter of @xmath204 , the column densities of , o vi , o vii , and o viii in the and the _ xmm - newton _ spectra can all be reproduced .
component # 1 also has an extraordinarily low covering fraction of 0.5 , and is consistent with covering only the continuum emission and none of the broad - line emission .
this suggests that it might arise in an accretion disk wind interior to the broad line region .
this work is based on data obtained for the guaranteed time team by the nasa - cnes - csa mission operated by the johns hopkins university .
financial support to u. s. participants has been provided by nasa contract nas5 - 32985 .
the u. k. authors acknowledge the support of the particle physics and astronomy research council .
g. kriss acknowledges additional support from nasa long term space astrophysics grant nagw-4443 . | we obtained far - ultraviolet spectra of the seyfert 1 galaxy ngc 7469 using the _ far ultraviolet spectroscopic explorer _ on 1999 december 6 .
our spectra cover the wavelength range 9901187 with a resolution of @xmath0 .
we see broad emission lines of , , o vi , and he ii as well as intrinsic absorption lines in the o vi @xmath1 resonance doublet .
the absorption arises in two distinct kinematic components at systemic velocities of @xmath2 and @xmath3 .
both components are very highly ionized no significant absorption is present .
the higher blueshift component is not quite saturated , and it has a total o vi column density of @xmath4 .
it covers more than 90% of the continuum and broad - line emission .
the lower blueshift component is heavily saturated and covers only @xmath5% of the continuum and broad - line emission .
it too has a column density of @xmath4 , but this is less certain due to the high saturation .
we set an upper limit of @xmath6 on the o vicolumn density of this component .
its line depth is consistent with coverage of only the continuum , and thus this component may lie interior to the broad emission - line gas .
the component at @xmath2 has a velocity comparable to the high - ionization x - ray absorption lines seen in the _ xmm - newton _ grating spectrum of ngc 7469 , and photoionization models show that the observed column densities of o vi and are compatible with their formation in the same gas as that causing the x - ray absorption .
the gas at @xmath3 has lower ionization and column density , and no significant x - ray absorption is associated with it . |
matrix completion amounts to estimate all the entries in a matrix @xmath0 from a partial and potentially noisy observation @xmath1 where @xmath2 is a binary matrix with 1/0 entries masking a portion of @xmath3 through the element - wise multiplication @xmath4 , and @xmath5 an additive noise . with matrix rows , columns , and entry values assigned to various attributes , matrix completion may have numerous applications .
for example , when the rows and the columns of @xmath3 are attributed to users and items such as movies and books , and an entry at position @xmath6 records a score given by the @xmath7-th user to the @xmath8-th item , matrix completion predicts users scores on the items they have not yet rated , based on the available scores recorded in @xmath9 , so that personalized item recommendation system becomes possible .
this is a classical problem in collaborative filtering . to solve an ill - posed matrix completion problem
, one must rely on some prior information , or in other words , some data model .
the most popular family of approaches in the literature assumes that the matrix @xmath3 follows approximately a low rank model , and calculates the matrix completion with a matrix factorization @xcite .
theoretical results regarding the completion of low - rank matrices have been recently obtained as well , e.g. , @xcite and references therein .
more elaborative probabilistic models and some refinement have been further studied on top of matrix factorization , leading to state - of - the - art results @xcite . in image processing , assuming that local image patches follow gaussian mixture models ( gmm ) , yu , sapiro and mallat have recently reported excellent results in a number of inverse problems @xcite . in particular , for inpainting , which is an analogue to matrix completion where the data is an image , the maximum a posteriori expectation - maximization ( map - em ) algorithm for solving the gmm leads to state - of - the - art results , with a computational complexity considerably reduced with respect to the previous state - of - the - art approaches based on sparse models @xcite , ( which are analogous to the low - rank model assumption for matrices ) . in this paper
, we investigate gaussian modeling ( a particular case of gmm with only a single gaussian distribution ) for matrix completion .
subparts of the matrix , typically rows or columns , are regarded as a collection of signals that are assumed to follow a gaussian distribution .
an efficient map - em algorithm is introduced to estimate both the gaussian parameters ( mean and covariance ) and the signals .
we show through numerical experiments that the fast map - em algorithm , based on the gaussian model which is the simplest probabilistic model one can imagine , leads to results in the same ballpark as the state - of - the - art in movie rating prediction , at significantly lower computational cost .
recent theoretical results @xcite further support the consideration of gaussian models for the recovery of missing data .
section [ sec : model : algo ] introduces the gaussian model and the map - em algorithm . after presenting the numerical experiments in section [ sec : numeric ]
, section [ sec : conclusion ] concludes with some discussions .
similar to the local patch decomposition often applied in image processing @xcite , let us consider each subpart of the matrix @xmath0 , the @xmath10-th row @xmath11 for example , as a signal .
let @xmath12 denote the @xmath10-th row in the binary matrix @xmath13 , and let @xmath14 be the number of non - zero entries in @xmath15 .
let @xmath16 denote a masking operator which maps from @xmath17 to @xmath18 extracting entries of @xmath19 corresponding to the non - zero entries of @xmath15 , i.e. , all but the @xmath20-th the entries in the @xmath21-th row of @xmath22 are zero , with @xmath20 the index of the @xmath21-th non - zero entry in @xmath15 , @xmath23 .
let @xmath24 and @xmath25 be respectively the _ sub - vector _ of the @xmath10-th row of @xmath9 and @xmath26 , where the entries of @xmath15 are non - zero . with this notation , we can rewrite in a more general linear model @xmath27 for all the signals @xmath19 , @xmath28 . in the reduced dimension
@xmath29 leads to a calculation in dimension @xmath29 instead of @xmath30 , which is considerably faster if @xmath31 . ]
note that @xmath19 can also be columns , or 2d sub - matrices of @xmath3 rendered in 1d .
the gaussian model assumes that each signal @xmath19 is drawn from a gaussian distribution , with a probability density function @xmath32 where @xmath33 and @xmath34 are the unknown covariance and mean parameters of the gaussian distribution .
the noise @xmath35 is assumed to follow a gaussian distribution with zero mean and covariance @xmath36 , here assumed to be known ( or calibrated ) . estimating the matrix @xmath3 from the partial observation @xmath9 can thus be casted into the following problem : * estimate the gaussian parameters @xmath37 , from the observation @xmath38 . * estimate @xmath19 from @xmath39 , @xmath28 , using the gaussian distribution @xmath40 .
since this is a non - convex problem , we present an efficient maximum a posteriori expectation - maximization ( map - em ) algorithm that calculates a local - minimum solution @xcite . following a simple initialization , addressed in section [ subsec : initialization ] ,
the map - em algorithm is an iterative procedure that alternates between two steps : 1 .
e - step : signal estimation . assuming the estimates @xmath41 are known ( following the previous m - step ) , for each @xmath10 one computes the maximum a posteriori ( map ) estimate @xmath42 of @xmath19 .
m - step : model estimation . assuming the signal estimates @xmath42 , @xmath43 ,
are known ( following the previous e - step ) , one estimates ( updates ) the gaussian model parameters @xmath44 .
it is well known that under the gaussian models assumed in section [ subsec : gaussian : model ] , the map estimate that maximizes the log a - posteriori probability @xmath45is a linear estimator and is optimal in the sense that it minimizes the mean square error ( mse ) , i.e. , @xmath46 = \min_{g}e_{{\mathbf{f}}_i , { \mathbf{w}}_i } [ \|{\mathbf{f}}_i- g({{\mathbf{y}}}_i)\|_2 ^ 2 ] , \nonumber $ ] as well as the mean absolute error ( mae ) , i.e. , @xmath47 = \min_{g } e_{{\mathbf{f}}_i , { \mathbf{w}}_i } [ \|{\mathbf{f}}_i - g({{\mathbf{y}}}_i)\|_1 ] , \nonumber $ ] where @xmath48 is any mapping from @xmath49 @xcite . the second equality of follows from the bayes rule , the third follows from the gaussian models @xmath50 and @xmath51 , and the last is obtained by deriving the third line with respect to @xmath19 .
the close - form map estimate can be calculated fast .
observe that @xmath16 is a sparse extraction matrix , each row containing only one non - zero entry with value @xmath52 , whose index corresponds to the non - zero entry in @xmath15 .
therefore , the multiplication operations that involve @xmath22 or @xmath53 can be realized by extracting the appropriate rows or columns at zero computational cost .
the complexity of is therefore dominated by the matrix inversion . as @xmath54 is positive - definite , @xmath55 can be implemented with @xmath56 flops through a cholesky factorization @xcite . in a typical case where
@xmath19 is the @xmath10-th row of the matrix @xmath3 , @xmath28 , to estimate @xmath3 the total complexity of the e - step is therefore dominated by @xmath57 flops . for typical rating prediction datasets that are highly sparse , among a large number of items @xmath30 , most users have rated more or less only a small number @xmath58 of items , where @xmath59 is large .
the total complexity of the e - step is thus dominated by @xmath60 flops .
the parameters of the model are estimated / updated with the maximum likelihood estimate , @xmath61 with the gaussian model @xmath50 , it is well known that the resulting estimate is the empirical one , @xmath62 the empirical covariance estimate may be improved through regularization when there is lack of data ( let us take an example of standard rating prediction , where there are @xmath63 items and @xmath64 users , the dimension of the covariance matrix @xmath65 is @xmath66 ) .
a simple and standard eigenvalue - based regularization is used here , @xmath67 where @xmath68 is a small constant .
the regularization also guarantees that the estimate @xmath69 of the covariance matrix is full - rank , so that is always well defined . to estimate @xmath3 , the computational complexity of the m - step
is dominated by the calculation of the empirical covariance estimate requiring @xmath70 flops , which is negligible with respect to the e - step . as the map - em algorithm iterates , the map probability of the observed signals @xmath71 increases .
this can be observed by interpreting the e- and m - steps as a coordinate descent optimization @xcite . in the experiments ,
the algorithm converges within 10 iterations .
the map - em algorithm is initialized with an initial guess of @xmath19 , @xmath43 .
the experiments show that the result is insensitive to the initialization for movie rating prediction . in the numerical experiments ,
all the unknown entries are initialized to 3 for datasets containing ratings ranging from 1 to 5 or 6 . in a typical case where the matrix row decomposition in is considered , the overall computational complexity of the map - em to estimate an @xmath72 matrix
is dominated by @xmath73 , with @xmath74 the number of iterations ( typically @xmath75 ) and @xmath76 the available data ratio , with @xmath59 typically large ( @xmath77 for the standard movie ratings data ) .
the algorithm is thus very fast . as each row @xmath19
is treated as a signal and the signals can be estimated in sequence , the memory requirement is dominated by @xmath78 ( to store the covariance matrix @xmath33 ) .
the experimental protocols strictly follow those described in the literature @xcite .
the proposed method is evaluated on two movie rating prediction benchmark datasets , the _ eachmovie _ dataset and the _ 1 m movielens _ dataset .
two test protocols , the so - called `` weak generalization '' and `` strong generalization , '' @xcite are applied . * * weak generalization * measures the ability of a method to generalize to other items rated by the _
users used for training the method .
one known rating is randomly held out from each user s rating set to form a test set , the rest known ratings form a training set . the method is trained using the data in the training set , and its performance is evaluated over the test set . * * strong generalization * measures the ability of the method to generalize to some items rated by _ novel _ users that have _ not _ been used for training .
the set of users is first divided into training users and test users .
learning is performed with all available ratings from the training set . to test the method ,
the ratings of each test users are further split into an observed set and a held - out set .
the method is shown the observed ratings , and is evaluated by predicting the held - out ratings .
the eachmovie dataset contains 2.8 million ratings in the range @xmath79 for 1,648 movies ( columns ) and 74,424 users ( rows ) . following the standard procedure @xcite ,
users with fewer than 20 ratings and movies with less than 2 ratings are removed .
this leaves us 36,656 users , 1,621 movies , and 2.5 million ratings ( available data ratio @xmath80 ) .
we randomly select 30,000 users for the weak generalization , and 5,000 users for the strong generalization .
the 1 m movielens dataset contains 1 million ratings in the range @xmath81 for 3,900 movies ( columns ) and 6,040 users ( rows ) . the same filtering leaves us 6,040 users , 3,592 movies , and 1 million ratings ( available data ratio @xmath82 ) .
we randomly select 5,000 users for the weak generalization , and 1,000 users for the strong generalization .
each experiment is run 3 times and the average reported .
the performance of the method is measured by the standard normalized mean absolute error ( nmae ) , computed by normalizing the mean absolute error by a factor for which random guessing produces a score of 1 .
the factor is 1.944 for eachmovie , and 1.6 for movielens .
in contrast to most exiting algorithms in the literature , the proposed method , thanks to its simplicity , enjoys the advantage of having very few intuitive parameters .
the covariance regularization parameter @xmath68 in is set equal to @xmath83 ( whose square root is one order of magnitude smaller than the maximum rating ) , the results being insensitive to this value as shown by the experiments .
the number of iterations of the map - em algorithm is fixed at @xmath84 , beyond which the convergence of the algorithm is always observed .
the noise @xmath35 in is neglected , i.e. , @xmath85 is set to @xmath86 , as the movie rating datasets mainly involve missing data , the noise being implicit and assumed negligible .
the experiments show that considering row @xmath11 of the matrix @xmath0 as signals leads to slightly better results than taking columns or 2d sub - matrices .
this means that each user is a signal , whose dimension is the number of movies . as in previous works @xcite , a post - processing that projects the estimated rating to an integer within the rating range
is performed .
a matlab code of the proposed algorithm is available at http://www.cmap.polytechnique.fr/~yu/research/mc/demo.html[_http://www.cmap.polytechnique.fr/@xmath87yu/research/mc/demo.html_ ] .
the results of the proposed method are compared with the best published ones including user rating profile ( urp ) @xcite , attitude @xcite , maximum margin matrix factorization ( mmmf ) @xcite , ensemble of mmmf ( e - mmmf ) @xcite , item proximity based collaborative filtering ( ipcf ) @xcite , gaussian process latent variable models ( gplvm ) @xcite , mixed membership matrix factorization ( m@xmath88f ) @xcite , and nonparametric bayesian matrix completion ( nbmc ) @xcite . for each of these methods ,
more than one results produced with different configurations are often reported , among which we systematically cite the best one .
all these methods are significantly more complex than the one here proposed
. tables [ tab : each : movie ] and [ tab:1m : movielens ] presents the results of various methods for both weak and strong generalizations on the two datasets .
nbmc generates most often the best results , followed closely by the proposed method referred to as gm ( gaussian model ) and gplvm , all of them outperforming the other methods .
the results produced by the proposed gm , with a by far simpler model and faster algorithm , is in the same ballpark as those of nbmc and gplvm , the difference with nmae being smaller than about 0.005 , marginal in the rating range that goes from 1 to 5 or 6 .
.nmaes generated by different methods for eachmovie database .
the smallest nmae is in boldface . [
cols="^,^,^",options="header " , ]
we have shown that a gaussian model and a map - em algorithm provide a simple and computational efficient solution for matrix completion , leading to results in the same ballpark as state - of - the - art ones for movie rating prediction .
future work may go in several directions .
on the one hand , the proposed conceptually simple and computationally efficient method may provide a good baseline for further refinement , for example by incorporating user and item bias or meta information @xcite . on the other hand ,
gaussian mixture models ( gmm ) that have been shown to bring dramatic improvements over single gaussian models in image inpainting @xcite , are expected to better capture different characteristics of various categories of movies ( comedy , action , etc . ) and classes of users ( age , gender , etc . ) .
however , no significant improvement by gmm over gaussian model has yet been observed for movie rating prediction .
this needs to be further investigated , and such improvement might come from proper grouping and initialization .
* acknowledgments : * work partially supported by nsf , onr , nga , aro , and nsseff . we thank s. mallat for co - developing the proposed framework , and m. zhou and l. carin for their comments and help with the data . | a general framework based on gaussian models and a map - em algorithm is introduced in this paper for solving matrix / table completion problems .
the numerical experiments with the standard and challenging movie ratings data show that the proposed approach , based on probably one of the simplest probabilistic models , leads to the results in the same ballpark as the state - of - the - art , at a lower computational cost .
matrix completion , inverse problems , collaborative filtering , gaussian mixture models , map estimation , em algorithm |
the australia telescope national facility pulsar catalog @xcite lists over 2300 radio pulsars .
several radio studies ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) have measured the polarization for many of these pulsars .
radio pulsars typically show moderate - to - strong linear polarization ( @xmath6 ) , being stronger for those of higher spin - down energy - loss rate ( * ? ? ?
* figure 8) .
the linear polarization sometimes exhibits a characteristic swing or sweep of the position angle in an s - like shape near the pulse center , which is routinely interpreted in terms of the rotating vector model ( rvm , * ? ? ?
* ) . for this model
the point of emission is assumed to be in the polar cap region of the pulsar where a dipolar magnetic - field line points with a small angle ( beamwidth ) towards the observer .
the two free parameters of this simple model are the angle between the axes of rotation and the orientation of the magnetic dipole , and the view angle between the line of sight and the rotation axis .
the variation of the radio position angle from some pulsars ( e.g. , * ? ? ? * and references therein ) can be described by this model .
the crab pulsar , the compact remnant of sn1054 , and its pulsar wind nebula ( pwn ) are amongst the most intensively studied objects in the sky .
the pulsar is one of the youngest and most energetic and its pulsed emission has been detected from 10 mhz @xcite up to 400 gev by veritas @xcite and by magic @xcite .
the pwn is detected at energies up to 100 tev @xcite .
both the pulsar and nebula are predominantly sources of non - thermal radiation ( synchrotron , curvature , and compton processes ) , indicated not only by the broadband spectral continua but also by strong polarization in many wavelength bands @xcite . in the visible band , spatially - resolved polarimetry of
the nebula , which began over a half century ago @xcite , continues ( e.g. , * ? ? ?
* and references therein ) .
owing to its brightness , phase - resolved optical polarimetry of the pulsar has also been possible @xcite . however , phase - resolved x- and @xmath7-ray polarimetry measurements of the crab pulsar require space - based instruments , which have had limited sensitivity .
oso-8 observations @xcite of the crab established only upper limits to the x - ray ( 2.6-kev and 5.2-kev ) polarization of the pulsed emission .
integral ibis observations @xcite also detect no significant pulsed @xmath7-ray ( 200800-kev ) polarization , although the off - pulse emission appears highly linearly polarized and is possibly associated with structures close to the pulsar rather than with the pulsar itself .
the crab pulsar s light curve exhibits different features at different wavelengths , but it is currently the only pulsar for which the principal features persist over all wavelengths , from radio to @xmath7-ray .
there are two principal components the main pulse ( mp ) and the interpulse ( ip ) .
this double - peak structure remains more - or - less phase - aligned over all spectral bands @xcite .
one of several additional features in the radio band is the low - frequency component ( lfc , e.g. , @xcite @xcite ) , having very low amplitude and occurring @xmath8 fractional pulse phase ( @xmath9 ) before the mp .
this component is most prominent around 1.4 ghz , in contrast with the `` precursor '' component @xcite , which precedes the mp by @xmath10 fractional pulse phase ( @xmath11 ) at 0.327 and 0.610 ghz ( table 2 of * ? ? ?
the mp and ip appear at roughly the same pulse phase from radio to @xmath7-ray wavelengths , suggesting that their emission originates from a similar location in the magnetosphere at all wavebands .
modeling of @xmath7-ray light curves from the many pulsars observed by the _ fermi gamma - ray space telescope _
@xcite strongly indicates that the high - energy emission originates in the outer magnetosphere , at altitudes comparable to the light - cylinder radius @xcite . outer magnetosphere emission models , such as the outer - gap @xcite , slot - gap @xcite , and current - sheet @xcite had been proposed and studied prior to the _ fermi _ observations , but their emission geometry seems to account for the characteristics and variety of observed @xmath7-ray light curves . in addition , _ fermi _ has discovered a number of @xmath7-ray millisecond pulsars whose radio peaks are nearly aligned with their @xmath7-ray peaks ( e.g. , * ? ? ?
* ) , like the crab .
modeling both @xmath7-ray and radio light curves of these pulsars with the same outer magnetosphere emission models used to model young pulsars has suggested that their radio emission may originate from very high altitudes @xcite .
thus , in this paper we compare the phase - resolved radio polarization observations ( [ s : obs ] ) that we have analyzed ( [ s : res ] ) with such models ( [ s : imp ] ) .
@xcite measured the linear polarization of the crab pulsar s mp and precursor components at two radio frequencies .
the 0.410-ghz measurements found the mp to be @xmath12 linearly polarized at position angle @xmath13 and the precursor to be @xmath14 linearly polarized at position angle @xmath13 .
the 1.664-ghz measurements found the mp to be @xmath1 linearly polarized at position angle @xmath15 and the precursor to be completely absent .
as these measurements had rather large uncertainties and were obtained with a time resolution @xmath16 of the pulse period , they were quite limited for detecting variation of the linear polarization degree or position angle within a feature .
however , @xcite noted a suggestion of rotation of the 1.664-ghz polarization position angle by about @xmath17 through the mp .
more recently , @xcite examined the pulse - profile morphology and polarization properties at three radio frequencies1.424 ghz , 4.885 ghz , and 8.435 ghz with a time resolution of @xmath18 ( about @xmath19 of the pulse period ) .
the 1.424-ghz measurements found the mp to be @xmath20 linearly polarized at position angle @xmath21 ; the ip , @xmath12 at position angle @xmath21 ; and the lfc , @xmath22 at position angle @xmath23 .
@xcite note that the polarization position angle `` changes across the full period , although not significantly within components '' .
here we first report our observations ( [ s : obs ] ) , using the westerbork synthesis radio telescope ( wsrt ) in the netherlands , of the full ( linear and circular ) 1.38-ghz polarization of the crab pulsar , at high time resolution .
we then describe the polarimetry analysis and results ( [ s : res ] and appendix [ s : stat ] ) for the three pulse components , with a primary objective of determining the sweep of the position angle across each .
next we discuss the implications ( [ s : imp ] ) of our measurements and analysis upon theoretical models for the pulsar emission .
finally , we summarize our conclusions ( [ s : sum ] ) .
the wsrt observations , on 2011 august 8 , used 14 25-m - diameter dishes combined coherently to form the equivalent of a 94-m dish for pulsar observations . owing to the interferometric nature of the wsrt
, the observations partially resolve out the radio - bright nebula , thus improving sensitivity over typical single - dish observations .
moreover , as the wsrt is an equatorially mounted telescope , there is no need to correct for parallactic angle . to combine coherently the dishes , correlated data from observation of a bright calibrator source
is used to determine phase delays amongst dishes .
this is accomplished using initially an unpolarized calibrator to determine delays between the two linear polarizations separately , followed by observation of a polarized calibrator to determine any residual delays between the two polarizations .
these procedures accurately calibrate the relative fluxes in the four stokes parameters hence , the polarization properties but not the absolute flux .
consequently , we express the stokes measurements ( e.g. , figure [ f : iquv ] ) in arbitrary units .
the puma - ii @xcite pulsar back - end was used to record nyquist - sampled voltages at 8-bit resolution , across a 160-mhz band centered on 1380 mhz , for psrs b0531 + 21 ( crab ) and b0355 + 54 , for a total of 144 and 18 minutes , respectively .
the data were subsequently coherently de - dispersed and folded using the dspsr @xcite software package .
polarization profiles were formed after correcting for ( frequency - dependent ) interstellar faraday rotation ( rotation measure rm = @xmath24 ) of the position angle , using the psrchive software package @xcite .
the polarization calibration was already carried out when forming the coherent sum of the dishes , nevertheless psr b0355 + 54 was observed to verify that no further polarization calibration was required .
comparison with the profile observed by @xcite showed that the polarization calibration matched exactly .
the crab - pulsar profile was folded using the jodrell bank ephemeris with 8192 bins ( about 4.1 @xmath25s / bin ) across the pulse profile , matching the time resolution of the data after dividing into frequency channels and coherently de - dispersing .
this time resolution was chosen also to match approximately the minimum broadening caused by scattering of the crab pulsed emission by free electrons in the crab nebula ( e.g. , * ? ? ?
* ; * ? ? ?
figure [ f : iquv ] displays our measurement of the four stokes parameters @xmath26 , @xmath27 , @xmath28 , and @xmath29which fully characterize the polarization folded on the pulse period . unfortunately , we were unable to determine the absolute polarization position angle for the crab pulsar observation . instead , we selected a coordinate system for the stokes parameters such that the mp has @xmath30 and @xmath31 .
inspection of figure [ f : iquv ] immediately shows that our 1.38-ghz observations detect the flux and polarization of three components mp , ip , and lfc . like the mp ,
the ip has @xmath32 and @xmath33 ; but the lfc has @xmath34 and @xmath35 : thus , the polarization position angles for the mp and the ip are roughly equal but differ from that of the lfc by about @xmath5 ( cf .
[ e : psi ] ) .
similarly , but less obviously , the circular polarization of the mp and the ip are comparable , but that of the lfc has opposite polarity .
the stokes parameters have the virtues that they are statistically independent , typically exhibit gaussian errors , and are directly superposable i.e . , each stokes component ( @xmath26 , @xmath27 , @xmath28 , or @xmath29 ) for multiple sources is the sum of the respective stokes component for each source .
these properties follow from the fact that the stokes parameters describe the polarization state in cartesian - like coordinates .
this has the added virtue that there is no coordinate singularity at the origin , as occurs for polar - like coordinates such as linear - polarization degree @xmath6 and position angle @xmath36 .
consequently , we perform all statistical analyses and model fitting ( appendix [ s : stat ] ) on ( pre - processed , appendix [ ss : proc ] ) raw stokes data .
it is , of course , straightforward to transform to more customary parameters
e.g . , linear - polarization degree @xmath6 ( eq . [ e : pl ] ) , position angle @xmath36 ( eq . [ e : psi ] ) , and circular - polarization ( signed ) degree @xmath37 ( eq . [ e : pc ] ) : @xmath38 @xmath39 @xmath40 for the three pulse features ( mp , ip , and lfc ) , we estimate @xmath41 , @xmath42 , and @xmath43 at each datum @xmath44 by substituting the measured @xmath45 , @xmath46 , @xmath47 , and @xmath48 into equations [ e : pl ] , [ e : psi ] , and [ e : pc ]
. figures [ f : pol_mp ] and [ f : pol_ip ] display the direct estimates of @xmath45 , @xmath49 , @xmath50 , and @xmath51 over the mp and ip , respectively . as
the lfc is quite weak relative to the mp and the ip , the plots for the lfc are too noisy to display legibly . even for the stronger features mp and ip the rms noise in the directly calculated polarization parameters ( @xmath49 , @xmath50 , and @xmath52 , which serves as an estimator of the statistical error , substantially increases away from the pulse center due to the low signal - to - noise ratio _ per sample _ in the pulse wings . in order to deal effectively with low - signal - to - noise data in the wings of the mp and ip and throughout the ( weaker ) lfc
, we adopt a more rigorous forward - modeling approach to fit the measured stokes data to the modeled @xmath53 , @xmath54 , @xmath55 , and @xmath56 : appendix [ s : stat ] describes in some detail our approach for fitting polarization models to the stokes data . as figures [ f : pol_mp ] and [ f : pol_ip ] indicate that neither @xmath60 , @xmath61 , nor @xmath62 varies rapidly across the pulse profile , the approach simply models @xmath60 , @xmath61 , and @xmath62 as taylor - series expansions in the phase - angle offset @xmath63 from the center @xmath64 of the respective pulse feature ( mp , ip , or lfc ) .
table [ t : fit ] tabulates the best - fit taylor - expansion coefficients for the polarization dependence upon phase - angle offset : @xmath65 @xmath66 @xmath67 .best - fit polarization coefficients for the mp , ip , and lfc , using a single gaussian for each pulse profile and up - to - quadratic variations in polarization functions @xmath60 , @xmath61 , and @xmath62 .
[ t : fit ] [ cols="^,^,^,^,^",options="header " , ] comparison of the column `` mp '' in table [ t : fit1 ] with that in table [ t : fit2 ] ( or , equivalently , with the column `` 1-gaussian '' in table [ t : fit6 ] ) finds that inclusion of the three quadratic polarization coefficients especially @xmath68reduces @xmath69 by 42 ( from 473 to 431 ) . while @xmath70 and @xmath71 within statistical uncertainties
, @xmath72 is statistically significant but small .
the main cause of the poor fit of the 1-gaussian model to the mp data , however , has nothing to do with polarization .
figure [ f : iquv_mp_2 + 4 ] illustrates that , for the fine time resolution and the high signal - to - noise ratio of the mp data , substructure in the pulse profile is quite evident . using a 6-gaussian ( 2 broad and 4 narrow ) profile for @xmath53
substantially improves the fit . comparing the column `` 6-gaussian '' with `` 1-gaussian '' in table [ t : fit6 ] finds that inclusion of @xmath73 additional ( gaussian ) parameters reduces @xmath74 by 1221 ( from 1909 to 688 ) .
even so , the fit to the stokes data is not formally acceptable .
it is important to note that the best - fit expectation values and uncertainties for the polarization coefficients ( @xmath75 , @xmath76 , @xmath68 ; @xmath77 , @xmath78 , @xmath79 ; @xmath80 , @xmath81 , @xmath82 ) are rather insensitive to details of the pulse profile .
thus , we compensate for fine substructure in the pulse profile by increasing the estimators for the measurement standard deviations until a statistically acceptable fit is achieved .
that is , we adjust @xmath83 , @xmath84 , @xmath85 , and @xmath86 until ( eq . [
e : chi2 ] ) @xmath87 , @xmath88 , @xmath89 , and @xmath90 , respectively , are close to unity .
the column `` 1-gaussian ( adj . ) '' in table [ t : fit6 ] shows the best - fit polarization parameters for a single - gaussian profile , with weightings adjusted as described .
the only noticeable effect of this adjustment upon the best - fit polarization parameters is a small change typically an increase in their uncertainties .
the uncertainties quoted in table [ t : fit ] ( [ s : res ] ) are the typically more conservative values obtained using the single - gaussian profiles and adjusted weightings .
abdo , a. a. , ajello , m. , allafort , a. , et al .
2013 , , 208 , 17 abdo , a. a. , allen , b. t. , atkins , r. , et al .
2012 , , 750 , 63 aharonian , f. , akhperjanian , a. g. , bazer - bachi , a. r. , et al .
2006 , , 457 , 899 aharonian , f. , akhperjanian , a. , beilicke , m. , et al .
2004 , , 614 , 897 aleksi , j. , alvarez , e. a. , antonelli , l. a. , et al .
2012 , , 540 , a69 aliu , e. , arlen , t. , aune , t. , et al .
2011 , science , 334 , 69 allen , b. t. , & yodh , g. b. 2007 , aip conf .
921 , 528 backer , d. c. , wong , t. & valanju , j. 2000 , 543 , 740 bai , n. & spitkovsky , a. 2010 , apj , 715 , 1282 bhat , n. d. r. , tingay , s. j. , & knight , h. s. 2008 , , 676 , 1200 blaskiewicz , m. , cordes , j. m. , & wasserman , i. 1991 , apj , 370 , 643 bridle , a. h. 1970 , , 225 , 1035 bhler , r. , & blandford , r. 2013 , arxiv:1309.7046 in preparation contopoulos , i. , & kalapotharakos , c. 2010 , , 404 , 767 dean , a. j. , clark , d. j. , stephen , j. b. , et al .
2008 , science , 321 , 1183 deutsch , a. j. 1955 , annales dastrophysique , 18 , 1 dyks , j. , harding , a. k. , & rudak , b. 2004 , , 606 , 1125 dyks , j. , & rudak , b. 2003 , , 598 , 1201 espinoza , c. m. , guillemot , l. , elik , . , et al .
2013 , , 430 , 571 forot , m. , laurent , p. , grenier , i. a. , gouiffs , c. , & lebrun , f. 2008 , , 688 , l29 gould , d. m. , & lyne , a. g. 1998 , , 301 , 235 hankins , t. h. , kern , j. s. , weatherall , j. c. , & eilek , j. a. 2003 , , 422 , 141 harding , a. k. , decesar , m. e. , miller , m. c. , kalapotharakos , c. & contopoulos , i. 2011 , proc . of 2011 fermi symposium , econf c110509 [ arxiv:1111.0828 ] hibschman , j. a. & arons , j. 2001 , apj , 546 , 382 jones , d. h. p. , smith , f. g. , & wallace , p. t. 1981 , , 196 , 943 kalapotharakos , c. , kazanas , d. , harding , a. , & contopoulos , i. 2012 , , 749 , 2 karastergiou , a. , & johnston , s. 2006 , , 365 , 353 karuppusamy , r. , stappers , b. , & van straten , w. 2008 , , 120 , 191 karuppusamy , r. , stappers , b. w. , & van straten , w. 2010 , , 515 , a36 kuiper , l. , hermsen , w. , cusumano , g. , et al .
2001 , , 378 , 918 kuzmin , a. , losovsky , b. y. , jordan , c. a. & smith , f. g. 2008 , , 483 , 13 li , j. , spitkovsky , a. , & tchekhovskoy , a. 2012 , , 746 , 60 lyne , a. g. , & graham - smith , f. 2006 , pulsar astronomy , 3rd ed .
( cambridge , uk : cambridge university press ) , isbn 0521839548 lyne , a. , graham - smith , f. , weltevrede , p. , et al .
2013 , science , 342 , 598 majid , w. a. , naudet , c. j. , lowe , s. t. , & kuiper , t. b. h. 2011 , , 741 , 53 manchester , r. n. 1971 , , 23 , 283 manchester , r. n. , hobbs , g. b. , teoh , a. , & hobbs , m. 2005 , , 129 , 1993 moffett , d. a. , & hankins , t. h. 1996 , , 468 , 779 moffett , d. a. , & hankins , t. h. 1999 , , 522 , 1046 moran , p. , shearer , a. , gouiffs , c. , & laurent , p. 2013a , arxiv:1302.3622 moran , p. , shearer , a. , mignani , r. p. , et al . 2013b , , 433 , 2564 muslimov , a. g. , & harding , a. k. 2004 , , 606 , 1143 ng , c .- y .
& romani , r. w. 2008 , apj , 673 , 411 oort , j. h. , & walraven , t. 1956 , , 12 , 285 ptri , j. , & kirk , j. g. 2005 , , 627 , l37 pierbattista , m. , et al .
2014 , submitted to a&a [ arxiv:1403.3849 ] radhakrishnan , v. , & cooke , d. j. 1969 , , 3 , 225 rankin , j. m. 1990 , , 352 , 247 romani , r. w. , & watters , k. p. 2010
, , 714 , 810 romani , r. w. , & yadigaroglu , i .- a . 1995 , , 438 , 314 silver , e. h. , kestenbaum , h. l. , long , k. s. , et al .
1978 , , 225 , 221 sowikowska , a. , kanbach , g. , kramer , m. , & stefanescu , a. 2009 , , 397 , 103 smith , f. g. , jones , d. h. p. , dick , j. s. b. , & pike , c. d. 1988 , , 233 , 305 van straten , w. , & bailes , m. 2011 , , 28 , 1 van straten , w. , demorest , p. , & oslowski , s. 2012 , astron .
research & technology , 9 , 237 venter , c. , johnson , t. j. , & harding , a. k. 2012 , , 744 , 34 weltevrede , p. , & johnston , s. 2008 , , 391 , 1210 wolfram research , inc .
2013 , mathematica , version 9.0 ( champaign , il : wolfram research ) woltjer , l. 1957 , , 13 , 301 | using the westerbork synthesis radio telescope ( wsrt ) , we obtained high - time - resolution measurements of the full ( linear and circular ) polarization of the crab pulsar .
taken at a resolution of 1/8192 of the 34-ms pulse period ( i.e. , @xmath0 ) , the 1.38-ghz linear - polarization measurements are in general agreement with previous lower - time - resolution 1.4-ghz measurements of linear polarization in the main pulse ( mp ) , in the interpulse ( ip ) , and in the low - frequency component ( lfc ) .
we find the mp and ip to be linearly polarized at about @xmath1 and @xmath2 , with no discernible difference in polarization position angle .
however , and contrary to theoretical expectations and measurements in the visible , we find no evidence for significant variation ( sweep ) in polarization position angle over the mp , the ip , or the lfc .
although , the main pulse exhibits a small but statistically significant quadratic variation in the degree of linear polarization . we discuss the implications which appear to be in contradiction to theoretical expectations .
in addition , we detect weak circular polarization in the main pulse and interpulse , and strong ( @xmath3 ) circular polarization in the low - frequency component , which also exhibits very strong ( @xmath4 ) linear polarization at a position angle about @xmath5 from that of the mp or ip .
the pulse - mean polarization properties are consistent with the lfc being a low - altitude component and the mp and ip being high - altitude caustic components . nevertheless , current models for the mp and ip emission do not readily account for the observed absence of pronounced polarization changes across the pulse .
finally , we measure ip and lfc pulse phases relative to the mp that are consistent with recent measurements , which have shown that the phases of these pulse components are evolving with time . |
r coronae borealis ( rcb ) stars are defined by their light curves displaying sudden and large drops in brightness with slower recoveries to the baseline level , with these events occurring randomly in time .
these spectacular dips are caused when the star forms dense dust clouds on the line of sight to earth hiding the star .
the rcb stars are all hydrogen deficient supergiants , with various abundance anomalies , including enriched nitrogen and carbon .
rcb stars are rare , with only 76 known in our milky way .
the evolutionary status of rcb stars is that they might be from recent coalescences of a double white dwarf binary or from a final helium shell flash in a born - again asymptotic giant branch ( agb ) star .
clayton ( 1996 ; 2012 ) presents full reviews of rcb stars .
most rcb stars are relatively cool , with surface temperatures of 5,000 - 7,000 k. however , four of the known rcb stars have a greatly hotter surface temperature , from 15,000 - 25,000 k , and these are called the ` hot rcb stars ' .
the four known hot rcb stars are v348 sgr , mv sgr , and dy cen in our milky way plus hv 2671 in the large magellanic cloud .
dy cen and mv sgr are hydrogen deficient ( just like the cool rcb stars ) and mostly composed of helium , while both have relatively infrequent drops in brightness .
v348 sgr and hv 2671 are very carbon - rich ( 55% carbon , most of the rest helium ) , and elemental abundances like in central stars of planetary nebulae , while both have frequent episodes of brightness declines .
a plausible idea is that the two greatly different composition indicate two formation mechanisms , with dy cen and mv sgr simply being the ` progeny ' of the normal rcb stars as they heat up , and with v348 sgr and hv 2671 being somehow formed during a final helium - shell flash on post - agb stars .
thus , the birth mechanism of the hot rcb stars could be either from born - again systems or white dwarf mergers . nevertheless , it is still unclear as to the relationship between the hot rcb stars and the cool rcb stars , as well as to other classes of stars ( the born - again stars and the wolf - rayet central stars of planetary nebulae ) .
de marco et al . ( 2002 ) present a full review of the hot rcb stars .
the key high - level science question is to understand the evolutionary state of the rcb stars , both hot and cold . for this , a critical piece of evidence is to watch them evolve in real time .
like for the born - again post - agb stars ( i.e. , v605 aql , sakurai s object , and fg sge ) , substantial movement across the hr diagram might be seen on time scales of a decade and a century .
for this , de marco et al .
( 2002 ) have pointed out that the baseline levels for mv sgr , v348 sgr , and dy cen are apparently fading over the last century , and this would only be from movement from right - to - left across the top of the hr diagram .
these century - long light curves have inevitable big problems for two reasons .
the light curves were compiled from magnitudes in the @xmath0 , @xmath1 , and @xmath2 systems , which makes for large systematic uncertainties between old and new magnitudes , with exactly this able to make apparent systematic declines when none are real .
de marco et al .
did attempt to correct for such effects .
the big effect not mentioned is that all the old photometry is always systematically in error by from 0.1 to @xmath31.0 mag simply because the old standard stars had these errors . in the days before photoelectric photometers
could reach to cover faint sequences ( i.e. , before the late 1970s ) , the standard stars and comparison sequences were all calibrated by photographic transfers from the harvard - groningen selected areas and the north polar sequence .
these photographic transfers always had problems , due to effects like reciprocity failure , in that claimed magnitudes were always reported systematically different than what we take on the modern magnitude scales . in general , the errors are small for stars brighter than tenth magnitude and start increasing steeply as the stars get fainter . thorough studies of errors for old standard fields and comparison stars are given by sandage ( 2001 ) and patat et al .
( 1997 ) , while i have many examples of poor old calibrations ( e.g. , schaefer 1994 ; 1995 ; 1996 ; 1998 ) the old ` photographic magnitude system ' ( i.e. , @xmath2 ) is really just a poorly calibrated b magnitude .
all the early magnitudes of hot rcb stars are photographic magnitudes , all from the harvard plates , all from old measures , and so there are inevitably possibly - large systematic errors in their long - term light curves .
it might be that these systematic errors have created the apparent secular fading of the hot rcb stars , or it might be that the errors are such that the claim of secular fading is still correct . in this paper , i will solve these problems for the old photometry , and answer the question of whether the hot rcb stars are fading or not . to do this
, i visited the harvard college observatory ( hco ) in october 2015 , and remeasured many magnitudes from plates dated 1896 to 1989 , all on the modern johnson b magnitude system . for the modern portions of the light curves ,
i used a variety of sources from the literature and from the _ american association of variable star observers _ ( aavso ) , all in the johnson b magnitude system .
my combined light curves , all in a single uniform system , have a larger time range and many more magnitudes than given in de marco et al .
i have further extended their work by adding the b - band light curve for the fourth and last - known hot rcb star , hv 2671 .
from around 1890 to the late 1970s , a large part of astronomy was from photometry as based on photographic sky pictures .
these pictures were recorded on blue - sensitive emulsion attached to one side of a glass plate , with the plate being exposed to star light in a special holder at the focus of the telescope .
the developed emulsion had the sky being nearly transparent , while stars were round black points .
photographic emulsions have only a small dynamic range between the sky and saturation , so star images are almost all completely saturated ( i.e. , black ) in the centers , with only a small annulus of grey in the outer tails of the star profile . with this , essentially , the only change in the star image as the magnitude varies is the diameter of the star image . given the variability from plate - to - plate and the non - linearity of the emulsion , the only way to calibrate the image - diameter - versus - magnitude relation is to use comparison stars nearby on the plate .
the procedure is to make some measure of the image diameter for the target star as well as for a sequence of nearby comparison stars with a tight spacing of magnitudes , both brighter and fainter than the target . with this
, the magnitude of the target can be measured by interpolation in the image - diameter - versus - magnitude relation as determined for each plate from the comparison sequence .
the image radii vary such that the square of the radii or the logarithm of the radii are linear with the magnitude , depending on the brightness of the star ( schaefer 1981 ) . in general
, the full calibration curve ( image radius versus magnitude ) is nonlinear with either magnitude or flux .
this condition violates one of the requirements ingrained in observers with ccds , because then magnitudes can not be calculated from any application of the magnitude equation with one comparison star .
the long - standing traditional solution is to use a whole sequence of comparison stars , strung out over a wide range of brightnesses , so that the radius - versus - magnitude relation is empirically determined for the brightness of the target star .
the image diameters can be measured with machines called ` iris diaphragm photometers ' ( first developed in the 1930s ) and with photoelectric scanners ( first developed in the 1970s ) . from the earliest days , the dominant method for measuring image diameters was simply for a trained observer to make size comparisons between the diameter of the target star and the diameters of nearby comparison stars .
the human eye is remarkably accurate at side - by - side comparisons of the sizes of round objects .
the procedure is to view the glass plate on a light table through a loupe or a low - power microscope , compare the target s size to the size of nearby stars , and judge by - eye the relative placement of the target star s diameter . to give a specific and typical example ,
if the target is judged to be halfway between two comparison stars of magnitude 12.2 and 12.6 , then the target has a magnitude of 12.4 . in practice ,
an inexperienced worker can produce magnitudes with an accuracy of 0.3 mag or so , while an experienced worker can produce magnitudes with a one - sigma uncertainty of 0.1 mag .
for many situations , where magnitudes from many plates can be averaged together , the real accuracy of the light curves can be 0.02 mag or better . for an experienced worker ,
the by - eye method provides equal or better accuracy as compared to iris diaphragm photometers or scanning .
the by - eye method is very simple , cheap , and fast , whereas the instrumented methods are always complex , slow , and costly .
harvard college observatory has a collection of roughly 500,000 glass plates recording the entire sky from 1889 to 1989 .
( there are few plates from 1954 to 1969 due to the notorious ` menzel gap ' . )
these are almost all blue - sensitive emulsion on glass plates 8@xmath410 inch in size , stored in paper envelopes , and placed on shelves in time order .
the plates were taken with a wide variety of telescope , from essentially camera lenses up to 24-inch apertures , with the plate sizes covering widths from 5@xmath5 to 42@xmath5 .
the limiting magnitudes for the normal - quality plates vary from b=12 to deeper than b=18 .
the harvard plates have 1000 - 4000 plates covering any given position on the sky .
harvard has about half the existing archival plates in the world , and is nearly the only source for targets in the southern skies .
historically , from 1890 to 1960 , the harvard plates dominated the world of variable stars for anything fainter than about eleventh magnitude . to take an example of the hot rcb stars , all four were discovered with the harvard plates , and the only published information of any type from before the 1950s
is the harvard light curves . for many questions of modern astrophysics , light curves with 0.1 mag or 0.02 mag accuracy
are more than adequate , so the accuracy attainable with ccds is completely irrelevant . in the world of variable stars , the stars are displaying phenomena on all time scales .
modern studies can cover variable star phenomena on time scales faster than the duration of a single telescope run , and multiple telescope runs can be pasted together to get a picture of phenomena up to a decade in time scale . but to measure phenomena on time scales from a decade to a century , the only means is to use archival data . for most stars ,
the only source of archival information older than a decade or two is from archival photographic plates , and that largely means the harvard plates . for the hot rcb stars , in looking for any secular trend ( as associated with the evolution of the stars ) ,
the only solution is to get fully calibrated light curves from harvard .
historically , the harvard plates were the predecessor of the johnson b system through the north polar sequence . in modern times , the native magnitudes of the harvard blue plates have always been found to have a near - zero color term with respect to the johnson b system .
this means that as long as the comparison star magnitudes are on the johnson b system , then the resultant magnitudes are exactly in the johnson b system . for my measures of the harvard plates ,
i have taken all my comparison star magnitudes from the b - band measures of the aavso photometric all - sky survey ( _ apass _ , henden & munari 2014 ) .
these magnitudes are tied to the johnson b magnitudes with high accuracy ( munari et al .
2014 ) , as calibrated from the standard stars of landolt ( 2009 ) .
thus , my modern measures of the harvard plates are accurately in the johnson b system . critically , both the very extensive _ dasch _ program ( grindlay et al .
2012 ) and my own extensive measures prove that long term light curves from harvard of normal ( i.e. , constant ) stars do _ not _ produce any measurable slope or trend ( i.e. , typically @xmath60.05 magnitude per century ) over the last century .
further , these check star magnitudes are consistent with the modern measures .
this is the proof that any observed secular trend is not some data or analysis artifact .
the goal of this paper is to get the century - long light curves for all four known hot rcb stars so as to test for any secular fading in the maximum light . for this , the only way to get the old data is from harvard , and these are only in the johnson b - band . to minimize the mixing of bands
, i will take the aavso and literature magnitudes for the b - band .
the johnson b magnitudes for the four hot rcb stars are listed in table 1 .
these do not include the magnitudes where the star was substantially fainter than the maximum brightness .
.b magnitudes from harvard plates . [ cols="<,<,<,<",options="header " , ] @xmath7as quoted in rao et al .
( 1993 ) [ fig : fig .
in essence , i have merely confirmed and extended the conclusion of de marco et al .
( 2002 ) that the hot rcb stars are secularly fading .
my improvements have been to use a single photometric system all with modern comparison stars , to collect many more magnitudes over a much wider time range , as well as to measure the decline rate for the fourth hot rcb star .
de marco et al . (
2002 ) have interpreted these secular declines as being due to the star evolving to hotter temperature at constant luminosity , such that the bolometric correction to the optical band makes for an apparent dimming .
( the only other plausible explanation is some sort of general increase in the circumstellar dust density , but such would lead to color changes that are not observed in the cases of dy cen and mv sgr . )
this interpretation matches with the general idea that the hot rcb stars are moving horizontally across the top of the hr diagram as part of their normal and fast evolution .
pandey et al .
( 2014 ) have explicitly tested this interpretation for dy cen , where archival spectra give surface temperatures of 19,400@xmath8400 k in 1987 , 23,000@xmath8300 k in 2002 , and 24,800@xmath8600 k in 2010 .
this is 5,400 k in 23 years , or 23,500 k per century .
this increase in the stellar temperature is confirmed and reflected in the dramatic change in the excitation of the nebula around dy cen ( rao et al .
. we can translate this rate of temperature change for dy cen into a magnitude decline rate .
the calculation of the change in bolometric corrections and the change of b - v color is presented in fig . 1 of pandey et al .
( 2014 ) for the relevant conditions . for a temperature of 19,400 k , they give v=12.78 and b - v=-0.80 ( with an arbitrary zero point ) , for b=11.98 .
for a temperature of 24,800 k , they give v=13.38 and b - v=-0.85 ( with the same arbitrary zero point ) , for b=12.53 .
thus , the observed temperature decline in 23 years corresponds to a fading by 0.55 mag , for a decline rate of 2.39 magnitudes per century .
this is close to the average decline rate for the years 1983 to 2015 ( see fig .
so the observed temperature change is consistent with the observed decline rate . for the evolution of dy cen
going back in time , the temperature must be relatively low in old times , resulting in a large bolometric correction .
a simple extrapolation back to 1905 puts the temperature to near zero , so the temperature changes can not be linear with time
. nevertheless , the temperature back in 1905 should be relatively quite cold .
the bolometric correction for the b band is minimized for a stellar temperature of 7,500 k , so that for evolution at constant luminosity , the b magnitude will be brightest at that temperature and dimmer as the temperature departs from this value to both hotter and colder temperatures .
so we then have a ready interpretation of the long - term trend in the maximum magnitude ( fig .
2a ) , with dy cen starting in 1906 out colder than 7,500 k , heating up to 7,500 k in 1932 when the star was at its brightest , then continuing heating of the star makes it dim over the next decades . the correction from a constant luminosity to the b band
can be taken from table 15.7 of cox et al .
( 2000 ) , where the minimum correction is at 7,500 k , and where the corrections of 0.6 mag are for temperatures of 5,800 k and 11,000 k. with this , dy cen had temperatures of 5,800 k around 1906 , 7,500 k around 1932 , and 11,000 k around the 1970s .
so we now have a simple explanation for why the dy cen maxima back around 1906 was substantially dimmer than around in 1932 . in all , a continuous temperature increase from 5,800 k in 1905 to 24,800 k in 2010 can account for both the observed change in maximum magnitudes and the observed changes in the temperature .
i have made a crude model that accounts for the stellar temperature and maximum magnitude as a function of time . from the models of saio ( 1988 ) ,
i take the logarithm of the temperature to be linear with time , with this being approximately right for a given star under hot rcb conditions .
i then set the linear relation by using the observed temperature in 2010 plus the 7,500 k condition for 1932 . with these temperatures ,
i get the bolometric corrections and b - v colors for supergiants from cox et al .
( 2000 ) , and add a constant to get the b - band magnitude outside of a decline .
this model light curve is displayed in fig .
this model result is not perfect , with the worst problem being that the bolometric correction for 24,800 k should make dy cen close to 2.0 mag fainter in 2010 than in 1932 , whereas it is observed to be more like 1.3 mag fainter .
this problem is easily solved if there is extra light in the dy cen system , perhaps from a wide binary companion or from the circumstellar material .
nevertheless , it is clear that the model captures the essence of a normal rcb stars heating up from around 5,800 k in 1906 to 24,800 k in 2010 .
so we have actually watched dy cen start out as an ordinary rcb star ( with temperature around 5,800 k ) , and then heat up to become a hot rcb star , and now appear as an extreme helium star with no dust dips .
this evolution has taken close to one century .
so we have a real measure of the duration of the hot rcb stage , and it is about one century .
this is a very fast phase of evolution .
this explains why so few hot rcb stars have been seen in our milky way galaxy , despite them being supergiant stars .
the heating up of dy cen is also associated with a sharp drop off in the frequency of dips .
de marco et al . ( 2002 ) note that dy cen had only four knowns dips from 1897 to 1927 , and zero known dips since 1960 .
with the harvard plates , i have identified eight additional dips , all from 1904 to 1934 .
apparently , the heating of the star s surface temperature is connected with the turn off of the dust formation , perhaps caused by a stoppage of pulsations as the star leaves some instability strip .
we realize that there is only a narrow time window over which the hot rcb phenomenon can be recognized , with only a few decades from the time when dy cen was sufficiently hotter than the upper limit for normal rcb stars up until the time when the dust dips turn off
. we can translate the typical decline rate of 1.11 magnitude per century into a temperature change rate . for a case with effective surface temperature of 15,000 k , pandey et al .
( 2014 ) give v=12.15 and b - v=-0.75 ( with the same arbitrary zero point ) , for b=11.40 .
for a temperature of 20,000 k , they give v=12.88 and b - v=-0.79 , for b=12.09 . for a 5,000 k temperature change over the range for hot rcb stars , the b magnitude changes by 0.69 magnitudes .
if this change happens over 62 years , then the b magnitude will fade at the rate of 1.11 magnitude per century .
dy cen is similar to extreme helium stars ( supergiants composed mostly of helium with near one percent carbon and temperatures 9000 - 35,000 k ) .
jeffery et al . ( 2001 ) found that four out of twelve extreme helium stars are heating up with rates from 20 to 120 degrees k per year .
( a useful program would be to search for b - band brightness changes from the 1890s to the present with the harvard plates for the two stars with the fastest temperature changes ; hd 160641 and bd -1@xmath53438 . )
such surface temperature changes are expected from models of extreme helium stars with masses of @xmath90.9 m@xmath10 ( saio 1988 ) .
the majority of stars with no measured change in surface temperature are presumably less massive , perhaps @xmath90.7 m@xmath10 .
dy cen is changing at a rate of 235 k per year from 1987 to 2010 .
if the models of saio ( 1988 ) are applicable to dy cen , then this star would be @xmath91.0 m@xmath10 . with the realization as to how some ` cold ' rcb stars should evolve on a time scale of a century , we can look for the same changes amongst the known normal rcb stars .
that is , the normal rcb stars are heating up , having their maximum magnitudes getting brighter , and their frequency of declines falling to near zero .
but such changes have never been seen for any star that is now a ` cold ' rcb star . a small number of cold rcb stars have century - long light curves with no apparent change in their brightness at maximum , while r crb itself has a 230 year record of unchanging peaks .
so the heating up of the cold rcb stars must usually be too slow to produce observable effects .
still , some fraction of the now - cold rcb stars might be like dy cen a century ago .
perhaps only the most - massive cold rcb stars will be evolving fast enough for the changes to be detectable .
a practical plan to search for fast evolving cold rcb stars is to construct a century - long light curve for many of them .
this could show secular changes in the magnitude at maximum as well as in the frequencies of declines . in practice ,
the primary sources are archival data from aavso and harvard . in any such study
, care must be used to place all magnitudes onto a consistent magnitude system .
( for example , old aavso magnitudes will require corrections for changes in the comparison sequences , and these can only be gotten from old charts archived at aavso headquarters . ) a substantial problem in seeking changes in the decline - frequency will be to adjust for the variations in time - coverage over the decades . with this , we have a plan for someone to make a systematic survey of century - long light curves for normal rcb stars so as to measure their evolution across the hr diagram . in general
, stars evolve on such slow time scales that astronomers have not been able to see the changes over time .
other than for supernovae , evolutionary changes have only been seen for a few post - agb stars , including the born - again stars and the stingray ( schaefer & edwards 2015 ) .
now , we can add the four hot rcb stars , with observed temperature rises of 8,000 k or more over the last century .
99 alcock , c. , allsman , r. a. , alves , d. r. , et al . , 1996 , apj , 470 , 583 clayton , g. c. , 1996 , pasp , 108 , 225 clayton , g. c. , 2012 , jaavso , 40 , 539 cox , a. n. , 2000 , allen s astrophysical quantities , 4th ed . , springer - verlag , new york de marco , o. , clayton , g. c. , herwig , f. , pollacco , d. l. , clark , j. s. , & kilkenny , d. , 2002 , apj , 123 , 3387 grindlay , j. , tang , s. , los , e. , & servillat , m. 2012 , in new horizons in time - domain astronomy ( iau symposium 285 ) , p. 29
- 34 , arxiv:1211.1051 heck , a. , houziaux , l. , manfroid , j. , jones , d. h. p. , & andrews , p. j .. 1985 , a&asupp , 61 , 375 henden , a. & munari , u. , 2014 , contrib .
skalnate pleso , 43 , 518 herbig , g. , 1964 , apj , 140 , 1317 hoffleit , d. , 1930 , harvard bulletin , 874 , 1 hoffleit , d. , 1958 , aj , 63 , 78 hoffleit , d. , 1959 , aj , 64 , 241 goldsmith , m. j. , evans , a. , albinson , j. s. , & bode , m. f. , 1990 , mnras , 245 , 119 jeffery , c. s. , starling , r. l. c. , hill , p. w. , & pollacco , d. , 2001 , mnras , 321 , 111 jones , k. , van wyk , f. , jeffery , c. s. , marang , f. , shenton , m. , hill , p. , & westerhuys , j. , 1989 , south african astronomical observatory circular , 13 , 39 kilkenny , d. , coulson , i. m. , laing , j. d. , jones , j. s. , & engelbrecht , c. , 1985 , south african astronomical observatory circular , 9 , 87 landolt , a. u. , 2009 , aj , 137 , 4186 marino , b. f. & walker , w. s. g. , 1971 , circular royal astron .
new zealand var .
star section , 184 , 1 munari , u. , henden , a. , frigo , a. et al .
, 2014 , aj , 148 , 81 pandey , g. , rao , n. k. , jeffery , c. , & lambert , d. l. , 2014 , apj , 793 , 76 patat , f. , barbon , r. , cappellaro , e. , & turatto , m. , 1997 , a&a , 317 , 423 pollacco , d. l. & hill , p. w. , 1991 , mnras , 248 , 572 rao , n. k. , giridhar , s. , & lambert , d. l. , 1993 , a&a , 290 , 201 rao , n. k. , lambert , d. l. , garcia - hernandez , d. a. , jeffery , c. s. , woolf , v. m. , & mcarthur , b. , 2012 , apjlett , 760 , l3 rao , n. k. , lambert , d. l. , garcia - hernandez , d. a. , & manchado , a. , 2013 , mnras , 431 , 159 saio , h. , 1988 , mnras , 235 , 203 sandage , a. , 2001 , pasp , 113 , 267 schaefer , b. e. , 1981 , pasp , 93 , 253 schaefer , b. e. , 1994 , apj , 426 , 493 schaefer , b. e. , 1995 , apjlett , 447 , l13 schaefer , b. e. , 1996 , apjlett , 460 , l19 schaefer , b. e. , 1998 , apj , 509 , 80 schaefer , b. e. & edwards , z. i. , 2015 , apj , 812 , 133 sherwood , v. e. , 1975 , in variable stars and stellar evolution , ed . v. e. sherwood and l. plaut ( dordrecht : reidel ) , 147 soszynski , i. , udalski , a. , szymanski , m. k. et al . , 2009 , acta astron . , 59 , 335 woods , i. e. , 1928 , bulletin harvard college obs . | the r coronae borealis ( rcb ) stars are cool supergiants that display irregular and deep dips in their light curves , caused by dust formation . there are four known hot rcb stars ( dy cen , mv sgr , v348 sgr , and hv 2671 ) , with surface temperatures of 15,00025,000 k , and prior work has suggested that three of these have secular fading in brightness .
i have tested this result by measuring century - long light curves in the johnson b - band with modern comparison star magnitudes , and i have extended this by measuring many magnitudes over a wide time range as well as for the fourth hot rcb star .
in all four cases , the b - band magnitude of the maximum light is now fast fading .
the fading rates ( in units of magnitudes per century ) are 2.5 for dy cen after 1960 , 1.3 for mv sgr , 1.3 for v348 sgr , and 0.7 for hv 2671 .
this secular fading is caused by the expected evolution of the star across the top of the hr diagram at constant luminosity , as the temperature rises and the bolometric correction changes .
for dy cen , the brightness at maximum light is rising from 1906 to 1932 , and this is caused by the temperature increase from near 5,800 to 7,500 k. before 1934 dy cen had frequent dust dips , while after 1934 there are zero dust dips , so there is some apparent connection between the rising temperature and the formation of the dust .
thus , we have watched dy cen evolve from an ordinary rcb star up to a hot rcb star and now appearing as an extreme helium star , all in under one century
. [ firstpage ] stars : agb and post - agb stars : chemically peculiar stars : individual ( v348 sgr , hv 2671 , dy cen , mv sgr ) stars : variable : general |
it is well established that the properties of galaxies are correlated with the environment ( e.g. * ? ? ? * ; * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
. at @xmath9 galaxy populations in rich clusters are dominated by ellipticals , s0s and at lower masses dwarf ellipticals ( des ) with few , if any , star - forming spirals @xcite .
cluster galaxies have not always been as inactive as they are at the present epoch .
@xcite showed that the fraction of blue ( star - forming ) galaxies among cluster members increases from almost zero in the local universe to @xmath020% by @xmath10 , while recent spitzer / herschel surveys have confirmed large numbers of starburst galaxies in clusters to @xmath11 and beyond ( e.g. * ? ? ?
* ) . while some of this rapid evolution of cluster galaxies can be explained by the cosmic @xmath12 decline in star formation among field galaxies since @xmath11 ( e.g. *
* ) , cluster populations have shown an accelerated evolution in star formation ( sf ) over the last four billion years , resulting in an overall @xmath13 reduction in the total star formation rates ( sfrs ) per unit halo mass since @xmath14 @xcite .
empirically , clusters accrete gas - rich , star - forming spirals at @xmath15.51.0 and then contribute to transforming them somehow into the passive s0s and des of local clusters . several mechanisms affecting the galaxy properties and dependent upon the environment have been proposed and investigated in detail and
all of them serve to kinematically disturb spiral galaxies and/or transform their structural properties and/or deplete their reservoirs of gas , and so quench star formation .
these physical processes include gravitational and tidal interactions amongst galaxies @xcite , between galaxies and the cluster gravitational field @xcite , galaxy mergers @xcite , group - cluster collisions @xcite , ram - pressure @xcite and viscous stripping @xcite , evaporation @xcite and ` starvation ' @xcite .
since these mechanisms are characterized by different time - scales and efficiencies which depend , in turn , on the properties of both the galaxies themselves ( e.g. their stellar masses , morphologies ) and their environment , they can affect the galaxy properties in different characteristic ways @xcite .
examples of this are the different effects of tidal and hydrodynamical interactions . the ram pressure exerted by the hot and dense intracluster medium ( icm )
can effectively remove the cold gas supply , truncating the gas disc and quenching star formation , in massive cluster galaxies passing through the centre of rich clusters in about one crossing time ( @xmath16yr ) , while the interstellar gas of a massive galaxy may never be completely stripped if the galaxy moves on a tangential orbit or is member of a poor cluster .
repeated high - velocity encounters ( harassment ) of cluster galaxies can destroy the fragile disc of dwarf galaxies , but not significantly affect the structure of a giant galaxy .
the development of large spectroscopic surveys such as the sloan digital sky survey ( sdss ) , plus the availability of panoramic far ultra - violet far infra - red ( fuv fir ) data from the _ galex _ and _ spitzer _ space telescopes have allowed the impact of environment on sf to be quantified in unprecedented detail . in these studies , however , the environment and its characteristics are usually treated as ` static parameters ' . while the galaxies move across the cluster , field and filament environments ( experiencing different ram pressures , encountering other galaxies and possibly being involved in merging ) , the cluster potential well , the galaxy density as well as the icm are usually not considered ` time - dependent ' .
of course , at first pass this approach is an unavoidable reduction on complexity .
on the other hand in a hierarchical universe with the assembly of the structures , the galaxies evolve and move , tending towards denser regions with time , while the environments change too , thus what we actually observe is _ galaxy evolution in an evolving environment_. the most massive structures in the local universe are superclusters , which are still collapsing with galaxy clusters and groups frequently interacting and merging , and where a significant number of galaxies are encountering dense environments for the first time .
the relative dynamical immaturity of superclusters and the presence of infalling dark matter halos make them ideal laboratories to test the predictions of hierarchical mass assembly models , and in particular on galaxy evolution .
superclusters are not so rare systems in the universe ( see * ? ? ?
the observations of superclusters is often considered a challenge to the hierarchical structure formation paradigm since such extreme dense structures , but also voids , are not reproduced by the @xmath17-body simulations .
@xcite pointed out that the reason of this discrepancy can be due to the method used to assess the probability of finding such events in the distribution of cold matter .
they proposed a new technique to analyse an ensemble of @xmath17-body simulations in a volume equal to that of the two - degree field galaxy redshift survey ( 2dfgrs , * ? ? ? * ) where the probability to find peculiar structures ( overdense and underdense ) was thus estimated to be @xmath02 per cent and not null .
superclusters allow in principle both to study dynamical processes such as cluster - cluster collisions and group - cluster mergers and to sample different environments from cluster cores to filaments and fields .
furthermore , within a dynamically active and locally dense structure the probability to observe evidence of environmental effects on galaxy evolution is dramatically enhanced making these systems a sort of _ magnifying glass _ to identify the different physical mechanisms which transform the properties of galaxies . in order to address all these aspects , a careful selection of the target
is fundamental .
the ideal structure should map different environments with evidence of cluster - cluster interactions .
finally , in order to study in detail the galaxy properties a resolution @xmath18kpc is required in a wide range of galaxy mass down to the dwarf regime where such galaxies are not quenched by internal processes , but are more susceptible to environmental transformations . with all this in mind , we have undertaken a study of the shapley supercluster which is the largest conglomeration of abell clusters in the local universe .
the shapley supercluster survey ( shass ) will map a 23deg@xmath2 region ( @xmath0 260mpc@xmath19 ) of the shapley supercluster at @xmath20 , containing filaments and embedded galaxy groups which form a dynamically - bound network connecting nine abell and two poor clusters , in order to identify the primary locations ( groups , filaments , clusters ) and mechanisms for the transformation of spirals into the s0s and des . among the observational studies of superclusters , the stages project @xcite addressed in particular the study of galaxy evolution in the abell901(a ,
b ) supercluster at redshift @xmath210.165 while the orelse survey ( lubin et al .
@xcite , but see also mei et al .
@xcite ) searched for structures on scales greater than 10mpc at higher redshifs ( [email protected] ) with the aim to investigate the properties of member galaxies .
the comparison between the stages , orelse and other similar studies and shass will be then unavoidable and useful to trace the evolution with redshift , although we notice that the area ( in mpc@xmath2 ) of shass is a factor 10 that of the stages survey and was chosen to map the filaments connecting the abell clusters .
the optical survey vst - access in four bands collected at the eso vlt survey telescope ( vst ) represents the core of this multi - band project . the infrared coverage with wise , the dedicated spectroscopic survey with aat / aaomega and the shapley - vista survey , together with other proprietary data , provide the fundamental data - set to achieve the scientific goals of shass .
in this article we will give an overview of the project and present the first results . in sect .
[ objec ] the motivations and main scientific objectives of the project are discussed .
the target is described in sect .
the characteristics and strategy of the survey are described in sect .
[ survey ] .
details of the data reduction and analysis of the data quality are given in sects .
[ dr ] and [ da ] , respectively .
the first results concerning the characterization of the environment are presented in sect .
[ results ] , where we derive the galaxy density across the whole shass region and the underlying dark matter distribution for the central 1deg@xmath2 field which allows to estimate the mass of the galaxy cluster a3558 . in sect .
[ sum ] we present the summary and conclusions . throughout the paper
we adopt a cosmology with @xmath23=0.3 , @xmath24= 0.7 , and h@xmath25=70kms@xmath26mpc@xmath26 . according to this cosmology
1arcsec corresponds to 0.941kpc at @xmath27=0.048 and the distance modulus is 36.66 .
the magnitudes are given in the ab photometric system .
in @xmath28cdm cosmological models structure formation occurs hierarchically , such that the most massive halos corresponding to galaxy clusters form latest , doubling their masses on average since @xmath29 @xcite , and are also the most dynamically immature .
preferentially located at the nodes of the complex filamentary web , clusters are continually accreting dark matter halos containing individual galaxies or galaxy groups . on average , 50% of galaxies in local clusters
have been accreted since @xmath10 @xcite , of which 40% are within groups @xcite .
the process of central galaxies becoming satellites , as their host halos are accreted into more massive halos , has been shown to strongly affect their evolution , with satellite galaxies more likely to be quenched than central galaxies of the same stellar mass ( e.g. * ? ? ? * ) .
star formation is suppressed within galaxy groups , with the fraction of star - forming galaxies declining steadily with increasing group mass ( at fixed stellar mass ) and proximity to the group centre @xcite .
galaxies which are quenched within such groups and later accreted into clusters are described as having been ` pre - processed ' .
numerical simulations and theoretical studies predict the effect of cluster - cluster mergers on galaxy properties .
@xcite claimed that merging of galaxy clusters may induce sf due to the increase of the external pressure of the icm compressing the cold gas in the cluster galaxies .
this results in a population of starburts or post - starburst galaxies having the same age and so dating the merging event .
the spatial distribution of these star - forming galaxies is expected to differ from the overall distribution of the other cluster galaxies .
similarly , for comparable group and cluster halo densities gravitational shocking as the group enters the cluster could temporarily increases its mass and pull the group members into a denser , more compact system inducing galaxy mergers @xcite
. simulations also showed that starbursts induced by galaxy mergers can be amplified by a factor @xmath30 @xcite if they take place in the tidal field of cluster / group of galaxies .
@xcite studied the excess population of radio galaxies in the cluster a2125 at @xmath27=0.25 , mostly located in groups outside the cluster core and with radio luminosities indicating the sf as the main mechanism responsible of the radio emission .
the authors related this observational evidence with the ongoing major cluster - cluster merger and explained the increased sf as due to the variation of the tidal field experienced by the member galaxies during a cluster - cluster interaction , probably being close to the core passage . studying a system of merging clusters , @xcite observed an excess of star - forming galaxies aligned along the bridge of galaxies connecting a3158 to the a3125/a3128 complex , suggestive of merger - induced sf .
in addition , they found that the fraction of radio - loud sources is lower with respect to a global cluster environment and similar to that measured by @xcite in the cluster a3558 which is considered as an example of a late merger stage .
they suggested that in the cluster cores radio emission from active galactic nuclei ( agns ) is suppressed in the late stages of cluster mergers .
the suppressed radio emission observed in both bright and faint cluster galaxies has been also associated to cluster merger by @xcite .
they explained their findings with the disruption of cool cores , i.e. the gas supply to the central agn for the bright galaxies , while the low radio loudness and fainter cluster galaxies have been affected by the enhanced ram pressure when crossing the shock front between the merging clusters .
further observational support of the important role that cluster mergers play in triggering the evolution of cluster galaxies is provided by @xcite studying the merging cluster a2744 .
they identified three rare ` jellyfish ' galaxies located in close proximity to the icm features associated with a merging subcluster and its shock front .
their interpretation is that the sf knots detected in the tails of stripped gas are due to the rapid increase in pressure experienced by the galaxies interacting with the shock .
hence , the observations seem to confirm that galaxy / group accretion and cluster - cluster merger affect the cluster galaxies and that the galaxy properties can be sensitive indicators of the merger stage .
the colour - density relation is observed in galaxy clusters well beyond @xmath32 ( e.g. * ? ? ? * ) and the fraction of star - forming galaxies steadily increases with cluster - centric radius . nevertheless even at 3 - 5@xmath33 r@xmath31
, the fraction of star - forming galaxies is below that seen in the field ( e.g. * ? ? ?
* ) suggesting that the environment plays a role inducing transformations well outside @xmath32 . by means of hydrodynamical cosmological simulations @xcite investigated the trend of cold and hot gas contents and sf as function of clustercentric distance in clusters and groups of galaxies .
they found a large - scale trend with cold and hot gas contents and sf increasing with the clustercentric distance , but approaching the values of the field galaxy sample only at @xmath34 ( corresponding to 10mpc for a massive cluster ) .
moreover , the sf in low - mass cluster galaxies ( m=10@xmath35m@xmath36 ) still shows a significant discrepancy even at this large radius with respect to that of the field galaxies . according to the authors , three main causes
can explain the observed trends : i ) ` pre - processing ' of galaxies within infalling groups ; ii ) ` overshooting ' for those galaxies that are not falling in for the first time ( also called ` backsplash ' galaxies ) ; iii ) ram - pressure stripping .
of course , the three mechanisms act differently depending on the galaxy and host halo mass and the clustercentric distance , but in any case the radial trend of the hot gas can not be explained solely by overshooting and pre - processing . a direct interaction with the host group / cluster is also required and their simulations suggest that ram - pressure stripping can strip the hot gas from low- and high - mass galaxies out to several times @xmath32 , although it can not affect the cold gas except for the low - mass galaxies .
the authors also claimed that tidal interaction does not play an important role in the gas removal , but it should be an important factor in the morphological transformation of these galaxies .
s0s differ from normal spirals by their higher bulge luminosities rather than fainter disks @xcite , disfavouring simple rps or starvation mechanisms ( but see * ? ? ?
instead , mechanisms such as merging or harassment are capable of channelling material to a central bulge , sufficient to produce the higher central mass densities seen in cluster spirals and ultimately the stellar phase densities found in s0s @xcite .
this channeling of material will also likely fuel rapid growth of the central supermassive black hole and trigger a period of nuclear activity .
deep observations are needed to reveal tidal streams or ` fans ' , characteristic of recent merging activity @xcite . in - situ transformation via merging in clusters is strongly suppressed due to the high encounter velocities .
low - velocity encounters and mergers should be much more frequent in galaxy groups , leading to the suggestion that the bulge growth required to form s0s occurs primarily via pre - processing in galaxy groups , which are subsequently accreted into clusters @xcite .
moreover , the s0 fractions of @xmath37 groups match those seen in clusters at the same redshifts , and are much higher than found in the field @xcite .
strong nuclear activity which is linked to bulge growth is also found to be effectively suppressed in cluster galaxies , with the distribution of x - ray agn in rich clusters revealing them to be an infalling population with high relative velocities @xcite .
this confirms that the transformation of galaxies often occurs ( or begins at least ) well outside the cluster cores , in groups or as infalling galaxies enter the virialized region for the first time @xcite .
hints about the distribution of galaxy luminosity , colours and morphology in the filaments have been given by @xcite , who found that brighter galaxies have a less filamentary distribution than the fainter ones .
with respect to the morphology , the early - type galaxies are concentrated in the vicinity of the nodes , while spiral galaxies are sparsely distributed across the filaments .
@xcite observed that the fraction of starburst galaxies in the filaments around the cluster a1763 is twice that in other cluster regions as detected by _
spitzer _ and suggested that filaments are a sort of _ galaxy reservoir _ for clusters . by studying a sample of supercluster filaments in the two - degree field galaxy redshift survey ,
@xcite observed a sudden enhancement in sf in faint dwarf galaxies outside the cluster virial radius .
they interpreted their findings with a close interaction and/or harassment with other infalling galaxies along the same filaments .
another possible explanation is the very first interaction that these galaxies experienced with the icm inducing tidal shocks and then a burst of sf .
actually , the dominant processes that quench sf in galaxies depend crucially on the galaxy mass .
dwarf galaxies , given their shallow potential wells , should be more susceptible to environmental processes such as tidal / ram - pressure stripping ( e.g. * ? ? ?
indeed their sf histories are completely defined by their local environment , since passive des are only found as satellites within massive halos ( i.e. cluster , group or massive galaxy ; * ? ? ?
* ; * ? ? ?
the formation of many cluster des is often rather recent , manifest by young stellar ages ( @xmath38gyr ) and the significant populations of starburst and post starburst dwarf galaxies in the outskirts of local clusters @xcite .
ram - pressure stripping ( rps ) , as originally proposed by @xcite requires , in principle , the presence of a dense icm .
thus , its effect would be limited to cluster cores where the gas discs of massive spirals are rapidly truncated .
nevertheless , with a 3d hydrodynamical simulation , @xcite showed that rps may extend to poorer environments for low - mass galaxies which , thanks to their lower escape velocities , are easier to strip .
@xcite , by means of high resolution 2d hydrodynamical simulations , demonstrated that ram pressure effects can be observed over a wide range of icm conditions .
in particular , in high density environments rps severely truncates the gas disc of l@xmath5 galaxies , while in low density environments , where moderate ram pressure is foreseen , their gas disc is clearly disturbed and bent @xcite .
the gas discs of these galaxies can be truncated to 15 - 20kpc in the first 20200myr of rps .
@xcite identified a bright ( l@xmath39l@xmath5 ) barred spiral galaxy 1mpc from the centre of the rich cluster a3558 in the shapley supercluster core , which is strongly affected by rps .
integral - field spectroscopic observations revealed ongoing gas stripping in the form of one - sided extraplanar ionized gas along the full extent of the disk , simultaneously with a starburst triggered by gas compression along the leading edge of the galaxy .
the galaxy is estimated to be being subjected to weak - moderate ram pressure , as defined by @xcite .
this adds a piece of evidence to the fact that rps is acting more efficiently on the galaxy interstellar medium ( ism ) than previously foreseen and also outside of the cluster cores as also observed in the virgo cluster by @xcite .
this new understanding of the rps supports the view that this mechanism is the principal transformation process to quench sf in spirals although probably helped by other processes affecting the structure of the galaxies .
@xcite drew a similar conclusion analysing a sample of 182 disk galaxies in the cluster system abell901/902 .
the fraction of galaxies showing asymmetric gas rotation curves , and thus probably affected by rps @xcite , turned out to be 75 per cent higher in the clusters than in the field , the majority of them being morphologically undisturbed .
although these galaxies seemed preferentially located at low clustercentric radii , they also observed a population of dusty and red spirals in the cluster outskirts , which could be also affected by ism - icm interactions .
the relevance of rps for quenching sf in cluster galaxies is further complicated by the fact that a non - homogeneous icm is expected for non virialized merging and post - merging clusters , thus the clustercentric distance is not the only parameter describing the icm density , i.e. the ram - pressure strength .
our study is based on characterizing a dynamically active environment in a multifaceted way and disentangling its effects on galaxy evolution .
it has the following main objectives .
- : : to investigate the role of cluster - scale mass assembly on the evolution of galaxies mapping the effects of the environment in the cluster outskirts and along the filaments with the aim to identify the very first interactions between galaxies and their environment .
- : : to identify and measure signs of ongoing transformation in galaxies belonging to a complex structure with the goal of improving our comprehension of what drives their star - formation quenching and structural modification .
- : : to obtain detailed maps of the dark matter and baryonic matter distributions ( galaxies , icm ) , combining weak lensing , x - ray and dynamical analyses .
- : : to quantify the variation in the stellar mass fractions going from cluster cores to groups , by comparing the near - infrared light distribution with the dark matter maps and dynamical masses .
- : : to build up a multi - band homogeneous data - set on this area of the sky made of sub - kiloparsec resolution imaging and magnitude - limited spectroscopy , thus providing the community with a solid background for studies of the shapley supercluster . to address the above objectives we will explore the global and internal properties of galaxy populations extending outside the cluster / group virial radius and aim for an accurate characterization of the environment .
this will be defined through galaxy density , dark matter distribution , dynamical substructure , and icm properties .
the different quantifications of the environment will allow us to disentangle the effects of local and large - scale density , cluster and group merging , dynamical state and mass of the host system on the properties of galaxies in different ranges of mass .
recent studies investigated the effects of the local and large - scale environment on galaxy evolution , mainly considering groups of galaxies .
some of these works defined the environment through the luminosity - density only ( e.g. * ? ? ?
* ) , or associate galaxy groups to a large - scale environment identified by a smoothed luminosity field ( e.g. * ? ? ?
* ) , or are limited to massive galaxies ( e.g. * ? ? ?
therefore , although there is a general agreement that the group environment is important for the evolution of the group members , a study of global and internal properties of galaxy population in groups and , in general , in different structures dynamically bound in a supercluster is still lacking .
for instance , @xcite found that both local ( distance from the centre ) and global ( mass ) group environments play a role in quenching star formation , while @xcite show that global group - scale mechanisms linked to the presence of a hot gas halo are dominant in quenching sf in group galaxies rather than purely density related processes . in the following we briefly outline how we will derive the quantities necessary for our study
( see also sect .
[ survey ] ) . as the main classification scheme of galaxies we will adopt the concentration , asymmetry and clumpiness parameters @xcite , complemented by the m20 and gini coefficients @xcite .
this parameter set ( hereafter ` cas+mg ' ) has proven to robustly link the internal light distribution of galaxies to their formation and evolution ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
besides being effective in separating the different morphological types and tracing star formation , the cas+mg scheme is particularly sensitive to their recent interaction or merging activities , making it the ideal tool to obtain a census of galaxies whose structure appears disturbed by the environment , which is crucial for our project . for a selected sample of ` normal ' galaxies ( as determined through the cas+mg parameters )
it will be possible to quantify the relevance of ( pseudo-)bulges and bars as a function of the environment by surface photometry fitting . in fig .
[ gals ] we show examples of ssc galaxies in the magnitude range from m*-1 to m*+4 ( @xmath40mag ) . for galaxies in this magnitude range
we plan to perform classical ( visual ) morphological classification , that will be used to calibrate the cas+mg parameters against the hubble sequence for our data .
the signal - to - noise ratio ( snr ) and spatial resolution of vst images will instead allow us to obtain reliable cas+mg parameters up to m*+6 ( @xmath41mag ) , i.e. well into the dwarf galaxy regime ( see sect .
[ optical ] ) .
finally , aaomega and literature spectra of more luminous galaxies will provide spectral classification through line indices .
galaxy global properties such as colours , stellar masses and star - formation rates will be derived .
the contribution of both unobscured and obscured sfr need to be considered since dust - obscured cluster galaxies are common in merging clusters ( e.g. * ? ? ? * ) .
the @xmath42band luminosity will provide a star - formation rate ( sfr ) indicator once calibrated using the multi
wavelength data already available in the shapley supercluster core ( see sect . [ cd ] ) , by obtaining correlations between luminosity , dust extinction , and metallicity .
the available mid - infrared data will provide a robust and independent indicator of obscured star formation ( sect . [ cd ] ) , and also allow us to identify agn via their unusually red w1-w2 colours ( e.g. * ? ? ? * ) .
agn will also be identified from their emission - line ratios or broad emission lines from the aaomega spectra , or their x - ray emission in the supercluster core where xmm data are available .
the depths of our survey are conceived to reach the necessary accuracy of all of the above quantities , as will be outlined in the next sections .
finally , we will not only measure statistically the environmental effects on galaxy properties on such large scales , but also to ` catch in the act ' the direct interaction of supercluster galaxies with their environment into the surrounding large - scale structure .
the latter is possible only using imaging with a sub - kiloparsec resolution and follow - up integral - field spectroscopy for a few individual cases ( e.g. * ? ? ?
the ` remote cloud of galaxies in centaurus ' first identified by @xcite is one of the richest superclusters in the nearby universe , consisting of as many as 25 abell clusters @xcite in the redshift range [email protected] @xcite .
the first spectroscopic study confirming the existence of the ` centaurus supercluster ' was carried out by @xcite , then it was re - discovered by ( * ? ? ?
* their @xmath44-region ) as a cluster overdensity in the ( * ? ? ?
* aco ) cluster catalogue and identified by ( * ? ? ?
* his shapley concentration ) as an excess in galaxy number counts in the uk schmidt telescope sky survey plates .
the shapley supercluster ( hereafter ssc ) because of its peculiar richness and location , lying in the direction of the dipole anisotropy of the cosmic microwave background ( cmb ) , was investigated as responsible of at least a fraction of the local group acceleration ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
to assess the role of this structure in the acceleration of the local group a robust estimate of its mass is required .
this gave a boost to spectroscopic and x - ray observations of the supercluster devoted to i ) map the whole structure and measure the density contrast in galaxy number counts and mass / number of associated clusters ; ii ) investigate its dynamical state and iii ) derive the underlying mass distribution .
a detailed review of the properties of the shapley supercluster based on previous investigations will be given elsewhere ( merluzzi et al .
, in preparation ) , here we will describe the key characteristics of the ssc - i.e. those features making the ssc , and above all its central region , _ a magnifying glass _ to investigate the effect of the environment and mass assembly on galaxy evolution . by analysing a spectroscopic sample across a region of 15deg in diameter
centred on the dominant cluster a3558 , @xcite concluded that the supercluster has a ` cigar - shape with the eastern side being the closest to us ' ( but see also * ? ? ?
* ) and from the flattened geometry they suggested that it is not spherical and virialized .
the complex morphology of the ssc comprises a main body at @xmath4515000kms@xmath26 together with walls / filaments / arms of galaxies connecting the three main systems of interacting clusters ( the a3558 , a3528 and a3571 complexes ) as well as a foreground structure connecting the ssc to the hydra - centaurus supercluster ( @xmath454000kms@xmath26 ) and hints of a background structure at @xmath4523000kms@xmath26(for details see * ? ? ?
* ; * ? ? ?
the main plane of the ssc ( @xmath4514800kms@xmath26 ) extends 10x20deg@xmath19 ( see * ? ? ?
is this vast and complex structure gravitationally bound ? in this case it would be the most massive bound structure known in the universe . applying a spherical collapse model
, @xcite found that the ssc is gravitationally collapsing at least in its central region within a radius of 8@xmath46mpc centred on a3558 including 11 aco clusters ; the very inner region , associated with the massive clusters , is likely in the final stages of collapse .
x - ray observations of clusters in the ssc confirmed the overdensity : a factor 10 to 50 cluster overdensity @xcite and factor 3 baryon overdensity @xcite over a region of 60x80mpc@xmath19 .
these results indicated that the structure is gravitationally bound at least in its central region . from the rosat all - sky survey in a region of 0.27sterad
, @xcite measured the cluster number density which turned out to be more than an order of magnitude greater than the mean density of abell clusters at similar latitudes - mainly due to an excess of low - luminosity x - ray systems in the outskirts , suggesting that the supercluster is still accreting low - mass systems .
@xcite analysing a mosaic of 15x20deg@xmath19 of rosat and einstein observatory x - ray observations centred on a3558 , concluded that the ssc core is dynamically bound within @xmath47mpc , approaching the maximum expansion before collapsing , i.e. the turnaround point . on the other hand
, the dynamical analysis of @xcite proposed a scenario where the ssc has already reached its turnaround radius and the final collapse will happen in @xmath48gyr . using the x - ray cluster sample of @xcite and the spherical collapse model , @xcite investigated the dynamics of the ssc .
their study showed that the ssc is not bound at a radius of 51mpc with an outer shell moving radially away from the centre , while the excess of mass becomes enough to bind a spherical region of @xmath010mpc radius .
recently , @xcite ran @xmath17-body numerical simulations of the ssc and other superclusters in order to determine what portions of the superclusters were potentially gravitationally bound .
the ssc showed the most extended bound structure among the other analyzed superclusters .
in particular , a3554 , a3556 , a3558 , a3560 and a3562 have a large number of close encounters in their simulations , while an additional pair , a1736 and a3559 , has also some chance of being bound . in such a crowded environment
it is very unlikely that galaxies have not been affected by cluster related processes such as rps , starvation and/or galaxy harassment and tidal interaction .
one of the key objectives of shass is to constrain the dark matter distribution and mass over this whole region via a weak lensing analysis , and in conjunction with x - ray and dynamical analyses estimate the extent of the region that is gravitationally bound to the ssc core , and determine whether clusters such as a3559 are currently falling towards the supercluster core .
we now zoom in to the very central dense region of ssc , namely the ` core ' , which consists of three abell clusters : a3558 , a3562 , and a3556 and two poor clusters ( sc1327 - 312 and sc1329 - 313 ) , resolved by @xcite by means of x - ray observations .
the evolutionary stage of the ssc core ( sscc ) is still matter of debate . [
cols="<,<,<,<,<,^,^,^,<,<,<,<",options="header " , ] [ zp ]
the vst images have been processed using the vst tube imaging pipeline @xcite , developed for the data produced by the vst . after applying the overscan correction and bias subtraction , we divide by the master flat - a normalized combination of the dome and twilight flats , in which the twilight flat is passed through a low - pass filter first . due to differences between the electronic amplifiers ,
the ccds do not have the same gain levels . in order to have the same photometric zeropoint ( zp ) for all the mosaic chips a gain harmonization procedure has been applied .
the procedure finds the relative ccd gain coefficients which minimizes the background level differences in adjacent ccds .
a further correction is needed for the light scattered by the telescope and instrumental baffling .
the resulting uncontrolled redistribution of light adds a component to the background and the flat field will not be an accurate estimate of the spatial detector response . indeed , after flat - fielding , the image background will appear flat but the photometric response will be position - dependent .
this error in the flat - fielding can be mitigated through the determination and application of the illumination correction ( ic ) map .
the ic map is determined by comparing the magnitudes of photometric standard fields with the corresponding sdss dr8 psf magnitudes .
the differences between the vst and sdss magnitudes are plotted vs. x and y axis in fig .
[ ic ] before ( top panels ) and after ( bottom panels ) the ic .
usually , the same ic map can be used for observations carried out on the time scale of a month . in the case of the _
i _ band it is required to correct for the fringe pattern due to thin - film interference effects in the detector of sky emission lines .
also this is an additive component that must be subtracted .
the fringing pattern is estimated using the @xmath49 ratio where _ superflat _ is obtained by overscan and bias correcting a sigma - clipped combination of science images and _ tflat _ is the sky flat . to this aim
, the dithering amplitudes of the science frames used must be larger than the extended object sizes in the same images .
this is mandatory to allow the sigma - clip procedure to efficiently remove such objects .
the difference between the above ratio across the image and its model obtained with a surface polynomial fit is the fringing pattern .
the fringe pattern is subtracted from the image applying a scale factor which minimizes the absolute difference between the _ peak _ and _ valley _ values ( maximum and minimum in the image background ) in the fringe corrected image .
the relative photometric calibration among the exposures contributing to the final mosaic image is obtained comparing the magnitudes of bright unsaturated stars in the different exposures , i.e. minimizing the quadratic sum of differences in magnitude using scamp @xcite .
the absolute photometric calibration is computed on the photometric nights comparing the observed magnitude of stars in photometric standard fields with sdss photometry . for those fields that are not observed in photometric but clear nights ,
we take advantage of the sample of bright unsaturated stars in the overlapping regions between clear and photometric pointings and , by using scamp , each exposure of the clear fields is calibrated onto the contiguous photometrically calibrated field . in table [ zp ] we give examples of the fitted values for the zp and colour term obtained using the photcal tool @xcite . in several cases , the photometric standard star fields
were observed with insufficient span in airmass to do a suitable fit .
the extinction coefficient was then taken from the extinction curve provided by eso .
the errors listed in table [ zp ] are those of the fit .
the actual errors for the zp is given by the r.m.s .
of the detection among the different exposures of each pointing and turns out to be less than 0.03mag in all bands .
-band magnitudes of stars in a standard field observed with vst and sdss magnitudes vs. x and y image axis .
_ bottom _ : the same @xmath50mag after the ic correction has been applied.,width=317 ] the relative astrometry among the pointings is derived using scamp , while for the absolute astrometry we refer to the astrometric catalogue 2mass obtaining @xmath51arcsec astrometric accuracy in all bands .
the image resampling , application of astrometric solution and co - addition is performed using the tool swarp @xcite which produces the final stacked image with a weighted average . at the stage of the co -
addition the scale factors for relative and absolute photometric calibration are applied .
once the mosaic images were obtained , we proceeded with the catalogue extraction and measured the photometric accuracy and completeness .
details of shass catalogue extraction and release will be given in a companion article ( mercurio et al . in preparation ) .
we used the software package sextractor @xcite estimating the background locally and using a gaussian filter for source detection . for each source
we measured magnitudes in different apertures as well as kron and psf magnitudes .
to distinguish between stars and galaxies , we adopted a progressive approach @xcite using the following parameters provided by sextractor : i ) the stellarity index to select point - like sources ; ii ) the half - flux radius as a measure of source concentration ; iii ) the new sextractor parameter which takes into account the difference between the model of the source and the model of the local psf ; iv ) the peak surface brightness above background ; v ) a final visual inspection for objects with ambiguous values of the stellarity index .
we estimated the completeness magnitudes using both trends of the galaxy number counts and the method by @xcite .
the catalogues turned out to be 100 per cent complete at the total magnitudes of 23.9 - 24.1 , 23.8 - 24.0 , 23.3 - 23.5 , 21.8 - 22.0mag in @xmath3 bands , respectively .
the ranges of magnitudes are due to small differences in seeing among the vst fields .
we measured the snr inside a 3arcsec aperture as function of magnitudes achieving snr=20 at @xmath3 at 22.8 , 23.3 , 22.8 , 21.7mag , respectively .
snr@xmath0100 is reached at 21.1mag in @xmath52 band ( m@xmath5 + 6 ) as required .
these depths enable us to study the galaxy population down to m@xmath5 + 6 .
.,width=302 ] in this section we derive the first quantitative characterization of the environment in the survey area in terms of stellar mass density
. we also derive the dark matter distribution in the very centre of the sscc including the cluster a3558 and the poor cluster sc1327 - 312 . to map the structure of the supercluster , and determine its extent in redshift space and across the plane of the sky , we take advantage of our redshift survey which allows us to demarcate the supercluster in redshift space as lying within the recession velocities 1130017000kms@xmath26 ( see fig . [ histo ] ) .
these cuts select a supercluster sample of 2281 galaxies across the shass area .
each galaxy was weighted ( @xmath53 ) according to the inverse probability of it having been observed spectroscopically .
firstly , each galaxy which could have been targeted for spectroscopy ( @xmath54 , @xmath55 ) or was a spectroscopic member of the supercluster , was given an initial equal weight of 1.0 . for each of these galaxies lacking a redshift ,
its weight was transferred equally to its ten nearest neighbours with known redshift that also had the same priority level in our aaomega spectroscopic survey .
this results in galaxies without redshifts having zero weight , while galaxies in regions where the spectroscopic survey is locally 50@xmath56 complete having weights of 2.0 .
the transferring of weight only within priority levels ensures that we can account statistically for the systematic differences in spectroscopic completeness from one level to another , as well as mapping the local spatial variations in completeness .
each galaxy @xmath57 belonging to the supercluster is represented by a gaussian kernel whose transverse width is iteratively set to @xmath58^{-1/2}$ ] , where @xmath59 is the geometric mean of the @xmath60 , and a fixed radial width of 700kms@xmath26 .
each galaxy is normalized by the weight parameter @xmath53 to account for spectroscopic incompleteness .
the transverse kernel width for each galaxy is initially set to @xmath61arcmin , and then iteratively adjusted to account for variations in the spatial density of galaxies , such that it typically encloses the 5 to 10 nearest neighbours of the galaxy , irrespective of its location within the supercluster .
figure [ d_map ] shows the resulting density map in which each galaxy is further weighted by its w1 flux as a proxy for its stellar mass ( e.g. * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
the density map draws our attention to several features of the whole structure .
a ) : : across the sscc the density of supercluster galaxies is always higher than 315mjydeg@xmath62 .
this corresponds to @xmath63 based on the w1stellar mass calibration obtained for the ssc galaxies ( see end of sect .
[ aao ] ) . assuming a depth of 5000kms@xmath26 or 71mpc the stellar mass density is @xmath64 .
both the stellar mass estimates include a 12% correction to account for the low - mass galaxies with @xmath65 or w@xmath66 not targeted in our spectroscopic survey , based on the stellar mass function of @xcite .
this is a mean overdensity of @xmath7 with respect to the cosmic total stellar mass density for galaxies in the local universe ( @xmath67 ) estimated by @xcite from the gama survey @xcite .
we obtain a similar overdensity of 3540@xmath68 if we simply compare the observed w1 flux density of sscc galaxies with the average w1 flux density of @xmath69 galaxies within the same redshift range selected from the sdss dr7 ( 8032deg@xmath19 ) .
although the sdss is @xmath52-band selected , it should be complete to w1=14.5 at @xmath20 , and only marginally incomplete to w1=15.0 .
the higher density peaks in the sscc , corresponding to the cluster cores but also the groups and cluster sub - structures , are interconnected forming a single system .
this implies that the galaxy distribution follows the hot gas distribution observed across the whole sscc ( e.g. * ? ? ?
* ) . + .
red diamonds ( @xmath70 ) and blue circles ( @xmath71 ) denote the correlation function for corrected galaxy ellipticities , @xmath72 and residual stellar ellipticities , @xmath73 after correcting the psf anisotropy .
green triangles ( @xmath70 ) and magenta squares ( @xmath71 ) are the correlation function for raw galaxy ellipticities , @xmath74 and raw stellar ellipticities , @xmath75 before the correction .
the separation radii are offset by @xmath76 . ]
( b ) : : there is clear evidence of a filament ( @xmath07mpc in projection ) heading north from the sscc , connecting it with a3559 , with w1 flux densities of @xmath7750mjydeg@xmath62 ( @xmath78 overdensities ) .
@xcite qualitatively indicated a ` broad arm running north ' from the sscc , but in their description this feature should point to a3557a which is located nw of the shass region . in this direction
we do not detect any clear overdensity .
c ) : : several overdensities are detected across the density map .
some of them can be associated to cluster dynamical substructures already identified by ( * ? ? ?
* see fig .
[ d_map ] ) , but other density peaks are also detected , i.e. that w of a3556 .
the positions of the system centres as defined by x - ray peak and galaxy overdensity appear slightly different , as expected in dynamically active systems .
the highest discrepancy between these determinations is found for as0724 , where the x - ray centre is located 35arcmin from the galaxy overdensity centre .
notice that as0726 was not detected by rosat and its centre was derived by a previous spectroscopic survey @xcite , likely affected by incompleteness .
our newly detected overdensity is actually located 16arcmin from the previous centre .
d ) : : in a3560 two main density clumps are sharply defined together with a western clump .
the northern clump can be related to the x - ray emission , elongated towards a3558 , detected in the very centre of the cluster which suggested a minor merger scenario between the main cluster and a group located @xmath79arcmin n of the cluster centre @xcite .
also the galaxy distribution shows an asymmetric shape elongated in the n direction , but extending a factor two in projection with respect to the x - ray emission ( c.f .
fig . 5 of * ? ? ?
* ) , taking into account only the two denser clumps in our map . thus the group involved in the claimed merger could be significantly more massive than that previously identified in the apm catalogue . in a companion paper
, we will assign galaxies to supercluster structures / overdensities . this will be achieved through the dynamical analysis enabling us to detect and measure the amount of cluster dynamical substructures , and to identify possible pre - merging clumps or merger remnants . in particular , we will i ) select cluster members ( e.g. * ? ?
* ; * ? ? ?
* ) ; ii ) derive the centre of the most significant peaks of each identified system applying the 2d adaptive kernel technique to galaxy positions ; iii ) detect possible subclumps and assign objects to groups @xcite ; iv ) identify the 3d substructures combining velocity and position information @xcite .
we conducted a weak lensing analysis of the vst - access field 8 ( see fig .
[ vst - access ] ) following @xcite as modified by @xcite .
we measure image ellipticity of objects detected in the @xmath52-band data , @xmath80 , from the weighted quadrupole moments of the surface brightness
. we then correct a psf anisotropy of galaxy ellipticities by the function of second - order bi - polynomials of the stellar anisotropy kernel .
a cross correlation function of residual stellar ellipticities and the corrected galaxy ellipticities does not show overcorrection or insufficient correction as shown in fig . [
fig : xi_ge ] .
then , a reduced distortion signal , @xmath81 , is estimated with a correction of isotropic smearing effect .
we select background galaxies following @xcite and adopt a new method for minimizing the contamination of member galaxies extending to the colour - colour plane .
we measure two colours , @xmath82 and @xmath83 , and fit the red sequence with a linear function .
a colour offset is defined for each galaxy by @xmath84 , where @xmath85 and @xmath86 and ` es0 ' denotes the red sequence galaxies .
we select background galaxies by @xmath87 which is the lower limit to allow @xmath88 contamination level .
the number density of background source galaxies is @xmath89 , and their mean redshift is @xmath90 .
the mass map is reconstructed as described in @xcite .
the reduced shear is pixelized into a regular grid with a gaussian smoothing , @xmath91 $ ] .
the fwhm resolution of the map is 11.7arcmin .
the smoothed shear pattern is estimated with a gaussian kernel , @xmath92 , and a statistical weight @xmath93 for the @xmath94th galaxy , where @xmath95 is the rms error of the shear estimate and the softening constant variance .
we choose @xmath96 as a typical value of the mean rms of @xmath95 .
we next invert the smoothed shear field with the kernel @xcite in fourier space to obtain the projected mass distribution .
the resulting mass map , @xmath97 , for field 08 of vst - access is shown in fig . [ fig : wlmap ] . the mass map is elongated along the east - west direction which is parallel to the large - scale filamentary structure . a main peak is associated with distributions of member galaxies and diffuse x - ray emission of a3558 .
another possible clump , located e of a3558 , is likely associated with diffuse x - ray emission of the galaxy group sc1327 - 3136 .
we compute the tangential distortion component , @xmath98 , with respect to the projected cluster - centric radius from the bcg of a3558 , shown in fig .
[ gt_profile ] .
the lensing signal @xmath99 is decreasing as the radius , @xmath100 increases , and becomes flat or increases at @xmath101arcmin .
to understand this feature , we fit the two - dimensional shear pattern with two mass components : a3558 and the east clump ( e.g. * ? ? ?
we use the navarro frenk & white ( nfw ) profile @xcite as the mass model represented by two parameters : @xmath102 which is the enclosed mass within a sphere of radius @xmath103 , and the halo concentration @xmath104 . here
, @xmath103 is the radius inside of which the mean density is @xmath50 times the critical mass density , @xmath105 , at the redshift @xmath27 . since it is difficult to constrain the concentration for the east clump , we assume a mass - concentration relation @xcite .
the centre of a3558 is fixed at the bcg position , while the position of the secondary mass component is treated as a free parameter .
the model is described by five parameters in total .
the best - fit m@xmath106 for a3558 is @xmath107m@xmath108 .
the total lensing signal of the two components ( solid line in fig . [ gt_profile ] ) well describes the observed lensing signals .
since a3558 and the east clump are embedded in the sscc , surrounding data are essential to further constrain the mass .
, as a function of the projected cluster - centric radius from the bcg , is estimated by azimuthally averaging the measured galaxy ellipticities . a bump in lensing signals
is found around @xmath109 .
the profile is well described by two nfw components of a3558 and the east clump . the red solid , green dashed and blue dotted lines are the best - fit nfw profile for the total mass , a3558 and the east clump , respectively . _
bottom panel : _ the @xmath110 degree rotated component , @xmath111 , is consistent with a null signal . ]
although this is a preliminary result , the derived dark matter mass can be compared with that evaluated from the xmm - newton data - set . to this aim , we computed the y@xmath112 mass proxy , defined as the product of gas mass m@xmath113 and average temperature @xmath114 @xcite .
as described in @xcite , we inverted a gas mass profile from the radially average surface brightness of a3558 , then iterated about the y@xmath112-m@xmath106 scaling relation calibrated from hydrostatic mass estimates in a nearby sample of clusters observed with xmm - newton @xcite .
note that the surface brightness of a3558 has been evaluated within an angular sector excluding the eastern filament connecting a3558 to sc1327 - 312 .
this yielded estimates of gas mass m@xmath115m@xmath108 , average temperature @xmath116kev , and total mass m@xmath117m@xmath108 within r@xmath118kpc , which is consistent with the wl estimate .
previous x - ray mass estimates based on rosat data gave for m@xmath106 : 8.7@xmath119m@xmath108 @xcite ; 6.09@xmath119m@xmath108 @xcite ; 8.4@xmath120m@xmath108 ( 30% uncertainty , * ? ? ?
the dynamical mass derived by @xcite of a3558 and listed in table [ clusters ] turns out to be a factor 1.7 higher than the wl mass , but also a factor 3 higher than x - ray mass determination previously available as the authors stated in their work .
@xcite computed the dynamical mass inside the virial radius obtaining m@xmath121=6.7@xmath119m@xmath108 .
the shapley supercluster survey ( shass ) aims to assess the role of cluster - scale mass assembly on galaxy evolution searching for possible connections between the properties of the cosmological structures ( density , dynamical status , hot - gas content , dark and luminous matter distribution ) and those of the associated galaxies ( morphology , internal structure , star formation , nuclear activity ) .
this requires that we extend the investigation from the cluster cores to their outskirts , to the infalling galaxies and groups along the filaments in a dynamically - bound network .
the centre ( @xmath03deg radius ) of the shapley supercluster at @xmath122 is the optimal target to undertake such a study enclosing a massive and dynamically active structure showing signs of cluster - cluster mergers , enhancing the probability to observe evidence of environmental effects on galaxy evolution , but also providing an extraordinary variety of environments concentrated in a small survey volume .
shass includes nine abell clusters ( a3552 , a3554 , a3556 , a3558 , a3559 , a3560 , a3562 , as0724 , as0726 ) and two groups ( sc1327 - 312 , sc1329 - 313 ) covering a region of @xmath0260mpc@xmath2 .
the survey includes the following data - sets .
- : : optical ( @xmath3 ) imaging acquired with the vlt survey telescope ( pi p. merluzzi ) provides a galaxy catalogue complete down to @xmath123 ( snr@xmath010 , limit for star / galaxy separation as measured from the collected data ) corresponding to @xmath0m@xmath5 + 8.3 at the supercluster redshift .
the multi
band @xmath124 catalogue is complete to @xmath0m@xmath5 + 7.1 and the @xmath42-band catalogue to @xmath0m@xmath5 + 6.7 .
the achieved snrs allow to study the galaxy population down to @xmath0m@xmath5 + 6 and derive morphological parameters ( cas+mg , see sect .
[ da ] ) in @xmath52 band to the same depth .
the vst - access survey is ongoing .
- : : near - infrared ( @xmath4 ) imaging acquired with the visible and infrared survey telescope for astronomy ( pi c. p. haines ) reaching the depth of @xmath125 ( @xmath0m@xmath5 + 7.9 at the supercluster redshift ) allows to study the galaxy population down to the magnitude limit of the optical catalogue .
the survey started in april 2014 .
- : : the spectroscopic survey with aaomega at the anglo - australian telescope ( pi p. merluzzi ) collected 4037 new redshifts across 21deg@xmath2 of shass . together with the already available redshifts
the spectroscopic sample is now 80% complete down to @xmath126 ( @xmath0m@xmath5 + 3 at the supercluster redshift ) .
- : : the above dedicated surveys are complemented by near / mid - infrared data from the wide - field infrared survey explorer ( wise ) in four bands ( w1-w4 ) 3.4 , 4.6 , 12 and 22@xmath127 m . over the shass area , we reach w1=16.96 ( vega magnitude ) and w3=11.12 mag ( 1.0mjy ) at snr=10 corresponding to @xmath0m@xmath5 + 5 and sfr of 0.46m@xmath108yr@xmath26 , respectively , for galaxies in the supercluster . for the central 2 - 3deg@xmath2 , xmm - newton ,
_ spitzer_/mips 24@xmath127 m and 70@xmath127 m , _ galex _ near - ultraviolet and far - ultraviolet data are also available , as well as targeted observations of single galaxies providing h@xmath44 imaging with the maryland - magellan tunable filter on the magellan - baade 6.5 m telescope at the las campanas observatory in chile and integral - field spectroscopy with wifes on the australian national university 2.3 m telescope at siding spring in australia . in this first article we derived the stellar mass density distribution based on supercluster members weighted by the w1 flux .
this first quantitative characterization of the environment for the whole region covered by shass shows a clumpy structure both in the sscc and the surrounding clusters with several substructures , most of them already identified in previous works .
all the clusters in the shass area are embedded in a common network .
this was suggested , but only for the sscc where supercluster galaxies continuously populate the core following the gas filaments connecting the abell clusters as mapped by x - ray observations @xcite .
we estimate the mean overdensity across the sscc being @xmath7 with respect to the cosmic total stellar mass density for galaxies in the local universe ( @xmath67 ) .
some new substructures with respect to previous works have been identified in the shass density map such as those associated with a3560
n from the cluster centre in the direction of a3558 , towards which also the x - ray emission is elongated @xcite .
the most important new feature is however the filament connecting the sscc and the cluster a3559 as well as the less pronounced overdensity extending from the sscc towards a3560 .
the other environment indicator analyzed here is the dark matter distribution derived from the weak - lensing analysis of vst imaging . using this approach
, we studied the central 1deg@xmath19 field including a3558 and sc1327 - 312 .
the derived wl map shows that the dark matter is concentrated in two peaks which correspond to the rich cluster and the group , although the centres seem slightly offset with respect to the x - ray emission and the galaxy density .
this can be due to the lower resolution of the wl map and/or ascribed to the complex dynamical state of the sscc .
the estimated mass of a3558 is @xmath8 , consistent with the x - ray estimate of m@xmath128m@xmath108 .
we notice that the wl mass determination will be improved extending the analysis to a larger region of the supercluster as planned in the shass project .
we conclude pointing out that the vst imaging is the first ccd photometry covering homogeneously and continuously such a large portion of the shapley supercluster , with this depth and resolution ( corresponding to 0.75kpc at the supercluster redshift ) . taking advantage of the @xmath94-band and wise w1 photometry ,
the aaomega survey has been designed and carried out to obtain a magnitude - limited redshift sample , which was never achieved before ( e.g. * ? ? ?
* ; * ? ? ?
* ) . with these characteristics , shass will build up the first multi - band homogeneous data - set of a vast region of the shapley supercluster and provide a fundamental local counterpart to the supercluster surveys at higher redshifts .
this work was conceived in the framework of the collaboration of the fp7-people - irses-2008 project access .
based on data collected with i ) the eso - vlt survey telescope with omegacam ( eso programmes 088.a-4008 , 089.a-0095 , 090.a-0094 , 091.a-0050 ) and the eso - visible and infrared survey telescope for astronomy with vircam ( eso programme 093a-0465 ) at the european southern observatory , chile and ii ) anglo - australian telescope and 2df+aaomega at the australian astronomical observatory , australia ( opticon proposal 2013a/014 ) .
the optical imaging is collected at the vlt survey telescope using the italian inaf guaranteed time observations .
the research leading to these results has received funding from the european community s seventh framework programme ( fp7/2007 - 13 ) under grant agreement number 312430 ( opticon ; pi : p. merluzzi ) and prin - inaf 2011 ` galaxy evolution with the vlt surveys telescope ( vst ) ' ( pi a. grado ) .
cph was funded by conicyt anillo project act-1122 .
n.okabe is supported by a grant - in - aid from the ministry of education , culture , sports , science , and technology of japan ( 26800097 ) and by world premier international research center initiative ( wpi initiative ) , mext , japan .
pm and gb would like to thank a. m. hopkins for his support during the spectroscopic observations at the australian astronomical observatory and the universidad de chile for the hospitality and support during their staying .
pm thanks m. petr - gotzens for her support in the vst observations .
the authors thank prof .
l. campusano for his helpful comments to the manuscript .
this publication makes use of data products from the wide - field infrared survey explorer , which is a joint project of the university of california , los angeles , and the jet propulsion laboratory / california institute of technology , funded by the national aeronautics and space administration .
the author thank the anonymous referee for her / his constructive comments and suggestions .
b. , mcintosh d. h. , barden m. , bell e. f. , rix h .- w .
, borch a. , beckwith s. v. w. , caldwell j. a. r. , heymans c. , jahnke k. , jogee s. , koposov s. e. , meisenheimer k. , snchez s. f. , somerville r. s. , wisotzki l. , wolf c. , 2007 , apjs , 172 , 615 m. j. , lewis j. , hodgkin s. , bunclark p. , evans d. , mcmahon r. , emerson j. p. , stewart m. , beard s. , 2004 , in quinn p. j. , bridger a. , eds , optimizing scientific return for astronomy through information technologies vol .
5493 of society of photo - optical instrumentation engineers ( spie ) conference series , vista data flow system : pipeline processing for wfcam and vista .
pp 411422 j. c. , gil de paz a. , zamorano j. , boissier s. , dale d. a. , prez - gonzlez p. g. , gallego j. , madore b. f. , bendo g. , boselli a. , buat v. , calzetti d. , moustakas j. , kennicutt jr .
r. c. , 2009 , apj , 703 , 1569 | we present an overview of a multi - wavelength survey of the shapley supercluster ( ssc ; [email protected] ) covering a contiguous area of 260@xmath1mpc@xmath2 including the supercluster core .
the project main aim is to quantify the influence of cluster - scale mass assembly on galaxy evolution in one of the most massive structures in the local universe .
the shapley supercluster survey ( shass ) includes nine abell clusters ( a3552 , a3554 , a3556 , a3558 , a3559 , a3560 , a3562 , as0724 , as0726 ) and two poor clusters ( sc1327- 312 , sc1329 - 313 ) showing evidence of cluster - cluster interactions . optical ( @xmath3 ) and near - infrared ( @xmath4 ) imaging acquired with vst and vista allow us to study the galaxy population down to m@xmath5 + 6 at the supercluster redshift .
a dedicated spectroscopic survey with aaomega on the anglo - australian telescope provides a magnitude - limited sample of supercluster members with 80% completeness at @xmath0m@xmath5 + 3 .
we derive the galaxy density across the whole area , demonstrating that all structures within this area are embedded in a single network of clusters , groups and filaments .
the stellar mass density in the core of the ssc is always higher than @xmath6 , which is @xmath7 the cosmic stellar mass density for galaxies in the local universe .
we find a new filamentary structure ( @xmath07mpc long in projection ) connecting the ssc core to the cluster a3559 , as well as previously unidentified density peaks .
we perform a weak - lensing analysis of the central 1deg@xmath2 field of the survey obtaining for the central cluster a3558 a mass of @xmath8 , in agreement with x - ray based estimates .
[ firstpage ] galaxies : evolution galaxies : clusters : general galaxies : clusters : individual : a3552 , a3554 , a3556 , a3558 , a3559 , a3560 , a3562 , as0724 , as0726 , sc1327 - 312 , sc1329 - 313 galaxies : photometry galaxies : stellar contents gravitational lensing : weak |
in this section , we provide in more detail the proof of proposition 1 , which gives us an upper bound for the truncation error . in order to quantify the error
, we use the trace norm of a matrix , defined as the sum of the singular values of the matrix : @xmath104 .
the following recursion relation holds : @xmath105 where @xmath106 is the ideal solution at time @xmath107 , @xmath108 is @xmath25th order truncation at time @xmath107 , @xmath18 , and @xmath109 , in which the superoperator norm is induced by the trace norm , i.e. @xmath110 . the truncation error can be thus evaluated by induction , by considering the zeroth - order truncation error , @xmath111 a bound on @xmath112 can be found by using a grnwall s inequality .
[ gro1 ] let @xmath113 be a continuous function defined on @xmath114 $ ] and let the function @xmath115 be continuous and nonnegative on the triangle @xmath116 and nondecreasing in @xmath107 for each @xmath117 .
let @xmath118 be a positive continuous and nondecreasing function for @xmath119 .
if @xmath120 then @xmath121 one can prove from the volterra equation that @xmath122 .
theorem [ gro1 ] implies that @xmath123 where we have set @xmath124 . here
, we have assumed that @xmath125 , to satisfy the hypothesis on @xmath126 , in order to apply theorem [ gro1 ] . by plugging eq . into eq . , we find that @xmath127 where in the second inequality
, we have used that @xmath128 for @xmath129 , allowing us to perform the integration . in the third inequality
we have made use of the bound on @xmath130 , @xmath18 .
we can now prove by induction that @xmath131 for any natural @xmath132 .
the case @xmath133 is just the inequality found in eq . .
let us assume that eq
. holds for @xmath134 .
then , @xmath135 which concludes the proof of eq . . in the following
, we will prove that @xmath136 holds , provided that @xmath137 .
indeed , we have that @xmath138 in the first inequality , we have used the lagrange error formula for the taylor expansion of the exponential series . in the second inequality ,
we have used the stirling inequality @xmath139 . in the third inequality
, we have used that @xmath140 . finally , in the last inequality of eq .
( 9 ) , we have used the lower bound on @xmath132 . by applying the last result to @xmath141
, we finish the proof of proposition 1 .
in proposition 3 , we estimate the error made when approximating the equation @xmath143 by the equation which corresponds to the semi - markovian process @xmath144\rho(s),\ ] ] where @xmath145 and @xmath146 , with the same initial condition for both equations .
let us denote by @xmath89 and @xmath90 the solutions to eqs . and , respectively .
considering the corresponding volterra equations , we can upper bound the distance between @xmath89 and @xmath90 , @xmath147}{\lambda } \rho_2(s)\right ) \right\|_1 \\ & & \nonumber = \left\| \int^{t}_{0}ds\ , k(t , s ) \left ( \mathcal{l}\left(\rho_1(s)-\rho_2(s)\right ) -\frac{1}{\lambda } \sum^{\infty}_{i=2 } \frac{(\lambda \mathcal{l})^{i}}{i ! } \rho_2 \right ) \right\|_1 \nonumber \\ & & \nonumber \leq \int^{t}_{0}ds\ , k(t , s ) \|\mathcal{l}\| \| \rho_1(s)-\rho_2(s)\|_1 \\ & & + \frac{\lambda^2 \|\mathcal{l}\|^2 e^ { \lambda \|\mathcal{l}\|}}{2}\int^{t}_{0}ds\ , h(t , s ) \|\rho_2(s)\|_1 , \label{eq3}\end{aligned}\ ] ] where we have used the definitions @xmath148 and @xmath149 . in eq .
, we have used the triangle inequality and , then , the lagrange bound for the taylor series truncation on the last term , i.e. , @xmath150 . as in proposition 1
, we can now bound @xmath151 by using the grnwall s inequality from theorem [ gro1 ] : @xmath152 , with @xmath124 . at this point
, we can bound the second term in eq .
, obtaining @xmath153 here , we have used that @xmath154 , for @xmath129 , and performed the integration . the second term in eq .
is positive and nondecreasing in time , so we can apply the grnwall s inequality from theorem [ gro1 ] , @xmath155 where we have set @xmath156 and we have assumed @xmath157 . finally , let us consider two parameter regimes : \(1 ) _ first regime _ : for @xmath158 , the expression in eq . is bounded by @xmath159 , provided that @xmath160 , @xmath161 where we have used in the first inequality that @xmath162 , in order to apply the inequality @xmath163 ( @xmath164 ) , and the last inequality holds for @xmath165 .
\(2 ) _ second regime _ : for @xmath166 , the expression in eq .
is bounded by @xmath159 , provided that @xmath167 .
in fact , for this parameter choice , we have that @xmath168 , which implies @xmath169 .
hence , the relation @xmath170 holds . here
, we have used in the first inequality that @xmath171 , applying it to @xmath172 , and the last inequality holds for @xmath165 .
any observable @xmath82 can be decomposed as a sum of two unitary matrices @xmath173 and @xmath174 , as @xmath175 , with @xmath176 and @xmath177 @xcite .
the first step is the diagonalization of @xmath82 , @xmath178 , and obtain the equations for @xmath179 and @xmath180 , the eigenvalues of @xmath181 and @xmath174 , as a function of @xmath182 and @xmath183 , the eigenvalues of @xmath82 , as follows : @xmath184 the eigenvalues are decomposed into real and imaginary parts , @xmath185 with @xmath186 , and the unitary matrices obtained , @xmath187 s. trotzky , y .- a .
chen , a. flesch , i. p. mcculloch , u. schollwck , j. eisert , and i. bloch , probing the relaxation towards equilibrium in an isolated strongly correlated one - dimensional bose gas .
* 8 * , 325 ( 2012 ) .
y. salath , m. mondal , m. oppliger , j. heinsoo , p. kurplers , a. potonik , a. mezzacapo , u. las heras , l. lamata , e. solano , s. filipp , and a. wallraff , digital quantum simulation of spin models with circuit quantum electrodynamics , phys .
x * 5 * , 021027 ( 2015 ) .
j. c. loredo , m. p. almeida , r. di candia , j. s. pedernales , j. casanova , e. solano , and a. g. white , measuring entanglement in a photonic embedding quantum simulator , phys .
lett . * 116 * , 070503 ( 2016 ) .
j. s. pedernales , r. di candia , p. schindler , t. monz , m. hennrich , j. casanova , and e. solano , entanglement measures in ion - trap quantum simulators without full tomography , phys .
a * 90 * , 012327 ( 2014 ) .
our concept of algorithmic quantum simulation is not equivalent to the one presented by d. wang in his phd thesis ( d. wang , algorithmic quantum channel simulation , phd thesis , university of calgary , 2015 ) , although both approaches involve a combination of classical processing and quantum simulation . while in the protocol of d. wang ,
the classical part is used for optimizing the design of quantum circuits , in our protocol the classical processing involves the numerical integration of certain functions depending on the measurement outcomes and memory kernels . | we propose a method for the algorithmic quantum simulation of memory effects described by integrodifferential evolution equations .
it consists in the systematic use of perturbation theory techniques and a markovian quantum simulator .
our method aims to efficiently simulate both completely positive and nonpositive dynamics without the requirement of engineering non - markovian environments .
finally , we find that small error bounds can be reached with polynomially scaling resources , evaluated as the time required for the simulation .
_ introduction .
_ fundamental interactions in nature are described by mathematical models that frequently overcome our analytical and numerical capacities .
this problem is especially challenging in the quantum realm , due to the exponential growth of the hilbert space with the number of particles involved .
richard feynman proposed @xcite that the desired calculations may be experimentally realized by codifying the model of interest into the degrees of freedom of another more controllable quantum system .
along these lines , in the last decade , this approach has been employed to simulate the dynamics of many - body quantum systems .
a machine performing this task is called quantum simulator , and it has been studied with increasing interest , theoretically and experimentally , in controlled quantum systems @xcite . it is expected that quantum simulators will solve relevant problems unreachable for classical computers . among them , we could mention complex spin , bosonic , and fermionic many - body systems @xcite , entanglement dynamics @xcite , and fluid dynamics @xcite , among others . in quantum mechanics , realistic situations in which the quantum system is coupled to an environment
are modeled in the framework of open quantum systems . in this description , an effective evolution equation for the system of interest
is obtained by disregarding the environmental degrees of freedom @xcite .
the resulting dynamics can be classified as markovian or non - markovian @xcite . in the former ,
the time evolution depends solely on the current state of the system , and there are several results concerning its quantum simulation @xcite . on the contrary , the non - markovian evolution depends on the history of the system , and it is more challenging to treat both analytically and numerically @xcite . in this sense , despite some recent results @xcite , including a work on the sufficient conditions for a completely positive and trace preserving ( cptp ) non - markovian dynamics @xcite , a general non - markovian quantum simulator has not been fully developed yet .
a paradigmatic feature of non - markovian dynamics is the existence of quantum memory effects as an extension of the classical history - dependent dynamics to the quantum domain .
moreover , a number of key applications in the quantum domain can be envisioned , such as quantum machine learning @xcite , neuromorphic quantum computing @xcite and quantum artificial life @xcite .
these can be implemented by mirroring the already existing results in memcomputing devices @xcite , intelligent materials @xcite and population dynamics @xcite .
therefore , the simulation of quantum memory effects would be a significant step forward to understand open quantum systems and , consequently , to employ them in the development of the aforementioned research fields . in this rapid communication
, we provide an efficient and general framework for an algorithmic quantum simulation @xcite of memory effects modeled by integrodifferential evolution equations .
the protocol algorithmically combines a markovian quantum simulator with perturbation theory techniques in order to retrieve the time evolution of an arbitrary initial state .
our method does not require the engineering of any additional environment , avoiding the challenging task of developing first - principle non - markovian quantum simulators .
moreover , the protocol works even when the evolution does not correspond to a cptp map , which is the case of most of time - delayed lindblad master equations . indeed , although the cptp character is not guaranteed , our approach circumvents this issue by splitting the simulation into two cptp parts . finally , we prove polynomial scaling error bounds for the proposed method .
_ integrodifferential equations with memory .
_ the model describing the memory effects we aim to simulate is based on the integrodifferential equation @xmath0 here , @xmath1 is a memory kernel modeling how the evolution of the state at a certain time is affected by its history , and @xmath2 is a general time - independent lindblad operator .
notice that @xmath3 corresponds to the standard markovian master equation written in the lindblad form . as noticed , for instance , in refs .
@xcite , it is not conceivable to simulate a general non - markovian dynamics efficiently .
the reason is that one could then imagine simulating a highly inefficient calculation in the environment , retrieving this information afterwards into the system due to the non - markovian information backflow , in an efficient manner .
however , eq . includes in the kernel @xmath1 the non - markovian aspects of the evolution , which gives only an effective description of the environment contribution . in order to simulate eq .
, we use as a tool the quantum simulation of the equation @xmath4\rho(s),\ ] ] where @xmath5 is a memory kernel , @xmath6 is a general cptp map and @xmath7 is the identity map .
equation describes the dynamics of a _ semi - markovian _
process @xcite .
it is noteworthy to mention that while eqs . and preserve the trace of the density matrix , they do not generally preserve positivity .
however , sufficient conditions for eq . to determine a cptp map have been studied when @xmath8 .
indeed , if the laplace transform of the memory kernel @xmath9 satisfies the relation @xmath10 for some waiting distribution @xmath11 , then eq
. corresponds to a cptp process @xcite .
moreover , if this condition is fulfilled , then the solution of eq . can be written as @xmath12 , where @xmath13 @xcite . in this case , by truncating the series , we can simulate eq . assuming that an efficient quantum simulator of @xmath6 and its powers is available . in the following
, we will consider processes @xmath6 corresponding to markovian evolutions , whose efficient quantum simulator has been already designed , e.g. , @xmath14-local lindblad equations @xcite .
we will show how to simulate a general kernel @xmath5 , including the case in which eq .
does not correspond to a cptp process .
finally , we illustrate how to employ this result to simulate eq . .
_ algorithmic quantum simulator .
_ let us consider the volterra version of eq .
, @xmath15 \rho ( s),\ ] ] where @xmath16 .
we assume that @xmath17 and @xmath18 , for a given constant @xmath19 , for all @xmath20 .
moreover , we quantify the results in terms of the trace norm for matrices , defined as the sum of their singular values @xmath21 , and the respective induced superoperator norm @xmath22 .
then , eq . can be solved iteratively , via the series @xmath23 , where @xmath24 \rho_{i-1}(s ) .
\label{cpe}\ ] ] this expansion can be truncated at order @xmath25 , @xmath26 , with a small truncation error given by the following estimation .
[ trunc ] @xmath27 provided that @xmath28 , with @xmath29 .
the proof of proposition 1 is provided in appendix a. this truncation allows us to write the approximated solution of eq . by a finite sum , with a number of terms growing linearly with the simulated time
indeed , we have that @xmath30^i\rho(0),\ ] ] with the corresponding parameter values @xmath31 and @xmath32 .
this truncated sum can be rewritten as @xmath33 , with @xmath34 .
proposition [ trunc ] tells us that we can directly simulate a semi - markovian dynamics by just implementing powers of the process @xmath6 , and numerically integrating the memory kernel .
as we need a number of terms which increases linearly with the simulated time , we have that this step is efficient if the implementation of the @xmath6 is efficient .
notice that , by construction , @xmath35 has trace @xmath36 , but it is not necessarily a density matrix , since it can have negative eigenvalues .
however , we can write @xmath37 as a weighted sum of two density matrices and introduce the quantities @xmath38 and @xmath39 . in consequence , we have that @xmath40 where the parameter values @xmath41 and @xmath42 , while @xmath43 holds due to trace preservation .
notice that @xmath44 are two density matrices , as their trace is @xmath36 and they are , by construction , positive .
indeed , we have approximated the dynamics , denoted by @xmath45 [ @xmath46 , corresponding to eq .
, as a weighted sum of two cptp maps : @xmath47 , with @xmath48 .
the form of the resulting cptp maps allows us to simulate eq . by making use of a markovian quantum simulator and numerical techniques .
in fact , all @xmath49 , and thus also @xmath50 , can be classically computed , and the states @xmath44 can be prepared assuming that the markovian operations @xmath51 ( @xmath52 ) are available .
[ semim ] let us consider the simulating dynamics @xmath53 , where @xmath54 , @xmath55 denotes an efficient quantum simulation of @xmath6 , and @xmath56 .
if @xmath57 requires a simulation time @xmath58 , then we can simulate the semi - markovian process in eq . within an error @xmath59 by using a simulation time @xmath60 .
we have that @xmath61 .
the first term is bounded by @xmath62 , as @xmath63 .
the second term can be bounded by @xmath64 .
we have that @xmath65 , assuming @xmath66 .
this requires a simulation time @xmath67 , where we have used that @xmath68 .
proposition [ semim ] allows us to compute approximately the evolution of expectation values of observables under the dynamics of eq . .
it is noteworthy to mention that our method does not require the engineering of any bath corresponding to a semi - markovian dynamics .
instead , we have written the formal solution of eq . , and
exploit the availability of a markovian quantum simulator generating @xmath6 and its powers .
this is possible due to the fast convergence of the exponential series , which limits the number of terms to be classically computed .
moreover , the truncation provided in proposition [ trunc ] implies also that an efficient markovian simulation is sufficient to approximatively generate the solution of eq . .
while in the cptp semi - markovian case we can directly sample from the probability distribution of a given observable , since we are directly implementing the solution , for more general non - markovian equations we only have access to expectation values , as this time the process is split into two parts .
a consequent question is whether we can compute interesting quantities beyond mere observables with our algorithmic quantum simulator . in the following , we study the example of the two - time correlation function of unitary operators , i.e. @xmath69 $ ] . in the last expression , @xmath70 , where @xmath71 is the dual of @xmath45 , defined as @xmath72\equiv { \rm tr}[(\lambda^*(t)\cdot a)\sigma]$ ] for arbitrary @xmath73 and @xmath74 .
first , let us notice that @xmath75 for a sufficiently large @xmath25 .
each resulting term can be computed with an extension to unitary dynamics of the protocol for the two - time correlation function proposed in ref .
@xcite .
indeed , we add a two - dimensional ancilla and initialize the joint system in the state @xmath76 .
first , we implement a controlled operation @xmath77 , then the evolution @xmath78 on the original system , and finally @xmath79 again . in the end , @xmath80 is retrieved by measuring the operator @xmath81 in the ancilla . notice that this protocol shows the same efficiency as the one in proposition [ semim ] .
moreover , the method can be straightforwardly extended to multi - time correlation functions of unitary operators by iterating the aforementioned steps .
lastly , the multi - time correlation functions of observables @xmath82 can be computed by decomposing it into @xmath83 , with @xmath84 and @xmath85 unitary matrices ( see appendix c and ref . )
. now , we are ready to show how to use the quantum simulation of eq . to simulate eq .
let us consider @xmath86 , where @xmath87 is a control parameter and @xmath2 an arbitrary lindblad operator , as in eq . .
in the following , we prove that the solution of eq .
, describing a semi - markovian process , approximates the solution of the memory process in eq . provided that @xmath88 is small .
[ prop3 ] let @xmath89 and @xmath90 be the solutions of eq . and
eq . respectively , with @xmath91 ( @xmath87 ) , @xmath92 with @xmath93 , and @xmath94 .
then , @xmath95 holds if @xmath96 , when @xmath97 , and if @xmath98 , when @xmath99 , provided that @xmath100 .
the bounds of proposition [ prop3 ] are rigorously found in appendix b. the result of proposition [ prop3 ] provides the error bound for a general simulation of a complex environment described by eq . , and it is rather general as it holds for any @xmath2 .
the algorithm consists in implementing the states defining the solution of the approximated semi - markovian process , together with the numerical integration of the memory kernel , as schematically depicted in fig .
[ aqs ] .
the method can be generalized to even more complicated dynamical equations .
for instance , the case of higher - order derivatives , as @xmath101 the solution of eq .
can be approximated analogously to eq . , and
proposition [ prop3 ] extended in order to find similar bounds . a further generalization
consists in introducing additional terms , increasing the versatility of the proposed algorithmic quantum simulator .
for instance , let us consider the equation @xmath102 where @xmath74 can be an arbitrary matrix . then , eq . can be simulated by approximating it with the equation @xmath103/\lambda\ , \rho ( s)$ ] , which can be rewritten and simulated similarly to eq . .
_ conclusions .
_ we have developed a flexible and efficient quantum algorithm for the solution of integrodifferential evolution equations describing quantum memory effects , including the case of non - markovian dynamics .
the proposed algorithmic quantum simulation is useful for mimicking the effective action of complex environments .
alternative situations that our approach may cover include quantum feedback , quantum machine learning , and neuromorphic quantum computation .
lastly , the results in this rapid communication can be exploited for the classical simulation of memory effect equations .
in fact , if the markovian process used as a tool is decomposed efficiently by gates with a positive wigner function , then expected values of observables can be estimated by using monte carlo techniques @xcite .
_ acknowledgments . _
the authors thank ryan sweke for useful discussions .
the authors acknowledge support from spanish mineco / feder fis2015 - 69983-p , upv / ehu ufi 11/55 and a postdoctoral fellowship , basque government grant it986 - 16 and bfi-2012 - 322 , the alexander von humboldt foundation and the aqus eu project . |
as is well known , deep inelastic scattering ( dis ) is one of the basic processes for probing the structure of hadrons . from the theoretical point of view , the inclusive cross section of dis is a convolution of the leptonic and hadronic tensors , with the information about the structure of the hadrons participating into dis coming from the hadronic tensor .
the forward compton amplitude , when a deeply off - shell photon with virtuality @xmath3 scatters off an on - shell hadron with momentum @xmath4 .
the spin - dependent part , @xmath5 , of the hadronic tensor is parametrized in terms of two structure functions , @xmath1 and @xmath6 , as @xmath7 \approx \imath\epsilon_{\mu\nu\lambda\rho } \frac{q_{\lambda } m}{pq } \big [ s_{\rho}^{|| } g_1 + s_{\rho}^{\perp}\big(g_1 + g_2 \big ) \big],\ ] ] so that @xmath1 is related to the longitudinal hadron spin - flip scattering , whereas the sum @xmath8 is relative to the transverse spin - flips . in eq .
( [ w ] ) , @xmath9 stands for the hadron mass , @xmath10 and @xmath11 are the longitudinal and transverse ( with respect to the plane formed by @xmath4 and @xmath12 ) components of the hadron spin @xmath13 .
both @xmath1 and @xmath6 depend on @xmath14 and @xmath15 .
obviously , small @xmath0 corresponds to @xmath16 . in this case ,
@xmath17 and therefore the part of @xmath5 related to @xmath1 does not depend on @xmath9 .
when @xmath18 , one can assume the factorization and regard @xmath5 as a convolution of two objects ( see fig . 1 ) .
the first one is the probability @xmath19 ( @xmath20 in fig .
1a and @xmath21 in fig .
1b ) to find a polarized parton ( a quark or a gluon ) within the hadron .
the second one is the partonic tensor @xmath22 defined and parametrized similarly to @xmath5 . whereas the partonic tensor @xmath22 , i.e. the partonic structure functions @xmath1 and @xmath6 , can be studied within perturbative qcd , @xmath23 are essentially non - perturbative objects , the lack of knowledge of @xmath19 is usually compensated by introducing initial parton distributions which can be found from phenomenological considerations . in doing so , the non - singlet component @xmath24 of @xmath1 is usually expressed as a convolution of a piece which we denote @xmath25 from purely evolution , and the initial polarized quark density @xmath26 : @xmath27 similarly , the singlet structure function @xmath28 is expressed in terms of the evolution pieces @xmath29 and @xmath30 and the densities of the polarized quarks and gluons , @xmath26 and @xmath31 : @xmath32 the subscripts @xmath33 in eqs .
( [ g1ns],[g1s ] ) refer to the kind of the initial partons , as is shown at fig . 1 .
we remind that there is no rigorous procedure for calculating @xmath26 and @xmath31 .
they have to be found from phenomenological considerations . on the contrary , there are regular perturbative methods for calculating the evolution parts @xmath25 and @xmath34 of @xmath1 . when @xmath35 is much greater than the starting point @xmath36 of the @xmath35-evolution and at the same time @xmath37 , it is convenient to rewrite eqs .
( [ g1ns],[g1s ] ) in the form of the mellin integral : @xmath38 e^{\gamma_s(\omega ) \ln(q^2/\mu^2 ) } ~,\end{aligned}\ ] ] with @xmath39 being the coefficient functions and @xmath40 the non - singlet and singlet anomalous dimensions respectively .
the anomalous dimensions control the @xmath35-evolution and the coefficient functions govern the @xmath41-evolution which , at fixed @xmath35 , is equivalent to the @xmath0-evolution . the best known instrument to calculate the dis structure functions is the dglap@xcite approach . in the dglap framework , both the coefficient functions and the anomalous dimensions are perturbatively known and represented by their one - loop , ( or leading order ( lo ) ) @xcite and two - loop ( next - to - leading order ( nlo))contributions .
the relevant ref.s concerning the nlo calculations can be found in the review @xcite .
the remaining ingredients to the rhs of eq .
( [ gmellin ] ) , @xmath26 and @xmath31 can be taken , for example , from ref .
dglap provides a quite good description of the experimental data@xcite .
the extrapolation of dglap into the small-@xmath0 region predicts an asymptotic behavior @xmath42 for all dis structure functions ( with different factors @xmath43 ) .
however , from a theoretical point of view , such an extrapolation at the small-@xmath0 is rather doubtful .
in particular , it neglects in a systematical way contributions ot the type @xmath44 which are small when @xmath45 but become relevant when @xmath37 .
the total resummation of these double - logarithmic ( dl ) contributions was made in refs .
@xcite and ref .
@xcite for the non - singlet @xmath46 and singlet @xmath1 respectively , and it leads to the regge ( power - like ) asymptotics @xmath47 , with @xmath48 being the intercepts calculated ib the double - logarithmic approximation ( dla ) .
the weak point of this resummation in refs .
@xcite,@xcite is the assumption that @xmath49 is kept fixed ( at some unknown scale ) .
it leads therefore to the value of the intercepts @xmath48 explicitly depending on this unknown coupling , while @xmath49 is well - known to be running .
the results of refs .
@xcite,@xcite had led many authors ( see e.g.@xcite ) to suggest that the dglap parametrization @xmath50 has to be used . however , according to results of ref .
@xcite , such a parametrization is indeed correct for @xmath45 only and can not be used for @xmath37 .
the appropriate dependence of @xmath49 suggested in ref .
@xcite , has been successfully used to calculate both @xmath24 and @xmath1 singlet at small @xmath0 in refs .
@xcite . in the present
talk we review these results . instead of a direct study of @xmath1 like it is done in dglap
, it is more convenient to consider the forward compton amplitude @xmath51 for the photon - parton scattering related to @xmath1 , with @xmath52 , as follows : @xmath53 as already stated above , we can not use dglap for studying @xmath1 or @xmath51 at small @xmath0 because it does not not include the total resummation of the double- and single - logarithms of @xmath0 for the anomalous dimensions and coefficient functions , and also the @xmath49-parametrization used is valid for @xmath0 not far from 1 .
then , in order to account for the double - logs of both @xmath0 and @xmath35 , we need to construct a kind of two - dimensional evolution equations that would combine both the @xmath0- and @xmath35- evolutions .
such equations should sum up the contributions of the feynman graphs involved to all orders in @xmath49 . some of those graphs have either ultraviolet or infrared ( ir ) divergences .
the ultraviolet divergences are regulated by the usual renormalization procedure . in order to regulate the ir ones
, we introduce an ir cut - off @xmath54 in the transverse momentum space for the momenta @xmath55 of all virtual quarks and gluons : @xmath56 where @xmath57 stands for the transverse ( with respect to the plane formed by the external momenta @xmath4 and @xmath12 ) component of @xmath55 .
this technique of regulating the ir divergences was suggested by lipatov and used first in ref .
@xcite for quark - quark scattering . using this cut - off @xmath54
, @xmath51 acquires a dependence on @xmath54 .
then , one can evolve @xmath51 with respect to @xmath54 , constructing the appropriate infrared evolution equations ( iree ) . as @xmath58 , @xmath59 where @xmath60 and @xmath61 .
( [ lhs ] ) represents the lhs of the iree for a. in order to write the rhs of the iree , it is convenient to use the sommerfeld - watson transform @xmath62 where @xmath63 is the negative signature factor , @xmath64/ 2 \approx \imath \pi \omega / 2 $ ] . it must be noted that the transform inverse to eq .
( [ mellin ] ) involves the imaginary parts of @xmath51 : @xmath65 notice that , contrary to the amplitude @xmath51 , the structure function @xmath1 does not have any signature and therefore @xmath66 when the transform ( [ mellin ] ) is applied directly to @xmath1 .
when the factorization depicted in fig . 1 is assumed , the calculation of @xmath1 ( for semplicity we will omit the superscript `` s '' for@xmath1 singlet , though we use the notation @xmath24 for the non - singlet @xmath1 ) is reduced to calculating the feynman graphs contributing the partonic tensor @xmath67 depicted as the upper blobs in fig . 1 . both cases , ( a ) , when the virtual photon scatters off the nearly on - shell polarized quark , and ( b ) , when the quark is replaced by the polarized gluon , should be taken into account .
therefore , in contrast to eq .
( [ g1a ] ) , we need to introduce two compton amplitudes : @xmath68 and @xmath69 corresponding to the upper blob in fig . 1a and fig .
1b respectively .
the subscripts `` q '' and `` g '' refer to the initial partons .
therefore , @xmath70 where @xmath71 let us now construct the iree for the amplitudes @xmath72 related to @xmath1 . to this aim ,
let us consider a virtual parton with minimal @xmath73 .
we call such a parton the softest one .
if it is a gluon , its dl contribution can be factorized , i.e. its dl contribution comes from the graphs where its propagator is attached to the external lines .
as the gluon propagator can not be attached to photons , this case is absent in iree for @xmath72 .
the second option is when the softest partons are a @xmath74-channel quark - antiquark or gluon pair .
it leads us to the iree depicted in fig . 2 .
applying the operator @xmath75 to it , combining the result with eq .
( [ lhs ] ) and using ( [ mellin ] ) , we arrive at the following system of equations : @xmath76 ~ , \nonumber \\
\big ( \omega + \frac{\partial}{\partial y}\big ) f_g(\omega , y ) & = & \frac{1}{8 \pi^2 } \big [ f_{gq}(\omega ) f_q(\omega , y ) + f_{gg}(\omega ) f_g(\omega , y ) \big ] ~.\end{aligned}\ ] ] the amplitudes @xmath77 are related to @xmath78 through the transform ( [ mellin ] ) .
the mellin amplitudes @xmath79 , with @xmath80 , describe the parton - parton forward scattering .
they contain dl contributions to all orders in @xmath49 .
we can introduce the new anomalous dimensions @xmath81 .
the subscripts `` q , g '' correspond to the dglap - notation . solving this system of eqs . and using eq .
( [ gqg ] ) leads to @xmath82 ~ , \\
\nonumber g_g(x , q^2 ) & = & \int_{- \imath \infty}^{\imath \infty } \frac{d \omega}{2 \pi \imath } ( 1/ x)^{\omega } \big[c_+(\omega ) \frac{x + \sqrt{r}}{2h_{qg } } e^{\omega_+ y } + c_-(\omega ) \frac{x - \sqrt{r}}{2h_{qg } } e^{\omega_- y } \big]~.\end{aligned}\ ] ] the unknown factors @xmath83 have to be specified and will be discussed later .
all other factors in eq .
( [ gqgg ] ) can be expressed in therms of @xmath84 : @xmath85~.\end{aligned}\ ] ] the anomalous dimension matrix @xmath84 was calculated in ref .
@xcite : @xmath86 where @xmath87}\big)/2}~,\ ] ] @xmath88 @xmath89 and @xmath90 with @xmath91 we have used here the notations @xmath92 and @xmath93 .
the quantities @xmath94 and @xmath95 account for the running of @xmath49 .
they are given by the following expressions : @xmath96 ~,\ ] ] @xmath97\ ] ] with @xmath98 and @xmath99 .
@xmath100 is defined as @xmath51 with the @xmath101 term dropped out .
finally we have to specify the coefficients functions @xmath102 appearing in eq .
( [ gqgg ] ) . when @xmath103 , @xmath104 where @xmath105 and @xmath106 are the input distributions of the polarized partons at @xmath107 .
they do not depend on @xmath35 .
( [ match ] ) allows us to express @xmath83 in terms of @xmath108 and @xmath109 , which are related to @xmath110 and @xmath111 through the ordinary mellin transform . indeed , @xmath112 with both @xmath26 and @xmath31 depending on @xmath113 .
this leads to the following expressions for @xmath29 and @xmath30 : @xmath114~,\ ] ] @xmath115\ ] ] with @xmath116 eqs .
( [ gqsol ] , ( [ ggsol ] ) express @xmath1 in terms of the parton distributions @xmath108 and @xmath109 , which are related to the distributions @xmath110 and @xmath111 at very low @xmath0 : @xmath117 .
therefore , they hardly can be found from experimental data . it is much more useful to express @xmath118 in terms of the initial parton densities @xmath119 and @xmath120 defined at @xmath45 .
we can do it , using the evolution of @xmath110 , @xmath111 with respect to @xmath41 .
indeed , the @xmath41-evolution of @xmath121 from @xmath122 to @xmath123 at fixed @xmath35 @xmath124 is equivalent to their @xmath0-evolution from @xmath45 to @xmath37 . in the @xmath113-space
, the system of iree for the parton distributions looks quite similar to eqs .
( [ system ] ) .
however , the eqs for @xmath125 are now algebraic because they do not depend on @xmath35 : @xmath126~ , \nonumber \\ \delta g(\omega ) & = & ( < e^2_q>/2)\hat{\delta g}(\omega ) + ( 1/\omega ) \left[h_{gq}(\omega ) \delta q(\omega ) + h_{gg}(\omega ) \delta g(\omega ) \right ] ~.\end{aligned}\ ] ] where @xmath127 is the sum of the quark electric charges ( @xmath128 for @xmath93 ) and @xmath129 is the starting point of the evolution of the gluon density @xmath130 .
it corresponds to fig .
1b where the upper blob is substituted by the quark box . solving eqs .
( [ systeminput ] ) , we obtain : @xmath131 } { \big[\omega^2 - \omega(h_{qq } + h_{gg } ) + ( h_{qq}h_{gg } - h_{qg}h_{gq})\big]}~,\ ] ] @xmath132 } { \big[\omega^2 - \omega(h_{qq } + h_{gg } ) + ( h_{qq}h_{gg } - h_{qg}h_{gq})\big ] } ~.\ ] ] then eqs . ( [ gqsol],[ggsol],[inputq],[inputg ] ) express @xmath1 in terms of the initial parton densities @xmath133 .
when we put @xmath134 and do not sum over @xmath135 , we arrive at the expression for the non - singlet structure function @xmath24 : obviously , in this case @xmath136
. however , the nonsinglet anomalous dimension @xmath137 should be calculated in the limit @xmath138 .
we denote such @xmath139 .
the explicit expression for it is : @xmath140~.\ ] ] therefore , we arrive at @xmath141
before making use of eqs .
( [ gqsol]),([ggsol ] and ( [ gns ] ) for calculating @xmath1 at small but finite values of @xmath0 , let us discuss its asymptotics .
when @xmath2 and @xmath142 , one can neglect contributions with @xmath143 in eqs .
( [ gqgg ] ) .
as is well known , @xmath144 at @xmath2 , with @xmath145 being the position of the leading singularity of the integrand of @xmath1 .
according to eqs .
( [ solh ] ) , the leading singularity , @xmath146 for @xmath24 is the rightmost root of the equation @xmath147 while the leading singularity , @xmath145 for @xmath1 is the rightmost root of @xmath148 in our approach , all factors @xmath149 depend on @xmath150 , so the roots of eqs .
( [ singns],[sings ] ) also depend on @xmath151 .
this dependence is plotted in fig . 3 for @xmath146 and in fig . 4 for @xmath145 . both the curve in fig . 3 and the curve 1 in fig .
4 have a maximum .
we denote this maximum as the intercept .
therefore , @xmath152 ( 1/x)^{\delta_s}(q^2/\mu^2)^{\delta_s/2},\end{aligned}\ ] ] and we find for the intercepts @xmath153 and @xmath154 .
this implies that @xmath24 is positive when @xmath2 whereas @xmath28 can be either positive or negative , depending on the relation between @xmath155 and @xmath130 . in particular , @xmath1 is positive when @xmath156 otherwise it is negative .
in other words , the sign of @xmath1 at small @xmath0 can be positive if the initial gluon density is negative and large .
let us estimate the impact of the total resummation of dl and sl contributions on @xmath24 . according to eq .
( [ gns ] ) , the value of @xmath157 depends both on the perturbative terms and on the inputs @xmath26 .
the latter can be obtained by fitting the experimental data and it is known ( see e.g. @xcite ) that widely different formulae for @xmath26 can be used . in order to avoid discussing the fitting procedure and as in this paper we present our results only the perturbative part of @xmath24 , we can assume @xmath158 then let us calculate the ratio @xmath159 where @xmath160 is the lo dglap non - singlet @xmath1 .
the results of a numerical calculations for @xmath161 at @xmath162gev@xmath163 , @xmath164gev are shown in table 1 .
._ratio @xmath161 of structure functions @xmath24 calculated in dla and dglap lo .
_ [ cols="^,^",options="header " , ] [ r ] it shows that the impact of the total resummation of dl contributions is negligible for @xmath165 but it grows fast with decreasing of @xmath0 , achieving the values @xmath166 at @xmath167 and @xmath168 at @xmath169 .
the total resummation of the most singular @xmath170 terms in the expressions for the anomalous dimensions and the coefficient functions leads to the expressions of eqs .
( [ g1qg],[gqsol],[ggsol],[gns ] ) for the singlet and the non - singlet structure functions @xmath1 .
it guarantees the regge ( power - like ) behavior ( [ g1asympt ] ) of @xmath171 when @xmath2 , with the intercepts given by eq .
( [ intercepts ] ) .
the intercepts @xmath172 are obtained with the running qcd coupling effects taken into account .
the value of the non - singlet intercept @xmath173 is now confirmed by several independent analysis @xcite of experimental data and our result @xmath174 is in a good agreement with the estimate of ref .
@xcite : @xmath175 obtained from analysis of the hermes data .
( [ gns ] ) states that @xmath24 is positive both at @xmath45 and at @xmath176 .
the situation concerning the singlet @xmath1 is more involved : being positive at @xmath45 , the singlet @xmath1 can remain positive at @xmath37 only if the initial parton densities obey eq .
( [ posit ] ) , otherwise it becomes negative .
the ratio of our results versus the dglap ones for non - singlet @xmath1 is given in table 1 .
it shows explicitly that the impact of high - order dl contributions is small at @xmath165 but it grows fast when @xmath0 is approaching @xmath177 .
the work is supported by grant rsgss-1124.2003.2 99 g. altarelli and g. parisi .
nucl.phys.b*126*(1977 ) 298 ; v.n . gribov and l.n .
sov.j.nucl.phys.*15*(1978 ) 438 and 675 ; yu.l .
sov.phys.jetp*46*(1977)641 . w.l .
van neerven .
hep - ph/9609243 .
g. altarelli , r. ball , s. forte and g. ridolfi .
acta phys.polon.b*29*(1998)1201 ; nucl.phys.b*496*(1999)337 .
ermolaev , s.i .
manaenkov and m.g .
z.phys.c*69*(1996)259 ; j. bartels , b.i .
ermolaev and m.g .
ryskin . z.phys.c*70*(1996)273 .
j. bartels , b.i .
ermolaev and m.g .
ryskin . z.phys.c*72*(1996)627 .
j. kwiecinski .
acta.phys.polon.b*29*(2001)1201 ; d. kotlorz and a. kotlorz .
acta.phys.polon.b*32*(2001 ) 2883 ; b. badalek , j. kiryluk and j. kwiecinski . phys.rev.d*61*(2000)014009 ; j. kwiecinski and b. ziaja .
phys.rev.d*60*(1999)9802386 ; b. ziaja .
phys.rev.d*66*(2002)114017 ; hep - ph/0304268 , d. kotlorz and a. kotlorz .
acta.phys.polon.b*35*(2004 ) 705 .
ermolaev , m. greco and s.i .
ermolaev , m. greco and s.i .
nucl.phys.b*594*(2001)71 ; ibid*571*(2000)137 ; phys.lett.b*579*(2004)321 .
r. kirschner and l.n .
sov.phys.jetp 56(1982)266 .
gribov , v.g .
gorshkov , g.v .
frolov , l.n .
sov.j.nucl.phys.*6*(1968)95 ; ibid * 6*(1968)262 .
j. soffer and o.v .
phys.rev.d*56*(1997)1549 ; a.l .
kataev , g. parente , a.v.sidorov .
cern - th-2001 - 058 ; phys.part.nucl.*34*(2003)20 ; fiz.elem.chast.atom.yadra*34*(2003)43 ; nucl.phys.a*666*(2000)184 .
kotikov , a.v .
lipatov , g. parente , n.p .
zotov , eur.phys.j.c*26*(2002)51 ; v.g .
krivokhijine , a.v .
kotikov , hep - ph/0108224 ; a.v .
kotikov , d.v .
peshekhonov , hep - ph/0110229 . n.i .
kochelev , k. lipka , w.d .
nowak , v. vento , a.v .
phys . rev.d*67*(2003)074014 . | accounting for double - logarithms of @xmath0 and running qcd coupling leads to expressions for both the non - singlet and singlet components of @xmath1 . these expressions manifest the regge asymptotics when @xmath2 and differ considerably from the dglap expressions at small values of @xmath0 .
= 11.6pt |
random walks model a host of phenomena and find application in virtually all areas of physics @xcite . in this paper
we consider a one - dimensional ( 1d ) system of non - interacting , unbiased diffusing walkers @xmath0 , with random initial conditions .
recently , we have studied the _ persistence _ properties of this model @xcite .
the persistence probability @xmath3 has been widely studied in recent years @xcite . in the present context @xmath3
is defined as the fraction of sites which have never been visited by a diffusing particle up to time @xmath2 .
such sites are termed `` persistent sites '' .
we showed @xcite that , in the case of the @xmath0 processes , @xmath12\ , \label{eq1}\ ] ] where @xmath13 is the particle density and @xmath14 is the diffusion constant .
the fraction of unvisited sites in this model can equally be thought of as the surviving fraction of immobile particles in the presence of randomly diffusing traps . at @xmath7 , consider each site to contain a static particle , @xmath5 , so that @xmath15 . then for @xmath16 , let particles of another type , @xmath6 , randomly distributed on the lattice with density @xmath17 , diffuse though the system with the @xmath6 particles acting as perfect traps for the @xmath5 particles so that @xmath18 .
this model is just the well - studied `` scavenger '' model of redner and kang @xcite .
the fraction of unvisited sites is then just @xmath19 .
the focus of this paper is to explore the relation between @xmath3 and the fraction of walkers @xmath1 that have never encountered another walker up to time @xmath2 . to facilitate our discussion , we introduce the following definition .
we consider all walkers to be _ uninfected _ at time @xmath7 .
the walkers are , however , considered to be mutually infectious , so that for @xmath20 any contact between walkers leads to mutual contamination .
we therefore define an uninfected walker to be one which has never encountered ( occupied the same site as ) another walker .
we can then refer to @xmath1 as the fraction of uninfected walkers .
the fraction of uninfected walkers is , in fact , isomorphic to the survival probability of @xmath5 particles in the reaction @xmath21 , with equal mobility for both species @xcite . at late times
, one expects the asymptotics of this model to be identical to those of the @xmath4 diffusion - limited process with @xmath22 . for this process
it was proved rigorously by bramson and lebowitz @xcite that @xmath23 in one - dimension , with unknown constant @xmath24 .
therefore we expect identical asymptotic behavior for @xmath1 .
the kinetics of the two species annihilation model has received considerable attention over the past two decades and is largely understood @xcite .
it is well appreciated that in this model , there exists an upper critical dimension @xmath25 , below which spatial fluctuations in the initial distribution of the reactants play a significant role in the evolution of the density of the particles .
this dependence on the microscopic fluctuations invalidates traditional approaches such as the mean - field approximation which yields @xmath26 .
it was first shown by toussaint and wilczek @xcite that , for equal initial densities ( @xmath27 ) of randomly diffusing @xmath5 and @xmath6 particles , the walker density decays anomalously , according to @xmath28 for @xmath29 , assuming that the particles are distributed homogeneously at @xmath7 . in the case of this model
we argued heuristically , on the basis of a toy model @xcite that , @xmath30 \label{eq3}\ ] ] where @xmath31 is a constant and @xmath32 .
we extend our study to include an investigation of the fraction of uninfected walkers in this model in one - dimension .
the fraction of uninfected walkers , @xmath1 , is related to the fraction of persistent sites , @xmath3 , in the following manner . in evaluating @xmath3 ,
one considers particles with diffusion constant @xmath14 and addresses the probability that a given site @xmath33 has never been visited by any walker up to time @xmath2 .
equivalently , one could consider the site @xmath33 to contain a particle with diffusion constant @xmath34 and address the probability that the particle remains uninfected ( i.e.unvisited ) . however , if the particle initially at position @xmath33 is given a diffusion constant equal to that of the other particles in the system i.e. @xmath35 , then the probability that it `` persists '' up to time @xmath2 is the probability that it remains uninfected .
this suggests that the number of uninfected particles is a function of the diffusion constants in the problem , with the limiting case @xmath34 corresponding to the standard persistence problem .
clearly , in order to avoid contamination , remaining stationary is a more effective strategy than diffusing .
therefore , we anticipate that the fraction of uninfected walkers decays more rapidly than the fraction of uninfected ( unvisited , i.e. persistent ) sites
. this faster decay of @xmath1 is clearly observed in the simulations .
given the similar decay of @xmath3 in the @xmath0 and @xmath4 processes , these models make a sensible combination to study together . we discover that in both cases , @xmath36^\gamma$ ] where @xmath37 .
in fact the relation @xmath38^\gamma$ ] , with @xmath39 close to unity and @xmath40 , describes the data rather accurately in both models over the whole range of times studied , although there are indications in both models that the asymptotic value of @xmath10 is smaller than @xmath41 .
we discuss the @xmath0 model in sec .
ii and follow with the @xmath4 reaction in sec .
we conclude in sec .
iv with a discussion and summary of the results .
we consider a system of non - interacting , unbiased , diffusing walkers , with random initial conditions .
our numerics are performed on a 1d lattice of size @xmath42 with lattice constant @xmath43 and periodic boundaries . at @xmath7 , @xmath44
random walkers are randomly distributed on the lattice with a maximum of one particle per site so that @xmath45 and @xmath46 , where @xmath47 .
clearly , @xmath48 is a consequence of our lattice description and in the continuum limit @xmath49 , where the mean distance between walkers is much larger than the lattice spacing , @xmath50 .
therefore , in order to place the fraction of persistent sites and the fraction of uninfected walkers on an equal footing in our numerical study , we define , @xmath51 so that both @xmath52 and @xmath53 are normalized to 1 .
we now use @xmath54 in the description of our numerical results .
our model is updated using the direct method @xcite i.e. at each computational step , a particle is picked at random and shifted with probability @xmath55 to a neighboring site , where @xmath14 is the diffusion constant . for @xmath56 ,
any site which is visited by a walker becomes nonpersistent for all remaining time @xmath2 .
mutual infection occurs when two or more particles simultaneously occupy a single lattice site . for each jump made by a particle , time @xmath2
is increased by @xmath57 , where @xmath44 is the current number of particles in the system ( which is constant in the present model ) .
our results are averaged over 100 independent runs .
dimensional analysis demands that , on a _ continuum _ , the fraction of unvisited sites is a function of the dimensionless combination of parameters in the problem , i.e.@xmath58 .
similarly , for the fraction of uninfected walkers , @xmath59 , where @xmath60 is some other function .
however , given that our numerical simulations take place on a discrete lattice , we expect the scaling of @xmath3 and @xmath1 as a function of @xmath61 to become strictly true only in the continuum limit , @xmath62 . in order to approximate this limit as closely as possible ,
we choose as our densities @xmath63 , 0.005 and 0.003 , so that @xmath64 .
the dimensionless combination of parameters , @xmath61 , also defines a natural characteristic time scale in the problem , namely @xmath65 .
therefore one expects to observe the asymptotic behavior of the system only after a time @xmath66 has been reached .
however , as far as the persistent properties are concerned , eq . ( [ eq1 ] ) is an exact relation in the continuum limit for all @xmath2 .
therefore , we expect to observe @xmath3 approaching asymptopia very quickly in our numerical data . clearly , in the discrete description of the model used in our simulations , eq .
( [ eq1 ] ) breaks down at the earliest times since , for very small @xmath2 , lattice effects play a dominant role .
( [ eq2 ] ) , however , which holds for @xmath1 is a truly asymptotic relation .
unfortunately , the extremely fast decay of @xmath1 leads to small - number effects at very early times in our simulations and we find the limit @xmath67 unattainable with good statistics .
therefore , we limit ourselves to studying the regime up to @xmath68 .
the fraction of unvisited sites decays , in the continuum limit , exactly as the stretched exponential form , eq .
( 1 ) @xcite . in order to make a direct comparison between the fraction of persistent sites , @xmath54 , and the fraction of uninfected walkers , @xmath1 , we present these two quantities on a log - linear plot in fig .
[ fig1 ] , using @xmath69 as abscissa .
[ fig1 ] shows clearly that the data scale as a function of the dimensionless quantity @xmath61 .
@xmath1 exhibits behavior characteristic of eq .
( [ eq2 ] ) , with a faster decay rate than for the fraction of persistent sites i.e. @xmath70 .
the data seem consistent with @xmath71^\gamma\ , \label{eq5}\ ] ] with @xmath72 and @xmath73 . the log - log plot of @xmath74 in fig .
[ fig2 ] indeed shows an almost linear relationship between @xmath75 and @xmath76 . in order to evaluate @xmath10 and @xmath77
we performed a linear regression on the data in fig.[fig2 ] in the region indicated by the arrows , thereby avoiding initial transients associated with our lattice description , and the onset of statistical fluctuations between different runs of the data .
we summarize our results for @xmath10 and @xmath77 in table 1 .
[ cols="^,^,^",options="header " , ] table 2 .
numerical values of @xmath78 and @xmath39 , where @xmath79^\delta$ ] , for various @xmath13 , in the @xmath4 model . in analogy to fig .
[ fig3 ] , we plot @xmath80 and @xmath81 against @xmath82 in fig . [ fig7 ] .
early - time lattice effects , where the different densities do not overlap , are evident in fig .
it is possible that all of our data for @xmath54 and @xmath1 are within the preasymptotic regime , since the @xmath54 curve has not yet saturated to the late - time asymptotic values impied by eq .
( [ eq3 ] ) .
note , however , that if the curves in fig .
[ fig7 ] were asympotically linearly increasing , rather than eventually saturating to a constant , the corresponding forms for @xmath54 and @xmath1 would be @xmath83 $ ] , with @xmath31 a constant ( different for @xmath84 and @xmath85 ) .
this form differs only by the logarithmic term from that predicted in @xcite on the basis of a toy model .
it should be noted that neither our present data nor that presented in @xcite can definitively rule out such logarithmic corrections , and the asympotic form of @xmath54 , as well as @xmath1 , can not be regarded as being definitively established . throughout the preasymptotic regime , however , the relation @xmath86^\delta$ ] , with @xmath87 , holds rather well . in fig .
[ fig8 ] we plot the effective exponent @xmath88 directly as a function of @xmath89 .
[ fig8 ] is remarkably similar to fig .
[ fig4 ] for the @xmath0 process .
indeed , except for the earliest times , the corresponding curves differ by less by @xmath90 everywhere . in a similar way we can bound the asymptotic @xmath78 ( assuming that the late - time trend evident in fig . [ fig8 ]
continues ) by @xmath91 . since we have argued elsewhere @xcite , using a toy model , that the asymptotic behavior of @xmath3 is described by eq .
( [ eq3 ] ) , we conjecture that the asymptotics of @xmath1 are given by @xmath92\ , \label{eq7}\ ] ] where @xmath31 is a constant and @xmath93 . for the @xmath0 model
, we expressed [ eq .
( [ eq6 ] ) ] @xmath1 in terms of the ratio @xmath94 , where @xmath95 is the diffusion constant of a tagged particle , @xmath14 is the diffusion constant of the other particles , and @xmath1 is the probability that the tagged particle remains uninfected at time @xmath2 . the analogous form for the @xmath4 process is naturally , @xmath96 \label{eq8}\ ] ] where @xmath97 is some unknown function .
as before , the case @xmath34 corresponds to the problem of unvisited sites , so that @xmath98 and @xmath99 .
for @xmath35 , the case studied above , our conjecture implies @xmath100 .
however , we stress that a logarithmic correction of the form discussed above can not be ruled out by the data . the main motivation for eq . ( [ eq8 ] )
is the toy model presented in @xcite , applied to the case @xmath34 .
in this paper we have studied two distinct models : ( i ) a one - dimensional system of non - interacting , unbiased diffusing walkers @xmath0 , with random initial conditions and ( ii ) the @xmath4 diffusion limited process in one dimension with equal initial densities of @xmath5 and @xmath6 particles distributed homogeneously at @xmath7 .
we have reduced the familiar study of persistent sites @xmath3 , to a limiting case in the study of uninfected walkers and shown numerically that in both models , @xmath101^\gamma\ , \label{eq9}\ ] ] where @xmath102 for the time regimes covered by our data and @xmath103 for asymptotically large time .
the common feature of the @xmath0 and @xmath4 processes is that the fraction of unvisited sites decays according to a stretched exponential ( possibly with logarithmic corrections for @xmath4 ) . in our earlier paper @xcite , we argued heuristically , on the basis of a toy model , that if the density of walkers decays asymptotically as @xmath104 , then the fraction of unvisited sites @xmath3 decays with a stretched exponential form if @xmath105 .
we showed that @xmath106^{(1/2)-\alpha } , \qquad \alpha < 1/2 \label{eq10}\ ] ] where @xmath5 is some function of @xmath107 .
this result is consistent with the results for models studied here .
remarkably , eq.([eq9 ] ) seems to hold for both the @xmath0 and the @xmath4 processes , with approximately the same bounds on @xmath10 .
furthermore , the effective time - dependent value of @xmath10 is very similar for both models ( see figs .
[ fig4 ] and [ fig8 ] ) .
an obvious goal for the future is to try to obtain a theoretical understanding of this simple relationship , and test it against other models with @xmath108 . for both models studied here we expressed the probability that a particle at site @xmath33 is uninfected in terms of its diffusion constant @xmath95 and the diffusion constant @xmath14 of the other particles [ eqs.([eq6 ] ) and ( [ eq8 ] ) ] , noting that @xmath3 corresponds to the limit @xmath34 . using eq .
( [ eq10 ] ) , we can generalize this description to any system of randomly diffusing particles for which the density decays as @xmath109 with @xmath110 .
we conjecture that , in this case , the fraction of uninfected walkers can be expressed , asymptotically , as @xmath111\ , \label{eq12}\ ] ] with @xmath112 and @xmath113 some unknown function .
clearly , there is much scope to test the validity of this form since we have only considered @xmath114 and @xmath115 .
note that for systems with @xmath116 , our toy model predicts @xcite that @xmath3 approaches a non - zero constant , while @xmath1 certainly behaves differently since it is bounded above by @xmath117 and must therefore vanish for large @xmath2 .
more interesting , however , is the borderline case where the particle density falls off like @xmath118 . in this case , our toy model predicts that @xmath3 @xcite decays as a power - law , @xmath119 .
a system where such behavior is observed is the well studied @xmath120-state potts model @xcite .
the results of our study of uninfected walkers in this model follows in a separate paper @xcite .
we thank peter grassberger for useful comments and pointing out references @xcite , and richard blythe for useful discussions .
this work was supported by epsrc ( uk ) . | we consider a system of unbiased diffusing walkers ( @xmath0 ) in one dimension with random initial conditions .
we investigate numerically the relation between the fraction of walkers @xmath1 which have never encountered another walker up to time @xmath2 , calling such walkers `` uninfected '' and the fraction of sites @xmath3 which have never been visited by a diffusing particle .
we extend our study to include the @xmath4 diffusion - limited reaction in one - dimension , with equal initial densities of @xmath5 and @xmath6 particles distributed homogeneously at @xmath7 .
we find @xmath8^\gamma$ ] , with @xmath9 , in both models , though there is evidence that a smaller value of @xmath10 is required for @xmath11 . 2 |
a large number of extensions of the sm predict the existence of color triplet particles carrying simultaneously leptonic and baryonic number , the so - called leptoquarks .
leptoquarks are present in models that treat quarks and leptons on the same footing , such as composite models @xcite , grand unified theories @xcite , technicolor models @xcite , and superstring - inspired models @xcite .
since leptoquarks are an undeniable signal for physics beyond the sm , there have been several direct searches for them in accelerators . at the cern large electron - positron collider ( lep ) , the experiments established a lower bound @xmath2@xmath3 gev for scalar leptoquarks @xcite . on the other hand , the search for scalar leptoquarks decaying into an electron - jet pair in @xmath4 colliders constrained their masses to be @xmath5 gev @xcite . furthermore , the experiments at the desy @xmath6 collider hera @xcite place limits on their masses and couplings , leading to @xmath7 gev depending on the leptoquark type and couplings .
there have also been many studies of the possibility of observing leptoquarks in the future @xmath8 @xcite , @xmath6 @xcite , @xmath9 @xcite , @xmath10 @xcite , and @xmath11 @xcite colliders . in this work
we study the constraints on scalar leptoquarks that can be obtained from their contributions to the radiative corrections to the @xmath0 physics .
we evaluated the one - loop contribution due to leptoquarks to all lep observables and made a global fit in order to extract the 95% confidence level limits on the leptoquarks masses and couplings @xcite .
the most stringent limits are for leptoquarks that couple to the top quark .
therefore , our results turn out to be complementary to the low energy bounds @xcite since these constrain more strongly first and second generation leptoquarks .
the masses and couplings of leptoquarks are constrained by low - energy experiments , since the leptoquarks induce two - lepton two - quark effective interactions , for energies much smaller than their masses @xcite .
the processes that lead to strong limits are : @xmath12 leptoquarks can give rise to flavor changing neutral current ( fcnc ) processes if they couple to more than one family of quarks or leptons @xcite . in order to avoid strong bounds from fcnc
, we assumed that the leptoquarks couple to a single generation of quarks and a single one of leptons .
however , due to mixing effects on the quark sector , there is still some amount of fcnc @xcite and , therefore , leptoquarks that couple to the first two generations of quarks must comply with some low - energy bounds @xcite . @xmath12 the analyses of the decays of pseudoscalar mesons , like the pions , put stringent bounds on leptoquarks unless their coupling is chiral that is , it is either left - handed or right - handed @xcite .
@xmath12 leptoquarks that couple to the first family of quarks and leptons are strongly constrained by atomic parity violation @xcite . in this case , there is no choice of couplings that avoids the strong limits .
it is interesting to keep in mind that the low - energy data constrain the masses of the first generation leptoquarks to be bigger than @xmath13@xmath14 tev when the coupling constants are equal to the electromagnetic coupling @xmath15 @xcite .
the bounds on scalars leptoquarks coming from low - energy and @xmath0 physics exclude large regions of the parameter space where the new collider experiments could search for these particles , however , not all of it @xcite .
notwithstanding , we should keep in mind that nothing substitutes the direct observation .
a natural hypothesis for theories beyond the sm is that they exhibit the gauge symmetry @xmath1 above the symmetry breaking scale @xmath16 .
therefore , we imposed this symmetry on the leptoquark interactions . in order to avoid strong bounds
coming from the proton lifetime experiments , we required baryon ( @xmath17 ) and lepton ( @xmath18 ) number conservation .
the most general effective lagrangian for leptoquarks satisfying the above requirements and electric charge and color conservation is @xcite @xmath19 where @xmath20 , @xmath21 ( @xmath22 ) stands for the left - handed quark ( lepton ) doublet , and @xmath23 , @xmath24 , and @xmath25 are the singlet components of the fermions .
we denote the charge conjugated fermion fields by @xmath26 and we omitted in ( [ lag : fer ] ) the flavor indices of the couplings to fermions and leptoquarks .
the leptoquarks @xmath27 and @xmath28 are singlets under @xmath29 while @xmath30 and @xmath31 are doublets , and @xmath32 is a triplet .
furthermore , we assumed in this work that the leptoquarks belonging to a given @xmath29 multiplet are degenerate in mass , with their mass denoted by @xmath33 .
local invariance under @xmath1 implies that leptoquarks also couple to the electroweak gauge bosons . to obtain the couplings to @xmath34 , @xmath0 , and @xmath35
, we substituted @xmath36 by the electroweak covariant derivative ( @xmath37 ) in the leptoquark kinetic lagrangian : @xmath38 \phi \ ; , \ ] ] where @xmath39 stands for the leptoquarks fields , @xmath40 is the electric charge matrix of the leptoquarks , @xmath41 is the sine of the weak mixing angle , and the @xmath42 s are the generators of @xmath29 for the representation of the leptoquarks .
the weak neutral charge is @xmath43 .
we employed the on - shell - renormalization scheme , adopting the conventions of ref .
we used as inputs the fermion masses , @xmath44 , @xmath45 , and the @xmath0 mass , and the electroweak mixing angle being a derived quantity that is defined through @xmath46 .
we evaluated the loops integrals using dimensional regularization and we adopted the feynman gauge to perform the calculations .
close to the @xmath0 resonance , the physics can be summarized by the effective neutral current @xmath47 \ ; , \label{form : nc}\ ] ] where @xmath48 ( @xmath49 ) is the fermion electric charge ( third component of weak isospin ) .
the form factors @xmath50 and @xmath51 have universal contributions , _
i.e. _ independent of the fermion species , as well as non - universal parts : @xmath52 leptoquarks can affect the physics at the @xmath0 pole through their contributions to both universal and non - universal corrections .
the universal contributions can be expressed in terms of the unrenormalized vector boson self - energy ( @xmath53 ) as @xmath54 \ ; , \end{aligned}\ ] ] where the factors @xmath55 are defined below .
the leptoquark contributions to the self - energies can be easily evaluated , yielding @xmath56 where @xmath57 is the number of colors and the sum is over all members of the leptoquark multiplet .
the coefficient @xmath58 is given by @xmath59 , @xmath60 , @xmath61 , and @xmath62 for @xmath63 , @xmath0 , @xmath64 , and @xmath65 respectively . the function @xmath66 is defined according to : @xmath67 \ ; , \ ] ] with @xmath68 and @xmath69 being the number of dimensions . the factors @xmath55 ( @xmath70 , @xmath71 ) stem from corrections to the effective coupling between the @xmath72 and fermions at low energy .
leptoquarks modify this coupling , inducing a contribution that we parametrize as @xmath73 where @xmath74 ( @xmath75 ) is the left - handed ( right - handed ) projector and @xmath22 stands for the lepton flavor .
since this correction modifies the muon decay , it contributes to @xmath76 , and consequently , to @xmath77 .
leptoquarks with right - handed couplings , as well as the @xmath78 ones , do not contribute to @xmath55 .
the analytical for @xmath55 due to left - handed leptoquarks in the @xmath79 sector can be found in ref . [ 14 ] .
corrections to the vertex @xmath80 give rise to non - universal contributions to @xmath50 and @xmath51 .
we parametrize the effect of leptoquarks to these couplings by @xmath81 \ ; , \ ] ] where for leptons ( @xmath22 ) and leptoquarks with @xmath79 @xmath82 \right . \\ & + 2 s_w c_w q_z^{j } \frac{m^2 - m_q^2 - \frac{1}{2 } m_z^2}{m_z^2 } \left [ - \ln \left ( \frac{m^2}{m_q^2 } \right ) + \bar{b_0 } ( m_z^2 , m^2 , m^2 ) \right ] \\ & + g_x^q \frac{m^2-m_q^2 - \frac{1}{2 } m_z^2}{m_z^2 } \bar{b_0 } ( m_z^2 , m_q^2 , m_q^2 ) + g^{\ell}_x \bar{b_1 } ( 0 , m_q^2 , m^2 ) \\ & + \left [ g_{-x}^q m_q^2 + g_x^q \frac{(m^2-m_q^2)^2}{m_z^2 } \right ] c_0 ( 0 , m_z^2 , 0 , m^2 , m_q^2 , m_q^2 ) \\ & \left .
- 2 s_w c_w q_z^{j } \frac{(m^2-m_q^2)^2 + m_q^2 m_z^2}{m_z^2 } c_0 ( 0 , m_z^2 , 0 , m_q^2 , m^2 , m^2 ) \right\ } \ ; , \end{array } \label{z : ll}\ ] ] where the @xmath83 @xmath84 corresponds to left- ( right- ) handed leptoquarks and @xmath85 with the neutral current couplings being @xmath86 and @xmath87 .
@xmath88 summarizes the couplings between leptoquarks and fermions .
the functions @xmath89 , @xmath90 , @xmath91 , and @xmath92 are the passarino - veltman functions @xcite .
we used the convention @xmath93 and @xmath94 ( @xmath95 ) .
we also defined @xmath96 with @xmath97 given by eq .
( [ delta ] ) . from this last expression
we can obtain the effect of @xmath79 leptoquarks on the vertex @xmath98 simply by the change @xmath99 .
moreover , we can also employ the expression ( [ z : ll ] ) to @xmath78 leptoquarks provided we substitute @xmath100 and @xmath101 .
with all this we have @xmath102 \ ; .\end{aligned}\ ] ] one very interesting property of the general leptoquark interactions that we are analyzing is that all the physical observables are rendered finite by using the same counter - terms as appear in the sm calculations @xcite .
for instance , starting from the unrenormalized self - energies ( [ sig : g ] ) and the mass and wave - function counter - terms we obtain finite expression for the two - point functions of vector bosons .
moreover , the contributions to the vertex functions @xmath80 and @xmath103 are finite . in order to check the consistency of our calculations , we analyzed the effect of leptoquarks to the @xmath104 vertex at zero momentum .
it turns out that the leptoquark contribution to this vertex function not only is finite but also vanishes at @xmath105 for all fermion species .
therefore , our expressions for the different leptoquark contributions satisfy the appropriate qed ward identities , and leave the fermion electric charges unchanged .
moreover , we also verified explicitly that the leptoquarks decouple in the limit of large @xmath33 .
in our analyses , we assumed that the leptoquarks couple to leptons and quarks of the same family . in order to gain some insight
on which corrections are the most relevant , let us begin our analyses by studying just the oblique corrections @xcite , which we parametrized in terms of the variables @xmath106 , @xmath107 , and @xmath108 .
these variables depend only upon the interaction of leptoquarks with the gauge bosons and it is easy to see that leptoquarks contribute only to @xmath107 . imposing that this contribution must be within the limits allowed by the lep data , we find out that the constraints coming from oblique corrections are less restrictive than the available experimental limits @xcite .
we then performed a global fit to all lep data including both universal and non - universal contributions . in table
[ lepdata ] we show the the combined results of the four lep experiments @xcite that were used in our analysis . in order to perform the global fit we constructed the @xmath109 function associated to these data
and we minimized it using the package minuit .
we expressed the theoretical predictions to these observables in terms of @xmath110 , @xmath111 , and @xmath76 , with the sm contributions being obtained from the program zfitter @xcite . in our fit we used five parameters , three from the sm : @xmath112 , @xmath113 , and @xmath114 , and two new ones : @xmath33 , and the leptoquark coupling denoted by @xmath115 .
furthermore , we have also studied the dependence upon the sm inputs @xmath116 , @xmath45 , and @xmath44 . @xmath117 } & 91.1888 \pm 0.0044 \\
\gamma_z \mbox{[gev ] } & 2.4974 \pm 0.0038 \\
\sigma_{\rm had}^0 \mbox{[nb ] } & 41.49 \pm 0.12\\ r_e = \frac{\gamma({\rm had})}{\gamma(e^+ e^- ) } & 20.850 \pm 0.067 \\
r_\mu = \frac{\gamma({\rm had})}{\gamma(\mu^+ \mu^- ) } & 20.824 \pm 0.059 \\ r_\tau = \frac{\gamma({\rm had } ) } { \gamma(\mu^+ \mu^- ) } & 20.749 \pm 0.070 \\
a_{fb}^{0e } & 0.0156 \pm 0.0034 \\ a_{fb}^{0\mu } & 0.041 \pm 0.0021 \\ a_{fb}^{0\tau } & 0.0228 \pm 0.0026 \\
a_{\tau}^0 & 0.143 \pm 0.010 \\
a_e^0 & 0.135 \pm 0.011 \\
r_b = \frac{\gamma(b \bar{b } ) } { \gamma({\rm had } ) } & 0.2202 \pm 0.0020\\ r_c = \frac{\gamma(c\bar{c } ) } { \gamma({\rm had } ) } & 0.1583 \pm 0.0098\\ a_{fb}^{0b } & 0.0967 \pm 0.0038 \\
a_{fb}^{0c } & 0.0760 \pm 0.0091 \\
\hline \hline \end{array}\ ] ] the first part of our analysis consisted of the study of the constraints on the leptoquark masses and couplings .
in order to determine the allowed region in the @xmath118@xmath119 plane , shown in fig.[contours ] for the different models , we obtained the minimum @xmath120 of the @xmath109 function with respect to the parameters above for each leptoquark model , and we then required that @xmath121 , with @xmath122 . in this procedure , the parameters @xmath112 , @xmath113 , and @xmath123 , as well as the sm inputs @xmath116 , @xmath45 , and @xmath44 were varied so as to minimize @xmath109 .
we must comment here that the dependence on @xmath45 and @xmath44 is negligible when they are allowed to vary in their @xmath124 cl range . on the other hand ,
the variation of @xmath116 in the interval @xmath125 leads to a change on the allowed values of leptoquarks parameters of at most 1% .
the contour plots exhibited in fig .
[ contours ] were obtained for third generation leptoquarks . from this figure
we can see that the bounds are much more stringent for the leptoquarks that couple to the top quark , _
i.e. _ for @xmath126 , @xmath32 , and @xmath127 , since their contributions are enhanced by powers of the top quark mass . moreover , the limits are slightly better for left - handed leptoquarks than for right - handed ones , given a leptoquark type , and the curve is symmetric around @xmath128 since the leptoquark contributions are quadratic functions of @xmath115 .
the contributions from @xmath129 and @xmath130 are not enhanced by powers of the top quark mass since these leptoquarks do not couple directly to up - type quarks .
therefore , their limits are much weaker , depending on @xmath112 only through the sm contribution , and the bounds for these leptoquarks are worse than the present discovery limits unless they are strongly coupled ( @xmath131 ) .
moreover , the limits on first and second generation leptoquarks are also uninteresting for the same reason .
nevertheless , if we allow leptoquarks to mix the third generation of quarks with leptons of another generation the bounds obtained are basically the same as the ones discussed above - channel leptoquark exchange to some observables .
] , since the main contribution to the constraints comes from the @xmath0 widths .
we next present our results as 95% cl lower limits in the leptoquark mass and study the dependence of these limits upon all other parameters .
for this , we minimized the @xmath109 function for fixed values of @xmath123 , @xmath113 , and @xmath112 and then required @xmath132 , with @xmath133 .
our results are shown in table [ res : top ] where we give the 95% cl limits obtained for a third generation leptoquark for several values of the coupling constants @xmath115 ( @xmath134 , @xmath14 , and @xmath135 ) .
the values given correspond to @xmath136 gev and variation of @xmath137 gev and @xmath138 , which is the range associated to the best values obtained from a fit in the framework of the sm @xcite .
for a fixed value of @xmath112 and leptoquark coupling constant , the dependence on @xmath114 and @xmath113 is such that the limits are more stringent as @xmath114 increases and @xmath113 decreases .
the sm parameters @xmath116 , @xmath45 , and @xmath44 have been also varied in their allowed range . however
, this did not affect the results in a noticeable way . .
lower limits ( 95% cl ) for the mass of third generation leptoquarks in gev for different values of the couplings , assuming @xmath139 gev , @xmath140 , and @xmath141 gev .
[ cols="^,^,^,^,^,^,^,^",options="header " , ] we would like to stress that the large apparent uncertainty associated with the value of @xmath123 and @xmath113 can be considered somehow fictitious as the value of @xmath120 grows very fast when we move from the central value @xmath142 , @xmath143 gev what means that the quality of the fit for the extreme values of these parameters is rather bad .
for instance , @xmath144 , results in a too high @xmath109 , even in the context of the sm ( @xmath145 ) .
i would like to thank alan sommerer for his hospitality .
this work was supported by conselho nacional de desenvolvimento cientfico e tecnolgico ( cnpq ) and by fundao de amparo pesquisa do estado de so paulo ( fapesp ) .
l3 collaboration , b. adeva _ et al .
_ , phys .
* b261 * ( 1992 ) 169 ; opal collaboration , g. alexander _ et al .
_ , phys . lett .
* b263 * ( 1992 ) 123 ; delphi collaboration , p. abreu _ et al .
_ , phys .
* b316 * ( 1993 ) 620 .
j. l. hewett and t. g. rizzo , phys .
* d36 * ( 1987 ) 3367 ; j. l. hewett and s. pakvasa phys .
* b227 * ( 1987 ) 178 ; j. e. cieza and o. j. p. boli , phys . rev . * d47 * ( 1993 ) 837 ; j. blmein and r. rckl , phys . lett .
* b304 * ( 1993 ) 337 ; j. blmlein and e. boos , nucl . phys .
( proc . suppl . )
( 1994 ) 181 .
`` combined preliminary data on @xmath146 parameters from the lep experiments and constraints on the standard model '' , the lep collaborations and the lep electroweak working group , report in preparation . | we obtain the constraints on scalar leptoquarks coming from radiative corrections to @xmath0 physics .
we perform a global fitting to the lep data including the contributions of the most general effective lagrangian for scalar leptoquarks , which exhibits the @xmath1 gauge invariance .
our bounds on leptoquarks that couple to the top quark are much stronger than the ones obtained from low energy experiments .
= cmssbx10 scaled 2 |
in the liquid hydrogen isotopes mixture , muons assist the fusion through the formation of a muonic molecule , since the size of the muonic molecule is much smaller than that of the ordinary molecules and the fusing nuclei tend to stay closer .
this whole process takes place at the thermal energies where the conventional measurements of the fusion cross section using a charged beam can not be performed .
this mechanism of the muon catalyzed fusion ( @xmath7cf ) might provide us an unique opportunity to investigate , in a rather direct way , the fusion cross section , i.e. , the astrophysical @xmath8-factor , at extremely low energy .
for this purpose we need to know quantitatively the effect that a muon shields the coulomb potential between colliding nuclei @xcite .
and then this shielding effects of muons should be removed from the @xmath8-factor data , in order to asses the bare reaction rate correctly . from this point of view
, the @xmath0-@xmath9 fusion could provide us an elucidating example .
it has been investigated experimentally in the gas , liquid @xcite and solid @xcite targets . especially in the latter experiments
the fusion neutron energy spectrum has been determined and its distribution suggests that the fusion is followed by a sequential decay of @xmath10he : @xmath11 where @xmath10*he*@xmath12 is in the 3/2@xmath13 and 1/2@xmath13 resonant states . the @xmath0-@xmath0 fusion with muons has not received much attention as a candidate of an energy source in contrast to the @xmath14-@xmath0 and the @xmath14-@xmath14 @xmath7cf . this is partly because of the difficulty of tritium handling .
moreover it is because its reaction rate is expected to be much lower than the others due to the lack of the resonant muonic complex formation .
in fact the cycling rate obtained experimentally : [email protected] @xmath7s@xmath3 @xcite ( 15 @xmath7s@xmath3 @xcite ) of the @xmath0-@xmath16 is smaller than that of the @xmath14-@xmath16 of the order of 100 .
another distinctive difference of the @xmath14-@xmath17 and the @xmath14-@xmath18 reactions from the @xmath0-@xmath17 reaction is that their cycling rates has target density and temperature effects , which are likely caused by 3-body collisions .
put another way , if the cycling rate of the @xmath0-@xmath17 reaction does not have the target density and temperature dependences , one can verify that the dependences originate from the formation of the resonant muonic complex .
on the other hand , the cross section of the reaction @xmath19h(@xmath19h,2@xmath20)@xmath21he has been measured in the triton beam energy range @xmath2230 - 300 ( kev ) @xcite .
this energy range is much higher than thermal energies .
the astrophysical @xmath8-factor of the reaction has been studied theoretically by means of dwba @xcite and the generator coordinate method @xcite .
we determine the reaction rate of the @xmath0-@xmath16cf by considering so - called in flight " fusion @xcite and compare it with the experimental muon cycling rate . at thermal energies , where the @xmath7cf takes place , fluctuations of the cross section might play an important role .
we investigate the influence of the fluctuations by using a semi - classical method , the constrained molecular dynamics ( comd ) approach @xcite .
the molecular dynamics contains all possible correlations and fluctuations due to the initial conditions(events ) . in the comd ,
the constraints restrict the phase space configuration of the muon to fulfill the heisenberg uncertainty principle .
the results are given as an average and a variance over ensembles of the quantity of interest , which is determined in the simulation .
we make use of the average and the variance of the enhancement factor of the cross section by the muon , that have been obtained in simulations of the @xmath14-@xmath9cf @xcite , and convert them into the average and the variance of the effective potential shift .
if so , one can use the same potential shift for the case of the @xmath0-@xmath9cf , because of the isotope independence of the screening effect .
we , thus , determine the reaction rate of the @xmath0-@xmath9cf as a function of the temperature , taking into account the effect of the fluctuation of the cross section by the presence of the muon .
this paper is organized as follows . in sec .
[ sec : mrcs ] we derive the variance of the effective potential shift of the reactions between hydrogen isotopes by the muons .
the reaction rate as a function of the temperature for the @xmath0-@xmath9cf is determined in sec .
[ sec : rr ] .
we discuss possible origins of the discrepancies between the experimental muon cycling rate and the obtained reaction rate in sec .
[ sec : rd ] . in sec .
[ sec : sum ] we summarize the paper .
the influence of the muonic degrees of freedom to the reaction cross section can be taken into account as an enhancement by the screening effect @xcite .
we determine the enhancement factor at the incident center - of - mass(c.m . )
energy @xmath23 as @xmath24 where @xmath25 , and @xmath26 are the cross sections in the presence , and in the absence , respectively , of the muon .
the @xmath25 fluctuates depending on the dynamics of the 3(or n)-body system @xcite .
this fluctuation of the cross section can be written in terms of the fluctuation of the enhancement factor : @xmath27 as @xmath28 where we have assumed that the fluctuation of the bare cross section is negligible at low temperature compared to the screened one . through the molecular dynamics simulation
we obtain this fluctuation as a variance of the enhancement factor . at the same time
we determine the average enhancement factor : @xmath29 . in our previous study
we have simulated this enhancement of the cross section by the muon in the case of the @xmath14-@xmath0 @xmath7cf reaction @xcite .
we utilize the results from this simulation for the case of @xmath0-@xmath0 @xmath7cf reaction . in the left - top panel of fig .
[ fig : eu ] , the ratio @xmath30 is shown as a function of the incident c.m . energy . in the high energy limit
the ratio approaches zero , i.e. , the @xmath31 distribution becomes a @xmath32-function ( @xmath33 ) and the average @xmath34 approaches 1 ; there is no effective enhancement . in the low energy limit
the ratio @xmath30 is much larger than 1 ; this fact implies that the system exhibits a sensitive dependence of the dynamics on the initial conditions , i.e. , the muonic motion becomes chaotic .
the energy dependence of the ratio @xmath30 is approximated well by the function 2.05@xmath35 , where @xmath36 is in units of kev .
this curve is shown by the dotted line in the left - top panel in fig .
[ fig : eu ] .
we write down the enhancement factor in terms of a constant shift of the potential barrier . here
we have assumed that the enhancement is represented in terms of a constant shift of the potential barrier due to the presence of the muon .
the average @xmath37 for the @xmath38 is in good agreement with the exact adiabatic limit with the screening potential : @xmath39 where @xmath40 and @xmath41 are the binding energies of muonic tritium and muonic @xmath10he ion , respectively @xcite .
we , therefore , assume that the average potential shift @xmath42 is equivalent to @xmath43 . from the commonly used expression of the bare cross section : @xmath44 where @xmath45 and @xmath46 are the astrophysical @xmath8-factor and sommerfeld parameter .
the cross section in the presence of the muon is expressed by @xmath47 in terms of the potential shift @xmath48 .
taking the derivative of the potential shift : @xmath49 hereafter we substitute the potential shift @xmath48 in eq .
( [ eq : dsigma ] ) by its average @xmath42 and thus @xmath50 one can deduce the average potential shift @xmath42 , and its fluctuation @xmath51 from the ratio @xmath30 .
@xmath52 this potential shift is independent of the difference of isotopes , so that we make use of the same value in both cases of the @xmath14-@xmath0 and the @xmath0-@xmath17cf . in the left - bottom panel of fig .
[ fig : eu ] the fluctuation of the potential shift divided by the average potential shift is shown as a function of the incident energy .
it is striking that the slope of the ratio @xmath53 changes at the ionization energy of the muonic tritium , which we indicate by the arrow in the figure . at this incident energy
the total energy of the system is zero .
the total system is unbound at the incident energies higher than this point , while the 3-body system might be bound at lower energies .
by contrast , as it is shown in the right - bottom panel in fig .
[ fig : eu ] , in the case of the bound electron screening the binding energy of the electron is much lower than the incident energy of our interest and the ratio @xmath54 decrease monotonically as the incident energy decreases .
again the arrow indicates the ionization energy of the deuterium atom .
the irregular muonic motion leads to smaller external classical turning point @xcite . as a consequence
the irregularity makes the enhancement factor larger opposed to the result of the electron screening @xcite , where the irregular(chaotic ) events give smaller enhancement factors .
this contradiction is accounted for the fact that the system remains bound in the present case at low incident energies , while in the previous case even the lowest incident energy that was investigated is much higher than the binding energy of the electrons .
therefore the chaotic dynamics of the electrons causes to dissipate the kinetic energy between the target and the projectile and lowers the probability of fusion .
in a liquid hydrogen tritium target at the temperature @xmath55 , the velocity distribution , @xmath56 , of a pair of colliding particles is written as the maxwellian distribution , @xmath57 where @xmath23 and @xmath58 are the relative energy and the velocity of the pair of colliding particles , in the present case two tritons , and @xmath59 is boltzmann constant .
although the experiment in @xcite has been performed using the solid target , in this paper we assume eq .
( [ eq : max ] ) as the velocity distribution for simplicity .
we will reconsider the validity of this assumption afterwards .
the reaction rate per pair of particles is given by @xcite @xmath60 where @xmath25 is the reaction cross section in the presence of the muon .
the reaction rate at the liquid hydrogen density @xmath61 : @xmath62 is obtained as a function of the temperature . as we have seen in sec .
[ sec : mrcs ] , the cross section in the presence of the muon is expressed effectively with a potential shift @xmath48 , hence we write down the reaction rate : @xmath63 where @xmath64 fluctuates depending on the variance of @xmath48 .
this fluctuation of the screening potential can be incorporated through the following equation : @xmath65 where @xmath66 is the distribution of the screening potential . as the first guess , the distribution is likely to be a normal ( gaussian ) distribution . in reality , however , we find that the distribution changes characteristically as a function of the incident energy .
this change is observed at the point where the incident energy coincides with the ionization energy of the muon . in fig .
[ fig : nnc ] the left panel shows a histogram of the screening potential at a relatively high incident energy @xmath6760 ( kev ) .
the abscissa is the screening potential normalized by the average value @xmath685.6 ( kev ) .
the distribution of the histogram is approximated well by the gaussian distribution , as we expected .
moreover one sees that there are some components with negative screening potentials .
this negative component of the screening potential means that muons can be kicked out to an unbound state in some cases and bring away the relative energy of the colliding nuclei . as the incident energy is reduced , the distribution changes .
the right panel in fig .
[ fig : nnc ] shows one of such a situation : a histogram of the screening potential at the incident energy @xmath670.3 ( kev ) . in this case , because the muon is forced to remain bound the whole process in the entrance channel , the screening potential can not be negative any more .
in fact , we can approximate the distribution of the histogram with the distribution of gaussian orthogonal ensemble ( goe ) rather than a normal distribution .
we , therefore , approximate this fluctuation of the screening potential by gaussian distribution around the average potential shift @xmath42 and with variance @xmath51 : @xmath69 in the energy region @xmath70 and by goe : @xmath71 in the energy region @xmath72 . using these distributions of the screening potential
, we assess the eq .
( [ eq : rrvar ] ) .
we substitute the bare cross section by eq .
( [ eq : sigma - s ] ) and we use the polynomial expression @xcite @xmath73 where the energy @xmath23 in mev , as the astrophysical @xmath8-factor in eq .
( [ eq : sigma - s ] ) .
the obtained reaction rate by eq . (
[ eq : rrvar ] ) is shown in fig . [
fig : rtt ] by circles linked with a solid line as a function of the temperature .
the circles linked with the dotted line are obtained by taking into account the enhancement of the cross section using the average potential shift @xmath42 alone , i.e. , eq .
( [ eq : rrtu ] ) substituted @xmath48 by @xmath42 .
the triangles show the bare reaction rate .
the circles deviate from eq .
( [ eq : rrtu ] ) at the temperature lower than 10@xmath74 k and do not show much temperature dependence from 0.1 k to 10@xmath10 k. the deviation of the reaction rate by eq .
( [ eq : rrvar ] ) from the one by eq .
( [ eq : rrtu ] ) in the low temperature region indicates that the fluctuation of the enhancement factor has a strong influence on the reaction rate at the thermal energy . including the fluctuation of the cross section , the reaction rate for the @xmath0-@xmath9cf reaches at 5.0@xmath1 10@xmath2s@xmath3 .
this is a factor of 10@xmath4 smaller than the experimental muonic cycling rate in the solid tritium target , [email protected] @xmath7s@xmath3 @xcite , which is marked by the square in fig .
[ fig : rtt ] .
we remark several possible explanations of this underestimation of the rate in the following .
the first one is that the assumption of the velocity distribution of colliding nuclei to be the maxwellian could be inappropriate .
we should make use of the velocity distribution in the solid target .
otherwise the second possible explanation is that assuming that the procedure of the estimation of the reaction rate is correct , the result could mean that the actual bare @xmath8-factor must be greater than that given by eq .
( [ eq : sfact ] ) in the low energy region where the measurement has not been performed .
in this paper we have estimated the reaction rate in the muon - catalysed @xmath0-@xmath0 fusion as a function of the temperature in the framework of the `` in flight '' fusion .
we have used the comd simulations to estimate the enhancement effect of the cross section due to the muon .
we found that the estimated reaction rate does not show temperature dependence in the low temperature region in contrast to the experimental data of the @xmath14-@xmath0 and the @xmath14-@xmath75cf .
the obtained reaction rate in the low temperature : 5.0@xmath1 10@xmath2s@xmath3 underestimates the experimental muonic cycling rate in the solid tritium target , [email protected] @xmath7s@xmath3 , a factor of 10@xmath4 .
this is either because our assumption that the velocity distribution is maxwellian is not correct in the case with the solid target or because the @xmath8-factor , which we used to estimate the rate , in the low energy region is lower than the actual value . | we discuss the reaction rate of the muon - catalysed @xmath0-@xmath0 fusion . the reaction rate is determined as a function of the temperature using the model of in flight " fusion .
we especially take into account the effect of the fluctuation of the cross section due to the existence of the muon .
the obtained reaction rate 5.0@xmath110@xmath2s@xmath3 is a factor of 10@xmath4 smaller than the experimental muonic cycling rate in the solid tritium target . *
importance of fluctuations of cross sections + in muon - catalysed @xmath0-@xmath0 fusion reactions * + _ @xmath5 laboratorio nazionale del sud , infn , via santa sofia , 62 , i-95123 catania , italy + @xmath6 libera universit kore , via della cittadella 1 , i-94100 enna , italy _ |
the thermopower of a two - dimensional electron gas ( 2deg ) @xcite has been attracting interest not only as a route to access its thermodynamic properties but also as a sensitive tool to probe various quantum phenomena that take place in a quantizing magnetic field ( see , e.g. , @xcite ) .
the thermopower in a 2deg contains contributions from two separate mechanisms : diffusion and phonon drag .
it is well known that the latter is by far the dominant contribution in standard experiments using an external heater to introduce temperature gradient @xcite .
this is because the heater raises both the lattice and the electron temperatures alike ; the heat current is thus predominantly carried by phonons , which generates the phonon - drag thermovoltage through the electron - phonon interaction .
however , it is the diffusion thermopower that is expected to be more sensitive to the phenomena taking place in a 2deg .
furthermore , the experimental results for diffusion thermopower will be much easier to interpret , since external complications , the phonons , are not involved .
therefore , it is desirable to have a method sensitive only to the diffusion contribution .
this can be achieved by employing current heating technique , which induces gradient only in the electron temperature @xmath6 , leaving the lattice temperature intact .
the technique was applied to a micro - scale ( 4@xmath78 @xmath8 ) hall bar by maximov _
@xcite to obtain diffusion contribution to the longitudinal ( @xmath9 ) and transverse ( @xmath3 ) thermopower in the low magnetic field regime @xmath10 t at a temperature @xmath0=1.6 k. their use of the micro - hall bar , however , resulted in rather large slowly - varying background attributable to the quasiballistic motion of electrons . in the present paper
, we describe our attempt to acquire diffusion thermopower at dilution - refrigerator temperatures @xmath5 40 mk , using a hall bar designed to be well suited for the measurement of the thermopower , and having dimensions larger than the mean - free path of the electrons to avoid the intervention by the ballistic electrons .
we make an attempt to extend the measurement to the quantum hall regime , @xmath11 t , employing a hall bar device equipped with a front gate on the section used as a heater to circumvent the problem ( to be discussed below ) encountered in the quantum hall regime .
279 @xmath12m@xmath13 ) contains three pairs of voltage probes ( 46 , 810 ) .
secondary ( vertical , from 3 to 11 ) hall bar ( 170@xmath750 @xmath12m@xmath13 ) is employed as the heater ; the voltage probes ( 2,11 ) are used for the sdh measurement .
a ( gray ) shaded rectangle is a metallic front gate to control the resistance of the area to be heated . ] a conventional gaas / algaas 2deg wafer with the carrier density and mobility @xmath14 @xmath15 @xmath16 m@xmath17 and @xmath12 @xmath15 70 m@xmath18/(vs ) , respectively , is patterned into the device geometry illustrated in fig .
the device is composed of two crossing hall bars .
the main hall bar ( between ohmic contacts 7 and 13 ) has a width @xmath19=50 @xmath12 m and length @xmath20= 279 @xmath12 m and contains three sets of the voltage probes ( with contacts 46 and 810 ) to measure the longitudinal ( @xmath21 ) and transverse ( @xmath22 ) voltages at three different locations ( or with different inter - probe distances for @xmath21 ) .
both @xmath19 and @xmath20 , and distances between voltage probes ( @xmath23=23 @xmath12 m or @xmath24=153 @xmath12 m ) , are designed to be much larger than the mean - free path @xmath25=7.3 @xmath12 m of the electrons
. the secondary hall bar ( between contacts 3 and 11 ) , 170 @xmath12m - long and 50@xmath12m - wide , is used as a heater by driving an ac heating current @xmath26=50200 na , with frequency @xmath27=13 hz .
the current @xmath26 used is much larger than that in the ordinary resistivity measurement ( @xmath28=0.510 na ) and raises the electron temperature @xmath29 through joule heating , but is kept small enough to prevent the heating of the lattice .
we can probe the electron temperature in the heater section by the voltage probes ( contacts 2 and 12 ) , exploiting the amplitude of the shubnikov - de haas ( sdh ) oscillation .
thus the difference in the electron temperature @xmath30=@xmath6(high)@xmath31(low ) is introduced to the main hall bar , between @xmath29(high ) at the crossing region shared with the secondary hall bar ( dubbed as `` heater area '' henceforth ) and @xmath29(low ) at the ohmic contact ( contact 7 ) composed of diffused niauge alloy , the latter assumed to be in equilibrium with the lattice temperature , or the temperature of the mixing chamber of the dilution refrigerator . strictly speaking , ohmic contacts for the voltage probes ( contact 46 , 810 ) are also at @xmath29(low ) , and therefore the temperature gradient is also directed from the heater area toward these contacts , resulting in rather complicated temperature distribution .
to minimize the disturbance by the voltage - probe contact pads , arms are designed to be thin ( 3 @xmath12 m ) and long ( 247 @xmath12 m ) , and the pads are made much smaller than that of contact 7 to diminish their efficiency as the heat sink . since @xmath32 , thermopower @xmath9 and @xmath3 are obtained by detecting the component of @xmath21 and @xmath22 having the frequency @xmath33=26 hz by using a lock - in amplifier ; then we have @xmath34 and @xmath35 .
the component of voltages with the frequency @xmath27 , on the other hand , yields non - local resistance .
we note in passing that since we are using a hall bar , we can , of course , also measure the longitudinal and the hall resistances ( which can readily be translated to the resistivities @xmath36 and @xmath37 ) for the same area of the sample as the thermovoltages are acquired , simply by passing ( small ) current between contacts 7 and 13 . with the method described so far , we succeeded in measuring the transverse ( nernst ) component @xmath3 of the diffusion thermopower for low magnetic fields @xmath38 t , as will be shown in the next section . the measurement of the longitudinal ( seebeck ) component @xmath9 is still suffering from the effect of electron deflection due to the magnetic field , which causes the mixing - in of the transverse component asymmetrically between @xmath39 and @xmath40 .
we therefore focus on the @xmath3 component , acquired by using contacts 4 and 10 as voltage probes , in the present paper .
a problem arises in applying the current heating technique in a magnetic field : since the longitudinal resistance varies with magnetic field also in the heater area , the temperature difference @xmath30 generated by the joule heating also varies with magnetic field .
this does not cause serious trouble in low magnetic fields where the resistance variation is not so large . in the quantum hall regime
, however , resistance variation is phenomenal , ranging from @xmath50 at the quantum hall states to @xmath5k@xmath41 in between . to avoid the difficulties
, we placed a metallic front gate on the heater area , as shown in fig .
this enabled us to control the carrier density , hence the resistance , of the heater area independent of whether or not the measured area is in the quantum hall states .
measured directly ( solid line ) and that calculated from the measured @xmath36 and @xmath37 using eq .
( [ eq : nernstb ] ) ( dotted line ) , at @xmath0 @xmath15 40 mk .
the latter is offset by 1.5 @xmath12v / k for clarity . ]
[ fig2 ] shows the transverse thermopower @xmath3 at low magnetic fields ( @xmath42 t ) measured by the method described in the previous section , with a heating current @xmath26=200 na .
( a sample without the metallic front gate is used for this measurement . ) note that @xmath3 oscillates around zero , without any noticeable background .
the diffusion thermopower is related to the longitudinal and transverse conductivities @xmath43 and @xmath44 by the generalized mott formulas @xcite , @xmath45 @xmath46 where @xmath47 is the lorenz number and @xmath48 the fermi energy .
if we assume that properties of the system are mainly determined by the location of the fermi energy with respect to the landau levels , we can identify the energy derivative with the derivative with respect to the magnetic field as @xmath49 using this relation , eqs .
( [ eq : seebecke ] ) and ( [ eq : nernste ] ) are rewritten as @xmath50 @xmath51 where we replaced the conductivities by the resistivities by inverting the tensor .
we can thus calculate @xmath9 and @xmath3 using @xmath36 and @xmath37 measured in the same hall bar . in fig .
[ fig2 ] , we also plot @xmath3 calculated by eq .
( [ eq : nernstb ] ) .
the good quantitative agreement between the two traces confirms that the measured @xmath3 is actually derived from the diffusion contribution .
discrepancy noted at very low magnetic fields ( @xmath520.05 t ) can be traced back to the negative magnetoresistance in @xmath36 , presumably originating from the electron interactions @xcite .
the absence of the similar effect in the thermopower is intriguing , but the reason is currently unknown .
measured with two different gate voltages @xmath53 , with the same heating current @xmath26= 50 na .
@xmath0 @xmath15 40 mk . ] next , we check the function of the front gate . in fig .
[ fig3 ] , we plot traces of the transverse thermovoltage @xmath22 measured with two different values of the gate voltages @xmath53 , using the same heating current @xmath26=50 na . with more negative @xmath53 , resistivity in the heater area increases accompanying the decrease in the electron density .
the heater area therefore achieves higher electron temperature with the same @xmath26 , resulting in larger amplitude in @xmath22 .
the two traces virtually overlap each other , when translated into @xmath3 by using measured values of @xmath30 . as a function of @xmath54 and @xmath53 .
@xmath26= 200 na .
( b ) cross sections of ( a ) at fixed magnetic fields , plotted as a function of @xmath55 .
solid line : @xmath54=1.29 t. dashed line : @xmath54 @xmath15 2.69 t. ( c ) cross section of ( a ) with a fixed filling factor @xmath56= 3.5 of the heater area , plotted as a function of @xmath54 .
cross sections in the @xmath54-@xmath53 plane are indicated in ( a ) by the same line - type as in ( b ) and ( c ) . ]
we now move on to the quantum hall regime .
[ fig4 ] ( a ) shows color - scale intensity plot of @xmath22 in the @xmath54-@xmath53 plane .
the plot was obtained by the repetitive @xmath54-sweeps with fixed @xmath53 , varying the @xmath53 step by step with the increment of 0.005 v. note that @xmath53 alters the filling factor @xmath56 of the heater area for a fixed @xmath54 ; the region of the sample to be measured is not affected by @xmath53 .
finite signal of @xmath22 appears only when both the measured area and the heater area are in the resistive ( the inter - quantum - hall transition ) regime ; temperature gradient is not generated unless the heater area possesses a finite resistivity , and @xmath3 equals zero for @xmath36=0 and a finite value of @xmath37 in the measured area ( see eq .
( [ eq : nernstb ] ) ) .
it can be seen that the plot considerably changes its appearance below and above @xmath57 t. below @xmath51.8 t , the @xmath22 signal is basically determined by @xmath54 and does not depend much on @xmath53 as long as the heater area is in the resistive regime . above @xmath51.8 t , by contrast , @xmath22 exhibits rather complicated pattern dependent both on @xmath54 and @xmath53 .
the difference is more apparent in the cross sections , shown in fig .
[ fig4](b ) , at fixed values of @xmath54 . for @xmath581.8 t
, the cross section shows trains of dips ( or peaks , depending on the value of @xmath54 ) , with the depth slightly varying with @xmath53 attributable to the slight difference in the resistivity of the heater area . for @xmath591.8 t ,
on the other hand , @xmath22 alternates sign with @xmath53 , with the sign reversal taking place roughly at the half - filling of the landau levels in the heater area , namely at @xmath60 half integer .
the latter is a rather anomalous behavior that defies simple interpretation , since , as noted above , @xmath53 affects only the heater area , leaving the measured area intact .
we currently have no clear explanation for this observation .
a possible origin is the non - uniform heating of the heater area in the quantum hall regime .
it has been shown , both experimentally @xcite and theoretically @xcite , that a 2deg subjected a large current develops a distribution in the electron temperature in the quantum hall regime owing to the ettingshausen effect .
the distribution varies rapidly with the filling factor ; the hot and cold region alternates roughly at integer and half integer filling factors @xcite . apparently , detailed knowledge of the temperature distribution both in the heater area and the measured area is necessary for the interpretation of the measured thermovoltage .
we would also like to point out that the border field @xmath611.8 t coincides with the onset of the spin splitting in the measured area , suggesting the possible involvement of the spins in the anomalous behavior . to avoid the complications arising from the behavior of the heater area , we look at the data at fixed @xmath56 . a cross section of fig . [ fig4 ] ( a ) at @xmath56= 3.5 is shown in fig .
[ fig4 ] ( c ) .
the trace is more or less the extension of the low - field traces , figs .
[ fig2 ] and [ fig3 ] , anticipated for the diffusion contribution to the @xmath3 , albeit with rather large noise in the high - field regime .
we have measured diffusion contribution to the transverse thermopower @xmath3 , by using a specially designed hall bar device having dimensions much larger than the mean - free path of the electrons .
we obtain good quantitative agreement with @xmath3 calculated from the resistivities using the generalized mott formula eq .
( [ eq : nernstb ] ) in the low magnetic field regime ( @xmath2 1 t ) . in the quantum hall regime ( @xmath4 1 t )
, we find , by mapping out the transverse thermovoltage @xmath22 in the @xmath54-@xmath53 plane , that @xmath22 signal appears only when both the measured and the heater area are in the resistive regime and that @xmath22 exhibits rather anomalous @xmath53 dependence above @xmath51.8 t. the results suggest that the diffusion thermopower can be obtained also in the quantum hall regime by picking out the @xmath22 at the constant filling factor @xmath56 at the heater area .
a. e. acknowledges the financial support by grant - in - aid for scientific research ( b ) ( 20340101 ) from the ministry of education , culture , sports , science and technology ( mext ) and by the national institutes of natural sciences undertaking forming bases for interdisciplinary and international research through cooperation across fields of study and collaborative research program ( no .
nifs08kein0091 ) .
r. fletcher , j. c. maan , k. ploog , g. weimann , phys .
b 33 ( 1986 ) 7122 .
b. l. gallagher , p. n. butcher , handbook on semiconductors , vol . 1 , elsevier , amsterdam , 1992 , p. 817 . r. fletcher , semicond .
technol . 14 ( 1999 ) r1 .
s. goswami , c. siegert , m. baenninger , m. pepper , i. farrer , d. a. ritchie , phys .
lett . 103 ( 2009 ) 026602 .
w. e. chickering , j. p. eisenstein , j. l. reno , phys .
lett . 103 ( 2009 ) 046807 .
k. yang , b. i. halperin , phys .
b 79 ( 2009 ) 115317 .
g. granger , j. p. eisenstein , j. l. reno , phys .
102 , ( 2009 ) 086803 .
s. maximov , m. gbordzoe , h. buhmann , l. w. molenkamp , d. reuter , phys .
b 70 ( 2004 ) 121308 .
m. jonson , s. m. girvin , phys .
b 29 ( 1984 ) 1939 .
l. li , y. y. proskuryakov , a. k. savchenko , e. h. linfield , d. a. ritchie , phys .
lett . 90 ( 2003 ) 076802 .
y. komori , t. okamoto , phys .
b 71 ( 2005 ) 113306 .
t. ise , h. akera , h.suzuura , j. phys .
74 ( 2005 ) 259 . | we have measured diffusion thermopower in a two - dimensional electron gas at low temperature ( @xmath0=40 mk ) in the field range 0 @xmath1 3.4 t , by employing the current heating technique .
a hall bar device is designed for this purpose , which contains two crossing hall bars , one for the measurement and the other used as a heater , and is equipped with a metallic front gate to control the resistivity of the areas to be heated . in the low magnetic field regime ( @xmath2 1 t ) , we obtain the transverse thermopower @xmath3 that quantitatively agrees with the @xmath3 calculated from resistivities using the generalized mott formula . in the quantum hall regime ( @xmath4 1 t ) , we find that @xmath3 signal appears only when both the measured and the heater area are in the resistive ( inter - quantum hall transition ) region .
anomalous gate - voltage dependence is observed above @xmath51.8 t , where spin - splitting in the measured area becomes apparent .
diffusion thermopower , quantum hall effect , two - dimensional electron gas |
the presence of a bar in a galaxy plays an important role in its evolution .
approximately two - thirds of all spiral galaxies are barred ( e.g. @xcite ; @xcite ) , so any complete picture of galaxies must include the impact of bars on both their nuclear and extra - nuclear properties .
secular ( i.e. internal ) galaxy evolution involving bars , oval distortions and spiral structure can have a significant effect when a galaxy is no longer merging or interacting strongly with other galaxies ( see @xcite ) .
for example , there is believed to be a causal connection between the existence of a bar and a galaxy s circumnuclear properties .
bars can drive gas from the outer disk to the inner regions of the galaxy ; gas clouds are shocked as they encounter the leading edge of the bar , lose angular momentum , and flow towards the galaxy s center @xcite .
an observable result of this is that a substantial fraction of galaxies with bars show enhanced star - formation activity in their central regions , and the inflow of material is also believed to build ` pseudobulges ' , which are distinct from classical galaxy bulges @xcite .
a strong global density pattern such as a bar can also produce a series of resonances between the angular velocity of the bar and the orbital motions of stars and gas in the disk , where rings can form @xcite .
indeed , some galaxies with bars are known to have circumnuclear rings of star formation ( e.g. @xcite ; @xcite ) , resulting from increased gas density at or near the inner lindblad resonance ( ilr ) .
a question that we are addressing with this study is , can the material driven to the center of a galaxy by the bar also fuel an active galactic nucleus ( agn ) ?
while there is clear observational evidence that bars in disk galaxies cause higher central concentrations of gas ( and hence centrally concentrated star formation ) compared to non - barred galaxies , it is unclear whether a large - scale bar is sufficient to transport material close enough for agn feeding ( typically on au scales , see @xcite for a review ) .
additionally , in an x - ray study by @xcite using _ ginga _ , _ asca _ and _ bepposax _ data , a strong correlation was found between absorbing column density toward type-2 seyfert nuclei and the presence of a strong bar , i.e. more than 80% of compton - thick seyfert 2s are barred .
this suggests that if a low - luminosity agn ( llagn ) is present , it may be obscured from view at optical wavelengths and/or diluted by strong star - formation activity , which could lead to a misclassification as a pure star - forming nucleus .
the topic of the influence of bars on galaxy evolution is not well studied at x - ray energies , but we are now able to utilize the excellent complementary imaging and spectral capabilities of _ chandra _ and _ xmm - newton _ to make detailed investigations of the x - ray characteristics of these systems . in particular , high - spatial - resolution x - ray observations with _
provide an ideal method for searching for low - level agn activity in nearby galaxies , in the form of a hard central point source that may not be visible at other wavelengths .
it also allows us to separate potential agn activity from other types of x - ray sources that may reside close to the nucleus , such as x - ray binaries ( xrbs ) associated with circumnuclear star - formation .
indeed , x - ray surveys utilizing _ chandra _ have proved successful in recent years in identifying previously unknown llagn activity in nearby galaxies .
in particular , x - ray nuclei spanning a wide range of x - ray luminosities ( @xmath5erg s@xmath2 , 210kev ) have been detected in a large fraction ( typically @xmath64570% ) of galaxies hosting low - ionization nuclear emission regions ( liners ; @xcite ) , as well as galaxies classed as having hii or transition nuclei ( e.g. @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ; @xcite ) .
studies at other wavelengths have also recently been used to uncover previously unknown agn activity in nearby galaxies .
for example , _ spitzer _ infrared ( ir ) spectroscopic surveys have revealed bonafide agn in both liners and ir - bright galaxies , some of which showed no sign of such activity in optical studies ( @xcite ; @xcite ) .
high - resolution radio imaging has been shown to be an efficient method for finding llagn activity ; compact radio nuclei have been found in @xmath7 50% of liners / llagn , which are thought to be attributable to relativistic jets produced by accreting super - massive black holes ( @xcite and references therein ) . to begin to address the nature of the nuclear ( and off - nuclear ) x - ray emission in barred galaxies ,
we have conducted a pilot study of the barred galaxy ngc 1672 , using x - ray observations from both _
chandra _ ( 40ks ) and _ xmm - newton _ ( 50ks ) . here
we report the results from these observations , and use supporting _
hst_/acs , _ spitzer _ and atca radio observations to assist in the interpretation of the x - ray results .
this paper is laid out as follows . in
[ sec:1672 ] we give background information on the target galaxy ngc 1672 . in
[ sec : data ] , we describe the _ chandra _ , _ xmm - newton _ , _
hst_/acs , _
spitzer _ and atca radio observations and data analysis techniques , and in
[ sec : results ] we present our results . in
[ sec : disc ] we discuss the nature of the nuclear emission , and in [ sec : compare ] we compare our results with those for other barred and non - barred galaxies in the literature . throughout this paper
we use a distance to ngc 1672 of 16.3mpc , corresponding to a recession velocity of 1140 km s@xmath2 relative to the centroid of the local group @xcite , with @xmath8=70 km s@xmath2 mpc@xmath2 .
for this pilot study , we have selected the nearby ( 16.3mpc ) late - type barred spiral galaxy ngc 1672 ( sb(r)bc ) .
it has a moderately low inclination angle ( @xmath9 , @xcite ) , and is known to have a strong bar ( 2.4 in length , corresponding to 11.4kpc at a distance of 16.3mpc ) , many hii regions in its four spiral arms , plus vigorous star formation at the ends of the bar ( @xcite , and references therein ) .
it has a high ir luminosity ( log @xmath10 ) , and a global star formation rate ( sfr ) of 2.7@xmath11 yr@xmath2 @xcite .
optical studies have given conflicting evidence as to the nature of the nuclear activity in ngc 1672 .
it was noted by @xcite as having a complex or ` amorphous ' nuclear morphology , and was subsequently shown by @xcite to possess a circumnuclear ring of star formation , measuring @xmath12 ( @xmath13kpc at a distance of 16.3mpc ) , located between two ilrs .
therefore , studies of its nuclear properties that use large apertures are strongly affected by the star - formation in the ring . for example
, @xcite showed that the optical emission spectrum in the central 17 was similar to those of normal hii nuclei , but with an h@xmath14/h@xmath15 ratio indicative of a large amount of dust reddening .
@xcite also measured optical emission - line ratios in a relatively large @xmath16 aperture , and classified the nucleus as ` composite ' , with ratios between those of typical starburst and seyfert values .
more recent optical line ratio diagnostics by @xcite of the emission within a @xmath17 slit are indicative of an hii - type nucleus , with no h@xmath14 line broadening detected .
all of these studies will have been strongly contaminated by the emission from the circumnuclear starburst ring ; however they do demonstrate that this component is dominant , and that any potential agn activity would be relatively weak .
more detailed , spatially resolved studies have also given ambiguous results .
@xcite detected possible broadening of [ oiii ] ( @xmath6300 km s@xmath2 ) compared to h@xmath15 lines ( @xmath6150 km s@xmath2 ) in the central @xmath18 , which they suggested was evidence of a composite hii / seyfert 2 nucleus .
@xcite performed a high - spatial - resolution spectral analysis of the nucleus , and found a strong increase in the [ oiii]/h@xmath15 ratio in the central 1 compared to its immediate surroundings .
however , both lines had the same fwhm of @xmath6300 km s@xmath2 , and the authors classified the nucleus as a liner .
@xcite also measured spatially - resolved ( 2 ) emission - line ratios , which were better modeled by a liner - type stellar photoionization model with @xmath19k , rather than photoionization by a strong agn continuum .
ngc 1672 has previously been observed with at x - ray wavelengths with _
rosat _ and _ asca_. three bright x - ray sources were detected in the soft 0.22.0kev band with the _ rosat _ hri and pspc , which were clearly associated with the galaxy ( @xcite ; @xcite ) .
the brightest source was located at the nucleus , and the other two were cospatial with the ends of the bar .
the nucleus had a soft x - ray spectrum consistent with thermal emission with a temperature of 0.68kev , with a soft x - ray luminosity of @xmath20erg s@xmath2 ( 0.22.0kev ) .
however , these studies were again unable to resolve the nuclear emission spatially due to the 5 fwhm of the hri point - spread - function ( psf ) , and so could not determine whether the bulk of the soft emission came from a starburst or agn . furthermore , the _ asca _ data showed no evidence of significant hard ( 210kev ) emission from the nuclear source , with the hard emission in the galaxy being dominated by two off - nuclear x - ray sources ( x-3 and x-7 , see figure 4b in @xcite ) .
thus , it was concluded that if an active nucleus is present in ngc 1672 , it must be compton - thick , with @xmath21 @xmath22 .
ngc 1672 was observed with _
chandra _ for 40ks on april 30 , 2006 ( obsid 5932 ) using the back - illuminated s3 chip of the advanced ccd imaging spectrometer ( acis ) .
the field of view of the observation is shown in figure [ fig : dss_fov ] ; the nucleus of the galaxy was placed at the aimpoint of the s3 chip , and the observation encompasses the entire @xmath0 = the apparent major isophotal diameter measured at the surface brightness level @xmath23 mag / arcsec@xmath24 extent of the galaxy .
the data were initially processed with the _ chandra _ interactive analysis of observations ( ciao ) tool suite version 3.3 , with caldb version 3.2.1 .
the processing included a charge transfer efficiency ( cti ) correction and time - dependent gain correction .
further data analysis was performed using ciao version 3.4 .
a full - field light curve was extracted from the level-2 event list to search for periods of high background , but no strong flares were found , leaving a net exposure time of 39.6ks . the basic data reduction was done using standard ciao data analysis threads .
images were created of an 8@xmath258 region ( encompassing the full @xmath0 area of the galaxy ) at full 0.492 pixel resolution in soft ( 0.31kev ) , medium ( 12kev ) , hard ( 210kev ) and broad ( 0.310kev ) energy bands .
monochromatic exposure maps were created at 0.8 , 1.0 and 2.0kev , corresponding to the peak energy in the soft , medium and hard band data .
the images were adaptively smoothed using the csmooth task , with a fairly low significance of 2@xmath26 above the local background ( corresponding to @xmath6130 pixels in the body of the galaxy ) to avoid over - smoothing of faint features in the central region ( see [ sec : results_morph ] ) .
initial detection of candidate sources was carried out in each energy - band image using the ciao task wavdetect , a mexican hat wavelet source detection algorithm .
wavelet scales of 1.0 , 1.414 , 2.0 , 2.828 , 4.0 , 5.657 & 8.0 pixels were used , and the significance threshold was set to @xmath27 corresponding to approximately one false detection per band over the 8@xmath258 field .
a merged source list was created that included sources detected in all bands , which was used as the input list for source parameterization in the acis extract idl software ( ae ; @xcite ) .
the main steps in the procedure are ( 1 ) the definition of source extraction regions based on the 90% contour of the local psf , ( 2 ) trial extraction of events within those regions , ( 3 ) improvement of positions if necessary ( see below ) , ( 4 ) full extraction of source and background spectra using final positions , and the creation of associated rmfs and arfs for each source .
source positions were improved ( in step 3 ) using the check_positions stage , which computes mean data " and correlation " positions for all sources .
the mean data " position corresponds to the central position of the extracted in - band events in the 0.310kev band , while the correlation " positions are obtained by convolving the data image with a position - dependent model psf .
the offsets between these and the original wavdetect positions were typically 0.10.8 .
the position estimates were visually reviewed , and the best was chosen for each source . where there was little difference between the estimates , the original position was retained .
the source extraction regions were based on the 90% contour of the local psf at the final positions .
background spectra were extracted in local annuli , with inner radii of 1.1 times of the 99% radius of the psf ( to exclude any residual source flux ) , and outer radii tailored so that each background spectrum contains at least 100 counts . for the central ( ring ) region , the whole area ( 17 radius ) was masked out from the background map due to the clumpy diffuse emission in the area , and the background subtraction for sources in this region ( except for the nucleus ) was performed using the area immediately outside this . due to the complexity of the central region , the spectrum of the nuclear x - ray source was extracted manually in ciao using specextract .
the source aperture was set to 1.2 radius , centered on the position from wavdetect ( 04@[email protected]@xmath30 , @xmath31591450.1 ) .
the local background flux was measured in an annulus immediately outside this ( 1.22.0 radius ) , and was chosen using the radio contours as a guide in order to cover as much of the area inside the ring , without encountering any ring emission ( see [ sec : results_morph ] ) .
any other typical non - local background region would have left residual soft emission in the nuclear source counts .
since there is significant diffuse emission in the circumnuclear ring area , it is possible that some of the candidate sources detected there may be clumpy gas rather than true point sources . to test this , we performed psf fitting using the ae task ae_radial_profile , which computes a 1-sided ks probability to determine whether the observed counts in the 12kev energy range are consistent with the model local 1.49kev psf encircled energy distribution .
six candidate sources in the ring region were inconsistent with the local psf profile with a significance of @xmath32 .
sources without detections in the hard 210kev band and with a ks probability @xmath32 were deemed to be diffuse and discarded .
the final _ chandra _ catalog contains 28 sources inside the @xmath0 ellipse , and their properties are listed in table [ tab : catalog ] .
sources with significance of @xmath33 were retained , including faint _
counterparts to two significantly detected _ xmm - newton _ sources ( for variability and completeness ) .
this detection threshold corresponds to @xmath3412 net source counts .
source fluxes were determined using spectral fitting ( see [ sec : source_fits ] ) .
the absolute astrometry of _ chandra _ is generally accurate to within @xmath61 . to confirm this ,
the x - ray positions were compared to the 2mass point source catalog .
three counterparts were found in the field ; the offsets between the x - ray and 2mass positions were small ( 0.14 , 0.16 & 0.6 ) , and in non - uniform directions , with the greatest offset for a source 4 off - axis at the edge of the field . given this accuracy , no astrometric correction was made to the _ chandra _ positions .
ngc 1672 was observed with _
xmm - newton _ for 50ks on november 27 , 2004 ( obsid 0203880101 ) .
the epic mos-1 , mos-2 and pn cameras were operated with medium filters in prime full window mode .
full - field light curves were accumulated for the three exposures in the 1015kev band to check for high background intervals of soft proton flares .
the data were found to be clean for the duration of the observation , and the full time intervals were used in subsequent analysis : 47.9ks ( pn ) , 49.4ks ( mos1 ) & 49.5ks ( mos2 ) . all data ( images and spectra ) were created using standard patterns corresponding to single and double pixel events for the pn ( 04 ) , and patterns 012 ( single to quadruple events ) for the mos cameras . for spectral extraction , the quality flag was set to zero to reject events from bad pixels and events too close to the ccd chip edges .
multi - band source detection was performed on the pn and mos data , using the methods detailed in @xcite . in brief , images were created in the following energy bands : soft ( 0.31kev ) , medium ( 12kev ) and hard ( 210kev ) .
exposure maps were created for each instrument / energy band using the sas task eexpmap , and detector masks were created for each camera with emask .
two methods of initial source recognition were used : eboxdetect , a sliding - box algorithm that simultaneously searches the three energy band images , and ewavelet , a mexican hat wavelet detection routine that works on individual images . both used background maps created with asmooth .
the source lists from both methods were then parameterized using emldetect , which performs maximum likelihood psf fits to each source .
these were then compared , and any additional sources detected with ewavelet were added to the eboxdetect source list , and the parameterization repeated .
all sources significant at @xmath34 4@xmath26 in at least one energy band , and within the @xmath0 ellipse of ngc 1672 , were retained and visually inspected .
spectra , arfs and rmfs were extracted from the data for each source using the sas task especget .
the final _ xmm - newton _ catalog contains 19 extra - nuclear sources , and their properties are listed in table [ tab : catalog ] along with the corresponding _ chandra _ properties .
the source fluxes were again determined using spectral fitting ( see [ sec : source_fits ] ) .
note that six sources ( including the nucleus ) within the central @xmath620 radius are not spatially resolved with _
xmm - newton _ ; these are marked with ( n)=nucleus and ( c)=central region source .
apart from # 10 , 17 and 24 , all other sources detected in the _ chandra _ observation are detected in the _ xmm - newton _ observation .
to estimate source fluxes reliably , we have performed spectral fitting on each source using xspec v12 .
detailed spectral fitting was performed on _ chandra _ sources with @xmath34 100 net counts , and on _ xmm - newton _ sources with @xmath34 100 net counts in either the pn or mos data .
these spectra were binned ( grouped to a minimum of 20 counts per bin ) and fitted , using @xmath35 statistics , with both powerlaw+absorption ( po*wabs ) and disk blackbody+absorption ( diskbb*wabs ) models . in all cases , the powerlaw model provided the best fit .
spectral parameters and corresponding 0.310kev fluxes and luminosities are shown in table [ tab : catalog ] .
for the _ xmm - newton _ data , the pn , mos1 and mos2 spectra were fitted simultaneously .
the fluxes and errors from the three instruments were combined to provide a weighted mean flux for each source using the method of @xcite.barlow / statistics.html ] for sources with @xmath36 100 net counts , a simple powerlaw+galactic absorption model was fitted to unbinned spectra using the c - statistic @xcite . since at such low count rates source parameters are difficult to constrain , the powerlaw and absorption parameters were frozen ( @xmath37 , @xmath38 @xmath22 ) , with the normalization as the only free parameter .
the resulting fluxes and luminosities are also shown in table [ tab : catalog ] .
the source properties are described in [ sec : results_sources ] .
ngc 1672 was observed by the _ hubble space telescope ( hst ) _ on august 1 , 2005 using the advanced camera for surveys ( acs ) wide field channel ( wfc ) in the broad - band f435w , f550 m and f814w filters , plus the narrow - band f658n ( h@xmath14 ) filter .
a large dither was implemented , covering the dimensions of one of the wfc chips in order to fit the visible disk of the galaxy into each observation ( see figure [ fig : dss_fov ] ) .
exposure times were 2444s in each observation , split into four separate observations corresponding to the visibility of ngc 1672 in each single orbit .
we also include in our analysis an archival acs high resolution channel ( hrc ) f330w uv observation of the nucleus of ngc 1672 , taken july 19 , 2002 , with an exposure time of 1140s .
all of the acs observations were processed through the standard _
multidrizzle _ pipeline @xcite . as sub - pixel dithers
were not used in the observations , the drizzle parameters pixfrac and scale were set to unity , preserving the native pixel scale of both the hrc and wfc .
each of the acs wfc images were astrometrically registered with 2mass stellar objects falling within the fov , yielding an astrometric fit with an rms of @xmath6 0.15 .
the f658n image is significantly contaminated by continuum emission within the h@xmath14 passband ; we therefore created a continuum - subtracted image ( denoted f658cs ) using a continuum image estimated from a linear combination of the f550 m and f814w images ( weighted by their separation from the h@xmath14 line , and multiplied by the relative sensitivity of each filter to obtain the same physical units in each image ) .
the acs images have a very high spatial resolution of @xmath60.1 .
they show that , while the optical emission peaks in the central 2 nuclear region , it is clearly non - point like with a fwhm of @xmath630 pixels ( 1.5 ) .
indeed it displays a complex / irregular morphology , and in a visual inspection we found no obvious point sources that could correspond to , for example , a nuclear star cluster ( which are generally only marginally resolved " in hst images of nearby galaxies , @xcite ) .
the complexity and extended nature of the region suggests that the use of any moderate sized photometric aperture will result in a significant overestimate of the true flux from the nucleus .
we investigated this in three ways : ( 1 ) via psf subtraction , ( 2 ) narrow aperture photometry , and ( 3 ) photometry using the _ chandra _ aperture . to estimate the maximum magnitude of any point source at the center of the galaxy
, psfs were constructed from isolated point sources in each image and subtracted from the centroid of the galaxy light in each band , scaled such that their cores resulted in a roughly zero flux value for the brightest pixel in the galaxy core . for two filters it was not possible to obtain actual psfs from the images . in the case of f330w ,
the hrc field of view is too small for any bright , isolated sources to be present , while in our f658n continuum - subtracted image the stars were largely subtracted . to overcome this
, we used tinytim to create artificial psfs for these filters , which were then subtracted from the data .
the resulting upper limits are shown in table [ tab : acs_photom ] .
an alternative to direct psf subtraction is to use a narrow aperture for the determination of the light within the nucleus , and then to apply the aperture corrections tabulated by @xcite .
we used a 0.1 ( 2 pixel ) radius aperture , and subtracted the background using an 1.22 annulus centered on the nucleus . to compare directly with the _ chandra _ observations , we have also performed photometry using the same physical aperture ( 1.2 radius ) and background annulus ( again 1.22 )
the resulting magnitudes from these methods are also shown in table [ tab : acs_photom ] .
the upper limits measured using the psf subtraction method are similar to the measurements obtained from the narrow aperture , though there is a good deal of scatter between these and the larger _ chandra _ aperture measurements . given that psf subtraction gives the most stringent constraints on the magnitude of a nuclear point source , we adopt these limits for the multiwavelength comparison ( see table [ tab : acs ] ) , and the construction of the nuclear spectral energy distribution ( sed , see [
sec : sed ] ) .
the ab magnitudes were converted to flux density using abmag = @xmath312.5 log @xmath39 @xmath31 48.6 , and then into luminosity using @xmath40central filter frequency .
we note that the centroid of the emission in the f330w images does not correspond with that seen in data taken at longer wavelengths , which suggests that there is some extinction along the line - of - sight .
ngc 1672 was observed with the _ spitzer _ infrared array camera ( irac , @xcite ) on november 29 2006 ( program i d 30496 , pi fisher ) .
basic calibrated data ( bcd ) were retrieved from the archive at 3.6 , 4.5 , 5.8 and 8.0@xmath41 m .
we performed post - bcd processing ( mosaicing , with a plate scale of 0.86 arcsec / pixel ) using the _ spitzer _ data - processing package mopex @xcite .
the images are flux calibrated in the pipeline in units of surface brightness ( mjy sr@xmath2 ) , and we converted them to flux density units by multiplying by a conversion factor of 1mjy sr@xmath2 = 17.38@xmath41jy per pixel .
this is a combination of the conversion given in the _ spitzer _ observers manual ( 1mjy sr@xmath2 = 23.50443 @xmath41jy arcsec@xmath42 ) , and the pixel solid angle ( 0.74 arcsec@xmath43 ) . a mips ( multi - band imaging photometer )
24@xmath41 m calibrated image was also retrieved from the archive ( from the same _ spitzer _ program ) , which was reduced using standard pipeline processing . a stellar - continuum
subtracted 8.0@xmath41 m ( pah ) image was created to highlight the star - forming regions , using a conversion based on a standard model of the starlight near - ir sed ; this is 8.0@xmath41 m @xmath31 [email protected]@xmath41 m @xcite .
photometry was performed on the nuclear source in the four irac bands using a 15 radius aperture centred on the _ chandra _ position of the nucleus , using a local annular background between 2.53.0 .
this aperture was chosen because of the larger psf of irac ( 1.41.7 fwhm , @xcite ) , and aperture corrections were applied as detailed in the irac data handbook v3.0 , section 5.5.1 . the resulting flux densities ( and ab magnitudes ) are shown in table [ tab : acs ] .
note that photometry is not performed on the mips data , as the spatial resolution ( @xmath66 ) is insufficient to separate the nuclear source from the circumnuclear ring emission .
radio continuum observations of ngc 1672 were obtained at 3-cm and 6-cm with the australia telescope compact array ( atca ) .
data reduction was carried out with the miriad software package , using standard procedures .
the 3-cm image was created using archival data from a 1.5c configuration obtained in may 2003 and a 6d configuration obtained in july 2003 , with on - source integration times of 9h and 10h respectively .
both observations consist of two frequency bands centered at 8640mhz and 8768mhz , with bandwidths of 128mhz divided into 32 channels . at 8700mhz
the atca primary beam is 5.5 .
the data for both observations / frequency bands were combined using ` robust ' weighting ( @xmath44 ) , resulting in an angular resolution of [email protected] and an rms of 0.03mjy / beam .
the 6-cm image was created using archival data in the 6a configuration obtained in may 2002 , with an on - source integration time of 11h .
the observation consists of two frequency bands centred at 4800mhz and 4928mhz , with bandwidths of 128mhz divided into 32 channels . at 4864mhz
the atca primary beam is 9.8 .
the data from both frequency bands were combined , resulting in an angular resolution of [email protected] and an rms of 0.04 mjy / beam .
the source pks b1934638 was used as the primary ( amplitude & bandpass ) calibrator , and pks b0522611 and pks b0420625 were used as the secondary ( phase ) calibrators at 3- and 6-cm , respectively . a point source fit to the nuclear source
gives a position of @xmath45(j2000 ) = 04@[email protected]@xmath30 , @xmath31591449.85 ( @xmath460.12 ) .
we measure peak fluxes of @xmath47mjy at 3-cm and @xmath48mjy at 6-cm .
we measure integrated flux densities of 34mjy ( 3-cm ) and 35mjy ( 6-cm ) for the radio continuum emission from both the nuclear source and surrounding star - forming ring .
it is likely that some extended flux has been resolved out by the interferometer .
single - dish radio continuum flux estimates of ngc 1672 ( pmn j04455914 , pks j04455915 ) at @xmath65ghz ( 6-cm ) are typically around 100mjy ( see @xcite ; @xcite ) .
there is also one other radio source in the field of view at @xmath45(j2000 ) = 04@[email protected]@xmath30 , @xmath31591428.00 ( @xmath460.03 ) , with flux densities of @xmath49mjy ( 3-cm ) and @xmath50mjy ( 6-cm ) .
this is positionally coincident with x - ray source # 8 ( 044536.73591428.0 ) .
figure [ fig : pretty](a c ) shows 3-color _ chandra _ , _ xmm - newton _ and _ spitzer _ irac images of the entire @xmath0 area of ngc 1672 .
the x - ray sources outside the nuclear region are well resolved in both the _ chandra _ and _ xmm - newton _ data , and are discussed in [ sec : results_sources ] .
there is also a diffuse hot x - ray gas component , which is shown in figure [ fig : pretty](d ) as extended red / green x - ray emission ( see [ sec : results_diffuse ] ) .
the nuclear region is very bright at all wavelengths , and will be discussed in detail in [ sec : results_morph ] .
the _ spitzer _
ir data show that the majority of the non - nuclear star - formation activity is confined to the bar and inner spiral arms ( illustrated in red and green ) , whereas older stellar populations are spread throughout the disk ( illustrated in blue ) .
figure [ fig : hst ] shows the high - resolution 3-color _ hst_/acs image , where the bar structure , numerous hii regions and dust lanes are visible .
figure [ fig : source_positions ] ( _ top left _ ) shows the dss optical image of ngc 1672 , overlaid with the positions of the 28 x - ray sources detected within the @xmath0 area of the galaxy ( _ blue crosses _ ) .
note that we do not plot the source positions on the _ hst _ image here since it does not cover the entire @xmath0 area .
the positions correlate well with the optical structure of the galaxy , with sources concentrated in the nuclear region ( see [ sec : results_morph ] ) and spread throughout the disk and spiral arms . the extra - nuclear source fluxes range between @xmath51 erg s@xmath2 @xmath22 over both _
xmm - newton _ and _ chandra _ observations , corresponding to @xmath52 erg s@xmath2 at the distance of ngc 1672 .
nine sources , almost a third of the population , reach ultraluminous x - ray source ( ulx ) luminosities ( @xmath53 erg s@xmath2 , 0.310kev ) in either one or both x - ray observations ; these are marked with red circles ( sources 1 , 4 , 5 , 6 , 7 , 13 , 24 , 25 and 28 ) .
three of the ulxs are very bright ( sources 4 , 5 , and 25 with @xmath1 erg s@xmath2 , marked with larger circles ) , and are co - spatial with the ends of the bar ( as previously noted by @xcite ) .
this is consistent with models that show that gas can accumulate at the tips of bars due to the corotation of the bar structure with the disk @xcite , thereby creating star - forming regions where x - ray binaries can form .
other examples of this phenomenon have been observed in the barred galaxies ngc 7771 @xcite and ngc 4303 @xcite .
figure [ fig : source_positions ] ( _ top right _ ) shows the source positions overlaid on the _ spitzer _
8@xmath41 m image , illustrating the coincidence of many x - ray sources with regions of star - formation in the bar and inner spiral arm regions .
we have correlated the x - ray source positions with the hii region catalog of @xcite .
taking into account both the absolute astrometric accuracy of the x - ray and hii region positions ( @xmath61 combined ) as well as the measured areas of the hii regions , we associate eleven x - ray sources with hii regions ( 2 , 4 , 5 , 7 , 13 , 14 , 18 , 21 , 25 , 27 , and 28 ) .
these include the three brightest ulxs , and three sources in the circumnuclear ring .
we have not performed any complex spectral fitting on the x - ray sources , as even the brightest are well - fit with simple absorption+powerlaw models , with @xmath54 close to 1 ( see table [ tab : catalog ] ) .
the spectra and model fits are shown in figures [ fig : xmm_spec_sources ] and [ fig : chan_spec_sources ] . however , we note that even though this simple model provides acceptable empirical fits to these data , this is likely due to the relatively low photon counts ( @xmath55 2000 counts , total pn+mos or acis - s ) , and the true underlying spectral shapes are likely to be more complex ( see @xcite for a discussion ) .
figure [ fig : source_positions ] ( _ top left _ ) also denotes variable sources ( marked with white squares ) .
six sources show significant long - term flux variability between the _ xmm - newton _ and _ chandra _ observations at levels of @xmath33 ; these are sources 1 , 4 , 6 , 8 , 21 and 25 .
most notable are the increases in flux in two of the brightest ulxs in the _ chandra _ observations ; # 4 increases by a factor of @xmath65 ( up to @xmath56 erg s@xmath2 ) , and # 25 increases by a factor of @xmath63 ( up to @xmath57 erg s@xmath2 ) , indicating that the high luminosity of these sources is not due to confusion of several fainter sources .
these flux increases are accompanied by a spectral hardening of source # 4 , but no significant spectral change in source # 25 .
the other four sources with spectral fits in both observations show no significant spectral change between observations .
another four sources ( 10 , 17 , 24 and 27 ) are only detected in one observation . to investigate longer - term behavior , in table [ tab : rosat ]
we show a comparison between _ xmm - newton _ , _ chandra _ and _ rosat _ hri soft x - ray luminosities for five sources detected in all three observations .
we quote the 0.22kev hri luminosities from @xcite for the combined 1992/1997 dataset ( which gives better photon statistics ) , together with 0.22kev luminosities derived from the best - fit _ xmm - newton _ and _ chandra _ spectral models corrected for galactic absorption .
note that we also include # 23 ( _ rosat _ source x-5 ) , which was originally believed to be coincident with a foreground star given the _ rosat _ positional errors ( @xmath65 ) .
however , with the _ chandra _ positional accuracy of @xmath60.5 , we show that this source is not consistent with the 2mass position of the star ( 04@[email protected]@xmath30 , @xmath31591254.1 ) with an offset of @xmath65 .
in fact , this is the source that shows the largest long - term variability with @xmath58 over this @xmath610 year period .
the remaining sources vary in luminosity by factors of @xmath59 .
we estimate the number of background x - ray sources in the @xmath0 area of ngc 1672 using the hard - band ( 210kev ) log @xmath60-log @xmath61 source distributions from _ chandra _ deep field studies ( we do not use the soft - band log @xmath60-log @xmath61 since the detection sensitivity at soft energies will be affected by absorption by neutral material in the galactic disk ) .
using a source detection threshold of 12 counts in the 0.310kev band ( corresponding to the faintest source detected in the _ chandra _ observation ) , and converting this to a 210kev flux for a typical agn ( @xmath62 erg s@xmath2 @xmath22 , using @xmath37 and @xmath38 @xmath22 ) , the log @xmath60-log @xmath61 relations of @xcite , @xcite and @xcite all predict between 23 background sources in the field . note that this is in fact an upper limit , since the ngc 1672 source catalog is not corrected for incompleteness .
inspection of the _
hst_/acs images allows us to identify two x - ray sources as background galaxies immediately ; these are # 26 , which appears to be a face - on disk galaxy with a bright nucleus , and # 8 , which looks like an edge - on galaxy with visible dust lanes seen _ through _ the disk of ngc 1672 ( see figure [ fig : source_positions ] , lower panel ) . as noted in
[ sec : data_radio ] , source # 8 is the only extra - nuclear x - ray source with a detected radio counterpart . at the distance of ngc 1672 , the 210kev x - ray luminosity ( at its brightest in the _ xmm - newton _ observation ) would be @xmath63 erg s@xmath2 , and the 5ghz radio luminosity would be @xmath64 erg s@xmath2 .
according to the x - ray / radio fundamental plane of @xcite , this is inconsistent with emission from a stellar mass black hole ( i.e. too radio - bright ) , and supports the hypothesis that this source is a background agn ( see [ sec : disc ] for further discussion of this relation with regards to the nuclear source ) .
we have used the _
chandra _ observation to study the diffuse x - ray component of ngc 1672 due to its @xmath61 spatial resolution , and hence ease of source subtraction .
the morphology of the diffuse emission is illustrated in figure [ fig : pretty ] ( d ) , where the x - ray point sources and central ring region have been masked , and the residual emission adaptively smoothed .
overlaid are contours from the continuum - subtracted _ spitzer _ irac 8.0@xmath41 m image , which highlights star - formation activity ; the diffuse emission is spatially correlated with the ir emission , following the bar / inner spiral arm structure .
a source spectrum was extracted from an elliptical region ( @xmath65 centered on the nucleus with a position angle of @xmath66 .
the detected point sources were excluded ( using regions with twice the radius of the 90% encircled energy ellipses ) , as well as the innermost 20 radius circumnuclear ring region .
the background regions were chosen to be ( 1 ) the region on the acis - s3 chip outside the @xmath0 radius and not containing point sources , and ( 2 ) the region on the acis - s1 chip not containing bright point sources . to further constrain the soft end of the background emission
, we also used the seven element _ rosat _ all sky survey ( rass ) spectral energy distribution @xcite from an annulus @xmath67 around ngc 1672 , extracted using the x - ray background web tool hosted by the heasarc .
for each _ chandra _ spectrum , we created a particle background spectrum using the `` stowed '' data from the appropriate chip and mode , scaled to produce the same 1013kev band count rate as observed in the obsid 5932 data .
the particle background spectra were then used as the `` background '' spectra within xspec , and the spectra of the galaxy and the remainder of the background emission were fitted simultaneously .
after subtraction of the particle background , there were 1580 counts in the source spectrum and 2890 counts in the background spectra .
the spectra and fits are shown in figure [ fig : spec_fit ] , and the fit parameters for both background and galaxy emission are shown in table [ tab : spec_fit ] .
the emission in the background regions consists of galactic foreground emission from the local hot bubble ( lhb ) and the galactic halo ( modeled with apec components ) , the extragalactic background due to unresolved agn ( modeled with a powerlaw ) , as well as contributions from soft protons ( sp ) .
the galactic halo and the extragalactic background are absorbed by the total neutral column along the line - of - sight ( @xmath68 @xmath22 ) .
the abundances were fixed to the @xcite solar values .
the photon index of the powerlaw was set to 1.46 @xcite , while the normalization of each spectrum was initially set to the value predicted by the point source detection limit , a total background value of 11.6 photons @xmath22
s@xmath2 sr@xmath2 kev@xmath2 at 1kev @xcite , and the 0.52.0kev luminosity function of @xcite .
the normalizations of the powerlaws were then allowed to vary , and the fitted values were not too different from the initial values .
a sp component was also included in the fit , which can be represented by a flat ( @xmath69 ) powerlaw with an exponential cutoff at 500kev @xcite .
the galaxy emission was fit with mekal thermal plasma models , split into components representing the unabsorbed emission coming from the front of the galactic disk , and absorbed emission coming from behind the disk , with their temperature parameters tied .
adequate fits could not be obtained from a single temperature model .
we found that a two - component model provided a very good fit , but can not rule out the existence of a distribution of temperatures or non - equilibrium plasmas ( see the discussion in @xcite ) .
the temperatures ( @xmath70=0.22/0.77kev ) are typical of those found in galactic disks , with a ratio of emission measures typical of late - type galaxies .
the abundance of the galaxy was assumed to be solar @xcite , and due to the uncertainty introduced by the intermixture of absorption and emission components , the disk absorption column density was initially set to zero and allowed to vary freely . in table [ tab : spec_fit ] we show fitted values for two different cases , the first assumes no absorbing galactic disk , and the second allows the absorbing column within the galaxy to vary .
the final fitted value for the disk absorption was @xmath71 @xmath22 , rather higher than a typical galactic disk , and which implies a near complete absorption of the flux from the far side of the disk .
however , since the spectrum was extracted from the bar region , this value may not be unreasonable .
the galactic absorption does not significantly change the fitted temperatures , and the normalizations vary by roughly a factor of two , as one might expect given the strong fitted absorption .
the total 0.52kev luminosity of the diffuse components is @xmath72 erg s@xmath2 for the model with no intrinsic absorption , and @xmath57 erg s@xmath2 for the model with high disk absorption ( see table [ tab : spec_fit ] ) . in order to place a spectroscopic limit on the contribution to this emission from unresolved point sources
, we introduced an absorbed power law component to the model with no intrinsic absorption , fixing the photon index to 1.7 , but allowing the absorption and normalization to vary .
the best fit allowed allowed a luminosity of @xmath73 erg s@xmath2 ( 0.52kev ) to be due to unresolved sources , with a 90% upper limit of @xmath74 erg s@xmath2 .
the new high - resolution data allow us to probe the nuclear structure of ngc 1672 on sub - kpc scales .
figure [ fig : ctr ] shows images of the central region of the galaxy in the x - ray , optical and 3-cm radio bands . for the first time , the _ chandra _ images ( figure [ fig : ctr]a ) spatially resolve the x - ray emission into a weak point source ( source # 15 ) surrounded by the circumnuclear ring ( white polygons highlight the detected point sources in the region ) .
the _ hst_/acs data ( figure [ fig : ctr]b ) also show that ngc 1672 possesses a nuclear spiral , which is one of the candidate fueling mechanisms for agn @xcite .
the radio data clearly show both the ring and a compact nuclear source ( figure [ fig : ctr]c ) , and in the _ spitzer _ irac image ( figure [ fig : ctr]d ) strong star - forming activity is seen in the ring ( 8.0@xmath41 m , green ) , while the region inside the ring is dominated by older stars ( 3.6@xmath41 m , blue ) .
the nucleus itself shows evidence of a combination of the two ( note there is still some pah emission in the nucleus itself ) . in figure
[ fig : ctr](e ) , radio contours are overlaid on the _ chandra _ image , which illustrates a strong spatial correlation between both the ring and position of the nuclear source .
the x - ray position of the nucleus agrees to within 0.25with the radio position .
chandra _ x - ray contours are overlaid on the h@xmath14 image in figure [ fig : ctr](f ) , and show a correlation between the location of the circumnuclear ring and the outer edge of the nuclear spiral .
the central x - ray source is also spatially co - incident with the peak of the h@xmath14 emission .
the nucleus is only detected in the medium band ( 12kev , 18 net counts ) and hard band ( 210kev , 17 net counts ) in the _ chandra _ observation , indicating that there is some obscuring material along the line - of - sight to the nucleus . to estimate the spectral shape ( and hence source flux ) of the nucleus
, an x - ray hardness ratio was calculated using the behr algorithm @xcite .
this is a bayesian method that takes into account the poissonian nature of low - count data , and deals with situations where no net source counts are detected in one or more bands ( as is the case here in the soft band ) .
the hardness ratio is plotted on a model grid in figure [ fig : hr ] , and indicates that the source is hard ( @xmath75 ) with fairly low absorption ( @xmath76 @xmath22 ) .
this translates into an unabsorbed 210kev luminosity of only @xmath77 erg s@xmath2 .
if we use the upper limit to the absorption in the flux calculation ( @xmath78 @xmath22 ) , the luminosity is only marginally increased to @xmath79 erg s@xmath2 .
this is very faint for a seyfert galaxy , but does just fall into the range found in the low - luminosity agn survey of @xcite ( @xmath80 erg s@xmath2 ) .
however , we have adopted a very simple model here ; more complex models may admit a significantly higher absorption column density ( see [ sec : cthick ] for further discussion ) .
figure [ fig : spec_ctr ] ( _ top , middle _ ) shows _ xmm - newton _ spectra of the central 22 radius region of the galaxy , which includes all contributions from the nucleus and circumnuclear ring .
the spectra contain @xmath610,500 background - subtracted counts in total ; @xmath66700 counts in the pn data , @xmath61800 counts in the mos1 data , and @xmath62000 counts in the mos2 data .
the data were fit simultaneously with two- and three - component spectral models , with combinations of a power - law , to model the non - thermal contributions from accreting sources ( xrbs / agn ) , and mekal thermal plasma components to model the hot gas ( see table [ tab : centre_fits ] ) . the simplest model ( mekal+po with equal absorbing columns , model 1 ) yields a reasonable fit ( @xmath811.3 ) , with a gas temperature of 0.60kev and a soft power - law component ( @xmath82=2.27 ) . to improve this , we fitted the data with three - component models .
we tried a model where all absorption columns were tied ( model 2 ) , and one where each component had it s own absorption ( model 3 ) .
of the two , model 3 provides a slightly superior fit ( @xmath811.16 ) , with thermal temperatures of @xmath70=0.24/0.60kev ) and a soft power - law ( @xmath83 ) .
the thermal components in this model show increasing absorption with increasing temperature , a pattern found previously in some starburst galaxies ( e.g. @xcite ; @xcite ) .
however , we have frozen the value for the powerlaw absorption to 0.5@xmath2510@xmath84 @xmath22 ( @xmath85 @xmath22 including galactic absorption ) , the value we get for model 2 where there is only one absorption column for the whole model ; if the absorption is left free in model 3 it tends to zero , which is not realistic if the point sources are related to the same stellar population as the thermal components .
the addition of the cooler mekal component produces a substantial improvement to the fit , with a @xmath86 of 43 for three additional degrees of freedom .
note that the temperatures of the thermal components are consistent with those measured in the diffuse emission analysis ( [ sec : results_diffuse ] ) , within the combined uncertainties .
the 210kev unabsorbed luminosity of the best - fit model is @xmath87 erg s@xmath2 .
note that there is no evidence for a bright agn - like component that would be absorbed at low energies and only transmitted at higher energies .
there is also no indication of a neutral fe-6.4kev emission line , typically seen in bright type-2 agn ( although such a line , if present , would likely be diluted by the emission from the starburst ring ) . inserting a narrow gaussian line at 6.4kev into the best - fit model , with all other model components frozen , results in a 90% upper limit on the equivalent width of @xmath6720ev . to search for spectral variability between the _ xmm - newton _ and _ chandra _ observations
, we extracted an acis - s spectrum from the same nuclear region ( figure [ fig : spec_ctr ] , _ bottom _ ) .
this spectrum has significantly fewer counts than the _ xmm - newton _ data ( 3070 ) , therefore we froze the mekal temperatures and @xmath88 values to those from the _ xmm - newton _ best fit model , and only fit the power - law component .
the slope is almost identical to the _ xmm - newton _ observation with @xmath89 , and the model has a similar unabsorbed 210kev luminosity of @xmath90 erg s@xmath2 , which is consistent with the sum of the hard emission from the nucleus and point sources in the circumnuclear ring .
the 0.22kev luminosity of the central source in the _ rosat _ hri observations ( x-1 ) remained almost constant at @xmath91 erg s@xmath2 between the 1992 and 1997 observations @xcite .
the equivalent 0.22kev luminosity in both the _ xmm - newton _ and _ chandra _ data is 8.2@xmath92 erg s@xmath2 ( corrected for galactic absorption ) ; there is therefore no evidence of significant variability ( in the soft band ) over a @xmath614 year period .
the nuclear x - ray source in ngc 1672 is observed to have a hard x - ray spectrum , with detections in the medium ( 12kev ) and hard ( 210kev ) bands only .
however , the spectral slope is not well constrained with the hardness ratio ( @xmath93 ) due to the low number of counts .
this , in conjunction with its low observed 210kev luminosity ( @xmath94 erg s@xmath2 ) , means that the underlying nature of the emission is not obvious .
given its location , it could be an accreting llagn , but there are several other possibilities .
for example , the emission may be produced by one , or a few , xrbs associated with star formation in the nuclear region . or it could be a composite object with some of the emission coming from an llagn and some from star - formation .
it may also be produced by compact supernovae ( sn)/supernova remnants ( snrs ) , also directly related to any star formation in the region .
we investigate these possibilities as follows .
@xcite have measured a stellar velocity dispersion of @xmath6110 km s@xmath2 from the calcium triplet ( @xmath958498 , 8542 , 8662 ) lines in the inner 300pc radius of ngc 1672 .
according to recent simulations and observational results ( @xcite ; @xcite ) , the @xmath96-@xmath97 relation for super massive black holes in galaxies with pseudobulges ( such as ngc 1672 , @xcite ) has a similar slope but _ lower _ normalization than that for galaxies with classical early - type bulges ( see figure 4 in @xcite ) .
this new relation predicts a central black hole mass of only @xmath98m@xmath99 for ngc 1672 .
if the central x - ray source is a llagn , then its luminosity implies a very low accretion rate , with @xmath100 ( or @xmath101 if a bolometric correction for llagn of @xmath102 for the 210kev band is assumed ; @xcite ) .
however , strong evidence of low - level agn activity comes from the radio imaging data , which shows that ngc 1672 has a compact 5ghz radio core coincident with both the x - ray and optical nucleus ( figure [ fig : ctr]c ) .
bright radio cores are frequently found in llagn ( e.g. @xcite ; @xcite ) , and are believed to be the signature of radio jets produced by accreting super - massive black holes with radiatively inefficient accretion flows .
the flux densities of the radio source are 0.95mjy ( @xmath103 erg s@xmath2 ) at 10ghz ( 3-cm ) , and 1.62mjy ( @xmath104 erg s@xmath2 ) at 5ghz ( 6-cm ) ; these translate into a steep spectral slope of @xmath105 ( where @xmath106 ) . both bright and
low - luminosity agn tend to have flat radio slopes ( @xmath107 ) , thought to result from self - absorption of synchrotron radiation from the base of a jet @xcite , although flat slopes can also be produced by thermal emission from optically - thin ionized gas in nuclear starbursts @xcite
. however , steeper slopes ( @xmath108 ) can be explained by optically - thin synchrotron emission from jets on larger scales @xcite , which is consistent with our result .
snrs are also expected to produce steep radio slopes ( @xmath109 ) , and we investigate this possibility below .
in contrast , radio emission from an advection - dominated accretion flow ( adaf ; @xcite ) is predicted to have an inverted radio spectrum ( @xmath110 , @xcite ) , which is not consistent with our result . to investigate whether the radio / x - ray properties of this source are consistent with a super - massive , or stellar - mass black hole ( i.e. a normal xrb ) , we can use the fundamental - plane relation of @xcite .
this relation shows that the 5ghz radio luminosity , 210kev x - ray luminosity and black hole mass are strongly correlated for both galactic ( stellar - mass ) black holes and central super - massive black holes for a range of accretion rates , which is believed to result from an intimate connection between accretion flow and jet activity .
the radio luminosity of the nuclear source in ngc 1672 is fully consistent with that expected from a jet from a @xmath98m@xmath99 super - massive black hole ( in combination with @xmath111 erg s@xmath2 ) , and approximately 5 orders of magnitude too luminous for a jet from a typical @xmath112m@xmath99 stellar - mass black hole ( see their figure 5 ) .
this effectively rules out the normal xrb scenario .
( note that a similar result is found for the one off - nuclear x - ray source with a radio counterpart , see
[ sec : results_sources ] ) .
we can also estimate the probability of a chance coincidence of an xrb with the nucleus , using the ` universal ' luminosity functions of high - mass xrbs ( hmxbs ; @xcite ) and low - mass xrbs ( lmxbs ; @xcite ) , which depend on star formation rate and stellar mass , respectively . to calculate the sfr for the nuclear region only ( excluding contributions from the ring )
, we use a spatially resolved , h@xmath14 line flux measured in the central 2 of the galaxy ( 9.2@xmath2510@xmath113 erg s@xmath2 ; @xcite ) together with the h@xmath14sfr calibration of @xcite ( hereafter k98 ) .
a color excess of @xmath114 was derived from the balmer decrement in the nuclear region , and the h@xmath14 flux is corrected for this .
note that we do not use the narrow - band acs h@xmath14 data for this purpose , since this filter ( f658n ) is known to be contaminated by the [ nii ] @xmath956583 and 6548 lines @xcite . using the k98 sfr calibration : @xmath115
the h@xmath14 luminosity is equivalent to a nuclear sfr of 0.007m@xmath99 yr@xmath2 ( if star formation is the source of the emission ) .
according to the relation of @xcite , this predicts only @xmath60.01 hmxbs with luminosities greater than @xmath4 erg s@xmath2 . to estimate the stellar mass in the nuclear region
, we use the _ spitzer _ irac 3.6@xmath41 m measurement as a proxy for the k - band luminosity , and convert this to stellar mass using the k - band mass - to - light ratio of @xcite ( @xmath116 ) , resulting in @xmath117 .
the universal lmxb luminosity function @xcite predicts only @xmath60.008 lmxbs above @xmath4 erg s@xmath2 for this stellar mass .
the other possibility is that the x - ray and radio core emission arises from sn / snrs .
although soft thermal x - ray emission is the characteristic signature of sn / snrs ( resulting from shock fronts interacting with circumstellar and ism material , e.g. @xcite ) , hard non - thermal emission is detected in some sn , possibly emitted from a young pulsar left behind in a core - collapse event .
an example of this is sn 1968d @xcite , where the 210kev component has @xmath118 erg s@xmath2 .
however , the soft component dominates the 0.310kev luminosity , with the hard emission only contributing @xmath610% to the total .
this is not the case here ; after careful background subtraction using the area inside the ring , there are no soft 0.31kev counts .
while this is likely due to absorption , the hard 210kev band luminosity dominates the detected emission . we can also use the radio data to investigate the sn / snr option .
have previously detected compact radio emission at 2.3ghz in the core of ngc 1672 , and speculated , by calculating the theoretical radio snr luminosity and comparing it to the observed luminosity , that it is possible that this can be produced by clumps of luminous radio sn , similar to those found in the nuclei of arp 220 @xcite and m82 @xcite .
however , they used a sfr ( derived from far - ir data ) for the _ whole _ galaxy rather than the nuclear region , and therefore this was not conclusive evidence for or against the presence of an agn .
what is required is a similar test for the nuclear region only .
following the recipe of @xcite , we take the sfr and calculate the expected supernova rate , @xmath119 , integrating the sfr over a salpeter initial mass function ( imf ) with lower and upper mass limits of 0.1@xmath11 and 100@xmath11 , and a minimum initial mass for supernova detonation ,
@xmath120@xmath11 : @xmath121 where @xmath122 is the sfr .
this results in @xmath123 yr@xmath2 .
we can then calculate the non - thermal radio luminosity expected for this sn rate for both young supernovae ( @xmath124 , with radio lifetime @xmath125yr ) , and old supernova remnants ( @xmath126 , with @xmath127yr ) , using the following equations derived by @xcite from observations of sn1979c : @xmath128 @xmath129 for the nuclear source in ngc 1672 , these predict @xmath130 erg s@xmath2 and @xmath131 erg s@xmath2 at 5ghz .
the total of the two is a factor of 34 lower than the measured luminosity of the compact central source at this frequency ( @xmath104 erg s@xmath2 ) , which indicates that snrs are not the primary source of the radio emission .
however , we must consider the uncertainties associated with the sfr conversion and extinction correction used here .
@xcite have made a detailed investigation of the systematic uncertainties associated with the balmer decrement correction , and find it to be accurate to 15% on average .
indeed , apart from variations in the imf , the amount of extinction present is the dominant source of uncertainty in the h@xmath14sfr conversion @xcite .
once the correction has been applied , additional uncertainties arise from variations in metallicity and stellar ages , but only at the @xmath620% level @xcite . taking these factors into consideration ,
we conclude that snrs can not easily account for the high radio output in the core of ngc 1672 .
figure [ fig : sed ] shows the sed of the nuclear source , constructed using the spatially - resolved measurements at radio , near - ir , optical , uv and x - ray wavelengths .
the x - ray spectral slope , plus upper / lower limits estimated from the hardness ratio , are shown in @xmath132 units as a ` bow - tie ' . for comparison
, we also plot the average llagn sed of @xcite , as well as the average seds of radio - quiet and radio - loud quasars from @xcite , which are all normalized to the 2kev nuclear x - ray luminosity of ngc 1672 .
these measurements allow us to derive some useful broad spectral properties of the source .
one such diagnostic is its radio ` loudness ' .
@xcite define a ratio of radio - to - x - ray luminosity , @xmath133 ( 5ghz)/@xmath134 , which is more sensitive to highly absorbed sources than the traditional radio - to-@xmath135-band luminosity ratio . a 5ghz radio luminosity of @xmath104 erg s@xmath2 and a 210kev x - ray luminosity of @xmath111 erg s@xmath2 gives log @xmath136 , which classifies the nucleus of ngc 1672 as radio _ loud _ ( see their figure 5 ) .
this is consistent with other llagn and the hypothesis that the majority of llagn possess relativistic jets ( e.g. @xcite ) .
given the scaling of the comparison seds , the radio luminosity of ngc 1672 is consistent with that of both the average llagn and radio - loud quasars .
however , there is a clear excess of emission at near - ir and optical wavelengths with respect to the average seds , which indicates that the photometry is contaminated by star - formation in the region .
indeed , the optical _ hst _ measurements are considered to be upper limits on the luminosity of a nuclear point source ( [ sec : data_hst ] ) , and we denote them , and the near - ir measurements , as such with arrows .
this excess indicates that the nuclear emission is a combination of both a low - accretion - rate central black hole _ and
_ star formation in the central @xmath6200pc around the black hole ; this agrees with the study of @xcite , who detected both narrow and slightly broadened weak optical emission lines in the central @xmath18 nuclear region ( see [ sec:1672 ] ) and concluded that they likely represent a composite hii / seyfert nucleus .
we can also test whether the h@xmath14 luminosity is consistent with the @xmath137 correlation of @xcite for agn .
the underlying physical reason for this relation , which holds from bright quasars all the way down to the llagn regime , is that the optical emission lines are powered by photoionization from a hard agn continuum , which naturally scales with the x - ray luminosity .
type-1 objects follow the relation strongly , while type-2 seyferts / llagn tend to be more under - luminous in x - rays ( see figure 2 in @xcite ) .
this suggests that either the h@xmath14 luminosity may also have a contribution from other processes ( e.g. , star formation ) in the vicinity of the central black hole in these objects , or that the agn is highly absorbed and therefore only _ appears _ under - luminous in the x - ray band @xcite .
the nuclear source of ngc 1672 , with @xmath138 erg s@xmath2 and @xmath13910@xmath113 erg s@xmath2 , is under - luminous in x - rays , consistent with other type-2 agn in the @xcite survey .
@xcite show a similar figure for a sample of liners observed with _
chandra _ , and also include luminosity ranges where a pure starburst would fall ( see their figure 7 ) .
the nuclear source in ngc 1672 falls in the region below the @xcite agn relation and above that expected for starbursts , again strongly suggesting that this is a type-2 agn .
we have also calculated _
irac colors to compare with the color - color diagram of @xcite , which separates pah - dominated ( star - forming ) , stellar - dominated ( passive ) and continuum - dominated ( agn ) sources .
the colors of the nuclear source of ngc 1672 ( log@xmath140=-0.27\pm 0.03 $ ] and log@xmath141=0.10\pm 0.01 $ ] ) fall at the intersection of all three source types , indicating again that there is some star - formation activity in the nuclear region ( which agrees with the imaging , see figure [ fig : ctr](d ) , where the pah emission is clearly visible )
. since there is evidence for a composite nucleus with some absorption , we can investigate whether it is compton - thick ( @xmath142 @xmath22 ) .
observationally , the nuclear source is faint and hard at x - ray wavelengths , with an upper limit on the absorption of a modest value of @xmath78 @xmath22 ( as derived from the hardness ratio ) .
however in compton - thick sources , the direct x - ray emission is completely suppressed below @xmath610kev ; the observed 210kev component represents x - rays reflected or scattered by surrounding material , which tends to produce a hard x - ray spectrum ( @xcite ; @xcite ) .
therefore , the observed luminosity and spectral slope will not be representative of the intrinsic properties of the source ( e.g. the archetypal seyfert 2 galaxy ngc 1068 , @xcite ; @xcite ) , and may lead to mis - classification as a moderately absorbed llagn .
an x - ray signature of a compton - thick agn is a strong fe k@xmath14 line at 6.4kev ( equivalent width typically @xmath7 1kev ) , due it being measured against a depressed x - ray continuum @xcite .
however , in the case of ngc 1672 we can not place any meaningful limits on the presence of such a line due to the low number of counts in the _ chandra _ detection .
the 720ev equivalent width upper limit from the _ xmm - newton _ spectral fits ( see [ sec : results_spectra ] ) does not provide a good constraint due to dilution by sources in the circumnuclear ring .
other multiwavelength measurements can also provide evidence for the presence or absence of strong absorption .
for example , an indicator that a source may be highly absorbed is the ratio of 210kev x - ray flux to the [ o iii ] @xmath1435007 line flux .
the [ o iii ] line is believed to be a measure of the intrinsic brightness of the source since it is produced by the photoionizing agn continuum , but in the narrow line region ( nlr ) far from the nucleus ( see @xcite and references therein ) .
@xcite measured the [ o iii ] flux from the central 2 of ngc 1672 as @xmath144 erg s@xmath2 @xmath22 .
together with the observed 210kev x - ray flux of @xmath145 erg s@xmath2 @xmath22 , this gives @xmath146 } = 0.99^{+0.47}_{-0.34}$ ] .
unfortunately , this is inconclusive , as it falls on the borderline of the compton - thick category ( typically @xmath36 1 , @xcite ; @xcite ) , and some confirmed compton - thick galaxies have ratios within our error range ( e.g. ngc 3079 ; @xcite ) .
however , it indicates that there is some absorption in the nucleus , which is consistent with type-2 classification .
how do the results for ngc 1672 fit into the broader picture of agn evolution ?
do more barred galaxies contain agn than non - barred galaxies because of the extra fuel provided , and what are their x - ray luminosities ? what effect does the presence of a circumnuclear ring have ?
a preliminary inspection of published results for nearby spiral galaxies observed with _
chandra _ to date reveals a large range of observed nuclear x - ray properties , regardless of morphological type .
for instance , while the strongly - barred galaxies ngc 1365 , ngc 1386 , ngc 3079 and ngc 3393 are classified as compton - thick , with absorption - corrected 210kev x - ray luminosities of @xmath147 erg s@xmath2 ( @xcite ; @xcite ; @xcite ) , other strongly - barred systems such as ngc 660 , ngc 1023 and ngc 2787 ( @xcite ; @xcite ) have observed nuclear luminosities of only @xmath148 erg s@xmath2 .
a similar range of x - ray luminosities is found for unbarred galaxies , such as m31 and m51 ( @xmath149 erg s@xmath2 , @xmath150 erg s@xmath2 ; @xcite ) , m33 ( @xmath151 erg s@xmath2 ; @xcite ) and ngc 7212 ( @xmath152 erg s@xmath2 ; @xcite ) .
the range of nuclear luminosities for galaxies with circumnuclear rings , which are predominantly found in barred galaxies ( e.g. @xcite ) , is also broad .
examples are ngc 4736 ( @xmath151 erg s@xmath2 ; @xcite ) , ngc 4303 ( @xmath153 erg s@xmath2 ; @xcite ) , ngc 1097 ( @xmath154 erg s@xmath2 ; @xcite ) , ngc 1365 ( @xmath155 erg s@xmath2 ; @xcite ) , and even non - detections in some cases such as ngc 3351 ( @xmath156 erg s@xmath2 ; @xcite ) and ngc 4314 ( @xmath157 erg s@xmath2 ; @xcite ) .
this indicates that the presence of a ring does not affect the x - ray luminosity of the nucleus in an obvious or straight - forward way .
for example , they do not all possess low - luminosity nuclei , which would be expected if a ring prevents the fuel from reaching the very center of the galaxy , a scenario suggested by @xcite .
it is also possible that galaxies with non- or low - luminosity detections have agn that are highly absorbed , or even compton - thick , which is more likely in barred systems due to the increase in gas / star formation activity in the central regions . in a recent survey of 187 nearby galaxies observed with _
chandra _ , @xcite found no correlation between the presence of a strong bar and the frequency or x - ray luminosity of nuclear sources .
indeed , they found a possible anti - correlation , in that strongly - barred galaxies tend to have less - active nuclei than unbarred or weakly - barred galaxies .
however , this study did not take into account the presence / absence of circumnuclear rings , or other smaller - scale features such as nuclear bars .
additionally , hydrodynamic mechanisms such as turbulence in the ism close to the nucleus ( e.g. nuclear spirals ) may play a significant role in agn fueling @xcite .
clearly , an extensive multiwavelength investigation is required to untangle the possible effects of large- and small - scale morphological features , as well as significant absorption , on the x - ray properties of agn . that is beyond the scope of this paper , but will be addressed in a future study .
this paper is part of a continuing study in which we are aiming to ascertain how both the large- and small - scale morphology of a galaxy may affect its nuclear activity .
we are studying this primarily from an x - ray perspective , as x - ray observations provide an ideal tool for searching for low - level nuclear activity that may not be visible at other wavelengths . here , we have performed a pilot study on the nearby barred spiral galaxy ngc 1672 , using the high - spatial - resolution imaging power of _ chandra _ , in conjunction with the spectral capabilities of _ xmm - newton _ , to study the x - ray properties of the nuclear emission .
we have supplemented this with multiwavelength imaging data from the _
hst_/acs , _ spitzer _
irac and atca ground - based radio observations .
we have also analyzed the x - ray point source and diffuse emission properties of this galaxy .
our main results can be summarized as follows : 1 .
we detect 28 x - ray sources within the @xmath0 area of the galaxy , ranging between @xmath52 erg s@xmath2 in the 0.310kev band .
the positions of many of the sources outside the nuclear region correlate with star - formation in the bar and spiral arms , and two are identified as background galaxies in the _ hst _ images .
nine of the x - ray sources are ulxs , with the three brightest ( @xmath1 erg s@xmath2 ) located at the ends of the bar .
there is also a diffuse component that traces star formation in the bar region , with a total x - ray luminosity of @xmath57 erg s@xmath2 in the 0.52kev band .
2 . with high - spatial - resolution _ chandra _ imaging ,
we have shown , for the first time , that ngc 1672 possesses a hard ( @xmath3 ) nuclear x - ray source , with a low 210kev luminosity of @xmath4 erg s@xmath2 ( figure [ fig : ctr ] ) .
this in turn is surrounded by an x - ray bright circumnuclear star - forming ring ( comprised of point sources and hot gas ) , which dominates the 210kev x - ray emission in the central region of the galaxy ( @xmath158 erg s@xmath2 within a radius of 22 ) .
the ring structure and nuclear source are also present in the radio and ir imaging .
the central ring region is not spatially resolved with _
xmm - newton _ , but spectral fits indicate that there is no significant long - term variability of the nuclear flux over the 2 year period between the _ xmm - newton _ and _ chandra _ observations , or with respect to _ rosat _ hri observations 9 and 14 years earlier .
spatially resolved , multiwavelength photometry of the nuclear source supports the hypothesis that it is an llagn accreting at a low rate ( @xmath100 ) .
the main discriminator is a bright radio counterpart , which can not be easily accounted for by other possibilities such as a normal xrb or the combined emission from young sn / snrs .
the broad - band sed is consistent with a radio - loud llagn , although an excess of optical and ir flux indicates the presence of star formation activity close to the central black hole .
the _ hst _ imaging also shows the presence of a nuclear spiral , which is a candidate mechanism for agn fueling @xcite .
while the x - ray hardness ratio indicates that there is some absorption present in the nuclear region ( @xmath159 @xmath22 ) , the ratio of the 210kev x - ray flux to the [ o iii ] @xmath1435007 line flux does not provide strong evidence that the nucleus is compton - thick .
a preliminary investigation of published results shows that both barred and unbarred spiral galaxies possess a broad range of observed nuclear x - ray luminosities . to further investigate the effect that bars , and their associated circumnuclear features , have on their nuclear activity
, we are conducting a multiwavelength survey of nearby galaxies . using optical / ir high - spatial - resolution imaging to resolve the circumnuclear structure in a well - matched sample of barred / unbarred galaxies
, we will use complementary _ chandra _ observations to measure their nuclear x - ray properties and search for a potentially absorbed agn .
morphological classification of galaxies is also vitally important , as newer ir surveys ( e.g. @xcite ) are discovering that more galaxies host bars than optical data suggest .
galaxies classed as ` oval ' should also be considered , since they may be just as effective as barred galaxies at moving gas from the disks of galaxies to their nuclear regions @xcite .
these issues will be addressed in our ongoing study , and will provide a crucial element in assessing the overall demography and evolution of supermassive black holes in the local universe .
we thank the referee for their helpful comments that have improved the manuscript .
lpj acknowledges funding for the majority of this project from the nasa postdoctoral fellowship program .
wnb acknowledges nasa adp grant nnx10ac99 g .
this research has made use of data obtained from the _ chandra _ data archive , and software provided by the _
chandra _ x - ray center ( cxc ) , which is operated by the smithsonian astrophysical observatory for and on behalf of the national aeronautics space administration under contract nas8 - 03060 .
the _ xmm - newton _ observatory is an esa science mission with instruments and contributions directly funded by esa member states and nasa .
observations made with the nasa / esa hubble space telescope were obtained from the data archive at the space telescope institute .
stsci is operated by the association of universities for research in astronomy , inc . under the nasa contract
nas 5 - 26555 .
the _ spitzer _
space telescope is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa .
the australia telescope compact array is part of the australia telescope which is funded by the commonwealth of australia for operation as a national facility managed by csiro .
this research has made use of the nasa/ ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with nasa .
llccccccccccccc 1 & [email protected] & 1.85/1.44/4.89 & @xmath36 1.02 & 1.25 @xmath160 & 15.80/13 & 3.98 @xmath161 & 1.27 @xmath162 & & 2.63 & 6.61 @xmath163 & 2.17 @xmath164 & 1.26/1 & 2.29 @xmath165 & 0.73 @xmath166 + 2 & [email protected] & 0.38/0.37/1.36 & & & & 0.73 @xmath167 & 0.23 @xmath168 & & 1.04 & & & & 0.93 @xmath169 & 0.29 @xmath170 + 3 & [email protected] & 0.83/0.67/1.90 & & & & 1.22 @xmath171 & 0.39 @xmath172 & & 1.49 & & & & 1.28 @xmath173 & 0.41 @xmath174 + 4 & [email protected] & 2.82/2.70/12.56 & 1.51 @xmath175 & 2.42 @xmath176 & 27.95/34 & 4.13 @xmath177 & 1.31 @xmath178 & & 23.28 & 1.67 @xmath179 & 1.75 @xmath180 & 35.71/36 & 21.74 @xmath181 & 6.91 @xmath182 + 5 & [email protected] & 10.43/10.00/36.23 & @xmath36 0.27 & 1.29 @xmath183 & 120.27/119 & 22.40 @xmath184 & 7.12 @xmath185 & & 19.10 & @xmath36 0.45 & 1.12 @xmath186 & 38.45/30 & 24.60 @xmath187 & 7.82 @xmath188 + 6 & [email protected] & 5.03/5.10/ ... & @xmath36 0.82 & 1.87 @xmath189 & 13.88/19 & 8.80 @xmath190 & 2.80 @xmath191 & & 1.51 & & & & 1.36 @xmath192 & 0.43 @xmath193 + 7 & [email protected] & 2.53/2.97/6.10 & @xmath36 1.32 & 1.81 @xmath194 & 21.18/19 & 4.88 @xmath195 & 1.55 @xmath196 & & 3.13 & @xmath36 5.24 & 1.33 @xmath197 & 0.91/2 & 3.96 @xmath198 & 1.26 @xmath199 + 8 & [email protected] ( b ) & 1.32/1.20/0.77 & & & & 2.48 @xmath200 & 0.79 @xmath201 & & 0.80 & & & & 0.69 @xmath202 & 0.22 @xmath203 + 9 & [email protected] & 0.42/0.44/1.06 & & & & 0.34 @xmath204 & 0.11 @xmath205 & & 0.36 & & & & 0.35 @xmath206 & 0.11 @xmath207 + 10 & [email protected] & & & & & & & & 0.53 & & & & 0.42 @xmath208 & 0.13 @xmath209 + 11 & [email protected] & 1.87/1.24/ ... & & & & 2.52
@xmath210 & 0.80 @xmath211 & & 1.07 & & & & 1.59 @xmath212 & 0.51 @xmath213 + 12 & [email protected] ( c ) & & & & & & & & 1.39 & & & & 1.18 @xmath214 & 0.38 @xmath215 + 13 & [email protected] ( c ) & & & & & & & & 3.83 & @xmath60.22@xmath216 & @xmath63.36@xmath216 & 8.94/3 & 3.20 @xmath217 & 1.02 @xmath218 + 14 & [email protected] ( c ) & & & & & & & & 0.83 & & & & 0.73 @xmath219 & 0.23 @xmath220 + 15 & [email protected] ( n ) & & & & & & & & 0.84 & & & & 1.37 @xmath221 & 0.44 @xmath222 + 16 & [email protected] ( c ) & & & & & & & & 1.89 & & & & 1.59 @xmath223 & 0.50 @xmath224 + 17 & [email protected] & & & & & & & & 0.88 & & & & 0.78 @xmath225 & 0.25 @xmath226 + 18 & [email protected] ( c ) & & & & & & & & 2.09 & & & & 1.73 @xmath227 & 0.55 @xmath228 + 19 & [email protected] & 0.83/0.99/2.71 & @xmath60.32@xmath216 & @xmath61.84@xmath216 & 12.36/5 & 1.19 @xmath229 & 0.38 @xmath230 & & 0.96 & & & & 0.82 @xmath231 & 0.26 @xmath232 + 20 & [email protected] & 0.60/0.34/1.74 & & & & 0.90 @xmath233 & 0.28 @xmath234 & & 0.78 & & & & 0.61 @xmath235 & 0.19 @xmath236 + 21 & [email protected] & 0.64/0.60/1.94 & & & & 1.06 @xmath237 & 0.34 @xmath238 & & 0.31 & & & & 0.23 @xmath239 & 0.07 @xmath240 + 22 & [email protected] & 0.93/1.24/4.02 & 1.46 @xmath241 & 2.87 @xmath242 & 11.21/8 & 1.12 @xmath243 & 0.36 @xmath244 & & 0.76 & & & & 0.64 @xmath245 & 0.20 @xmath207 + 23 & [email protected] & 0.69/0.68/ ... & & & & 1.44 @xmath246 & 0.46 @xmath170 & & 1.21 & & & & 1.02 @xmath247 & 0.33 @xmath170 + 24 & [email protected] & & & & & & & & 5.25 & @xmath36 1.25 & 1.80 @xmath248 & 4.03/6 & 4.50 @xmath249 & 1.43 @xmath250 + 25 & [email protected] & 7.77/7.15/26.82 & 1.12 @xmath251 & 1.91 @xmath252 & 86.23/89 & 10.47 @xmath253 & 3.33 @xmath254 & & 31.90 & 2.52 @xmath255 & 1.65 @xmath256 & 42.89/51 & 28.21 @xmath257 & 8.97 @xmath258 + 26 & [email protected] ( b ) & 1.05/0.77/2.83 & @xmath36 3.00 & 1.57 @xmath259 & 4.46/5 & 1.78 @xmath260 & 0.57 @xmath261 & & 1.69 & & & & 1.45 @xmath262 & 0.46 @xmath204 + 27 & [email protected] & 0.58/0.77/2.55 & @xmath36 1.63 & 2.10 @xmath263 & 1.90/3 & 1.18 @xmath264 & 0.38 @xmath265 & & & & & & & + 28 & [email protected] & 1.34/1.60/5.47 & 1.14 @xmath266 & 1.25 @xmath267 & 16.04/13 & 3.79 @xmath268 & 1.21 @xmath269 & & 3.38 & @xmath36 44.45 & 2.51 @xmath270 & 0.42/2 & 4.65 @xmath271 & 1.48 @xmath272 + lccc @xmath273 & @xmath274 & [email protected] & [email protected] + @xmath275 & @xmath276 & [email protected] & [email protected] + @xmath277 & @xmath278 & [email protected] & [email protected] + @xmath279 & @xmath280 & [email protected] & [email protected] + @xmath281 & @xmath282 & [email protected] & [email protected] + @xmath283 & @xmath284 & [email protected] & [email protected] + lcccc @xmath273 & @xmath274 & @xmath285 & 8.94@xmath286 & @xmath287 + @xmath275 & @xmath276 & @xmath288 & 6.98@xmath286 & @xmath289 + @xmath277 & @xmath278 & @xmath290 & 5.38@xmath286 & @xmath291 + @xmath279 & @xmath280 & @xmath292 & 4.56@xmath286 & @xmath293 + @xmath281 & @xmath282 & @xmath294 & 4.56@xmath286 & @xmath295 + @xmath283 & @xmath284 & @xmath296 & 3.60@xmath286 & @xmath297 + 3.6@xmath41 m & [email protected] & [email protected] & 8.45@xmath298 & [email protected]@xmath299 + 4.5@xmath41 m & [email protected] & [email protected] & 6.68@xmath298 & [email protected]@xmath299 + 5.8@xmath41 m & [email protected] & [email protected] & 5.23@xmath298 & [email protected]@xmath300 + 8.0@xmath41 m & [email protected] & [email protected] & 3.81@xmath298 & [email protected]@xmath300 + 5ghz ( 6-cm ) & & [email protected] & 4.80@xmath301 & [email protected]@xmath302 + 10ghz ( 3-cm ) & & [email protected] & 8.64@xmath301 & [email protected]@xmath302 + llccc 2 & 1.16@xmath303@xmath304 & 1.49@xmath305@xmath304 & x-9 & [email protected]@xmath304 + 4 & 7.92@xmath306@xmath304 & 2.48@xmath307@xmath92 & x-7 & [email protected]@xmath304 + 5 & 2.03@xmath308@xmath92 & 1.91@xmath309@xmath92 & x-3 & [email protected]@xmath92 + 23 & 3.27@xmath310@xmath304 & 1.60@xmath311@xmath304 & x-5 & [email protected]@xmath304 + 25 & 1.37@xmath312@xmath92 & 2.99@xmath307@xmath92 & x-2 & [email protected]@xmath92 + lccl @xmath313 & @xmath314 & @xmath314 & kev + @xmath315 & @xmath316 & @xmath317 & @xmath318 @xmath319pc + @xmath320 & @xmath321 & @xmath322 & kev + @xmath323 & @xmath324 & @xmath325 & @xmath318 @xmath319pc + @xmath326 & @xmath327 & @xmath328 & kev + @xmath329 & @xmath330 & @xmath331 & @xmath318 @xmath319pc + @xmath332 & @xmath333 & @xmath334 & kev + @xmath335 & @xmath336 & @xmath337 & @xmath318 @xmath319pc + @xmath88 & 0 ( fixed ) & @xmath338 & @xmath339 @xmath22 + @xmath35/dof & 310.6/268 & 307.7/260 & + @xmath340 & @xmath341 & @xmath342 & + @xmath343 & @xmath344 & @xmath345 + @xmath346 & @xmath347 & @xmath348 & + @xmath349 & @xmath350 & @xmath351 + lccccccccc + 0.59@xmath352 & 2.27@xmath353 & & & @xmath354 & 0.60@xmath355 & 435.8/343 & 3.07@xmath356 & 12.29@xmath357 + + 0.76@xmath358 & 2.21@xmath359 & @xmath354 & 0.20@xmath360 & @xmath354 & 0.61@xmath361 & 401.4/341 & 3.01@xmath362 & 12.86@xmath363 + + 0.72@xmath364 & 2.17@xmath365 & @xmath360.54 & 0.24@xmath366 & 2.01@xmath367 & 0.60@xmath361 & 392.7/340 & 3.12@xmath368 & 15.14@xmath369 + | we have performed an x - ray study of the nearby barred spiral galaxy ngc 1672 , primarily to ascertain the effect of the bar on its nuclear activity .
we use both _
chandra _ and _ xmm - newton _ observations to investigate its x - ray properties , together with supporting high - resolution optical imaging data from the _ hubble space telescope _ ( _ hst _ ) , infrared imaging from the _ spitzer space telescope _ , and atca ground - based radio data .
we detect 28 x - ray sources within the @xmath0 area of the galaxy , many of which correlate spatially with star - formation in the bar and spiral arms , while two are identified as background galaxies in the _ hst _ images .
nine of the x - ray sources are ulxs , with the three brightest ( @xmath1 erg s@xmath2 ) located at the ends of the bar . with the spatial resolution of _ chandra _ , we are able to show for the first time that ngc 1672 possesses a hard ( @xmath3 ) nuclear x - ray source with a 210kev luminosity of @xmath4 erg s@xmath2 .
this is surrounded by an x - ray bright circumnuclear star - forming ring , comprised of point sources and hot gas , which dominates the 210kev emission in the central region of the galaxy .
the spatially resolved multiwavelength photometry indicates that the nuclear source is a low - luminosity agn ( llagn ) , but with star formation activity close to the central black hole .
a high - resolution multiwavelength survey is required to fully assess the impact of both large - scale bars and smaller - scale phenomena such as nuclear bars , rings and nuclear spirals on the fueling of llagn . |
the term correlation has come to represent all electronic or quasi - particle interactions that go beyond a simple single slater determinant description .
it can be divided into two parts , classical and quantum correlation .
the former gives rise to mixed states , and the latter to entangled states .
the four different combinations of classical and quantum correlation result in the following classifications of an n - electron state in terms of wave - functions and density operators @xmath56 where @xmath57 is the vacuum state .
here @xmath58 and @xmath59 are sets of probabilities normalized to one , and @xmath60 are expansion coefficients .
the vectorial indices @xmath61 and @xmath62 are composed of the ordered components @xmath63 and @xmath64 referring to a generic one - particle basis .
for example the state @xmath65 corresponds to a many body index @xmath66 .
classical correlations can derive from physical processes , e.g. thermal decoherence , but also be induced by a local projection upon some set of orbitals , e.g. the d- or f - orbitals of single atom in a solid . to see this ,
let us consider a set orbitals @xmath67 labelled by the indices @xmath68 , and define the operator subspaces @xmath69 the full hilbert space @xmath70 can be partitioned as @xmath71 where @xmath72 the local projection of a pure state @xmath73 upon @xmath67 is obtained by taking the partial trace over @xmath74 : @xmath75 where @xmath76 .
the locally projected state @xmath77 is in general a mixed state , unless @xmath78 is idempotent .
it becomes harder to distinguish two states after a local projection since some information is lost in the partial trace over @xmath74 . for example , the highly mixed state @xmath15 , the pure and separable state @xmath25 , and the pure entangled state @xmath79 given by @xmath80 become indistinguable after a local projection onto the orbitals 1 and 2 ( @xmath81 ) .
it is therefore in general not possible to draw any conclusions about the classical or quantum correlations in the original unprojected state from its local projection .
quantum correlations can stem from physical electron - electron interaction , e.g. coulomb interaction , or be induced by a non - local projection .
a non - local projection is qualitatively different from a local projection in that it projects onto a set of many - body states , and not a set of orbitals .
the entanglement of the projected state may be larger than that of the original state .
a trivial example is when the non - local projection is defined with respect to a single entangled state @xmath14 , as any separable state with a non - zero overlap with @xmath14 becomes entangled after the projection .
a less trivial example is given by the projection operator @xmath82 the projection with @xmath83 makes the pure separable spin - coherent state @xmath84 entangled .
in contrast to the first example it should be noted that @xmath83 projects onto two separable states .
moreover , the projection itself corresponds to a simple stern - gerlach experiment of a spin-1 particle with a single slit to remove the spin @xmath85 components .
the problem of finding the maximal overlap @xmath86 between an n - electron pure state @xmath14 and any pure separable state can be written as @xmath87 = \max_{{\tilde{\mathbf{v } } } } \left|\langle 0|\sum_{i_1\cdots i_n}^k { \hat{\mathbf{c}}}_{i_n } { \tilde{\mathbf{v}}}_{i_n n } \cdots{\hat{\mathbf{c}}}_{i_1 } { \tilde{\mathbf{v}}}_{i_1 1 } |\psi\rangle\right|^2,\label{eqn : omax}\ ] ] where @xmath88 is the number of orbitals and @xmath89 is a unitary matrix .
a direct numerical optimization of the overlap , with respect to the parameters in @xmath89 , becomes quickly very cumbersome as the number of orbitals and electrons increase . a practical alternative to a brute force numerical optimization
is the search function @xmath90 , schematically presented in pseudocode form in fig .
[ fig : search ] .
the algorithm mimics an experiment where a sequence of one - electron removal measurements are performed .
it is easy to visualize the search for the optimal @xmath89 as an extremization of the intermediate intensities at each step in the measurement process .
as the intensity monotonically decreases at each step , it is advantageous to cast the algorithm as a recursive depth first search function @xmath91 $ ] , where @xmath92 is the currently known maximal overlap . 1 .
_ # measurement optimization _ : + diagonalize the one - particle density matrix @xmath93 , where @xmath94 , such that the eigenvalues @xmath95 are sorted from largest to smallest .
_ # detect the vacuum state : _
+ if ( @xmath96 ) * return * @xmath97 3 .
_ # perform the measurement and add another one - particle detector : _
+ for @xmath98 + if = ( @xmath99 ) then + @xmath100\right)$ ] + end if + done 4 .
* return * @xmath92 although @xmath101 $ ] may not return the maximal overlap for a general state , it is guaranteed to be exact if @xmath14 is separable or represents a 2-electron system . in the former case @xmath14
can be represented as a single slater determinant .
its natural orbitals are therefore either completely occupied or empty , and the algorithm will return an overlap equal to one . the latter case can be proved by noting that the first electron removal leads to a state with one single electron , which is necessarily separable . hence , no intensity will be lost in the second measurement , which implies that the maximal overlap is equal to the maximal probability of successfully removing the first electron .
a reason the recursive search algorithm is not exact for systems with three or more electrons is that the optimal state is not always a superposition of pair - wise entangled states .
this can be illustrated by applying a spin ladder operator to a maximally spin - polarized separable 3-electron state @xmath102 , giving @xmath103 the three slater determinants on the right hand side in eq .
( [ eqn : sp3 ] ) are all pair - wise entangled , which implies that the corresponding reduced one - particle density matrix @xmath104 is diagonal .
if the slater search algorithm is applied to this state it will therefore keep the current basis and yield a maximal squared overlap of 1/3 .
however , the unitary spin transformation @xmath105 gives @xmath106 the first slater determinant on the right hand side of eq .
( [ eqn : sptrans ] ) has a squared overlap of 4/9 , which is larger than the overlap of 1/3 found for the untransformed basis .
the spin transformation in eq . ( [ eqn : spoptimal ] ) is related to the number of ways the electrons can arrange their spins . for an arbitrary spin - resolved state @xmath107 it takes the form @xmath108 where @xmath109 and @xmath110 . to check the accuracy of our results for the transition metal oxides we applied a numerical conjugate gradient ( cg ) scheme set to maximize the overlap in eq .
( [ eqn : omax ] ) directly via the parameters in the unitary transformation @xmath89 .
the obtained maximal overlaps were all in exellent agreement with the results obtained from the search algorithm once the extra spin transformations in eq .
( [ eqn : spgoptimal ] ) had been applied to the spin - resolved states .
the search space of the algorithm can of course be extended further , e.g. by shifting the order of the electron removal measurements or using the final state as an input to a cg scheme , but such development is beyond the scope of this article . for a general many - body density operator @xmath15 , the linear entropy @xmath111 = { \mathrm{tr}}[{\hat{\mathbf{\rho}}}({\hat{\mathbf{1}}}-{\hat{\mathbf{\rho}}})]\ ] ] measures how far such a state
is from being pure .
it has been known for a long time @xcite that @xmath15 can be represented by a single slater determinant , i.e. a pure and separable state , if and only if the one - particle reduced density matrix @xmath112 is idempotent
. the linear entropy of the one - particle reduced density matrix @xmath113 = { \mathrm{tr}}[{\tilde{\mathbf{\rho}}}({\tilde{\mathbf{1}}}-{\tilde{\mathbf{\rho}}})]\ ] ] measures the deviation from idempotency , which makes it sensitive to both classical correlations and entanglement in @xmath15 .
a lower bound on the contribution from the entanglement can be obtained by removing the maximal contribution from a mixed but separable ensemble , under the constrain that and .
let us consider a fictitious extended system , with @xmath114 electrons and @xmath115 orbitals , for which we define @xmath116 the concavity of @xmath117 implies that @xmath118 \leq s_l[{\tilde{\mathbf{\rho}}}'_{e}].\ ] ] evaluating @xmath119 $ ] gives @xmath118 \leq ns_l[{\hat{\mathbf{\rho } } } ] = { \mathrm{tr}}[{\tilde{\mathbf{\rho}}}]s_l[{\hat{\mathbf{\rho}}}]\ ] ] an analagous construction can be made for holes and @xmath120 orbitals , giving that @xmath118 \leq { \mathrm{tr}}[{\tilde{\mathbf{1}}}-{\tilde{\mathbf{\rho}}}]s_l[{\hat{\mathbf{\rho}}}].\ ] ] hence , the state @xmath15 is necessarily entangled if @xmath121 & = & s_l[{\tilde{\mathbf{\rho } } } ] - \min({\mathrm{tr}}[{\tilde{\mathbf{\rho}}}],{\mathrm{tr}}[{\tilde{\mathbf{1}}}-{\tilde{\mathbf{\rho}}}])s_l[{\hat{\mathbf{\rho}}}]\nonumber\\ & = & s_l[{\tilde{\mathbf{\rho } } } ] - \min(n , m)s_l[{\hat{\mathbf{\rho } } } ] > 0.\end{aligned}\ ] ] however , the reversed implication is not true in general .
any mixed state of the form @xmath122 where @xmath123 and @xmath124 larger than the number of orbitals , gives @xmath125 < 0 $ ] .
it should be noted that @xmath126 and the fidelity based entanglement measure @xmath127 \equiv 1 - \max_{{\hat{\mathbf{\rho } } } ' } { \mathrm{tr}}\left[\sqrt{\sqrt{{\hat{\mathbf{\rho}}}}{\hat{\mathbf{\rho}}}'\sqrt{{\hat{\mathbf{\rho}}}}}\right]^2,\ ] ] where @xmath25 is separable , are not equivalent even for pure states , in the sense there exist states @xmath128 and @xmath129 such that @xmath130 & < & e_f[|\psi_b\rangle\langle|\psi_b|],\\ e_{l}[|\psi_a\rangle\langle|\psi_a| ] & > & e_{l}[|\psi_b\rangle\langle|\psi_b|].\end{aligned}\ ] ] for example let us set @xmath131 @xmath130 & = & \frac{1}{2 } > 1-x = e_f[|\psi_b\rangle\langle|\psi_b|],\\ e_{l}[|\psi_a\rangle\langle|\psi_a| ] & = & 1 < ( 1-x)(1 + 3x ) = e_{l}[|\psi_b\rangle\langle|\psi_b|].\end{aligned}\ ] ] the difference comes from that @xmath132 only focuses on the largest overlapping pure separable state @xmath16 , while @xmath55 also takes into account the entanglement between the states orthogonal to @xmath16 .
any maximally spin - polarized n - electron state with spin quantum number @xmath133 can be decomposed as @xmath134 where @xmath57 is the vacuum state and @xmath135 are the coefficients defined above . applying the spin - coherent operator @xmath136 to @xmath38
gives @xmath137 where we have used that @xmath138 the unitary transformation @xmath139 brings eq .
( [ eqn : sp ] ) back to exactly the same form as eq .
( [ eqn : p ] ) , which proves that @xmath140 conserves the entanglement of @xmath38 .
the 2s+1 roots of unity @xmath141 $ ] satisfy the important relation @xmath142 for all @xmath143 .
applying @xmath144 to @xmath145 yields @xmath146 where @xmath41 and @xmath147 through eq .
( [ eqn : rhoin ] ) the thermal paramagnetic density operator @xmath148 can be written as @xmath149 we can now introduce a set of maximally spin - polarized separable states @xmath150 and use eq .
( [ eqn : pms ] ) to define @xmath151 and @xmath152 the entanglement preserving properties of @xmath140 entails that @xmath25 is separable as well . from eq .
( [ eqn : pms ] ) we also have that @xmath153 which implies that the fidelity between @xmath154 and @xmath155 becomes spin - independent @xmath156 \equiv { \mathrm{tr}}\!\left[\sqrt{\sqrt{{\hat{\mathbf{\rho}}}^t_{m_s}}{\hat{\mathbf{\rho}}}'_{m_s}\sqrt{{\hat{\mathbf{\rho}}}^t_{m_s}}}~\right]^2 \!\!= f[{\hat{\mathbf{\rho}}}_{s}',{\hat{\mathbf{\rho}}}^t_{s}].\label{eqn : fnos}\ ] ] substituting eq .
( [ eqn : ort ] ) and ( [ eqn : fnos ] ) into @xmath157 $ ] gives @xmath158 & = & { \mathrm{tr}}\!\left[\sqrt{\sqrt{{\hat{\mathbf{\rho}}}^t}{\hat{\mathbf{\rho}}}'\sqrt{{\hat{\mathbf{\rho}}}^t}}~\right]^2\nonumber\\ & = & { \mathrm{tr}}\!\left[\sqrt{\sum_{m = -s}^s \frac{\sqrt{{\hat{\mathbf{\rho}}}^t_m } { \hat{\mathbf{n}}}_z^{-1 } { \hat{\mathbf{\rho}}}_m ' { \hat{\mathbf{n}}}_z^{-1 } \sqrt{{\hat{\mathbf{\rho}}}^t_m}}{(2s+1)^2}}~\right]^2\nonumber\\ & = & { \left(\sum_{m_s =- s}^s \sqrt{\frac{f[{\hat{\mathbf{\rho}}}^t_{m_s},{\hat{\mathbf{\rho}}}_{m_s}']\binom{2s}{s - m_s}}{2^{2s}(2s+1)}}\right)\!\!}^2 \nonumber\\ & = & \frac{f[{\hat{\mathbf{\rho}}}_s',{\hat{\mathbf{\rho}}}_s^t]}{2^{2s}(2s+1)}{\left(\sum_{m_s =- s}^s \sqrt{\binom{2s}{s - m_s}}\right)\!\!}^2.\label{eqn : frrt}\end{aligned}\ ] ] finally , under the assumption that a separable density operator of the same form as @xmath25 gives the maximal fidelity in eq .
( [ eqn : frrt ] ) , the fidelity based entanglement measure can be written as @xmath159 = 1 - \frac{1-e_f[{\hat{\mathbf{\rho}}}^t_s]}{2^{2s}(2s+1)}{\left(\sum_{m_s =- s}^s \sqrt{\binom{2s}{s - m_s}}\right)\!\!}^2.\ ] ] | here we present a study of the entanglement in the electronic structure of the late transition metal monoxides - mno , feo , coo , and nio - obtained by means of density - functional theory in the local density approximation combined with dynamical mean - field theory ( lda+dmft ) .
the impurity problem is solved through exact diagonalization ( ed ) , which grants full access to the thermally mixed many - body ground state density operator .
the quality of the electronic structure is affirmed through a direct comparison between the calculated electronic excitation spectrum and photoemission experiments .
our treatment allows for a quantitative investigation of the entanglement in the electronic structure .
two main sources of entanglement are explicitly resolved through the use of a fidelity based geometrical entanglement measure , and additional information is gained from a complementary entropic entanglement measure .
we show that the interplay of crystal field effects and coulomb interaction causes the entanglement in coo to take a particularly intricate form .
entanglement is a fundamental aspect of quantum mechanics , responsible for a large range of complex phenomena not present in a classical setting .
the entanglement of distinguishable particles was historically seen as something spooky , but is now considered a valuable resource .
it plays an essential role in quantum information theory , and has been studied in great detail both theoretically and experimentally @xcite .
entanglement of indistinguishable particles has received less explicit attention , but it has been studied indirectly in both the quantum chemistry and condensed matter communities . describing the electronic structure of materials with a pure separable state , represented as a single slater determinant , is at the very heart of the modern computational approaches , and the inability to do so is known under the term ` correlation ' .
the term encompasses both classical and quantum correlations ( entanglement ) , where the former results in mixed states and the latter in entangled states .
the presence of entanglement is usually considered as a computational complication as it prevents the use of these standard approaches .
the development of reliable computational methods and entanglement measures are of key importance to turn also the entanglement of indistinguishable particles from a complication into a potential resource .
well known examples of strongly correlated materials are the late transition metal monoxides ( tmo )
mno , feo , coo and nio which have been under intense experimental and theoretical attention for a long time @xcite . here
we report on a theoretical description of the late tmos using the local density approximation plus dynamical mean field theory ( lda+dmft ) @xcite where the effective impurity model is solved by exact diagonalization ( ed ) @xcite .
the quality of the results is assessed through a direct comparison between the computed density of states ( dos ) and photoemission experiments ( xps / bis ) .
we then take advantage of the direct access to the local many - body ground state of the impurity problem to analyse the entanglement in detail .
+ the lda+dmft scheme is built around the mapping of the local lattice problem to an effective impurity problem .
the impurity system is described in ed through the local projected lda hamiltonian @xmath0 , a double counting term @xmath1 , the on - site coulomb interaction @xmath2 , and a few auxiliary bath states , giving the hamiltonian @xmath3 where the indices @xmath4 run over the local correlated orbitals and @xmath5 runs over the auxiliary bath states .
the energies @xmath6 and the hybridization strength @xmath7 of the auxiliary bath states mimic the hybridization between the tm-3d and the o-2p and o-2s orbitals .
the hamiltonian of the finite system is diagonalized numerically to produce an analytical self - energy .
the calculations were carried out in the paramagnetic ( pm ) phase , using a finite temperature ( @xmath8 ry ) fully charge self - consistent lda+dmft implementation @xcite .
further technical details , including the parametrization of @xmath9 and the fitting of @xmath6 and @xmath7 , can be found in the supplemental material @xcite . in fig .
[ fig : tmo ] the calculated projected spectral function of the tm-3d and o-2p states are compared to experimental xps ( electron removal ) and bis ( electron addition ) data @xcite showing mainly the contribution of the tm-3d states , due to the photon energies used . the overall agreement for mno , coo , and nio is excellent , and even minor experimental features like the high energy satellites are found in our theory . the coo sample used in the experiment
was doped with 1% li to avoid charging effects .
however , this doping gives rise to co@xmath10o@xmath11-like domains in the sample , which contribute to the early onset seen in the bis spectrum the projected spectral function of feo show an acceptable agreement with the experimental spectrum , although the relative intensities could be improved .
a greater concern is that the initial shoulder at -0.5 ev is missing and that the peak at 3.5 ev is not evident in the experimental data .
however , this may be related to the fact that the experiment was performed with a non - stoichiometric sample ( fe@xmath12o ) including fe@xmath10o@xmath11-like domains @xcite , in contrast to the conventional feo unit cell used in the calculation .
nevertheless , it can not be ruled out that these features are beyond what can be described with the current method .
given the satisfactory comparison of the spectral properties with experimental data , we now turn our attention to the entanglement in the thermal many - body ground state of the impurity problem , described by the density operator @xmath13 . before we start let us for clarity briefly introduce the concept of pure , mixed , separable , and entangled n - electron many - body states .
a pure state is a state which can be described by a state vector @xmath14 , while a mixed state requires the use of a density operator @xmath15 .
a mixed state is said to be classically correlated as its components are related through classical probabilities . a separable
( non - entangled ) pure state @xmath16 can be written as a single slater determinant @xcite , while an entangled pure state @xmath14 requires a superposition of several slater determinants .
an entangled state is said to be quantum correlated as the superposition between its components is a quantum mechanical phenomenon . finally , a separable mixed state can be written in a diagonal form @xmath17 using only separable component states @xmath18 , while an entangled mixed state requires at least one entangled component @xcite .
a key point in the definition of mixed state entanglement is that a classical mixture of any two separable states remains separable .
there exist several ways to measure the entanglement in an n - electron many - body state @xcite . in the case of a pure state @xmath14 , we first look at the geometric entanglement measure @xcite @xmath19 = 1 - \max_{|\psi'\rangle } \left| \langle\psi'|\psi\rangle \right|^2,\label{eqn : efpure}\ ] ] where @xmath16 is restricted to be pure and separable . from a computational point of view
this measure is natural since @xmath16 corresponds to the best possible single slater determinant description of the system .
we perform the search for the optimal @xmath16 by recursively removing one electron at the time from the natural orbitals of the many - body state @xcite .
although this procedure is in general not guaranteed to find the optimal @xmath16 , it is exact for separable states and for pure 2-electron systems , where it gives @xmath20 = 1 - p_{max}$ ] with @xmath21 the largest eigenvalue of the corresponding one - particle reduced density matrix @xmath22 . in order to analyse the thermal density operator @xmath23 it is necessary to generalize the entanglement measure @xmath24 in eq .
( [ eqn : efpure ] ) to mixed states .
this can be achieved by replacing the overlap with the fidelity @xcite between @xmath23 and any separable state @xmath25 , @xmath26 = 1 - \max_{{\hat{\mathbf{\rho}}}'}{\mathrm{tr}}\left[\sqrt{\sqrt{{\hat{\mathbf{\rho}}}^t}{\hat{\mathbf{\rho}}}'\sqrt{{\hat{\mathbf{\rho}}}^t}}\right]^2,\label{eqn : ef}\ ] ] where @xmath17 , @xmath27 , and @xmath18 are separable states . performing the restrained maximization in eq .
( [ eqn : ef ] ) is in general a formidable task , as one has to consider the effect of mixing several non - orthogonal separable states .
however , the problem can be simplified by noting that @xmath28 of a pm system with negligible spin - orbit interaction commutes with @xmath29 , @xmath30 and its ladder operators @xmath31 .
this implies that there is a common eigenbasis in which the many - body eigenstates @xmath32 can be indexed by , and , and that the eigenvalues @xmath33 are independent of @xmath34 .
if the system has a well - defined spin moment , e.g. due to a strong on - site hund s coupling , its thermal density matrix is composed of a mixture of several degenerate eigenstates , all with the same spin quantum number @xmath35 @xmath36 we also observe that the spin - coherent operator @xmath37 conserves the entanglement @xcite of the maximally spin - polarized state @xmath38 .
setting @xmath39 yields @xmath40 where @xmath41 .
when applying the proposed slater search algorithm to the pure state @xmath42 of the tmos , the maximal overlap is always obtained for @xmath43 , where @xmath44 is the closest separable state to @xmath38 .
hence the separable state @xmath45 gives a local minimum of @xmath46 $ ] .
moreover each term @xmath42 contains the same amount of entanglement , originating from the material specific @xmath38 and the action of the renormalization operator @xmath47 .
therefore we conjecture that the local minimum given by @xmath48 is in fact a global minimum .
the geometric entanglement measure for the pm state @xmath23 can then be written as @xcite @xmath26 = 1 - \frac{1-e_g[{\hat{\mathbf{\rho}}}^t_s]}{2^{2s}(2s+1 ) } \left[\sum_{m =- s}^s \sqrt{{\binom{2s}{s - m}}}\right]^2,\label{eqn : eft}\ ] ] where @xmath49 . even when @xmath50 $ ] is zero there is still a non - trivial term remaining in eq .
( [ eqn : eft ] ) .
this source of entanglement is very robust as it does not depend on the details of the electronic structure of the system but only on @xmath51 . as a complement to the geometric entanglement measure we have used an entropic measure , based on the reduction of the many - body density matrix @xmath15 to the one - particle reduced density matrix @xmath52 .
this reduction converts the entanglement in @xmath15 to entropy , which can be measured directly @xcite , e.g. in form of linear entropy @xmath53 = { \mathrm{tr}}[{\tilde{\mathbf{\rho}}}({\tilde{\mathbf{1}}}-{\tilde{\mathbf{\rho}}})]$ ] .
however , quantifying entanglement in form of entropy requires care , as also the classical correlation in @xmath15 is converted to entropy .
we propose the following entropic entanglement measure @xmath54 = s_{l}[{\tilde{\mathbf{\rho } } } ] - \min\big({\mathrm{tr}}[{\tilde{\mathbf{\rho}}}],{\mathrm{tr}}[{\tilde{\mathbf{1}}}-{\tilde{\mathbf{\rho}}}]\big)s_{l}[{\hat{\mathbf{\rho}}}]\label{eqn : el},\ ] ] based on the fact that the entropy gained from the classical correlations is less than the number of electrons or holes times the entropy in @xmath15 @xcite .
removing this upper bound of the classical contribution simplifies the evaluation of @xmath55 , but at the same time makes it less sensitive to detect entanglement in mixed states .
apart from the entanglement measures proposed above , other choices are possible .
an intermediate scheme is to evaluate the geometric entanglement by using the relative von neumann entropy @xcite .
however , the current lack of an efficient minimization procedure for the relative entropy with respect to the separable reference state @xmath25 reduces its applicability to setting upper bounds on the entanglement @xcite . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^ " , ] |
the compound , bairo@xmath0 has drawn significant attention in the recent time due to the observation of charge density wave ( cdw ) transition despite insulating transport observed above and below the cdw transition temperature ( @xmath3 185 k).@xcite interestingly , it also exhibits a ferromagnetic transition at the same temperature ( 185 k ) where cdw ground state sets in .
the saturation magnetic moment in the ferromagnetic phase , however , is found to be very small ( @xmath40.03 @xmath5 ) compared to that expected from the @xmath6 electronic configuration corresponding to ir@xmath7 .
= 4.0 in the crystal structure of bairo@xmath0 is monoclinic ( space group b2/m ; unique axis @xmath8 ) . in this structure , iro@xmath9 octahedra form ir@xmath0o@xmath10 trimers by face sharing as shown in fig . 1 .
the ir@xmath0o@xmath10 trimers are inter - linked by corner - sharing to form columns parallel to @xmath11-axis .
the axis of the connecting trimers are marginally tilted ( @xmath412@xmath12 ) as shown in the figure.@xcite this leads to the monoclinic structure with the crystallographic angle , @xmath13 = 103.27@xmath12 .
this typical structure makes it a quasi - one dimensional system as has clearly been manifested in the anisotropic transport , magnetic and optical properties.@xcite the structure becomes more complex due to the twisting and buckling of the trimers resulting in the multiplicity of ir - o and ba - o bond distances , thereby creating 4 types of ir and 3 types of ba sites.@xcite different non - equivalent ir atoms are represented by ir1 , ir2 , ir3 and ir4 in the figure .
interestingly , each trimer consists of two non - equivalent ir atoms .
for example , if ir2 forms the central octahedron of a trimer , it is connected by face sharing to the other two octahedra formed by ir1 .
similarly , the neighboring ir@xmath0o@xmath10 trimer is formed by ir3 and ir4 , where ir4 forms the central octahedron and ir3 forms the other two octahedra .
thus , the inter - trimer link is established along @xmath11 direction via ( ir1)-o-(ir3 ) connectivity . despite very short intra - trimer ir - ir distances compared to ir - metal and ir - o - ir angle close to 180@xmath12 for inter - trimer connections , bairo@xmath0 exhibits insulating transport behavior .
in addition , there are multiple metal - insulator ( mi ) transitions within the cdw ground state .
while the occurrence of cdw transitions in the insulating phase is surprising , a recent high - resolution photoemission study@xcite reveals the signature of finite density of states at the fermi level , @xmath14 , at room temperature and it was shown that the cdw ground state evolves by opening a soft gap at @xmath14 across cdw transition temperature .
interestingly , the intensity at @xmath14 in the photoemission spectra at room temperature ( @xmath15 ) is significantly small consistent with the observation of very small electronic specific heat coefficient@xcite ( @xmath4 1 mj / mole.k@xmath16 ) .
in contrast , previous theoretical study using extended hckel tight - binding electronic band structure calculations@xcite exhibits large intensity at @xmath14 as expected for a @xmath6 electronic configuration corresponding to ir@xmath7 .
in addition , the analysis of the core level spectra in the photoemission measurements@xcite predicts the existence of two different kinds of ba sites at room temperature , which become similar before the onset of cdw ground state .
all these anomalies@xcite in the transport , magnetic and spectroscopic data reveal that bairo@xmath0 is truly an exotic material . since
, the 5@xmath1 orbitals in 5@xmath1 transition metal oxides are highly extended compared to that in 3@xmath1 or 4@xmath1 systems , it is expected that the electron correlation strength will be significantly weak and that the _ ab initio _ approaches will be successful to describe the electronic properties in this system . in this study , we therefore , investigate the electronic structure of bairo@xmath0 using _ state - of - the - art ab initio _ approaches .
the calculations for various magnetic configurations suggest that the ground state of bairo@xmath0 is ferromagnetic exhibiting small magnetic moments consistent with the experimental observations.@xcite the analysis of the valence band density of states in comparison with the experimental photoemission results@xcite suggests that the small intensity at @xmath14 in the photoemission spectra and the insulating transport behavior are presumably related to the pinning of @xmath14 at the upper edge of @xmath17 band due to the intrinsic non - stoichiometry of the samples@xcite and these edge states are localized due to disorder / crystallographic distortions.@xcite the calculations for the ba 4@xmath2 core levels exhibit signature of two non - equivalent ba sites with respect to the madelung potential at the ba sites in the crystal structure at room temperature .
the electronic band structure calculations were carried out using full potential linearized augmented plane wave ( fplapw ) method ( wien2k software@xcite ) within the local spin density approximations ( lsda ) .
the crystal structure considered for these calculations has b2/m space group with the @xmath8-axis as the unique axis .
the lattice parameters ( @xmath11 = 10.005 , @xmath18 = 15.174 , @xmath8 = 5.571 , @xmath19 = @xmath20 = 90@xmath12 and @xmath13 = 103.27@xmath12 ) and the atomic positions used for all the calculations correspond to the room temperature structure reported for well characterized polycrystalline and single crystalline samples in references [ 1,3 ] .
it is to note here that most of these studies for the crystal structure analysis report the space group as c2/m with @xmath18 axis as the unique axis , which has been transformed to b2/m space group by coordinate transformations ( @xmath21 ) in order to satisfy the requirement of the wien2k software .
the muffin - tin radii ( @xmath22 ) for ba , ir and o were set to 1.217 , 1.111 and 0.847 , respectively .
the convergence for different calculations were achieved considering 512 @xmath23 points within the brillouin zone .
the error bar for the energy convergence was set to @xmath24 1 mev .
in fig . 2 , we show the density of states calculated for the non - magnetic ground state using the lattice parameters corresponding to room temperature structure as mentioned in the previous section . the total density of states ( tdos ) , ir 5@xmath1 partial density of states ( pdos ) , o 2@xmath2 pdos and ba 5@xmath1 pdos are plotted in fig .
2(a ) , 2(b ) , 2(c ) and 2(d ) , respectively .
ba 5@xmath1 electronic states contribute in the energy range 4 ev above the fermi level as shown in fig .
2(d ) with almost no contribution at lower energies . o 2@xmath2 contribution in the ba 5@xmath1 dominated energy region is very little , which is expected for the 5@xmath1 energy level in this heavy alkaline earth atom compared to the cases observed for lighter alkaline earth atoms such as sr in sr - compounds or ca in ca - compounds having unoccupied 4@xmath1 and 3@xmath1 levels , respectively.@xcite this suggests that the role of ba - o covalency in the electronic / crystal structure is not significant in this system .
= 4.5 in there are three groups of features observed in tdos in the occupied part of the electronic structure shown below @xmath14 (= 0 ev ) . the energy range between -5.5 ev to -2.3 ev is primarily dominated by the o 2@xmath2 pdos with negligible contributions from ir 5@xmath1 pdos .
this suggests that the electronic states in this energy range are not bonded with the ir 5@xmath1 states and are known as non - bonding o 2@xmath2 states . in the energies below -5.5 ev ( -5.5 ev@xmath25 - 8.5 ev )
, o 2@xmath2 states have the largest intensity with finite contributions from ir 5@xmath1 electronic states .
thus , these features can be assigned as bonding bands having primarily o 2@xmath2 character .
the anti - bonding bands having @xmath17 symmetry appears in the energy range -2.5 ev to + 0.5 ev ( total width @xmath4 3 ev ) .
the @xmath26 band appears above the fermi level with a minimum excitation gap of about 1.2 ev from the upper edge of @xmath17 band .
the dos in these regions are essentially contributed by the ir 5@xmath1 electronic states with minimal contributions from o 2@xmath2 states .
thus , the electronic states at and around @xmath14 have primarily ir 5@xmath1 character , which naturally determines the transport , magnetic and other thermodynamic properties .
interestingly , dos at @xmath14 is found to be significantly large as also observed before@xcite suggesting highly metallic behavior of the system in contrast to the insulating transport , small specific heat coefficient observed@xcite and the high - resolution photoemission results recently published.@xcite in order to investigate the contribution from various non - equivalent ir sites separaterly , we plot the @xmath17 and @xmath26 band in fig .
3(a ) and 3(b ) , respectively . the total @xmath17 bandwidth for each of the ir atoms
is found to be close to 3 ev spanning the same energy region .
the pdos corresponding to ir2 and ir4 are almost uniformly distributed over the whole energy range .
however , ir1 and ir3 exhibit largest intensity close to the fermi level suggesting largest contribution to the conduction electrons .
such a scenario is expected since inter ir@xmath0o@xmath10 trimer connectivity is established via ( ir1)-o-(ir3 ) superexchange interaction as shown in fig . 1 and has largest extended character . ir2 and ir4
provide intra - trimer connectivity via @xmath27 interactions with ir1 and ir3 , respectively .
thus , there is a relative shift between the center of gravity of ir1 , ir3 contributions and ir2 , ir4 contributions .
the features in the energy range @xmath28 -0.75 ev has largest contribution from ir2 and ir4 sites and for @xmath29 -0.75 ev , the intensity arises primarily from ir1 and ir3 contributions .
= 5.0 in the effective width of the @xmath26 bands is significantly narrower in this system as shown in fig .
while ir2 and ir4 are primarily contributing in the energies 2.5 ev - 3.5 ev above @xmath14 , the ir1 and ir3 contributions appear below 2.5 ev .
interestingly , ir1 and ir3 are found to be very similar in both the energy regions , and ir2 and ir4 have similar dos distributions . since the axis of any of the iro@xmath9 octahedra are not aligned along any of the crystal axis as shown in fig . 1
, the @xmath17 bands will have contributions from all the @xmath1-orbital states ( @xmath30 , @xmath31 , @xmath32 , @xmath33 and @xmath34 ) defined in the axis system shown in fig . 1 .
the projected partial density of states corresponding to these orbital are shown separately for each of the ir sites in fig .
interestingly , the primary contribution ( @xmath4 35.77% ) at @xmath14 comes from @xmath32 electronic states located at ir1 and ir3 sites . @xmath30 and
@xmath34 at ir1 and ir3 sites also exhibit large contributions ( 14.3% and 16.6% respectively ) at @xmath14 .
the largest contributions from ir2 and ir4 , however , arise from @xmath32 ( 8.4% ) and @xmath33 ( 7.6% ) electronic states .
these numbers clearly manifest that the electronic states containing the quasi - one dimensional axis of the crystal structure ( @xmath35-axis ) have largest contribution at @xmath14 and hence clearly manifests the signature of large anisotropy observed in various bulk properties.@xcite = 5.0 in we now turn to the understanding of the magnetic behavior of bairo@xmath0 .
we carried out the spin - polarized calculations using the same crystal structure as in the case of non - magnetic calculations and use local spin density approximations .
the magnetic moment has been defined by the difference in occupancy of the majority ( up ) and minority ( down ) spin states .
spin - orbit interaction was not considered in these calculations .
interestingly , the calculations for various magnetic configurations reveal that the ferromagnetic ground state is 3.25 mev lower than ( error bar used for convergence is @xmath24 1 mev ) the total energy found for the non - magnetic ground state .
this energy difference is about 20% of @xmath36 ( @xmath37 = 185 k ) .
this calculation , thus , once again establish that lsda approximations are quite accurate in capturing the magnetic ground state of various transition metal oxides as has also been observed in 3@xmath1 transition metal oxides@xcite despite the fact that the electron correlation effects are underestimated in these calculations . the total spin magnetic moment is found to be about 1.34@xmath3810@xmath39 @xmath5 .
this is significantly small compared to that expected for @xmath6 electronic configuration for ir@xmath7 .
such a small value of the spin moment appears presumably due to the highly extended nature of the 5@xmath1 orbitals leading to large kinetic energy of the associated electrons .
the calculated spin moment is consistent with the experimental observation of small saturation magnetic moment ( @xmath4 0.03 @xmath5 ) in the magnetic measurements of single crystalline bairo@xmath0 ; this moment could be measured only after applying a magnetization field as high as 20 t.@xcite no magnetic contribution is found from the ba - related bands , as all these bands are either completely occupied and appear much below @xmath14 or completely unoccupied appearing far above @xmath14 .
the contribution from o 2@xmath2 states is of the order of 10@xmath40 .
the primary contribution to the total magnetic moment arises from the spin polarization of ir 5@xmath1 partial density of states .
interestingly , all the ir atoms are found to order ferromagnetically and the magnetic moment at the central ir site of the ir@xmath0o@xmath10 trimer is almost the half of that observed at the other ir sites .
= 4.5 in in fig . 5
, we show the spin - polarized ir 5@xmath1 partial density of states for the ferromagnetic ground state , where 5(a ) , 5(b ) , 5(c ) and 5(d ) show the results corresponding to ir1 , ir2 , ir3 and ir4 respectively .
the down - spin contributions are shown by inverting the @xmath41-axis in the same figure with same energy scale for up - spin contributions shown along @xmath35-axis .
no signature of exchange splitting is observed at any of the ir sites .
the up- and down - spin dos reveal almost identical structures in all the cases . in order to investigate the origin of the spin moment
, we show the total ir 5@xmath1 pdos for both up- and down - spin states in fig .
6(a ) and the difference between up- and down - spin contributions in fig .
it is clear that the observed magnetic moment appears due to small asymmetry of the up- and down - spin density of states as a function of energy .
thus , the occupancy of the up- and down - spin contributions is different leading to a small spin - moment .
now , we compare the calculated results with the experimental photoemission spectra published recently for bairo@xmath0.@xcite in fig .
7(a ) , we show the ir 5@xmath1 band obtained from he ii spectra@xcite and overlap with the calculated @xmath17 band of ir 5@xmath1 pdos . while the calculated results exhibit large intensity at the fermi level suggesting highly metallic behavior of the system , the experimental ir 5@xmath1 feature at room temperature exhibit only very small intensity .
the total width of the occupied part in the calculated pdos is also substantially smaller compared to the ir 5@xmath1 signal in the experimental spectra .
such observations of significant reduction of spectral weight at the fermi level and the population of the higher binding energy region often attributed to the underestimation of electron - electron coulomb repulsion effect in the _ ab initio _ approaches
. however , such a strong correlation effect in these highly extended 5@xmath1 electrons is unexpected .
= 4.5 in it is to note here that the difference between experimental bulk spectral functions and the _ ab initio _ results could well be explained in the 4@xmath1 transition metal oxides@xcite using perturbative approach to introduce the correlation effect in the self energy of the system .
the effective correlation strength , @xmath42 ( @xmath43 = electron - electron coulomb repulsion strength , @xmath44 = bandwidth ) was found to be small ( @xmath45 0.2 ) in these systems .
we thus calculate the spectral function for bairo@xmath0 in the same way , using the second order perturbation method employed by treglia _
et al._@xcite in this method , the spectral function can be expressed as , @xmath46 where , @xmath47 is the retarded green s function representing the many electron system and is given by @xmath48 where , @xmath49 is the self energy of the system . for small @xmath43 , @xmath49 can be calculated using perturbation approach upto the second order term@xcite within the local approximations .
we have used ir @xmath17 pdos shown in fig .
7(a ) as the band dos for these calculations .
the calculated spectral functions for different @xmath42 are shown in fig .
7(b ) and the calculated self energy is shown in the inset of the same figure .
the increase in @xmath42 leads to a significant spectral weight transfer to higher energies .
the new features appearing at higher binding energies due to such spectral weight transfer is known as incoherent feature and represent the electronic states essentially localized due to such correlation effect .
the feature at the fermi level represent the delocalized states and is known as coherent feature . the total weight of the coherent feature diminishes gradually with the increase in @xmath42 .
however , the intensity at the fermi level remains almost the same for @xmath42 as large as 1.0 .
the spectral function obtained in this process could not generate the spectral lineshape observed in the experimental spectrum for any value of @xmath42 .
this suggests that this procedure may not be adequate to determine the experimental results .
interestingly , a rigid shift of the calculated @xmath17 band by 0.5 ev towards higher binding energies provide a significantly good description of the experimental spectrum . the width and spectral distribution in the shifted spectral function shown in fig .
7(a ) is remarkably similar to that observed in the ir 5@xmath1 band in the he ii spectrum . in order to investigate this
further we compare the o 2@xmath2 pdos with the experimental spectra as it is well known that the _ ab initio _ calculations are quite successful to provide a good description of the o 2@xmath2 band even in 3@xmath1 transition metal oxides.@xcite we show the experimental al @xmath50 and he i spectra in fig . 8
adopted from ref .
the changes in these experimental spectra in the energy range between 3 ev to 9 ev binding energies concomitant to the change in relative photoemission cross section@xcite of the o 2@xmath2 and ir 5@xmath1 electronic states suggest that the features in 3 - 6 ev binding energies appear primarily due to the excitations of o 2@xmath2 non - bonding states and the intensity at higher binding energies are due to the bonding electronic states having large o 2@xmath2 character .
interestingly , the calculated o 2@xmath2 pdos shows the intense non - bonding o 2@xmath2 features around 2.3 - 5.5 ev binding energies with the peak at 2.8 ev binding energy ( see fig .
2 ) , where the dip appears in the he i spectrum .
this clearly suggests that it is necessary to shift the o 2@xmath2 pdos by at least 0.5 ev towards higher binding energies to compare with the experimental spectra .
the shifted o 2@xmath2 pdos shown by solid line in fig .
8 provide a good description of the experimental results .
= 4.5 in thus , the results in fig . 7 and fig .
8 suggest that the fermi level is shifted to the upper edge of the @xmath17 band presumably due to the electron doping into the system .
such a shift would stabilize the system due to the closeness to the fully filled electronic configuration resulting to lowering in energy .
interestingly , various experimental results indeed indicate that the oxygen content is often found to be less than 3.0,@xcite which effectively leads to electron doping into the ir 5@xmath1 band .
in addition , the transport measurements always exhibit insulating temperature dependence.@xcite such a behavior is not unusual since it is natural that the small intensity at the upper edge of the @xmath17 band will be localized due to the disorder and/or crystallographic distortions@xcite in this quasi - one dimensional structure leading to an insulating nature .
all these observations , thus , clearly indicate that the electron correlation may not be the primary origin for the differences between experimental and theoretical results .
the electron doping due to the oxygen non - stoichiometry intrinsic to bairo@xmath0 possibly plays the key role in determining the electronic structure and subsequently various physical properties in this system .
= 4.4 in we now turn to the question of multiple ba core level signals in the experimental spectra.@xcite we show the dos corresponding to ba 5@xmath2 core levels in fig .
there are three non - equivalent ba sites in the structure with respect to the ba - o bond length and the symmetries .
it is clear that the dos corresponding to ba1 and ba2 appear in the same energy range .
while the width of the dos corresponding to ba2 5@xmath2 states is similar to those corresponding to ba1 and ba3 , the peak position appear at higher binding energies .
the energy difference between these spectral features is about 0.25 ev .
this observation clearly suggests that the madelung potential at all the ba sites are not same and is consistent with the experimental observation of two non - equivalent features corresponding to each ba core level photoemissions . such relative shift of the dos corresponding to different non - equivalent ir atoms
is not observed in these calculations ( e.g. see fig . 3 for the @xmath17 bands corresponding to 4 non - equivalent ir sites ) .
in summary , we investigate the detailed electronic structure of bairo@xmath0 in this study using full potential linearized augmented plane wave method within the local spin density approximations .
the calculations for various magnetic configurations reveal lowest energy for the ferromagnetic ground state and is consistent with the experimental observations .
the ferromagnetic ground state energy calculated using room temperature crystal structure is found to be 3.25 mev lower than the non - magnetic ground state .
no exchange splitting is observed in any of the electronic states involved in forming the valence band .
the small magnetic moment appears due to unequal population of the different ir 5@xmath1 spin density of states .
comparison with the photoemission spectra suggests that electron- electron coulomb repulsion may not be the key factor in determining the electronic structure as expected for the highly extended nature of the 5@xmath1 electronic states .
the discrepancy between the experimental spectra and the calculated results appears to arise primarily from the intrinsic non - stoichiometry of bairo@xmath0 leading to electron doping into the @xmath17 band .
the calculations for ba 5@xmath2 core levels exhibit the signature of multiple non - equivalent ba sites in the room temperature structure as observed experimentally . thus , these results provide a significant advance towards understanding the electronic properties of the exotic compound , bairo@xmath0 .
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 .
isbn 3 - 9501031 - 1 - 2 . | we investigate the electronic structure of bairo@xmath0 , an interesting compound exhibiting charge density wave transition in its insulating phase and ferromagnetic transition at the same temperature , using full potential linearized augmented plane wave method within the local spin density approximations .
the ferromagnetic ground state could exactly be described in these calculations and the calculated spin magnetic moment is found to be small as observed in the magnetic measurements .
interestingly , no signature of exchange splitting is observed in the density of states corresponding to ir 5@xmath1 and/or any other electronic states .
the small spin moment appears essentially due to unequal population of the up- and down - spin ir 5@xmath1 bands .
comparison of the valence band density of states with the experimental spectral functions suggests that a rigid shift of the fermi level towards higher energies in the calculated density of states provides a good description of the experimental spectra .
this indicates that the intrinsic oxygen non - stoichiometry leads to electron doping in the system and plays the primary role in determining the electronic structure rather than the electron correlation effects as often observed in other systems .
the calculated results for ba 5@xmath2 core levels show that the madelung potential of one of the three non - equivalent ba atoms is different from that of other two as predicted in the recent experiments . |
the standard cosmological paradigm , while successful in describing our observable universe , is plagued with horizon and flatness problems . moreover , despite being able to explain the large scale structure formation due to some seed fluctuations in our universe , standard cosmology fails to provide a mechanism that can produce such seed fluctuations .
inflationary models are at present the only way to provide solutions for these shortcomings in standard cosmology @xcite . according to this paradigm ,
the universe at an early epoch experienced an exponentially rapid expansion for a very brief period due to some apparently repulsive gravity - like force .
such a scenario not only can successfully address the horizon and flatness problems but at the same time , provides a theoretical set up to produce the primordial fluctuations which later may act as a seed for large scale structure formation in the universe .
amazingly the predicted primordial fluctuations in any inflationary model @xcite can be tested accurately through the measurement of temperature anisotropies in the cosmic microwave background radiation as recently done by planck experiment @xcite .
the construction of a viable models for inflation , which are consistent with cosmological observations like planck experiment , therefore is of utmost importance and is a subject of study of the present work . among various models for inflation ,
the models with extra dimensions have been discussed by many authors @xcite .
such models are independently considered in particle phenomenology due to their promise of resolving the well - known naturalness / fine tuning problem in connection with stabilising the mass of higgs boson against large radiative corrections @xcite . in this context , the 5-dimensional warped geometry model due to randall and sundrum ( rs ) @xcite is very successfull in offering a proper resolution to the naturalness problem without incorporating any intermediate scale other than plank / quantum gravity scale .
the radius associated with the extra dimension in this model ( known as rs modulus ) acts as a parameter in the effective 4-dimensional theory and from a cosmological point of view , such a modulus can be interpreted as a scalar field which , due to it s time evolution , may drive the scale factor of our universe before getting stabilized to a desired value . the well - known methodology to extract an effective or induce theory on a 3-brane from a 5-dimensional warped geometry model is demonstrated in @xcite where using the gauss - codazzi equation with appropriate junction condition in a two - brane warped geometry model and implementing a perturbative expansion in terms of the brane - bulk curvature ratio , the effective einstein s equation is obtained on a lower dimensional hypersurface .
this eventually results in the form of a scalar - tensor gravity theory in our brane ( known as visible brane ) @xcite .
here we try to explore the role of such a scalar - tensor theory in stabilizing the modulus of the bulk geometry as well as to generate inflation in the visible brane which is consistent with the results obtained by planck .
we show that an effective potential for the modulus field ( often called radion ) automatically emerges from the construction of the model .
the stabilisation requirements put further constrains on various parameters of the modulus / scalar potential . after deriving these constraints we study the cosmological evolution in the einstein frame in presence of such a potential and show that it gives a viable model for inflation with required number of e - folds ( to solve the horizon and flatness problem ) and also gives a primordial fluctuations which is perfectly consistent with the planck results .
the inflation is shown to end with the modulus attaining it s stable value .
hence our set up not only provides a mechanism to stabilise the modulus in the bulk but also provides a viable model for inflation which is consistent with the recent observational results .
the structure of the paper is as follows : in section ii , we briefly review the shiromizu and koyama set up @xcite for the low energy effective gravity in curved branes ; in section iii , we investigate the constraints on the potential that is necessary for moduli stabilisation ; in section iv , we describe the inflationary behaviour in our model and constrain it using the recent result from the planck experiment ; in section v , we briefly comment about the recent bicep2 results ; we end with conclusions in section vi .
we start with a configuration which contains two 3-branes embedded in a five dimensional @xmath0 symmetric ads spacetime containing a bulk cosmological constant @xmath1 .
the branes are located at two orbifold fixed points .
one has positive tension and is placed at @xmath2 in the fifth dimension ( the `` hidden brane '' ) while the other has negative tension and is placed at @xmath3 ( the `` visible brane '' ) , @xmath4 being the distance between the two branes .
next we assume the five dimensional action as @xcite : @xmath5 here @xmath6 is the five dimensional gravitational constant , @xmath7 is the bulk curvature radius which is related to the bulk cosmological constant as @xmath8 . @xmath9 and @xmath10 are the tensions of the hidden and the visible branes .
the 5d line element is taken as : @xmath11 fixing the bulk curvature scale to be @xmath12 , we define a parameter @xmath13 and assume @xmath14 which is legitimate as the scale of the cosmological evolution in brane is considerably smaller than the planck s scale namely the natural scale for the bulk curvature .
this ensures that the classical solutions of the effective einstein s equation can be trusted .
one can perturbatively expand the extrinsic curvature of the brane at fixed @xmath15 in terms of @xmath16 . at the zeroth order one retrieves the rs model with the corresponding brane tensions : @xmath17 the einstein tensor can be calculated from the given action ( see @xcite for detail derivation ) and
in the first order , one can get the effective einstein s equation on the visible brane as : @xmath18 where @xmath19 and @xmath20 . here
@xmath21 is the proper distance between the two branes and is the modulus field in the effective 4-dimensional theory . @xmath22 and @xmath23 are the respective energy momentum tensors in the hidden and the visible branes .
if we now assume that the two branes are endowed with only cosmological constants i.e @xmath24 and @xmath25 then for a spatially flat frw metric with scale factor @xmath26 , the einstein equations in the visible brane are given by : @xmath27 where @xmath28 is the hubble parameter .
the term @xmath29 can be interpreted as the potential for the scalar field @xmath30 in this model .
to study the dynamical evolution of our system , it is convenient to write the equations in the einstein frame which can be obtained using the following conformal transformations : @xmath31 we use a field redefinition @xmath32 , such that @xmath33 on solving the above equation we arrive at , @xmath34 where @xmath35 is a constant of integration .
we can also write jordan frame field @xmath30 in terms of field @xmath36 in the einstein frame as @xmath37 where @xmath38 with @xmath39 potential @xmath40 in einstein frame is now related to potential @xmath41 in jordan frame as@xcite , @xmath42 we further define two parameters @xmath43 in terms of the parameters @xmath44 and @xmath45 , @xmath40 has the form : @xmath46 moreover with these transformations , the einstein equations , simplifies to @xmath47 there are several restrictions on the form of the potential so that the model can simultaneously address the following issue : + firstly to achieve the stabilisation of the brane motion , the potential should have a minimum and field value at this minimum should be non - zero to avoid any brane collision .
secondly , the potential should also satisfy the necessary slow - roll conditions to trigger the inflation on the visible brane and finally the spectrum of the primordial fluctuations produced in this case should also be consistent with its recent measurement by planck experiment .
we address these issues one by one to ascertain the viability of the model .
first let us examine the extremum of the potential given by equation ( 11 ) .
the equation @xmath48 gives the following set of conditions : @xmath49 } \end{aligned}\ ] ] where @xmath50 .
the first and the third conditions result the corresponding @xmath30 in the jordan frame to be either infinity or zero . neither of these are acceptable as they imply infinite or zero separation between the two branes .
the second condition is not possible as @xmath45 is strictly positive .
so the acceptable @xmath51 is given by equation ( 20 ) .
also it is easy to check that @xmath52 is necessary in order to have a real @xmath51 .
now if one further calculates @xmath53 , one gets @xmath54 in order to have a minimum of the potential at @xmath55 , one further needs @xmath56 .
this , together with the condition on @xmath57 , implies that @xmath58 should be negative and @xmath59 should be positive which is similar to the randall - sundrum set up of warped geometry .
further , it is easy to show that at the minimum , @xmath60 but the cosmological observations actually is consistent with a de - sitter universe .
hence we need to add an uplifting term @xmath61 in our potential with the condition such that @xmath62 this feature is similar to the de - sitter lifting by fluxes in kklt model @xcite where the supersymmetry preserving ads minima is lifted to a de - sitter one from the energy of the background fluxes of higher form tensor fields in type iib string - based @xmath63 supergravity model in presence of brane and anti - brane .
presence of anti - brane breaks the supersymmetry giving rise to a positive definite vacuum energy by compensating the ads value of the scalar potential at the minimum .
the mechanism here is to generate some extra energy from background fluxes which show up in the scalar potential .
we should stress that adding this uplifting term does not disturb the process of radion stabilisation and the subsequent solution of the hierarchy problem .
so the final form of the potential is @xmath64 where @xmath56 , @xmath52 .
we now study the inflationary solution induced by the potential as given in equation ( 24 ) .
we define two new variables @xmath65 and @xmath66 , where @xmath67 is the reduced planck mass .
the form of the potential @xmath68 is shown in figure 1 .
as @xmath69 rolls over from the flat part near the region @xmath70 towards the minimum at the either side , inflation continues to occur .
the equations ( 15 ) and ( 16 ) now become , @xmath71 we further define slow roll parameters as@xcite , @xmath72 and @xmath73 inflation takes place when @xmath74 and @xmath75 which are also called the slow - roll conditions .
the inflation ends when any one of these conditions breaks down . in figure
( 2 ) , we show the behaviour of the slow - roll parameters for the potential shown in figure ( 1 ) .
the inflation ends at @xmath76 and hence it ends before the scalar field settles down at its minima at @xmath77 .
the total amount of inflation is measured through the number of efolding @xmath78 , defined as @xcite @xmath79 to solve the flatness problem in standard cosmology , we need at least 70 efolds of inflation . for the potential shown in figure ( 1 ) and with @xmath76 ,
the required initial value of @xmath69 is @xmath80 to achieve the desired number of efolding .
the figure ( 1 ) clearly depicts that at such a @xmath81 , the field is initially displaced slightly from the flat part of the potential .
after that it slowly rolls down and one gets enough number of efolds before it finally settles at the minimum of the potential @xmath68 .
the relevant observational quantities related to the spectrum of the primordial fluctuations are @xcite @xmath82 where @xmath4 is the tensor to scalar ratio , @xmath83 is the scalar spectral index and @xmath84 is amplitude of the scalar fluctuation . to comply with our purpose all these quantities here
must be calculated at the time of hubble exit @xmath85 .
when the scale @xmath86 leaves the hubble radius , the number of efolding before the end of inflation , @xmath87 , is given by @xmath88 so the quantities @xmath4 , @xmath89 and @xmath90 should be estimated at @xmath87 .
the value of @xmath87 depends crucially on the reheating mechanism . for reasonable inflationary models ,
one can show that @xmath91 .
in our case we take @xmath92 which is consistent with the planck s analysis . the constraints obtained by the planck s measurements are as follows @xcite : @xmath93 [ cols="^,^ " , ] + -1.2 & -0.96 & 50 & 0.955 & 0.003 & 3.09 & 0.02041 & 57.6 & -60 + -1.0 & -0.97 & 55 & 0.959 & 0.003 & 3.09 & 0.01523 & 64.7 & -66.7 + -0.8 & -0.98 & 55 & 0.961 & 0.003 & 3.08 & 0.01010 & 78.4 & -80 + -0.6 & -0.985 & 57 & 0.962 & 0.003 & 3.07 & 0.00756 & 78.8 & -79.9 +
cmb polarization is one of the most important observational signatures that can give important clues about the physics of very early universe .
the e - mode polarization was first detected by dasi in 2001 .
but the b - mode polarization which is a clear evidence of primordial gravitational waves generated during inflation has not been detected until recently .
just few months before , bicep2 experiment @xcite has announced the detection of the b - mode signal for the cmb polarization thereby confirming the existence of the primordial gravitational waves . their measured value for the tensor - to - scalar ratio @xmath4 turns out to be @xmath94 where as they rule out zero tensor fluctuation at @xmath95 confidence level .
this result brings in huge conflict with the planck results on measurement of @xmath4 which is @xmath96 .
but the contribution from galactic foregrounds to this b - mode signal has been an issue which has to be settled .
people have shown that although the bicep2 data is consistent with @xmath97 with negligible galactic foreground , it is also consistent with negligible @xmath4 with significant polarization due to dust @xcite . just recently by using the planck hfi polarization data for @xmath98 to @xmath99 ghz , adam et al . @xcite
have shown that polarization signal due to dust over the multipole @xmath100 is roughly the same as that obtained by bicep2 over this @xmath101 range .
this shows that it is entirely possible that the polarization signal bicep2 has measured is not due to the primordial gravitational wave but due to dust . as
this issue is still not settled , we have not considered the bicep2 results in our analysis .
we believe , we should wait till we get the results from ongoing joint analysis of planck and bicep2
. however we should stress that it is difficult to get high value of @xmath4 in the present model .
but added contributions coming from cosmic defects @xcite , primordial magnetic fields @xcite as well as cosmic birefringence caused by the coupling between scalar field and the cmb photons through chern - simons term @xcite can cause an enhancement to the total contribution for the tensor components .
to summarize , we study a scalar tensor theory that can explain the moduli stabilisation in the bulk geometry as well as can produce an inflationary universe in the visible brane which is consistent with the recent measurements by planck experiment @xcite
. the scalar tensor theory can naturally arise as an effective 4-dimensional theory through perturbative corrections of the brane curvature in a two - brane rs - like set up as obtained earlier by shiromizu and koyama @xcite .
the potential for the inflaton field is not an ad hoc one but emerges from the construction of the model through the effective energy momentum tensor of the modulus field .
the dynamics of radion facilitates the inflation and thus offers a natural explanation for the origin of inflation . such an inflaton field
( i.e. the radion ) needs to be stabilised and stabilisation of the radion in turn is related to the scalar field sitting at the minimum of the potential . to inflict a de - sitter character to this minimum
, we have to add an uplifting term to the potential which is similar to the de - sitter lifting by adding fluxes in the kklt set up .
we show that one gets enough e - folding in this model to solve the flatness and horizon problems .
moreover the primordial fluctuations produced by the inflaton field is consistent with the planck s measurements for @xmath89 , @xmath4 and @xmath90 .
hence the present set up not only provides a viable inflationary scenario which is consistent with the planck data but also offers a possible resolution to the modulus stabilisation mechanism concomitantly .
s.k . thanks the ugc , govt . of india for financial support .
a.a.s . acknowledges the funding from serc , dept . of science and technology , govt .
of india through the research project sr / s2/hep-43/2009 .
ssg thanks ctp , jmi for the local hospitality during his stay where part of the work was done .
a. a .starobinsky , jetp lett . * 30 * , 682 ( 1980 ) ; + v .f .
mukhanov , g. v. chibisov , soviet phys .
jetp lett .
* 33 * 532 ( 1981 ) ; + a. h. guth , phys .
d , * 23 * , 347 ( 1980 ) ; + a. a. linde , phys . lett .
b , * 108 * , 389 ( 1982 ) ; + a. .a . albrecht and p. steinhardt , phys . rev .
lett . , * 48 * , 1220 ( 1982 )
. v. f. mukhanov , h. a. feldman and r. h. bradenberger , phys .
rept . , * 215 * , 203 ( 1992 ) ; + a. riotto , arxiv : hep - th/0210162 ; + d. baumann , arxiv:0907.5424 ; + j. lidsey et al . , rev . of mod ,
phys . , * 69 * , 373 ( 1997 ) .
a. lukas , b. a. ovrut and d. waldram , phys .
d , * 60 * , 086001 ( 1999 ) ; t. nihei , phys .
, * b 465 * , 81 ( 1999 ) ; n. kaloper and a. linde , hep - th/9811141 ; d. ida , jhep , * 0009 * , 014 ( 2000 ) ; d. solomons and p. dunsby and g. ellis , gr - qc/0102016 ; e. j. copeland , a. r. liddle and j. e. lidsey , astro - ph/0006421 ; s. nojiri and s. d. odintsov , phys .
, * b 484 * , 119 ( 2000 ) ; r. maartens , d. wands , b. a. bassett and t. heard , phys .
d , * 62 * , 041301 ( 2000 ) ; s. bhattacharya , d. choudhury , d. p. jatkar and a. a. sen , hep - th/0103248 . v. a. rubakov and m. e. shaposhnikov , phys .
lett . , * b125 * , 139 ( 1983 ) ; n. arkani - hamed , s. dimopoulos and g. dvali , phys . lett . , * b 429 * , 263 ( 1998 ) ; i. antoniadis , _ et al .
_ , phys .
lett . , * 436 * , 257 ( 1998 ) ; n. arkani - hamed , s. dimopoulos and g. dvali , phys . rev .
d. , * 59 * , 086004 ( 1999 ) . | we study an effective 4-dimensional scalar - tensor field theory , originated from an underlying brane - bulk warped geometry , to explore the scenario of inflation .
it is shown that the inflaton potential naturally emerges from the radion energy - momentum tensor which in turn results into an inflationary model of the universe on the visible brane that is consistent with the recent results from the planck s experiment .
the dynamics of modulus stabilization from the inflaton rolling condition is demonstrated .
the implications of our results in the context of recent bicep2 results are also discussed . |
to date , luminous extragalactic h@xmath7o masers can be grouped into three classes : those tracing accretion disks ( e.g. ngc 4258 ) ; those in which the emission is either the result of an interaction between the radio jet and an encroaching molecular cloud or an accidental overlap , along the line - of - sight , between a warm dense molecular cloud and the radio continuum of the jet ( e.g. ngc 1052 and mrk 348 ) ; those related to prominent sites of star formation , such as the ones observed in m 33 ( the earliest known extragalactic h@xmath7o masers ) , and later also in ic 10 .
extragalactic h@xmath7o masers are preferentially detected in nearby galaxies that are bright in the infrared ( @xcite ) .
while nuclear masers are of obvious interest , non - nuclear masers are also important for a number of reasons : these sources allow us to pinpoint sites of massive star formation , to measure the velocity vectors of these regions through vlbi proper motion studies , and to determine true distances through complementary measurements of proper motion and radial velocity ( e.g. @xcite ) .
we have therefore observed the nearby spiral galaxies ic 342 and maffei 2 , both of which exhibit prominent nuclear bars and strong molecular , infrared , and radio continuum emission . in the following we report the results of our observations .
the @xmath8 line of h@xmath7o ( rest frequency : 22.23508ghz ) was observed with the 100-m telescope of the mpifr at effelsberg toward maffei 2 and ic 342 .
the full width to half power beamwidth was 40 .
the observations were carried out in a dual beam switching mode with a beam throw of 2 and a switching frequency of @xmath01hz .
the pointing accuracy was always better than 10 . on may 12 , 2001 , ic 342
was observed with the very large array ( vla ) in its b configuration .
the 103 channels used , out of the 128 observed , cover a range in velocity of @xmath08kms@xmath1 centered at 16kms@xmath1 lsr ( the velocity of the line detected at effelsberg ) .
no continuum subtraction was needed .
the data were deconvolved using the clean algorithm ( @xcite ) .
the restoring beam is @xmath9 and the rms noise per channel is @xmath010mjy , consistent with the expected thermal noise .
our single - dish observations towards maffei 2 yielded no detection , with a 5@xmath10 upper limit of 25mjy ( channel spacing : 1.05kms@xmath1 ; velocity range : 250kms@xmath1 @xmath11 @xmath12 @xmath11 + 230kms@xmath1 ; epoch : apr 3 , 2001 ; position : @xmath13 = 02@xmath14 41@xmath15 55@xmath162 , @xmath17 = + 59@xmath18 36@xmath19 11@xmath20 ) . o maser feature observed with high velocity resolution ( channel spacings after smoothing four contiguous channels are 0.26kms@xmath1 ) on * a ) * april 3 , * b ) * april 22 , * c ) * may 7 , and * d ) * june 22 .
the first spectrum has been taken at the ( 0,0 ) position , the others at ( 10,0 ) relative to the position given in footnote ` a ' of table 1 .
the dashed line indicates @xmath21 = 16kms@xmath1 .
[ flare],width=302 ] o spectra obtained toward the central region of ic 342 .
positions are offsets relative to and .
note that the spacing between two individual spectra ( 20 ) is approximately half of the size of the 100-m effelsberg telescope beam at 22ghz ( 40 ) . averaging four contiguous channels ,
the spectra have been smoothed to a spacing of 0.26 kms@xmath1 .
[ map],width=453 ] on april 2 , 2001 , we obtained the first definite detection of water vapor emission in ic 342 . during the next night
the detection was confirmed with a velocity resolution sufficient to resolve the line profile ( fig.[flare]a ) .
the detected feature lies at @xmath21 = 16kms@xmath1 and has a linewidth of @xmath00.5kms@xmath1 ; on april 2 , no other component was seen at velocities 175kms@xmath1 @xmath11 @xmath12 @xmath11 + 310kms@xmath1 ( channel spacing : 1.05kms@xmath1 ; 5@xmath10 noise level : 16mjy ) . a high line intensity ( @xmath0100mjy ) and good weather conditions allowed us to map the emitting region ( fig.[map ] ) .
further measurements were performed on april 22 and may 7 ( fig.[flare]b&[flare]c ) .
it is noticeable a fading in the peak and integrated intensities and a blue - shift in the velocity of the line by @xmath01kms@xmath1 .
unfortunately , no emission above @xmath030mjy ( 3@xmath10 level ; 0.08kms@xmath1 channel spacing ) was detected in the 22 ghz vla b - array data taken on may 12 .
this fading by at least a factor of 3 within 5 days implies a size scale of @xmath22@xmath23 cm ( @xmath22900au ) or @xmath220.5mas at a distance of 1.8mpc ( @xcite ; see also sect.4.3 ) ; the corresponding brightness temperature is @xmath2410@xmath25k .
the most recent spectrum was obtained on june 22 at effelsberg ( fig.[flare]d ) . confirming the vla result ,
no maser signal was seen above 30mjy ( 3@xmath10 ; channel spacing : 1.05kms@xmath1 ) .
fitting a synthetic gaussian to the data taken on april 3 ( see fig.[map ] ) , i.e. minimizing the sum of the difference squared between calculated and observed peak and integrated flux densities , we can obtain an accurate position of the emitting region : @xmath13 = 03@xmath14 46@xmath15 463 , @xmath17 = + 68@xmath18 0546 .
this is 13 to the west of our center position that coincides with the optical nucleus and the 2@xmath26 m peak @xcite ; @xcite .
the accuracy of the derived maser position has been estimated to be @xmath0 5 .
arising from the central region of ic 342 , but being displaced from the nucleus , the h@xmath7o maser is likely associated with a prominent star forming region . ,
width=302 ] fig .
[ mappadss ] shows an optical b - band image of the central region of ic 342 , taken from the xdss .
the maser emitting region ( white circle ) coincides with an arc - like structure extending e - s and is associated with a chain of sources that appear to be hii regions .
occasionally , prominent galactic star forming regions show narrow ( @xmath00.5 kms@xmath1 ) flaring components that are exceptionally bright .
such flares were observed in w31a and w49 ( @xcite ; @xcite ) .
the prototype of these flares is the 8kms@xmath1 super maser in orion - kl ( e.g. @xcite ) . during several months ,
the narrow highly linearly polarized maser reached flux densities in excess of 5mjy . surpassing the flux of any other velocity component by more than an order of magnitude and reaching a peak luminosity of @xmath27 @xmath0 10@xmath2l@xmath3 , the feature seems to be similar to that seen in ic 342 . while luminosity , linewidth , and flux variations
are reminiscent of the 8kms@xmath1 super maser in orion - kl , our h@xmath7o profiles show a significant velocity shift between april 22 and may 7 ( figs .
[ flare]b and c ) that has not been seen in the orion - kl flaring component . a simple model for explaining both the flare and the shift is , in our opinion , that the latter has a kinematic origin . adopting a scenario of a chance alignment of two masing clouds along the line - of - sight , motion of the foreground relative to the background cloud along the plane of the sky and velocity structure in the foreground cloud
could explain the observed data .
a velocity gradient in the foreground cloud would then first shift the line velocity ; once velocities are reached that are not matched within the background cloud , the flux density of the maser rapidly drops . assuming that the distance between these clouds is @xmath281.8mpc and that their relative velocity in the plane of the sky is at the order of 100kms@xmath1 , this implies a cloud velocity gradient of up to 1kms@xmath1/au during the time the source was monitored .
an almost identical scenario was proposed by @xcite to account for the velocity shift of the flaring water maser component at 66kms@xmath1 in w49n .
the similarity in velocity shift ( @xmath0 0.5kms@xmath1 ) and time scale ( 58 days ) of the event in w49n w.r.t the one in ic 342 , indicates a common origin .
while typical searches for extragalactic maser sources have only yielded detection rates between zero ( e.g. @xcite ) and a few percent ( e.g. @xcite ; @xcite ) , there exists one sample with detection rates @xmath610% : these are the northern ( @xmath29 @xmath6 30@xmath18 ) extragalactic iras ( infrared astronomy satellite ) point sources with 100@xmath26 m fluxes in excess of 50jy ( for a source list , see @xcite ; maffei 2 with @xmath5 @xmath0 200jy should be added to the list ) .
there is a total of 44 galaxies , two ultraluminous galaxies at intermediate distances ( ngc3690 and arp220 ) and 42 nearby sources ( @xmath12 @xmath11 3000kms@xmath1 ) . among these ,
seven ( 16% ) are known to contain h@xmath7o masers in their nuclear region .
two of these contain megamasers ( ngc 1068 and ngc 3079 ) , two are possibly nuclear kilomasers ( ngc 253 and m 51 ; for details , see sect.1 ) , and three are associated with prominent sites of star formation ( ic 10 , ic 342 , m 82 ) . among the subsample of 19 sources with 100@xmath26 m fluxes in excess of 100jy , five
were so far detected in h@xmath7o , yielding a detection rate in excess of 20% . since
few deep integrations have been obtained toward these sources , more h@xmath7o detections can be expected from this promising sample in the near future . we wish to thank nikolaus neininger for useful discussion .
we are also endebted to the operators at the 100-m telescope , and to michael rupen and the nrao analysts , for their cheerful assistance with the observations . | we report the first results of an ongoing survey at 22ghz with the 100-m effelsberg telescope to search for water maser emission in bright iras sources .
we have detected water vapor emission in ic 342 .
the maser , associated with a star forming region @xmath01015 west of the nucleus , consists of a single 0.5kms@xmath1 wide feature and reaches an isotropic luminosity of 10@xmath2l@xmath3 ( @xmath4=1.8mpc ) .
our detection raises the detection rate among northern galaxies with iras point source fluxes @xmath5 @xmath6 50jy to 16% .
.5 cm |
the standard framework for the description of continuous phase transitions is the landau - ginzburg , or @xmath1-model like , effective action @xcite . in particular , in that approach
, a phase transition is governed by the parameter @xmath2 where @xmath3 is the effective potential and @xmath4 is an order parameter connected with the phase transition . when @xmath5 , the symmetric ( non - symmetric ) phase is realized .
the value @xmath6 defines the critical point .
thus , as @xmath7 changes , one phase smoothly transforms into another .
in particular , masses of light excitations are continuous ( though non - analytic at the critical point ) functions of such parameters as coupling constants , temperature , etc . if @xmath8 , the parameter @xmath9 plays the role of @xmath7 , etc . in this paper
, we will describe a non-@xmath1-model like , though continuous , phase transition , which is relevant for the description of non - perturbative dynamics in gauge field theories . because , as will become clear below , this phase transition
is intimately connected with a nonperturbative breakdown of the conformal symmetry , we will call it the conformal phase transition ( cpt ) . in a @xmath1-model like phase transition , around the crirical point @xmath10 ( where @xmath11 is a generic notation for parameters of a theory , as the coupling constant @xmath12 , number of particle flavors @xmath13 , etc )
, an order parameter @xmath4 is @xmath14 ( @xmath15 is an ultraviolet cutoff ) , where @xmath16 has such a non - essential singularity at @xmath10 that @xmath17 as @xmath11 goes to @xmath18 both in symmetric and non - symmetric phases .
the standard form for @xmath16 is @xmath19 , @xmath20 , around @xmath21 .. by a @xmath1-model like phase transition , we understand a more general class , when fields may have anomalous dimensions @xcite . ]
the cpt is a very different continuous phase transition .
we define it as a phase transition in which an order parameter @xmath4 is given by eq .
( [ 2 ] ) where @xmath16 has such an @xmath22 singularity at @xmath10 that while @xmath23 as @xmath11 goes to @xmath18 from the side of the non - symmetric phase , @xmath24 as @xmath25 from the side of the symmetric phase ( where @xmath26 ) .
notice that since the relation ( [ 3 ] ) ensures that the order parameter @xmath27 as @xmath25 , the phase transition is continuous .
there actually exist well - known models in which such a phase transition is realized . as an example of the cpt is the phase transition at @xmath28 ( @xmath29 is the bare coupling constant ) in massless qcd with a small ,
say , @xmath30 , number of fermion flavors . in this case , the order parameter @xmath4 , describing chiral symmetry breaking , is @xmath31 and @xmath32 where @xmath33 ( @xmath34 is the first coefficient of the qcd @xmath35 function ) .
the function @xmath36 goes to zero only if @xmath37 from the side of @xmath38 .
the above example is somewhat degenerate : the critical point @xmath39 is at the edge of the physical space with @xmath40 .
a more regular example of the cpt is given by the phase transition at @xmath41 in the @xmath42-dimensional gross - neveu model : in that case both positive and negative values of @xmath43 are physical ( see sec.3 ) .
there may exist more sophisticated realizations of the cpt . as will be discussed in sec.7 , an example of the cpt
may be provided by the phase transition with respect to the number of fermion flavors @xmath13 in a @xmath44 vector - like gauge theory in @xmath45 dimensions , considered by banks and zaks long ago @xcite . in that case , unlike the phase transition at @xmath46 in qcd , the critical value @xmath47 separates two physical phases , with @xmath48 and @xmath49 .
there may exist other examples of the cpt .
also there may exist phase transitions in @xmath50-dimensional theories which `` imitate '' the cpt ( see sec.5 ) .
the main goal of this paper is to reveal the main features of the cpt ( common for its different realizations ) and to apply the conception of the cpt to concrete models .
the cpt is @xmath51 a @xmath1-model like phase transition , though it is continuous . in particular , in the cpt , one can not introduce the parameters @xmath52 governing the phase transition .
another characteristic feature of the cpt is an abrupt change of the number of light excitations as the critical point is crossed ( though the phase transition is continuous ) .
while evident in qcd and the gross - neveu model , it is realized in a more subtle way in the general case .
this feature implies a specific form of the effective action describing light excitations in theories with the cpt , which will be discussed in this paper .
the paper is organized as follows . in sec.2
we consider the properties of the spectrum of light excitations around the critical point @xmath10 in the cpt .
we show that , though the cpt is a continuous phase transition , there is an abrupt change of the spectrum of light excitations as the critical point is crossed . in sec.3
we describe the chiral phase transition in the @xmath53-dimensional ( @xmath54 ) nambu - jona - lasinio ( gross - neveu ) model .
we study the cpt in the 2-dimensional gross - neveu model .
this allows to illustrate main features of the cpt in a very clear way . in sec.4
we study the cpt in quenched qed4 , which is relevant for the study of the phase transition with respect to fermion flavors in a 4-dimensional @xmath44 gauge theory . in sec.5
the main features of the cpt are summarized and the realization of a pseudo - cpt in qed3 is considered . in sec.6
the structure of the effective action in theories with the cpt and the realization of the dynamics of the partially conserved dilatation current in these theories are discussed . in sec.7 the phase diagram with respect to the bare coupling constant @xmath55 and the number of fermion flavors @xmath13 in a @xmath44 gauge theory is considered . in particular , we suggest a modified ( as compared to that suggested in ref . [ 3 ] ) phase diagram .
possibilities of the examination of this phase diagram in lattice computer simulations are discussed . in sec.8
we summarize the main results of the paper . in the appendix
some useful relations are derived .
as was already pointed out in introduction , in the case of the @xmath1-model like phase transition , masses of light excitations are continuous functions of the parameters @xmath11 around the critical point @xmath10 ( though they are non - analytic at @xmath10 ) .
let us show that the situation in the case of the cpt is different : there is an abrupt change of the spectrum of light excitations , as the critical point @xmath10 is crossed .
let us start from a particular , and important , case of the cpt connected with dynamical chiral symmetry breaking . in this case , in the non - symmetric phase , amongst light ( with masses much less than cutoff @xmath15 ) excitations
, there are massless nambu - goldstone ( ng ) bosons @xmath56 , their chiral partners , @xmath1 bosons , and light ( with @xmath57 ) fermions .
the masses of @xmath1 and fermions are given by scaling relations : @xmath58 where @xmath59 and @xmath60 are some positive constants , and @xmath16 is a universal scaling function .
because of the assumption ( [ 3 ] ) , @xmath61 and @xmath62 are indeed much less than @xmath63 , when @xmath11 is near @xmath18 from the side of the non - symmetric phase .
now , are there light @xmath56 and @xmath1 resonances in the symmetric phase , with @xmath64 ?
since , as was assumed , @xmath24 as @xmath25 in that phase , one should expect that there are no light resonances .
let us show that this is indeed the case .
one might think that in the symmetric phase the mass relation for @xmath56 and @xmath1 is yielded by the analytic continuation of the relation ( [ 5 ] ) for @xmath61 .
however this is not the case .
the point is that while in the non - symmetric phase , @xmath56 and @xmath1 bosons are described by bethe - salpeter ( bs ) equations with a non - zero fermion mass , in the symmetric phase they are described by bs equations with @xmath65 . because of that , bs equations ( and , more generally , all the schwinger - dyson equations for green s functions ) in the symmetric phase are not yielded by an analytic continuation of the equations in the non - symmetric phase . to overcome this obstacles
, we shall use the following trick . in the non - symmetric phase , besides the stable solution with @xmath66 , there is also an unstable solution with @xmath67 . in that solution ,
@xmath56 and @xmath1 bosons are tachyons : @xmath68 .
since the replacement of @xmath69 by @xmath64 ( at fixed values of the parameters @xmath11 ) does not change the ultraviolet properties of the theory , the scaling relation for the tachyon masses has the same form as that in eqs .
( [ 5 ] ) and ( [ 6 ] ) : @xmath70 since now @xmath67 , the bs equations for tachyons have the same form as the bs equations for @xmath56 and @xmath1 in the symmetric phase ; the difference between these equations is only in the values of @xmath11 ( for convenience , we shall assume that @xmath71 in the non - symmetric ( symmetric ) phase ) .
then , in the symmetric phase , @xmath72 with @xmath73 and @xmath74 from eq .
( [ 7 ] ) .
notice that because in the symmetric phase @xmath56 and @xmath1 bosons decay to massless fermions and antifermions , @xmath75 and @xmath61 are complex , i.e. @xmath56 and @xmath1 are now resonances , if they exist at all .
since , by definition , in the cpt , @xmath24 as @xmath76 , we conclude from eq . ( [ 8 ] ) that there are no light resonances near the critical point from the side of the symmetric phase : @xmath77 as @xmath76 .
so far , for concreteness , we have considered the case of dynamical chiral symmetry breaking .
but it is clear that ( with minor modifications ) this consideration can be extended to the general case of the cpt connected with spontaneous breakdown of other symmetries .
notice also that the relation in eq .
( [ 8 ] ) can be useful for general phase transitions and not just for the cpt .
the point is that the scaling function @xmath16 can be determined from the gap equation for the order parameter ( @xmath62 , in the case of chiral symmetry ) which is usually much simpler than the bs equation for massive composites .
for example , an abrupt change of the spectrum at the critical point @xmath10 have been revealed in some models : in quenched qed4 @xcite and qed3 @xcite .
this conclusion was based on an analysis of the effective action @xcite and the bs equation @xcite , considered in a rather crude approximation . on the other hand , since the determination of the scaling function @xmath16 in these models is a much simpler task , this conclusion can be firmly established in the present approach ( see secs.4 and 5 ) . thus the present consideration yields a simple and general criterion of such a peculiar behavior of the spectrum of light excitations .
it is clear that the abrupt change of the spectrum discussed above implies rather peculiar properties of the effective action for light excitations at the critical point .
below we shall consider this problem in more detail .
we shall also reveal an intimate connection between this point and the essential difference of the character of the breakdown of the conformal symmetry in different phases of theories with the cpt .
in this section we consider the dynamics in the @xmath53-dimensional @xmath78 nambu - jona - lasinio ( gross - neveu ) model and , in particular , describe the cpt in the gross - neveu ( gn ) model at @xmath79 .
this will allow to illustrate main features of the cpt in a very clear way .
the lagrangian density of the @xmath53-dimensional gn model , with the @xmath80 chiral symmetry , is @xmath81 + \frac{g}{2 } \left [ ( \bar{\psi } \psi)^2 + ( \bar{\psi } i \gamma_5 \psi)^2 \right ] , \label{9}\ ] ] where @xmath82 , and the fermion field carries an additional `` color '' index @xmath83 .
the theory is equivalent to the theory with the lagrangian density @xmath84 -\bar{\psi } ( \sigma + i \gamma_5 \pi ) \psi - \frac{1}{2 g } ( \sigma^2 + \pi^2 ) .
\label{10}\ ] ] the euler - lagrange equations for the auxiliary fields @xmath85 and @xmath56 take the form of constraints : @xmath86 and the lagrangian density ( [ 10 ] ) reproduces eq . ( [ 9 ] ) upon application of the constraints ( [ 11 ] ) . the effective action for the composite fields @xmath85 and @xmath56 is obtained by integrating over fermions in the path integral : @xmath87 - \frac{1}{2 g } \int d^d x ( \sigma^2 + \pi^2 ) .
\label{12}\ ] ] the low energy dynamics are described by the path integral ( with the integrand @xmath88 ) over the fields @xmath1 and @xmath56 .
as @xmath89 , the path integral is dominated by the stationary points of the action : @xmath90 let us look at the effective potential in this theory .
it is @xcite @xmath91 + o(\frac{\rho^4}{\lambda^4 } ) , \label{13}\ ] ] where @xmath92 the dimensionless coupling constant @xmath93 is @xmath94 and the critical coupling @xmath95 . at @xmath96 , one finds that @xmath97 the sign of @xmath7 defines two different phases : @xmath98 @xmath99 corresponds to the symmetric phase and @xmath100 @xmath101 corresponds to the phase with spontaneous chiral symmetry breaking , @xmath102 .
the value @xmath103 defines the critical point @xmath104 .
therefore at @xmath96 , a @xmath85-model like phase transition is realized .
however the case @xmath79 is special : now @xmath105 and @xmath106 as @xmath107 in this case the effective potential is the well - known potential of the gross - neveu model @xcite : @xmath108 . \label{16}\ ] ] the parameter @xmath7 is now : @xmath109 therefore , in this model , one can not use @xmath7 as a parameter governing the continuous phase transition at @xmath110 : the phase transition is not a @xmath1-model like phase transition in this case . indeed , as follows from eq .
( [ 16 ] ) , the order parameter , which is a solution to the gap equation @xmath111 , is @xmath112 in this model . the function @xmath16 , defined in eq .
( [ 2 ] ) , is now @xmath113 , i.e. , @xmath114 , and therefore the cpt takes place in this model at @xmath115 : @xmath116 goes to zero only if @xmath117 from the side of the non - symmetric phase .
let us discuss this point in more detail . at @xmath118 , the spectrum of the @xmath1 and @xmath56 excitations in the symmetric solution , with @xmath119 ,
is defined by the following equation ( in leading order in @xmath120 ) @xcite : @xmath121 therefore at @xmath96 , there are tachyons with @xmath122 at @xmath123 , and at @xmath124 there are `` resonances '' with @xmath125 which agrees with eq .
( [ 8 ] ) .. notice that , as follows from eq .
( [ 19 ] ) , narrow resonances occur near @xmath126 : @xmath127 . ]
( [ 21 ] ) implies that the limit @xmath128 is special .
one finds from eq .
( [ 19 ] ) that at @xmath79 @xmath129 at @xmath130 , and @xmath131 at @xmath132 , i. e. , in agreement with the main feature of the cpt , there are no light resonances in the symmetric phase at @xmath79
. the effective potential ( [ 16 ] ) can be rewritten as @xmath133 \label{24}\ ] ] ( with @xmath134 given by eq .
( [ 18 ] ) ) in the non - symmetric phase .
that is , in this phase @xmath135 is finite in the continuum limit @xmath136 after the renormalization of the coupling constant , @xmath137 ( see eq .
( [ 18 ] ) ) .
but what is the form of the effective potential in the continuum limit in the symmetric phase , with @xmath138 ? as eq .
( [ 16 ] ) implies , it is infinite as @xmath136 : indeed at @xmath138 , there is no way to cancel the logarithmic divergence in @xmath3 .
it is unlike the case with @xmath139 : in that case , using eq .
( [ 15 ] ) , the potential ( [ 13 ] ) can be put in a @xmath1-model like form : @xmath140 however , since @xmath141 at @xmath79 , the @xmath1-model like form for the potential is not available in the gross - neveu model .
what are physical reasons of such a peculiar behavior of the effective potential at @xmath79 ? unlike the case with @xmath96 , at @xmath79 the lagrangian density ( [ 9 ] ) defines a conformal theory in the classical limit . by using the conventional approach , one can derive the following equation for the conformal anomaly in this model ( see the appendix ) : @xmath142 \label{27}\ ] ] where @xmath143 is the dilatation current , @xmath144 is the energy - momentum tensor , and the @xmath35 function @xmath145 .
it is @xmath146 both in the non - symmetric and symmetric phases .
while the non - symmetric phase corresponds to asymptotically free dynamics , the symmetric phase ( with @xmath147 ) defines infrared free dynamics : as @xmath136 , we are led to a free theory of massless fermions , which is of course conformal invariant . on the other hand , in the non - symmetric phase the conformal symmetry
is broken , even as @xmath136 .
in particular , eq . ( [ 24 ] ) implies that @xmath148 in leading order in @xmath120 in that phase .
the physics underlying this difference between the two phases is clear : while negative @xmath93 correspond to repulsive interactions between fermions , attractive interactions at positive @xmath93 lead to the formation of bound states , thus breaking the conformal symmetry .
notice the following interesting point .
as follows from eq .
( [ 26 ] ) , at @xmath149 the conformal symmetry is broken by a relevant ( superrenormalized ) mass operator : its dynamical dimension is @xmath150 at all @xmath151 . on the other hand , at @xmath79 the symmetry
is broken by a marginal ( renormalized ) operator with the dynamical dimension @xmath150 .
this point is reflected in that while at @xmath79 the expression for the order parameter @xmath134 has an essential singularity at the critical point @xmath152 , at @xmath96 , the singularity at @xmath104 in @xmath134 is not essential : as follows from eq .
( [ 13 ] ) , the solution to the gap equation @xmath153 is @xmath154 in that case . as is known , the essential singularity implies the absence of of fine tuning for bare parameters .
this is another reason why the cpt is so interesting .
thus the cpt , in accordance with its name , describes the two essentially different realizations of the conformal symmetry in the symmetric and non - symmetric phases .
if one adds a fermion mass term , @xmath155 , in the 2-dimensional gn model , the conformal and chiral symmetries will be of course broken in both phases .
however , there remains an essential trace of the cpt also in this case : an abrupt change of the spectrum of light excitations still takes place .
while now in the subcritical @xmath156 phase repulsive interactions between massive fermions are take place ( and there are no light resonances there ) , in the supercritical @xmath157 phase the pcac dynamics , describing interactions between fermions and light @xmath56 and @xmath1 bosons , is realized
. still defines the fermion dynamical mass , and the would - be ng boson @xmath56 transforms into a bkt gapless excitation @xcite . ] besides the point that in the 2-dimensional gn model both subcritical and supercritical phases are physical , this picture is similar to that in qcd .
it is hardly surprising : in both models the dynamics in the supercritical phases are asymptotically free .
we will however argue that the main features of the cpt found in the gn model will retain valid ( with appropriate minor modifications ) in the general case .
in this section we shall describe the main features of the cpt in quenched qed4 .
the dynamics in this model is relevant for some scenarios of dynamical electroweak symmetry breaking and has been intensively discussed in the literature ( for a review see ref . [ 10 ] ) . in the present paper
the emphasis of the discussion will be on the points relevant for the general cpt in gauge theories .
we shall consider the ladder ( rainbow ) approximation in massless qed4 .
since the contribution of fermion loops is omitted , the perturbative @xmath35-function equals zero in this approximation .
however , as is well known @xcite , beyond the critical value @xmath158 , there are nonperturbative divergences which break the conformal symmetry in the model . moreover , since at @xmath159 ,
the anomalous dimension @xmath160 of the chiral operators @xmath161 and @xmath162 is @xmath163 @xcite , the four - fermion operators @xmath164 and @xmath165 become ( marginally ) relevant : their dynamical dimension @xmath166 is @xmath167 , where @xmath168 is their canonical dimension . therefore , it is appropriate to include these four - fermion operators in the qed action .
this leads to the gauged nambu - jona - lasinio model @xcite : @xmath169 + \frac{g}{2 } \left [ ( \bar\psi \psi)^2 + ( \bar\psi i \gamma_5 \psi)^2 \right ] , \label{29}\ ] ] where @xmath170 ( for simplicity , we consider the chiral symmetry @xmath171 ) . in this model ,
the gauge interactions are treated in the ladder approximation and the four - fermion interactions are treated in the hartree - fock ( mean field ) approximation .
since the coupling constant @xmath172 is dimensional , one may think that the four - fermion interactions in eq .
( [ 29 ] ) explicitly break the conformal symmetry .
the real situation is however more subtle . in fig . 1
, we show the critical line in this model @xcite , dividing the symmetric phase , with the unbroken @xmath171 , and the phase with the spontaneously broken chiral symmetry @xmath173 ) .
each point of the critical line corresponds to a continuous phase transition .
we distinguish two parts of the critical line : @xmath174 ^ 2 , \ : \ : \alpha_c = \frac{\pi}{3 } , \label{30}\ ] ] at @xmath175 , and @xmath176 at @xmath177
. the anomalous dimension @xmath160 of the operators @xmath178 and @xmath179 along the critical line is @xcite @xmath180 in this approximation , the anomalous dimension of the four - fermion operator @xmath181 $ ] equals @xmath182 .
therefore while this operator indeed breaks the conformal symmetry along the part ( [ 30 ] ) of the critical line , it is a marginal ( scale invariant ) operator along the part of the critical line with @xmath183 : its dynamical dimension is @xmath184 there .
thus the part ( [ 31 ] ) of the critical line with @xmath159 is special . in this case
the symmetric phase is not only chiral invariant but also conformal invariant . on the other hand , in the non - symmetric phase , both these symmetries are broken : while the chiral symmetry is broken spontaneously , the conformal symmetry is broken explicitly ( see below ) @xcite . unlike the case of the njl model , where the lagrangian density ( [ 10 ] ) with the auxiliary fields @xmath1 and @xmath56 was used to derive the effective action @xmath185 $ ] , now we will derive another effective action : generating functional for proper vertices of the local composite operators @xmath186 and @xmath187 .
the point is that the trick , used in eq .
( [ 10 ] ) , to introduce the fields @xmath1 and @xmath56 does not work in pure qed ( @xmath188 ) .
it is also unclear how well this effective action describes the collective excitations @xmath1 and @xmath56 in the case of weakly coupling four - fermion interactions ( when qed forces dominate ) .
we shall discuss the connection between these two actions below ( the effective action based on using the auxiliary fields @xmath1 and @xmath56 was considered in this model in refs .
@xcite ) . let us describe the effective action ( generating functional for proper vertices ) for the local composite operators @xmath178 and @xmath179 in the gauged njl model .
this effective action was derived in ref . [ 4 ] . in the present paper
, we will describe in more detail those features of the action which are relevant for understanding the nature of the cpt .
the effective action is constructed in the standard way .
first , one introduces a generating functional for green s functions of the operators @xmath189 : @xmath190\ ] ] @xmath191 , \label{33}\ ] ] where the @xmath192 integration is functional , @xmath193 is the source for @xmath194 , and @xmath195 is the lagrangian density ( [ 29 ] ) ( the symbol @xmath196 represents all fields of the model ) .
the effective action for the operators @xmath194 is a legendre transform of the functional @xmath197 : @xmath198 where @xmath199 ( we use `` tilde '' here in order to distinguish this effective action from that for the auxiliary fields @xmath1 and @xmath56 ) .
> from eqs .
( [ 33 ] ) and ( [ 34 ] ) one finds that the following relations are satisfied : @xmath200 the effective action @xmath201 can be expanded in powers of derivatives of the fields @xmath202 : @xmath203 \label{37}\ ] ] where @xmath204 is the effective potential .
the calculation of the effective potential is reduced to finding the legendre transform of the functional @xmath205 with the sources @xmath206 independent of coordinates @xmath207 .
because of chiral symmetry , w depends on the chiral invariant @xmath208 . therefore to determine the form of w
, it is sufficient to consider the source term with @xmath209 . then , owing to the relation ( [ 35 ] ) , one finds that @xmath210 where @xmath211 and @xmath212 . here
@xmath213 is the fermion mass function ( the fermion propagator is @xmath214 ) .
since the functional @xmath215 corresponds to the source term with @xmath216 , the condensate @xmath217 is related to the gauged njl model with the bare mass @xmath218 .
now we need to know the following information concerning the fermion propagator in rainbow qed4 ( for a review see ref .
@xcite ) . in the landau gauge , the functions @xmath219 and @xmath213 satisfies the equations @xmath220
@xmath221 @xmath222 \sigma(q^2 ) , \label{41}\ ] ] where @xmath223 , @xmath15 is an ultraviolet cutoff and @xmath224 is the bare mass of fermions ( equation ( [ 41 ] ) is written in euclidean space ) .
differentiating eq .
( [ 41 ] ) with respect to @xmath225 , one finds that @xmath213 satisfies the differential equation @xmath226 + \frac{3\alpha}{4 \pi } \frac{\sigma(p^2)}{p^2 + \sigma^2(p^2 ) } = 0 \label{42}\ ] ] and two boundary conditions : @xmath227 \right| _ { p^2=\lambda^2 } , \label{43}\ ] ] @xmath228 = 0 . \label{44}\ ] ] differentiating eq .
( [ 41 ] ) with respect to @xmath225 at @xmath229 , we find that the chiral condensate @xmath230 is @xmath231 @xmath232 as was mentioned above , eqs .
( [ 40 ] ) and ( [ 41 ] ) are written in the landau gauge .
this gauge is preferable in the rainbow approximation from the viewpoint of satisfying the ward - takahashi identities @xcite .
the transformation to other gauges changes the vertex , i.e. it leads to an approximation beyond the rainbow one .
however , the main results of the analysis remain essentially the same @xcite .
the function @xmath213 , satisfying eqs .
( [ 42 ] ) - ( [ 44 ] ) , has the following ultraviolet asymptotics ( at @xmath233 ) : @xmath234 \label{46}\ ] ] at @xmath235 , @xmath236 at @xmath159 , and @xmath237 \label{48}\ ] ] at @xmath238 . here
@xmath239 and @xmath240 are known functions of @xmath12 .
and @xmath241 are @xmath242^{1/2}\ ] ] and @xmath243\ ] ] at @xmath244 . at @xmath245 ,
@xmath246 is replaced by @xmath247 in these expressions . ]
we will calculate the effective potential along the critical line .
our aim is to show that while along the part of the critical line ( [ 30 ] ) , with @xmath248 , the potential has the conventional , @xmath1-model like , form , its form is rather unusual along the critical line with @xmath159 , where the cpt occurs .
we begin by considering the potential along the part of the critical line ( [ 30 ] ) . from eqs .
( [ 45 ] ) and ( [ 46 ] ) , we find the chiral condensate in this case : @xmath249 .
\label{49}\ ] ] the function @xmath250 is determined from eq .
( [ 43 ] ) with @xmath251 . substituting the condensate ( [ 49 ] ) in equation ( [ 38 ] ) , we find : @xmath252 @xmath253 \label{50}\ ] ] where @xmath254 @xmath255 and @xmath256 has to be considered here as a function of @xmath257 .
realizing the legendre transform of @xmath258 , we find the effective potential : @xmath259 @xmath260 @xmath261 , \label{52}\ ] ] where now @xmath256 has to be considered as a function of the condensate @xmath262 a solution @xmath263 to the gap equation @xmath264 ( or @xmath265 ) defines the dynamical mass @xmath266 as a function of @xmath267 and g : @xmath268^{1/\omega}. \label{54}\ ] ] the limit @xmath269 defines the critical line ( [ 30 ] ) : @xmath270 .
let us show that the potential ( [ 52 ] ) can be rewritten as a conventional , @xmath1-model like , potential .
it is convenient to define the renormalized fields @xmath271 and @xmath272 as @xmath273 @xmath274 where the renormalization constant is @xmath275 ( see eq .
( [ 53 ] ) ; notice that the renormalized composite fields @xmath276 and @xmath277 are defined in such a way as to have canonical dimension equal to 1 ) . then we find that @xmath278 at @xmath279 and @xmath280 .
we will express the potential ( [ 52 ] ) through the chiral invariant @xmath281 ( where @xmath282 ) and the mass parameter @xmath283 it is easy to check that , along the critical line ( [ 30 ] ) , the parameter @xmath284 is @xmath285 one can see that @xmath286 and @xmath287 from the side of the symmetric and non - symmetric phases , respectively . as @xmath288 , with @xmath284
being fixed , the potential ( [ 52 ] ) is @xmath289 thus , as was promised , we derived a @xmath1-model like potential : the sign of @xmath284 defines two different phases , and the value @xmath290 corresponds to the critical line with @xmath291 .
notice that the parameter @xmath284 appears in @xmath292 as a result of the dynamical transmutation : in the continuum limit , @xmath288 , the dimensional parameter @xmath284 replaces the dimensionless coupling constant @xmath93 .
note also that , because the anomalous dimension @xmath160 of @xmath178 and @xmath179 is @xmath293 ( see eq .
( [ 32 ] ) ) , the dynamical dimension of these operators , and therefore of @xmath294 , is @xmath295 .
therefore the second term in the potential ( [ 60 ] ) has the dynamical dimension equal to 4 , i.e. it preserves the conformal symmetry . the first , massive , term in the potential breaks this symmetry .
one can check that , taking @xmath296 , the potential in the non - symmetric phase can be rewritten as @xmath297 , \label{61}\ ] ] where @xmath298 . as shown in refs .
@xcite , the kinetic term and terms with a larger number of derivatives in the effective action are also finite in the continuum limit .
a heuristic explanation of this fact is simple : the most severe ultraviolet divergences always occur in the effective potential ; therefore the finiteness of the potential implies the finiteness of other terms in the action .
this implies that around the critical line ( [ 30 ] ) , both in the symmetric and non - symmetric phase , the light excitations include , besides fermions and photon , composite @xmath1 and @xmath56 particles .
let us now turn to the part of the critical line with @xmath299 , where the cpt takes place , and show that the character of the phase transition is essentially different there .
we begin by considering the critical point @xmath300 . calculating the potential in the same way as the potential ( [ 52 ] ) for @xmath244
, we find that at @xmath159 , around @xmath301 , it is @xmath302 \label{62}\ ] ] where @xmath303 .
the gap equation @xmath304 yields the dynamical mass at @xmath175 : @xmath305 at @xmath159 , the relation between @xmath306 and @xmath256 is @xmath307 ( compare with eq .
( [ 57 ] ) ; in this case the renormalization constant is @xmath308 ) . taking the continuum limit @xmath288 in eq .
( [ 62 ] ) ( with @xmath309 being fixed ) we come to the expression for the potential in the non - symmetric phase : @xmath310 but what is the form of the potential ( in the continuum limit ) from the side of the symmetric phase ?
let us show that there the potential @xmath292 goes to infinity as @xmath288 . in the @xmath288 limit @xcite .
this may define a different continuum theory than that discussed in this paper .
] indeed eq .
( [ 62 ] ) implies that to get a finite @xmath292 as @xmath311 , the coupling constant @xmath93 must have the behavior @xmath312 i.e. it _ must _ go to the critical value @xmath313 from the side of the non - symmetric phase . in the symmetric phase the potential @xmath314 as @xmath288 .
what is the physical meaning of this result ? as in the case of the gross - neveu model , it implies that there are no composite light resonances in the symmetric phase at @xmath315 .
indeed , the relations ( [ 8 ] ) and ( [ 63 ] ) imply that in the symmetric phase @xmath316 this conclusion correlates with the point that , unlike the case with the critical line with @xmath244 , the parameter @xmath284 is not well defined at @xmath317 : it is @xmath318 and @xmath319 in the symmetric and non - symmetric phase , respectively ( see eq .
( [ 59 ] ) ) .
thus the cpt takes place at @xmath320 .
let us show that a similar situation takes place along the whole critical line with @xmath159 , where the cpt takes place . around the critical line with @xmath159 , the condensate @xmath321 is @xmath322 at @xmath244 , ( the symmetric phase ) , and @xmath323 at @xmath324 ( the non - symmetric phase ) . here
@xmath325 and @xmath326 , and eqs . (
[ 45 ] ) , ( [ 46 ] ) and ( [ 48 ] ) were used in the derivation of expressions ( [ 68 ] ) and ( [ 69 ] ) .
now the effective potential can be derived in the same way as before .
it is @xmath327 @xmath328 \simeq - \frac{\tilde{a}^2 \sigma_0 ^ 4}{24 \pi \alpha } \left [ \left ( 4g-1 \right ) l^2 - 8gl + 4 g + \frac{3}{2 } \right ] \label{70}\ ] ] at @xmath244 and @xmath329 @xmath330 , and @xmath331 @xmath332 \label{71}\ ] ] at @xmath333 and @xmath334 @xmath330 .
notice that at @xmath335 , the expression ( [ 70 ] ) for @xmath292 coincides with the expression ( [ 62 ] ) . therefore we conclude that in the symmetric phase , along the whole critical line with @xmath299 , the collective @xmath1 and @xmath56 excitations decouple from the infrared dynamics . in this phase ,
the conformal dynamics of massless fermions and photons is realized .
let us now consider the potential ( [ 71 ] ) in the non - symmetric phase . the gap equation @xmath336 yields the following solutions for @xmath256 at @xmath337 : @xmath338 @xmath339 where @xmath340 .
one can check that while all the solutions ( [ 72 ] ) correspond to minima of @xmath292 , the solutions ( [ 73 ] ) correspond to maxima of the potential .
actually , only the global minimum , corresponding to @xmath341 , defines the stable vacuum .
therefore the dynamical mass is @xmath342 the expression ( [ 74 ] ) is valid at @xmath343 . at @xmath344 , we find from eq .
( [ 71 ] ) that @xmath345 in the continuum limit @xmath288 , with @xmath309 being fixed ( i.e. , lim @xmath346 as @xmath311 and @xmath347 ) , we find the expression for @xmath292 : @xmath348 .
\label{76}\ ] ] here @xmath256 is considered as a function of a renormalized field @xmath349 which is defined in the following way .
one finds from eq .
( [ 69 ] ) that , in the continuum limit , the condensate is : @xmath350 therefore it is appropriate to define the renormalized fields @xmath351 and @xmath352 with the renormalization constant @xmath353 then one finds that @xmath354 .
\label{79}\ ] ] the function @xmath355 in eq .
( [ 76 ] ) is defined from eq .
( [ 79 ] ) , with @xmath356 replaced by @xmath357 . the potential ( [ 76 ] )
describes composite @xmath1 and @xmath56 particles in the non - symmetric phase .
thus there is an abrupt change of the number of light excitations and the character of the dynamics , as the critical line is crossed , along the whole critical line with @xmath159 , where the cpt takes place .
it is instructive to compare the effective action ( generating functional for proper vertices of the operators @xmath178 and @xmath179 ) considered here with the effective action @xmath358 for auxiliary fields , whose derivation in the gauged njl model was considered in refs .
the action @xmath359 is derived by rewriting the lagrangian density ( [ 29 ] ) as @xmath360 - \bar\psi ( \sigma + i \gamma_5 \pi ) \psi - \frac{1}{2 g } ( \sigma^2 + \pi^2 ) , \label{80}\ ] ] and then integrating out both the fermion and photon fields . in the mean field approximation , the path integral over @xmath1 and @xmath56 is dominated by the stationary points of @xmath358 : @xmath361 .
it is clear , however , that the trick of introducing the fields @xmath1 and @xmath56 in eq .
( [ 80 ] ) can not be applied to pure qed @xmath362 .
it is also unclear how well is the mean field approximation around the part of the critical line with @xmath159 , where the qed dynamics dominate . comparing the expression for the potential @xmath292 with that for the potential @xmath3 , derived in refs .
@xcite , one finds that while they coincide around the part ( [ 30 ] ) of the critical line , with @xmath363 and @xmath175 ( where the four - fermion interactions dominate ) , they are different around the critical line with @xmath299 , where the cpt takes place .
in particular , the counterpart of the expressions ( [ 62 ] ) and ( [ 76 ] ) for @xmath292 are @xcite @xmath364 , \label{81}\ ] ] and @xcite @xmath365 , \label{82}\ ] ] respectively . unlike the expressions ( [ 62 ] ) and ( [ 76 ] ) for @xmath292 , @xmath3 is singular at @xmath366 ( pure qed ) .
this feature of @xmath3 reflects the point that at @xmath367 , @xmath177 , when qed interactions dominate , the mean field approximation is not good enough and one has to consider quantum fluctuations of the fields @xmath1 and @xmath56 . on the other hand
, the generating functional @xmath368 adequately describes the dynamics along the whole critical line . a similarity between the dynamics in quenched qed4 and @xmath53-dimensional gn model considered in sec
. 3 is evident . at @xmath369 @xmath370 a @xmath1-model like phase transition
is realized in quenched qed4 ( gn model ) ; at @xmath159 @xmath371 the cpt takes place in these models .
however , there is an essential difference between the cpt phase transitions in these two models . while in the gn model , the symmetric phase , with @xmath138 , is infrared free , the symmetric phase in quenched qed is a coulomb phase , describing interactions between massless fermions and photons . as was indicated in sec .
3 , a marginal operator is responsible for the breakdown of the conformal symmetry in the non - symmetric phase in the 2-dimensional gn model ( see eq .
( [ 27 ] ) ) .
this leads to an essential singularity in the expression for the order parameter @xmath134 ( [ 18 ] ) .
this in turn cures fine tuning problem which takes place at @xmath96 , where relevant ( superrenormalized ) operators break the conformal symmetry .
a similar situation takes place in quenched qed4 . while at @xmath244 , the ( relevant ) mass operator breaks the conformal symmetry ( see eq .
( [ 60 ] ) ) , at @xmath159 , it is broken ( in non - symmetric ) phase by a marginal operator . actually ,
as shown in the appendix , the equation for the conformal anomaly has the following form at @xmath372 : @xmath373 \nonumber \\ & = & \lim_{\lambda \to \infty , \alpha \to \alpha_c } \lim \frac{\beta ( \alpha)}{4\alpha } f_{\mu \nu}f^{\mu \nu } , \label{83}\end{aligned}\ ] ] where @xmath374 is determined from eq .
( [ 63 ] ) , and @xmath375 is determined from eq .
( [ 75 ] ) .
as was pointed above , at @xmath376 the dynamical dimension of the operator @xmath377 $ ] is @xmath378 and , therefore , it is indeed a marginal operator ( the operator @xmath379 in eq .
( [ 83 ] ) is also of course marginal ) . at @xmath159 and @xmath380 ,
the equation for the conformal anomaly is ( see the appendix ) @xmath381 where @xmath382 is determined from eq .
( [ 74 ] ) . here , again , @xmath383 is a marginal operator . in the next section , we shall summarize the main features of the cpt .
we shall also discuss the phase transition in qed3 .
now we are ready to summarize the main features of the cpt .
there is an abrupt change of the spectrum of light excitations , as the critical point @xmath21 is crossed , in the cpt . as was shown in sec .
2 , this property is general and reflects the presence of an essential singularity at @xmath10 in the scaling function @xmath16 .
the cpt is ( though continuous ) a non-@xmath1-model like phase transition .
this implies a specific form of the effective action , in particular , the effective potential , for the light excitation near @xmath10 .
while the potential does not exist in the continuum limit in the symmetric phase , it has infrared singularities at @xmath384 in the non - symmetric phase ( @xmath385 is a generic notation for fields describing the light excitations ) . as a result , unlike the @xmath1-model like phase transition , one can not introduce parameters @xmath386 which would govern the phase transition : all of them are equal either to zero or to infinity .
the infrared singularities in the effective potential imply the presence of long range interactions as @xmath387 .
this is turn connected with an important role of the conformal symmetry in the cpt . in the examples considered in sec . 3 and 4 , while the symmetric phase is conformal invariant , there is a conformal anomaly in the non - symmetric phase : the conformal symmetry is broken by a marginal operator .
the latter allows to get rid of the fine tuning problem in such a dynamics .
we shall return to the problem of the effective action in the cpt in the next section . because of the abrupt changing the spectrum of light excitations at @xmath10 , the very notion of the universality class for the dynamics with the cpt seems rather delicate .
for example , in both gn model and qcd , at the critical point @xmath388 and @xmath46 , respectively ) , and at finite cutoff @xmath15 , the theories are free and their infrared dynamics are very different from the infrared dynamics in the non - symmetric phases of these theories ( at @xmath389 and @xmath390 , respectively ) .
this is a common feature of the cpt : around the critical point , the infrared dynamics in the symmetric and non - symmetric phases are very different .
however , in the non - symmetric phase , the hypothesis of universality has to be applied to the region of momenta @xmath391 satisfying @xmath392 , where @xmath134 is an order parameter . in that region , critical indices ( anomalous dimensions ) of both elementary and composite local operators in near - critical regions of symmetric and non - symmetric phases are nearly the same : the critical indices are continuous functions of @xmath11 around @xmath10 . ) in green s functions in that phase . ] on the other hand , since the infrared dynamics ( with @xmath393 and @xmath394 ) abruptly changes as the critical point @xmath10 is crossed , the low energy effective actions in the symmetric and non - symmetric phases are different .
if there is a perturbative running of the coupling in the symmetric phase , it will lead to perturbative violation of the conformal symmetry ( see sec .
however , this does not change the most characteristic point of the cpt : the abrupt change of the spectrum of light excitations at @xmath10 discussed above .
the reason is that there is an additional , nonperturbative , source of the breakdown of the conformal symmetry in the non - symmetric phase , which provides the creation of light composites .
the conception of the cpt , in a slightly modified form , can be also useful for a different type of dynamics . as an example , let us consider qed3 with massless four - component fermions @xcite .
it is a superrenormalizable theory where ultraviolet dynamics plays rather a minor role . as was shown in refs .
@xcite , when the number of fermion flavors @xmath13 is less than @xmath395 , with @xmath396 , there is dynamical breakdown of the flavor @xmath397 symmetry in the model , and fermions acquire a dynamical mass : @xcite . for a recent discussion supporting the relation ( [ 85 ] ) , see ref . @xcite . ]
@xmath398 , \label{85}\ ] ] where the coupling constant @xmath399 is dimensional in qed3 .
though this expression resembles the expression ( [ 74 ] ) for the dynamical mass in quenched qed4 , where @xmath15 plays the role of @xmath400 and @xmath12 plays the role of @xmath13 , the phase transition at @xmath401 is , strictly speaking , not the cpt .
indeed , because of superrenormalizability of qed3 , the ultraviolet cutoff @xmath15 is irrelevant for the dynamics leading to relation ( [ 85 ] ) .
also , since @xmath400 is dimensional , the conformal symmetry is broken in both symmetric @xmath402 and non - symmetric @xmath403 phases .
nevertheless , the consideration of the spectrum of light ( with @xmath404 ) excitations in this model can be done along the lines used in sec .
2 . in agreement with the result of ref .
@xcite , where the bs equation was used , one concludes that there are no light resonances ( with @xmath404 ) in the symmetric phase of qed3 and that there is an abrupt change of the spectrum of light excitations at @xmath405 .
it is appropriate to call the phase transition in qed3 a pseudo - cpt : in the non - symmetric phase , at @xmath406 , a new , nonperturbative , source of the breakdown of the conformal symmetry occurs .
in this section we shall discuss the properties of the effective action in theories with the cpt in more detail .
in particular we shall consider a connection of the dynamics of the cpt with the hypothesis of the partially conserved dilatation current ( pcdc ) @xcite .
the effective potentials derived in the 2-dimensional gn model ( see eqs .
( [ 16 ] ) ) and ( [ 24 ] ) ) and in quenched qed4 with @xmath407 ( see eqs .
( [ 62 ] ) and ( [ 65 ] ) ) have a similar form .
moreover , one can show that the kinetic term and terms with higher number of derivatives in both the gn model and quenched qed4 are conformal invariant @xcite .
in other words , the conformal anomaly comes only from the effective potential in both these models .
this point is intimately connected with the pcdc dynamics . in order to see this ,
let us determine the divergence of the dilatation current in these models .
( [ 24 ] ) implies that @xmath408 in the gn model , and eq .
( [ 65 ] ) yields @xmath409 in quenched qed4 with @xmath159 , where the cpt takes place @xmath410 .
now , recall that the dynamical dimension @xmath411 of the field @xmath412 is @xmath413 and @xmath414 in the gn model and in quenched qed4 ( with @xmath159 ) , respectively . therefore eqs .
( [ 86 ] ) and ( [ 87 ] ) assure that the dynamical dimension of the operator @xmath415 coincides with its canonical dimension : @xmath416 and @xmath417 in the 2-dimensional gn model and quenched qed4 , respectively .
this implies the the realization of the pcdc hypothethis in these models @xcite : the operator @xmath415 has the correct transformation properties under dilatation transformations . in the renormalization group language
, this means that the conformal symmetry in these models is broken by marginal ( renormalized ) operators and not by relevant ( superrenormalized ) ones ( irrelevant ( nonrenormalized ) operators contribute only small corrections in the infrared dynamics ) .
though these two models are very special , one may expect that at least some features of this picture will survive in the general case of theories with the cpt .
in particular , one may expect that in the general case the effective potential has the form @xmath418 where @xmath419 is a dimensionless constant and @xmath420 is a ( presumably ) smooth function .
the contribution of @xmath421 ( [ 88 ] ) into the conformal anomaly is of the form @xmath422 where @xmath423 , i.e. , in the general case , logarithmic factors may destroy the covariance ( with respect to dilatation transformations ) of the relation for the conformal anomaly .
actually this takes place already in quenched qed4 with @xmath424 but @xmath177 . indeed , as follows from eqs .
( [ 76 ] ) and ( [ 79 ] ) , the logarithmic factors occur in the equation for the conformal anomaly in that case .
also , one should expect that the conformal invariance of the kinetic term and terms with higher number of derivatives may also be destroyed by logarithmic terms .
it is clear that the effective action in theories with the cpt are very different from that in the 4-dimensional linear @xmath1-model and nambu - jona - lasinio model , where the conformal symmetry is broken by relevant operators and the chiral phase transition is a mean - field one .
this point can be relevant for the description of the low energy dynamics in qcd and in models of dynamical electroweak symmetry breaking .
in particular , as was already pointed out in ref .
@xcite , the low energy dynamics are very sensitive to the value of the dynamical dimension @xmath411 .
in this section , we will consider the phase diagram with respect to the bare coupling constant @xmath425 and the number of fermion flavors @xmath13 in a 4-dimensional @xmath44 vector like gauge theory @xcite .
in particular , we will discuss a recent suggestion @xcite that the phase transition with respect to @xmath13 in that theory resembles the phase transition ( with respect to the coupling constant ) in quenched qed4 at @xmath159 . a starting point of the analysis of refs .
@xcite and @xcite is the presence of an infrared fixed point in the two - loops @xmath35 function of a @xmath44 theory , when the number of fermion flavors @xmath13 is large enough . recall that the perturbative @xmath35 function in that theory is @xmath426 in the case of the @xmath13 fermions in the fundamental representation , the first two coefficients are @xcite : @xmath427 while these two coefficients are invariant under change of a renormalization scheme , the higher - order coefficients are scheme - dependent . actually , there is a renormalization scheme in which all the higher - order coefficients vanish @xcite .
therefore there is at least one renormalization scheme in which the two - loop @xmath35 function is ( perturbatively ) exact .
we will use such a renormalization scheme .
the theory is asymptotically free if @xmath428 @xmath429 .
if @xmath428 and @xmath430 , the @xmath35 function has a zero , corresponding to a infrared - stable fixed point , at @xmath431 when @xmath13 is close to @xmath432 , the ratio @xmath433 , and therefore the value of @xmath434 , is small .
the value of the fixed point @xmath434 increases with decreasing @xmath13 , and this fixed point disappears at the value @xmath435 , when the coefficient @xmath436 becomes positive ( @xmath437 is @xmath438 for @xmath439 ) .
it is convenient to consider @xmath13 as a continuous parameter and to study the dynamics as @xmath13 is varied .
note that since @xmath13 appears analytically in the path integral of the theory , one can give a non - perturbative meaning to the theory with non - integer @xmath13 .
unlike ultraviolet - stable fixed points , defining dynamics at high momenta , infrared - stable fixed points ( defining dynamics at low momenta ) are very sensitive to nonperturbative dynamics leading to the generation of particle masses .
for example , if fermions acquire a dynamical mass , they decouple from the infrared dynamics .
then , only gluons will contribute to the @xmath35 function , and as a result , the perturbative infrared - stable fixed point in the @xmath35 function will disappear .
thus the crucial question is the interplay between the value @xmath434 of the infrared - stable fixed point and the chiral dynamics . in fig .
2 , the phase diagram suggested in ref .
@xcite is shown .
the authors considered the hamiltonian lattice gauge theories with kogut - susskind fermions .
the main features of this diagram are the following .
the vertical line ( with @xmath440 ) corresponds to a first - order phase transition , dividing a weak - coupling phase , possessing a continuum limit @xmath136 , from a strong - coupling phase .
notice that there is spontaneous chiral symmetry breaking in both these phases : the chiral order parameter jumps without vanishing at this phase transition : note the following points @xcite : \a ) the vertical line occurs because the chiral @xmath35 function is positive for large @xmath43 and negative for small @xmath43 , where it coincides with the perturbative @xmath35 function .
thus @xmath35 function has an infrared - stable zero at some intermediate @xmath43 .
the line of these zero is the vertical line at which the first - order phase transition occurs ( it can not be second - order because @xmath35 has no ultraviolet - stable zeroes ) ; \b ) the number @xmath13 refers to the number of fermion fields in the formal continuum theory which is twice the number of single component lattice fields .
therefore the minimal ( non - zero ) value of @xmath13 is @xmath441 .
the vertical line ends at this value since in pure gluodynamics there is apparently no phase transition between weak - coupling and strong - coupling phases @xcite .
the right - hand portion of the curve on the diagram , separating symmetric and non - symmetric phases occurs due to the following reason . at large enough values of the coupling , spontaneous chiral symmetry breaking takes place for any number @xmath13 of kogut - susskind fermions .
then the authors of ref .
@xcite argue that it is not reasonable to allow spontaneous chiral symmetry breaking to persist bellow some finite @xmath442 . as a result
they suggest the existence of that right - hand portion of the curve on the diagram , describing a chiral first - order phase transition .
the form of this curve reflects the fact that polarization screening effects become stronger with increasing @xmath13 , and therefore the value of @xmath43 , at which the first - order chiral phase transition occurs , increases with @xmath13 .
note that it is called a bulk phase transition in the literature . at last
, the left - hand portion of the curve , separating symmetric and non - symmetric phases , coincides with the line of the infrared - stable fixed points @xmath443 in eq .
( [ 92 ] ) .
it separates the symmetric , coulomb , phase describing interactions of massless gluons and fermions , and the non - symmetric ( and confinement ) phase .
since it is a line of infrared - stable fixed points , it describes a first - order phase transition .
thus , in the banks - zaks picture , spontaneous chiral symmetry breaking and confinement occur in the weak - coupling phase at all @xmath444 . notice that the left - hand part of the curve in the phase diagram describes a rather unusual situation : at @xmath445 spontaneous chiral symmetry breaking and confinement disappear with @xmath446 the bare coupling constant @xmath43 .
as we shall show , the phase diagram changes dramatically if one adopts the suggestion of ref .
@xcite concerning the dynamics of chiral symmetry breaking in this model .
the suggestion is that since the value of the infrared - stable fixed point @xmath434 is small at @xmath447 , one should expect that there is a critical value of @xmath13 , @xmath448 , above which the chiral symmetry is restored . in order to estimate the critical value @xmath47 , the authors of ref .
@xcite use the dynamical picture of chiral symmetry breaking corresponding to the rainbow ( ladder ) approximation . as in known @xcite , in a @xmath44 gauge theory , this picture coincides with that in quenched qed4 , with the replacement the coupling constant @xmath449 by @xmath450 .
therefore , in this approximation , spontaneous chiral symmetry occurs when the gauge coupling exceeds a critical value @xmath451 .
then , the estimate for the critical value @xmath47 is @xmath452 where @xmath443 is @xmath453 with @xmath34 and @xmath436 from eq .
( [ 91 ] ) .
the estimate ( [ 93 ] ) leads to the critical value @xmath454 for @xmath439 , for example , @xmath47 is just below 12 .. of approximately @xmath455 @xcite . for @xmath456 , the factor
@xmath457 is @xmath458 .
therefore , if this factor reflects the contribution of higher orders , the estimate ( [ 93 ] ) may be reliable . ]
when @xmath459 , the value of the infrared - stable fixed point @xmath434 is less than the critical value @xmath460 and there is no chiral symmetry breaking .
there are two possibility : the bare coupling constant @xmath425 is @xmath461 or @xmath462 . let us first consider the case with @xmath463 . then , if @xmath464 , the running gauge coupling @xmath465 is equal to the value @xmath434 for all @xmath466
. then , as @xmath136 , one gets a conformal theory describing interactions of massless fermions and gluons . on the other hand , if @xmath467 , the running coupling @xmath468 changes from @xmath469 at @xmath470 through @xmath471 at @xmath472 .
nonperturbative effects , such as chiral symmetry breaking , are now absent , though the conformal symmetry is broken by ordinary perturbative contributions leading to running of @xmath473 .
thus , in this case , an interacting non - abelian coulomb phase of massless quarks and gluons is still realized .
let us now consider the case with @xmath474 .
if the value of @xmath425 is close to @xmath434 , then the interactions are still weak , and chiral symmetry is unbroken .
therefore there is still an interacting coulomb phase in this case , though , unlike the case with @xmath467 , the running coupling decreases with @xmath475 , and @xmath476 as @xmath477 . as @xmath136
, we recover the conformal theory , with @xmath478 for all @xmath475 , discussed above .
as @xmath425 becomes sufficiently large , one comes again to the first - order chiral ( bulk ) phase transition .
the above consideration leads us to suggesting the phase diagram shown in fig .
3 . notice that , as before , the form of the right - hand part of the curve , describing the bulk phase transition , reflects the point that the polarization screening effects becomes stronger with increasing @xmath13 .
in particular , the rainbow approximation , used at @xmath479 , ceases to be good at larger @xmath13 .
the left - hand portion of the curve in fig .
3 still coincides with the line of the infrared - stable fixed points @xmath443 in eq .
( [ 92 ] ) .
however , now it separates two symmetric ( with unbroken chiral symmetry ) phases , and , besides that , its lower end point is @xmath448 and not @xmath480 as in fig . 2 : at @xmath481 the infrared - stable fixed point is washed out by generating a dynamical fermion mass .
these two symmetric phases are distinguished by their dynamics at short distances : while the dynamics of one phase is asymptotically free , another is not . on the other hand ,
their long distances dynamics , governed by the infrared - stable fixed point , are similar . at last , the horizontal , @xmath448 , line describes the cpt - like phase transition in this model .
the relation ( [ 74 ] ) suggests the following scaling law for @xmath62 : @xmath482 where @xmath419 is some constant @xmath483 in eq .
( [ 74 ] ) ) and @xmath484 is a scale at which the running coupling is of order @xmath460
. the dynamics in a @xmath485 gauge theory with @xmath486 may be relevant for the realization of the scenario of `` walking''-like technicolor for electroweak symmetry breaking @xcite : the `` walking''coupling constant @xmath487 governs the chiral symmetry breaking dynamics . in this case
the effective action in quenched qed4 , considered in sec.4 , should be relevant for the description of this dynamics .
let us now turn to data of lattice computer simulations of a @xmath44 gauge theory .
lattice computer simulations of the @xmath488 theory with @xmath489 and @xmath490 of staggered fermions in ref .
@xcite show the presence of the first - order , bulk transition separating strong- and weak - coupling phases .
while at @xmath490 there is a clear signature of the chiral phase transition at which the order parameter jumps to zero , at @xmath491 the situation is less clear .
more recent , and refined , simulations of this theory with @xmath491 indicates that in this case there is a first - order phase transition at which the order parameter jumps without vanishing @xcite .
thus the data of both these simulations seem indicate on the presence of the vertical line ( at least at @xmath492 ) shown in fig . 2 and fig .
it is still impossible at present to discriminate between these two phase diagrams .
note that since the bulk phase transition is a lattice artifact , the form of the phase diagram can depend on the type of fermions used in the simulations .
the simulations of the @xmath488 theory with wilson fermions @xcite show that theories which satisfy both quark confinement and spontaneous chiral symmetry breaking in the continuum limit exist only for @xmath493 . when @xmath494 , the theory is non - trivial ( interacting ) , however , without quark confinement
. moreover , at @xmath495 , chiral symmetry is unbroken at all values of the bare coupling @xmath425 .
these data seem favor the phase diagram in fig . 3 in which the right - hand part of the curve is replaced by the horizontal line @xmath496 .
it is clear that more data are needed in order to establish firmly the phase diagram in a @xmath44 gauge theory .
in this paper we introduced the conception of the conformal phase transition ( cpt ) which provides a useful framework for studying nonperturbative dynamics in gauge ( and also other ) field theories .
we described the general features of this phase transition .
the cpt is intimately connected with the nonperturbative breakdown of the conformal symmetry , in particular , with the pcdc dynamics . in the non - symmetric phase
the conformal symmetry is broken by marginal operators .
this in turn yields a constraint on the form of the effective action in theories with the cpt . in all the examples of the cpt considered in this paper , the conformal symmetry was explicitly broken by the conformal anomaly in the phase with spontaneous chiral symmetry breaking .
is it possible to realize dynamics with both chiral and conformal symmetries being broken spontaneously ?
although at present this question is still open , we would like to note that long ago arguments had been given against the realization of such a possibility @xcite the conception of the cpt can be useful for strong - coupling gauge theories , in particular , for qcd and models of dynamical electroweak symmetry breaking . in connection with that , we note that the effective action considered in sec .
6 may be relevant for the description of @xmath1 meson @xmath497 @xcite .
if it is rather light ( with @xmath498 mev ) as some authors conclude @xcite , it can dominate in the matrix elements of the operator @xmath499 in low energy dynamics , i. e. , it can be considered as a massive dilaton , as was already suggested some time ago @xcite .
it is also clear that the conception of pcdc and massive dilaton can be useful for the description of the dynamics of composite higgs boson .
another application of the cpt ( or pseudo - cpt ) may be connected with non - perturbative dynamics in condensed matter .
here we only mention the dynamics of non - fermi liquid which might be relevant for high - temperature surperconductivity : some authors have suggested that qed3 may serve as an effective theory of such a dynamics @xcite .
there has been recently a breakthrough in understanding non - perturbative infrared dynamics in supersymmetric ( susy ) theories ( for a review see ref .
it would be worth considering the realization of the cpt , if any , in susy theories , thus possibly establishing a connection between susy and non - susy dynamics .
v.a.m . thanks v. p. gusynin and i. a. shovkovy for useful discussions . he wishes to acknowledge the jsps ( japan society for the promotion of science ) for its support during his stay at nagoya university .
this work is supported in part by a grant - in aid for scientific research from the ministry of education , science and culture ( no .
08640365 ) .
in this appendix the relation for the conformal anomaly is derived .
since the relation for the conformal anomaly is an operator one , one can consider its realization for any matrix element .
we will consider the vacuum expectation value @xmath500 .
as @xmath501 , where @xmath502 is the canonical expression for the trace of the energy - momentum tensor and @xmath503 is its vacuum expectation value in the perturbative ( massless ) vacuum .
for the models in question , this definition guarantees the finiteness of @xmath500 in the continuum limit , after renormalizations of coupling constants . ] then , because of the lorentz invariance , @xmath504 , where @xmath505 is the vacuum energy density .
it has the form : @xmath506 where @xmath507 is some function of dimensionless coupling constants @xmath508 .
let us assume that the renormalization of @xmath508 leads to a finite @xmath505 in the continuum limit .
this implies that @xmath509 where @xmath510 .
therefore @xmath511 then , using the path integral representation for @xmath505 , we obtain the relations for the conformal anomaly considered in secs . 3 and 4 .
99 l. d. landau , phys . z. sowjet . * 11 * ( 1937)26 ; v. l. ginzburg and l. d. landau , jetp * 20 * ( 1950)1064 ; j. zinn justin , `` quantum field theory and critical phenomena '' ( clarendon press , oxford , ( 1989 ) ) . k. g. wilson , phys . rev . * b4 * ( 1971)3174 ; k. g. wilson and j. b. kogut , phys . rep .
* 12 * ( 1974)75 .
t. banks and a. zaks , nucl . phys .
* b196 * ( 1982)189 .
v. a. miransky , int . j. mod
a8 * ( 1993)135 .
t. appelquist , j. terning , and l. c. r. wijewardhana , phys .
rev . lett . * 77 * ( 1996)1214 .
t. appelquist , j. terning , and l. c. r. wijewardhana , phys .
* 75 * ( 1995)2081 .
kondo , m. tanabashi , and k. yamawaki , prog .
* 89 * ( 1993)1249 .
d. gross and a. neveu , phys . rev . * d10 * ( 1974)3235 . e. witten , nucl . phys .
* b145 * ( 1978)110 . ,
k. yamawaki ( world scientific co , singapore , 1992 ) ; v. a. miransky , _ dynamical symmetry breaking in quantum field theories _ ( world scientific co , singapore , 1993 ) .
t. maskawa and h. nakajima , prog .
* 52 * ( 1974)1326 ; p. i. fomin and v. a. miransky , phys .
b64 * ( 1976)166 ; r. fukuda and t. kugo , nucl
b117 * ( 1976)250 .
p. i. fomin , v. p. gusynin , v. a. miransky , and yu .
a. sitenko , rivista del nuovo cimento , * 6 * n5 ( 1983)1 ; v. a. miransky , nuovo cimento , * 90a * ( 1985)149
. w. a. bardeen , c. n. leung , and s. t. love , phys .
* 56 * ( 1986)1230 ; c. n. leung , s. t. love , and w. a. bardeen , nucl .
phys . * b273 * ( 1986)649 .
kondo , h. mino , and k. yamawaki , phys .
* d39 * ( 1989)2430 ; t. appelquist , m. soldate , t. takeuchi , and l. c. r. wijewardhana,_in proceedings of john hopkins workshop on current problems in particle theory _ , eds . g. domokos and s. kovesi - domokos ( world scientific co , singapore , 1988 ) .
v. a. miransky and k. yamawaki , mod .
. lett . * a4 * ( 1989)129 .
w. a. bardeen and s. t. love , phys .
* d45 * ( 1992)4672 ; w. a. bardeen , s. t. love , and v. a. miransky , phys . rev .
* d42 * ( 1990)3514 .
v. p. gusynin and v. a. miransky , mod .
. lett . * a6 * ( 1991)2443 ; sov .
jetp * 74 * ( 1992)216 .
t. appelquist , u. mahanta , d.nash , and l. c. r. wijewardhana , phys . rev .
* d43 * ( 1991)646 ; d. c. curtis and m. r. pennington , phys . rev .
* d48 * ( 1993)4933 ; d. atkinson , j. c. r. bloch , v. p. gusynin , m. r. pennington , and m. reenders , phys . lett . *
b329 * ( 1994)117 ; z. dong , h. munczek , and c. roberts , phys .
b333 * ( 1994)536 .
r. pisarski , phys . rev .
* d29 * ( 1984)2423 ; t. appelquist , m. bowick .
d. karabali , and l. c. r. wijewardhana , phys . rev . * d33 * ( 1986)3704 .
t. appelquist , d. nash , and l. c. r. wijewardhana , phys .
* 60 * ( 1988)2575 ; d. nash , phys .
* 62 * ( 1989)3024 .
e. dagotto , j. b. kogut , and a. koci , phys .
* 62 * ( 1989 ) 1083 ; e. dagotto , a. koci , and j. b. kogut , nucl
* b344 * ( 1990)279 ; s. hands and j. b. kogut , nucl . phys . * b335 * ( 1990)455 . m. r. pennington and d. walsh , phys . lett .
* b253 * ( 1991)246 ; d. c. curtis , m. r. pennington , and d. walsh , phys . lett . *
b295 * ( 1992)313 ; r. pisarski , phys . rev .
* d44 * ( 1991)1866 . v.
p. gusynin , a. h. hams , and m. reenders , phys . rev .
* d53 * ( 1996)2227 ; p. maris , phys . rev . *
d54 * ( 1996)4049 .
r. j. crewther , phys .
* 28 * ( 1972)1421 ; m. s. chanowitz and j. ellis , phys . rev .
* d7 * ( 1973)2490 .
j. schechter , phys .
rev . * d21 * ( 1980)3393 ; a. a. migdal and m. a. shifman , phys . lett . * b114 *
( 1982)445 .
v. a. miransky and v. p. gusynin , prog .
* 81 * ( 1989)426 ; v. a. miransky and m. d. scadron , sov .
* 49 * ( 1989)922 .
s. shuto , m. tanabashi , and k. yamawaki , _ in proceedings of 1989 workshop on dynamical symmetry breaking , nagoya , 1989 _ ( department of physics , nagoya university , nagoya,1989 ) .
d. r. t. jones , nucl .
b75 * ( 1974)531 ; w. e. caswell , phys .
lett . * 33 * ( 1974)244 .
g. t hooft , _ in recent developments in gauge theories , nato advanced study inst .
b , v. 59 _ , eds . g. t hooft et al .
( plenum press , new york , 1980 ) . g. bhanot and m. creutz ,
phys . rev .
* d24 * ( 1981)3212 .
t. appelquist , k. lane , and u. mahanta , phys .
* 61 * ( 1988)1553 .
b. holdom , phys . rev . * d24 * ( 1981)1441 ; phys . lett .
* b150 * ( 1985)301 ; k. yamawaki , m. bando , and k. matumoto , phys .
lett . * 56*(1986)1335 ; t. akiba and t. yanagida , phys .
* b169 * ( 1986)308 ; t. appelquist , d. karabali , and l. c. r. wijewardhana , phys .
* 57 * ( 1986)957 .
j. b. kogut and d. r. sinclair , nucl . phys .
* b295 * [ fs 21 ] ( 1988)465 .
f. r. brown , h. chen , n. h. christ , z. dong , r. d. mawhinney , w. shafer , and a. vaccarino , phys
d46 * ( 1992)5655 .
y. iwasaki , k. kanaya , s. sakai , and t. yoshi , phys .
rev . lett .
* 69 * ( 1992)21 ; y. iwasaki , k. kanaya , s. kaya , s. sakai , and t. yoshi , tsukuba preprint uthep-342 , 1996 ; hep - lat/9608125 .
r. jackiw , phys . rev .
* d3 * ( 1971)1347 ; _ ibid . _
* d3 * ( 1971)1356 ; v. p. gusynin , v. a. kushnir , and v. a. miransky , phys . lett .
* b213 * ( 1988)177 .
r. m. barnett et al . , ( particle data group ) , phys . rev . * d54 * ( 1996)1 .
n. a. trnqvist and m. roos , phys . rev .
* 76 * ( 1996)1575 ; m. harada , f. sannino , and j. schechter , phys .
* d54 * ( 1996)1991 ; s. ishida et al .
* 95 * ( 1996)745 .
r. j. furnstahl , h .- b .
tang , and b. d. serot , phys .
rev . * c52 * ( 1995)1368 .
a. kovner and b. rosenstein , phys.rev .
* b42 * ( 1990)4748 ; n. dorey and n. e. mavromatos , nucl . phys .
* b368 * ( 1992)614 ; i. j. r. aitchison and n. e. mavromatos , phys . rev .
* b53 * ( 1996)9321 .
k. intriligator and n. seiberg , nucl .
b ( proc . suppl . )
* 45b&c * ( 1996)1 .
figure 1 . the phase diagram in the gauged njl model .
the marks s and a denote symmetric and asymmetric phases , respectively . | the conception of the conformal phase transiton ( cpt ) , which is relevant for the description of non - perturbative dynamics in gauge theories , is introduced and elaborated .
the main features of such a phase transition are established . in particular
, it is shown that in the cpt there is an abrupt change of the spectrum of light excitations at the critical point , though the phase transition is continuous .
the structure of the effective action describing the cpt is elaborated and its connection with the dynamics of the partially conserved dilatation current is pointed out .
the applications of these results to qcd , models of dynamical electroweak symmetry breaking , and to the description of the phase diagram in ( 3 + 1)-dimensional @xmath0 gauge theories are considered .
= 20.5 cm = 16.5 cm -1.35 cm = .285 in dpnu-96 - 58 + november 1996 |
water is essential in biology , because it participates in nearly every process necessary for life ( including cell metabolism , transport of nutrients and residues , protein conformation changes , etc . ) , and is the most common solvent in chemistry .
it regulates a large variety of processes , including atmospheric phenomena , the formation of geophysical structures , the propagation of cracks in stones and cement , the sliding of glaciers , the transport in plants , and is ubiquitous in the universe as ice in the interstellar space . in all these examples the properties of water are essential to understand what is observed .
nevertheless , water has proven to be a complex puzzle to many researchers for its anomalous thermodynamic and dynamic properties at room temperature and pressure .
for example , by decreasing temperature @xmath0 at pressure @xmath1 atm , water s volume fluctuations , proportional to the isothermal compressibility @xmath2 , increase below @xmath3c , and entropy fluctuations , proportional to the isobaric specific heat @xmath4 increase below @xmath5c , while in normal liquid any fluctuation decreases when @xmath0 is decreased @xcite .
these water s anomalies grow upon cooling and increase in number .
for example below @xmath6c the cross - fluctuation of volume and entropy , proportional to the isobaric thermal expansion coefficient @xmath7 , becomes negative @xcite , while it is always positive in normal liquids where the entropy decreases when the volume decreases @xcite . by decreasing @xmath0
even more , it is experimentally possible to supercool liquid water down to @xmath8c at 1 atm and to @xmath9c at 2000 atm , where the liquid is metastable with respect to crystal phases @xcite .
these extreme conditions are not unusual in nature , where water exists in its liquid form at @xmath10c in insects , @xmath11c in clouds or @xmath12c in plants @xcite .
below @xmath13 the homogeneous nucleation of crystal ice occurs in a time too short to allow any measurements .
but even the crystal phase of water is not simple .
in fact , water is a polymorph with at least sixteen forms of crystal ices , the last one , ice xv , was discovered in 2009 @xcite .
however , at very low @xmath0 , crystal water is not the only possible kind of ice . by rapidly quenching liquid water below @xmath14c @xcite , or by condensing the vapour at low @xmath0 @xcite , or by compressing crystal ice at low @xmath0 @xcite , or by irradiation ( with ions for example @xcite ) , it is possible to solidify water as an amorphous , or glass , i. e. a form that has the elastic properties of a solid , but the structure of a liquid with no long - range order @xcite . as for the crystal state ,
the amorphous state of water is also not unique .
water is a polyamorph with at least three different amorphous states : low density amorphous ( lda ) , discovered in 1935 @xcite , high density amorphous ( hda ) , discovered in 1984 @xcite , and very high - density amorphous ice ( vhda ) , discovered in 2001 @xcite .
all these anomalies of water are a consequence of the properties of the hydrogen bond network that they form .
the hydrogen bond interaction is characterized by a preferred geometrical configuration , that at low @xmath0 and @xmath15 is approximately a tetrahedron of four molecules around a central one , with an angle varying around @xmath16 ( slightly smaller than a tetrahedral angle of @xmath17 ) and a distance oscillating around @xmath18 @xcite .
the local arrangement , including the number of nearest neighbours , can change with @xmath0 and @xmath15 .
in particular , in 2000 soper and ricci observed at @xmath19 k , compressing from @xmath20 to @xmath21 mpa , a continuous transformation from low
density liquid ( ldl ) local arrangement of water with an open , hydrogen - bonded tetrahedral structure , to high density liquid ( hdl ) local arrangement with nontetrahedral o - o - o angles and a collapsed second coordination shell with broken hydrogen bonds , and a change in density of about 73% @xcite .
all the above results are consistent with theories that propose different mechanisms and different phase behaviors in the supercooled region .
they can be summarized in four possible scenarios for the @xmath22 phase diagram .
\(i ) in the _ stability limit _ ( sl )
scenario @xcite the behavior of the superheated liquid spinodal , i. e. the limit of stability of the liquid with respect to the gas , and the stretched water , i. e. water under tension as in a plant fibers , are related . in particular , it is hypothesized that the limits of stability of these two regions are continuously connected at negative pressure , forming a re - entrant curve toward the positive @xmath15 region below @xmath23 .
the response functions , including @xmath2 , @xmath4 and @xmath7 , diverge when @xmath0 is decreased a positive @xmath15 as a consequence of the approaching of the re - entrant spinodal line .
\(ii ) in the _ liquid liquid critical point _ ( llcp ) scenario @xcite it is hypothesized the existence of a ldl hdl first
order phase transition line with negative slope in the @xmath22 plane and terminating in a critical point @xmath24 . below the critical pressure
@xmath25 the response functions increase on approaching the widom line ( the locus of correlation length maxima emanating from @xmath24 into the one phase region ) @xcite , and for @xmath26 by approaching the hdl to ldl spinodal line .
the possibility with @xmath27 have also been proposed @xcite .
\(iii ) in the _ singularity free _ ( sf ) scenario @xcite it is hypothesized that the low-@xmath0 anticorrelation between volume and entropy gives rise to response functions that increase upon cooling and display maxima at non
zero @xmath0 , but do not display any singular behavior .
specifically , sastry et al .
@xcite show that this is a direct consequence of the fact that water s line of temperatures of maximum density ( tmd ) has a negative slope in the @xmath28 plane .
\(iv ) in the _ critical point free _ ( cpf ) scenario @xcite it is hypothesized that a first order phase transition line separates two liquid phases and extends to @xmath29 toward 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 .
since all these scenarios are consistent with the available experimental data , a natural question is if we can design an experiment that would discriminate among them .
unfortunately , many scientists have discovered that an answer to this question is a difficult challenge @xcite .
in fact , experiments on bulk water are hampered by freezing below @xmath13 , and no measurements on bulk liquid water can be performed with our present technology below this temperature . nevertheless , the different proposed theories have different implications on phenomena such as the cold denaturation ( and stop of activity ) of proteins at low @xmath0 , an important issue in cryopreservation , cryonics , cryostasis and cryobiology .
one possible strategy to probe supercooled water at very low @xmath0 is to consider water at an interface .
water adsorbed onto the surface of proteins or confined in nanopores freezes at much lower @xmath0 than bulk water , giving access to a low - temperature region where interfacial water is still liquid , while bulk water would not be @xcite . in many cases of interest for practical purposes in biology , geology or industrial applications , water is hydrating a surface or is confined . as a consequence , fundamental research in physics and
chemistry has been performed in recent years with experiments @xcite , theories and simulations @xcite . during the last years experiments on water in vycor micropores @xcite , in nanopores of mcm-41 silica @xcite , of sol gel silica glass @xcite , of naa zeolites @xcite , or of double - wall carbon nanotubes @xcite
have contributed to the investigation of water dynamics in confinement .
in particular , confinement in hydrophilic mcm-41 silica nanopores of @xmath30 and @xmath31 nm diameter allows to study water dynamics down to @xmath32 k where quasielastic neutron scattering reveals a crossover at @xmath33 k in the average translational relaxation time from a non - arrhenius behavior at high @xmath0 to an arrhenius behavior at low @xmath0 @xcite .
a similar crossover is also observed for the self - diffusion coefficient of water by nuclear magnetic resonance at @xmath34 k @xcite . by increasing from @xmath21 bars to @xmath35
bars the external pressure applied on a sample of mcm-41 silica nanopores with @xmath31 nm diameter at full hydration level of @xmath36 g of h@xmath37o per g of silica , it has been observed that the crossover occurs at lower @xmath0 , reaching @xmath38 k at @xmath39 bars and disappears at higher @xmath15 @xcite .
quasielastic neutron scattering experiments show the same crossover for the average translational relaxation time of at least three different systems : ( i ) water hydrating lysozymes , at hydration level @xmath40 g of h@xmath37o per g of dry lysozyme , for @xmath41 k , a temperature below which the protein has a glassy dynamics @xcite , ( ii ) dna hydration water , at hydration level corresponding to about 15 water molecules per base pairs , for @xmath42 k , at which dna displays the onset of anharmonic molecular motion @xcite , and ( iii ) rna hydration water , at a similar hydration level , for @xmath41 k , where both rna and its hydration water exhibit a sharp change in slope for the mean - square displacements of the hydrogen atoms @xcite , all these results can be interpreted as a consequence of a structural rearrangement of water molecules associated with a ldl - hdl critical point @xcite . in fact , along the widom line in the supercritical region of the ldl - hdl critical point , the changes in the hydrogen bond network are consistent with the dynamic behavior observed in the experiments , as we will discuss in the following sections
. this interpretation has been criticized @xcite on the base of similar crossover observed for water confined in molecular sieves @xcite or for water mixtures @xcite and water solutions @xcite .
it has also been proposed a possible interpretation as a consequence of the dynamics of ( bjerrum - type ) defects due to orientationally disordered water molecules that are hydrogen bonded to less than four other water molecules @xcite .
@xmath43h - nmr studies on hydrated proteins , at a comparable hydration level as in @xcite , show no evidence for the crossover at @xmath44 k and indicate that water performs thermally activated and distorted tetrahedral jumps at @xmath45 k , which may be related to a universal defect diffusion @xcite .
also , dielectric spectroscopy studies of hydrated protein show a smooth temperature variations of conductivity at @xmath44 k and ascribe the crossover observed in neutron scattering to a secondary relaxation that splits from the main structural relaxation @xcite .
on the other hand , numerical simulations for bulk water show that crossing the widom line emanating from a ldl - hdl critical point , the structural change in water is maximum , as emphasized by the maximum in specific heat , and the diffusion constant has a crossover @xcite .
this result is observed also in water hydrating lysozyme or dna , where the dynamic transition of the macromolecules occurs at the temperature of dynamic crossover in the diffusivity of hydration water and also coincides with the maximum change in water structure @xcite .
a crossover from high-@xmath0 non - arrhenius to low-@xmath0 arrhenius behavior is observed also in simulation of water hydrating lysozyme powder in the translational correlation time and in the inverse of the self - diffusion constant , in agreement with the neutron scattering experiments , at about @xmath46 k @xcite . the activation energy for the arrhenius regime
is found to be of about @xmath47 ev , as in the neutron scattering experiments @xcite .
also , simulations of water hydrating elastin - like and collagen - like peptides show this crossover , but with a weaker change in the slope and an arrhenius activation energy of about @xmath48 ev , consistent with dielectric spectroscopy and nuclear magnetic resonance studies @xcite .
it is therefore difficult not only in the experiments , but also in the models to get a clear answer about the relevant dynamic mechanisms and their relation with the thermodynamics in water at interfaces .
moreover , the relation between confined water and bulk water remains not fully understood .
for this reason models that are tractable with a theoretical approach are particularly appealing in this context . with these models , in fact , simulations can be compared with analytic calculations to develop a consistent theory .
we consider the case of water in two dimensions ( 2d ) .
this case can be considered as an extreme confinement of one single layer of water between two repulsive ( hydrophobic ) walls when the distance between the walls is such to inhibit the crystal formation @xcite .
in fact , it has been shown that the relevant parameter to avoid the transition to a crystal phase is the distance between the confining wall and not the characteristics of the hydrophobic interaction with the wall @xcite .
another case in which the study of a monolayer of water is relevant is when a substrate of protein powder is , on average , hydrated only by a single layer of water , and the proteins do not undergo any configurational transformation and/or large scale motion @xcite . in these conditions , for a hydrophilic protein surface , we can assume that the effect of the water
protein interaction is to attract water on a surface that , by constraining the water molecule positions , inhibits its crystallization .
a very desirable feature of a model for a liquid is transferability .
the parameters and effective interactions of a model are optimized to precisely reproduce static and dynamic properties of the liquid at one particular thermodynamic state point .
the quality of the model is measured by the range of validity of its predictions in other state points .
unfortunately , there is no water model that is truly transferable , nor can reproduce all the properties of water @xcite .
many routes have been explored to solve this issue .
molecular polarizability @xcite is one way to introduce effects not considered by standard pairwise additive potentials . however , polarizable models are computationally very expensive and provide only a partial solution @xcite .
an alternative way is to include many
body effects into the potential . in the following
we define a model with an effective many
body interaction introduced through a cooperative hydrogen bond term .
we consider @xmath49 molecules in a volume @xmath50 with periodic boundary conditions ( p.b.c . ) in two dimensions , and the size of about one single molecule ( and no p.b.c . ) in the third dimension .
we initially consider the case in which the molecules are distributed in a homogeneous way , with each molecule @xmath51 $ ] occupying the same volume @xmath52 larger than a hard core volume @xmath53 @xmath54 due to short - range electron clouds repulsion .
we consider the case in which each molecule has coordination number four , consistent with the tendency of a water molecule to minimize its energy by forming four hydrogen bonds .
the interaction hamiltonian for water molecules is @xcite @xmath55 where @xmath56 denotes the sum of all the isotropic interactions ( e. g. van der waals ) between molecules at distance @xmath57 , represented by a lennard jones potential with attractive energy @xmath58 plus a hard core at distance @xmath59 . the second term ( with @xmath60
if @xmath61 and @xmath62 otherwise , and @xmath63 denoting that @xmath64 and @xmath65 are nearest neighbors ) accounts for the directional contribution to the hydrogen bond energy with strength @xmath66 , where @xmath67 is a ( potts ) variable representing the orientational state of the hydrogen ( or the lone e@xmath68 ) of molecule @xmath64 facing the lone e@xmath68 ( or the hydrogen , respectively ) of the molecules @xmath65 . for the sake of simplicity
we do not distinguish between hydrogen and lone e@xmath68 , associating to each molecule four equivalent bond indices @xmath69 .
we choose the parameter @xmath70 by selecting 30@xmath71 as the maximum deviation from a linear bond , i. e. the o
h .... o angle is less than 30@xmath71 , as estimated from debye - waller factors @xcite .
hence , @xmath72 and every molecule has @xmath73 possible orientations .
the effect of choosing a different value for @xmath70 has been analyzed in @xcite . the third term ( with @xmath74 indicating each of the six different pairs of the four bond indices of molecule @xmath64 ) represents an interaction accounting for the hydrogen bond cooperativity and giving rise to the o o o correlation @xcite , locally driving the molecules toward an ordered ( tetrahedral in the bulk ) configuration with lower energy .
by defining the energy per hydrogen bond ( between @xmath69 and @xmath75 ) as the sum of the interactions in which two bonded molecules ( @xmath64 and @xmath65 ) are participating , we obtain @xmath76 , where @xmath77 is the number of cooperative interactions in which that bond variables ( @xmath69 and @xmath75 ) are partaking .
if we choose as parameters @xmath78 kj / mol ( consistent with the value @xmath79 kj / mol of the estimate of the van der waals attraction based on isoelectronic molecules at optimal separation @xcite ) , @xmath80 kj / mol and @xmath81 kj / mol , the values of @xmath82 ranges between @xmath83 and @xmath84 kj / mol depending on @xmath85 .
however , a definition of @xmath82 based on a cluster of @xmath86 or @xmath87 bonded molecules in @xmath883-dimensions increases this range up to @xmath89 or @xmath90 kj / mol , respectively .
therefore , @xmath82 depends on the environment ( the value of @xmath85 and the number of molecules bonded in a cluster ) , as observed in computer simulation of the crystalline phases of ice @xcite , and has values within the range 6.3 @xcite 23.3 kj / mol @xcite , proposed on the base of experiments .
experiments show that formation of the hydrogen 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 that a full bonded molecule , i. e. a molecule with four hydrogen bonds , has a molecular volume larger than a non - bonded molecule by an amount @xmath91 where @xmath92 is the volume increase per h bond .
hence , if @xmath93 is the total number of hydrogen bonds in the system , the hydrogen bond contribution to the total volume is @xmath94 we adopt @xmath95 consistent with the expected value of the van der waals radius @xcite , and @xmath96 , with @xmath97 , corresponding to a maximum hydrogen bond distance of about @xmath98 , consistent with the range of a water molecule s first coordination shell , @xmath99 , as determined from the oxygen - oxygen radial distribution function @xcite .
the model is studied using both mean field ( mf ) analysis and monte carlo ( mc ) simulations .
the mf approach has been describe in details in ref.s @xcite .
it consists of expressing the molar gibbs free energy in terms of an exact partition function for a portion of the system made of a treatable number of degrees of freedom .
we take into account the effect of all the rest of the system as a mean field acting on the border of this portion @xcite .
mc simulations are performed at constant @xmath49 , @xmath15 , @xmath0 , allowing the volume @xmath100 of the system to fluctuate as a stochastic variable .
the average distance @xmath101 between the molecules is then calculated as @xmath102 , where @xmath103 in two dimensions .
the total volume of the system is by definition @xmath104 where @xmath105 is the hydrogen bond volume contribution in eq.([d_vol ] ) .
note that @xmath105 is not included in the calculation of @xmath101 to avoid mf
type long range correlation in volume fluctuations in the mc simulations . for the parameters choice @xmath106 , @xmath107 and @xmath108
, we find that the density @xmath109 as a function of @xmath0 at constant @xmath15 displays a discontinuous change at high @xmath0 and low @xmath15 corresponding to the gas - liquid first order phase transition anding in a critical point where the discontinuity disappears ( fig .
[ fig1]a ) . by decreasing @xmath0 ,
the density reaches a maximum , that in real water at atmospheric pressure occurs at @xmath110c . at lower @xmath0 , in the supercooled state , and higher @xmath15
we find another discontinuity in density , this time with a lower density at lower @xmath0 ( fig .
[ fig1]b ) .
the system at these supercooled @xmath0 displays a first order phase transition from hdl to ldl , as hypothesized in the llcp scenario ( fig . 2 ) .
\(a ) ( in units of @xmath111 ) as a function of the temperature @xmath0 ( in units of @xmath112 where @xmath113 is the boltzmann constant ) for different values of pressure @xmath15 ( in units of @xmath114 ) as calculated from mc simulation of a system with @xmath115 water molecules .
the parameters of the model are @xmath106 , @xmath107 and @xmath108 .
( a ) at high @xmath0 and for ( from bottom to top ) values of @xmath116 from @xmath117 to @xmath118 , we observe for @xmath119 a discontinuity in the density corresponding to the first - order gas - liquid phase transition , with a critical @xmath15 at about @xmath120 and critical @xmath0 at about @xmath121 . note that if we choose as model parameters @xmath122 kj / mol and @xmath123 , we get an estimate @xmath124 mpa and @xmath125 k consistent with the real water critical point at about @xmath126 mpa and @xmath127 k. ( b ) at low @xmath0 and for ( from bottom to top ) @xmath116 from @xmath36 to @xmath118 , for @xmath128 a discontinuity in @xmath109 marks the first - order ldl - hdl phase transition.,title="fig:",width=264 ] ( b ) ( in units of @xmath111 ) as a function of the temperature @xmath0 ( in units of @xmath112 where @xmath113 is the boltzmann constant ) for different values of pressure @xmath15 ( in units of @xmath114 ) as calculated from mc simulation of a system with @xmath115 water molecules .
the parameters of the model are @xmath106 , @xmath107 and @xmath108 .
( a ) at high @xmath0 and for ( from bottom to top ) values of @xmath116 from @xmath117 to @xmath118 , we observe for @xmath119 a discontinuity in the density corresponding to the first - order gas - liquid phase transition , with a critical @xmath15 at about @xmath120 and critical @xmath0 at about @xmath121 .
note that if we choose as model parameters @xmath122 kj / mol and @xmath123 , we get an estimate @xmath124 mpa and @xmath125 k consistent with the real water critical point at about @xmath126 mpa and @xmath127 k. ( b ) at low @xmath0 and for ( from bottom to top ) @xmath116 from @xmath36 to @xmath118 , for @xmath128 a discontinuity in @xmath109 marks the first - order ldl - hdl phase transition.,title="fig:",width=264 ] [ fig2 ] as a function of the number of water molecules @xmath49 for pressure @xmath129 at low @xmath0 shows a linear increase as expected at a first order phase transition.,width=302 ] the experiments for confined water have boosted the debate over the supercooled phase diagram of water , motivating the proposal of the cpf scenario hypothesized by c. a. angell @xcite , as described above .
this new scenario leads to questions such as * how to understand the new angell hypothesis ? * how to connect it to the other three existing hypotheses ? a recent work by stokely et al .
@xcite succeeds in answering both questions ( i ) and ( ii ) .
specifically , it is shown that all four existing hypotheses are cases of the cooperative water model .
thus no matter which hypothesis may be correct ( if any is correct ) , it is possible that the underlying mechanism is basically the same the thermodynamic properties of water can be accounted for by considering two main contributions to the hydrogen bond interaction : ( a ) the directional ( or covalent ) contribution ( parametrized by @xmath66 in the model ) and ( b ) the three - body ( or cooperative ) contribution ( parametrized by @xmath130 in the model ) . by mf calculations and mc simulations , stokely et al .
@xcite demonstrate that the balance between contributions ( a ) and ( b ) determines which of the four hypotheses presented in section 1.1 .
best describes experimental facts .
since the characteristic energy associated with these two contributions can be estimated , the work allows to begin to validate or contradict each hypothesis on an experimental basis .
specifically , by fixing the parameters @xmath106 and @xmath131 , and varying the parameter @xmath132 , it is possible to observe that the cooperative model reproduces all four scenarios of section 1.1 .
the overall picture that emerges is one in which the amount of cooperativity among h bonds ( @xmath132 ) , in relation to the h bond directional strength ( @xmath133 ) , governs the location of a llcp , hence which scenario is realized . for zero cooperativity ,
the temperature @xmath134 where @xmath135 and @xmath136 diverge is at zero temperature , and no liquid - liquid transition exists for @xmath137the sf scenario .
for very large cooperativity , @xmath24 lies outside the region of stable liquid states , and a liquid - liquid transition extends to the entire ( supercooled and superheated ) liquid phase the cpf / sl scenario . for intermediate values of h bond cooperativity , @xmath134 varies in a smooth way between these two extremes the llcp scenario . due to the anticorrelation between the volume and entropy associated to the h bonds , the larger @xmath134 , the smaller @xmath25 , eventually with @xmath27 for larger cooperativity .
these cases are summarized in fig .
[ fig3 ] .
\(a ) vs temperature @xmath0 phase diagram of the cooperative water model for different values of @xmath132 .
( a ) mean field results showing the low @xmath0 phase diagram with the ldl - hdl phase transition lines ( solid lines , where @xmath2 is discontinuous ) and widom lines ( dashed lines , where @xmath2 has a finite maximum ) for varying @xmath132 from ( rightmost ) 0.04 , 0.03 , 0.02 , 0.01 , 0.005 and 0 ( leftmost ) .
for each value of @xmath132 solid circles indicate the ldl - hdl critical point @xmath24 where the response function ( such as @xmath2 ) diverge .
hashed circle indicates the state point at @xmath138 where @xmath2 diverges when @xmath139 .
( b ) the mc phase diagram for varying @xmath132 for @xmath140 water molecules . at high
@xmath0 the system displays a liquid
gas first order phase transition line ( continuous line with full circles ) ending in a liquid gas critical point @xmath141 ( full circle ) , from which departs the liquid - to - gas spinodal line ( dashed line with open diamonds ) . at lower @xmath0
, the retracing line with open squares marks the temperatures of maximum density ( tmd ) along isobars .
all these loci do not change in an appreciable way with the value of @xmath132 .
the phase diagram at lower @xmath0 , instead , show a strong dependence from @xmath132 . for @xmath142 ,
we find for any @xmath15 above the spinodal line a first - order phase transition line between a hdl ( at high @xmath15 ) and a ldl ( at low @xmath15 ) phase ( continuous line with open circles ) .
this is the cpf scenario @xcite .
the analysis of the hdl - to - ldl ( not shown ) reveals that the limit of stability of the hdl phase retraces to positive @xmath15 as in the sf scenario @xcite .
for @xmath143 we observe that the hdl - ldl phase transition ends to a critical point @xmath24 ( continuous line with open circles ending in a full large circle ) at negative @xmath15 , as in the llcp scenario suggested in ref . @xcite . from @xmath24
a widom line ( dashed line with full circles ) departs . for @xmath107
, @xmath24 occurs at positive @xmath15 , has hypothesized in ref . @xcite and the widom line ( dashed line with full squares ) extends to lower @xmath15 . for @xmath144
, @xmath24 approaches the @xmath138 axis , as well as the widom line ( dashed line with full diamonds ) , going toward the limit of the sf scenario @xcite.,title="fig:",width=264 ] ( b ) vs temperature @xmath0 phase diagram of the cooperative water model for different values of @xmath132 .
( a ) mean field results showing the low @xmath0 phase diagram with the ldl - hdl phase transition lines ( solid lines , where @xmath2 is discontinuous ) and widom lines ( dashed lines , where @xmath2 has a finite maximum ) for varying @xmath132 from ( rightmost ) 0.04 , 0.03 , 0.02 , 0.01 , 0.005 and 0 ( leftmost ) . for each value of @xmath132 solid circles indicate the ldl - hdl critical point @xmath24 where the response function ( such as @xmath2 ) diverge . hashed circle indicates the state point at @xmath138 where @xmath2 diverges when @xmath139 .
( b ) the mc phase diagram for varying @xmath132 for @xmath140 water molecules . at high @xmath0
the system displays a liquid
gas first order phase transition line ( continuous line with full circles ) ending in a liquid gas critical point @xmath141 ( full circle ) , from which departs the liquid - to - gas spinodal line ( dashed line with open diamonds ) . at lower @xmath0
, the retracing line with open squares marks the temperatures of maximum density ( tmd ) along isobars .
all these loci do not change in an appreciable way with the value of @xmath132 .
the phase diagram at lower @xmath0 , instead , show a strong dependence from @xmath132 . for @xmath142 ,
we find for any @xmath15 above the spinodal line a first - order phase transition line between a hdl ( at high @xmath15 ) and a ldl ( at low @xmath15 ) phase ( continuous line with open circles ) .
this is the cpf scenario @xcite .
the analysis of the hdl - to - ldl ( not shown ) reveals that the limit of stability of the hdl phase retraces to positive @xmath15 as in the sf scenario @xcite .
for @xmath143 we observe that the hdl - ldl phase transition ends to a critical point @xmath24 ( continuous line with open circles ending in a full large circle ) at negative @xmath15 , as in the llcp scenario suggested in ref .
@xcite . from @xmath24
a widom line ( dashed line with full circles ) departs . for @xmath107
, @xmath24 occurs at positive @xmath15 , has hypothesized in ref . @xcite and the widom line ( dashed line with full squares ) extends to lower @xmath15 . for @xmath144 ,
@xmath24 approaches the @xmath138 axis , as well as the widom line ( dashed line with full diamonds ) , going toward the limit of the sf scenario @xcite.,title="fig:",width=264 ]
to elucidate the relation between the protein dynamic crossover at about @xmath44 k and the dynamic crossover observed for the average translational correlation time in the first layer of protein hydration water @xcite , we perform mf calculations and mc simulations of the cooperative model of water of section 2.1 .
since we are interested in cases at low humidity , we consider the case of a water monolayer hydrating an immobile surface of globular protein that , forcing the water molecules out of place with respect to crystal configurations , inhibits the crystallization .
we focus on the hydrogen bonds dynamics , regardless if the hydrogen bonds are formed with the protein or among water molecules . following the work of kumar et al .
@xcite , mazza et al .
@xcite study the orientational correlation time @xmath145 associated with the hydrogen bond dynamics of the model is in section 2.1 .
they confirm the occurrence of a dynamic crossover at a temperature of about @xmath146 corresponding to the @xmath0 of maximum variation of the number of hydrogen bonds , that in turn corresponds to the widom line .
they also confirm that the crossover is from a non arrhenius behavior at high @xmath0 to another non arrhenius behavior @xcite , that closely resembles an arrhenius behavior around the crossover .
these results are consistent with those from simulations of other models for hydrated proteins where a crossover in the translational dynamics is observed @xcite .
the difference here is that ( i ) the crossover is for the dynamics of the hydrogen bonds , ( ii ) in the cooperative water model the crossover can be calculated from mf and an exact relation can be found between the crossover and the widom line , and ( iii ) the model can test different hypotheses . in particular ,
kumar et al . have shown that the crossover at @xmath146 is independent of whether water at very low temperature is characterized by a llcp or is sf .
in fact , the crossover is a consequence of the sharp change in the average number of hydrogen bonds at the temperature of the specific heat maximum , that occurs in both scenarios .
kumar et al .
were able also to make predictions about the @xmath15dependence of quantities characterizing the crossover at @xmath146 : ( i ) the time scale of the crossover , showing that it is independent of @xmath15 ( isochronic crossover ) ; ( ii ) the activation energy of the apparent arrhenius behavior at low @xmath0 and ( iii ) the crossover temperature , showing that both ( ii ) and ( iii ) decrease linearly upon increasing @xmath15 @xcite .
these predictions have been confirmed by chu et al .
@xcite in a study on the dynamics of a hydrated protein under moderately high pressures at low temperatures using the quasielastic neutron scattering method .
they relate these predictions ( i)(iii ) to the mechanical response of the protein to an external force , that is the average elastic constant calculated from the mean square displacement of the protein atoms . in particular , the degree of `` softness '' of the protein , related to the enzymatic activity , is preserved at lower @xmath0 if the pressure is increased @xcite . however , a criterion proposed in ref .
@xcite for discriminating which scenario better describe water on the basis of the crossover at @xmath146 , can not be tested in the experiments since the predicted difference between the two scenarios ( of the order of 1% ) is within the error bars of the measurements @xcite .
the answer to the puzzle of which of the scenarios better describe water might be related to the very recent experimental discovery of another crossover for the hydrogen bond @xmath145 to an arrhenius behavior at very low @xmath0 , of the order of @xmath147 k at hydration @xmath40 g h@xmath37o / g .
this crossover has been observed by mazza et al . at @xmath148 @xcite , in relation to an ordering process of the hydrogen bonds leading to the hdl - ldl critical point .
the study has been possible thanks to the use of a highly efficient cluster mc dynamics @xcite .
this very low @xmath0 crossover would reduce even more the @xmath0 at which the proteins preserve their `` softness '' , essential for their correct functionality . by considering partially hydrated hydrophobic plates at a distance such to inhibit the crystallization of water at low @xmath0 , franzese and de los santos @xcite
have show that water has a glassy behavior @xcite for both the translational and rotational degrees of freedom when cooled down to a low @xmath15 .
this result is consistent with simulations of tip4p water forming a quasi-2d amorphous when confined in a hydrophobic slit pore with wall - to - wall separation just enough to accommodate two molecular layers @xcite . at higher @xmath15
the hydrogen bond network builds up in a less gradual way , allowing the system to equilibrate the rotational degrees of freedom also at very low @xmath0 , but not the translational degrees of freedom .
this effect is emphasized by the appearance of many dehydrated regions @xcite , as also observed in water confined between two protein - like hydrophobic flattened surfaces at distances ranging from @xmath149 to @xmath150 nm @xcite .
when @xmath15 is close to the llcp value , the cooperativity of the hydrogen bond network induces a strong non - exponential behavior @xcite for the hydrogen bond correlation function .
however , both rotational and translational degrees of freedom equilibrate within the simulation time . at higher @xmath15
the rotational correlation function recovers the exponential behavior and the diffusion of the system allows the formation a large dry cavity , while the rest of the surface is well hydrated .
it is interesting to observe that the cooperative model allows to calculate , in the mf approximation , the diffusion constant at any @xmath0 and @xmath15 @xcite .
the hydrophobic confinement has effects also on water thermodynamics .
it shifts the hdl - ldl phase transition to lower temperature and lower pressure compared to bulk water when the confinement is between plates @xcite .
moreover , it shifts both the line of maximum density and the liquid - to - gas spinodal toward higher pressures and lower temperatures with respect to bulk when the confinement is in a hydrophobic disordered matrix of soft spheres @xcite .
this result is confirmed also in the analysis performed by using the cooperative water model in confinement between hydrophobic hard spheres @xcite .
however , the effect of the matrix on the hdl - ldl critical point is less clear and is presently under investigation .
the effect of confinement is of great interest to biology , chemistry , and engineering , yet the recent experimental and simulations results are object of an intense debate .
a better understanding of the physico chemical properties of liquid water at interfaces is important to provide accurate predictions of the behavior of biological molecules @xcite , including the folding - unfolding transitions seen in proteins @xcite , and the dynamical behavior of dna @xcite .
however , it is still unclear whether such behaviors are inherent in the structure of such molecules , or an effect of water in which they are always found , or due to the interactions between the two . to get insight into this subject the formulation of a model that allows
the development of a theory could be useful to find functional relations connecting different observables .
the advantage of this approach is to have two independent ways of approaching the problem , one theoretical and the other numerical .
we have presented here several recent results obtained with a cooperative water model suitable for studies with mean field theories and with @xmath49 , @xmath15 , @xmath0 simulations with thousands of molecules . the model has been studies in the context of water monolayers on hydrated proteins , between hydrophobic surfaces or in a hydrophobic matrix . some of the conclusions reached with this model are the following . * the different scenarios proposed to interpret the low-@xmath0 behavior of water are instances of the same mechanism , with different values of the directional ( covalent ) strength and the cooperative ( many - body ) interaction of the hydrogen bonds .
the parameters that can be estimated from the experiments suggest that the scenario with the llcp is the most plausible for water .
* previous experiments showing one dynamic crossover in the water monolayer hydrating proteins , rna and dna are consistent with ( at least ) two scenarios . *
the possibility of a second dynamic crossovers detectable at lower @xmath0 and lower hydration level would be consistent only with the llcp scenario , because its origin would be related to the ordering of the hydrogen bond network .
* a consequence of the occurrence of a llcp should be detectable when the translational and rotational dynamics of water are studied for a monolayer in a hydrophobic confinement .
in particular , the rotational dynamics should appear with the strongest non - exponential behavior in the vicinity of the llcp , as an effect of the cooperativity .
moreover , the slow increase of the number of hydrogen bonds at ow @xmath0 and low @xmath15 is the cause of the formation of an amorphous glassy state when the confinement is such that to inhibit the crystallization of water . under this conditions ,
the dehydration of hydrophobic surfaces is characterized by the appearance of heterogeneities and cavitation . *
the hydrophobic confinement affects the thermodynamics of water by lowering the @xmath0 and increasing the @xmath15 of the liquid - gas phase transition and of the tmd line .
it also affects the ldl - hdl phase transition in a way that is possibly more complex .
all these results are potentially relevant in problems such as the protein denaturation or the protein aggregation .
works are in progress to underpin and build up a theory of water at interfaces that could help us to acquire a better understanding of these subjects .
this work was partially supported by the spanish ministerio de ciencia e innovacin grants nos .
fis2005 - 00791 and fis2009 - 10210 ( co - financed feder ) , junta de andaluca contract no .
10 url # 1#1urlprefix[2][]#2 debenedetti p g and stanley h e 2003 _ physics today _ * 56 * 4046 swenson j , jansson h and bergman r 2008 _ aspects of physical biology : biological water , protein solutions , transport and replication _ ( _ lecture notes in physics _ vol 752 ) ed franzese , g and rubi , m pp 2342 stokely k , mazza m g , stanley h e and franzese g 2010 _ metastable systems under pressure _ ( springer ) chap metastable water under pressure , pp 197216 nato science for peace and security series a : chemistry and biology de los santos f and franzese g 2009 _ modeling and simulation of new materials : proceedings of modeling and simulation of new materials : tenth granada lectures _ ( _ aip conf .
_ vol 1091 ) ed marro j , garrido p l and hurtado p i ( granada ( spain ) : aip ) pp 185197 | new experiments for water at the surface of proteins at very low temperature display intriguing dynamic behaviors . the extreme conditions of these experiments make it difficult to explore the wide range of thermodynamic state points needed to offer a suitable interpretation .
detailed simulations suffer the same problem , where equilibration times at low temperature become unreasonably long .
we show how monte carlo simulations and mean field calculations of a tractable model of water help interpret the experimental results . here
we summarize the results for bulk water and investigate the thermodynamic and dynamic properties of supercooled water at an interface . |
discovery of superconductivity in layered materials or quasi - two - dimensional systems in the past several decades has brought up great interest in the physics of low dimensional superconductors . among those are the high @xmath4 cuprates,@xcite a ruthenate sr@xmath5ruo@xmath6,@xcite a cobaltate na@xmath7coo@xmath8 h@xmath5o,@xcite mgb@xmath5,@xcite a heavy electron system cecoin@xmath9,@xcite and organic conductors such as ( bedt - ttf)@xmath5x
( x is an anion).@xcite theoretically , it is well known by mermin - wagner s theorem that no off - diagonal long range order takes place at finite temperature in purely two - dimensional(2d ) systems.@xcite in the case of pure 2d , superconducting transition is expected to be of the kosterlitz - thouless ( kt ) type.@xcite the superconducting kt transition has previously been studied using finite temperature auxiliary field quantum monte carlo ( afqmc ) technique for the attractive hubbard model , that is , the hubbard model with a negative on - site @xmath10 , on a square lattice@xcite and on a triangular lattice.@xcite nowadays , a renewed interest for the kt transition on triangular lattices has arisen because unconventional superconductivity has been observed in materials having ( anisotropic ) triangular lattice structure such as na@xmath7coo@xmath8h@xmath5o and ( bedt - ttf)@xmath5x . for the square lattice
, it has been shown that the kt superconducting transition temperature @xmath11 , @xmath12 is the hopping integral , for square lattice with @xmath13 ( @xmath14 is the band filling ) away from the half filling @xmath15 , where charge density wave ( cdw ) ordering takes place.@xcite as for the triangular lattice , it has been concluded that @xmath16 takes its maximum at around half - filling due to the absence of cdw ordering , while @xmath16 is low near @xmath17,where
the density of states becomes large due to the van hove singularity at @xmath18 .
since previous studies have been restricted to small system sizes , here we revisit the problem of the superconducting kt transition of the attractive hubbard model on a triangular lattice using afqmc technique .
combining three methods of analyzing the numerical data , we find that the kt transition temperature stays almost constant within the band filling region of @xmath19.our results suggest that the estimation of the kt transition temperature using finite size scaling requires some caution when the density of states near the fermi level is small and the calculation is restricted to small system sizes .
the hamiltonian of the attractive hubbard model is given in standard notation as @xmath20 where @xmath21 is a fermion creation ( annihilation ) operator at site @xmath22 with spin @xmath23 , and @xmath24,@xmath25 denotes a pair of nearest neighbors on square or triangular lattice having periodic boundary condition.the chemical potential @xmath26 controls the band filling @xmath14 ( average number of electrons per site ) .
the hopping parameter @xmath12 is taken as the unit of the energy and is set equal to unity throughout the study .
as for the lattice structure , we mainly concentrate on the triangular lattice , but we also perform calculation on the square lattice at the band filling of @xmath13 in order to make comparison with previous studies and with the triangular lattice . the on - site attraction @xmath10 is fixed at @xmath27 on both lattices , which also enables us to make comparison with the previous studies . above the superconducting kt transition temperature @xmath16
, the on - site ( @xmath28-wave ) pairing correlation function decays exponentially , @xmath29 by contrast , the pairing correlation function exhibits a power - law decay @xmath30 where @xmath31 for @xmath32 . in the actual afqmc calculation
, we calculate a summation of the pairing correlation function , @xmath33 , given as @xmath34 due to the decaying behavior of the pairing correlation function mentioned above , @xmath33 increases as @xmath35 is increased below @xmath16 , while it remains constant above @xmath16 . in the present study ,
finite temperature afqmc is used to calculate @xmath33 .
@xcite the number of trotter decomposition slices @xmath36 is chosen to satisfy the condition @xmath37 ( @xmath38 ) , so that @xmath39 is satisfied , where @xmath40 .
@xmath41 monte carlo sweeps , depending on the temperature and system size , have been taken to assure sufficiently small statistical errors .
we have performed calculation on system sizes from @xmath42 to @xmath43 sites . in order to obtain @xmath16 from the afqmc data of @xmath33
, we use three methods , two of which are based on finite size scaling , while the other one is a more straightforward method .
as shown in eq.[eqrawcor ] , the pairing correlation decays as @xmath44 below @xmath16 for a large system size .
hence @xmath33 is proportional to @xmath45 with system size @xmath46 . in a finite size system , scaling variable @xmath47 , where @xmath48 is the correlation length , becomes more important.taking this into account , the scaling hypothesis assumes the following behavior for @xmath33 .
@xmath49 where @xmath50 is a constant and @xmath51 is a certain scaling function . since @xmath52 for @xmath32 , @xmath53 regardless of the system size @xmath35 at @xmath54
thus , if we plot @xmath55 as functions of @xmath56(or @xmath57 ) for different system sizes , they should coincide regardless of the system size at @xmath54 .
@xcite another method to identify @xmath16 is to plot @xmath55 as a function of @xmath47 , where @xmath16 and @xmath50 are chosen so that all the data points fall on a single curve ( namely @xmath58 ) regardless of the system size.@xcite however , as we shall see in the following , these scaling methods eq.[eqfss ] turns out to suffer from finite size effects when the system size is small and/or the density of states at the fermi level is small . in fact , in the former method mentioned above , @xmath33 for small system sizes do not coincide with those for large system sizes even at @xmath54 . as for the latter method ,
@xmath55 for small system sizes deviate from those for larger system sizes at low temperatures . due to this problem , here we identify @xmath16 as the temperature at which @xmath33 for the largest two system sizes coincide in the former method , while in the latter method , we search the values for @xmath16 and @xmath50 so that as many data points as possible fall on a single curve , although the data for small system sizes deviate from the curve at low temperatures .
since the scaling method suffers from finite size effects on some cases , we have adopted the third method , namely the extrapolation method .
this method has been adopted for the study of the ( real ) superconducting transition temperature for the three dimensional attractive hubbard model.@xcite since @xmath33 should increase sharply at @xmath54 in the thermodynamic limit , it is reasonable to assume that @xmath59 , where @xmath60 is the inflection point of @xmath61 for the @xmath35 site system .
@xmath60 can be obtained by fitting @xmath61 by an appropriate function . in this study
we use a gaussian form @xmath62+d_n$ ] as a fitting function , where @xmath63-@xmath64 are fitting parameters . here , we find that @xmath60 scales as @xmath65 in the range of @xmath35 considered in the present study .
we may then obtain @xmath16 by plotting @xmath60 as a function of @xmath65 and linearly extrapolating it to the @xmath66 limit .
@xmath67 plotted against @xmath57 for the square lattice .
dashed lines are guides to the eyes . ] a plot similar to fig.[sqa ] but for triangular lattice with several band filling .
band filling @xmath14 is @xmath68(a),@xmath69(b),@xmath70(c),@xmath71(d ) . ]
we first present finite size scaling analysis , where we identify the temperature at which @xmath67 coincides between different system sizes . before going into the results for the triangular lattice , we show the result for the square lattice with @xmath13 , where a scaling analysis has been performed for system sizes up to @xmath72 in ref .. in fig.[sqa ]
, @xmath67 is plotted as functions of @xmath57 for system sizes from @xmath73 to @xmath0 .
we can see that the results for the largest two system sizes , @xmath74 and @xmath0 merges at around @xmath75 .
the results for smaller systems do not cross or merge with each other within the present temperature range , but if we assume that the results for small systems are strongly affected by finite size effects , we may adopt @xmath76 for this band filling , which is roughly consistent with what has been concluded @xmath77 in ref .. we now move on to the results for the triangular lattice . in fig.[tria ] , @xmath67 is plotted as functions of @xmath57 . for @xmath17 , the results for the largest two systems , @xmath74 and @xmath0 merges at around @xmath78 .
the results for smaller systems do not cross or merge with each other within the present temperature range as in the case of the square lattice , but if we again assume that the results for small systems are strongly affected by finite size effects , we may adopt @xmath79 for this band filling . for @xmath80 ,
the results for the largest three system sizes cross at @xmath81 . again assuming that results for smaller system sizes are affected by finite size effects
, we may take @xmath82 for this band filling . as for @xmath15 , although the results for @xmath0 and @xmath72 crosses at around @xmath78 , those of the largest two sizes do not cross with each other .
therefore , @xmath83 can not be estimated from these results at this band filling .
the situation is even worse in the case of @xmath84 , where @xmath67 of the largest two system sizes do not even come close at low temperatures . here
again we can not evaluate @xmath83 for this band filling from these results .
the afqmc data of pairing correlation function @xmath33 for square lattice with @xmath13 . dashed curves are the fitting results . ]
@xmath85 plotted against @xmath65 for the square lattice with @xmath13 .
the dashed line is a least squares fit . ]
we now evaluate @xmath16 using an alternative method , that is , the extrapolation method . here again , we first show the result for the square lattice with @xmath13 . in fig.[rawpssq ] , the raw @xmath33 data is plotted as functions of temperature for each system size .
@xmath60 extracted from these data are plotted against @xmath65 in fig.[sqb ] .
the results for @xmath86 turn out to be too much affected by finite size effects , so we have omitted these data in the extrapolating process .
@xmath16 obtained by extrapolating the results to @xmath87 is @xmath88 , which is a little bit higher than , but in fair agreement with the value obtained by the scaling method . a plot similar to fig.[rawpssq ] but for the triangular lattice with @xmath84(a ) , @xmath69(b ) , @xmath70(c ) , @xmath71(d ) .
dashed lines are fitting results . ]
a plot similar to fig.[sqb ] but for the triangular lattice with @xmath15 , @xmath70 , @xmath71 . ] we move on to the triangular lattice . in fig.[rawpstri ] , the raw @xmath33 data as well as the fitting curves are plotted as functions of @xmath56 for each system size and band filling .
( we do not show the fitting curves for @xmath84 since @xmath60 turns out to be negative . )
@xmath60 obtained fromthese data are plotted against @xmath65 for each band filling in fig.[trib ] .
@xmath16 obtained by extrapolation is @xmath89 , @xmath90 , @xmath89 for @xmath17 , @xmath80 , @xmath15 , respectively . for @xmath17 and @xmath80
, @xmath16 obtained using the scaling method coincides fairly well with the values obtained here . for @xmath84
, @xmath33 turns out to _ decrease _ upon increasing the system size from @xmath72 to @xmath0 , which seems to indicate that there is no symptom of kt transition within the temperature range @xmath91 investigated in the present study , so @xmath16 , if any , should be lower than @xmath92 . in fig.[t0-n ] , @xmath16 obtained by the present method is plotted against the band filling .
the band filling dependence of the kt transition temperature @xmath16 .
solid and dashed lines are guides to the eyes . ]
@xmath67 plotted against @xmath47 for the triangular lattice . ] the kt transition temperature determined by the above method can be further justified by checking whether @xmath33 actually scales as @xmath93 with @xmath94 . in fig.[tric ] , we plot @xmath95 as functions of @xmath47 for all the system sizes .
we find for @xmath15 , @xmath70 , and @xmath71 that the numerical results for @xmath72 , @xmath74 and @xmath0 fall on a single curve by adopting the @xmath16 obtained above and choosing appropriate values of @xmath50 .
the results for smaller systems also fall on the same curves at high temperatures ( namely for large @xmath47 ) , but at low temperatures they start to deviate due to finite size effects . as for @xmath84 , if @xmath96 and @xmath97 are chosen , the results for @xmath98 and @xmath99 fall on a similar curve .
these results further justify the values of @xmath16 obtained in the preceding sections .
similar results are also obtained for the square lattice as seen in fig.[sqc ] . a plot similar to fig.[tric ] for the square lattice with @xmath13 .
our results show that the scaling method using the crossing point of @xmath67 is strongly affected by finite size effects , and that it is difficult to know from the beginning the system size necessary to obtain an accurate @xmath16 .
the present analysis suggests that this method seems to work well at band fillings where the density of states at @xmath100 is relatively large , namely at @xmath80 and @xmath17 in the present case .
there , the agreement with the results of the extrapolation method is also good .
this may be because the discreteness of the energy levels due to the finite system size is small when the density of states is large . in order to look into a possible correlation between @xmath16 and the density of states at the fermi level
, we compare the present results of @xmath16 with the transition temperature obtained by mean field approximation ( fig.[mf ] ) .
although the values of @xmath101 themselves are much larger than @xmath16 , we find a similar band filling dependence , namely , @xmath101 is almost constant for @xmath102 , and smaller for @xmath3 .
the origin of this similarity is not clear , but this does seem to suggest that @xmath16 is roughly correlated with the density of states near the fermi level at least for the value of @xmath103 adopted in the present study
. it would be an interesting future study to analyze this point from the viewpoint of the study by timm _
et al _ , where the relation between the mean field @xmath4 and the kt superconducting transition temperature in the _ repulsive _ hubbard model has been investigated by calculating the superfluid density as a function of temperature and combining it with the berezinskii - kosterlitz - thouless theory.@xcite the density of states of the square and the triangular lattices on the tight binding model . ]
the band filling dependence of the @xmath104 for the triangular lattice obtained by applying bcs mean field approximation . ]
@xmath16 for @xmath105 is still larger than @xmath16 for the square lattice at @xmath13 despite the fact that the density of states near the fermi level for the square lattice with @xmath13 is larger than for the triangular lattice with @xmath15 ( see fig.[dos ] ) .
if @xmath16 is indeed positively correlated with the density of states near the fermi level , this `` inversion '' of @xmath16 between the square and the triangular lattice may be due to the absence of charge density wave ordering in the latter lattice due to frustration , as discussed previously.@xcite
in the present study , we have investigated the superconducting kt transition of the attractive hubbard model on a two - dimensional triangular lattice using auxiliary field quantum monte carlo method for system sizes up to @xmath0 sites . combining three methods to analyze the numerical data for the pairing correlation function
, we find that the transition temperature stays almost constant within the band filling range of @xmath106 , while it is found to be much lower in the @xmath3 region . among the three methods
, the extrapolation method is found to work well regardless of the band filling , while the methods relying on finite size scaling require some caution when the density of states near the fermi level is small and the calculation is restricted to small system sizes .
we thank yoichi yanase for valuable discussions .
part of the numerical calculations has been done at the computer center , issp , university of tokyo .
99 j.g . bednorz and k.a .
mller , z. phys .
b * 64 * , 189 ( 1986 ) .
y. maeno , h. hashimoto , k. yoshida , s. nishizaki , t. fujita , j.g .
bednorz , and f. lichtenberg , nature * 372 * , 532 ( 1994 ) .
k. takada , h. sakurai , e. takayama - muromachi , f. izumi , r. a. dilanina , and t. sasaki , nature * 422 * , 53 ( 2003 ) .
j. nagamatsu , n. nakagawa , t. muranaka , y. zenitani , and j. akimitsu , nature * 410 * , 63 ( 2001 ) . c. petrovic , p.g .
pagliuso , m.f .
hundley , r. movshovich , j.l .
sarrao , j.d .
thompson , z. fisk , and p. monthoux , j. phys.:condens.matter * 13 * , l337 ( 2001 ) . for a review , see e.g. r.h .
mckenzie , comments cond .
. phys . * 18 * , 309 ( 1998 ) .
n.d . mermin and h. wagner , phys .
lett * 17 * , 1133 ( 1966 ) .
kosterlitz and d.j thouless , j. phys .
c * 6 * , 1181 ( 1973 ) .
a. moreo and d.j .
scalapino , phys .
lett * 66 * , 946 ( 1991 ) .
raimundo r. dos santos , phys .
b * 48 * , 3976 ( 1993 ) .
white , d.j .
scalapino , r.l .
sugar , e.y .
loh , j.e .
gubernatis , and r.t .
scalettar , phys . rev .
b * 40*,506 ( 1989 ) .
j. e. hirsch , phys .
b * 31 * , 4403 ( 1985 ) .
a. sewer , x. zotos , and hans beck , phys .
b * 66*,140504(r ) ( 2002 ) . c. timm , d. manske , and k. h. bennemann , phys . rev .
b * 66*,094515 ( 2002 ) . | we study the superconducting kosterlitz - thouless transition of the attractive hubbard model on a two - dimensional triangular lattice using auxiliary field quantum monte carlo method for system sizes up to @xmath0 sites . combining three methods to analyze the numerical data , we find , for the attractive interaction of @xmath1 , that the transition temperature stays almost constant within the band filling range of @xmath2 , while it is found to be much lower in the @xmath3 region . |
it is sufficient to do the calculation in the limit @xmath52 is thin , i.e. , close to weyl surface @xmath63 . for the model given in eq .
( [ hyp ] ) , such a 4d manifold @xmath52 can be given by @xmath180 , where @xmath181 , and @xmath182 .
using gamma matrices defined in @xcite , the wave function @xmath183 is given by @xmath184 while @xmath185 is well - approximated by @xmath186 where we have defined the angles @xmath187 by @xmath188 , and @xmath189 .
this approximation basically ignores the dependence of @xmath185 on @xmath190 and @xmath138 , which is valid since @xmath191 , and @xmath185 is nonsingular at @xmath63 .
the nonzero components of u(2 ) berry connection @xmath192 can then be shown to be @xmath193 with @xmath194 .
it is then straightforward to calculate the non - abelian field strengths @xmath195 .
in particular , one can prove that @xmath196/8 $ ] , which gives @xmath40 when integrated over the four angles .
therefore , the off - diagonal components of @xmath35 have no contribution to the second chern number @xmath32 .
further , one can show @xmath197 and @xmath198 are the only rest nonzero terms , where @xmath199 and @xmath200 are the u(1 ) berry connection in band @xmath5 and @xmath3 , respectively .
therefore , we have @xmath99 , and @xmath201 .
the non - abelian second chern number is then @xmath202 if we rewrite the non - abelian field strength as @xmath203 where @xmath204 ( @xmath205 are the generator of u(2 ) , the non - abelian second chern number on @xmath52 can be expressed as @xmath206 where we have defined @xmath207 . in the limit
@xmath52 is close to @xmath63 , the above calculations tell us that @xmath208 , and @xmath209 . in contrast , in the yang monopole case which is rotationally symmetric , one can show @xmath210 , and @xmath211 .
therefore , the @xmath212 symmetry breaking also breaks the symmetry between @xmath213 , @xmath214 and @xmath215 . | we study two weyl semimetal generalizations in five dimensions ( 5d ) which have yang monopoles and linked weyl surfaces in the brillouin zone , respectively , and carry the second chern number as a topological number .
in particular , we show a yang monopole naturally reduces to a hopf link of two weyl surfaces when the @xmath0 ( time - reversal combined with space - inversion ) symmetry is broken .
we then examine the phase transition between insulators with different topological numbers in 5d . in analogy to the 3d case , 5d weyl semimetals emerge as intermediate phases during the topological phase transition .
the discovery of topological states of matter has greatly enriched the variety of condensed matter in nature @xcite .
these states usually undergo phase transitions involving a change of topology of the ground state wave function , which are called topological phase transitions ( tpts ) . in three dimensions ( 3d ) ,
an significant topological state is the weyl semimetal @xcite , which plays a key role in tpts of 3d insulators .
an example is the time - reversal invariant ( tri ) transition between a noncentrosymmetric topological insulator ( ti ) @xcite and a ni in 3d , during which an intermediate tri weyl semimetal phase inevitably occurs @xcite .
another example is the tpt between different 3d chern insulators ( ci ) @xcite , where an intermediate weyl semimetal phase is also required @xcite . in both examples , the topological numbers of the insulators are transferred via weyl points of the weyl semimetal phase , which behave as `` dirac monopoles '' of the berry curvature in the brillouin zone ( bz ) .
the electrons around each weyl point obey the weyl equation , with a chirality equal to the first chern number @xmath1 of the berry curvature around the weyl point .
recently , there has been a revival of interest in gapless topological phases in higher dimensions , aimed at understanding roles of higher - dimensional topological numbers @xcite . in particular ,
the weyl semimetal can be generalized to 5d in two ways : the first is to promote weyl fermions in 3d to chiral fermions in 5d , which are described by a 4-component spinor and have a 2-fold degenerate linear energy spectrum .
the dirac monopoles associated with the weyl points in 3d become the yang monopoles in 5d @xcite , which carry a non - abelian second chern number @xmath2 of the su(@xmath3 ) berry curvature of the 2-fold degenerate valence ( conduction ) band @xcite .
the yang monopole was first introduced into condensed matter physics in the construction of the four dimensional quantum hall effect @xcite .
the second way is to keep the energy spectrum non - degenerate , while promoting the weyl points to linked 2d weyl surfaces in the 5d bz @xcite . in this case
, each weyl surface carries an abelian second chern number @xmath4 of the u(@xmath5 ) berry curvature , which is equal to the sum of its linking number with all the other weyl surfaces @xcite .
two natural questions are then whether the two 5d weyl semimetal generalizations are related , and whether they play the role of intermediate phases during the tpt of certain gapped topological states of matter in 5d . in this letter
, we show the two 5d weyl semimetal generalizations , namely , the yang monopole and the linked weyl surfaces in 5d , are closely related via the @xmath0 symmetry breaking , where @xmath6 and @xmath7 stand for time - reversal and space - inversion , respectively .
we then demonstrate they also arise as intermediate phases in the tpt between 5d ci and ni , and between 5d ti and ni with particle - hole symmetry @xmath8 that satisfies @xmath9 @xcite . in analogy to 3d cases
, the weyl arcs on the boundary of the 5d weyl semimetal @xcite naturally interpolate between the surface states of different gapped topological phases . _
yang monopoles and linked weyl surfaces_. in 3d , a weyl semimetal is known as a semimetal which is gapless at several points in the bz , i.e. , weyl points . the low energy bands near a weyl point
is generically given by a @xmath10 weyl fermion hamiltonian @xmath11 up to an identity term , where @xmath12 is the momentum , and @xmath13 ( @xmath14 ) are the pauli matrices .
the weyl point is located at @xmath15 , while the velocities @xmath16 ( @xmath14 ) play the role of light speed . by defining the u(@xmath5 ) berry connection @xmath17 of the valence ( conduction ) band wavefunction @xmath18
, one can show the first chern number of the berry curvature @xmath19 on a 2d sphere enclosing @xmath15 is @xmath20 , where @xmath21 is the sign of @xmath22 .
therefore , the weyl point @xmath15 can be viewed as a dirac monopole of the berry connection .
the first way of generalizing the weyl semimetal to 5d is to replace the weyl fermions above by the chiral dirac fermions in 5d : @xmath23 where @xmath12 is now the 5d momentum , and @xmath24 ( @xmath25 ) are the @xmath26 gamma matrices satisfying the anticommutation relation @xmath27 .
the band structure of such a hamiltonian is 4-fold degenerate at @xmath28 , and is 2-fold degenerate everywhere else with a linear dispersion .
the 2-fold degeneracy enables us to define a u(@xmath3 ) berry connection @xmath29 , where @xmath30 ( @xmath31 ) denote the two degenerate wavefunctions of the valence bands @xcite .
one can then show the non - abelian second chern number @xmath32 on a 4d sphere enclosing @xmath28 is @xmath33}{32\pi^2}=\pm1\ , \ ] ] where @xmath34 $ ] is the non - abelian u(@xmath3 ) berry curvature . in this calculation , only the traceless su(2 ) part of @xmath35 contributes .
therefore , @xmath28 can be viewed as a yang monopole in the bz , which is the source of su(@xmath3 ) magnetic field in 5d @xcite .
however , the generic 2-fold degeneracy of hamiltonian @xmath36 requires the system to have certain symmetries .
a common symmetry of this kind is the combined @xmath0 symmetry of time - reversal and inversion , which is anti - unitary and satisfies @xmath37 for fermions .
therefore , the yang monopole 5d generalization is not in the same symmetry class as that of the generic 3d weyl semimetal .
we remark here that the above 5d yang monopole , together with the 3d weyl point and the 2d dirac point ( e.g. , in graphene ) , correspond exactly to the quaternion ( pseudoreal ) , complex and real classes of the wigner - dyson threefold way @xcite , and the anti - unitary @xmath0 symmetry plays a key role in the classification .
basically , a matrix hamiltonian @xmath38 falls into these three classes if @xmath39 , respectively ( @xmath40 stands for no @xmath0 symmetry ) , and one can show @xmath41 are the corresponding spacial dimensions where point - like gapless manifold in the bz are stable .
the minimal hamiltonians of the three classes are listed in tab .
[ threefold ] .
in particular , @xmath42 is possible for systems with a negligible spin - orbital coupling such as graphene , where the electrons can be regarded as spinless . .a
hamiltonian @xmath38 with @xmath43 is in the real ( @xmath44 ) , complex ( @xmath45 ) and quaternion ( @xmath46 ) classes of the wigner - dyson three - fold way , respectively , and the hamiltonians of 2d dirac point , 3d weyl point and 5d yang monopole shown here exactly fall into these three classes . [ cols="^,^,^,^",options="header " , ] the second 5d weyl semimetal generalization requires no symmetry ( other than the translational symmetry ) , thus is in the same symmetry class with the 3d weyl semimetal .
its band structure is non - degenerate except for a few closed submanifolds @xmath47 called weyl surfaces where two bands cross each other@xcite .
the effective hamiltonian near each @xmath47 involves only the two crossing bands and takes the @xmath10 form @xmath48 .
therefore , @xmath47 is locally determined by 3 conditions @xmath49 ( @xmath14 ) . in one band @xmath50 of the two associated with @xmath47
, one can define a u(@xmath5 ) berry connection @xmath51 with its wavefunction @xmath30 , and define the u(@xmath5 ) second chern number of @xmath47 in band @xmath50 on a 4d closed manifold @xmath52 that only encloses @xmath47 as @xmath53 where @xmath54 is the berry curvature of @xmath55 .
remarkably , we showed in an earlier paper that @xcite @xmath56 where @xmath57 is the linking number between @xmath47 and @xmath58 in the 5d bz , and @xmath58 runs over all the weyl surfaces associated with band @xmath50 .
( @xmath59 ) .
( b ) the band structure and weyl surfaces in the presence of the @xmath0 breaking @xmath60 term .
( c ) yang monopole enclosed by @xmath61 shown in the 3d subspace @xmath62 .
( d ) the hopf link of weyl surfaces @xmath63 and @xmath64 shown in 3d subspace @xmath62 . ,
width=307 ] the relation between the above two 5d generalizations can be most easily seen in the following @xmath65-band model with hamiltonian @xmath66}{2}\ , \ ] ] where @xmath60 is a real parameter that breaks the @xmath0 symmetry .
when @xmath67 , the hamiltonian reduces to the yang monopole hamiltonian @xmath36 in eq .
( [ hy ] ) , where we have set all the velocities to @xmath68 .
when @xmath69 , the @xmath0 symmetry is broken , and the yang monopole necessarily evolves into linked weyl surfaces .
this can be seen explicitly by deriving the energy spectrum @xmath70^{1/2}$ ] , where we have defined @xmath71 ( @xmath25 ) . here
@xmath72 denotes the @xmath50-th band in energies .
fig .
[ yang](a ) and fig .
[ yang](b ) show the band structures for @xmath67 and @xmath69 , respectively , where @xmath73 are assumed zero . in the @xmath69 case , one can readily identify three weyl surfaces : @xmath63 between bands @xmath74 and @xmath75 , @xmath64 between bands @xmath76 and @xmath74 and @xmath77 between bands @xmath75 and @xmath78 ( see fig .
[ yang](b ) ) .
@xmath63 is a 2d sphere given by @xmath79 and @xmath80 , while @xmath64 and @xmath77 coincide and are a 2d plane given by @xmath81 .
in particular , the second band @xmath82 ( thick red line in fig . [ yang](b ) )
is associated with @xmath63 and @xmath64 , which form a hopf link in 5d as can be seen in the 3d subspace @xmath83 plotted in fig .
[ yang](d ) . in the limit @xmath84 , the radius of @xmath63 contracts to zero ,
so @xmath63 collapses onto @xmath64 ( and @xmath77 ) and becomes the 4-fold degenerate yang monopole in fig .
[ yang](c ) .
one can add other small @xmath0 breaking terms to eq .
( [ hyp ] ) , and the above picture remains topologically unchanged . due to the @xmath0 symmetry breaking , the u(2 ) gauge field @xmath85 is broken down to two u(1 ) gauge fields @xmath86 and @xmath87 in bands @xmath88 and @xmath82 .
one can easily check the abelian second chern number of @xmath63 calculated from @xmath87 is @xmath89 , which is defined on 4d manifold @xmath52 with topology @xmath90 as shown in fig .
[ yang](d ) @xcite .
this is closely related to the non - abelian second chern number @xmath91 of the yang monopole before symmetry breaking .
in fact , ignoring the gauge invariance , we can still define the u(2 ) gauge field @xmath92 using the two valence bands of hamiltonian @xmath93 , which is singular on @xmath63 but not on @xmath64 ( since @xmath64 is between the two bands defining the u(2 ) berry connection ) , and still satisfies @xmath91 on a sphere @xmath61 enclosing @xmath63 .
can be deformed adiabatically into @xmath52 in fig .
[ yang](d ) , so we also have @xmath91 on @xmath52 . to see @xmath32 is equal to @xmath94 , we can take the limit @xmath52 is a thin `` torus '' @xmath90 , i.e. , its smaller radius ( distance to @xmath63 ) tends to zero . in this limit , one will find @xmath95 , namely , the off - diagonal elements of field strength @xmath35 do not contribute ( see appendix [ a1 ] ) .
so @xmath96 is solely given by the diagonal field strengths @xmath97 and @xmath98 , which can be roughly identified with u(1 ) berry curvatures of band @xmath5 and @xmath3 . by calculations
, one can show @xmath99 .
a heuristic understanding of this is the berry curvature @xmath97 of band @xmath5 sees only @xmath64 , while the u(1 ) trace berry curvature @xmath100 sees only @xmath63 , so both of them do not see linked weyl surfaces and have zero contribution to the second chern number .
one can then readily show @xmath101 .
we note that in this limit where @xmath52 is closely attached to @xmath63 , only the diagonal elements of @xmath35 contributes , while in the yang monopole case which is spherically symmetric , the diagonal and off - diagonal elements are equally important @xcite . in high energy physics , a u(2 ) gauge symmetry can be spontaneously broken down to u(1)@xmath102u(1 ) via the georgi - glashow mechanism @xcite with an isospin @xmath5 higgs field . in 5d space ,
su(2 ) gauge fields are associated with point - like yang monopoles , while u(1 ) gauge fields are associated with monopole 2-branes ( codimension 3 objects ) .
we conjecture that a gauge symmetry breaking from u(2 ) to u(1)@xmath102u(1 ) in 5d will always break an su(2 ) yang monopole into two linked u(1 ) monopole @xmath3-branes @xmath63 and @xmath64 , where @xmath63 is coupled to one of the two u(1 ) gauge fields , while @xmath64 is coupled to both u(1 ) gauge fields with opposite monopole charges .
_ tpt in 5d .
_ it is known that 3d weyl semimetals play an important role in 3d tpts .
an example is the tpt of 3d chern insulator ( ci ) with no symmetry , which is characterized by three integers @xmath103 , with @xmath104 being the first chern number in the plane orthogonal to @xmath105 in the bz @xcite .
the ci becomes a normal insulator ( ni ) when all @xmath106 .
the tpt from a 3d ni to a @xmath107 ci involves an intermediate weyl semimetal phase as shown in fig . [
3dweyl](a)-(c ) .
by creating a pair of weyl points with opposite monopole charges and annihilating them after winding along a closed cycle in @xmath108 direction , one creates a berry flux quanta in the @xmath109-@xmath110 plane , and @xmath111 increases by one @xcite . at the same time , a fermi arc arises on the real space boundary connecting the projections of the two weyl points @xcite , which finally becomes a closed fermi loop along @xmath108 .
another example is the tpt from ti to ni , which are the two phases in the @xmath112 classification of 3d tri insulators @xcite .
when the inversion symmetry is broken , an intermediate tri weyl semimetal arises @xcite , which contains ( multiples of ) 4 weyl points as shown in fig .
[ 3dweyl](d)-(f ) .
the tpt is done by creating two pairs of weyl points with opposite charges , winding them along a loop that encloses a tri point ( e.g. , @xmath113 point ) , then annihilating them in pairs with their partners exchanged . meanwhile , the fermi surface loop of the dirac surface states of ti breaks into two fermi arcs connecting the 4 weyl points , which vanish when all the weyl points are gone . .
the blue loops are weyl surfaces between bands @xmath3 and @xmath114 , while the red loop is that between bands @xmath5 and @xmath3.,width=307 ] similarly , the 5d tpts involve creation of 5d weyl semimetal phases .
we first examine the tpt of 5d cis with no symmetry , which are characterized by 5 second chern numbers @xmath104 in the 4d hyperplanes of the bz orthogonal to @xmath105 ( @xmath25 ) , and 10 first chern numbers @xmath115 in the 2d planes parallel to @xmath105 and @xmath116 ( @xmath117 ) .
the second chern numbers @xmath104 are even under both @xmath6 and @xmath7 transformations , while the first chern numbers @xmath115 are odd under @xmath6 and even under @xmath7 . here
we shall show that changes of the five @xmath104 will involve creation and annihilation of linked weyl surfaces in the bz .
a simple example without @xmath0 symmetry is the following 4-band hamiltonian @xmath118}{2}\ , \ ] ] where @xmath119 for @xmath120 , and @xmath121 . here
@xmath122 is a tuning parameter , while @xmath123 .
we shall label each band by its order in energies , and assume the lower two bands are occupied . through an analysis similar to we did below eq .
( [ hyp ] ) , the weyl surfaces between bands @xmath3 and @xmath114 are given by @xmath124 and @xmath125 , while those between bands @xmath5 and @xmath3 ( also @xmath114 and @xmath65 ) are given by @xmath126 .
these two kinds of weyl surfaces are drawn as blue and red in the 3d subspace @xmath127 of the bz shown in fig .
[ qahi ] , respectively , where they appear as 1d loops . when @xmath128 , the system is a 5d ni with all @xmath104 and @xmath115 zero and no weyl surfaces ( fig .
[ qahi](a ) ) .
the tpt to a 5d ci with @xmath129 is driven by decreasing @xmath122 , which experiences the following stages : when @xmath130 , a weyl surface between bands @xmath3 and @xmath114 arise around the origin , which is topologically a 2d sphere in the @xmath131 hyperplane ( fig .
[ qahi](b ) , ( c ) ) . when @xmath132 , as shown in fig .
[ qahi](d ) , the 2d sphere between band @xmath3 and @xmath114 splits into two smaller spheres @xmath63 and @xmath133 ( blue ) in the @xmath131 hyperplane , while another 2d sphere weyl surface @xmath64 ( red ) between bands @xmath5 and @xmath3 is created in @xmath134 plane , which is linked to both @xmath63 and @xmath133 .
as @xmath122 is further decreased , @xmath63 and @xmath133 will move along @xmath135 , respectively , and finally merge into a single weyl surface when @xmath136 ( fig .
[ qahi](e ) ) .
this weyl surface then shrinks to zero , and the system becomes a 5d ci with @xmath129 for @xmath137 , leaving a cylindrical weyl surface @xmath64 between bands @xmath5 and @xmath3 ( also one between bands @xmath114 and @xmath65 , fig . [ qahi](f ) ) .
we note that if @xmath67 , the @xmath0 symmetry is restored , and the two blue weyl surfaces @xmath63 and @xmath133 will collapse into two yang monopoles of opposite monopole charges @xmath32 .
the tpt process then becomes the creation , winding and annihilation of two yang monopoles .
this tpt is also accompanied with a surface state evolution from trivial to nontrivial .
it has been shown @xcite that a 5d weyl semimetal with linked weyl surfaces contain protected weyl arcs in the 4d momentum space of surface states , which have linear dispersions in the other @xmath114 directions perpendicular to the arc . by taking an open boundary condition along @xmath108 direction ,
one can obtain a weyl arc on the 4d boundary connecting the projections of the two weyl surface hopf links ( fig .
[ qahi](d ) , ( e ) ) .
when the system becomes a ci , the weyl arc develops into a non - contractible weyl loop along @xmath138 as expected .
the second example is the tpt between @xmath0 breaking 5d insulators with particle - hole symmetry @xmath8 satisfying @xmath9 , which are shown to be classified by @xmath112 into 5d tis and nis @xcite . here
we consider an eight - band model hamiltonian of a 5d ti as follows : @xmath139+h_a\\ = \sum_{i=1}^6\zeta_i(\mathbf{k})\gamma^i+h_a\ , \ ] ] where @xmath140 ( @xmath141 ) are the @xmath142 gamma matrices so chosen that @xmath143 , @xmath144 , @xmath145 , @xmath146 are real and @xmath147 , @xmath148 , @xmath149 are imaginary , @xmath150 for @xmath25 , @xmath151 with @xmath152 , and @xmath153 is a symmetry breaking perturbation .
the @xmath6 , @xmath7 and @xmath8 transformation matrices are given by @xmath154 , @xmath155 and @xmath156 , and a hamiltonian @xmath38 will have these symmetries if @xmath157 , @xmath158 and @xmath159 , respectively .
it is then easy to see that @xmath160 respects the @xmath8 symmetry but breaks @xmath6 , @xmath7 and @xmath0 symmetries . in particular , only @xmath161 and @xmath162 breaks the @xmath0 symmetry . and
@xmath163 .
the small blue loops are the weyl surfaces at the fermi energy between bands @xmath65 and @xmath164 , while the large red loop is that between bands @xmath114 and @xmath65 .
( g)-(h ) the evolution of surface states from a weyl ring to two weyl arcs during the tpt.,width=307 ] in the absence of @xmath160 , the system is a 5d ni if @xmath165 , and is a 5d ti if @xmath166 . with the symmetry breaking term @xmath160 ,
the calculation of the band structure of @xmath167 becomes more complicated . for simplicity
, we shall only examine the limiting case where @xmath168 , @xmath169 , @xmath170 and @xmath171 .
we shall label each band by its order in energies , and keep the fermi energy at zero , i.e. , between band @xmath65 and @xmath164 , as required by the @xmath8 symmetry . to a good approximation ,
the weyl surfaces between bands @xmath65 and @xmath164 are given by @xmath172 , @xmath173 and @xmath174 , while those between bands @xmath164 and @xmath175 ( also between @xmath114 and @xmath65 ) are given by @xmath176 . the tpt can be driven by tuning @xmath122 from negative ( ti ) to positive ( ni ) , and the evolution of these low energy weyl surfaces are illustrated in fig .
[ tini ] ( a)-(f ) in the 3d subspace @xmath177 and @xmath163 .
the small blue loops are the images of weyl surfaces between bands @xmath65 and @xmath164 , while the red loop at @xmath83 is that between bands @xmath164 and @xmath175 . at first ,
two pairs of blue weyl surfaces arise unlinked ( panel ( b ) ) . as @xmath122 increases
, they merge into four new weyl surfaces linked with the red weyl surface , which then wind around the red weyl surface and merge into unlinked pairs again with their partners exchanged ( panels ( c)-(e ) ) .
finally , the four unlinked blue weyl surfaces contract to zero , and the system becomes a 5d ni . similar to the ci case , if @xmath178 , the @xmath0 symmetry is recovered , and the 4 blue weyl surfaces will collapse into 4 yang monopoles .
the tpt process then involves winding of yang monopoles instead of linked weyl surfaces .
the topological surface states of the system also involves a topological transition during the tpt .
the topological surface states of a noncentrosymmetric 5d ti is generically a `` weyl ring '' as shown in fig.[tini](g ) .
the tpt then breaks into two weyl arcs ( panel ( h ) ) , which finally vanish when entering the ni phase . in conclusion , we show that 5d weyl semimetals with yang monopoles are protected by the @xmath0 symmetry , and generically reduce to 5d weyl semimetals with linked weyl surfaces in the presence of @xmath0 symmetry breaking .
we therefore expect that yang monopoles generically break into linked u(1 ) monopole 2-branes in 5d theories of gauge symmetry breaking from u(2 ) to u(1)@xmath102u(1 ) .
as gapless states carrying the second chern number , they emerge as intermediate phases in the tpts between cis and nis or between tis and nis in 5d space , generalizing the connection between gapless and gapped topological phases in 3d @xcite . _
acknowledgments_. we would like to thank jing - yuan chen for helpful discussions during the research .
this work is supported by the nsf under grant number dmr-1305677 . |
this work was supported in part by the u.s .
department of energy under grant no .
doe91er40651 taskb .
[ 1 ] t. kinoshita , _ quantum electrodynamics _ , t.k
. editor , ( world scientific , 1990 ) , pp .
it should be stressed that one should not confuse our @xmath4 here with the limits on the lepton or quark form factors in qcd . in axiomatic qft
even the deuteron , which certainly is composite , has an interpolating field which is strictly causal , i.e. two field operators at x and y commute if @xmath124 , i.e. spacelike .
this leads to dispersion relations for forward @xmath125 scattering .
if @xmath126 does not satisfy strict causality , this will have little to do with whether @xmath127 or the electron are composite or not .
[ 2 ] n.n .
khuri , proceedings of les recontres de physiques da la vallee daoste : _ results and perspectives in particle physics " _ ( m. greco , ed . ) , pp * 701 - 708*. editions frontieres , gif - sur - yvette , france , 1994 . | we show that , even for a soft collision like forward elastic scattering , the phase of the amplitude is extremely sensitive to a breakdown of strict causality in local @xmath0 .
this is especially the case when the breakdown is manifested by a failure of polynomial boundedness which leads to amplitudes that in some complex direction have order 1 exponential growth in @xmath1 .
-0.24 truecm -0.24 truecm -2.94 truecm 24.94 truecm 16.89 truecm 1.5 truecm 4 truecm on the importance of measuring @xmath2 at @xmath3 n.n .
khuri _ department of physics _ _ the rockefeller university , new york , new york 10021 _ 18 pt at present the best experimental limit on the existence of a fundamental length " , signifying a breakdown in the local nature of quantum field theory ( qft ) , comes from ( g-2 ) calculations and experiments in qed .
if there is a fundamental length , @xmath4 , then from the results on the muon ( g-2 ) we have @xmath5 this leads us to the following estimate @xmath6 with model dependent arguments one can accommodate an @xmath4 such that @xmath7 , but not much better .
however , the following assertion can be safely made : _ today , we have no experimental evidence that can rule out the existence of a fundamental length , " @xmath4 , such that @xmath8_.@xmath9 one should add that we have already reached the end of the road as far as learning more from qed and ( g-2 ) regarding this issue . at the level of @xmath10
both electroweak and hadronic contributions to ( g-2 ) become important .
the forward dispersion relations for @xmath11 and @xmath12 scattering represent one of the few general rigorous consequences of local quantum field theory .
one tests their validity by measuring @xmath13 , and comparing the results with the calculated @xmath2 obtained from the dispersion relations with @xmath14 as an input .
these tests have been carried out at increasing energies for both @xmath11 , and @xmath12 .
the highest energy , @xmath12 only , is @xmath15 .
the agreement between theory and experiment is good , as can be seen from figure 1 .
a measurement of @xmath2 at @xmath16 will explore a short distance domain about which we have little previous knowledge .
this is in contrast to the preceeding measurements which dealt with length scales that had already been pre - explored by qed . however , there is a problem .
the particle physics folklore includes statements to the effect that short distance structure shows up first , and in more easily detectable ways , in , and not in such as forward elastic scattering .
unfortunately , when statements like the above are repeated for 35 years , people get to be rather sloppy in using them . in this brief talk
i will clarify this problem and show that in a domain where the dispersion relations no longer hold , @xmath2 is extremely sensitive to the existence of a fundamental length , @xmath4 , and changes drastically even when @xmath17 .
the dilemma between hard and soft collisions can be clarified as follows . given some kind of structure at short distance , @xmath4
, we can make the following correct statement : _ as we approach the energy region such that @xmath18 is non - negligible , then hard cross - sections will in general show a more detectable and larger change than soft cross - sections_. this statement , though not extremely precise , is in general qualitatively correct .
however , if in the same statement we replace by then the statement is false , as we shall demonstrate by counterexample . of course when local qft is valid , then for the forward scattering amplitude , @xmath19 , if @xmath20 does not change much in a certain region so will @xmath21 and @xmath2 because one can use the dispersion relation to get @xmath2 from @xmath14 .
but , in a world in which the dispersion relations or qft fail , @xmath2 as we shall show below can be extremely sensitive to a breakdown in qft .
the main purpose of this talk is to give a counterexample " to the statement about soft amplitudes being insensitive to a short distance breakdown in qft .
indeed , i will show that when traditional qft is changed by a breakdown in polynomial boundedness , which in many models is the main feature of the existence of a fundamental length , then the phase , @xmath2 , could change by more than a factor of 2 even for energies as low as @xmath22 .
this happens without any significant change in @xmath14 or @xmath23 .
before we present our counterexample " , we summarize briefly the well established properties of @xmath19 in local qft .
we have the following : i ) @xmath19 is analytic in the doubly cut s - plane ; ii ) @xmath19 is crossing symmetric , relating @xmath11 for @xmath24 to @xmath12 for @xmath25 ; iii ) @xmath26 satisfies the optical theorem , @xmath27 ; iv ) @xmath19 is polynomially bounded in all directions in the cut complex s - plane , @xmath28 we have argued in ref .
2 , that the most likely property to change is iv ) , i.e. equation ( 3 ) . properties ii ) and iii ) are on a solid footing .
while even if property i ) , analyticity , fails due to some new complex singularities on the physical sheet , these will not lead to a strong signal except for @xmath29 ( new singularities ) . the breakdown in iv )
replaces equ .
( 3 ) by an exponential bound : @xmath30 in fact we claim more .
namely , that along some complex s direction , @xmath31 grows exponentially like @xmath32 .
note that a behavior of the type @xmath33 , with @xmath34 , is excluded due to some technical mathematical arguments .
hence one goes from @xmath35 behavior to a first order exponential in @xmath36 .
there are several examples which exhibit the behavior ( 4 ) .
these are : a. ) non - local potential scattering here one replaces the interaction term in the schrodinger equation , by a non - local one , i.e. , @xmath37 where @xmath38 b. ) some nonlocal field theories .
c. ) string - theory@xmath39 ( @xmath40 once we have exponential behavior in @xmath41 we have a natural way to a fundamental length " .
noting that @xmath42 momentum for large @xmath43 , then the @xmath4 we use in the bound to make a dimensionless exponent is by definition our fundamental length .
we take the smallest @xmath4 such that for some complex sequence @xmath44 , as @xmath45 , @xmath26 actually grows like @xmath46 .
our main counterexample is a representative of what happens in non - local potential scattering as defined by equ .
( 5 ) and ( 6 ) .
it also mimics the behavior of some non - local field theories , although these admittedly have serious problems . in non - local potential scattering one
can easily prove the analyticity of @xmath47 , but @xmath26 is not polynomially bounded as @xmath48 .
however , one can prove that @xmath49 as @xmath48 in all directions .
the physical sheet here corresponds to @xmath50 , and @xmath4 is defined in equ.(6 ) .
these facts lead us to the following ansatz : @xmath51 where @xmath52 is the true amplitude , and @xmath53 is a false " amplitude defined by ( 7 ) .
the true amplitude satisfies the optical theorem , @xmath54 , but @xmath55 is not necessarily positive .
however , by definition it is @xmath56 that is polynomially bounded .
also @xmath56 satisfies the dispersion relations albeit with a non - positive @xmath55 .
finally , @xmath57 has no dispersion relation because of its exponential growth when @xmath58 , and @xmath59 .
at first the ansatz(7 ) looks like a tautology , or at best a definition of @xmath53.nevertheless , because of the special properties of @xmath2 for @xmath60 we can still learn much from this ansatz .
we consider two energy regions .
a low " energy region @xmath61 , and a transitional " region @xmath62 .
we also note the fact that for @xmath63 , @xmath2 is small and decreases logarithmically . at @xmath64 , we have @xmath65 .
given the fact that cosmic ray data tell us that @xmath14 will continue to increase , @xmath66 , the dispersion relation fit shown by the curve in figure 1 , will continue to decrease slowly as @xmath41 increases .
we consider the two energy regions separately . a. low energy region , @xmath67 in this region , we have @xmath68 there is no observable difference between the polynomially bounded amplitude and the exponentially bounded one . b. transitional region here
@xmath69 is small but not negligible , @xmath70 note that @xmath41 is still well below @xmath71 . from the ansatz(7 ) , we obtain @xmath72,\ ] ] at this stage it is important to remark that the second term in the bracket in ( 10 ) is doubly small , @xmath73 is small and @xmath74 .
we obtain @xmath75.\ ] ] to proceed further let us assume that in the transitional region , @xmath76 , a bound almost 2.5 times larger than the value of @xmath2 at @xmath77 .
we hasten to add that this assumption is not really needed , but it gives us a quick way to arrive at our result .
later we shall show how one can get the same result without this assumption .
we now have , @xmath78,\ ] ] for @xmath79 .
thus in the transitional region @xmath80 , and hence @xmath81 for @xmath82 and the optical theorem holds for @xmath55 to a very good approximation . but @xmath53 satisfies a dispersion relation , and we can therefore use the approximate derivative form of the dispersion relations@xmath83 to get @xmath84 where @xmath85 is the dispersion relation fit to @xmath2 as the one shown by the continuous curve in figure 1 . the true " and false " phases are related by @xmath86 with @xmath87 .
after some simple algebra we get @xmath88 but @xmath89 in the transitional region . replacing @xmath90 by @xmath66 in ( 15 ) and expanding in powers of @xmath91 , we get @xmath92 the error , @xmath93 , is determined by @xmath94 there are two important features of our final result ( 16 ) which should be stressed .
first , while @xmath66 decreases logarithmically the additional term @xmath95 increases linearly with @xmath41 , and even when @xmath96 it leads to a 75% increase in @xmath2 .
the second feature is the fact that @xmath4 appears linearly in ( 16 ) , unlike the correction to the muon ( g-2 ) in equ .
( 1 ) which was @xmath97 , with @xmath98 appearing in the correction .
since @xmath4 is very small , this makes the qed case less sensitive to a fundamental length . in figure 1
we plot @xmath73 for @xmath99 .
the fact that @xmath2 is sensitive is amply demonstrated . in closing
, we explain briefly how our assumption on @xmath100 can be removed .
one has to divide the interval , @xmath101 , into @xmath102 small intervals , @xmath103 .
then one carries out the same calculation we did above step by step starting at @xmath104 , and calculating @xmath73 at the next point . at each step one uses the approximation @xmath105 . the result given by equ.(16 )
remains the same .
there are other examples which lead to a decrease in @xmath2 rather than the increase given by the ansatz ( 7 ) .
for example one could have instead of ( 7 ) , @xmath106 this leads to @xmath107 with the result shown in figure 1 for @xmath108 , and labeled as case ii .
our ansatz is an example of a whole class where one can write @xmath109 , where @xmath110 is an entire function of order 1 .
all these lead , via similar arguments , to a dramatic change in @xmath2 , @xmath111 where @xmath112 . in conclusion , @xmath2 is extremely sensitive to a short distance breakdown of qft , and more so if that breakdown is manifested by the failure of polynomial boundedness and its replacement by exponential growth which is order 1 in the variable @xmath113 . here
@xmath114 is the c.m .
momentum in forward scattering , and @xmath4 we take , by definition , to be the fundamental length " . strictly within the context of this picture
, we can set the following limits on a breakdown in polynomial boundedness : 1 . ) @xmath115 gives us a lower limit for @xmath116 , @xmath117 . 2 .
) @xmath118 if redone with smaller errors could improve this bound to @xmath119 ; here @xmath120 . 3 .
) agreement between @xmath121 and @xmath122 at lhc could get us @xmath123 . |
core collapse supernovae ( sne ) mark the spectacular death of massive stars . among these sne types
ib and ic ( collectively referred to as ib / c due to their perceived similarities ) are associated with the explosion of massive stars which were stripped of their h envelopes by strong stellar winds .
types ib show he features in their spectra , whereas types ic do not .
our current understanding of type ib / c sne suggests that wolf - rayet ( wr ) stars are a primary progenitor path @xcite , although the relative numbers of wr stars fall well short of the observed type ib / c rates and thus a lower - mass binary progenitor path is most likely required to account for a significant fraction of these events ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
archival imaging of pre - sne locations , however , have failed to produce direct associations between type ib / c sne and their possible progenitors .
thus we must rely both on theory and indirect information obtained after the explosion .
fortunately , x - ray and radio emission are detectable by - products of the shock interaction between the high - velocity ejecta and the low - velocity progenitor wind in sne ( whereas the optical emission is generated by lower - velocity ejecta and radioactive decay ) .
these observables crudely scale with the density of the circumstellar medium ( csm ) , and hence provide constraints on the wind properties of the sn progenitor ( e.g. , * ? ? ?
* ; * ? ? ?
* ) . because massive stars are expected to have strong episodic mass - loss ( e.g. , * ? ? ?
* ) and line driven winds ( e.g. , * ? ? ?
* ) , radio monitoring of type ib / c sne provide interesting constraints on possible progenitor types .
the past decade has witnessed increased attention on type ib / c sne due to the association of type ic sn 1998bw with gamma - ray burst ( grb ) 980425 @xcite , as these objects may help elucidate how the central engines of both grbs and their associated sne work . following this historic connection ,
several other type ic sne have been linked to grbs ; all of these objects show broad emission lines in their optical spectra , with velocities above @xmath10 . to date
so far 5 grbs have been spectroscopically confirmed to have a sne counterpart , while 14 show photometric evidence of a sne counterpart @xcite , and hence a larger sample is needed in order to shed some light on the explosion mechanisms involved . given the large number of sne type ic compared to grbs
, one key question is whether these sne could simply be off - axis grbs .
initial efforts concluded that the vast majority were not ( e.g. , * ? ? ? * ) .
recently new evidence was presented for the detection of at least one possible off - axis grb with no direct observation of the gamma - ray emission : sn 2009bb @xcite ; as well as the more controversial sn 2007gr @xcite . here
we report on a particularly interesting broad - lined type ic , sn 2007bg , which was optically discovered on 2007 april 16.15 ut at @xmath11 , @xmath12 ( j2000 ) at redshift @xmath13 @xcite ( @xmath14 mpc , @xmath15 km s@xmath16 mpc@xmath16 , @xmath17 , and @xmath18 in a @xmath19cdm cosmology ) . a spectrum taken on 2007 april 18.3 with the 9.2 m hobby - everly telescope
let @xcite classify sn 2007bg as a type ic broad - lined ( ic - bl ) supernova .
the classification of this sn as a type ic and the fact that it resides in a faint host encouraged @xcite to think about this object as a likely off - axis grb .
@xcite also noted that sn 2007bg was one of the brightest radio sne one year after explosion , making it an even better candidate for an off - axis grb .
however , based on the radio light curve of sn 2007bg known at the time , @xcite contended that the observed radio emission could be explained by the presence of a density enhancement in the csm , rather than a grb like central engine driving the explosion . using the archival radio data for sn 2007bg
, we further address this issue below .
the article outline is the following . in
[ s : obs ] we introduce the observations taken with the very large array ( vla ) .
our results and modelling of the radio light curves are presented in
[ s : results ] . finally in
[ s : discussion ] we summarise our findings and provide a brief discussion on the future of radio observations of sne .
the vla radio observations were carried out between 2007 april 19 and 2009 august 28 as part of the programs as887 , as929 and as983 ( pi a. m. soderberg ) .
the data were taken in standard continuum observing mode , with a bandwidth of @xmath20 mhz .
data were collected at @xmath21 , @xmath22 , @xmath23 , @xmath24 , @xmath25 and @xmath26 ghz . as primary flux density calibrators 3c286 and 3c147 were used .
we used these calibrators to scale our flux density measurements to the perley - buttler 2010 absolute flux scale . for secondary calibrators
three different sources were used ; sdss j114856.56 + 525425.2 in most epochs , as well as sdss j114644.20 + 535643.0 and sbs 1150 + 497 .
the derived flux density for the secondary calibrators is shown in table [ tab : phase ] .
traditionally one has been able to cross - check the derived secondary calibrator fluxes with archival data , but in this case the vla stopped regular monitoring of calibrators since 2007 .
this , in combination with the weak calibrator for the @xmath24 and @xmath25 ghz data , makes this data less reliable than the rest at lower frequencies , even adopting larger error bars for it .
lccccc obs .
& @xmath27 & @xmath28 & @xmath29 & @xmath30 & @xmath31 + date & ( jy ) & ( jy ) & ( jy ) & ( jy ) & ( jy ) + 2007-apr-19 & & & 0.4525 & & + 2007-apr-23 & & & 0.4306 & & 0.2732 + 2007-apr-24 & & 0.4571 & 0.433 & & + 2007-apr-25 & & & 0.453 & & 0.282 + 2007-apr-26 & & 0.4552 & 0.4521 & 0.3644 & + 2007-apr-30 & & 0.467 & 0.4583 & & + 2007-may-3 & & & & 0.355 & 0.2752 + 2007-may-5 & & 0.458 & 0.4427 & & + 2007-may-12 & & 0.412 & 0.396 & 0.3607 & 0.2733 + 2007-may-17 & & 0.4582 & 0.4623 & 0.3590 & 0.268 + 2007-may-27 & 0.4533@xmath32 & 0.4507 & 0.4173 & 0.345 & 0.2784 + 2007-jun-11 & & 0.45 & 0.464 & 0.359 & 0.2781 + 2007-jun-22 & & 0.4587 & 0.4372 & & + 2007-jul-3 & & & & 0.30 & 0.29 + 2007-jul-7 & & 0.4599 & 0.4479 & & + 2007-jul-24 & 0.465@xmath32 & 0.453 & 0.4462 & 0.349 & 0.266 + 2007-aug-18 & & 0.4545 & 0.4410 & 0.348 & 0.2720 + 2007-sep-7 & 0.465@xmath32 & 0.464 & 0.4475 & 0.352 & 0.30 + 2007-sep-23 & 0.476@xmath32 & 0.4668 & 0.4624 & & + 2007-oct-23 & 0.459@xmath32 & 0.4514 & 0.4410 & & + 2007-nov-17 & 0.4562@xmath32 & 0.4332 & 0.4449 & & + 2007-dec-23 & 0.452@xmath32 & 0.4583 & 0.4513 & & + 2008-jan-3 & & & & & 0.2685 + 2008-jan-28 & & 0.50 & 0.475 & & 0.10 + 2008-feb-25 & & 0.4642 & 0.4561 & & 0.304 + 2008-apr-19 & 0.476@xmath32 & 0.6247@xmath32 & 0.587@xmath32 & & + 2008-may-7 & & & 0.449 & & + 2008-jun-9 & & 1.197@xmath33 & 1.05@xmath33 & & 1.050@xmath33 + 2008-nov-3 & 0.5160@xmath32 & 0.6480@xmath32 & 0.6069@xmath32 & & 0.491@xmath32 + 2008-dec-30 & & & 0.481 & 0.351 & + 2009-apr-6 & 0.5100@xmath32 & 0.575@xmath32 & 0.5430@xmath32 & & 0.3491@xmath32 + 2009-may-31 & 0.503@xmath32 & 0.5763@xmath32 & 0.523@xmath32 & & + 2009-aug-27 & 0.530@xmath32 & 0.5411@xmath32 & 0.5162@xmath32 & & + + + + + + during this period of time , the vla incorporated the new evla correlator to some of the antennas , so some of the antennas did not have all of the different wavelength receivers .
this greatly reduced the sensitivity of observations during some epochs , especially at @xmath21 and @xmath24 ghz .
we flagged , calibrated , and imaged the observations using the casa ( common astronomy software applications ) software following standard procedures .
we combined vla and evla baselines , and to reduce the effects of the differences between vla and evla antennas , we derived baseline - dependent calibrations on the primary flux calibrator . to measure the flux density of the source we fitted a gaussian at the source position in our cleaned images , as for upper limits these were computed from the dirty map _ rms _ in the central region . since the _ rms _ noise in each cleaned map only provides a lower limit on the total error we added a systematic uncertainty @xmath34 due to possible inaccuracies of the vla flux density calibration and deviations from an absolute flux density scale resulting in a final error for each flux density measurement of @xmath35 , with @xmath36 at @xmath26 , @xmath25 , @xmath24 , @xmath23 , @xmath22 , and @xmath21 ghz .
the factors were chosen based on the quality of the secondary calibrator , and on the observed phase scatter after calibration of the data .
the resulting flux density measurements of sn 2007bg are shown in table [ tab : flux ] along with the image _
rms_. lcccccccc obs .
& @xmath37 & @xmath27 & @xmath28 & @xmath29 & @xmath30 & @xmath31 & @xmath38 & array + date & ( days)@xmath39 & ( @xmath40jy ) & ( @xmath40jy ) & ( @xmath40jy ) & ( @xmath40jy ) & ( @xmath40jy ) & ( @xmath40jy ) & configuration + 2007-apr-19 & 2.9 & & & @xmath41264 & & & & d + 2007-apr-23 & 7.0 & & & @xmath41348 & & @xmath411421 & & d + 2007-apr-24 & 7.8 & & @xmath41390 & @xmath41387 & & & & d + 2007-apr-25 & 8.8 & & & @xmath41417 & & 1293 @xmath42 220 & & d + 2007-apr-26 & 9.8 & & @xmath41339 & @xmath41555 & @xmath41666 & & & d + 2007-apr-30 & 13.8 & & @xmath41291 & 480 @xmath42 102 & & & & d + 2007-may-03 & 16.8 & & & & @xmath41978 & 1032 @xmath42 75 & & d + 2007-may-05 & 19.2 & & @xmath411692 & 753 @xmath42 107 & & & & d + 2007-may-12 & 26.1 & & @xmath411905 & 804 @xmath42 100 & 1900 @xmath42 300 & 1062 @xmath42 245 & & d + 2007-may-17 & 30.9 & & @xmath41657 & 728 @xmath42 87 & 1480 @xmath42 289 & @xmath411050 & & d + 2007-may-27 & 41.3 & @xmath41171 & 487 @xmath42 96 & 1257 @xmath42 53 & 1250 @xmath42 270 & @xmath411020 & & a + 2007-jun-11 & 55.9 & & @xmath411191 & 1490 @xmath42 104 & 2000 @xmath42 308 & 1246 @xmath42 127 & & a + 2007-jun-22 & 66.8 & & 709 @xmath42 60 & 1390 @xmath42 45 & & & & a + 2007-jul-03 & 77.8 & & & & @xmath41843 & @xmath41978 & & a + 2007-jul-07 & 81.8 & & 915 @xmath42 62 & 1325 @xmath42 47 & & & & a + 2007-jul-24 & 98.8 & @xmath41201 & 886 @xmath42 80 & 1131 @xmath42 59 & @xmath41909 & @xmath41870 & & a + 2007-aug-18 & 124.0 & & 1050 @xmath42 90 & 957 @xmath42 64 & @xmath41957 & @xmath41693 & & a + 2007-sep-07 & 144.0 & @xmath41336 & 1103 @xmath42 88 & 621 @xmath42 81 & @xmath411137 & @xmath41825 & & a + 2007-sep-23 & 159.8 & @xmath41525 & 792 @xmath42 77 & 316 @xmath42 57 & & & & anb + 2007-oct-23 & 189.9 & @xmath41585 & 685 @xmath42 62 & 379 @xmath42 59 & & & & anb + 2007-nov-17 & 214.9 & @xmath41342 & 479 @xmath42 65 & 404 @xmath42 62 & & & & b + 2007-dec-23 & 250.9 & @xmath41387 & 685 @xmath42 65 & 783 @xmath42 57 & & & & b + 2008-jan-03 & 261.8 & & & & & 1723 @xmath42 77 & & b + 2008-jan-05 & 263.8 & & & & & & @xmath411662 & b + 2008-jan-28 & 286.8 & & @xmath41594 & 1669 @xmath42 90 & & @xmath41999 & & b + 2008-feb-25 & 314.8 & & 1059 @xmath42 70 & 2097 @xmath42 73 & & 2570 @xmath42 160 & & cnb + 2008-apr-19 & 368.8 & @xmath41822 & 1304 @xmath42 62 & 2200 @xmath42 59 & & & & c + 2008-may-07 & 386.8 & & & 2852 @xmath42 58 & & & & c + 2008-jun-09 & 419.9 & & 2039 @xmath42 88 & 3344 @xmath42 81 & & 2023 @xmath42 140 & & dnc + 2008-nov-03 & 566.9 & @xmath41321 & 2812 @xmath42 60 & 3897 @xmath42 78 & & 1854 @xmath42 128 & & a + 2008-dec-30 & 623.8 & & & 3891 @xmath42 78 & @xmath412358 & & & a + 2009-apr-06 & 720.8 & 746 @xmath42 168 & 3390 @xmath42 136 & 3842 @xmath42 69 & & 1629 @xmath42 115 & & b + 2009-may-31 & 775.8 & @xmath411167 & 3460 @xmath42 80 &
3641 @xmath42 79 & & & & cnb + 2009-aug-27 & 863.8 & @xmath411503 & 3373 @xmath42 84 & 3408 @xmath42 109 & & & & c + + sn 2007bg was observed seven times with the _ swift _ x - ray telescope ( xrt ) and once with the _ chandra _ x - ray observatory , as listed in table [ tab : xray ] .
processed data were retrieved from the _ swift _ and _ chandra _ archives , and analysis was performed using the heasoft ( v6.12 ) and ciao ( v4.4.1 ) software packages , along with the latest calibration files available at the time . in the case of _ chandra
_ , we reprocessed the data to apply the latest calibration modifications , apply positional refinements , and correct for charge transfer inefficiency ( cti ) .
all datasets were then screened for standard grade selection , exclusion of bad pixels and columns , and intervals of excessively high background ( none was found ) .
the seven _ swift _ datasets were combined into three rough epochs .
we adopted a 35radius aperture ( corresponding to @xmath4385% encircled energy ) as the source extraction region and modelled the local diffuse and scattered background using an annulus of 35100 .
no statistically significant detection was found within our adopted aperture .
for the _ chandra _ dataset , we adopted a 15 radius aperture ( corresponding to @xmath4395% encircled energy ) as the source extraction region and modelled the local diffuse and scattered background using an annulus of 1.510 .
although five ( 5 ) counts fell within our aperture ( a @xmath44 signal ) , no statistically significant detection could be made .
upper limits in all cases were determined using the bayesian technique from @xcite .
we used pimms v4.6 to determine a count - to - flux conversion , adopting an absorbed apec thermal plasma model with @xmath45 kev and @xmath46 @xmath47 @xcite .
lrrcp0.9cmcc date@xmath39 & instrument & obsid & exposure & 0.5 - 8.0 & @xmath48 & @xmath49 + & & ( ks ) & & counts & ( erg s@xmath16 @xmath47 ) & ( erg s@xmath16 ) + 2007-apr-22@xmath50 & _ swift _ xrt & 00030920001 & 9.4 & @xmath514.6 & @xmath52 & @xmath53 + 2007-apr-30@xmath54 & _
chandra _ acis - s3 & 7643 & 34.0 & @xmath5111.2 & @xmath55 & @xmath56 + 2008-mar-05@xmath57 & _ swift _ xrt & 00030920002 & 2.8 & & & + 2008-mar-17@xmath57 & _ swift _ xrt & 00030920003 & 4.2 & & & + 2008-apr-23@xmath57 & _ swift _ xrt & 00030920004 & 2.9 & & & + 2009-jun-09@xmath58 & _ swift _ xrt & 00030920005 & 3.5 & & & + 2009-jun-15@xmath58 & _ swift _ xrt & 00030920006 & 3.1 & & & + 2009-jun-17@xmath58 & _ swift _ xrt & 00030920007 & 2.0 & & & + +
the resulting light curves for sn 2007bg are shown in fig . [
f : lc ] . the evolution of the light curves can be divided in three phases .
_ phase 1 _ : this phase starts after the explosion and lasts until day @xmath59 . here we observe the typical turn - on of radio supernovae ( rsne ) , first starting at shorter wavelengths and cascading to longer wavelengths with time as wavelength - dependent absorption becomes less important .
the peak luminosity reached during phase 1 is @xmath60 observed on day @xmath61 at @xmath23 ghz .
this makes sn 2007bg one of the most luminous type ib / c rsne on comparable time - scales .
the first detections at @xmath22 and @xmath23 deviate from the rest of the evolution , and it could be due to interstellar scintillation , a real deviation from pure ssa due to clumpy csm , or deceleration ; unfortunately there is not enough early - time data to confirm it robustly . on day
@xmath61 we also observe a rise in luminosity at @xmath24 and @xmath25 ghz , this could be the effect of a small density enhancement , since that would produce the achromatic variation observed , and as the lower frequencies are still optically thick it would not be noticed at these frequencies . _ phase 2 _ : if the shock is expanding into a constant ( or even smoothly - varying ) stellar wind , then after the absorption turn - on one expects the source to fade in intensity following a power - law decline with time .
however on day @xmath59 the flux density consistently deviates from this at @xmath23 and @xmath22 ghz , showing a discontinuous drop in the observed flux density from days @xmath62 .
_ phase 3 _ : the luminosity of sn 2007bg begins a second turn - on around day @xmath63 , becoming brighter than in phase 1 and reaching a peak luminosity of @xmath64 at @xmath23 ghz on day @xmath65 .
this late time behaviour is comparable to that of other type ib / c sne like sn 2004cc , sn 2004dk and sn2004gq @xcite , albeit much stronger and more well - defined than these other objects .
there are also some type iib s sne which show late time modulations in their radio light curves .
one of these is sn 2001ig @xcite in which case the late time modulations could be explained by a binary companion of the progenitor .
another example is sn 2003bg @xcite , where the late time variations can not rule out a binary companion or different mass - loss episodes of the progenitor . in both cases the modulations in the light curve result from modulations in the csm density . a comparison between sn 2007bg , other sne and grbs is shown in fig .
[ f : msne ] . during phase 1
the light curve of sn 2007bg resembles that of other sne .
it shows the characteristic turn - on , but with a higher luminosity than other normal ib / c sne and comparable to grb - sne .
then during phase 3 we observe a strong rise in its luminosity , which is unprecedented for sne , as it reaches such high luminosities , @xmath4 times larger than during phase 1 . the spectral index , @xmath66 , evolution of the source is shown in fig .
[ f : si ] .
we observe a trend towards a negative value between the best sampled frequencies @xmath22 ghz and @xmath23 ghz during phase 1 , but it could be that during this phase the source is still optically thick at @xmath22 ghz . if we assume that at this time the emission has reached its peak luminosity at @xmath22 ghz , then the late observed values of the spectral index point to a value of @xmath67 .
even the deviant points at @xmath24 and @xmath25 ghz on day @xmath61 are consistent with this spectral index value . during phase 2
the source spectrum begins to invert .
this behaviour continues during the first days of phase 3 , indicating that some absorption mechanism becomes important again in phase 3 . from the earliest spectral indices available we observe that the values are @xmath68 , which is lower than the typical synchrotron self absorption ( ssa ) value of @xmath69 and the spectrum is shallower than the exponential cut - off expected from free - free absorption ( ffa )
this suggests may be observing a variety of emission regions from sn 2007bg ; some being absorbed more highly than others .
since we can not resolve the source , we get the combined flux resulting in a shallower optically thick part of the spectrum .
given that we do not have strong constraints on how the intrinsic spectral index varies with time , we adopt a fixed spectral slope for each phase of @xmath67 . our first data point at @xmath23 ghz allows us to evaluate the brightness temperature of the source at @xmath70 days after explosion
. the brightness temperature of the source should not exceed @xmath71 k if it is to remain under inverse compton catastrophe ( icc ) , where the cooling time due to inverse compton scattering is shorter than the synchrotron cooling time @xcite ( see also ; * ? ? ? * ) . using the measured flux density of @xmath72 at @xmath23 ghz , and
if we assume that the optical expansion speed of @xmath73 km s@xmath16 , measured from the optical ca ii triplet , at @xmath74 days after explosion also pertains the radio emitting region , we obtain @xmath75 k. this value is above the icc limit , so one of our assumptions must be wrong . as discussed in the introduction , typically the radio ejecta from sne
have higher velocities than that of the photosphere , which produces the optical emission . based on this we adopt a radius for the source such that the brightness temperature remains below the icc limit constraining the source size to be @xmath76 cm .
the reported errors come from the flux density measurement uncertainty . at @xmath70 days after explosion
this implies a bulk expansion speed of @xmath77 , where @xmath78 and @xmath79 is the lorentz factor .
this speed @xmath80 is @xmath81 times larger than that of the photosphere .
we stress that this only a lower limit on the actual velocity at which the radio emitting region expands , as the radio ejecta could be expanding at a higher velocity and be well below the icc brightness temperature .
the resulting brightness temperature is shown in fig .
[ f : tb ] where we assumed that the radio emitting region does not decelerate i.e. , @xmath82 .
another way to estimate the expansion speed of the radio emitting region can be obtained by placing sn 2007bg in a luminosity peak time plot , which as discussed by @xcite provides information about the average expansion speed of the radio emitting region .
the plot is shown in fig .
[ f : vel ] and we note that sn 2007bg has an average expansion speed of @xmath83 typical of type ib / c rsne and it shows one of the highest luminosities among type ib / c s and comparable to those of grbs .
it is important to note that other absorption mechanisms , such as ffa , or deviations from equipartition , would increase the expansion velocity ( and hence source energy , derived later ) .
synchrotron emission from sne originates from the shock region between the sne ejecta and the pre - existing csm .
this shock amplifies the existing magnetic fields via turbulence .
the electrons in the shock region then interact with this magnetic field producing the observed synchrotron emission .
this emission is subject to self - absorption , which is caused by ssa @xcite in the case of low density environments . to describe the time evolution of the flux density during phase 1 we adopt the parametrization of @xcite in which the radio emission originates from a spherical thin shell of width @xmath84 expanding at a velocity @xmath85 in which the intrinsic synchrotron emission is absorbed by ssa .
the parameter @xmath86 is the number of shells that fit into the size of the source with a value of @xmath87 @xcite .
we adopt the assumptions of the standard model of @xcite where the hydrodynamic evolution of the ejecta is self - similar across the shock discontinuity . with this assumption
the radius of the source evolves as @xmath88 where the parameter @xmath89 is related to the density profile of the outer sn ejecta @xmath90 and the density profile of the radiating electrons in the shock region @xmath91 such that @xmath92 .
the self similar model also requires that @xmath93 , @xmath94 and @xmath89 in the range @xmath95 .
the model also assumes that the magnetic energy density and the relativistic electron energy density scale as the total post - shock energy density .
this implies that the fraction of energy contained in the relativistic electrons @xmath96 and magnetic field @xmath97 remains fixed .
this in turn relates the magnetic field evolution @xmath98 , to the radius of the source and the density profile of the radiating electrons @xmath99 .
this means that for a freely expanding wind @xmath100 , the magnetic field evolves as @xmath101 . for our analysis
we adopt a source in equipartition with @xmath102 .
equipartition of energy minimizes the total energy . using this model and requiring the time evolution of the magnetic field
to go as @xmath101 we find that the radius of the source evolves as @xmath103 .
we will refer to this as model 1 .
this is consistent with a shock expanding freely through the csm .
the normalization constants for the flux density and ssa are @xmath104 g s@xmath16 and @xmath105 s@xmath106 , giving a reduced chi square value of @xmath107 for 22 degrees of freedom .
these values require the size of the source to go as @xmath108 cm , which in turn means that the expansion velocity is @xmath109 .
the magnetic field goes as @xmath110 g and the energy is @xmath111 erg . for this we assumed a synchrotron characteristic frequency of @xmath112 ghz . using these values we can also estimate the mass - loss rate of the progenitor .
following @xcite we estimate a mass - loss rate of @xmath113 .
this mass - loss rate lies at the low end of those observed for wr stars in low metallicity environments @xcite .
the derived mass - loss rates from this model depend on the fraction of energy contained in relativistic electrons and magnetic fields , deviations from equipartition would require higher mass - loss rates and energy requirements . the early data hint at a shallower slope of the light curve ,
so we also fit a second model , model 2 , where we allow the magnetic field evolution to change as @xmath114 and the source size as @xmath115 . in this model
the reduced chi square value is @xmath116 for 21 degrees of freedom . in this case
the normalization constants are @xmath117 g s@xmath16 and @xmath118 s@xmath106 .
this implies a source size evolution of @xmath119 cm , and an expansion velocity of @xmath120 .
given this expansion speed the source would remain below the icc limit .
the magnetic field goes as @xmath121 g. the source energy with these values is @xmath122 erg , and the mass - loss rate is @xmath0 . here
again we have used the same parameters for @xmath123 and @xmath86 .
given this time evolution for the magnetic field and source size , the density radial profile goes as @xmath124 and the number density of emitting electrons as @xmath125 .
we adopt the result from model 2 , as it provides a better fit to the data , although we can not truly rule out model 1 due to the lack of early detections .
given this mass - loss rate , the circumstellar density for a wind stratified medium @xmath126 , defined as @xmath127 g cm@xmath16 @xcite , is @xmath129 . under these conditions
a relativistic jet from a grb would become non - relativistic , and hence observable on a time - scale of @xmath130 . with the kinetic energy of the ejecta of @xmath131 erg , derived from the optical light curve ,
this time - scale is @xmath132 yr . the lack of clear afterglow signatures ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) in our observations thus provide strong evidence against an expanding grb jet being the source of the observed luminosity for sn 2007bg . as can be seen in fig .
[ f : lc ] , at around day @xmath59 the flux density at @xmath23 and @xmath22 ghz suddenly decreases creating a break in the power law flux decay of the optically thin radio emission . at @xmath23 ghz the observed flux is a @xmath133 less than the predicted one at the beginning of phase 2 on day @xmath134 .
based on the derived size of the radio emitting region , from the ssa modelling , this drop occurs at a distance of @xmath135 cm .
the most likely cause for this decrease in the radio emission is a drop in the csm density . here
we consider two scenarios that would produce a change in csm density such as a binary companion or two different wind components from the progenitor star .
in the first scenario , there should be evidence of this interaction from the optical spectra as balmer series recombination lines .
the late time optical spectra for sn 2007bg was too faint to be detected .
the early optical spectra evidence points to a broad - lined sne , most likely associated with sne like 1998bw , and the latest optical spectra of sn 2007bg where it is detected does not show evidence for a transition from a type ib / c to a type ii . for this kind of sne
the most likely progenitor is a single wr .
also if there was a binary companion we would expect to see coherence between the emission before and after the drop / increase ( depending on whether the csm density dropped / increased at the region of interest ) . in the case of sn 2007bg
, we see a @xmath136 dex jump in the normalizations of the unabsorbed phase 1 and phase 3 light curves , as well as a strong enhancement in absorption , making a binary companion csm scenario highly unlikely .
the second scenario considered is that the drop in flux density is due to a change in the progenitor wind .
based on the distance of the drop and with our assumed wind speed of @xmath137 km
s@xmath16 the change occurred @xmath138 yr before the explosion and lasted for @xmath139 yr . since @xmath140
@xcite a flux density decrease of @xmath133 implies that the mass - loss rate / wind speed dropped by a factor of @xmath141 .
for this estimate we have used @xmath142 .
our modelling of the light curves during phase 1 suggests that the density profile of the csm is shallower ( i.e. , smaller @xmath89 ) , but as discussed bellow , this relation overestimates the mass - loss rate in such cases .
phase 3 begins around day @xmath63 with an increase in the flux density at @xmath25 , @xmath23 and @xmath22 ghz . as discussed in
[ s : results ] the luminosity of the source during this phase is larger than in phase 1 in all of the bands where we detect it .
this flux density rise shows the characteristics of an absorption turn - on . at @xmath25 ghz
we observe what appears to be the peak of the turn - on on day @xmath143 , and at longer wavelengths the radio spectrum begins to turn - on near the end of observations .
this turn - on is accompanied by an increase in the spectral index suggesting that the source has become optically thick again .
the luminosity increase implies a sharp csm density enhancement , presumably due to evolution of the progenitor wind , while the observed re - absorption requires either ssa or ffa .
following @xcite we can estimate the mass - loss rate from the progenitor assuming that the absorption is purely ffa .
adopting the time of peak at @xmath22 ghz as the time when the free - free optical depth is unity , and taking the expansion velocity of the shock from phase 1 then the mass - loss rate is @xmath144 .
this rate is suspiciously high and inconsistent with our estimate ( later in this section ) of the mass - loss rate based on the unabsorbed portion of the spectrum .
this high rate favours ssa .
the absorption slope is relatively flat , and thus pure ssa with a similar model as fit to phase 1 does a poor job of fitting the turn - on .
one possibility is differential ssa arising from small dense clumps in the csm , which are preferentially affected by ssa while sparser regions are less affected ; this naturally flattens the optically thick slope .
the existence of clumpy winds in wr stars has been confirmed by observations @xcite , and thus provides a natural way of explaining the observed radio emission .
alternatively , the density jump would naturally produce strong x - ray emission , which could ionize csm material in front of the shock and provide the necessary free electrons for ffa .
however , without constraints on the x - ray luminosity and phase 3 csm density distribution for sn2007bg , it is difficult to assess the feasibility of ffa . given the poor fit of the pure ssa model ( @xmath145 ) , we simply adopt the parametrization of including only a clumpy csm with the presence of internal ssa . in this case the flux density is normalised by @xmath146 , the clumpy csm absorption by @xmath147 and ssa by @xmath148 .
we fit the above model to the light curve during phase 3 .
the best - fit values are given in table [ t : phase3 ] .
the resulting reduced chi square of this fit is @xmath116 for 18 degrees of freedom .
lc parameter & adopted value + @xmath149 & @xmath150 + @xmath151 & @xmath152 + @xmath89 & @xmath153 + parameter & best - fit value + @xmath146 & @xmath154 + @xmath147 & @xmath155 + @xmath156 & @xmath157 + @xmath148 & @xmath158 + @xmath159 & @xmath160 + + + + as shown in fig .
[ f : lc ] , this model provides a reasonable fit to the light curves at all frequencies and also to the spectral index , fig .
[ f : si ] .
the deviant point at @xmath21 ghz could be easily affected by the host galaxy luminosity .
the data did not allow us to constrain the deceleration rate , so we adopted a constant expansion by fixing @xmath161 , which in turn requires that @xmath162 given the spectral index .
using our best - fit values we can estimate the mass - loss rate from the progenitor during this period following equations ( 11 ) and ( 13 ) of .
the effective optical depth at the time of our latest observation , @xmath163 days , at @xmath22 ghz is @xmath164 and @xmath165 .
if we extrapolate the velocity derived in [ ssec : ssa ] to phase 3 , @xmath166 at @xmath167 days , then the derived mass - loss rate is @xmath168 , where an electron temperature of @xmath169 k was used . this mass - loss rate is higher than the one observed during phase 1 , which is a natural explanation for the high luminosity observed during phase 3 .
however , such a large mass - loss rate has not yet been directly observed , and it is at odds with the metallicity mass - loss rate relation derived from the milky way and magellanic clouds wr samples ( see for e.g. , * ? ? ?
this relation predicts orders of magnitude lower @xmath170 in such a low metallicity host . on the other hand , this large mass - loss rate is also roughly consistent with that of the saturation limit for line - driven winds , @xmath171 , and is of the same order of magnitude as values found in other type ib / c sne ( e.g. , * ? ? ?
we should caution however that the dramatic flux increase between phases 1 and 3 implies a strong density contrast that is inconsistent with one of the primary assumptions of the @xcite self - similar model ( namely a csm density profile with a power - law slope ) .
this density discontinuity could naturally lead to an overestimate of the mass - loss rate ; to properly account for this type of case , one should generally model the hydrodynamics in detail ; unfortunately , such simulations are not well - constrained by radio data alone due to ambiguities in how the b - field and electron density distribution might change across the density discontinuity and are beyond the scope of this work .
thus this phase 3 mass - loss rate should be regarded with caution and viewed as a likely upper limit .
the strong mass - loss variations on such short time - scales seen here imply that the progenitor of sn 2007bg may have undergone unstable mass - loss just prior to explosion or a dramatic change in wind properties .
such rapid variations have been seen in a handful of objects before .
for instance , in the case of type ib sn 2006jc @xcite , a bright optical transient was observed two years prior to explosion .
this was interpreted as a luminous blue variable - like outburst event from the progenitor , which the later sn explosion surely interacted with .
then there are other possibilities like the case of type iin sn 1996cr @xcite . in this event
, the radio light curve also exhibits a strong late - time rise in emission associated with a density enhancement likely arising from the interaction of the sn blast - wave with a wind - blown bubble @xcite .
the progenitor of sn 1996cr is proposed to be either a wr or blue supergiant star .
we still know relatively little about why such global changes in the progenitors occur , but it seems clear that a subset of sne progenitors have them .
x - ray emission from supernovae can arise from three mechanisms ; non - thermal emission from synchrotron radiating electrons , thermal bremsstrahlung emission from the material in the circumstellar and reverse shocks , and inverse compton emission . since x - ray emission from sn 2007bg could not be detected it is difficult to assess the feasibility of each of these mechanisms , so we will compare the derived mass - loss rates from the previous subsections with that predicted by the thermal bremsstrahlung process .
if the x - ray emission comes from synchrotron radiating electrons then the x - ray luminosity can be obtained by extrapolating the radio synchrotron spectrum up to higher frequencies . to do this
we must consider that the synchrotron spectrum changes its slope at the cooling frequency @xmath172 by a factor @xmath173 @xcite .
using the derived magnetic field @xmath174 from our ssa analysis of the radio light curve [ ssec : ssa ] we find that @xmath175 ghz .
extrapolation of the luminosity as @xmath176 from the cooling break on gives a x - ray luminosity 5 orders of magnitude bellow our upper limit @xmath177 days after explosion ( epoch 2 ) .
extrapolation of the cooling frequency up to epochs 3 and 4 of the x - ray data yields x - ray fluxes @xmath74 orders of magnitude bellow the upper limits .
however , as mentioned in previous sections , the csm discontinuity between phases 1 and 3 could cause a break in the magnetic field evolution .
x - ray emission from thermal bremsstrahlung comes from the interaction of the shocked electrons with the circumstellar material .
this interaction heats the electrons in the medium which then cool via free - free emission . in this process
the x - ray luminosity is related to the density of the emitting material as @xmath178 @xcite .
we can obtain an upper limit on the mass - loss rate from the x - ray luminosity upper limits using equation ( 30 ) of @xcite . at @xmath179 days
the x - ray luminosity is @xmath180 .
we use the density profile of the outer sne ejecta derived from our ssa analysis , @xmath124 . given this luminosity and csm properties thermal bremsstrahlung requires a mass - loss rate of @xmath181 . at epochs 3 and 4 of the x - ray data we obtain mass - loss rates of @xmath182 and @xmath183 respectively .
these mass - loss rates are consistent with the one required by ssa during phase 1 and the light curve during phase 3 . inverse compton scattering of the optical photons by the relativistic synchrotron emitting electrons can also be a cause of x - ray emission in sne .
it has been shown that this process played an important role in the case of sn 2002ap @xcite , sn 2003l @xcite and 2003bg @xcite . using equation ( 15 ) of @xcite
we obtain a ratio of @xmath184 @xmath177 days after explosion . for this
we have used a bolometric luminosity of @xmath185 erg s@xmath16 based on the pseudo - bolometric light curves for sn 2007bg of . given the ratio @xmath186 and the radio flux of @xmath187 erg s@xmath16 @xmath47
the derived x - ray luminosity of @xmath188 erg s@xmath16 @xmath47 is in agreement with our upper limits . for later epochs we can not estimate the bolometric luminosity of sn 2007bg
in this article we presented the radio light curves and x - ray observations of type ic - bl sn 2007bg .
the radio emission is characterised by three different phases .
phase 1 being characterised by a ssa turn - on . based on the brightness temperature of sn 2007bg the expansion speed of the radio emitting region
is @xmath189 .
an expansion speed similar to what has been found in the small sample of radio - emitting type ib / c sne .
the derived mass - loss rate from the progenitor during phase 1 is @xmath0 , a low mass - loss rate among the observed in wr stars . during phase 2
we observe a drop in the flux density while the spectral index seems to follow a smooth evolution .
this drop is presumably due to the radio ejecta entering a region of lower csm density , probably arising because of a change in the mass - loss rate of the progenitor or a variable wind speed .
then during phase 3 we observe a rise in the radio flux density along with a re - absorption of the radio emission .
the shallow observed slope of the re - absorption could be explained by small clumps of material in the progenitor wind which would become more strongly self - absorbed than the surrounding csm when shocked and produce a shallower optically - thick spectrum . from our modelling of the radio light curves during phase 3 we put an upper limit on the mass - loss rate of @xmath5 .
this second wind component could arise from a different stellar evolution phase of the progenitor prior to explosion , possibly a lbv phase . or it could be the effect of stellar rotation on the stellar wind properties . if the radio flux density from sn 2007bg declines in time like @xmath190
. it should still be detectable , with a flux density at @xmath25 ghz of @xmath191 , and higher at lower frequencies .
if the progenitor star went through more mass - loss episodes , then the observed flux density would deviate from this value .
recent observations taken with the karl g. jansky very large array ( jvla ) should help to establish this .
very few broad - lined type ic sne have radio detections , and none of them show signs of a complex csm like that of sn 2007bg .
these different wind components , along with the derived mass - loss rates will eventually help to constrain the progenitors of type ib / c sne , from their inferred pre - sne properties .
currently two main evolutionary paths has been proposed to account for the rate of these explosions : a single wr origin , and a binary system .
however , the single wr scenario itself still needs significant clarification .
for instance , what effects do evolution ( mass - loss , rotation , etc . ) and metallicity have in shaping the csm around massive type ib / c progenitors and ultimately their explosions . and
more generally , how does this fold into the connection between type ii s , type ib / c , and grbs . compounded by their relative rarity , a key impediment to improving
our knowledge of broad - lined ic sne has been the lack of high - quality , multi - frequency observations from early through late epochs .
sensitive high - frequency radio observations of ic sne within 1 - 2 days of explosion are needed to efficiently identify the best objects for follow - up and allow further refinement of the physical properties surrounding these unique explosions . both alma and
the newly retooled jvla will play critical roles here , providing rigorous probes of the synchrotron - emitting shock ( to constrain evolution of the shock velocity , any potential beaming , etc . ) and ultimately new insights into the properties of their csm ( such as density , variability , clumpiness , and perhaps even dust content for these massive systems ) .
sensitive , long - term x - ray and optical / nir spectroscopic follow - up are also needed in order to break potential degeneracies and provide consistency checks against the shock - csm interaction models .
if such complete datasets can be assembled in the next several years , they may afford us a leap in our physical understanding of sne and grbs .
we thank the vla for regularly monitoring this object , the vla archive for making this data publicly available , a. m. soderberg for her telescope time , to m. krauss for her help , k. stanek and s. schulze for their useful comments while drafting this article .
we acknowledge support from programa de financiamiento basal ( f. e. b. ) , the iniciativa cientfica milenio through the millennium center for supernova science grant p10 - 064-f ( f. e. b. ) , and conicyt - chile under grants fondecyt 1101024 ( p. s. , f. e. b. ) , alma - conicyt 31100004 ( f. e. b. ) , and fondap - cata 15010003 ( f. e. b. ) , j. l. p. acknowledges support from nasa through hubble fellowship grant hf-51261.01-a awarded by stsci , which is operated by aura , inc . for nasa , under contract
nas 5 - 2655 .
r. a. , fransson c. , 2003 , in weiler k. , ed . , supernovae and gamma - ray bursters vol .
598 of lecture notes in physics , berlin springer verlag , supernova interaction with a circumstellar medium .
pp 171194 p. a. , 2006 , in lamers h. j. g. l. m. , langer n. , nugis t. , annuk k. , eds , stellar evolution at low metallicity : mass loss , explosions , cosmology vol .
353 of astronomical society of the pacific conference series , observed metallicity dependence of winds from wr stars . | in this paper we present the results of the radio light curve and x - ray observations of broad - lined type ic sn 2007bg .
the light curve shows three distinct phases of spectral and temporal evolution , implying that the sne shock likely encountered at least 3 different circumstellar medium regimes .
we interpret this as the progenitor of sn 2007bg having at least two distinct mass - loss episodes ( i.e. , phases 1 and 3 ) during its final stages of evolution , yielding a highly - stratified circumstellar medium . modelling the phase 1 light curve
as a freely - expanding , synchrotron - emitting shell , self - absorbed by its own radiating electrons , requires a progenitor mass - loss rate of @xmath0 for the last @xmath1 yr before explosion , and a total energy of the radio emitting ejecta of @xmath2 erg after @xmath3 days from explosion .
this places sn 2007bg among the most energetic type ib / c events .
we interpret the second phase as a sparser `` gap '' region between the two winds stages .
phase 3 shows a second absorption turn - on before rising to a peak luminosity @xmath4 times higher than in phase 1 . assuming this luminosity jump is due to a circumstellar medium density enhancement from a faster previous mass - loss episode
, we estimate that the phase 3 mass - loss rate could be as high as @xmath5 .
the phase 3 wind would have transitioned directly into the phase 1 wind for a wind speed difference of @xmath6 . in summary ,
the radio light curve provides robust evidence for dramatic global changes in at least some ic - bl progenitors just prior ( @xmath7 yr ) to explosion .
the observed luminosity of this sn is the highest observed for a non - gamma - ray - burst broad - lined type ic sn , reaching @xmath8 , @xmath9 days after explosion .
[ firstpage ] stars : mass - loss supernovae : general supernovae : individual : sn 2007bg |
a connected @xmath0-manifold with non - empty boundary has a handle - decomposition with a @xmath3-handle , some @xmath4-handles and some @xmath5-handles .
we call @xmath3 , @xmath4 and @xmath5-handles balls beams and plates respectively
. a beam is homeomorphic to @xmath6 $ ] and is attached to the ball along @xmath7 .
we shall call @xmath7 the ` sticky end ' of a beam .
likewise a plate is homeomorphic to @xmath6 $ ] and is attached to the union of balls and beams along @xmath8 $ ] , which we shall call the ` sticky end ' of the plate .
the components of intersection of beams and plates with the ball shall be called islands and bridges respectively .
it is clear that a handle - decomposition as above is completely determined by the beams and plates together with their ` sticky ends ' , for there is an essentially unique way to stick the ` sticky ends ' to the @xmath3-handle .
further , suppose we are given beams and plates , with the plates glued to beams along strips in their sticky end as they would be in a @xmath0-manifold .
then these are the beams and plates of a handle - decomposition of a @xmath0-manifold iff the sticky end of these is planar .
this is an unnecessary assumption for many applications of @xmath0-manifold techniques , in particular many results in normal surface theory .
this motivates the definition a _ handle - structure _ is a union of beams and plates , which intersect in unions of discs that when viewed as subsets in either the beams or the plates are of the form @xmath9 $ ] where @xmath10 is an arc in @xmath11 .
the horizontal boundaries @xmath7 of the beams together with the portion of the horizontal boundary of the plates that is not glued to a beam form a surface consisting of islands and bridges that we call the _ sticky end _ ( see figure [ f : sticky ] ) .
the space consisting of the beams and plates will be called the _ total space_. -handles and @xmath5-handles in a handle structure ] our aim in introducing handle - structures is to associate these with finite presentations of groups .
we shall use the notation @xmath12 , where @xmath13 and @xmath14 are topological spaces with @xmath15 , to denote the quotient space obtained form @xmath13 by identifying @xmath14 to a point .
briefly , we can associate with any presentation a singular @xmath0-manifold , with the only singularities being transversal intersections and self - intersections of the bridges .
we delete from this the closure of the ball to get a non - singular open @xmath0-manifold .
the completion of this gives a handle - structure , with the sticky - part being the points added to complete the space . [ p : gphan ] given a finite presentation of a group @xmath16 , we can associate to it a handle - structure .
this construction is canonical up to finitely many choices .
further , if @xmath17 is the total space and @xmath18 the sticky end , then @xmath19 .
take a beam corresponding to each generator and a plate corresponding to each relation .
successively glue the plates to the beams according to the word that the relation represents .
namely , start with a base point that will not be glued to any @xmath4-handle , and for each successive letter in the word of the relation attach a strip in the horizontal boundary @xmath8 $ ] of the plate to the corresponding beam , with the direction determined by the sign of the exponent .
thus , strips attached to @xmath4-handles alternate with bridges as in figure [ f : sticky ] , with the attaching maps being governed by the word that the @xmath5-handle represents .
the only essential choices involved are in choosing where to attach the strip to the beam relative to previous intersections with plates .
clearly @xmath20 deformation retracts onto a @xmath5-complex that is well known to have fundamental group @xmath16 .
more precisely , each one handle deformation retracts to its core , and in the quotient the two islands of a @xmath4-handle are identified to a single point .
thus , the @xmath4-handles deformation retract onto a wedge of circles .
the @xmath5-handles in turn deformation retract on to their cores , which are @xmath5-discs . by construction ,
the attaching maps of the disc corresponding to a given @xmath5-handle to the wedge of circles formed by the @xmath4-handles gives the corresponding relation in the fundamental group .
for the presentation @xmath21 of a free group on @xmath22 generators , we have @xmath22 @xmath4-handles with @xmath23 islands , @xmath5 on each @xmath4-handle , and no @xmath5-handles .
thus @xmath20 is a thickened wedge of @xmath22 circles ( see figure [ f : eg ] ) . and
@xmath18 ] for the presentation @xmath25 of the trivial group , we have one @xmath4-handle and a @xmath5-handle , with the intersection of these connected . thus @xmath17 is the union of two @xmath0-balls glues together along a @xmath5-disc , and is therefore a @xmath0-ball .
@xmath18 consists of two islands joined together by a bridge , and hence is @xmath26 ( see figure [ f : eg ] ) .
the standard balanced presentation of the trivial group with @xmath22 generators and relations gives a handle - structure with @xmath22 components , each of them as in the previous example .
we shall recognise handle - structures that correspond to a class of presentations of the trivial group .
we assume for the next two sections that we are given a _ balanced _ presentation of a group . to this
we associate a finite collection of handle structures as above .
we apply the following algorithm to each of these . given a handle - structure , let @xmath17 be the total space and @xmath18 the sticky end . 1
. [ loop ] attach thickened discs to @xmath17 along each component of @xmath27 to get a closed manifold .
define the sticky part of this to be the union of the components of the boundary that intersect @xmath18 .
[ rtm ] if each component of the resulting manifold is @xmath28 $ ] , with one of its boundary components being the sticky part , then answer ` yes ' and terminate .
[ ann ] cut @xmath29 along a maximal family of pairwise disjoint and non - parallel annuli such that each annulus has exactly one boundary component in @xmath18 . if there are no such annuli , answer ` no ' and terminate .
4 . repeat from step [ loop ] using the new @xmath29 .
we shall call annuli which are as in step [ ann ] of the algorithm _ vertical annuli_. step [ rtm ] can be achieved by using the rubinstein - thompson - matveev algorithm to recognise @xmath2 .
step [ ann ] uses basic normal surface theory .
the algorithm terminates because the boundary gets simpler at each iteration .
[ l : gptriv ] if the algorithm terminates with a ` yes ' , then the @xmath30 is the trivial group .
this follows by induction on the number of steps needed .
if @xmath17 is as in step [ rtm ] then this is true .
it is easy to see that this remains true if @xmath28 $ ] is obtained after cutting along vertical annuli and filling in discs as in steps [ ann ] & [ loop ] .
we need to see that if @xmath30 is trivial after cutting along an annulus or capping off , then it is trivial before making these changes . in the latter case
, this follows as there is no relation added - the boundary of the disc is already trivial as it is in @xmath18 . in the case where we cut along a vertical annulus ,
let @xmath31 and @xmath32 be obtained after cutting .
now an arc that in @xmath17 that intersects the annulus can be homotoped in the quotient off this - it breaks into an arc ending in @xmath18 just before the annulus and an arc starting from s just after the annulus .
this is in @xmath31 , hence trivial in @xmath33 , and so all the more in @xmath20 .
we thus have an algorithm to recognise a certain class of presentations of the trivial group .
namely , given a presentation , we apply the above algorithm to all the handle - structures associated to it .
if the algorithm says ` yes ' for at least one of these , we conclude that the presentation corresponds to the trivial group .
we shall call presentations where this happens , and also the handle - structures for which the algorithm says ` yes ' _ sphere - like_. in case the presentation and the handle - structure come from a @xmath0-manifold , after capping off @xmath18 is a union of @xmath5-spheres , and handle - structure is sphere - like iff the manifold is a sphere .
we have constructed above an algorithm to recognise certain presentations of the trivial group , which we call _ sphere - like_. for this to be useful , we need to see that this class of presentations is reasonably large .
in fact , we shall see we shall show this in two stages .
first , we multiply relations merely by concatenation without any cancellation .
we shall make a choice of handle - structures that correspond to the new ( in general not reduced ) presentation such that the new presentation is sphere - like .
we then consider the effect of cancellation and show that even after cancellation , the handle - structure is sphere - like .
let @xmath35 and @xmath36 be two points on @xmath27 . then by _ gluing _ @xmath35 and @xmath36 we mean the identification of disjoint regular neighbourhoods of these points , with the new @xmath18 being the image of @xmath18 under the quotient map . on making an appropriate choice of handle - structure corresponding to a presentation obtained by multiplying words without cancellation ,
the effect of the andrews - curtis moves is a sequence of gluings and puncturings .
next consider conjugation of a relation by a generator .
this , again with an obvious choice of handle structure , amounts to a handle slide together with puncturing a hole along an arc with one boundary point on the sticky part .
it may be helpful to consider the change in handle structure on making the andrews - curtis move @xmath39 .
the resulting @xmath40 is @xmath41 , annulus)$ ] ( see figure [ f : eg3 ] ) ) .
the transformation in general is very similar to this particular case .
if we attach a thick disc along the new bridge that forms a loop , we see that the resulting @xmath29 is obtained from the original one by gluing .
puncturing along the arc dual to the thick disc now gives us the handle - structure obtained by conjugation .
next , we consider the effect of multiplying a relation by another
. first consider the particular example @xmath42 .
the new handle - structure is @xmath43 ( see figure [ f : eg2 ] ) and is clearly obtained by a gluing .
suppose that we are replacing the relation @xmath44 by @xmath45 .
if @xmath46 has only one bridge , then we have a similar situation .
namely , in the new handle - structure , @xmath44 has a bridge replaced by two bridges and has an additional component of intersection with the only beam that @xmath46 intersects .
choose the new handle structure so that this component of intersection is adjacent to the attachment of @xmath46 to the beam .
then the handle - structure is clearly obtained topologically by gluing .
suppose now @xmath46 has more than one bridge , i.e. , bridges in addition to the one along which gluing has taken place .. we attach the relations other than @xmath44 as before .
the new @xmath44 is the old one with additional bridges and gluings to @xmath4-handles .
namely first pick bridges of @xmath46 and @xmath44 corresponding to the base points , and replace the bridge of @xmath44 by two bridges to the islands bounding the bridge of @xmath46 and adjacent to this bridge . then attach @xmath44 to the @xmath4-handles adjacent to the where @xmath46 has been attached .
the new bridges of @xmath44 will thus be adjacent to those of @xmath46 . attaching thick discs to these gives , as in the example above , a handle - structure obtained by gluing .
as in the case of conjugation , the final structure is obtained by puncturing along dual arcs to these discs . to see the above , note that when we have a cancelling pair of letters , the @xmath5-handle goes over a @xmath4-handle and then returns over the same @xmath4-handle immediately .
thus , succesive bridges of the @xmath5-handle are one ending in the @xmath4-handle , one beginning and ending in the @xmath4-handle ( which shall often be referred to as a _ loop _ in what follows ) and then one beginning in the @xmath4-handle .
we cut along a quadrilateral ( see figure [ f : quad ] ) consisting of separating arcs along the first and last of these bridges and arcs in the vertical boundary of the @xmath5-handle .
then we cut along a bigon whose boundary has an arc in the bridge that begins and ends in the @xmath4-handle .
the result is homeomorphic to a handle - structure corresponding to the cancelled presentation .
note that in the figure , after splitting along the quadrilateral , we declare the left side , i.e. , the side away form the bigon as part of the sticky end .
thus what we have left is indeed a bigon
. it may be helpful here to see these in figure [ f : eg3 ] . here
, the quadrilateral splits the two bridges joining the two @xmath4-handles .
it is thus the dotted arc joining these bridges in the figure thickened .
the side towards the lower handle in the figure becomes part of @xmath18 .
the bigon is the dotted line joining the third bridge to the quadrilateral thickened .
the effect of puncturing along a vertical arc is undone in the first step of the algorithm , where the components of @xmath18 are capped off by attaching discs .
thus , there is nothing to prove in this case . when we glue along a bigon , there are two cases to consider , namely where we glue two distinct boundary components and where we glue a boundary component to itself . in the latter case ,
the manifold after capping off does not change , we merely glue two discs instead of one . in the former case , after gluing and capping off , we end up with an essential vertical annulus .
this is because gluing together and capping off has the same effect as attaching a cylinder between the two boundary components . after cutting along this and capping off
, we end up with the manifold obtained by capping off the original annulus .
finally , we note that as though the maximal family of vertical annuli need not be unique , the complementary region does not depend on the family chosen .
hence we may assume that the family chosen in the algorithm contains the abovementioned annulus .
note that there are at most two punctures intersecting the bigon , corresponding to the two ` vertices ' of the bigon .
if the vertices are on different components of @xmath27 , then we see that cutting along the bigon and capping off has the same effect as capping off the original manifold . thus , we need to consider only the case where both vertices are on the same component .
we first consider the case where we have a punctured product @xmath48 , with the only punctures being ones that intersect the bigon .
we shall then reduce to this case .
note that we can assume that punctures disjoint from the bigon were made after cutting along the bigon and thus the latter part of the reduction is immediate .
thus , the vertices are either on a puncture or a component @xmath38 of @xmath49 . in the former case ,
we call a neighbourhood of the puncture the _ boundary annulus _ and in the latter case we use this term for @xmath50
. we shall see that we have two possibilities , a bigon parallel to an ( in general ) knotted puncture and a possibly essential one in a product ( i.e. with vertex on @xmath49 or an unknotted puncture ) .
we have a natural vertical annulus in the manifold capped off at the component of @xmath27 containing the vertex ( see figure [ f : bigon ] ) .
namely , we add to the bigon the neighbourhood of an arc in @xmath27 , i.e. , a strip in the boundary annulus . if this is an essential vertical annulus ,
then as the arc along which the manifold was punctured is a vertical arc on this annulus , puncturing and then cutting along the bigon still gives a product . otherwise , the boundary curves of the annulus must bound discs , giving a @xmath5-sphere .
this must bound a @xmath0-ball ( which follows from irreducibility of the punctured product ) .
thus , cutting along the bigon gives one component @xmath51 and the other homeomorphic to the original manifold . in either case , the manifold is still sphere - like . now to reduce to the case
considered , we need to consider the intersection of the annulus with the bigons along which gluing has taken place .
this intersection , after making things in general position , is a collection of inessential arcs ( i.e. those with both endpoints on the same boundary component of the annulus ) , vertical arcs and horizontal circles .
we take an outermost such component .
if it is an inessential arc , then we can isotope the annulus to remove this intersection . to do this , note that the arc separates a disc in each of the annulus and the bigon , and these together form a compressing disc .
as the boundary is incompressible , we have a sphere , which must bound a ball by irreducibility .
if we have a horizontal circle , then this bounds a disc in the annulus , and we see that we have a bigon parallel to a puncture . finally , in case of a vertical arc , our bigon decomposes into a piece between two parallel punctures ( perhaps the same one and a bigon with simpler intersection .
again , the first bigon is as in one of the above cases , and in either case , it is easy to see that we get a sphere - like pair .
first , observe that when cutting along a quadrilateral , there is a choice involved .
namely , the disc along which @xmath17 has been cut now corresponds to two discs in @xmath52 , and one of these is to be included in @xmath18 . if the wrong choice is made , the result is not even homologically @xmath53 .
it is easy to see that in our case we do have a product homologically , and so the choice has been made correctly .
namely , if we consider the handle - structure obtained with the other choice , then we have two bridges , one corresponding to the earlier loop and one a new one , so that the bridges end on both sides in the two islands corresponding to the one - handle where the cancellation takes place . for each bridge
, we can take an arc along it and close it up in the one handle to get a closed curve .
these curves bound an annulus in @xmath17 , and are thus homologically equivalent . but an arc across one of the bridges intersects one of the curves once , but not the other , hence the curves are not homologous in @xmath18 .
thus we do not have a ( homological ) product with this wrong choice . with the correct choice
, this claim is well known in the case of products ( and more generally for taut sutured manifolds @xcite ) . in our situation , we claim that cutting along a quadrilateral has the effect of cutting along two bigons and then gluing along a bigon . more specifically ,
we can find @xmath37-compressing disc for the quadrilateral , i.e. , a disc whose boundary consists of an arc in the quadrilateral and an arc in @xmath18 ( see figure [ f : compress ] ) at some stage in the algorithm .
boundary - compressing along such a disc gives two bigons , and it is easy to see that cutting along the quadrilateral is equivalent to cutting along each of these discs and then re - gluing along the @xmath37-compressing disc .
we show that such a disc exists by induction . in the case of a product
, this follows by cutting up @xmath17 along vertical annuli and discs .
an outermost disc of the vertical discs and annuli cut by the quadrilaterals gives the compression , except in the trivial case where the quadrilateral can be made disjoint from these .
similarly , if we have a quadrilateral after gluing along bigons , we look at outermost arcs of the bigons after cutting by the quadrilateral to reduce to the situation before gluing along the bigon . as with bigons
, we may assume that punctures not involving the vertices of the quadrilateral were made after cutting along the quadrilateral .
further , we may again assume that we are in a product .
for , we again consider the intersections with the bigon .
inessential arcs can be handled as before .
an arc that separates one vertex from the other three is like a vertical arc in the case of the bigons .
an arc that separates the vertices into two pairs separates a disc in the bigon which gives a compression , and so we are done in this case . to find the compression ,
consider an arc @xmath54 along the quadrilateral joining a pair of edges in @xmath18 ( see figure [ f : compress ] ) .
this is homotopic to an essentially unique arc in @xmath18 before puncturing ( as we are in a product ) .
we claim that the arc is unknotted , i.e. , it is isotopic to this arc and unlinked from the punctures on which the vertices lie .
this follows since the quadrilateral corresponds to a cancellation , and hence we can find an embedded , once - punctured disc that the arc @xmath54 together with an arc in @xmath18 bounds .
namely , we can join the end points of this arc by an arc in @xmath18 running through the bridges that have been cut and the common island in which these end .
the original arc together with this bound a tube , one half of which is in the chopped off portion of the plate and the other in the beam corresponding to the cancellation .
the other boundary of this tube is a circle in @xmath18 , whose boundary is an arc along the bridge that forms a loop and the other in the other island that our beam bounds . if the total space is @xmath55 with @xmath56 a punctured sphere , as the arc is unknotted and unlinked from the punctures containing its vertices , the arc bounds a compressing disc . in the case of a general product ,
there exists a family of annuli , which we can assume are disjoint from the arc and the circle in @xmath18 that are part of the boundary of the quadrilateral , so that the complement of these is @xmath55 , with @xmath56 a punctured sphere . by considering outermost components of intersection as before ,
we reduce to the previous case .
one may use the algorithm given here to attempt to find particular presentations that violate the ( balanced ) andrews - curtis conjecture . to do so
, we take a candidate presentation of the trivial group and apply the algorithm to it . if it turns out not to be sphere - like we have a counterexample .
in particular , this method could show that the andrews - curtis conjecture ( in the balanced case ) is false .
furthermore , if this method works , for instance , for presentations corresponding to the akbulut - kirby @xmath57-sphere , we would be able to conclude that property - r for links is false . | we relate the andrews - curtis conjecture to the triviality problem for balanced presentations using algorithms from @xmath0-manifold topology .
implementing this algorithm could lead to counterexamples to the andrews - curtis conjecture .
the andrews - curtis conjecture @xcite says that a balanced presentation of the trivial group can be transformed to a standard presentation @xmath1 by using ` andrews - curtis moves ' , i.e. multiplying one relation by another , inverting a relation and conjugating a relation by a generator . here
, we use algorithmic methods from @xmath0-manifold topology to relate this conjecture to a fundamental algorithmic question in group theory , namely the triviality problem for balanced presentations .
thus , our main result is at least one of the following holds * there is an algorithm to recognise balanced presentations of the trivial group , or * the ( balanced ) andrews - curtis conjecture is false .
moreover , our methods could , after extensive computations , show the failure of the ( balanced ) andrews - curtis conjecture for balanced presentations , and also ` property - r ' for links .
our approach is to construct an algorithm , using algorithms from @xmath0-manifold topology , in particular the rubinstein - thompson algorithm ( @xcite,@xcite and @xcite ) for recognising the @xmath0-sphere , to recognise certain presentations of the trivial group . to do this
, we associate with a group presentation a ` handle - structure ' , a natural generalisation of a handle - decomposition of a @xmath0-manifold .
just as the rubinstein - thompson - matveev recognises a certain class of handle - decompositions of homotopy @xmath0-spheres , namely those corresponding to a @xmath0-sphere , our algorithm recognises handle - structures corresponding to certain presentations of the trivial group .
it is not clear exactly what this class of presentations is , but one can see that it includes those related to the standard presentations by the andrews - curtis moves . we can immediately conclude that either the andrews - curtis conjecture is false , or we have an algorithm to solve the triviality problem for balanced presentations . indeed , if the andrews - curtis conjecture were true , then our algorithm would recognise the trivial group given a balanced presentation .
further , we can apply the algorithm to candidate counterexamples .
if the answer is in the negative , then we know that we have a presentation that violates the andrews - curtis conjecture . in particular ,
if such a presentation comes from a kirby diagram for @xmath2 then property - r for links is false .
i would like to thank david gabai for his indispensable help and encouragement at all stages of this work .
i am indebted to michael freedman , andrew casson , robion kirby , peter teichner , dave letscher and tao li for many valuable comments and conversations . |
9 schmitt - rink s , miyake k and varma c m 1986 _ phys .
lett . _ * 56 * 2575 miyake k 1987 _ j. magn .
_ * 63&64 * 411 hirschfeld p vollhardt d wlfle p 1986 _ solid state commun . _ * 59 * 111 hirschfeld p j wlfle p einzel d 1988 _ phys .
_ b * 37 * 83 anderson p w 1984 _ phys .
30 * 4000 volovik g e gorkov l p 1985 _ sov . phys.@xmath3jetp _ * 61 * 843 ueda k rice t m 1985 _ phys . rev .
_ b * 31 * 7114 blount e i 1985 _ phys . rev .
_ b * 32 * 2935 sigrist m ueda k 1991 _ rev .
phys . _ * 63 * 308 miyake k 1985 _ theory of heavy fermions and valence fluctuations _ eds .
kasuya t and saso t ( springer series in solid state sciences * 62 * ) p 256 ozaki m machida k 1989 _ phys .
_ b * 39 * 4145 leggett a j 1975 _ rev .
_ * 47 * 331 machida k nishira t and ohmi t 1999 _ j. phys .
jpn . _ * 68 * 3364 sauls j a 1994 _ j. low temp .
phys . _ * 95 * 153 | it is shown that the density of states ( dos ) proportional to the excitation energy , the so - called polar like dos , can arise in the odd - parity states with the superconducting gap vanishing at points even if the spin - orbit interaction for cooper pairing is strong enough .
such gap stuructures are realized in the non - unitary states , f@xmath0 , f@xmath1 , and f@xmath2 , classified by volovik and gorkov , _ sov .
phys.@xmath3jetp _ * 61 * ( 1985 ) 843 .
this is due to the fact that the gap vanishes in quadratic manner around the point on the fermi surface .
it is also shown that the region of quadratic energy dependence of dos , in the state f@xmath4 , is restricted in very small energy region making it difficult to distinguish from the polar - like dos . in early stage of research of the heavy fermion superconductors
, it has been important to infer the gap anisotropy from the power - law of the temperature dependence of a series of physical quantities @xcite .
it was a sort of the golden rule there that point node(s ) of the superconducting gap leads to the density of states ( dos ) , @xmath5 , while line node(s ) leads to @xmath6 .
it was also emphasized that all the odd - parity pairings would have only point node(s ) if the spin - orbit coupling for the pairing interaction were so strong that the spin- and orbital degrees of freedom of the gap function can not change independently @xcite .
however , it is not so self - evident whether the spin - orbit coupling for pairing is really so strong to quench technically the independent variation in spin- and orbital space @xcite . in any case
, the classification scheme proposed by volovik and gorkov ( vg ) has been believed to rule out the polar - like dos for the odd - parity states .
the purpose of this letter is to point out that three of the non - unitary states in vg scheme has the polar - like dos because the @xmath7-dependence of the gap around the point node is quadratic rather than linear . in the odd - parity manifold , the quasiparticle energy is the matrix in the representation of spin eigen state as follows @xcite : @xmath8^{1/2 } , \label{dispersion}\ ] ] where the superconducting gap is also the 2@xmath92 matrix in the presentation of spin eigenstate , and is represented in terms of the @xmath10-vector as @xmath11 where @xmath12 is the pauli matrix of the @xmath13-the component , with @xmath13=@xmath14 , @xmath15 , and @xmath16 .
the eigen values of the magnitude of the gap matrix is given as @xcite @xmath17 .
\label{eigenvalue}\ ] ] it is remarked that the time reversal symmetry is broken in the non - unitary state where @xmath18 the dos in the superconducting state @xmath19 is expressed as follows : @xmath20 where @xmath21 is the heaviside function .
the @xmath10-vector of the state f@xmath22 , the class of group theoretical representaion @xmath23 , is given as @xcite @xmath24 , \label{f1u1i}\ ] ] where @xmath25 .
then , the magnitude of the gap is calculated , leading to the expression @xmath26 it is remarked that the amplitude of the smaller gap @xmath27^{1/2}$ ] has point nodes in the direction @xmath28 , and has a quadratic dependence as @xmath29 around the node on the the fermi sphere .
therefore , the dos is proportional to the excitation energy @xmath30 . with the use of ( [ eigenvaluef1u1i ] ) ,
the dos is calculated numerically by means of the formula ( [ dos ] ) .
the result is shown in fig .
[ fig:1 ] .
the shape of the dos is similar to those for the polar state .
the @xmath10-vector of the state f@xmath31 , the class of group theoretical representaion @xmath23 , is given as @xcite @xmath32 .
\label{f2u1i}\ ] ] the magnitude of the gap is calculated , leading to the same expression as ( [ eigenvaluef1u1i ] ) .
therefore , the dos @xmath19 is the same as shown in fig .
[ fig:1 ] , the polar - like one . in the states
f@xmath22 and f@xmath31 , @xmath33 being the dos in the normal state at the fermi level .
the dashed curve is for the dos of the polar state @xmath34 . ]
the @xmath10-vector of the state f@xmath1 , the class of group theoretical representaion @xmath35 , is given as @xcite @xmath36 , \label{f1uee}\ ] ] where @xmath37 .
then , the magnitude of the gap is calculated , leading to the expression @xmath38 the smaller gap @xmath27^{1/2}$ ] has point nodes in the direction @xmath39 , and has also a quadratic dependence as ( [ eigenvaluef1u1i ] ) .
so , the dos is proportional to the excitation energy @xmath30 , and its explicit dependence is calculated numerically by means of the formula ( [ dos ] ) .
the result is shown in fig .
[ fig:2 ] .
the dos is the same as those for the polar state @xmath34 . in the state f@xmath1 ,
@xmath33 being the dos in the normal state at the fermi level .
this is the same as that of the polar state @xmath34 within numerical errors .
] the @xmath10-vector of the state f@xmath4 , the class of group theoretical representaion @xmath35 , is given as @xcite @xmath40 .
\label{f2uee}\ ] ] the magnitude of the gap is calculated as @xmath41^{1/2}\biggr\ } \label{eigenvaluef2uee}\end{aligned}\ ] ] the smaller gap @xmath27^{1/2}$ ] has point nodes in the direction @xmath39 .
the gap vanishes linearly around the nodes so that the dos is proportional to the square of the energy @xmath30 .
however , the region of @xmath30 where @xmath42 holds is very restricted , i.e. , @xmath43 , as shown in fig
. [ fig:3 ] .
therefore , the temperature dependence of physical quantities is hard to distinguish from those of the polar - like state . in the state f@xmath4 , @xmath33 being the dos in the normal state at the fermi level . ]
other example , in which the point node(s ) give the polar - like dos , is the so - called planar state with e@xmath44 symmetry which is unitary state and was proposed as a candidate of that of upt@xmath45 @xcite .
such a state gives a magnitude of the gap as @xmath46^{1/2}. \label{eigenvaluee2u}\ ] ] this gap has point nodes at @xmath28 and shows the quardatic behaviour around the node on the fermi surface .
so , the quasiparticles around the point nodes should also give the polar - like dos if there exists the fermi surface around the nodes . in conclusion ,
we have pointed out by explicit calculations that the superconducting gap with point nodes in the non - unitary states , f@xmath0 , f@xmath1 , and f@xmath2 , classified by volovik and gorkov , exhibits the polar - like dos which is proportional to the excitation energy itself rather than its square .
this results arises from the fact that the @xmath7-dependence of the gap around the point nodes is quadratic rather than linear as expected in general .
the author acknowledges y. aoki for cummunications urging the author to calculate the density of states of various states in the volovik and gorkov classification scheme . this work was supported in part by a grant - in - aid for coe research program ( no .
10ce2004 ) by the ministry of education , culture , sports , science and technology . |
stellar clusters have been long recognized as important laboratories for astrophysical research .
their study has played an important role in developing an understanding of the universe . for example
, clusters contain statistically significant samples of stars spanning a wide range of stellar mass within a relatively small volume of space .
since stars in such groups share the common heritage of being formed more or less simultaneously from the same progenitor molecular cloud , observations of cluster color - magnitude ( cm ) diagrams can be , and indeed , have been used to provide classical tests of stellar evolution theory .
moreover , clusters offer the smallest physical scale over which a meaningful determination of the stellar initial mass function ( imf ) can be made . because a cluster is held together by the mutual gravitational attraction of its individual members , its evolution is determined by newton s laws of motion and gravity . in many body systems
these interactions are inherently complex and thus clusters are also important testbeds for studies of stellar dynamics . the spatial distribution of clusters has also played a vital role in our understanding of galactic structure .
the distribution of globular clusters , for example , was critical for determining the location of the galactic center , establishing the existence of a galactic halo and setting the overall scale of the galaxy .
young open clusters have provided an important tracer of recent star formation in galaxies and of spiral structure in galactic disks .
such clusters are also of interest for understanding the origin of the solar system , since the presence of rare short - lived radio nuclides in meteoritic samples has long suggested that the sun itself was formed in near proximity to a massive star , and thus most likely within in relatively rich cluster .
little is known or understood about the origin of clusters .
globular clusters in the galaxy were formed billions of years ago . because they are not being formed in the milky way in the present epoch of galactic history , direct empirical study of their formation process is not possible ( except perhaps in certain extragalactic systems and at cosmological distances ) . on the other hand ,
open clusters appear to be continuously forming in the galactic disk and , in principle , direct study of the physical processes leading to their formation is possible .
however , such studies have been seriously hampered by the fact that galactic clusters form in giant molecular clouds ( gmcs ) and during their formation and earliest stages of evolution are completely embedded in molecular gas and dust , and thus obscured from view .
given the constraints imposed by traditional techniques of optical astronomy , direct observation and study of young embedded clusters had been extremely difficult , if not impossible .
however , during the last two decades the development of infrared astronomy and , more recently , infrared array detectors , has dramatically improved this situation .
figure [ rcw38 ] shows optical and infrared images of the southern embedded cluster rcw 38 and amply illustrates the power of infrared imaging for detecting such heavily obscured young clusters .
6.0 in = 3.8 in the deployment of infrared imaging cameras and spectrometers on optical and infrared optimized telescopes has provided astronomers the ability to survey and systematically study embedded clusters within molecular clouds .
almost immediately such studies indicated that rich embedded clusters were surprisingly numerous and that a significant fraction , if not the vast majority , of all stars may form in such systems .
consequently , it is now recognized that embedded clusters may be basic units of star formation and their study can directly address a number of fundamental astrophysical problems .
these include the issues of cluster formation and early evolution as well as the more general problems of the origin and early evolution of stars and planetary systems .
because most stars in the galactic disk may originate in embedded clusters , these systems must play a critical role in understanding the origins of some of the most fundamental properties of the galactic stellar population , such as the form and universality of the stellar imf and the frequencies of stellar and planetary companions .
the purpose of this review is to summarize the current status of observational knowledge concerning young embedded clusters in the galaxy
. we will consider both embedded and partially embedded clusters .
the embedded phase of cluster evolution appears to last between 2 - 3 myrs and clusters with ages greater than 5 myrs are rarely associated with molecular gas ( leisawitz , bash & thaddeus 1989 ) therefore , this review deals with clusters whose ages are typically between 0.5 - 3 million years .
particular emphasis will be placed on embedded clusters within @xmath0 2 kpc of the sun since this presents a sample which is most statistically complete and for which the most detailed observational data are available .
previous reviews of embedded clusters , some with slightly different emphasis , can be found in various conference proceedings ( e.g. , lada & lada 1991 ; zinnecker , mccaughrean & wilking , 1993 ; lada 1998 ; clarke , bonnell & hillenbrand 2000 , elmegreen et al .
2000 , lada et al .
for the purposes of this review we consider clusters to be groups of stars which are physically related and whose observed _ stellar _ mass volume density would be sufficiently large , if in a state of virial equilibrium , to render the group stable against tidal disruption by the galaxy ( i.e. , @xmath3 pc@xmath2 ; bok 1934 ) , and by passing interstellar clouds ( i.e. , @xmath4 pc@xmath2 ; spitzer 1958 ) . furthermore we adopt the additional criterion ( e.g. , adams & myers 2001 ) that the cluster consist of enough members to insure that its evaporation time ( i.e. , the time it takes for internal stellar encounters to eject all its members ) be greater than 10@xmath5 yrs , the typical lifetime of open clusters in the field .
the evaporation time , @xmath6 , for a stellar system in virial equilibrium , is of order @xmath7 , where the relaxation time is roughly @xmath8 and @xmath9 is the dynamical crossing time of the system and @xmath10 the number of stars it contains ( binney & tremaine 1987 ) . the typical crossing time in open clusters is of order 10@xmath11 yrs , so if such a cluster is to survive disintegration by evaporation for 10@xmath5 yrs , its relaxation time must be comparable or to or greater than its crossing time or @xmath12 .
this condition is met when @xmath13 .
therefore , for this review , we define a stellar cluster as : _ a group of 35 or more physically related stars whose stellar mass density exceeds 1.0 pc@xmath2 . _ with our definition we distinguish clusters from multiple star systems , such as small ( n @xmath14 6 ) hierarchical multiples and binaries which are relatively stable systems and small multiple systems of the trapezium type , which are inherently unstable ( ambartsumian 1954 ; allen & poveda 1974 ) .
we also distinguish clusters from stellar associations , which we define as loose groups of physically related stars whose stellar space density is considerably below the tidal stability limit of 1 yr@xmath2 ( blaauw 1964 ) .
clusters , as defined above , can be classified into two environmental classes depending on their association with interstellar matter . _
exposed clusters _ are clusters with little or no interstellar matter within their boundaries .
almost all clusters found in standard open cluster catalogs ( e.g. , lynga 1987 ) fall into this category .
_ embedded clusters _ are clusters which are fully or partially embedded in interstellar gas and dust .
they are frequently completely invisible at optical wavelengths and best detected in the infrared .
these clusters are the youngest known stellar systems and can also be considered protoclusters , since upon emergence from molecular clouds they will become exposed clusters .
a similar classification can be applied to associations .
our definition of a cluster includes stellar systems of two dynamical types or states .
_ bound clusters _ are systems whose total energy ( kinetic + potential ) is negative .
when determining the total energy we include contributions from any interstellar material contained within the boundaries of the cluster .
we define a classical open cluster as a bound , exposed cluster , such as the pleiades , which can live to be at least 10@xmath5 yrs in the vicinity of the sun .
_ unbound clusters _ are systems whose total energy is positive .
that is , unbound clusters are clusters of 35 or more stars whose space densities exceed 1 pc@xmath2 but whose internal stellar motions are too large to be gravitationally confined by the stellar and non - stellar material within the boundaries of the cluster . infrared surveys of molecular clouds are necessary to reveal embedded clusters , since many if not all of their members will be heavily obscured .
the initial identification of an embedded cluster is typically made by a survey at a single infrared wavelength ( e.g. , 2.2@xmath15 m or k - band ) .
the existence of a cluster is established by an excess density of stars over the background . in general
the ease of identifying a cluster depends sensitively on the richness of the cluster , the apparent brightness of its members , its angular size or compactness , its location in the galactic plane and the amount of obscuration in its direction .
for example , it would be particularly difficult to recognize a spatially extended , poor cluster of faint stars located in a direction where there is a high background of infrared sources , ( e.g. , l = 0.0 , b = 0.0 ) .
identification of the individual members of a cluster is considerably more difficult than establishing its existence . in particular , for most clusters the source density of intrinsically faint members is usually only comparable to or even significantly less than that of background / foreground field objects . in such circumstances cluster membership can be determined only on a statistical basis , by comparison with star counts in nearby control fields off the cluster . however , determining whether or not a specific star in the region is a cluster member or not is not generally possible from a star counting survey alone . in situations where field star contamination is non - negligible , other independent information ( e.g. , proper motions , spectra , multi - wavelength photometry ) is required to determine membership of individual stars .
the first deeply embedded cluster identified in a molecular cloud was uncovered in near - infrared surveys of the ophiuchi dark cloud first made nearly thirty years ago using single - channel infrared photometers ( grasdalen , strom & strom 1974 ; wilking & lada 1983 ) .
however , it was nt until the deployment of infrared imaging cameras in the late 1980s , that large numbers of embedded clusters were identified and studied . in a search of the astronomical literature since 1988
we have found that well over a hundred such clusters have been observed both nearby the sun ( e.g. , eiroa & casali 1992 ) and at the distant reaches of the galaxy ( e.g. , santos et al . 2000 ) . to date embedded clusters
have been discovered using three basic observational approaches : 1 ) case studies of individual star forming regions , such as for example , ngc 2282 ( horner , lada & lada 1997 ) , lkh@xmath16 101 ( barsony , schombert & kis - halas 1991 ) , and ngc 281 ( megeath & wilson 1997 ) , 2 ) systematic surveys of various signposts of star formation , such as outflows ( hodapp 1994 ) , luminous iras sources ( e.g. , carpenter et al . 1993 ) , and herbig aebe stars ( testi , natta & palla 1998 ) , and 3 ) systematic surveys of individual molecular cloud complexes ( e.g. , lada et al .
1991b ; carpenter snell & schloerb 1995 ; phelps & lada 1997 ; carpenter , heyer & snell 2000 ; carpenter 2000 ) .
to date most known embedded clusters have been found in surveys of star formation signposts ( 2 ) , in particular the hodapp ( 1994 ) survey of outflows has had by far the most prolific success rate . in the near future , we expect surveys conducted using the data generated by the all sky near - infrared surveys ( i.e. , denis and 2mass ) will likely provide the the most systematic and complete inventory of the embedded cluster population of the galaxy . we have compiled a catalog of embedded clusters within @xmath0 2 kpc of the sun .
the catalog is based on a search of the astronomical literature since 1988 .
this search produced information on well over 100 clusters , most of which were identified in various systematic surveys ( e.g , lada et al .
1991b ; hodapp 1994 : carpenter heyer & snell 2000 ) . from this list we selected 76 clusters which met the following criteria : 1 ) evidence for embedded nature by association with a molecular cloud , hii region or some significant degree of optical obscuration or infrared extinction , 2 ) identification of 35 or more members above field star background within the cluster field , and 3 ) location within @xmath0 2 kpc of the sun .
due to large distance uncertainties of regions slightly beyond 2.0 kpc , such as the w3 molecular clouds ( 1.82.4 kpc ) we have included clusters with published distance estimates of up to 2.4 kpc .
our catalog of nearby embedded clusters is presented in table 1 which lists the cluster name , approximate location , distance , radius , number of members , and absolute magnitude limits of the corresponding imaging observations .
these data were compiled from the references listed in the last column . given the heterogeneous nature of the observations from which this sample is drawn , this catalog can not be considered complete . in particular ,
southern hemisphere regions , such as the vela complex are not well represented since little observational data exists for this portion of the galaxy .
in addition at least 24 additional clusters have been identified in the rosette gmc ( phelps & lada 1997 ) , north american and pelican nebula ( cambresy , beichman & cutri 2002 ) and cygnus x region ( dutra & bica 2001 ) but no properties have been presented in the literature for them and they are not included in our cluster catalog .
also there is a general incompleteness for the more distant clusters due to sensitivity limitations .
we estimate later , that this 2 kpc sample is complete to only factors of 3 - 4 . although not complete , the catalog is , however , likely representative of the basic statistical properties of embedded clusters within @xmath0 2 kpc .
this is because a significant portion of the catalog is comprised of clusters drawn from systematic imaging surveys of individual cloud complexes and a survey ( hodapp 1994 ) which is reasonably complete for clusters associated with outflows , a primary tracer of very recent star formation activity in molecular clouds .
moreover , the subset of clusters found in systematic surveys of nearby gmcs ( orion , monoceros and perseus ) is also likely to be reasonably complete for clusters with 35 or more members .
masses were derived for each cluster in the catalog by assuming a universal imf ( initial mass function ) for all the clusters .
we adopted the imf of the trapezium cluster that was derived by muench et al .
( 2002 ) from modeling of the cluster s k - band luminosity function ( klf ) .
we then used the klf models of muench et al .
( 2002 ) to predict infrared source counts as a function of differing limiting magnitudes for two model clusters whose ages correspond to that of the trapezium ( 0.8 myr ) and ic 348 ( 2 myr ) clusters .
this was necessary to attempt to account for the expected luminosity evolution of the pms populations of embedded clusters ( see discussion below , section 4 ) .
a conversion factor from total source counts ( for a given limiting magnitude ) to total mass was then determined for each synthetic cluster .
the infrared source counts listed for each observed cluster in the catalog were adjusted for distance and variable detection limits and then directly compared with the two model predictions . in most cases
the near - ir limits are faint enough ( i.e. , the imf is reasonably sampled ) that the both models yielded cluster masses that agree extremely well for the two different ages .
given that no age information is available for the bulk of the clusters in the catalog , we adopted a conversion factor that was the average of the trapezium and the ic 348 cluster ages .
additionally , we assumed that all clusters have an average extinction of 0.5 magnitudes in the k band .
the masses we have derived are probably uncertain to less than a factor of 2 for most clusters . the derived cluster masses in our sample range from about 20 to 1100 . in figure [ clusmassfn ] ( left panel )
we present the embedded cluster mass distribution function ( ecmdf ) for all the clusters in our sample .
the ecmdf was derived by summing individual embedded cluster masses ( @xmath17 ) in evenly spaced logarithmic mass bins , 0.5 dex in width , beginning at log(@xmath17)=1.2 .
( the boundaries of the bins were selected to insure that the least populated bin would have more than one object . )
the ecmdf is equal to @xmath18 dn / dlog@xmath17 , and thus differs by a factor of @xmath17 from the mass function ( dn / dlog@xmath17 ) of embedded clusters .
the histogram with the solid line represents the ecmdf for the entire cluster catalog ( i.e. for clusters having n@xmath1935 and d @xmath14 2.4 kpc ) .
the mass distribution function displays two potentially significant features .
first , the function is relatively flat over a range spanning at least an order of magnitude in cluster mass ( i.e. , 50 @xmath20 1000 ) .
this indicates that , even though rare , 1000 clusters contribute a significant fraction of the total stellar mass , the same as for the more numerous 50100 clusters .
moreover , more than 90% of the stars in clusters are found in clusters with masses in excess of 50 corresponding to populations in excess of 100 members .
the flat mass distribution corresponds to an embedded cluster mass spectrum ( dn / d@xmath17 ) with a spectral index of -2 over the same range .
this value is quite similar to the spectral index ( -1.7 ) typically derived for the mass spectrum of dense molecular cloud cores ( e.g. lada , bally & stark .
the fact that the embedded cluster mass spectrum closely resembles that of dense cloud cores is very interesting and perhaps suggests that a uniform star formation efficiency characterizes most cluster forming dense cores .
the index for the mass spectrum of embedded clusters is also essentially the same as that ( -1.5 to -2 ) of classical open clusters ( e.g. , van den berg & lafontaine 1984 ; elmegreen & efremov 1997 ) .
the second important feature in the ecmdf is the apparent drop off in the lowest mass bin ( @xmath0 20 - 50 ) . given that our cluster catalog only included clusters with more than 35 stars ,
it is likely that we will be considerably more incomplete for clusters in the 20 to 50 m@xmath21 range than for the higher mass clusters . to test the significance of this fall off to low cluster masses we consider the mass distribution function of a subset of clusters drawn from a sample of local clouds where observations are reasonably complete .
these were selected from systematic large scale nir surveys of 4 molecular clouds ( l1630 , l1641 , perseus and mon r2 ) without applying any lower limit to the size of the cluster population .
therefore , this sample should be sensitive to the full mass range of clusters in these representative gmcs .
this local molecular cloud sample is plotted as a dashed line in the right panel of figure [ clusmassfn ] .
while the statistical errors due to the small sample size are large , the local sample confirms that there is indeed a drop off in total cluster mass for the lowest mass clusters . as a further check
, we also plot the ecmdf for the hodapp ( 1994 ) sample , again without applying any lower limit to the richness of the cluster .
we choose the hodapp sample since it is selected from a complete sample of outflows , indicative of very young stellar objects , and should not be biased to any particular mass range of clusters .
all three samples are consistent with a fall off in the cluster mass spectrum below about 50 m@xmath21 . even if the cluster samples are not formally complete they should be representative of the total local cluster population within 2 kpc .
therefore , we conclude that the drop off in the ecmdf at masses less than 50 is significant . consequently , there appears to be a characteristic cluster mass ( 50 m@xmath21 ) above which the bulk of the star forming activity in clusters is occurring .
recently , adams and myers ( 2000 ) suggested , based on dynamical modeling of open clusters and knowledge of the cluster formation rate , that most clustered star formation occurs in clusters with between 10 and 100 stars .
however , our results imply that no more than about 10% of all stars are formed in such small clusters .
the discrepancy results from adams & myers use of the battinelli & capuzzo - dolcetta ( 1991 ) catalog of open clusters which undercounts clusters with ages less than 3 myrs and underestimates the cluster formation rate as discussed below . using the masses in table 1 we can estimate the contribution to the star formation rate made by embedded clusters . because of the incompleteness of our sample , this estimate necessarily will be a lower limit .
to minimize the effect of incompleteness we can calculate the star formation rate for the local ( d @xmath14 500 pc ) subset of clusters for which we are likely to be reasonably complete . for this subsample
we calculate a local star formation rate of @xmath22 1 - 3@xmath2310@xmath24 m@xmath21 yr@xmath25 pc@xmath26 assuming typical embedded cluster ages of @xmath01 - 2 myrs .
this rate is in reasonable agreement with the local star formation rate derived from field stars by miller and scalo ( 1979 ) of between 3 - 7 @xmath2310@xmath24 m@xmath21 yr@xmath25 pc@xmath26 .
this suggests that embedded clusters may account for a large fraction of all star formation occurring locally as has been suggested by other considerations ( lada et al .
1991b ; carpenter 2000 ) .
extending our sample to 1 and 2 kpc gives star formation rates of 1 - 0.7 @xmath2310@xmath24 m@xmath21 yr@xmath25 pc@xmath26 , respectively ( for @xmath27 1 myr ) . the systematic drop in the star formation rate with distance likely reflects progressively more incomplete cluster surveys as we move to greater distances .
if we assume that we are nearly complete for the local 0.5 kpc sample , then the drop in the calculated birthrates would imply that we are incomplete by factors of at least 3 to 4 for the 1 to 2 kpc samples .
the embedded cluster catalog can be used to estimate a lower limit to the birthrate of embedded clusters in molecular clouds .
early estimates of the embedded cluster birthrate , based primarily on the number of clusters in the orion cloud complex , found the rate to be extremely high compared to the birthrate of classical open clusters suggesting that only a small fraction of embedded clusters survived emergence from molecular clouds to become classical open clusters ( lada & lada 1991 ) .
our more extensive embedded cluster catalog with an order of magnitude more clusters , allows for a straightforward but much more meaningful estimate of this important formation rate . for ( 53 ) clusters within 2.0 kpc we estimate the formation rate to be between 2 - 4 clusters myr@xmath25 kpc@xmath26 for assumed average embedded cluster ages of 2 and 1 myrs , respectively .
although this rate is a lower limit , it is a factor of 816 times that ( 0.25 myr@xmath25 kpc@xmath26 ) estimated for classical open clusters by elmegreen & clemens ( 1985 ) and 5 - 9 times that ( 0.45 myr@xmath25 kpc@xmath26 ) estimated by battinelli & capuzzo - dolcetta ( 1991 ) for a more complete open cluster sample within 2 kpc of the sun .
this difference in birthrates between embedded and open clusters represents an enormous discrepancy and is of fundamental significance for understanding cluster formation and evolution . by combining our embedded cluster catalog with the open cluster catalog of battinelli & capuzzo - dolcetta ( 1991 ) we can examine the age distribution of all clusters , open and embedded , within 2 kpc of the sun .
the battinelli & cappuzzo - dolcetta catalog contains about 100 classical open clusters and is thought to be complete out to a distance of 2 kpc from the sun for clusters with m@xmath28 -4.5 . in figure [ clustages ]
we plot the distribution of ages of all known clusters both embedded and open within 2 kpc .
embedded clusters populate the lowest age bin .
we have included only those embedded clusters with masses greater than 150 to correspond to the magnitude - limited selection of battinelli & cappuzzo - dolcetta .
this represents roughly one - third of our sample of clusters with published distances of 2 kpc or less .
the average mass of these embedded clusters is 500 , the same as that estimated for the open cluster sample by battinelli & cappuzzo - dolcetta ( 1991 ) .
the number of clusters is found to be roughly constant as a function of age for at least 100 myr . in figure [ clustages ]
we also compare the merged cluster age distribution with the expected age distribution for a constant rate of cluster formation .
our prediction also includes an adjustment for the expected luminosity fading of clusters below the detection limits following the perscription of battinelli & cappuzzo - dolcetta ( 1991 ) .
there is a large and increasing discrepancy between the expected and observed numbers .
these distributions clearly confirm earlier speculations that the vast majority of embedded clusters do not survive emergence from molecular clouds as identifiable systems for periods even as long as 10 myr .
_ figure [ clustages ] suggests an extremely high infant mortality rate for clusters . _
less than @xmath0 4% of the clusters formed in molecular clouds are able to reach ages beyond 100 myr in the solar neighborhood , less than 10% survive longer than 10 myr .
indeed , most clusters may dissolve well before they reach an age of 10 myr . if we consider our entire sample of embedded clusters we predict , after similarly adjusing for fading , that at least 4100 clusters would be detected within 2 kpc of the sun with ages less than 300 myr .
the webda open cluster catalog lists roughly 300 open clusters of this age or less within 2 kpc , suggesting that only about 7% of all embedded clusters survive to pleiades age .
it is likely that only the most massive clusters in our catalog are candidates for long term survival .
roughly 7% of embedded clusters in our catalog have masses in excess of 500 , and this likely represents a lower limit to the mass of an embedded cluster that can evolve to a pleiades - like system .
moreover , figure [ clustages ] also indicates that the disruption rate for bound clusters between 10100 myrs of age is significant , probably due to encounters with gmcs .
many of the observed open clusters in this age range may also not be presently bound ( battinelli & capuzzo - dolcetta 1991 ) .
the discovery of large numbers of embedded clusters coupled with the high birthrates and star formation rates we have inferred for them from our analysis of the data in table 1 , suggests that such clusters may account for a significant fraction of all star formation in the galaxy .
however , because of the incompleteness of our sample it is difficult to produce an accurate estimate of the actual fraction of stars born in embedded clusters from statistical analysis of the data in our catalog .
the best estimates of this quantity are derived from systematic , large scale surveys of individual gmcs . the first systematic attempt to obtain an inventory of high and low mass ysos in a single gmc
was made by lada et al .
( 1991b ) who performed an extensive near - infrared imaging survey of the central regions ( 1 square degree ) of the l1630 gmc in orion .
their survey produced the unexpected result that the vast majority ( 60 - 90% ) of the ysos and star formation in that cloud occurred within a few ( 3 ) rich clusters with little activity in the vast molecular cloud regions outside these clusters .
a subsequent survey by carpenter ( 2000 ) using the 2mass database to investigate the distribution of young stars in 4 nearby molecular clouds , including l1630 produced similar results with estimates of 50 - 100% of the clouds embedded populations be confined to embedded clusters . in both studies the lower limits were derived with no correction for field star contamination which is substantial .
consequently , it is likely that the fraction of stars formed in clusters is very high ( 70 - 90% ) .
subsequent near - infrared surveys of l1630 ( li , evans & lada 1997 ) as well as other molecular clouds such as mon ob1 ( lada , young & greene 1993 ) , the rosette ( phelps & lada 1997 ) and gem ob1 ( carpenter , snell & schloerb 1995 ) have yielded similar findings suggesting that formation in clusters may be the dominant mode of formation for stars of all masses in gmcs and that embedded clusters may be the fundamental units of star formation in gmcs . since gmcs account for almost all star formation in the galaxy , most field stars in the galactic disk may also have originated in embedded clusters .
the intimate physical association with interstellar gas and dust is the defining characteristic of embedded clusters .
embedded clusters can either be partially ( i.e. , a@xmath29 @xmath0 15 mag . ) or deeply ( i.e. , a@xmath29 @xmath0 5100 mag ) immersed in cold dense molecular material or hot dusty hii regions .
the degree of their embeddedness in molecular gas is related to their evolutionary state .
the least evolved and youngest embedded clusters ( e.g. , ngc 2024 , ngc 1333 , ophiuchi , monr2 , and serpens ) are found in massive dense molecular cores , while the most evolved ( e.g. , the trapezium , ngc 3603 , ic 348 ) within hii regions and reflection nebulae or at the edge of molecular clouds .
our present understanding of the relation of dense cores and embedded clusters is largely guided by the coordinated surveys of such clouds as l1630 ( orion b ) , gem ob1 and the rosette ( mon ob2 ) .
these are the clouds for which the most systematic and complete surveys for both embedded clusters and dense molecular material exist ( lada 1992 ; carpenter , snell & schloerb 1996 ; phelps and lada 1997 ) .
these studies all show that embedded clusters are physically associated with the most massive ( 1001000 ) and dense ( n(h@xmath1 ) @xmath0
10@xmath30 cm@xmath2 ) cores within the clouds .
these cores have sizes ( diameters ) typically on the order of 0.5 - 1 pc .
the typical star formation efficiencies range between 10 - 30% for these systems .
the gas densities correspond to mass densities of 10@xmath32 pc@xmath2 suggesting that clusters with central densities of a few times 10@xmath33 pc@xmath2 can readily form from them . typically less than 10% of the area and mass of a gmc is in the form of dense gas .
this gas is non - uniformly distributed through the cloud within numerous discrete and localized cores .
these cores range in size between about 0.1 - 2 pc and in mass between a few solar masses to up to a thousand solar masses .
the largest cores which spawn clusters are highly localized and occupy only a very small fraction ( a few % ) of the area of a gmc .
numerous studies have indicated that the mass spectrum ( @xmath34 ) of dense molecular cloud cores is a power - law with an index of @xmath16 @xmath0 -1.7 ( e.g , lada , bally & stark 1991 , blitz 1993 , kramer et al .
for such a power - law index , most of the mass of dense gas in a cloud will be found in its most massive cores , even though low mass cores outnumber high mass cores .
stars form in dense gas and it is not surprising therefore that a high fraction of all stars form in highly localized rich clusters , since most of a cloud s dense gas is contained in its localized massive cores . moreover , as discussed earlier , the mass spectrum of cores is very similar to that of both embedded and classical open clusters .
not all massive dense cores in molecular clouds are presently forming clusters ( e.g. , lada 1992 ) .
however , in the l1630 cloud , the cores with clusters appear to contain more gas at very high density ( n(h@xmath1 ) @xmath35 cm@xmath2 ) and to be more highly clumped or structured than those cores without clusters ( lada , evans & falgarone 1997 ) . whether this difference in physical properties is
a cause or a result of the formation of a cluster in a massive core is unclear .
studies of the distribution of dust continuum emission in the ophiuchi ( motte , andre & neri 1998 ) , serpens ( testi & sargent 1998 ) and the ngc 2068/2071 ( motte et al .
2001 ) cluster forming cores reveal numerous small scale ( @xmath0 5000 au ) clumps whose mass spectra are characterized by power - law slopes steeper than those of cloud cores but very similar to those which characterize the stellar imf ( see below ) .
this would suggest that there is a direct mapping of clump mass to stellar mass and that the substructure of cluster forming cores reflects the initial conditions of the star formation process in dense cores . on the other hand , detailed observations of the ngc 1333 cluster forming core paint a very different picture . here
the density structure appears to be defined by numerous shells and cavities associated with intense outflow activity ( sandell & knee 2001 ; lefloch et al .
this suggests that much of the structure in the core is the result of excavation by outflows from the young stars themselves and is a post- rather than pre- star formation condition in this and perhaps other cores where outflow activity is occurring . in this regard
it would certainly be interesting to obtain high resolution dust continuum maps of the massive cores that are not forming clusters to see if the mass spectra of pre - stellar clumps in cores for which active star formation has not yet taken place is similar to that in cores in which star formation is already under way .
a significant difference could implicate outflow activity as an agent that transforms an initial cloud - like clump mass spectrum into a clump spectrum more similar to that of stars .
the structure of an embedded cluster is of great interest since it likely possesses the imprint of the physical process responsible for its creation .
in particular , structure in the youngest embedded clusters reflects the underlying structure in the dense molecular gas from which they formed .
although all embedded clusters appear to display structure at some level , they can be characterized by two basic structural types : 1 ) hierarchical - type clusters exhibit surface density distributions with multiple peaks and often significant structure over a large range of spatial scale .
2 ) centrally condensed - type embedded clusters exhibit highly concentrated surface density distributions with relatively smooth radial profiles which can be described to a good approximation by simple power - law functions ( e.g. , @xmath36 ) or king - like ( isothermal ) potentials . in this sense
they are similar to classical open clusters .
the relative frequency of these two types of structure in clusters is presently unknown , although there are clear examples of each in the literature .
examples of hierarchical - type clusters include the deeply embedded double cluster ngc 1333 ( lada , alves & lada 1996 ) and the partially embedded cluster ngc 2264 which is highly structured ( lada , green & young 1992 , piche 1993 ) .
figure [ n2264map ] shows a map of the spatial distribution of infrared sources in the ngc 2264 cluster constructed from the data of lada , greene & young ( 1992 ) .
the cluster appears to be a double - double or quadruple cluster containing at least two levels of hierarchy in its spatial structure .
the existence of hierarchical structure over large scales in star forming regions has been well documented and is thought to be a signature of the turbulent nature of the interstellar gas and dust out of which gmcs , their dense cores and ultimately stars form ( e.g. , elmegreen et al . 2000 ) . 6.0 in = 5.8 in the trapezium - onc cluster ( hillenbrand & hartmann 1998 ) , ic 348 ( lada & lada 1995 ; muench et al 2002 ) , ngc 2024 , ngc 2071 ( lada et al . 1991b ) and ngc 2282 ( horner , lada & lada 1997 ) have been shown to have strong central concentrations and radial surface density profiles that can typically be fit by simple power - laws ( @xmath37 ) as well as king models .
such structure is a signature of the global dominance of gravity ( over turbulence , for example ) in the formation of these systems . whether this structure is a primordial property of these clusters or a result of evolution from an initially more structured and hierarchical state is not clear .
it is interesting to note in this context that each of these clusters is about the spatial extent of a single subcluster in ngc 2264 .
overall , the centrally condensed clusters do exhibit some structure , but it is considerably less dramatic and more subtle than that observed in clusters such as ngc 2264 .
for example , lada & lada ( 1995 ) documented a handful of small , satellite subclusters in the outer regions of ic 348 .
however , given the small numbers of stars within them , one can not rule out the hypothesis that these structures are merely the expected spatial fluctuations in the overall power - law fall off of the cluster s radial density distribution .
images of the trapezium cluster in l band ( 3.4 @xmath15 m ) revealed that about 10% of the sources in the cluster belong to a deeply embedded population , whose surface density distribution differs from that of the main ( less buried ) cluster ( lada et al . 2000 ) .
in particular , the surface density distribution of the embedded population has a different orientation being more closely aligned with the ridge of molecular gas and dust at the back of the cluster .
this population is also not nearly as centrally condensed .
another indication of subtle structure in the trapezium - onc and ic 348 clusters is the evidence for a spatial variation in the clusters mass functions , both of which appear to exhibit an excess of the lowest mass stars in their outer regions perhaps suggestive of some degree of mass segregation ( hillenbrand & carpenter 2001 ; muench et al .
2002 ) .
the question of mass segregation in embedded clusters is of great interest .
evidence for mass segregation in open clusters is well documented ( e.g. , elmegreen et al . 2000 ) and likely due to dynamical evolution and equipartition of energy in those systems .
but is mass segregation also a primordial property of clusters or only achieved after significant dynamical evolution ?
bonnell & melvyn ( 1998 ) have argued that embedded clusters , like the trapezium , are too young to have dynamically evolved significant mass segregation .
if there is mass segregation in this cluster it must be imprinted by the star formation process .
the strongest indication of possible mass segregation in embedded clusters derives from observations which suggest that the most massive stars in some of these systems are preferentially found near the cluster centers .
this phenomenon has been observed in the trapezium ( hillenbrand & hartmann 1998 ) , ngc 2071 and ngc 2024 ( lada et al .
nurnberger & petr - gotzens ( 2002 ) found a steepening of the mass function in the outer region of ngc 3603 and jiang et al .
( 2002 ) found an exponential , rather than power - law , radial decline of massive ob stars in m17 suggesting perhaps more extensive mass segregation in these more massive clusters .
in other embedded clusters , such as monr2 , no evidence has been found for any significant mass segregation ( carpenter et al . 1997 ) . indeed , there are even examples where the high mass stars are found both in the outer regions as well as the central regions of a cluster ( e.g. , ic 5146 ; herbig & dahm 2002 ) . in this context
it is interesting to note that a study of massive young clusters in the lmc also found little evidence for mass segregation ( elson , fall & freeman 1987 ) .
the extent of the phenomenon of mass segregation in embedded clusters remains far from clear .
unfortunately , the inherent uncertainties due to small number statistics make it difficult to definitively investigate this issue in all but the most rich and massive clusters , which themselves are rare and typically very distant .
the ages and age spreads of embedded clusters and their members are fundamental parameters which are among the most uncertain and difficult to determine for such young systems .
knowledge of these two timescales are critical for understanding the evolutionary appearance and state of a cluster and its star formation history .
for example , the ratio of cluster age , @xmath38 , to such timescales as the crossing time ( @xmath9 ) , the relaxation time , ( @xmath39 ) , the evaporation time ( @xmath6 ) , etc . determines the dynamical state of the cluster .
the relation between @xmath38 and the various timescales of early stellar evolution determine the evolutionary demographics of cluster members ( e.g. , the number of cluster members that are protostars , or pms stars with and without disks , etc . ) .
indeed , as will be discussed later , the protostellar , @xmath40 , and disk , @xmath41 , evolution timescales can be inferred from knowledge of @xmath38 and an observational census of either the protostars or disk bearing stars ( respectively ) within a cluster , or a sample of clusters of varying age .
the age spread , @xmath42 , gives the duration of star formation or the gestation timescale for the cluster population .
also of interest is the star formation rate which is essentially given by the ratio of the number of cluster members to the gestation timescale ( i.e. , @xmath43 .
these latter timescales provide important constraints for understanding the physical process of star formation within the cluster . finally , embedded cluster ages are critical for dating molecular clouds , since cloud ages can not be determined from observations of the dust and gas within them .
the age of an embedded cluster provides an interesting lower limit to the age of the molecular cloud from which it formed .
indeed , the relative absence of molecular emission from around clusters with ages in excess of 5 myr , has long suggested that the lifetimes of molecular clouds typically do not exceed 5 - 10 myr ( leisawitz et al 1989 ) .
the most reliable method for age dating clusters and their members is through use of the hertzberg - russell diagram ( hrd ) where the positions of member stars are compared with the locations of theoretical pms evolutionary tracks .
unfortunately , the theoretical trajectories of pms stars on the hrd can be highly uncertain , particularly for cluster ages of 1 myr or less and for low mass stars and substellar objects ( e.g , baraffe et al .
in addition , empirical measurements of the two stellar parameters ( i.e. , luminosities and effective temperatures ) that are necessary for placement of individual stars on the hrd can be very difficult due to such factors as extinction , stellar variability , binarity , veiling and infrared excess which are common characteristics of pms stars .
moreover , often such measurements can only be carried out at infrared wavelengths . as a result of these various factors ,
the ages of embedded clusters are inherently uncertain and often very poorly constrained by observations . 6.0 in in practice it is often more straightforward to place stars on the color - magnitude diagram ( cmd ) and then transform the theoretical tracks to that observational plane for comparison and cluster age determinations .
there exist only a small number of embedded clusters , such as ngc 2024 ( meyer 1996 ) , ic 348 ( herbig , 1998 ) , the trapezium ( hillenbrand 1997 ) , ngc 2264 ( park et al .
2000 ) , ic 5146 ( herbig & dahm 2002 ) etc .
, for which sufficient observations are available to place a significant sample of sources on the cmd and estimate their ages . published estimates for the mean ages of these clusters vary between 0.5 - 3 myr . unfortunately , the uncertainties introduced by using pms tracks can be on the order of the derived age of the cluster .
for example , park et al . ( 2000 ) derived the age of ngc 2264 using 4 different pms models with resulting values of 0.9 , 2.1 , 2.7 , and 4.3 myrs for the age of the cluster .
another problem is that an individual pms model can give different ages for high and low mass stars in the same cluster ( e.g. , hillenbrand 1997 ) . on the other hand , the relative mean
ages of young clusters can be established to much greater precision by using a single or consistent set of pms models to extract the cluster ages ( e.g. , haisch et al . 2001a )
. published comparisons of embedded cluster hrds or cmds with theoretical pms tracks indicate age spreads that are usually of the same order or even greater than the mean cluster ages . for example , age spreads of 5.5 , 8.0 , 10.0 and 15.3 myr were found by park et al . for ngc 2264 using four different pms models for the 2 - 3 myr old cluster .
a good example of an embedded cluster cmd is shown in figure [ cmdic348 ] which displays the cmd for the ic 348 cluster constructed by herbig ( 1998 ) .
the positions of suspected cluster members form a well defined , but relatively wide , pms .
comparison with the isochrones derived from a single set of pms tracks gives a cluster mean age of @xmath0 2 myr and an age spread of approximately 5 myrs . in principle , the detailed distribution of stars within the pms band in the plane of the cmd reflects the star formation history of the embedded population . indeed , from a detailed examination of the distribution of stars in the hrds of a number of embedded clusters ( including the trapezium , ic 348 , ngc 2264 and rho ophiuchi ) , palla and stahler ( 2000 ) produced intriguing evidence for a strongly time dependent star formation rate in these regions . using a consistent analysis and a single set of pms tracks
, they found that star formation appeared to be accelerating with time , with the star formation rate reaching its peak in the last 1 - 2 myr in all clusters still associated with significant molecular gas .
however , it is difficult to evaluate the significance of age spreads and distributions estimated from cmds of embedded clusters because differential extinction , source variability , infrared excess , binarity , and contamination by field stars can contribute significantly to the intrinsic scatter in the diagram ( e.g. , hartmann 2001 ) .
the uncertainties due to such factors as variability , infrared excess and extinction are expected to be greater for younger clusters .
figure [ cmdn2362 ] shows the cmd obtained for ngc 2362 , a 5 myr old , exposed , open cluster where such uncertainties should be minimized ( moitinho et al .
the pms of this cluster is very well defined and relatively narrow indicating a clear upper limit to its age spread of @xmath14 3 myr . in this cluster , where the total and differential extinction are barely measurable , and stellar activity associated with the youngest stars minimal , the cmd indicates a simple star formation history characterized by cluster formation in a rapid , coeval burst of activity less than 3 myr in duration .
these observations also may suggest that a significant portion of the observed scatter in the cmds of other younger and embedded clusters is due to factors other than age .
unfortunately , it is presently not possible to determine whether the large spreads in embedded cluster cmds result from a wide variety of gestation times , accelerating star formation or other factors . since the number of embedded clusters with age determinations is small , a systematic and detailed examination of a larger sample of embedded and young open cluster cmds would be useful in resolving this issue . at present , self - consistent determinations of the mean ages and , in particular , the relative mean ages of clusters may be the most robust information about star formation histories that can be extracted from cmd or hrd analysis . 6.0 in
a fundamental consequence of the theory of stellar structure and evolution is that , once formed , the subsequent life history of a star is essentially predetermined by one parameter , its birth mass . consequently , detailed knowledge of the initial distribution of stellar masses at birth ( i.e. , the imf ) and how this quantity varies through time and space is necessary to predict and understand the evolution of stellar systems , such as galaxies and clusters .
detailed knowledge of the imf and its spatial and temporal variations is also particularly important for understanding the process of star formation , since it is the mysterious physics of this process that controls the conversion of interstellar matter into stars .
unfortunately , stellar evolution theory is unable to predict the form of the imf .
this quantity must be derived from observations .
however this is not a straightforward exercise , since stellar mass is not itself an observable quantity .
stellar radiant flux or luminosity is the most readily observed property of a star .
determination of stellar masses therefore requires a transformation of stellar luminosities into stellar masses which in turn requires knowledge of stellar evolutionary states .
numerous techniques have been employed in an attempt to determine the imf both for the galactic field star population and in open clusters . these techniques and results
have been extensively reviewed in the literature ( e.g. , scalo 1978 , 1986 ; gilmore & howell 1998 ; meyer et al .
2000 ; kroupa 2002 ) .
imfs derived from these studies appear to exhibit two similar general properties .
first , for stars more massive than the sun the imf has a nearly power - law form with the number of stars increasing as the stellar mass decreases . if we adopt the classical definition that the imf ( @xmath44 ) is the number of stars formed per unit volume per unit _ logarithmic _ mass interval , the slope at any point is then : @xmath45 , and @xmath46 1.3 for masses greater than one solar mass ( e.g. , massey 1998 ) .
this is very similar to the value ( 1.35 ) originally derived for field stars by salpeter ( 1955 ) .
second , the imf breaks and flattens near but slightly below 1 , departing significantly from a salpeter slope . at the lowest masses ( i.e. , 0.5 - 0.1 )
, however , there is considerable debate concerning whether the imf declines , rises or is flat and whether or not it extends smoothly below the hydrogen burning limit ( hbl ) to substellar masses .
however , imf determinations for local field stars and in open clusters are hampered by a number of serious difficulties .
to deduce the imf for field stars requires compilation of a volume limited sample of nearby stars .
this in turn requires accurate distance measurements , usually parallaxes , for all stars in the sample . to obtain the necessary complete sample to as low a mass as possible ,
necessitates that this volume be limited to stars relatively nearby the sun ( d @xmath0 5 - 25 pc ) , because of the extreme faintness of the lowest mass stars and the limitations inherent in the distance determinations .
such samples suffer from incompletness for both the highest mass stars , due to their rarity and complete absence in the solar neighborhood , and the lowest mass stars due to their faintness . moreover , such samples contain stars formed over a time interval encompassing billions of years ( essentially the age of the galactic disk ) . therefore , the mass function derived directly from observations of field stars is a present day mass function ( pdmf ) and must be corrected for the loss of higher mass stars due to stellar evolution in order to derive the imf of the sample .
this , in turn , requires the assumptions of both a star formation rate , usually taken to be constant , and a time independent functional form of the imf .
the standard final product is an imf which is time - averaged over the age of the milky way disk .
details regarding any dependence of the imf on either space or time over the history of the galaxy are necessarily lost in the time - averaged imfs derived from nearby field stars .
finally , it is very difficult to measure the imf below the hbl from magnitude - limited studies of the field .
this is because the luminosities of brown dwarfs continue to fade throughout their entire life history , so the number of substellar objects at any brightness is always a time - dependent mixture of brown dwarfs of varying mass and age and it depends sensitively on the formation history of brown dwarfs in the galactic disk .
stellar clusters have played an important role in imf studies because they present equidistant and coeval populations of stars of similar chemical composition .
compared to the disk population , clusters provide an instantaneous sampling of the imf at different epochs in galactic history ( corresponding to the different cluster ages ) and in different , relatively small volumes of space .
this enables investigation of possible spatial and temporal variations in the imf .
however some of these advantages are mitigated by the larger distances of visible open clusters ( compared to local field stars ) which reduces the sensitivity to faint low mass stars and by field star contamination which seriously hampers determination of cluster membership and achievement of completeness , especially at low masses .
in addition dynamical evolution produces both mass segregation and evaporation and depletes the low mass population of clusters requiring uncertain corrections to be applied in order to obtain their imfs .
also , stellar evolution depletes the high mass end of the imf in clusters older than ten million years or so and thus must be accounted for as well .
using young embedded clusters for imf determinations alleviates many of these issues .
for example , embedded clusters are often significantly more compact than visible open clusters minimizing field star contamination , except at the very lowest masses .
moreover , the molecular gas and dust associated with such clusters can screen background stars even at faint magnitudes further reducing background contamination and associated difficulties with membership determinations .
in addition , _ embedded clusters are too young to have lost significant numbers of stars due to stellar evolution or dynamical evaporation , thus their present day mass functions are , to a very good approximation , their initial mass functions_. embedded clusters are also particularly well suited for determining the nature of the imf for low mass stars and substellar objects .
this is because low mass stars in embedded clusters are primarily pre - main sequence stars , and thus are brighter than at any other time in their lives prior to their evolution off the main - sequence . at these young ages , substellar objects or brown dwarfs
are also significantly more luminous than at any other time in their subsequent evolution , and moreover have brightnesses comparable to the lowest mass stars .
indeed , infrared observations of modest depth are capable of detecting objects spanning the entire range of stellar mass from 0.01 to 100 in clusters within 0.5 1.0 kpc of the sun .
however , the study of embedded clusters suffers from two disadvantages : 1 ) the clusters are often heavily obscured and can not be easily observed at optical wavelengths ; 2 ) the stars in such clusters are mostly pre - main sequence stars and the timescale for forming them is an appreciable fraction of the cluster age .
consequently , uncertain corrections for pre - main sequence evolution and non - coevality must be applied to the members to derive mass spectra from luminosity functions .
advances in infrared detectors have enabled the direct observation of such embedded clusters and helped to minimize the first disadvantage .
the second disadvantage requires modeling and is more difficult to overcome ( e.g. , zinnecker et al .
1993 ; comeron et al .
1993 ; fletcher and stahler 1994a ; lada & lada 1995 ; comeron , rieke & rieke 1996 ; lada , lada & muench 1998 ; luhman et al . 2000 ; meyer et al .
2000 ; muench , lada & lada 2000 ; muench et al .
2002 and others ) . finally ,
although the underlying mass functions of most embedded clusters are likely to be a fair representation their initial mass functions , some embedded clusters can be expected to be in extremely early stages of evolution in which active star formation is still contributing to building the ultimate cluster imf .
appropriate caution must be taken when interpreting the mass functions derived in such circumstances .
two basic methods have been generally employed to derive mass functions for embedded clusters .
the first method involves modeling the observed luminosity function of an embedded cluster to derive the form of its underlying mass function ( e.g. , zinnecker et al .
1993 ; fletcher & stahler 1994a , b ; lada & lada 1995 ; megeath 1996 ; muench et al . 2000 , 2002 ) .
the second method involves use of spectroscopy and/or multi - color photometry to place individual stars on the hrd .
comparison of the locations of these stars in the hrd with the predictions of pms evolutionary tracks results in the determination of their individual masses from which the cluster imf is then directly constructed ( e.g. , hillenbrand 1997 , hillenbrand & carpenter 2000 ) . as will be discussed below
, these methods have their own advantages and disadvantages , but in general are complementary .
indeed , in a few studies , imfs for embedded clusters have been derived using a combination of these and similar techniques ( e.g. , comeron et al 1993 , 1996 ; luhman et al
. 1998 , 2000 ) . deriving the imf of an embedded cluster by modelling its luminosity function
first requires the construction of the cluster s luminosity function .
although determining the bolometric luminosity function of a cluster would be most desirable for comparison with theoretical predictions ( e.g. , lada & wilking 1984 ; fletcher & stahler 1994a , b ) , obtaining the multi - wavelength observations necessary to do so would require prohibitive amounts of observing time on telescopes both on the ground and in space .
on the other hand , the monochromatic brightness of a star is its most basic observable property and infrared cameras enable the simultaneous measurement of the monochromatic brightnesses of hundreds of stars .
thus , complete luminosity functions , which span the entire range of stellar mass , can be readily constructed for embedded stellar clusters with small investments of telescope time .
the monochromatic ( e.g. , k band ) luminosity function of a cluster , @xmath47 , is defined as the number of cluster stars per unit magnitude interval and is the product of the underlying mass function and the derivative of the appropriate mass - luminosity relation ( mlr ) :
@xmath48 where @xmath49 is the apparent stellar ( k ) magnitude , and @xmath50 is the stellar mass .
the first term on the right hand side of the equation is the underlying stellar mass function and the second term the derivative of the mlr .
with knowledge of the mlr ( and bolometric corrections ) this equation can be inverted to derive the underlying mass function from the observed luminosity function of a cluster whose distance is known .
this method is essentially that originally employed by salpeter ( 1955 ) to derive the field star imf .
however , unlike main sequence field stars , pms stars , which account for most of the stars in the an embedded cluster , can not be characterized by a unique mlr .
indeed , the mlr for pms stars is a function of time .
moreover , for embedded clusters the duration of star formation can be a significant fraction of the cluster s age .
consequently , to invert the equation and derive the mass function one must model the luminosity function of the cluster and this requires knowledge of both the star formation history ( i.e. , age and age spread ) of the cluster as well as the time - varying pms mass - luminosity relation .
this presents the two major disadvantages for this technique .
first , a priori knowledge of the age or star formation history of the cluster is required and this typically can be derived by placing cluster stars on an hrd . however , this in turn requires additional observations such as multi - wavelength photometry or spectroscopy of a representative sample of the cluster members .
second , pms models must be employed to determine the time varying mass - luminosity relation .
the accuracy of the derived imf therefore directly depends on the accuracy of the adopted pms models which may be inherently uncertain , particularly for the youngest clusters ( @xmath51 yrs ) and lowest mass objects ( @xmath52 ) .
in addition , most pms models predict bolometric luminosities as a function of mass and time , and thus bolometric corrections must be used to transform the theoretical predictions to monochromatic fluxes and magnitudes . despite these complexities , monte carlo modeling of the infrared luminosity functions of young clusters ( muench
, lada & lada 2000 ) has demonstrated that _ the functional form of an embedded cluster s luminosity function is considerably more sensitive to the form of the underlying cluster mass function than to any other significant parameter _
( i.e. , stellar age distribution , pms models , etc . ) .
in fact , despite the significant differences between the parameters that characterize the various pms calculations ( e.g. , adopted convection model , opacities , etc . ) , model luminosity functions were found to be essentially insensitive to the choice of the pms mass - to - luminosity relations .
as discussed below , this reflects the robust nature of pms luminosity evolution .
there are however other limitations of this technique .
in particular , the observed luminosity function of a cluster will always contain unrelated foreground and background field stars along with cluster members .
such field star contamination can be straightforwardly corrected for using imaging observations of nearby control fields .
however at the faintest magnitudes , often corresponding to the substellar mass regime , the field star contamination can be severe and may introduce significant uncertainty in the faint end of the field star - corrected luminosity function . deriving the imf of an embedded population via the hrd requires simultaneous knowledge of both the luminosities and effective temperatures of all the stars in a cluster so that they can be individually placed on the hrd .
this , in turn , requires both photometry to determine stellar luminosities and either colors or a spectrum of each star to determine effective temperature .
for embedded clusters spectroscopy is the preferred method of obtaining a stellar effective temperature because the infrared colors of stars are not intrinsically very sensitive to effective temperature and in addition are significantly altered by extinction and infrared excess associated with the young stars .
the advantage of this method is that the final product is the set of individual masses for all stars for which both spectra and photometry were obtained . in other words
this procedure provides a more detailed determination of the mass function than the first method .
in addition , this procedure also yields the ages of the stars and the star formation history of the cluster .
the major disadvantage of this method is that it requires spectra to be obtained for a complete sample of stars across the entire spectrum of stellar masses . as a result
, a significant investment of integration and telescope time is required to obtain a complete sampling of the imf , particularly at the faint , low mass end . as with the first method ,
field star contamination , particularly at the lowest masses is a serious limitation .
however , this limitation can be overcome with the acquisition of spectra for all stars ( unrelated foreground and background stars plus members ) within a cluster .
but , at faint magnitudes , field stars can easily dominate cluster members , sharply decreasing observing efficiency .
thus sensitivity limitations inherent in spectroscopic observations ultimately restrict the application of this technique to a small number of nearby clusters
. 0.6 in -3.0 in similar to the first method , the technique of deriving a pms star s mass from its location on the hrd is also fundamentally limited by uncertain knowledge of pms evolutionary tracks .
indeed , the imf derived by this technique is more sensitive to uncertainties in pms models than is the imf derived by modeling stellar luminosities .
this is because existing pms models are able to predict the effective temperatures of pms stars with considerably less certainty than their luminosities .
this is illustrated in figure [ mlrteff ] which compares the predictions of standard pms models for luminosities and effective temperatures of pms stars of the same age but varying mass .
the predicted luminosities are essentially degenerate with respect to the pms models used .
although perhaps surprising at first glance , this result can be understood by considering the fact that the luminosity of a pms star is determined by very basic physics , simply the conversion of gravitational potential energy to radiant luminosity during kelvin - helmholtz contraction .
this primarily depends on the general physical conditions in the stellar interior ( e.g. , whether the interior is radiative or partially or fully convective ) .
the close agreement of the model predictions reflects the robust nature of pms luminosity evolution . on the other hand , predicting the effective temperature of such stars , which depends on detailed knowledge of uncertain characteristics ( e.g. , opacities ) of the stellar atmospheres , is a more difficult exercise .
these same models can predict very different locations for such stars on the hrd , corresponding to significant differences in the predicted masses and mass functions .
other limitations that hinder determinations of the imfs of embedded clusters via any of the standard approaches include differential reddening of cluster members , presence of infrared excess and veiling continuum emission .
these effects need to be accounted for either in the modeling or by direct correction of the observed photometry and spectroscopy of individual sources using additional observations .
embedded clusters are also at sufficiently large distances that binary systems within them are almost always unresolved in typical observations .
consequently , the imfs derived by these methods do not include the masses of any unresolved companions .
binary companions can effect the derivations of the imfs in two ways .
first , they can contribute additional flux to the system luminosity .
however , because the vast majority of binaries are not equal mass ( brightness ) systems this contribution is typically small ( 0.1 - 0.2 magnitudes ; e.g. , see simon et al .
1995 ) compared to the typical bin sizes ( 0.5 mag ) used to construct the infrared luminosity functions .
second , the presence of unresolved binaries can result in an underestimate of the numbers of low mass stars in a cluster compared to that expected for a system of stars in which all binaries are resolved , since companion stars are not directly observed or counted ( kroupa et al .
thus , the imfs that are derived are system or primary star imfs . whether or not such a primary star imf should be adjusted by adding in the masses of companion stars depends on the question being considered .
for example , for comparisons with imfs derived for field stars as well as open and globular clusters , the primary star imf is the appropriate imf to use .
if one desires to exactly weigh the amount of interstellar medium transformed into stars by the star formation process , then a primary + companion star imf would be the more appropriate mass function to consider .
unfortunately , the imf of companion stars is not very well known or constrained by existing observational data and determination of a total imf including primary and companion stars is not presently possible . finally ,
since the imf is a statistical property of an ensemble of stars , it can only be meaningfully derived over a mass interval which is statistically well sampled by observations .
the richness of the observed cluster thus sets a basic limit on the level of uncertainty in any derived imf .
the trapezium cluster in orion is the best studied of all embedded clusters .
first identified by trumpler ( 1931 ) , the trapezium cluster is a rich cluster of faint ( mostly pms ) stars embedded within the great orion nebula with an age of about 10@xmath11 yrs ( prosser et al .
1994 ; hillenbrand 1997 ) .
the cluster is approximately 0.3 - 0.4 pc in diameter ( e.g. , lada et al .
2000 ) and contains approximately 700 stars ( hillenbrand & carpenter 2000 ; muench et al 2002 ) .
it is thought to be the highly concentrated core of the more extended orion nebula cluster ( onc ) which contains nearly 2000 stars spread over a region roughly 4 pc in extent ( e.g. , hillenbrand & hartmann 1998 ) . at its center
is the famous trapezium , a close grouping of four ob stars which excite the nebula .
it is a superb target for imf studies because of its youth , richness , compactness , location in front of and partially within an opaque molecular cloud , and its proximity to the sun ( @xmath0 450 pc ) .
these factors combine to enable a statistically significant sampling of the imf from ob stars to substellar objects near the deuterium burning limit ( 0.01 ) with minimal field star contamination .
indeed , this cluster is particularly well suited for investigating the substellar portion of the imf and determining the initial distribution of masses for freely floating brown dwarfs .
deep infrared surveys of this cluster have been performed using the hst ( luhman et al 2000 ) , the keck telescope ( hillenbrand & carpenter 2000 ) , ukirt ( lucas & roche 2000 ) and the ntt ( muench et al 2002 ) and have produced infrared luminosity functions and mass functions which sample well into the substellar mass range . 6.0 in = 7.5 in figure [ trapntt ] shows a three - color infrared image of the cluster resulting from the ntt survey .
muench et al .
( 2002 ) used this data along with observations of the same region obtained with a 1.2 meter telescope to recover the brighter stars typically saturated in deep exposures with the larger telescopes and they produced a complete sampling of the k - band ( 2.2 @xmath15 m ) luminosity function ( klf ) of this cluster spanning the mass range from ob stars to substellar objects near the deuterium burning limit .
figure [ trapklf ] shows the field - star corrected , complete , extinction - limited , klf derived from the muench et al . study .
it counts all stars within a cloud depth of 17 magnitudes of visual extinction with luminosities corresponding to million year old objects with masses @xmath0 0.010 0.015 and greater and is representative of the infrared luminosity functions obtained in all similarly sensitive investigations of this cluster . in particular ,
the klf is found to rise steadily from the brightest stars to around m@xmath53 @xmath0 11 -12 mag where it flattens before clearly falling again to fainter magnitudes .
a clear secondary peak is present at approximately 15th magnitude , which is well into the brown dwarf luminosity range . at lower luminosities the klf rapidly drops off .
= 5.0 in muench et al . derived the imf of the trapezium cluster by using a suite of monte carlo calculations to model the cluster s klf .
the observed shape of a cluster luminosity function depends on three parameters : the ages of the cluster stars , the cluster mass - luminosity relation , and the underlying imf ( i.e. , equation [ eq1 ] ) . with the assumptions of a fixed age distribution , derived from the spectroscopic study of the cluster by hillenbrand ( 1997 ) , a composite theoretical mass - luminosity relation adopted from published pms calculations ( i.e. , bernaconi 1996 ; burrows et al .
1997 ; dantona & mazzitelli 1997 ; schaller 1992 ) , and an empirical set of bolometric corrections , muench et al . varied the functional form of the underlying imf to construct a series of synthetic klfs .
these synthetic klfs were then compared to the observed trapezium klf in a chi - squared minimization procedure to produce a best - fit imf . as part of the modeling procedure ,
the synthetic klfs were statistically corrected for both variable extinction and infrared excess using monte carlo probability functions for these quantities derived directly from multi - color ( jhk ) observations of the cluster .
the best - fit synthetic klf is plotted in figure [ trapklf ] .
the corresponding underlying mass function is displayed in figure [ trapimf ] in the form of a histogram of binned masses of the stars in the best - fit synthetic cluster .
this model mass function represents the imf of the young trapezium cluster .
the main characteristics of this imf are : 1 ) the sharp power - law rise of the imf from about 10 ( ob stars ) to 0.6 ( dwarf stars ) with a slope ( i.e. , @xmath54 = 1.2 ) similar to that of salpeter ( 1955 ) , 2 ) the break from the single power - law rise at 0.6 followed by a flattening and slow rise reaching a peak at about 0.1 , near the hydrogen burning limit , 3 ) the immediate steep decline into the substellar or brown dwarf regime and finally 4 ) the prominent secondary peak near 0.015 or 15 @xmath55 ( jupiter masses ) followed by a very rapid decline to lower masses beyond the deuterium burning limit ( at @xmath0 10 @xmath55 ) .
= 5.8 in the most significant characteristic of this imf is the broad peak , extending roughly from 0.6 to 0.1 .
_ this structure clearly demonstrates that there is a characteristic mass produced by the star formation process in orion .
_ that is , the typical outcome of the star formation process in this cluster is a star with a mass between 0.1 and 0.6 .
the process produces relatively few high mass stars and relatively few substellar objects .
_ indeed , no more than @xmath0 22% of all the objects formed in the cluster are freely floating brown dwarfs_. the overall continuity of the imf from ob stars to low mass stars and across the hydrogen burning limit strongly suggests that the star formation process has no knowledge of the physics of hydrogen burning .
substellar objects are produced naturally as part of the same physical process that produces ob stars ( see also najita , tiede & carr 2000 ; muench et al .
2001 ) . in this respect
the secondary peak at 0.015 is intriguing .
the existence of such a peak may imply a secondary formation mechanism for the lower mass brown dwarfs , similar to suggestions recently advanced by reipurth and clarke ( 2000 ) and thus is potentially very important .
however , the significance that should be attached to this feature depends on the accuracy of the adopted mass - luminosity relations for substellar objects used in the modelling
. these mlrs may be considerably more uncertain than those of pms stars .
indeed , observations of an apparent deficit of stars in the m6-m8 range of spectral types in a number of open clusters are suggestive of the existence of a previously unknown opacity feature in the mlr for such cool stars ( dobbie et al .
the presence of such a feature could produce a peak in the luminosity function and a corresponding artificial peak in the derived mass function if not included in the theoretical mlrs ( e.g. , kroupa tout & gilmore 1990 , 1993 ) . given that this spectral type range corresponds to the temperature range predicted for young low mass brown dwarfs , it is quite plausible that the secondary feature in the derived imf is artificial and does not represent a true feature in the underlying imf .
clearly more data , both observational and theoretical is needed to assess the reality and significance of this intriguing feature .
= 5.0 in figure [ trapcomp ] shows a comparison of trapezium imfs recently derived from a number of different deep infrared imaging surveys using a variety of methods ( lucas & roche 2000 ; hillenbrand and carpenter 2000 ; muench et al .
2000 ; luhman et al .
2000 ; muench et al .
the general agreement of the derived imfs is impressive for @xmath56 0.015 .
the fundamental features ( 1 - 3 ) described above are evident in all the imfs . in the region of the secondary peak ( 4 ) , the agreement is less impressive , likely reflecting the inherent uncertainties in the modeling at the lowest substellar masses .
nonetheless , there is general agreement that the imf turns over and falls off below the hydrogen burning limit and into the substellar mass regime .
however , the precise details , such as the steepness of the falloff and the amplitude of the secondary peak , remain somewhat uncertain .
other differences in details between the various imfs likely result from the uncertainties inherent in the different techniques used in the imf determinations and provide some measure of the overall uncertainty in our present ability to measure the exact form of the imf in this cluster . clearly , however , these infrared studies of the trapezium cluster have established the fundamental properties of its imf .
the derived imf of the trapezuim cluster spans a significantly greater range of mass than any previous imf determination whether for field stars or other clusters ( e.g. , kroupa 2002 ) .
its statistically meaningful extension to substellar masses and the clear demonstration of a turnover near the hbl represents an important advance in imf studies . for masses in excess of the hbl
the imf for the trapezium is in good agreement with the most recent determinations for field stars ( kroupa 2002 ) .
this is to some extent both remarkable and surprising since the field star imf is averaged over billions of years of galactic history , assuming a constant star formation rate , over the age of the galaxy , and over stars originating from very different locations of galactic space .
the trapezium cluster , on the other hand , was formed within the last million years in a region considerably less than a parsec in extent .
moreover , there is evidence that this region is not yet finished producing stars as significant star formation appears to be continuing in the molecular cloud behind the cluster ( lada et al 2000 ) .
taken at face value this agreement suggests that the imf and the star formation process that produces it is very robust , at least for stellar mass objects .
although few other embedded clusters have been as completely studied as the trapezium , the luminosity and mass functions of a number of such clusters have been investigated in various levels of detail .
these include such clusters as ic348 ( lada & lada 1995 ; luhman et al .
1998 ; najita , tiede & carr 2000 ; muench et al .
2003 ) , ngc 1333 ( aspin , sandell & russell 1994 ; lada , alves & lada 1996 ) , ngc 2264 ( lada , young & greene 1993 ) , ngc 2024 ( lada et al . 1991b ; comeron et al . 1996
; meyer 1996 ) , rho ophiuchi ( lada and wilking 1984 ; comeron et al .
1993 ; green and meyer 1995 ; bontemps et al .
2001 ) , serpens ( eiroa & casali 1992 ; giovannetti et al . 1998 ) , m17 ( lada et al .
1991a ) , w3 ( megeath et al .
1996 ) , and ngc 3603 ( eisenhauer et al . 1998 ; brandl et al . 1999 ; nurnberger & petr - gotzens 2002 ) . due varying distances , sizes , sensitivities , methodologies , etc .
, these klfs and corresponding imfs were not uniformly sampled nor investigated using a common systematic approach . as a result
only limited conclusions can be drawn from comparison of all these results with each other and with the imf derived for the trapezium cluster .
on the other hand , when homogeneous data is analyzed with similar methodology more meaningful comparisons of embedded cluster imfs are possible ( e.g. , lada et al .
1996 ; luhman et al .
2000 ; muench et al .
2002 ) .
= 5.0 in = 5.0 in the first conclusion that can be drawn from such studies is that the klf for embedded clusters is not a universal function and statistically significantly variations are present in observed clusters ( e.g. , lada , alves & lada 1996 ) .
this is illustrated in figure [ klfcomp ] which displays the klfs of the trapezium , ic 348 and ngc 2362 clusters .
such variations in the cluster luminosity functions are not unexpected for embedded clusters which consist mostly of pms stars .
even if such clusters were characterized by a universal imf , they would experience significant luminosity evolution as their pms population evolved and collectively approached the main sequence
. this luminosity evolution would be particularly rapid during the first 5 myrs of a cluster s existence when the luminosities of its pms stars experience their most rapid declines .
for example , figure [ klfevo ] shows the expected klfs for three different aged synthetic clusters with identical imfs .
these klfs were calculated from monte carlo simulations by muench et al .
the systematic evolution of the klfs to lower brightness is clearly evident and certainly would be significant enough to be observable .
indeed , modeling of the klfs of clusters such as ic 348 , ngc 2362 and the trapezium , indicates that observed differences in their klfs can be explained by the expected luminosity evolution in clusters that have very similar underlying mass functions ( lada & lada 1995 ; muench et al .
2002 ; alves et al .
this can also be inferred empirically , without resorting to modeling .
comparison of the klfs of young clusters of known age shows that clusters of similar age display klfs of very similar shape while clusters of differing age show the greatest variation of klf forms ( lada , alves & lada 1996 ; alves et al .
2003 ) . 2.0 in to illustrate this further we display in figure [ ic348imfklf ] the background corrected klf and corresponding imf derived for the ic 348 cluster from monte carlo modeling ( muench et al .
ic 348 is the best studied embedded cluster after the trapezium .
although it is not as rich and suffers appreciably more background contamination , this cluster is closer to the solar system , older ( @xmath57 2 - 3 myr . ) and more evolved than the trapezium cluster ( lada & lada 1995 ; herbig 1998 ) .
this places additional constraints for imf modeling making the cluster a good candidate for comparison with the trapezium . as seen in the figure , the imf for this cluster derived from klf modeling compares very well with that of the trapezium over the entire range of mass ( @xmath58 ) over which it was determined . in particular
, the imf displays a broad peak between 0.1 - 0.5 with a clear turn down near the hbl .
this indicates that there is a characteristic mass produced by the star formation process in ic 348 and this mass is essentially the same as that suggested for star formation in the trapezium .
moreover , the size of the substellar population of ic 348 is relatively well determined and constitutes only about @xmath0 25 % of the cluster membership , again similar to the trapezium . because of poor statistics , muench et al
. did not fit the klf in the luminosity range that corresponds to the location of the secondary peak in the substellar imf of the trapezium .
however examination of the ic 348 klf shows a marginally significant bump at the luminosity corresponding to that mass .
the relatively poor statistics in the ic 348 klf at these faint magnitudes is due to a combination of the fact that ic 348 is not as rich as the trapezium and suffers more contamination from background stars since it is not foreground to a similarly opaque background molecular cloud . using independent observations and different methodology to extract imfs from the observations , luhman et al .
( 2000 ) derived imfs for the trapezium , ic 348 and ophiuchi clusters and also found good agreement in the shapes of the imfs of these clusters above the hydrogen burning limit .
these comparisons again suggest a relatively robust or universal imf .
another indication of the robustness of the imf is found in the observations of ngc 3603 , perhaps the most luminous and richest young cluster in the galaxy .
this distant ( @xmath0 7 kpc ) , massive and very dense ob star cluster has been shown to contain low mass ( 0.5 - 1 ) stars in roughly the proportion predicted for a normal imf ( nurnberger & petr - gotzens 2002 ) .
this is interesting given the significant differences in cluster richness ( and central densities ) which likely corresponds to differences in formation environment and internal star formation rates .
indeed , it is entirely possible that the formation process for the ophiuchi , ngc 1333 and even perhaps the trapezium clusters is far from being completed .
yet , it seems that once even a hundred or so stars are formed the fundamental form of the imf is already determined .
differences in the underlying imfs of these clusters may exist , but they are not large enough to be detectable given the existing uncertainties in the measurements , which are likely dominated by the small sizes of the existing samples .
it is interesting as discussed earlier that a consensus appears to be building that imfs for open clusters can also be described by a universal form similar to the field star imf ( massey 1998 ; kroupa 2002 ) .
here we note that recent studies of the pleiades ( bouvier et al .
1998 ; adams et al . 2001 ) and @xmath16 per ( barrado y navascues et al . 2002 ) clusters have extended the imfs of these older ( 10@xmath5 myr . )
clusters to the substellar regime and thus an initial comparison can be made with the embedded clusters .
these open cluster imfs display the same 4 fundamental characteristics described above for the trapezium cluster , including a secondary peak in the low mass portion of the substellar imf ! the imfs do , however , differ in the location of the substellar peak . in the two older open clusters this peak
is found between 40 - 50 @xmath55 , a significantly higher mass than suggested ( @xmath0 15 @xmath55 ) by observations of the trapezium and ic 348 clusters .
this perhaps indicates that the feature is indeed the result of a hitherto unknown feature in the brown dwarf mlr ( dobbie et al .
2002 ) as discussed earlier .
strong similarities have emerged in the imfs derived for field stars , open clusters and embedded clusters suggesting a robust and universal imf . however , there is some evidence that this universal imf may not characterize all star formation events .
for example , hst observations of the extragalactic star burst cluster r136 ( 30 dor ) in the large magellanic cloud suggest that the imf of this o star rich cluster flattens and departs from a salpeter - like power - law rise at a mass of about 2 ( sirianni et al . 2000 ) . this is higher than the mass ( @xmath0 0.6 ) of the corresponding inflection point in the trapezium ,
open cluster and field star imfs and suggests a relative deficit of lower mass stars in this rich o cluster .
however , it is likely that the numbers of solar mass stars in r136 has been underestimated due to the severe crowding in the cluster center .
it is still quite possible that even in this cluster , which contains 1000 o stars , the underlying imf is characterized by the same universal form as that derived for the trapezium . a more significant indication of a departure from a universal imf may be present in recent observations of the nearby taurus clouds .
luhman ( 2000 ) has found a significant ( factor of 2 ) deficit of substellar mass objects in this region relative to the embedded population in the trapezium and ic 348 clusters .
the embedded population of the taurus clouds is an embedded t association consisting of isolated stars and small loose groupings of stars formed over a relatively large area .
these conditions are decidedly different than those which characterize embedded cluster formation . since for taurus
the imf above the hbl appears to be similar to that of clusters and the field ( e.g. , kenyon & hartmann 1995 ) , the finding of a deficit of brown dwarf stars may indicate that the substellar imf is less robust than the stellar imf and thus may be a sensitive function of formation environment and/or initial conditions .
however , more observations would be necessary to test the significance of this possibility .
nearly half a century ago , walker s ( 1956 ) observations of the partially embedded cluster ngc 2264 showed that its late type ( i.e , f and later ) stars were characterized by subgiant luminosities which placed them well above the main sequence on the hrd .
thus , these observations empirically established the pre - main sequence , pre - hydrogen burning , nature of young low mass stars and provided the critical data needed to test and constrain the theory of pms evolution ( e.g. , hayashi 1966 ) .
thirty years later infrared observations of the embedded cluster in ophiuchus enabled the first systematic classification of infrared protostars and young stellar objects based on emergent stellar energy distributions ( wilking & lada 1983 ; lada & wilking 1984 , lada 1987 ) .
such observations were very influential in constructing the early framework for a theoretical understanding low mass star formation ( e.g. , shu , adams & lizano 1987 ) . today , embedded clusters continue to play an important role for the development and testing of theories dealing with the formation and early evolution of both stars and planetary systems .
embedded clusters contain young stellar objects in various evolutionary states , including deeply buried protostars with their infalling envelopes and associated bipolar outflows and more exposed pms stars surrounded by protoplanetary accretion disks .
clusters are particularly useful for statistical studies of the evolution of such objects .
for example , infrared photometric and spectroscopic observations have been used in comparative studies to investigate the physical and evolutionary natures of embedded populations in clusters .
observational investigations of the embedded ophiuchi cluster led to the identification of four classes of young stellar objects corresponding to four phases of early stellar evolution ( lada & wilking 1984 ; adams , lada & shu 1987 ; andre , ward - thompson & barsony 1993 ) . the ratios of the numbers of objects in the various stages coupled with cluster ages have led to estimates for the lifetimes of the various states of early stellar evolution .
for example , the rarity of protostars in even the youngest clusters suggests that the protostellar phases are relatively short , @xmath0 10@xmath59 yrs to @xmath0 10@xmath60 yrs ( e.g. , wilking , lada & young 1989 ; greene et al 1994 , andre & montmerle 1994 ) . coupled with observation of source luminosities and estimates of the masses of the underlying stars ,
these timescales yield mass accretion rates for protostellar evolution which constrain theoretical predictions ( shu , adams & lizano 1987 ) .
more recently , spectroscopic surveys of embedded clusters have been employed to systematically investigate the more detailed physical properties of embedded populations .
low resolution infrared spectroscopic surveys of the embedded population in the rho ophiuchi cluster revealed significant differences between protostellar objects and pms stars indicating that protostellar photospheres were heavily veiled .
the large measured veilings are likely the result of excess emission originating in infalling envelopes and surrounding accretion disks ( casali & matthews 1992 ; casali & eiroa 1996 ; greene & lada 1996 )
. deep , high resolution spectral observations of the embedded populations of the ophiuchi cluster further found the rotation rates of classical protostars ( class i sources ) to be systematically greater than those of disk bearing pms stars ( class ii sources ) in the cluster suggesting that rotation rates in young stellar objects may be modulated by their accretion rates ( greene & lada 1997 , 2002 ) .
the demographics of the protostellar populations in embedded clusters are not well known and poorly constrained by existing observations .
although generally rare , protostars are most abundant in the youngest embedded clusters where as many as 15 - 20% of the members can be protostellar in nature .
examples of embedded clusters that are protostar rich include ngc 1333 , serpens and rho ophiuchi .
likely no more than 10@xmath11 yrs old , these clusters represent the least evolved embedded clusters known .
indeed , both ngc 1333 and the serpens cluster each contain relatively large populations of class 0 sources , the most deeply embedded and least evolved protostellar class identified ( sandell & knee 2001 ; hurt & barsony 1996 ) .
these two clusters are also rich in bipolar outflows and both appear to be undergoing recent bursts in outflow activity ( davis et al .
1999 ; bally , devine & reipurth 1996 ; knee & sandell 2000 ) . both ngc 1333 and serpens are experiencing vigorous star formation and
are likely true protoclusters , still largely in the process of being constructed from molecular gas and dust .
this is in contrast to the situation for older embedded clusters such as ic 348 , which , at an age of @xmath0 2 - 3 myr . , appears to have few protostellar sources and only one outflow ( lada & lada 1995 ; mccaughrean , rayner , & zinnecker 1994 ) .
lada et al .
( 1996 ) have argued that ngc 1333 and ic 348 have similar present day rates of star formation and suggested that ngc 1333 would evolve into a cluster similar to ic 348 if star formation were to continue in ngc 1333 at the same rate for another 2 - 3 myr .
however , once a cluster s age exceeds the timescale ( @xmath0 10@xmath60 yrs ) for protostellar evolution , the older and more evolved the cluster becomes , the smaller the fraction of protostars and outflows it will contain ( e.g. , fletcher & stahler 1994b ) .
since most embedded clusters have ages between 13 myrs , their membership is typically dominated by pms stars .
pms stars come in two varieties : those with circumstellar disks ( class ii ) and those without such disks ( class iii ) .
the frequency of disks within a cluster is directly related to the physical processes of disk formation and evolution . since circumstellar disks
can be the progenitors of planetary systems , knowledge of the cluster disk fraction ( cdf ) and how it evolves with time also have important consequences for understanding the origin of planetary systems .
since most stars probably formed in embedded clusters , the measurement of the cdf in the youngest embedded clusters produces a determination of the initial disk frequency ( idf ) which in turn directly measures the probability of disk formation around newly formed stars . coupled with knowledge of the probability of planet formation in circumstellar disks , the idf can provide an estimate and indirect census of extrasolar planetary systems in our galaxy .
the variation of the cdf with cluster age sets the timescale for disk evolution and thus the duration or lifetime of the circumstellar disk ( and planet building ) phase of early stellar evolution .
this therefore provides a critical constraint for determining the probability of planet formation in circumstellar disks and so therefore also directly bears on the question of the ubiquity of extrasolar planetary systems . do stars of all masses form with circumstellar disks ? is the likelihood of forming planetary systems from disks dependent on the mass of the central star , or on the environment in which the star formed , or both ? these and similar questions concerning the origins of planetary systems may be best addressed by observations of embedded clusters .
= 5.0 in the cdf and its dependence on stellar mass and cluster age , in principle , can be directly measured by obtaining the infrared spectral energy distributions of all or a representative fraction of a cluster population .
this is because stars with circumstellar disks emit excess infrared emission which displays a clear and specific spectral signature in the star s optical - infrared energy distribution ( e.g. , lynden - bell & pringle 1974 ; adams , lada & shu 1987 ) . in practice
, it is prohibitive to obtain complete ( i.e. , 1 - 1000 @xmath15 m ) infrared spectral energy distributions of a cluster population since to do so requires multi - wavelength ground - based and space - based observations .
however , an excess from a disk can be measured at any infrared wavelength sufficiently longward of the peak of the underlying stellar energy distribution of the central star .
although , the longer the wavelength the more unambiguous the infrared excess , observations at wavelengths as short as 2 @xmath15 m ( k - band ) can detect infrared excesses in the majority of disk bearing stars . indeed , essentially all stars with circumstellar disks containing 10@xmath24 of hot dust or more can be detected using ground - based l - band ( 3.4 @xmath15 m ) observations ( lada et al .
2000 ; wood et al .
2002 ) .
near - infrared jhk imaging surveys of numerous embedded clusters have suggested that the idf is relatively high ( @xmath61 50% ; e.g. , stauffer et al .
1994 ; lada , alves & lada 1996 ; carpenter et al 1997 ) . indeed , longer wavelength ( i.e , 3 - 10 @xmath15 m ) surveys , which are more sensitive to disk excess emission , have determined idfs of @xmath0 8085% for the trapezium ( lada et al .
2000 ) and ngc 2024 ( haisch , lada & lada 2000 ; haisch et al .
2001c ) clusters in orion , two of the youngest clusters studied .
these results suggest that circumstellar disks are a natural byproduct of the star formation process and that most stars , independent of mass , are therefore born with the ability to form planetary systems .
however , infrared studies of embedded populations and clusters have also suggested that the duration of the accretion or protoplanetary disk phase may be relatively brief ( 3 - 15 @xmath23 10@xmath11 yrs ; strom et al .
1989 ; skrutskie et al . 1990
; strom , edwards & skrutskie 1993 ; lada & lada 1995 ; brandner et al . 2000 ) .
since the building times for giant gaseous planets are estimated to be 10@xmath62 yrs or longer ( lissauer 2001 ) , it is critically important to accurately constrain the empirical disk lifetime measurements .
young clusters offer an excellent laboratory for investigating disk evolution timescales .
these clusters present statistically significant samples of stars whose mean ages are well determined . moreover , by combining observations of embedded as well as revealed clusters one can create a sample that spans a range in age much greater than that characterizes any individual star forming region .
recently , haisch , lada & lada ( 2001a ) performed the first systematic and homogeneous observational survey for circumstellar disks in young clusters .
they used jhkl imaging observations to determine the cdf in six clusters whose ages varied between 0.3 - 30 myrs .
figure [ diskfreq ] shows the cdf of their sample along with cdf measurements for two embedded t associations ( taurus , kenyon & hartmann 1995 ; chamaeleon i , kenyon & gomez , 2001 ) .
the observations show a rapid decline in the cdf with cluster age .
half of the disks in a cluster appear to be lost within only 2 - 3 myrs and essentially all the disks are gone in about 5 - 6 myrs .
moreover , observations also indicate that disk lifetimes are also functions of stellar mass with disks around higher mass stars evolving more rapidly than disks around low mass stars ( haisch , lada & lada 2001b ) such a rapid timescale for disk evolution places stringent constraints on the timescale for building planets , particularly giant gaseous planets .
however , the jhkl observations of haisch et al ( 2001a ) trace the infrared excess arising from small ( micron - sized ) , hot ( 900 k ) dust grains located in the inner regions of a circumstellar disk ( @xmath0 0.25 au ) .
it is possible for substantial amounts of material to be still present in the disk if either this material is purely gaseous or if the disk has a large inner hole .
the former situation could arise if significant grain growth occurs due dust settling and rapid grain coagulation and growth , as is the expected first step in the formation of planetesimals .
however , any turbulence in the disk would be expected to keep detectable amounts of small dust grains suspended within the gas ( ruden 1999 ) , so dust should still remain a good tracer of total disk mass as the disks evolve .
indeed , recent observations of dust and h@xmath1 emission in more evolved and much less massive debris disks appears to support this assumption ( thi et al .
2001 ) the latter possibility could occur if protoplanetary disks evolved large inner holes as they aged .
it has been noted that in the taurus clouds , near - infrared l - band excess is closely correlated with millimeter - continuum emission which traces outer disk material suggesting that the evolution of the inner and outer disks is homologous and occurs on the same timescale ( haisch et al .
similarly , a recent millimeter continuum survey of 4 young embedded clusters ( ngc 2071 , ngc 2068 , ngc 1333 , and ic 348 ) has found that the variation in the fraction of detected millimeter sources from cluster to cluster is similar to the variation in the fraction of near - ir excess sources ( lada , haisch and beckwith 2003 ) .
this implies that the lifetimes of the inner disk , as detected by near - ir excess , and the outer disk , as detected at millimeter wavelengths , are coupled for embedded clusters as they were for the taurus population .
these observations strengthen the previous l band excess determination of a short disk lifetime of @xmath14 6 million years , and further suggest that the lifetime for massive outer disks may be only as long as 3 million years .
indeed , the dearth of strong millimeter continuum emission from the disk population of the very young trapezium cluster implies even more rapid outer disk evolution ( mundy , looney & lada 1995 ; bally et al . 1998 ; lada 1999 ) .
it may also be possible that disk lifetimes depend on internal cluster environment ( e.g. , the degree to which a cluster produces o stars ) .
systematic surveys of a larger population of clusters are needed to assess this possibility .
in particular , mid - infrared surveys that will be carried out by the sirtf mission , should more definitively address this question .
as discussed earlier in this review , embedded clusters contain significant populations of substellar objects .
indeed , the trapezium cluster alone contains nearly as many brown dwarfs as have been identified in the galactic field to this point in time .
however , in the two clusters ( trapezium and ic 348 ) with the most complete substellar census , substellar objects account for only about 20 - 25% of the total cluster population .
thus , if embedded clusters such as these , supplied the galactic field population , then we would expect to find only 1 brown dwarf for every 3 - 4 stars in a typical volume of galactic space .
embedded cluster research has provided one of the most interesting discoveries concerning the nature of brown dwarfs .
examination of young brown dwarfs and brown dwarf candidates in a number of embedded clusters , including ic 348 , ophiuchus and the trapezium and in other star forming regions , such as the chamaeleon clouds , has produced evidence that some brown dwarfs emit excess emission at near- and mid - infrared wavelengths similar to that emitted by stars with circumstellar disks ( e.g. , wilking et al .
1999 , luhman 1999 ; nata & testi 2001 ; natta et al .
moreover , deep near - infrared images of the trapezium have enabled the first statistically significant determination of the disk frequency for an embedded _ sub_stellar population .
muench et al .
( 2001 ) found a very high fraction ( 65% ) of the substellar objects in the cluster to display infrared excess emission at 2 @xmath15 m suggestive of the presence of circumstellar disks .
moreover , muench et al ( 2001 ) discovered that about 20% of the substellar population were optical proplyds on hst archive images , independently confirming the presence of circumstellar disks around a significant fraction of the substellar objects in the cluster . because observations of 2 @xmath15 m excess undercounts disk bearing stars , the actual fraction of substellar objects with disks is likely greater than 65% and thus very similar to the disk fraction ( 80% ) observed for the stellar population of the cluster ( lada et al .
2000 ) . like stars ,
brown dwarfs appear to be formed surrounded by disks and as a result possess the ability to form planetary systems .
the detection of infrared excess around the faintest sources in the klf of the trapezium cluster also conclusively established the nature of these sources as young objects and cluster members and thus confirmed their status as bona fide substellar objects .
furthermore , the detection of a high disk fraction ( similar to that of the stellar population ) coupled with the smooth continuity of the imf across the hydrogen burning limit provides strong evidence that freely floating brown dwarfs are a natural product of the star formation process in embedded clusters .
the formation process for brown dwarfs is essentially identical to that of stars .
since most field stars appear to be binary systems understanding the origin of binary star systems is of fundamental importance for developing a general theory of star formation .
clusters are important laboratories for investigating binary formation and evolution . in this context
it is interesting that the binary fractions of well studied embedded clusters such as the trapezium ( prosser et al .
1994 ; petr et al . 1998
; simon , close & beck 1999 ) , ic 348 ( duchene , bouvier & simon 1999 ) and rho ophiuchi ( simon et al . 1995 ) are found to be indistinguishable from that of the galactic field . on the other hand
, it is well established that the binary fraction of the embedded population in the taurus - auriga association is significantly ( a factor of 2 ) in excess of that of the galactic field ( e.g. , leinert et al .
1993 ; ghez , neugebauer , & matthews 1993 ; duchene 1999 ) .
this difference in binary fraction provides an important clue relating to the origins of galactic field stars .
specifically , it supports the notion that most field stars originated in embedded clusters rather than in embedded associations such as the taurus - auriga clouds .
however , it has been suggested that the taurus - auriga binary fraction may represent the initial binary fraction even for stars that form in embedded clusters .
kroupa ( 1995a , b ) performed n - body experiments to simulate the evolution of a binary population in a young cluster .
these experiments began with clusters containing 100% binaries and showed that stellar encounters could disrupt binaries and reduce the overall binary fraction with time .
moreover , kroupa found that the field star binary population could be produced by such a model , if most stars formed in what he identified as a dominant - mode cluster , a cluster with roughly 200 systems and a half mass radius of 0.8 pc .
the fact that such systems are relatively common ( e.g. , table 1 ) lends support to this notion .
furthermore , the concept that disruption of multiple systems can occur in clusters appears to be supported by the observation of a deficit of wide binaries in the trapezium cluster ( scally , clarke & mccaughrean 1999 ) .
however , the binary fraction of embedded clusters is not significantly different from that of much older open clusters ( patience & duchene 2001 ) .
this indicates that any evolution of the binary fraction must have occurred on timescales of order 1 myr or less .
such rapid evolution in the binary population can occur for a very dense embedded cluster such as the trapezium ( kroupa , petr & mccaughrean 1999 , kroupa aarseth , & hurley 2001 ) .
it is not clear , however , whether binaries can be as efficiently disrupted in the lower density clusters which account for most star formation .
the question of a universal initial binary fraction remains open .
star formation may produce a variety of outcomes for the emerging binary fraction due to a corresponding variety of initial conditions .
determinations of the binary properties of the protostellar populations in embedded clusters would provide an important test of this question .
if there is a universal initial binary fraction then essentially all protostellar objects must be nascent binary systems .
high resolution infrared imaging and spectroscopic monitoring of class i sources in embedded clusters could resolve this issue .
to understand how an embedded cluster forms we must understand two basic physical processes : 1 ) the formation of a massive , dense core in a gmc and 2 ) the subsequent development of stars from dense gas in the core .
molecular clouds form from the turbulent , diffuse and atomic interstellar medium by a physical process or collection of processes that are far from understood .
overall this process likely involves the complex interplay of such things as spiral density waves , supernova explosions , the galactic dynamo , phase transitions , and various types of instabilities ( e.g , thermal , gravitational , magneto - hydrodynamic , etc . )
( e.g. , elmegreen 1991 , 1993 ) .
the vast majority of gmcs are observed to contain dense gas and signposts of star formation suggesting that the formation of dense cores and then stars proceeds very rapidly after the cloud has formed from the diffuse interstellar medium .
the gmcs that form from the ism are gravitationally bound entities with highly supersonic and turbulent velocity fields .
the turbulent dissipation timescales for gmcs are thought to be shorter than the cloud lifetime suggesting that on global scales the clouds are stabilized against collapse by internal turbulent pressure .
numerical simulations ( e.g. , klessen , heitsch & mac low 2000 ) suggest that under such conditions supersonic turbulent flows can collide , shock and dissipate energy . under the right conditions these collisions can produce dense cores which are gravitationally unstable and decouple from the overall turbulent flow .
the largest and most massive of these fragments are then the potential sites of cluster formation .
the second step of the cluster formation process , the rapid evolution of dense gas in a massive core to form stars , likely involves the continued dissipation of turbulence in the dense gas which is followed by , fragmentation , gravitational instability and the formation of protostellar seeds which grow by accreting their infalling envelopes and then perhaps other surrounding dense gas from the general potential in which they are embedded ( e.g. , see reviews by clarke et al .
2001 , elmegreen et al .
this scenario appears to be quite different than that which has successfully explained the formation of isolated low mass stars from individual low mass cores .
such solitary stars form from initially turbulent , magnetically supported , dense cores which evolve through ambipolar diffusion of magnetic fields to be dynamically unstable and then collapse from the inside out ( shu , adams & lizano 1987 ) .
the cores which form isolated low mass stars in this manner have sizes that are considerably larger than the separation of stars in an embedded cluster .
evidently protostellar cores in clusters must have smaller radii than those which form in isolation .
this suggests that cluster forming cores must experience significant fragmentation in their evolution to form stars .
the physical mechanism that produces this fragmentation is not well understood .
this process likely involves progressive cooling of a marginally stable or collapsing massive core which continuously reduces the jean s mass . in turbulent dense cores
this cooling takes place as a result of dissipation or loss of turbulence . an elegant possibility to account for fragmentation
was proposed by myers ( 1998 ) for the case of mhd turbulence .
if the ionization rate in a massive core is low enough ( i.e. , the extinction is high enough that cosmic rays are the sole source of ionization ) then mhd waves greater than a certain frequency can not couple well to the neutral gas .
this corresponds to a cutoff wavelength , below which turbulence can no longer be sustained ( e.g. , mouschovias 1991 ) .
this situation can lead to the formation of a matrix of critically stable bonnor - ebert condensations or kernels confined by the pressure in the surrounding gas .
myers ( 1998 ) finds that for typical conditions the sizes of these kernels can be comparable to the separation of stars in embedded clusters .
fragmentation can also be produced in the turbulent decay process as flows collide and shock , creating density enhancements , which if massive enough can become gravitationally bound and separate from the general turbulent velocity field ( e.g. , klessen & burkert 2000 , 2001 ) .
once these fragments or kernels become gravitationally unstable they collapse , and gaining mass through infall of surrounding material become protostars .
however , the rates at which protostellar condensations grow must vary significantly within the cluster .
this is because the star formation process must produce a range of stellar and substellar masses spanning three orders of magnitude within a timescale of only a few ( 1 - 2 ) million years in order to reproduce the stellar imf .
as they move through the cluster core protostellar fragments also accrete additional material from the reservoir of residual gas not locked up in other protostellar objects ( bonnell et al .
since all these stellar embryos share a common envelope , a process of competitive accretion begins with initially more massive protostellar clumps or clumps closer to the center of the cluster experiencing higher accretion rates .
the process is highly nonlinear and , even for a cluster with initially equal mass protostellar fragments , can lead to the development of a protostellar mass spectrum similar to that of the stellar imf ( bonnell et al .
2001b ; klessen 2001 ) . in this picture
the more massive stars tend to be formed in the central regions of the cluster leading to some degree of primordial mass segregation .
it is also possible for protostellar fragments in the dense inner regions of the cluster to collide and coalesce leading to the production of very massive stars ( bonnell , bate & zinnecker 1998 ) .
it otherwise would be difficult to built up a massive star from general accretion since radiation pressure from embryonic stars more massive than about 10 can reverse infall and stunt the growth of the star ( e.g. , adams , lada & shu 1987 ) . as indicated in table 1 embedded clusters
are found to span a range in mass of at least two orders of magnitude .
yet their ages are all probably within a factor of 2 of 2 myr .
this fact suggests that the star formation rates in clusters vary widely .
this is even more apparent when one considers the more distant `` star burst '' clusters such ngc 3603 and r 136 which contain almost as many o stars as the entire memberships of embedded clusters like ngc 1333 and ophiuchi .
yet all these clusters are likely within a factor of 2 of the same age .
the finding of a significantly higher star formation rate for the trapezium cluster compared to that in the two embedded clusters ( ic 348 & ngc 1333 ) in the perseus molecular complex led lada et al .
( 1996 ) to suggest that the star formation in the trapezium cluster was externally triggered .
compressive triggering increases the external pressure and thus the density of a core and in doing so speeds up the star formation .
the close association of other clusters with adjacent hii regions has also suggested that triggering may have played an important role in the formation of at least some clusters ( e.g. , s 255 , howard et al .
1997 ; s 281 , megeath & wilson 1997 , w3-w4 , carpenter et al .
2000 ) . at some point ,
when molecular gas is either depleted or expelled from the cluster , star formation ceases and the cluster emerges from its molecular womb . [
[ emergence - from - molecular - clouds - dynamical - evolution - infant - mortality ] ] emergence from molecular clouds : dynamical evolution & infant mortality ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ although the origin of embedded clusters remains a mystery , the subsequent dyanamical evolution of embedded clusters and their emergence from molecular clouds are well posed theoretical problems which have been studied extensively both analytically ( e.g. , hills 1980 ; elmegreen 1983 ; verschueren & david 1989 ) and numerically ( lada , margulis & dearborn 1984 ; goodwin 1997 ; geyer & burkert 2001 ; kroupa , aarseth & huley 2001 ; kroupa & boily 2002 ) . as described earlier , clusters form in massive , dense cores of molecular gas which are strongly self - gravitating .
star formation is an inherently destructive process and upon formation , new stars will immediately begin to disrupt their surrounding gaseous environments
. the birth of high mass stars can be particularly destructive and not only lead to the rapid disruption of a cluster forming core but in addition to the complete dispersal of an entire gmc ( e.g , whitworth 1979 ) .
moreover , outflows generated by a population of low mass stars are also capable of disrupting a massive cloud core in a relatively short time ( e.g. , matzner & mckee 2000 ) . as a result of these effects ,
star formation turns out to be a relatively inefficient process .
thus , the gravitational glue that binds together the system of stars and gas in an embedded cluster may be largely provided by the gas .
stars are then expected to orbit in a deep potential well of the dense core with orbital velocities ( i.e. , @xmath63/r)^{0.5}$ ] ) characteristic of the virial velocities of the dense gaseous material .
as it emerges from a cloud the evolution of an embedded stellar cluster is consequently sensitively coupled to the evolution of its surrounding gas .
the two physical parameters that determine the evolution of an emerging embedded cluster are the star formation efficiency and the timescale of gas dispersal from the cluster .
the star formation efficiency ( sfe @xmath64 ) is a fundamental parameter of both the star and cluster formation processes . since the measurement of the sfe requires a reliable and systematic determination of _ both _ the gaseous and stellar mass within a core , accurate sfe measurements are not generally available for cluster forming regions . in table 2
we list the star formation efficiencies for the sample of nearby embedded clusters ( drawn from table 1 ) which appear to be fully embedded and for which reasonable empirical determinations for the gaseous and stellar mass exist . the sfes range from approximately 10 - 30% .
these estimates are typical of those sometimes estimated for other embedded clusters but are also significantly higher than the global sfes estimated for entire gmcs which are typically only 1 - 5 % ( e.g. , duerr , imhoff & lada 1982 ) .
the fact that the clusters with the lowest sfes ( serpens , ophiuchi , ngc 1333 ) appear to be the least evolved , suggests that the sfe of a cluster increases with time and can reach a maximum value of typically 30% by the time the cluster emerges from its parental cloud core .
whether all clusters can reach sfes as high as 30% before emerging from a molecular cloud is unclear , however , it does seem apparent that clusters rarely achieve sfes much in excess of 30% before emerging from molecular clouds .
references : 1-hodapp 1994 ; 2-carpenter et al . 1993 ; 3-carpenter , heyer & snell 2000 ; 4-megeath et al . 1996 ; 5-lada , alves & lada 1996 ; 6-muench et al . 2003 ; 7-lada & lada 1995 ; 8-aspin & barsony 1994 ; 9-barsony et al .
1991 ; 10-strom , strom & merrill 1993 ; 11-carpenter 2000 ; 12-hillenbrand & carpenter 2000 ; 13-muench et al . 2002 ; 14-lada e. et al . 1991 ; 15-testi , palla & natta 1998 ; 16-horner , lada & lada 1997 ; 17-carpenter et al .
1997 ; 18-carpenter , snell & schloerb 1995 ; 19-marshal , van altena & chiu 1992 ; 20-lada , young & greene 1993 ; 21-alves 2002 , private comm . ; 22-persi et al . 1994 ; 23-kenyon , lada & barsony 1998 ; 24-tapia et al .
1996 ; 25-rho et al . 2001 ; 26-chini , kruegel & wargau 1992 ; 27-lada c. et al .
1991 ; 28- van den ancker et al . 1997 ; 29-eiroa & casali 1992 ; 30-chen et al . 2002 ; 31-wilking et al 1997 ; 32 hodapp & rayner 1991 ; 33-herbig & dahm 2002 . | stellar clusters are born embedded within giant molecular clouds ( gmcs ) and during their formation and early evolution are often only visible at infrared wavelengths , being heavily obscured by dust . over the last 15 years advances in infrared detection capabilities have enabled the first systematic studies of embedded clusters in galactic molecular clouds . in this article
we review the current state of empirical knowledge concerning these extremely young protocluster systems . from a survey of the literature
we compile the first extensive catalog of galactic embedded cluster properties .
we use the catalog to construct the mass function and estimate the birthrate for embedded clusters within @xmath0 2 kpc of the sun .
we find that the embedded cluster birthrate exceeds that of visible open clusters by an order of magnitude or more indicating a high infant mortality rate for protocluster systems .
less than 4 - 7% of embedded clusters survive emergence from molecular clouds to become bound clusters of pleiades age . the vast majority ( 90% ) of stars that form in embedded clusters form in rich clusters of 100 or more members with masses in excess of 50 .
moreover , observations of nearby cloud complexes indicate that embedded clusters account for a significant ( 70 - 90% ) fraction of all stars formed in gmcs .
we review the role of embedded clusters in investigating the nature of the imf which , in one nearby example , has been measured over the entire range of stellar and substellar mass , from ob stars to subsellar objects near the deuterium burning limit .
we also review the role embedded clusters play in the investigation of circumstellar disk evolution and the important constraints they provide for understanding the origin of planetary systems .
finally , we discuss current ideas concerning the origin and dynamical evolution of embedded clusters and the implications for the formation of bound open clusters .
epsf.tex psfig.sty 2h@xmath1 3pc@xmath2 |
exchanges in networks have been studied for a long time in both sociology and economics . in sociology , they appear under the name of _ network exchange theory _ , a field which studies the behaviour of agents who interact across a network to form bilateral relationships of mutual benefit .
the goal is to determine how an agent s location in the network influences its ability to negotiate for resources @xcite . in economics , they are known as _ cooperative games _ and have been used for studying the distribution of resources across a network , for example in the case of two - sided markets @xcite @xcite . from a theoretical perspective
the most commonly used framework for studying such exchanges is that of _ network bargaining games_. the model consists of an undirected graph @xmath0 with edge weights @xmath1 and vertex capacities @xmath2 .
the vertices represent the agents , and the edges represent possible pairwise contracts that the agents can form .
the weight of each edge represents the value of the corresponding contract .
if a contract is formed between two vertices , its value is divided between them , whereas if the contract is not formed neither vertex receives any profit from this specific contract .
the capacity of each agent limits the number of contracts it can form .
this constraint , together with an agent s position in the network determine its bargaining power .
a _ solution _ for the network bargaining model specifies the set of contracts which are formed , and how each contract is divided . specifically , a solution consists of a pair @xmath3 , where @xmath4 is a @xmath5-matching of the underlying graph @xmath6 , and @xmath7 is a vector which assigns each edge @xmath8 two values @xmath9 corresponding to the profit that agent @xmath10 , respectively agent @xmath11 , earn from the contract @xmath8 . to be a valid solution , the two values @xmath12 and @xmath13 must add up to the value of the contract whenever the edge @xmath8 belongs to the @xmath5-matching @xmath4 , and must be zero otherwise .
solutions to network bargaining games are classified according to two main concepts : _ stability _ and _ balance_. a solution is stable if the profit an agent earns from any formed contract is at least as much as its outside option .
an agent s _ outside option _ , in this context , refers to the maximum profit that the agent can rationally receive by forming a new contract with one of its neighbours , under the condition that the newly formed contract would benefit both parties .
the notion of balance , first introduced in @xcite , @xcite , is a generalization of the nash bargaining solution to the network setting .
specifically , in a balanced solution the value of each contract is split according to the following rule : both endpoints must earn their outside options , and any surplus is to be divided equally among them .
balanced solutions have been shown to agree with experimental evidence , even to the point of picking up on subtle differences in bargaining power among agents @xcite .
this is an affirmation of the fact that these solutions are natural and represent an important area of study .
there is a close connection between network bargaining games and _ cooperative games_. specifically given a solution @xmath3 to the network bargaining game @xmath14 we can define a corresponding payoff vector @xmath15 , where @xmath16 is just the total profit earned by vertex @xmath10 from all its contracts in the solution @xmath3 .
then this vector @xmath15 can be seen as a solution to a corresponding cooperative game @xmath17 defined as follows : we let @xmath18 denote the set of players , and for every subset @xmath19 of players , we define its _ value @xmath20 _ as the weight of the maximum weight @xmath5-matching of @xmath21 $ ] .
this game is also known as the _ matching game _ @xcite .
the subsets @xmath19 are referred to as _ coalitions _ , and the value @xmath20 of each coalition is interpreted as the collective payoff that the players in @xmath22 would receive if they decide to cooperate .
the players are assumed to be able choose which coalitions to form , and their objective is to maximize their payoffs .
the underlying assumption in cooperative game theory is that the grand coalition @xmath23 will form , and the question becomes how to distribute the payoff @xmath24 among the players .
a vector @xmath15 describing such a distribution is referred to as an _
allocation_. given an allocation @xmath15 , the excess of a coalition @xmath22 is defined as @xmath25 .
intuitively , a negative excess means that the players of the coalition have no incentive to break away from the grand coalition , since the collective payoff they could achieve by forming the new coalition is less then what they are currently receiving .
the _ power _ of player @xmath10 over player @xmath11 with respect to the allocation @xmath15 is the maximum excess achieved by a coalition which includes @xmath10 but excludes @xmath11 .
two important concepts from cooperative games are those of the _ core _ and _ prekernel_. the core consists of allocations for which no coalition has a negative excess , whereas the prekernel consists of all allocations with symmetric powers . * our contribution and results . *
our main result is providing the first polynomial time algorithm for computing balanced solutions for network bargaining games with general capacities and fully characterizating the existence of balanced solutions for these games .
specifically we show the following results in sections [ sec1 ] and [ sec2 ] respectively : result 1 .
_ there exists a polynomial time algorithm which given an instance of a network bargaining game with general capacities and a maximum weight @xmath5-matching @xmath4 , computes a balanced solution @xmath3 whenever one exists . _ result 2 . _
a network bargaining game with general capacities has a balanced solution if and only if it has a stable one .
_ our method relies on a new approach of reducing a general capacity instance to a network bargaining game with unit capacities defined on an auxiliary graph .
this allows us to use existing algorithms for obtaining balanced solutions for unit capacities games , which we can then transform to balanced solutions of our original instance .
this represents a departure from previous approaches of @xcite which relied on proving an equivalence between the set of balanced solutions and the intersection of the core and prekernel of the corresponding matching game . in section [ sec3 ]
we show that such an approach can not work for our case , since this equivalence does not extend to all instances of general capacity games : result 3 .
_ there exists an instance of a network bargaining game with general capacities for which we can find an allocation in the intersection of the core and prekernel such that there is no corresponding balanced solution for this allocation_. despite this result , we provide two necessary conditions which ensure that the correspondence between the set of balanced solutions and allocations in the intersection of the core and prekernel is maintained . using the definition of _ gadgets _ from section [ gadgets ] we have the following result given in section [ sec4 ] : result 4 . _
if the network bargaining game has no gadgets and the maximum @xmath5-matching @xmath4 is acyclic , the set of balanced solutions corresponds to the intersection of the core and prekernel .
_ * related work . * kleinberg and tardos @xcite studied network bargaining games with unit capacities and developed a polynomial time algorithm for computing the entire set of balanced solutions .
they also show that such games have a balanced solution whenever they have a stable one and that a stable solution exists if and only if the linear program for the maximum weight matching of the underlying graph has an integral optimal solution .
bateni et al .
@xcite consider network bargaining games with unit capacities , as well as the special case of network bargaining games on bipartite graphs where one side of the partition has all unit capacities .
they approach the problem of computing balanced solutions from the perspective of cooperative games .
in particular they use the matching game of shapley and shubik @xcite and show that the set of stable solutions corresponds to the core , and the set of balanced solutions corresponds to the intersection of the core and prekernel . like we do here ,
kanoria et al .
@xcite also study network bargaining games with general capacities .
they show that a stable solution exists for these games if and only the linear program for the maximum weight @xmath5-matching of the underlying graph has an integral optimal solution .
they are also able to obtain a partial characterization of the existence of balanced solutions by proving that if this integral optimum is unique , then a balanced solution is guaranteed to exist .
they provide an algorithm for computing balanced solutions in this case which uses local dynamics but whose running time is not polynomial .
an instance of the _ network bargaining game _ is a triple @xmath26 where @xmath27 is a undirected graph , @xmath28 is a vector of edge weights , and @xmath29 is a vector of vertex capacities .
a set of edges @xmath30 is a @xmath5-matching of @xmath6 if @xmath31 for all @xmath32 .
given a @xmath5-matching @xmath4 , we let @xmath33 denote the degree of vertex @xmath10 in @xmath4 . we say that vertex @xmath10 is _ saturated _ in @xmath4 if @xmath34 . a _ solution _ to the network bargaining game
@xmath14 is a pair @xmath35 where @xmath4 is a @xmath5-matching of @xmath6 and @xmath36 assigns each edge @xmath8 a pair of values @xmath37 such that @xmath38 if @xmath39 and @xmath40 otherwise .
the _ allocation _ associated with the solution @xmath3 is the vector @xmath41 where @xmath16 represents the total payoff of vertex @xmath10 , that is for all @xmath32 we have @xmath42 .
the _ outside option _ of vertex @xmath10 with respect to a solution @xmath3 is defined as @xmath43 } \min_{vw \in m } z_{vw } { \right)}{\right ) } , \notag\end{aligned}\ ] ] where @xmath44 is the indicator function for the event @xmath45 , which takes value one whenever the event holds , and zero otherwise . if @xmath46 then we set @xmath47 .
we write @xmath48 instead of @xmath49 whenever the context is clear .
a solution @xmath3 is _ stable _ if for all @xmath39 we have @xmath50 , and for all unsaturated vertices @xmath10 we have @xmath51 .
a solution @xmath3 is _ balanced _ if it is stable and in addition for all @xmath39 we have @xmath52 .
the definitions from the previous section simplify in the case where all vertices have unit capacities .
specifically a _ solution _ to the unit capacity game @xmath53 is a pair @xmath54 where @xmath4 is now a matching of @xmath6 and @xmath55 assigns a value to each vertex such that for all edges @xmath39 we have @xmath56 and for all @xmath32 not covered by @xmath4 we have @xmath57 .
since each vertex has at most one unique contract , the vector @xmath15 from the solution @xmath58 is also the allocation vector in this case .
the _ outside option _ of vertex @xmath10 can now be expressed as @xmath59 where as before we set @xmath60 whenever @xmath61 .
a solution @xmath58 is _ stable _ if for all @xmath32 we have @xmath62 and _ balanced _ if it is stable and in addition @xmath63 for all @xmath39 .
given an instance @xmath14 of the network bargaining game we let @xmath64 and define the _ value _ @xmath20 of a set of vertices @xmath19 as @xmath65 $ } } w(m ) . \notag\end{aligned}\ ] ] then the pair @xmath17 denotes an instance of the _ matching game _ of shapley and shubik @xcite .
we will refer to this as the matching game associated with the instance @xmath14 .
given @xmath66 and two vertices @xmath67 we define the _ power _ of vertex @xmath10 over vertex @xmath11 with respect to the vector @xmath15 as @xmath68 where @xmath69 .
we write @xmath70 instead of @xmath71 whenever the context is clear .
the _ core _ of the game is defined as the set @xmath72 the _ prekernel _ of the game is the set @xmath73
the first attempt towards computing balanced solutions for the network bargaining game with general capacities is to use the connection to cooperative games presented in @xcite . for the special class of unit capacity and
constrained bipartite games , bateni et al .
show that the set of stable solutions corresponds to the core , and the set of balanced solutions corresponds to the intersection of the core and prekernel of the associated matching game .
this implies that efficient algorithms , such as the one of @xcite , can be used to compute points in the intersection of the core and prekernel from which a balanced solution can be uniquely obtained .
the first question of interest is whether this equivalence between balanced solutions and the intersection of the core and prekernel extends to network bargaining games with arbitrary capacities .
the following lemma proves that this is not always the case .
[ lem : 1 ] there exists an instance @xmath14 of the network bargaining game and a vector @xmath75 such that there exists no balanced solution @xmath3 satisfying @xmath76 . consider the following graph where every vertex has capacity 2 and the edge weights are given above each edge consider the vector @xmath15 defined as @xmath77 for all @xmath32 .
we now show that the vector @xmath15 is in the intersection of the core and prekernel and there exists no balanced solution @xmath3 corresponding to @xmath15 .
let @xmath78 denote the outer cycle on vertices @xmath79 and let @xmath80 and @xmath81 denote the inner cycles on vertices @xmath82 and @xmath83 respectively .
the unique optimal @xmath84-matching is the cycle @xmath78 with weight @xmath85 . since any stable , and therefore balanced
, solution must occur on a maximum weight @xmath5-matching @xcite , any balanced solution @xmath3 will have @xmath86 .
it can be easily checked that @xmath87 . to check that @xmath88 we compute the powers @xmath89 for all pairs of vertices @xmath90 . for the pair @xmath91 we have : @xmath92 similarly for the pair @xmath93 we have : @xmath94 and for the pair @xmath95 : @xmath96
hence the pairs @xmath97 and @xmath98 satisfy the prekernel condition . by symmetry
so do @xmath99 and @xmath100 .
note that for any pair @xmath101 of non - adjacent vertices , one of the two cycles @xmath80 or @xmath81 will contain @xmath10 but not @xmath11 , and viceversa .
therefore @xmath102 for all non - adjacent pairs @xmath101 .
this proves that @xmath15 is in the prekernel .
we now show that there is no vector @xmath7 such that @xmath3 is a balanced solution corresponding to the vector @xmath15 .
first note that vertices @xmath103 and @xmath104 have an outside option of zero in any solution , since there are no edges in @xmath105 incident with these vertices .
hence the contracts @xmath98 and @xmath106 have to be split evenly in any balanced solution .
since each vertex must have a total profit of @xmath107 from its two contracts in @xmath4 , this uniquely determines all values of the vector @xmath7 , which are shown in the figure below the minimum contract of both @xmath108 and @xmath45 is @xmath109 and therefore @xmath110
. however , the edge @xmath111 ( and also the edge @xmath112 by symmetry ) violates the balance condition since @xmath113 while @xmath114 .
note that this instance does possess a balanced solution as shown in the figure below here the outside option of both @xmath108 and @xmath45 is @xmath115 and all edges in the matching satisfy the balance condition .
we also remark that the allocation associated with this balanced solution is also in the intersection of the core and prekernel . in view of lemma [ lem : 1 ]
, we can not hope to extend the correspondence between balanced solutions and allocations in the intersection of the core and prekernel to all network bargaining games .
however we can generalize the results of @xcite by characterizing a larger class of network bargaining games , including unit capacity and constrained bipartite games , for which this correspondence holds .
we achieve this by defining a certain gadget whose absence , together with the fact that the @xmath5-matching @xmath4 is acyclic , will be sufficient for the correspondence to hold .
let @xmath14 be an instance of the network bargaining game and @xmath3 a solution .
consider a vertex @xmath32 with @xmath116 and let @xmath11 be a neighbour of @xmath10 in @xmath4 .
let @xmath117 be vertex @xmath10 s best outside option and if @xmath117 is saturated in @xmath4 , let @xmath118 be its weakest contract . using these definitions we have @xmath119}z_{v'u'}. \notag\end{aligned}\ ] ]
we say that @xmath10 is a @xmath120 vertex in the solution @xmath3 if at least one of the following two conditions holds : 1 .
there is a @xmath121 path in @xmath4 , 2 .
there is a @xmath122 path in @xmath4 , that does not pass through vertex @xmath117 .
we refer to such @xmath121 or @xmath122 paths as @xmath123 of the solution @xmath3 .
the following figure depicts these two types of gadgets , solid lines denote edges in @xmath4 and dashed lines denote edges in @xmath105 .
we can now state our main theorem of this section .
[ thm : 1 ] let @xmath14 be an instance of the network bargaining game .
let @xmath124 and @xmath3 be a corresponding stable solution so that @xmath42 for all @xmath32 . if the following two conditions are satisfied 1 .
@xmath4 is acyclic , 2 .
there are no bad vertices in the solution @xmath3 , then , the following statement holds @xmath125 fix @xmath39 .
note that it suffices to show @xmath126 , since this would imply that @xmath127 if and only if @xmath128 .
our strategy is to first show that @xmath70 is upper bounded @xmath129 , after which it will be sufficient to find a set @xmath130 for which @xmath131 achieves this upper bound .
we start with the following lemma .
[ lem : 2 ] @xmath132 .
let @xmath133 such that @xmath134 and @xmath135 .
let @xmath136 be a maximum weight @xmath5-matching in @xmath137 $ ] .
then @xmath138 define the set of ordered pairs @xmath139 so that @xmath140 since @xmath3 is a stable solution it follows that @xmath4 is a @xmath5-matching of maximum weight .
hence any edge in @xmath141 must have at least one saturated endpoint .
let @xmath142 be the set of vertices in @xmath130 which are saturated in @xmath4 .
for each @xmath143 $ ] define the sets of ordered pairs @xmath144 note that all these sets are pairwise disjoint , and @xmath145 for all @xmath143 $ ] .
now since each @xmath146 is saturated in @xmath4 it follows that @xmath147 .
therefore we can fix an arbitrary mapping @xmath148 so that each element of @xmath149 is mapped to a distinct element of @xmath150 .
we now assign to each edge @xmath151 a set of ordered elements @xmath152 .
as previously observed , each such edge @xmath153 must have at least one saturated endpoint . hence let @xmath154 and define : @xmath155 \backslash { \left\ { } \newcommand{\rc}{\right\}}i \rc$ , } \\ { \left\ { } \newcommand{\rc}{\right\}}\phi_i { \left(}a_i , y { \right)}\rc & \text{otherwise . } \\
\end{cases } \notag\end{aligned}\ ] ] it follows from the definition of the sets @xmath149 and the choice of the mapping @xmath156 that the sets @xmath157 are well defined , are pairwise disjoint , and are all subsets of @xmath158 .
let @xmath159 then from equation and the fact that each ordered pair of @xmath158 belongs to at most one @xmath157 set we obtain @xmath160 now it follows from stability that for all @xmath151 we have @xmath161 to see this , consider @xmath162 . if @xmath163 for some @xmath164 \backslash { \left\ { } \newcommand{\rc}{\right\}}i \rc$ ] then @xmath165 .
now @xmath166 represents the profit that @xmath146 gets from one of his contracts in @xmath4 , and similarly @xmath167 represents the profit that @xmath168 earns from one of his contracts in @xmath4 .
since the edge @xmath169 is not in @xmath4 , by stability we must have @xmath170 .
in the other case where @xmath171 is not a saturated vertex in @xmath130 , we have @xmath172 .
since @xmath171 is an outside option for @xmath146 and @xmath173 represents the profit that @xmath146 gets from one of his contracts in @xmath4 , stability for vertex @xmath146 implies that @xmath174 as required .
now suppose that @xmath175 .
then from and we have @xmath176 since @xmath177 , this proves the lemma in this case .
if on the other hand , there exists a set @xmath178 such that @xmath179 . then using and again we have @xmath180 now @xmath179 implies that @xmath181 must be an edge in @xmath141 that is incident to vertex @xmath10 . hence @xmath182 where @xmath183 is an outside option for vertex @xmath10 .
if @xmath183 is not saturated then the set @xmath184 is empty .
otherwise if @xmath183 is saturated , this set contains a unique ordered pair @xmath185 such that @xmath186 .
therefore it follows from the definition of @xmath48 that @xmath187 as desired .
hence it suffices to find a set @xmath188 such that @xmath134 , @xmath135 and show that @xmath189 .
given a set of vertices @xmath22 we let @xmath190 denote the edges of @xmath4 which have both endpoints in @xmath22 .
note that for any set of vertices @xmath22 we have @xmath191 we define @xmath192 to be the set of components of @xmath6 induced by the edges in @xmath4 .
since @xmath10 and @xmath11 are neighbours in @xmath4 they will be in the same component , call it @xmath193 .
now suppose we remove the edge @xmath8 from @xmath193 .
since @xmath4 is acyclic , this disconnects @xmath193 into two components @xmath194 and @xmath195 , containing vertices @xmath10 and @xmath11 respectively .
now @xmath196 is a valid @xmath5-matching of @xmath194 hence applying equation to the vertex set the component @xmath194 we obtain @xmath197 if @xmath198 then setting @xmath130 to be the vertex set of component @xmath194 completes the proof for this case . hence it remains to consider the case where @xmath199 .
then by stability of the solution @xmath3 vertex @xmath10 must be saturated in @xmath4 .
let @xmath117 be vertex @xmath10 s best outside option .
* case 1 * : @xmath200 and @xmath117 is not saturated in @xmath4 . since @xmath201 and @xmath117 is not saturated in @xmath4 the set of edges @xmath202 is a valid @xmath5-matching of @xmath194 and therefore @xmath203 * case 2 * : @xmath200 and @xmath117 is saturated in @xmath4 .
let @xmath118 be the weakest contract of @xmath117 and suppose we remove the edge @xmath204 from @xmath194 .
since @xmath4 is acyclic , this disconnects @xmath194 into two components . from condition
@xmath205 we know that @xmath118 is not on the @xmath206 path in @xmath4 .
hence @xmath10 and @xmath117 are in the same component of @xmath207 .
denote this component by @xmath208 .
now @xmath209 is a @xmath5-matching of @xmath208 and thus @xmath210 [ fig ] * case 3 * : @xmath211 and @xmath117 is not saturated in @xmath4 .
condition @xmath205 implies that there is no @xmath121 path in @xmath4 , hence the fact that @xmath211 implies that @xmath212 .
let @xmath213 be the component in @xmath192 that contains vertex @xmath117 .
now @xmath214 is a @xmath5-matching of @xmath215 and therefore @xmath216 * case 4 * : @xmath211 and @xmath117 is saturated in @xmath4 . as in case 3 we let @xmath213 be the component in @xmath192 that contains vertex @xmath117 .
let @xmath118 be the weakest contract of @xmath117 and suppose we remove the edge @xmath204 from @xmath213 .
since @xmath4 is acyclic , this disconnects @xmath213 into two components and we let @xmath217 be the one that contains vertex @xmath117 . now @xmath218 is a @xmath5-matching of @xmath215 thus @xmath219 hence in all cases we have @xmath220 as required .
this completes the proof of the theorem .
we note that all network bargaining games studied in @xcite satisfy conditions @xmath221 and @xmath205 of theorem [ thm : 1 ] .
in addition to these , theorem [ thm : 1 ] also covers the case of network bargaining games where the underlying graph is a tree , but the vertices are allowed to have arbitrary capacities . hence starting with a maximum weight @xmath5-matching @xmath4 we can use the polynomial time algorithm of @xcite to compute a point in the intersection of the core and prekernel for these games , from which we can obtain a corresponding solution @xmath3 . then using theorem [ thm : 1 ] we know that @xmath3 will be balanced .
while we were able to generalize the class of network bargaining games for which balanced solutions can be obtained by computing a point in the intersection of the core and prekernel , we were not able to apply this technique to all network bargaining games . in this section we show that balanced solutions can be obtained to any network bargaining game @xmath14 by a reduction to a unit capacity game defined on an auxiliary graph .
+ suppose we are given an instance @xmath26 of the network bargaining game together with a @xmath5-matching @xmath4 of @xmath6 .
we describe below how to obtain an instance @xmath222 of the unit capacity game together with a matching @xmath4 of @xmath6 . 1 . for each @xmath32 :
fix an arbitrary labelling @xmath224 and create @xmath225 copies @xmath226 in @xmath227 .
2 . for each @xmath8 in @xmath228 : add the edge @xmath229 to @xmath230 and set its weight to @xmath231 .
3 . for each edge @xmath232 : add all edges @xmath233 to @xmath234 for all @xmath235 $ ] and @xmath236 $ ] , and set all their weights to @xmath231 .
+ * example 1 : * consider the instance depicted on the left hand side of the figure below .
the solid edges are in the @xmath5-matching @xmath4 and the dotted edges are in @xmath105 .
node @xmath10 has capacity four , nodes @xmath15 and @xmath171 have capacity two and all other nodes have capacity one .
all edges have unit weight .
we make four copies of @xmath10 in @xmath237 , two copies of @xmath15 and @xmath171 , and one copy of every other node .
each edge in @xmath4 corresponds to a unique edge in @xmath136 . for the edges @xmath8 and @xmath238 which are not in @xmath4
, we connect every copy of @xmath10 to every copy of @xmath11 and @xmath183 with edges in @xmath239 .
the resulting graph is on the right .
suppose we are given an instance of the network bargaining game @xmath14 with a @xmath5-matching @xmath4 .
let @xmath240 $ ] be obtained using the construction given in section [ const ] .
note that @xmath4 and @xmath136 have the same number of edges and each edge @xmath39 is mapped to the unique edge @xmath241 where @xmath242 and @xmath243 .
this allows us to go back and forth between solutions on @xmath4 and @xmath136 by dividing the weight of each edge in the same way as its corresponding pair .
note that @xmath251 if and only if @xmath252 .
the following lemma shows that the mapping given by the function @xmath253 and its inverse @xmath254 defines a bijection between the @xmath255 and @xmath256 let @xmath257 and @xmath258 .
we show that @xmath259 .
take @xmath261 .
suppose @xmath262 and @xmath243 .
then from the construction of @xmath237 and @xmath136 it follows that @xmath241 .
we have @xmath263 furthermore if @xmath232 then from the definition of @xmath264 we have @xmath265 .
now let @xmath259 and @xmath260 .
we show that @xmath257 .
take @xmath266 . from the construction of @xmath237 and @xmath136 there must exist an edge @xmath261 such that @xmath262 and @xmath243 .
we have @xmath267 furthermore if @xmath268 is uncovered in @xmath136 then @xmath269 by definition .
[ lem : key ] let @xmath14 be an instance of the network bargaining game and @xmath4 a @xmath5-matching on @xmath6 .
suppose the auxiliary instance @xmath222 and the matching @xmath136 were obtained using the construction given in section [ const ] .
let @xmath3 be a solution to @xmath14 and @xmath271 a solution to @xmath222 such that @xmath270 .
then for any @xmath32 and any @xmath272 $ ] we have @xmath273 we first show that @xmath274 .
we may assume that @xmath275 since otherwise there is nothing to show .
let @xmath11 be vertex @xmath10 s best outside option in @xmath3 .
that is @xmath232 and @xmath276 } \min_{vw \in m } z_{vw } \notag .\end{aligned}\ ] ] 1 .
@xmath11 is not saturated in @xmath4 .
then the vertex @xmath278 is in @xmath227 and it is not covered by @xmath136 .
since @xmath279 we have @xmath280 2 .
@xmath11 is saturated in @xmath4 .
let @xmath281 .
suppose that @xmath282 . then @xmath283 is covered in @xmath136 and @xmath284 .
since @xmath285 we have @xmath286 we now show that @xmath287 .
we may assume that @xmath288 .
let @xmath283 be vertex @xmath268 s best outside option in @xmath289 .
that is , @xmath290 such that @xmath291 and @xmath292 since @xmath291 we must have @xmath232 .
again , we have two cases : 1 . @xmath283 is not covered in @xmath136 .
then the vertex @xmath11 is not saturated in @xmath4 and @xmath293 2 .
@xmath283 is covered in @xmath4 .
then there exists @xmath294 such that @xmath295 and @xmath296 .
we have @xmath297 using lemma [ lem : key ] we can now prove that stability and balance are preserved when mapping between solutions of the network bargaining game and the corresponding unit capacity game of the auxiliary instance .
[ thm : main ] let @xmath14 be an instance of the network bargaining game and @xmath4 a @xmath5-matching on @xmath6 .
suppose the auxiliary instance @xmath222 and the matching @xmath136 were obtained using the construction given in section [ const ] .
let @xmath3 be a solution to @xmath14 and @xmath271 a solution to @xmath222 such that @xmath270 . then
: let @xmath39 .
suppose that @xmath262 . then @xmath298 and using lemma [ lem : key ] we have @xmath299 it remains to show that
if @xmath289 is stable then @xmath47 for any unsaturated vertices @xmath10 of @xmath6 .
suppose @xmath10 is such a vertex
. then the vertex @xmath300 is not covered in @xmath136 and therefore @xmath301 . if @xmath289 is stable then @xmath302 and by lemma [ lem : key ] we have @xmath47 as desired .
this completes the proof of the first statement . to prove the second statement
let @xmath39 and suppose that @xmath303 and @xmath304 .
then @xmath305 , @xmath306 and by lemma [ lem : key ] we have : @xmath307 this completes the proof . 1 .
find a maximum @xmath5-matching @xmath4 in @xmath6 .
2 . obtain unit capacity game @xmath222 with matching @xmath136 using the construction from section [ const ] .
3 . find a balanced solution @xmath15 on the matching @xmath136 in @xmath237 .
4 . set @xmath308 and return @xmath3 .
we note that step 3 of the algorithm can be implementing using the existing polynomial time algorithm of kleinberg and tardos @xcite .
given any instance of a network bargaining game with unit capacities together with a maximum weight matching , their algorithm returns a balanced solution on the given matching , whenever one exists . using theorem [ thm : main ] we know that stable solutions of the original problem map to stable solutions of the matching problem and viceversa . since any stable solution must occur on a @xmath5-matching , respectively matching , of maximum weight we have the following corollary
let @xmath14 be an instance of the network bargaining game and @xmath4 a @xmath5-matching on @xmath6 .
suppose the auxiliary instance @xmath222 and the matching @xmath136 were obtained using the given construction
. then 1 .
@xmath4 is a maximum weight @xmath5-matching for @xmath14 if and only if @xmath136 is a maximum weight matching for @xmath222 .
2 . there exists a balanced solution for @xmath14 on the @xmath5-matching @xmath4 if and only if there exists a balanced solution for @xmath222 on the matching @xmath136 .
it was previously shown in @xcite that a unit capacity game possesses a balanced solution if and only if it has a stable solution , which in turn happens if and only if the linear program for the maximum weight matching of the underlying graph has an integral optimal solution .
for the case of network bargaining game with general capacities , @xcite have shown that a stable solution exists if and only if the linear program for the maximum weight @xmath5-matching of the underlying graph has an integral optimal solution . in terms of existence of balanced solutions ,
they only obtain a partial characterization by proving that if this integral optimum is unique , then a balanced solution is guaranteed to exist .
our results imply the following full characterization for the existence of balanced solutions , thus extending the results of @xcite : , devanur , n. , jain , k. , and rabani , y. 2010 .
monotonicity in bargaining networks . in _ proceedings of the twenty - first annual acm - siam symposium on discrete algorithms_. society for industrial and applied mathematics , 817826 . | we study _ balanced solutions _ for _ network bargaining games _ with general capacities , where agents can participate in a fixed but arbitrary number of contracts .
we provide the first polynomial time algorithm for computing balanced solutions for these games .
in addition , we prove that an instance has a balanced solution if and only if it has a stable one .
our methods use a new idea of reducing an instance with general capacities to a network bargaining game with unit capacities defined on an auxiliary graph .
this represents a departure from previous approaches , which rely on computing an allocation in the intersection of the _ core _ and _ prekernel _ of a corresponding _ cooperative game _ , and then proving that the solution corresponding to this allocation is balanced .
in fact , we show that such cooperative game methods do not extend to general capacity games , since contrary to the case of unit capacities , there exist allocations in the intersection of the core and prekernel with no corresponding balanced solution .
finally , we identify two sufficient conditions under which the set of balanced solutions corresponds to the intersection of the _ core _ and _ prekernel _ , thereby extending the class of games for which this result was previously known . |
in this paper , all graphs are simple , finite and undirected .
we refer to the book @xcite for undefined notation and terminology in graph theory . throughout this paper ,
let @xmath3 and @xmath4 denote the order ( number of vertices ) and size ( number of edges ) of a graph , respectively .
moreover , a vertex of a connected graph is called a _ leaf _ if its degree is one ; otherwise , it is an _
internal vertex_. let @xmath13 and @xmath14 denote the number of leaves and the number of internal vertices of a tree @xmath15 , respectively , and let @xmath16 @xmath15 is a spanning tree of @xmath0 @xmath17 and @xmath18 @xmath15 is a spanning tree of @xmath0 @xmath17 for a connected graph @xmath0 .
note that the sum of @xmath19 and @xmath20 is @xmath3 for any connected graph @xmath0 of order @xmath3 .
a path in an edge - colored graph is a _ monochromatic path _ if all the edges on the path have the same color .
an edge - coloring of a connected graph is a _ monochromatically - connecting coloring _ ( _ mc - coloring _ , for short ) if any two vertices of the graph are connected by a monochromatic path of the graph .
for a connected graph @xmath0 , the _ monochromatic connection number _ of @xmath0 , denoted by @xmath21 , is defined as the maximum number of colors used in an mc - coloring of @xmath0 .
extremal mc - coloring _ is an mc - coloring that uses @xmath21 colors .
note that @xmath22 if and only if @xmath0 is a complete graph .
the concept of @xmath21 was first introduced by caro and yuster @xcite and has been well - studied recently .
we refer the reader to @xcite for more details . in @xcite
, the authors introduced the concept of total monochromatic connection of graphs .
a graph is said to be _ total - colored _ if all the edges and the vertices of the graph are colored . a path in a total - colored graph is a _ total monochromatic path _ if all the edges and internal vertices on the path have the same color .
a total - coloring of a graph is a _ total monochromatically - connecting coloring _ ( _ tmc - coloring _ , for short ) if any two vertices of the graph are connected by a total monochromatic path of the graph . for a connected graph @xmath0 , the _ total monochromatic connection number _ , denoted by @xmath1 , is defined as the maximum number of colors used in a tmc - coloring of @xmath0 .
note that a tmc - coloring does not exist if @xmath0 is not connected , in which case we simply let @xmath2 .
extremal tmc - coloring _ is a tmc - coloring that uses @xmath1 colors .
it is easy to check that @xmath23 if and only if @xmath0 is a complete graph .
actually , these concepts are not only inspired by the concept of monochromatic connection number but also by the concepts of monochromatic vertex connection number and total rainbow connection number of a connected graph . for details about them
we refer to @xcite . from the definition of the total monochromatic connection number ,
the following results follow immediately .
[ pro1]@xcite if @xmath0 is a connected graph and @xmath24 is a connected spanning subgraph of @xmath0 , then @xmath25 .
[ thm1]@xcite for a connected graph @xmath0 , @xmath26 . in particular , @xmath27 if @xmath0 is a tree .
the authors @xcite also showed that there are dense graphs that still meet this lower bound .
[ thm2]@xcite let @xmath0 be a connected graph of order @xmath28 .
if @xmath0 satisfies any of the following properties , then @xmath27 .
@xmath29 the complement @xmath30 of @xmath0 is @xmath31-connected .
@xmath32 @xmath0 is @xmath33-free .
@xmath34 @xmath35 .
@xmath36 @xmath37 .
@xmath38 @xmath0 has a cut vertex . moreover , the authors @xcite gave an example to show that the lower bound @xmath39 is not always attained .
[ example1]@xcite let @xmath40 be a complete multipartite graph with @xmath41 and @xmath42 .
then @xmath43 .
let @xmath0 be a connected graph and @xmath44 be an extremal tmc - coloring of @xmath0 that uses a given color @xmath45 .
note that the subgraph @xmath24 formed by the edges and vertices with color @xmath45 is a tree where the color of each internal vertex is @xmath45 @xcite .
now we define the _ color tree _ as the tree formed by the edges and vertices with color @xmath45 , denoted by @xmath46 .
if @xmath46 has at least two edges , the color @xmath45 is called _ nontrivial _ ; otherwise , @xmath45 is _
trivial_. we call an extremal tmc - coloring _ simple _ if for any two nontrivial colors @xmath45 and @xmath47 , the corresponding trees @xmath46 and @xmath48 intersect in at most one vertex .
if @xmath44 is simple , then the leaves of @xmath46 must have distinct colors different from color @xmath45 .
moreover , a nontrivial color tree of @xmath44 with @xmath49 edges and @xmath50 internal vertices is said to _ waste _ @xmath51 colors .
for the rest of this paper we will use these facts without further mentioning them .
in addition , we list a helpful lemma below .
[ lem1]@xcite every connected graph @xmath0 has a simple extremal tmc - coloring .
this paper is organized as follows . in section @xmath52 , we characterize all graphs @xmath0 with @xmath53 , respectively . in section @xmath54 , we show that for any function @xmath8 satisfying @xmath9 , if @xmath55 , where @xmath56 , then @xmath57 is a sharp threshold function for the property @xmath58 ; if @xmath59 , then @xmath60 is a sharp threshold function for the property @xmath58 . in section @xmath31 , we prove that for a given connected graph @xmath0 , and a positive integer @xmath10 with @xmath11 , it is np - complete to decide whether @xmath12 .
in this section , we characterize all graphs @xmath0 with @xmath53 , respectively . we call a connected graph @xmath0 _ unicyclic _ , _ bicyclic _ , or _ tricyclic _ if @xmath62 or @xmath63 , respectively
. let @xmath64 denote the set of the trees with @xmath65 , where @xmath66 .
note that if @xmath0 is a connected graph with @xmath67 , then @xmath0 is either a path or a cycle .
[ thm3 ] let @xmath0 be a connected graph .
then @xmath68 if and only if @xmath0 is a path .
if @xmath0 is a path , then @xmath69 . hence it remains to verify the converse .
let @xmath0 be a connected graph with @xmath68 . by theorem [ thm1 ]
, we get that @xmath70 and then @xmath71 as @xmath72 .
since @xmath0 is a connected graph , it follows that @xmath73 and @xmath67 .
thus @xmath0 is a path .
[ thm4 ] let @xmath0 be a connected graph
. then @xmath74 if and only if @xmath75 or @xmath0 is a cycle except for @xmath33 .
if @xmath75 or @xmath0 is a cycle except for @xmath33 , then @xmath74 by theorem [ thm2]@xmath32 .
conversely , let @xmath0 be a connected graph with @xmath74 .
first , we have @xmath76 by theorem [ thm1 ] .
since @xmath72 and @xmath77 , it follows that @xmath67 or @xmath54 .
if @xmath78 , then @xmath73 and so @xmath75 . otherwise , from theorem [ thm3 ] we have that @xmath0 is a cycle and @xmath79 since @xmath80 . [ thm5 ] let @xmath0 be a connected graph
. then @xmath81 if and only if @xmath82 or @xmath83 , where @xmath84 ; see figure [ fig.1 . ] .
[ 0.8].,title="fig : " ] if @xmath82 or @xmath83 , where @xmath84 , then @xmath0 has a cut vertex and so @xmath81 by theorem [ thm2]@xmath38 .
hence it remains to verify the converse .
let @xmath0 be a connected graph with @xmath81 .
first , we have @xmath85 by theorem [ thm1 ] .
since @xmath72 and @xmath77 , it follows that @xmath86 or @xmath31 . if @xmath87 , then @xmath73 and so @xmath82 . if @xmath78 , then we have @xmath88 from theorem [ thm4 ] and so @xmath0 is a unicyclic graph with @xmath78 ; see figure [ fig.1 . ] .
if @xmath67 , then we have @xmath89 from theorems [ thm3 ] and [ thm4 ] .
however , @xmath80 , a contradiction . [ thm6 ]
let @xmath0 be a connected graph .
then @xmath90 if and only if @xmath89 , @xmath91 or @xmath92 , where @xmath93 ; see figure [ fig.2 . ] . [ 0.8 ] it is easy to verify the sufficiency by theorem [ thm2 ] .
next we just need to prove the necessity .
let @xmath0 be a connected graph with @xmath90 .
first , we have @xmath94 by theorem [ thm1 ] .
since @xmath72 and @xmath77 , it follows that @xmath95 or @xmath96 . if @xmath97 , then @xmath73 and so @xmath91 . if @xmath87 , we have that @xmath88 from theorem [ thm5 ] and so @xmath0 is a unicyclic graph with @xmath87 ; see @xmath98 in figure [ fig.2 . ] .
similarly , from theorems [ thm4 ] and [ thm5 ] , we have that @xmath99 if @xmath78 and then @xmath0 is a bicyclic graph with @xmath78 except for @xmath100 since @xmath101 ; see @xmath102 in figure [ fig.2 . ] .
if @xmath67 , we have that @xmath89 from theorems [ thm3 ] and [ thm4 ] .
recall that @xmath23 if and only if @xmath103 .
in fact , there does not exist a graph such that @xmath104 .
we are given a connected graph @xmath0 with @xmath105 and a simple extremal tmc - coloring of @xmath0 .
then there must be two nonadjacent vertices in a nontrivial color tree .
since every nontrivial color tree wastes at least 2 colors , we get that @xmath106 . hence in the following , we characterize all graphs @xmath0 having @xmath107 .
let @xmath108 be @xmath109 nonadjacent edges of @xmath110 , where @xmath111 . given a graph @xmath24 ,
let @xmath112 denote the graph obtained from @xmath110 by deleting the edges of @xmath24 .
[ thm7 ] let @xmath0 be a connected graph .
then @xmath113 if and only if @xmath114 . clearly , @xmath115
. then @xmath116 .
conversely , let @xmath0 be a connected graph with @xmath113 .
we are given a simple extremal tmc - coloring @xmath44 of @xmath0 .
suppose that @xmath44 consists of @xmath117 nontrivial color trees , denoted by @xmath118 . since each nontrivial color tree wastes at least two colors , it follows that @xmath119 and @xmath120 .
thus , @xmath114 .
[ thm8 ] let @xmath0 be a connected graph .
then @xmath121 if and only if @xmath0 is either @xmath122 or @xmath123 .
note that @xmath124 and then @xmath125 .
note that @xmath122 is a spanning subgraph of @xmath123
. then @xmath126 by proposition [ pro1 ] .
now we just need to prove that @xmath127 .
let @xmath44 be a simple extremal tmc - coloring of @xmath123 .
suppose that @xmath44 consists of @xmath117 nontrivial color trees .
since there are two pairs of nonadjacent vertices in two nontrivial color trees or in a common nontrivial color tree , it wastes at least three colors and then @xmath127 .
hence @xmath128 .
now it remains to verify the converse .
let @xmath0 be a connected graph with @xmath121 .
we are given a simple extremal tmc - coloring @xmath44 of @xmath0 .
suppose that @xmath44 consists of @xmath117 nontrivial color trees , denoted by @xmath118 .
since each nontrivial color tree wastes at least two colors , we get that @xmath119 and @xmath129 .
thus , @xmath122 is a spanning subgraph of @xmath0 . from theorem [ thm7 ]
, it can be checked that @xmath0 is either @xmath122 or @xmath123 .
[ thm9 ] let @xmath0 be a connected graph .
then @xmath130 if and only if @xmath131 . clearly , @xmath132 and @xmath133 .
thus we have that @xmath134 .
if @xmath135 , there are three pairs of nonadjacent vertices and let @xmath44 be a simple extremal tmc - coloring of @xmath0 .
suppose that @xmath44 consists of @xmath117 nontrivial color trees .
then it wastes at least 4 colors and so @xmath136 . since @xmath137 is a spanning subgraph of @xmath0 , @xmath138 by proposition [ pro1 ] .
thus we get that @xmath139 .
similarly , it can be verified that @xmath140 .
conversely , let @xmath0 be a connected graph with @xmath130 .
we are given a simple extremal tmc - coloring @xmath44 of @xmath0 .
suppose that @xmath44 consists of @xmath117 nontrivial color trees , denoted by @xmath118 .
since each nontrivial color tree wastes at least two colors , we get the following two cases . * case 1 .
* @xmath119 .
then @xmath141 or @xmath142 .
if @xmath141 , then @xmath143 is a spanning subgraph of @xmath0 . from theorems [ thm7 ] and [ thm8 ]
, we obtain that @xmath0 is either @xmath143 or @xmath144 .
if @xmath145 , then @xmath137 is a spanning subgraph of @xmath0 . from theorems [ thm7 ] and [ thm8 ]
, we get that @xmath131 .
* case 2 . * @xmath146 .
then @xmath147 . from theorem
[ thm8 ] , @xmath148 and @xmath149 have not a common leaf .
thus @xmath144 is a spanning subgraph of @xmath0 .
since @xmath116 and @xmath150 , we have that @xmath151 .
let @xmath152 denote the random graph with @xmath3 vertices and edge probability @xmath153 @xcite . for a graph property @xmath154 and for a function @xmath155
, we say that @xmath156 satisfies @xmath154 _ almost surely _ if the probability that @xmath157 satisfies @xmath154 tends to @xmath158 as @xmath3 tends to infinity .
we say that a function @xmath8 is a _ sharp threshold function _ for the property @xmath154 if there are two positive constants @xmath159 and @xmath45 so that @xmath156 satisfies @xmath154 almost surely for all @xmath160 and @xmath161 almost surely does not satisfy @xmath154 .
let @xmath0 and @xmath24 be two graphs on @xmath3 vertices .
a property @xmath154 is said to be _ monotone _
if whenever @xmath162 and @xmath0 satisfies @xmath154 , then @xmath24 also satisfies @xmath154 .
it is well - known that all monotone graph properties have sharp threshold functions ; see @xcite and @xcite . for any graph @xmath0 with @xmath3 vertices and any function @xmath8 ,
having @xmath7 is a monotone graph property ( adding edges does not destroy this property ) , so it has a sharp threshold function . in the following ,
we establish a sharp threshold function for the graph property @xmath7 .
[ thm10 ] let @xmath8 be a function satisfying @xmath9 .
then @xmath163 is a sharp threshold function for the property @xmath58 .
[ rem1 ] note that if @xmath164 , then @xmath156 is a complete graph @xmath110 and @xmath165 .
hence we only concentrate on the case @xmath166 . before proving theorem [ thm10 ]
, we need some lemmas .
[ lem2]@xcite let @xmath167
. then @xmath168\rightarrow \begin{cases } e^{e^{-a } } & if\ |a|=o(1 ) , \cr 0 & a\rightarrow -\infty , \cr 1 & a\rightarrow + \infty . \end{cases}\end{aligned}\ ] ] [ lem3 ] @xcite * ( chernoff bound ) * if @xmath169 is binomial variable with expectation @xmath170 , and @xmath171 , then @xmath172\leq exp(-\frac{\delta^2\mu}{2})\ ] ] and @xmath173\leq exp(-\frac{\delta^2\mu}{2+\delta}).\ ] ] [ lem4 ] let @xmath0 be a noncomplete connected graph of order @xmath3 with minimum degree @xmath174
. then @xmath175 . for a noncomplete graph @xmath0
, we have that @xmath176 whose proof is contained in the proof of theorem 6 in @xcite .
moreover , @xmath177 by proposition 12 in @xcite .
thus @xmath175 .
* proof of theorem 10 : * we divide our proof into two cases according to the range of @xmath8 .
* @xmath55 , where @xmath178 .
we first prove that there exists a constant @xmath159 such that the random graph @xmath179 with @xmath57 almost surely has @xmath180 .
let @xmath181 it is easy to check that @xmath179 is almost surely connected by lemma [ lem2 ] .
let @xmath182 denote the expectation of the number of edges in @xmath179 .
then @xmath183 moreover from lemma [ lem3 ] , it follows that @xmath184\leq\exp(-\frac{1}{8}\mu_1)=o(1)$ ] .
suppose that @xmath185 . by theorem [ thm1 ]
we have that for @xmath3 sufficiently large , @xmath186 @xmath187 @xmath188 @xmath189 thus , we conclude that @xmath180 holds with the probability at least @xmath190 .
next we show that there exists a constant @xmath45 such that the random graph @xmath191 with @xmath57 almost surely has @xmath192 .
let @xmath193 and @xmath194 denote the expectation of the number of edges in @xmath191 .
then we have @xmath195 furthermore by lemma [ lem3 ] , it follows that @xmath196\leq\exp(-\frac{1}{10}\mu_2)=o(1)$ ] .
if @xmath191 is not connected , then @xmath197 .
otherwise , let @xmath174 be the minimum degree of @xmath191 .
suppose that @xmath198 . from lemma [ lem4 ]
, we have that for @xmath3 sufficiently large , @xmath199 @xmath200 @xmath201 @xmath202 hence , we conclude that @xmath203 holds with the probability at least @xmath204 . * case 2 .
* @xmath59 .
let @xmath205 and @xmath206 .
by lemma [ lem2 ] , we have that @xmath179 is almost surely connected and @xmath191 is almost surely not connected .
it can be checked that @xmath180 almost surely holds in a similar way as case 1 . on the other hand , since @xmath191 is almost surely not connected , @xmath197 almost surely holds .
given a graph @xmath0 , a set @xmath207 is called a _ dominating set _ of @xmath0 if every vertex of @xmath0 not in @xmath208 has a neighbor in @xmath208 . if the subgraph induced by @xmath208
is connected , then @xmath208 is called a _ connected dominating set_. the _ connected dominating number _ , denoted by @xmath209 , is the minimum cardinalities of the connected dominating sets of @xmath0 .
note that the sum of @xmath209 and @xmath19 is @xmath3 because a vertex subset is a connected dominating set if and only if its complement is contained in the set of leaves of a spanning tree . in this section , we mainly prove the following result .
[ lem5]@xcite the first problem defined below is polynomially reducible to the second one : + problem 1 . given a graph @xmath0 and a positive integer @xmath210 , decide whether there is a dominating set of size @xmath117 or less .
+ problem 2 . given a connected graph @xmath0 with a cut vertex and a positive integer @xmath117 with @xmath210 ,
decide whether there is a connected dominating set of size @xmath117 or less .
* proof of theorem 11 : * given a connected graph @xmath0 with a cut vertex , and a positive integer @xmath210 .
note that @xmath211 if and only if @xmath212 by theorem [ thm2]@xmath38 .
then problem 2 can be polynomially reducible to problem 3 : given a connected graph @xmath0 with a cut vertex and a positive integer @xmath10 with @xmath11 , decide whether @xmath12 .
thus , problem 1 can be reducible to problem 3 by lemma [ lem5 ] .
moreover , problem 1 is known as a np - complete problem in @xcite .
hence the problem in theorem [ thm11 ] is np - hard .
next we prove that given a connected graph @xmath0 and a nonnegative integer @xmath213 , to decide whether @xmath214 is np .
recall that a problem belongs to np - class if given any instance of the problem whose answer is yes " , there is a certificate validating this fact which can be checked in polynomial time . for any fixed integer @xmath215 , to prove the problem of deciding whether @xmath214 is np , we choose a tmc - coloring of @xmath0 with @xmath216 colors as a certificate . for checking a tmc - coloring with @xmath216 colors ,
we only need to check that @xmath216 colors are used and for any two vertices @xmath217 and @xmath218 of @xmath0 , there exists a total monochromatic path between them .
notice that for any two vertices @xmath217 and @xmath218 of @xmath0 , there are at most @xmath219 paths of length @xmath220 , since if we let @xmath221 , then there are less than @xmath3 choices for each @xmath222 . clearly , the path @xmath154 wastes at least @xmath223 colors . then @xmath224 and so @xmath225 which implies that @xmath226 .
therefore , @xmath0 contains at most @xmath227 @xmath217-@xmath218 paths of length at most @xmath228 .
then , check these paths in turn until one finds a path whose edges and internal vertices have the same color .
it follows that the time used for checking is at most @xmath229 .
since @xmath215 is a fixed integer , we conclude that the certificate can be checked in polynomial time .
then the problem of deciding whether @xmath214 belongs to np - class and so is the problem in theorem [ thm11 ] . | a graph is said to be _ total - colored _ if all the edges and the vertices of the graph are colored .
a total - coloring of a graph is a _ total monochromatically - connecting coloring _ ( _ tmc - coloring _ , for short ) if any two vertices of the graph are connected by a path whose edges and internal vertices on the path have the same color . for a connected graph @xmath0 , the _ total monochromatic connection number _ , denoted by @xmath1 , is defined as the maximum number of colors used in a tmc - coloring of @xmath0 .
note that a tmc - coloring does not exist if @xmath0 is not connected , in which case we simply let @xmath2 . in this paper , we first characterize all graphs of order @xmath3 and size @xmath4 with @xmath5 and @xmath6 , respectively
. then we determine the threshold function for a random graph to have @xmath7 , where @xmath8 is a function satisfying @xmath9 .
finally , we show that for a given connected graph @xmath0 , and a positive integer @xmath10 with @xmath11 , it is np - complete to decide whether @xmath12 .
* keywords * : total - colored graph , total monochromatic connection , random graphs , np - complete * ams subject classification 2010 * : 05c15 , 05c40 , 05c75 , 05c80 , 68q17 . |
the magnetic , electronic and transport behaviour of the @xmath0 high @xmath7 superconductor have been intensely studied , and one avenue of research that might lead to an understanding of the anomalous normal state properties , as well as possibly the key to the superconducting instability , involves the examination of the evolution of the antiferromagnetic ( afm ) insulator ( @xmath8 ) to a spin glass ( sg ) phase ( @xmath9 ) , and ultimately to a superconductor with anomalous normal state properties ( @xmath10 ) .
each of these doping ranges involves new and exciting physics with many mysteries to resolve @xcite , and in this paper we focus on the microscopic details of the spin texture of the sg phase .
early experimental work demonstrated that the @xmath11 materials had a low temperature sg phase , as evidenced in the freezing of the magnetic moment seen in @xmath12sr studies @xcite .
shortly afterward it was noted that quasielastic neutron scattering results for this phase @xcite could be interpreted in terms of the freezing of microdomains roughly of a size of 20 .
subsequent nqr experiments supported this conjecture by showing how the la relaxations rates could be understood in terms of afm domains of small , finite size magnetic domains @xcite .
these interpretations made it quite natural to refer to this phase as a cluster sg .
more recent experiments have provided a much more complete characterization of these compounds .
most importantly , the measurements show evidence of canonical sg behaviour @xcite .
further , samples in this doping range are found to have an anomalously small curie constant as one approaches the sg transition from above @xcite .
also , studies of the anomalous transport properties @xcite , _
e.g. _ , a conductivity that rises logarithmically with temperature over a small temperature range , and an isotropic , negative magnetoresistance @xcite , serve to emphasize the unusual behaviour of lasrcuo in this doping regime . in this paper
we discuss how the spin texture of the sg phase at a microscopic level ( i ) necessitates the identification of this phase as a cluster sg , and ( ii ) explains how the small finite size domains characterizing this spin texture are consistent with the anomalously small curie constant found experimentally . to do this
, we must begin with a model for the effect of the dopant sr impurities and the holes they induce into the cuo@xmath13 planes .
however , if a model is to be successful in describing the physics of this compound , it must be able to describe the physics in _ all _ doping ranges , not simply the sg regime .
we employ such a candidate model it was introduced previously @xcite and is based on the strong coupling of the hole s motion to the background antiferromagnetic cu spin correlations .
to be specific , at low temperatures , when the carriers are localized in the region of the sr impurities , they generate spin distortions that one can model using a spin only hamiltonian .
our model is only strictly justified at low temperatures where the carriers are localized near the sr impurities however , as has been argued elsewhere @xcite , and as we elaborate on below , we believe that the spin distortions produced by the holes that gradually become mobile as the temperature is raised will resemble those generated by the localized carriers that we focus on in this paper , and thus the qualitative aspects of the arguments presented here should be applicable at higher temperatures .
this model accounts for the crucial and largely ignored role ( in previous literature ) of dopant
_ disorder_. to be specific , when comparing its predictive powers to the physics of the weakly
doped afm phase this model has been shown to be capable of describing ( i ) the traverse spin freezing temperature _ vs. _ doping @xcite observed by chou _
@xcite , and ( ii ) provides a better fit to the conductivity _ vs. _
@xmath14 data than 3d variable
range hopping , as well as correctly predicting the critical concentration @xmath15 at which the metal non metal transition takes place @xcite .
further , when compared to the behaviour of the intermediate doping regime for the sg phase , preliminary work @xcite showed that this model is able to track the behaviour of the doping and temperature dependent spin correlation length , as measured by keimer _
_ @xcite and characterized by the empirical relationship @xmath16 in this paper we continue the comparison of our candidate model by studying its spin texture _ vs. _ doping in the sg regime .
we focus on understanding the spin texture at a microscopic level and find that disorder induces meandering domain walls connecting sr sites ; we argue that this is the correct way to think of the underlying spin texture of a cluster sg .
a state qualitatively similar to this but one produced by electronic phase separation in the absence of sr impurity disorder was previously proposed by emery and kivelson @xcite . in our model
the possibility of electronic phase separation does not exist , and so electronic phase separation is not necessarily the progenitor of the cluster spin sg . instead , we believe that the distortions generated by the holes and the sr disorder are the key aspects of this problem .
a summary of our new theoretical and experimental results for this spin texture are as follows : i clusters of antiferromagnetically correlated regions are formed , the boundaries of which are defined by the sr impurities .
these domains have afm order parameters which vary in direction , and thus long ranged afm order in this doping regime is destroyed by the lack of orientational order of the moments of these clusters .
the smallest collection of impurities which forms a small cluster of afm correlated spins is three non collinear sr impurities , and from such three impurity configurations it is established that domain walls separating the clusters are regions of disordered spins which have a larger local ferromagnetic ( fm ) correlation than anywhere else in the lattice .
also , since carriers are much more mobile in fm regions , it is reasonable to assume that this spin texture leads to meandering rivers of charge connecting the impurity sites namely , at the higher temperatures at which the holes move the domain walls constrain and are populated with the mobile holes .
this spin texture thus explains why the @xmath17 material does not develop incommensurate magnetic fluctuations : since the holes are expelled " from the afm clusters by their preference for moving in a fm region , there is no impetus for the local magnetic order to peak at any wave vector other than @xmath18 .
we demonstrate this for the low temperature spin texture by the explicit evaluation of @xmath19 .
iia the susceptibility experiments of chou _ et al .
_ on the sg phase of @xmath20 between 20 k and 100 k displayed an anomalously small curie constant @xcite .
we have grown a new set of samples using the travelling
solvent , float zone method , for a variety of concentrations in the sg regime .
we have determined the curie constants for all of these samples , and have determined how this constant changes with doping concentration in the sg phase .
iib modelling the topology of the sg phase in terms of small domains of afm correlated spins separated by disordered domain walls , and then treating the clusters as independent low
spin magnetic units , we show that such a small moment follows immediately , and that it increases with doping by one over the correlation length squared .
this result is in quantitative agreement with experiment , and thus our comprehensive experimental study of this quantity serves as a critical test of the cluster sg phenomenology .
iii as mentioned previously , this model accounts for some observations in the afm phase , and in this paper we show that it also describes some results found in the sg phase .
if this model is an accurate representation of the spin texture in all doping regimes it should also be able to explain the appearance of incommensurate correlations for @xmath21 @xcite .
indeed , we find that this is the case , providing very strong support for this model .
we associate this phenomenon with the degradation of the integrity of the domains of the cluster sg as doping levels are increased , and relate the delay in the appearance of the incommensurate correlations to impurity disorder effects .
we propose that our results are suggestive that the mechanism behind the incommensurate phase follows from the spiral phase theory of shraiman and siggia @xcite .
our paper is organized as follows .
firstly , in ii we discuss our model of the spin distortions introduced by a single sr impurity and a single hole , and explain how this may be used to study the effects of a nonzero density of holes all localized around sr impurities .
then we present an analysis of a small number of sr impurities ( three , in fact ) that leads to the simplest understanding of how domains of afm correlated spins are produced , and survey results from our extensive numerical simulations that produce the ground states for our model hamiltonian . in
iii we present new experimental results for the susceptibilities of this compound just above the sg transition temperature , and then show how the above mentioned small afm domains adequately describe the observed curie constants . in
iv we present a discussion of the commensurate to incommensurate transition , and relate the effects of spin distortions that we have included explicitly to theoretical mechanisms for the spiral phase .
lastly , we summarize our results and relate them to other outstanding problems in the study of weakly doped lasrcuo . a brief , preliminary report on this work
has already appeared elsewhere @xcite .
it is well known that the near neighbour heisenberg hamiltonian provides a very accurate representation of the magnetic properties of the 2d cuo@xmath13 planes @xcite apart from small anisotropies which we ignore in the present paper ( _ e.g. _ , these anisotropies only affect the correlation length when it becomes of the order of 80 lattice constants , a distance that greatly exceeds any length scale discussed in this paper ) . thus , we represent the magnetic interactions between the cu spins using @xmath22 where @xmath23 denote neighbouring cu sites , and @xmath24 is the exchange constant , known to be of the order of 1550 k. at low doping levels and at low temperatures , transport measurements @xcite suggest that the carriers move by hopping .
presumably their motion corresponds to the holes hopping from one impurity site to another , such as occurs in doped semiconductors before the metal
nonmetal transition .
then , it is appropriate to consider the response of the magnetic background to carriers localized in the vicinity of the sr impurity ions . a variety of previous theoretical studies @xcite
have all shown that the ground state for a single hole in such a situation corresponds to a doubly degenerate ground state for which the hole circulates either clockwise or counterclockwise in the plaquette directly above or directly below the sr impurity .
the hole motion couples to the transverse fluctuations of the cu spins and produces a spiraling twisting of the afm order parameter @xcite , a state that is topologically similar to the singly charged skyrmion excitations of the 2d classical nonlinear sigma model @xcite .
one of us and mailhot @xcite proposed that a simple way of representing of the effects of this circulating hole motion on the afm background was to integrate out the hole motion and replace it by a purely magnetic interaction . for ( 1,2,3,4 ) denoting the four spins in the plaquette bordering the sr impurity , the interaction hamiltonian for a single hole is given by @xmath25~~. \label{eq : ham int}\end{aligned}\ ] ] one may show that for classical spins , as long as @xmath26 the ground state of @xmath27 has the same topology as that of the circulating hole ground state .
the factor of @xmath28 ensures that this hamiltonian scales as @xmath29 , the same as the near neighbour heisenberg hamiltonian ( which facilitates the inclusion of quantum fluctuations in a straightforward manner ) , and the ratio @xmath30 has been suggested @xcite on the basis of ( i ) comparisons to raman scattering results , and ( ii ) a numerical simulation that has been shown to be consistent with eq .
( [ eq : keimercl ] ) . from now on we shall assume that this interaction hamiltonian represents the spin texture in the neighbourhood of any sr impurity , and that we can represent the low temperature multiply doped state with such an interaction present at every plaquette bordering a sr site . one may examine the interaction between pairs of these skyrmion like impurity states and explain the transverse spin freezing temperature found experimentally @xcite . in this section
we wish to consider the interactions between more than two such impurity states , as this immediately leads one to an understanding of the possible origin of the cluster spin glass phase . around each impurity site
the ground state is chiral , with the spin distortions corresponding to either clockwise or counterclockwise circulations . on the basis of only this fact
it is straightforward to predict the ground state for two sr impurities : as shown in fig . 1a
, if one chooses opposite chiralities around the two sr sites , then the currents " ( either those of the motion of the hole , or the concomitant spin currents see ref .
@xcite ) add constructively in the region between the impurities .
one may show , either numerically or approximately analytically , that this is indeed the ground state .
now consider three non collinear sr impurities .
it is clear that there is no configuration for which the chiralities assigned to each impurity do not destructively interfere with one another three such impurity states necessarily lead to a frustrated interaction between the spin distortions created by each sr impurity , as shown in fig . 1b , and
the question then arises : what is the ground state of this configuration of impurity sites ?
figure 2 shows the ground state that we have found ( our method for finding the ground state is detailed in the next subsection @xcite ) for three impurity sites placed in the positions of an isosceles triangle with the lengths of the sides roughly equivalent to the separations between randomly placed sr impurities for a doping level in the middle of the sg phase .
figure 2a shows the actual spin directions , whereas fig .
2b shows the spins on all b sublattice sites flipped , as this substantially improves the visualization .
the hamiltonian is isotropic in spin space , and thus we have complete freedom to orient the spins in any direction we have chosen that the afm order parameter would point directly out of the page infinitely far from these impurities .
figure 2b most clearly shows the very important result that the state that is produced by the frustrated current interactions is a small domain of afm correlated spins whose orientation is rotated with respect to that of the bulk afm . for example , in the configuration depicted in the figures the domain s afm order parameter points toward the top of the page whereas the order parameter of the bulk afm points out of the page .
also , there are narrow domain walls separating this domain from the bulk of the lattice , although as expected theoretically @xcite the spins distortions decay algebraically as one moves away from the sr impurities . with many sr impurities distributed randomly above and below a plane
, we thus expect to find many such domains , and we now analyze the results from our numerical determination of the ground states showing that this is indeed an appropriate way to think of this phase : these domains are the predominant components of the cluster sg spin texture . as explained above
, the ground state for a random distribution of sr impurities is frustrated ( _ viz .
_ , the interactions between impurities are frustrated ) .
thus , standard numerical approaches such as the conjugate gradient method usually find low lying states , but not the ground state .
we have numerically searched for the true ground states using a technique first put forward to examine a random distribution of frustrating fm bonds doped into a 2d afm xy model on a square lattice @xcite , a technique that from now on we shall refer to as the grempel algorithm .
this approach was formulated after earlier work @xcite showed that the lower energy states of this latter model could be thought of in terms of small domains of afm correlated spins separated by domain walls , very similar to the spin texture that we are proposing for our perturbed heisenberg model .
given that the grempel algorithm is helpful in illuminating the physics of the cluster sg phase , we briefly explain its key features with regards to the sr impurity problem discussed in this paper .
firstly , for a given lattice size and distribution of the sr impurities we start by assigning random directions to all spins .
then , using the conjugate gradient technique we relax " the system and find a state that is a local minimum on the sg energy surface .
we repeat this for up to fifteen initial conditions ( our results did not change when we experimented and used more than fifteen ) .
then , we applied the grempel algorithm to find lower energy configurations : we begin with the lowest energy relaxed state and first calculate the local field at each lattice site , the latter being defined by @xmath31 ( although @xmath32 is not the exact local field for the spins on the sr
doped plaquettes , it nonetheless supplies an adequate description of the disorder and frustration in the spin morphology . ) if these classical spins are assumed to be of a unit magnitude , then the maximum magnitude of the local field is 4 . then we calculate the local field s magnitude for each spin , and if it is lower than some critical field @xmath33 , we change its orientation to be any direction . for those sites which have a large local field ,
only a small random angle is used ( chosen from no more than 5 % of the 4 @xmath34 steradians on the unit sphere ) .
this method ensures that large domains of afm correlated spins have their integrity preserved at this step , whereas the domain walls , the regions of more strongly disordered spins ( and thus smaller local fields ) , are allowed to more fully explore the full spectrum of spin directions .
then , another local minimum is found using the conjugate gradient method and the grempel algorithm is repeated with this new local minimum . this procedure allows the system to more easily access its lower energy states .
we have varied the critical field @xmath33 from @xmath35 up to @xmath36 after each application of the conjugate gradient method , and find that after only 3 or 4 applications of the grempel algorithm the ( presumably ) true ground state is found .
a full critique of the success of this method may be found in ref .
@xcite .
we have found ground states ( or at the very least , very low energy states ) for sr doping concentrations corresponding to values of @xmath37 and 0.08 .
different lattice sizes , @xmath38 with @xmath39 20 , 30 , 40 , 50 and 60 , and different distributions of sr impurities , at least five for each @xmath40 , are implemented as described in the appendix . in order to obtain better statistics we also included @xmath41 clusters for @xmath42 and @xmath43 .
( we impose one restriction on each distribution of sr impurities in an attempt to mimic the coulomb repulsion between impurities that would tend to force them apart during the crystal growth process , we do not allow for two near neighbour plaquettes to both contain sr impurities . ) as a check that we are indeed finding ground states consistent with the actual spin texture of lasrcuo , from these ground states we obtained the following quantities : \(i ) the bulk value of the staggered magnetization . assuming that the finite size scaling for the staggered magnetization is the same as it is for the undoped lattice _ if _ the system possesses long ranged order @xcite , _ viz .
_ @xmath44 , we find that for @xmath45 long ranged order is absent ; for @xmath42 finite size effects are too large for us to determine if long
ranged order is present .
thus , it is unclear whether or not our @xmath42 simulations are properly reflecting the absence of long ranged afm order found experimentally @xcite for @xmath46 .
\(ii ) the zero temperature correlation length @xmath47 doping .
we have used the finite size scaling ideas that are covered in the appendix to extract this quantity from our ground states .
our results are shown in fig .
3 , along with the compilation of all two axis experimental data contained in ref . @xcite .
clearly , over the @xmath48 range of doping our numerically determined correlation lengths agree with experiment quite well . for @xmath42
we find a correlation length that is significantly less than the approximately 40 lattice constants found in the experiments of keimer em et al . @xcite .
the possible reasons behind this failure are similar to those discussed above : for such a large correlation length we would have to study systems that were of a linear dimension that were at the very minimum twice the correlation length .
also , because of the low concentration , the average should be over more distributions of disorder to properly sample what a low concentration bulk system would be like .
thus , for the @xmath42 simulations both finite size effects and a restricted averaging over disorder potentially limit the reliability of our simulations and thus prevent a proper comparison with experiment .
we have been able to fit our @xmath49 data for @xmath50 to the following empirical form ( setting the lattice constant to be one from now on ) : @xmath51 we find that the best fit occurs for @xmath52 and @xmath53 , and thus our data are reasonably consistent with the scaling @xmath54 .
this implies the vanishing of long ranged afm order in the plane at t = 0 for any non
zero density of sr impurities , a result that may be understood in terms of the algebraic decay of the spin distortions produced by a single hole localized near a sr impurity .
the resulting curve is shown in fig . 3 , and clearly represents a good fit to both our numerical data and the data found in the study of the 3 and 4 % samples by keimer _ et al .
_ @xcite .
the @xmath42 results described in the two points listed above suggest that , due to finite size effects , we are unable to reliably simulate the @xmath42 phase , and thus from now on we focus on samples with a minimum concentration of @xmath43 . encouraged by the @xmath50 successes , we now describe the spin texture , at a microscopic level , that is associated with these ground states .
we proceed in a manner that most clearly displays the existence of the domains . the first information that details the kind of inhomogeneous spin texture that exists in the sg phase is the local field defined in eq .
( [ eq : local field ] ) . in fig .
4a we show the local fields for a @xmath43 distribution of sr impurities ( this concentration is in the middle of the sg phase ) on a @xmath55 lattice .
the plaquettes having a sr impurity above or below them are shown as open squares , and the size of the local field at the cu sites is indicated by the radius of a solid circle .
for this figure all local fields less than 99 % of the maximum local field are not included .
( smaller values of this field do not affect the qualitative features of this figure : the domain wall widths become smaller and the continuity of the domain walls connecting neighbouring sr sites is not as clear , so here we use this relatively large local field to make the domain walls appearance as obvious as possible . )
this figure displays some evidence of clusters of strongly afm correlated spins ; clearly , some regions of spins are no longer strongly afm correlated with other regions .
however , what this figure most clearly shows is that there are channels of disordered spins ( defined as having a reduced local field ) that are set up connecting neighbouring sr impurities these paths of disordered spins are the domain walls of the clusters .
more complete information concerning the actual spin texture that is found follows from the kind of figure shown in fig .
this diagram corresponds to the same ground state shown in fig .
4a , and is constructed in the same way as fig .
2 the projection of each spin onto the xy plane is shown , and every spin on a b sublattice site is flipped so that a completely ordered state would appear to be a fm state with all spins pointing up out of the page .
a careful examination of this figure clearly shows the existence of the domains .
for example , in the upper middle portion of this figure a region bounded by six sr impurities is seen to produce a domain of strongly afm correlated spins .
further , one may characterize the sizes of the domains using figures such as fig .
4b we have done this for a number of the ground states found in @xmath55 lattices , and observed that the average number of spins per domain varies as @xmath56 to within 10 % .
many other examples of these domains can be found , some being three , four , @xmath57 sided domains .
these assemblies are always associated with the sr impurities being at the edges of the clusters , and thus it is clear that these afm correlated regions are a result of the dopant disorder .
it is the domain like spin texture of this phase which motivates us to refer to it as a cluster sg .
one question that this spin texture answers is the elimination of long ranged afm order : we find that the orientations of the local afm order parameters of these clusters is random , and thus the disappearance of afm order may be thought of in terms of the lack of orientational order of the local cluster moments .
thus , the long ranged afm order is replaced by the sg order .
one may calculate @xmath58 for the ground states of the @xmath42 or @xmath43 states , and it is always found that the peak occurs at the commensurate afm wave vector @xmath18 .
( for higher doping concentrations incommensurate correlations develop , a subject that we shall discuss in iv . )
this is consistent with the experimental observation that the spin correlations remain commensurate in the sg phase @xcite . to see that this may still be the case even if the holes are
mobile simply note that holes like to move in fm regions rather than afm ones . then , since the domain walls are more disordered than the interior of the domains , they necessarily have a greater fm correlation present ; thus , the holes will tend to move along the domain walls .
put another way , the holes are _ expelled _ from the interiors of the domains , and only move along the domain walls .
( a complete theoretical discussion of the transport in the low @xmath40 compounds will be presented elsewhere @xcite . )
this provides a plausible explanation of why the spin correlations remain commensurate both at low temperatures , when the hole s motion is dominated by the sr impurity potentials , or at higher temperatures , where the holes are mobile ( and a resistivity that varies linearly with @xmath14 is found ) .
we note that a similar explanation of the commensurability of the sg phase was pointed out by emery and kivelson @xcite .
an important experimental result contained in the study of ref .
@xcite involves the behaviour of the susceptibility _ vs. _ temperature just above the sg transition temperature , @xmath59 . to be specific , @xmath60k and it was found that from roughly 20 k to 100 k the susceptibility @xmath61 could be described via a curie form with the weiss constant equal to zero : @xmath62 for one well studied @xmath63 sample it was found that the curie constant @xmath64 was very small , corresponding to only @xmath65 of the cu spins present in the sample .
other samples studied in the work associated with ref .
@xcite were found to have similar susceptibilities with curie constants corresponding to between @xmath66 and @xmath67 of the cu spins present in the sample @xcite . in this section
we present new experimental results that detail the manner in which @xmath68 varies with doping level , and then show how the spin texture described at the end of the previous section leads to a straightforward and plausible explanation of this behaviour .
single crystals of @xmath0 ( @xmath40 = 0.02 , 0.025 , 0.035 , and 0.045 ) of masses around 2.0 @xmath69 have been grown using a crucible free , contamination
free , travelling
solvent floating zone ( tsfz ) method @xcite .
the stoichiometric feed rods are prepared from the proportional amount of high purity originals ( @xmath70 ) @xmath71 and cuo .
the solvent rods are prepared using cuo as self flux ( @xmath72 ) .
the optical floating
zone furnace uses four elliptical mirrors with halogen lamps at one focal point and the solvent zone is melted at the second focal point .
the growth rate is about 1 mm / hour , the rotation rate is 50 r.p.m .
opposite , and the growth is done under air flow . the as grown single
crystal samples were prepared using identical preparation techniques it was hoped that this would produce a reliable trend of the curie constant with doping ( see below ) .
for the first three concentrations , this was the case .
unfortunately , the @xmath73 sample showed signs of superconductivity , and this concentration will only be addressed in the next section .
the as grown crystals were subsequently exposed to additional treatments : oxygen annealing or vacuum annealing .
the oxygen annealing was performed at a temperature of 600 @xmath74c for 10 hours .
the vacuum annealing has been done at 850 @xmath74c under @xmath75 torr for one hour .
the susceptibility results for these samples were very similar to those of the as grown samples , thus supporting our contention that these are excellent single crystals with a high degree of sr impurity homogeneity .
the magnetization as a function of temperature has been measured using a quantum design squid magnetometer .
the @xmath76 _ vs. _ @xmath77 data indicate that there are no ferromagnetic impurities present within our instrumental resolution .
the obtained @xmath76 _ vs. _ @xmath14 data were replotted as @xmath78 _ vs. _ @xmath14 to extract the curie constants from the region above @xmath59 in which the slope is a constant .
the data for our samples are presented in fig .
5 ; the 0.5 % value for the curie constant quoted for the @xmath63 crystal studied in ref .
@xcite is similar to that extracted from this figure .
the curie constant is expressed as follows : for a given mass of lasrcuo we calculate the curie constant that would be expected if all of the cu spins were isolated @xmath79 spins with @xmath80 , and call this @xmath81 .
then , we express the curie constant found in the region above @xmath59 as a fraction of @xmath81 .
the errors in our data are roughly constant and equal to 10 % of the measured values . in the as
grown samples there is a clear indication of a linear variation with doping concentration , approximately given by @xmath82 for the o@xmath13 annealed samples it could be argued that the curie constant grows quadratically with doping , whereas for the vacuum annealed samples it appears that the curie constant is nearly independent of @xmath40 ( we note , however , that the oxygen depletion is expected to be largest for the higher sr dopings for vacuum annealing , and thus our identification of hole doping levels with @xmath40 is least accurate for this preparation technique ) . despite this variation ,
it is pleasing that in all cases the magnitude of the curie constant is roughly similar , and we now present a heuristic argument that accurately tracks this behaviour .
the results from the work of iic suggested that the spin texture of the low temperature moderately doped lasrcuo system can be thought of in terms of small clusters of afm correlated spins separated by narrow domain walls of disordered spins .
the linear dimension of these domains tracks the zero temperature correlation length .
also , in the above section we showed how the curie constant found in the temperature range just above @xmath59 varied with @xmath40 , with an anomalously small number of effective cu spins contributing to this constant . in this section
we show that these two results are compatible with one another , and thus establish the consistency between phenomenologies promoting the existence of a cluster sg @xcite and the experiments that showed that this material has a canonical sg phase @xcite .
we describe the cluster sg spin arrangements in terms of domains of spins , the sizes of the domains corresponding , on average , to the zero temperature correlation length , and narrow domain walls composed of strongly disordered spins . due to the domain walls we make the assumption that there is no interaction between the domains
this is consistent with the vanishing weiss constant found experimentally .
then , the effective curie constant for this spin texture follows immediately : as a simple picture of such spin configurations , let each domain be of a @xmath38 geometry ; further , take @xmath83 . noting that the quantum
mechanical ground state for such a cluster is a @xmath84 singlet for @xmath85 even , and a @xmath86 doublet for @xmath85 odd , _ and _ that the gaps to the first excited states are large @xcite , as a first approximation we take the singlet clusters to be magnetically inert ( in the temperature interval over which the susceptibility was measured ) .
accordingly , for a lattice with @xmath87 @xmath88 spins , the susceptibility arises from @xmath89 spin@xmath90 magnetic units . finally , for a cluster sg we predict @xmath91 the result of using eq .
( [ eq : xi critical ] ) for the correlation length is also shown on fig . 5 , and displays the quantitative agreement between this admittedly highly simplified theory and experiment .
given that we have ignored any contribution from the domain walls , it is not surprising that our value for the curie constant is somewhat less than what is observed experimentally .
further , eqs .
( [ eq : xi critical],[eq : curie constant ] ) together approximately yield @xmath92 , consistent with the overall trend of an increasing curie constant with @xmath40 .
in _ all _ ground states we have found some evidence for domains .
these clusters are antiferromagnetically correlated , so it is not surprising that in the sg phase we found that the dominant magnetic correlations are commensurate , as we described in iid . however , as @xmath40 increases we find that the areal fraction associated with the domain walls starts to dominate at the expense of the clusters , and thus the persistence of afm correlations is no longer guaranteed .
this is shown most clearly in fig .
6 wherein we show a @xmath93 result for the local fields the @xmath43 result for the same distribution of sr impurities ( at a lower concentration see the appendix ) is shown in fig .
as is seen from a comparison of these two figures , the concept of domain walls is lost : only in a small number of regions can one say that there are any clusters , and thus the concept of domain walls separating the clusters is irrelevant . instead
, almost everywhere there is a more homogeneously distorted spin texture , and effectively dopant disorder is being averaged out .
this is consistent with the experimental result @xcite that for @xmath94 one only observes the curie behaviour over a much smaller temperature range ( @xmath95 ) , and the curie constant that is found is much smaller , whereas for higher @xmath40 no curie like behaviour is found at any temperature .
we now consider this behaviour in relation to the observed commensurate
incommensurate transition .
recent experimental work by yamada _ at al . _
@xcite has shown that the incommensurate ( ic ) correlations found in more strongly doped lasrcuo @xcite develops continuously as the sr doping concentration is increased , with a critical concentration corresponding to @xmath96 .
it is very intriguing that these correlations first develop at the @xmath40 value for which superconductivity first appears , and thus makes the mechanism behind the appearance of the ic phase an important question .
here we show that our model , based on spin distortions produced by holes localized around sr impurities at low temperatures , reproduces this commensurate incommensurate transition .
most importantly , any model of mobile holes that induces spin distortions analogous to those discussed here may also produce a similar ic phase ( see below ) , and it can be argued @xcite that what we are analyzing are snapshots " of the kinds of effects produced by mobile holes generating dipolar backflow spin distortions . we have only calculated the ground state spin textures for this system .
thus , we can not directly compare our results to the measured dynamic structure factors @xmath97 found experimentally .
instead , we examine @xmath58 , and look for signs of incommensurate correlations in the snapshots of the ground state spin texture .
we find that at @xmath98 two of the five distributions of sr impurities generate ic correlations . for
@xmath99 three of the five distributions generate ic correlations , whereas for @xmath100 all five of the distributions of sr impurities generated ic correlations . from these numbers
we conclude that the c ic is seemingly continuous and happens close to @xmath101 .
this number is very close to that found experimentally .
also , in _ both _ experiment and our simulations the correlations develop along @xmath102 or @xmath103 .
this agreement between our work and experiment is striking , and suggests that mechanisms which produce the ic phase by spin distortions , such as the spiral phase mechanism of shraiman and siggia @xcite , should be re examined with an eye to understanding the stability of this phase in the presence of strong sr impurity disorder ( albeit with a bias towards maximum spin distortions around the sr impurities ) .
we have considered a simple model of the effect of holes at low temperatures that are localized around the sr impurities in lasrcuo and have found evidence for clusters of afm correlated spins in the sg phase we believe that this spin texture provides ample reason to refer to this state as a cluster sg . that this two fold degenerate chiral ground state leads to a sg phase
lends support to the prescient speculations of villain @xcite .
we briefly note that we have also considered another related model of spin distortions produced by localized o holes , the so called aharony model , and we still find strong evidence for such clusters in this model of sr impurity disorder .
thus , our result is seemingly model independent _ providing _ that the sr disorder is included explicitly .
we have examined in greater detail , using new tsfz grown crystals , the anomalously small curie constant that was measured in ref .
@xcite , and have characterized both its magnitude and the manner in which this constant changes with doping . we have thereby provided a quantitatively accurate model of this curie constant by utilizing the idea that the clusters behave like independent low
spin quantum units .
we have also shown that the c ic transition in our model is continuous , and occurs at roughly the same @xmath40 as is found experimentally .
we thus attribute the critical value of @xmath104 to sr impurity disorder .
also , our model induces incommensurability in the same axial directions ( @xmath6 ) as is found experimentally .
given that the spin distortions produced by the partially localized holes studied in this paper are similar to the dipolar backflow spin distortions induced by mobile holes in the theories of shraiman and siggia , it is possible that a model of such holes including the effects of sr disorder for @xmath105 will provide an accurate representation of the physics of the intermediately doped lasrcuo system studied here .
we note that previous theoretical claims of an incipient phase separation of the spiral phase must be reinvestigated using a model that includes explicitly the sr disorder ; it is likely that disorder will completely suppress this instability .
we will report on the results of related studies in a future publication . in none of the theoretical work presented here
have we included the quantum fluctuations of the cu spins , instead relying on a classical model of the spins .
support for this idea follows in part from the success of the renormalized classical model in describing the undoped phase .
however , we are now utilizing inhomogeneous , linearized , spin wave theory , including the modified spin
wave theory constraint to enforce the fact that there is no long ranged order in the sg phase , to assess the importance of quantum fluctuations to these results .
further , upon completion of these results , we should better be able to test if the dynamical fluctuations of such a cluster sg phase are responsible for the unusual scaling of @xmath106 found by keimer _
et al._@xcite .
we have also not included the xy anisotropy , a component that could possibly affect our results .
however , the topology of a cluster sg was first found in a 2d xy model @xcite , and thus it seems highly unlikely that this small interaction would qualitatively affect our conclusions , especially given the small afm correlation length in the sg phase .
other magnetic anisotropies would serve to sharpen the domain walls separating the clusters , and thus could only serve to strengthen our arguments .
lastly , we note that the ideas expressed in this paper are not necessarily limited to the sr doped 214 compound .
recent work by budnick , _ et al . _
@xcite , has shown that sr
doped 123 ( with the sr substituting for the y ions ) produces a similar sg phase , and by symmetry we expect similar impurity ground states in this material .
also , li doped 214 , studied originally by endoh _
@xcite , and more recently examined in ref .
@xcite , potentially has chiral impurity states , as recently discussed by haas , _
so , it is possible that similar physics might also be found in this compound .
we wish to thank david johnston for providing us with a collection of the measured correlation lengths , and for other valuable comments .
we also wish to thank barry wells , martin greven , marc kastner , pakwo leung and ferdinando borsa for helpful comments .
this work was initiated while one of us ( rjg ) was visiting mit , and he wishes to thank them for their warm hospitality .
this work was supported by the nsf , grant number 9315715 and by the nserc of canada .
one of the outstanding problems of theoretical solid state physics involves the determination of accurate finite
size scaling forms for disordered systems .
we have attempted to circumvent this problem in the following manner : we have performed our numerical simulations of this spatially disordered system by imagining that we have @xmath107 large samples of the crystal , all having the same average density of sr impurities above and below a cuo@xmath13 plane .
then , for one of the @xmath107 large samples , choose some square subsystem of the sample of dimension @xmath108 , and find its ground state spin texture .
then , as schematically shown in fig . 7 , we have studied increasingly larger subsystems , namely @xmath109 and then @xmath110 , _ etc .
_ , with @xmath111 that include exactly the same distribution of impurities in the smaller subsystems as are present in the larger subsystems . then , as one increases the density of sr impurities we add more impurities to the initial distribution to raise @xmath40 to the desired value .
from such sets of data we have found that we can best obtain good statistics for each sample , and then we perform an average over all @xmath107 samples .
further , the manner in which we increase the density of impurities leads to smoother variations of quantities with @xmath40 .
we repeat this process for a number of differing initial , low density distributions of impurities , and we find that quantities like the correlation length are very well converged with only five samples per @xmath40 .
r. j. gooding , n. m. salem , r. j. birgeneau , and f. c. chou , contributed paper at the 10th anniversary workshop on high @xmath7 superconductors : materials , mechanisms , and applications ( houston , march 1996 ) . | we examine the low temperature magnetic properties of moderately doped @xmath0 , paying particular attention to the spin glass phase and the commensurate incommensurate transition as they are affected by sr impurity disorder .
a model of the carriers , believed to be appropriate at low temperatures , is employed in the @xmath1 planes that accounts for both the strong coupling of the hole s motion to the antiferromagnetically correlated spins and the pinning potential associated the @xmath2 impurities .
this model has been shown to explain quantitatively several magnetic and transport features of the antiferromagnetic region of the phase diagram .
preliminary indications that this model can also explain the @xmath3 region of the phase diagram follow from its success in accounting for the doping and temperature variation of the spin correlation length . here
we further scrutinize this model by attempting to explain various features of the spin texture of the spin glass phase .
new measurements with travelling
solvent float zone grown crystals of the low temperature susceptibility show an increase of an anomalously small curie constant with doping .
this behaviour is modelled in terms of our numerical simulation results that find small clusters of antiferromagnetically aligned regions separated by disordered domain walls produced by the impurities the domain walls lead to a percolating sequence of paths connecting the impurities .
we predict that for this spin morphology the curie constant should scale as @xmath4 , @xmath5 being the spin correlation length , a result that is quantitatively in agreement with experiment . also , we find that the magnetic correlations in the ground states produced by our simulations are commensurate in the spin glass phase , consistent with experiment , and that this behaviour will persist at higher temperatures where the holes should move along the domain walls being effectively expelled from the antiferromagnetically correlated domains .
however , our results show that incommensurate correlations develop continuously around 5 % doping with an incommensurate wave vector along ( @xmath6 ) , consistent with recent measurements by yamada , _
et al .
_ at this doping level the domain walls and not the clusters start to become the dominant feature of the spin texture , and thus the curie behaviour should disappear in the incommensurate phase , again consistent with experiment .
thus , we find that this model is capable of describing the low temperature disorder induced magnetic spin morphologies of the @xmath0 system in the low and intermediate doping regions of the phase diagram . coupling this understanding of the magnetic correlations with the observed transport features ,
it is clear that |
the authors gratefully acknowledge discussions with s. gavin , d. kharzeev , c. loureno , j. moss , and h. satz .
c.g . thanks the school of physics and astronomy at the university of minnesota for its hospitality during a sabbatical leave . c.g . and j.k .
thank the institute for nuclear theory at the university of washington for its hospitality and the department of energy for partial support during the program
probes of dense matter in ultrarelativistic heavy ion collisions " .
this work was also supported by the u. s. department of energy under grant de - fg02 - 87er40328 , by the natural sciences and engineering research council of canada , by the fonds fcar of the quebec government , and by the director , office of energy research , office of high energy and nuclear physics , division of nuclear physics , of the u.s .
department of energy under contract no .
de - ac03 - 76sf00098 and de - fg03 - 93er40792 . | measured @xmath0 production cross sections for 200 and 450 gev / c protons incident on a variety of nuclear targets are analyzed within a glauber framework which takes into account energy loss of the beam proton , the time delay of particle production due to quantum coherence , and absorption of the @xmath0 on nucleons .
the best representation is obtained for a coherence time of 0.5 fm / c , previously determined by drell - yan production in proton - nucleus collisions , and an absorption cross section of 3.6 mb , which is consistent with the value deduced from photoproduction of the @xmath0 on nuclear targets .
pacs numbers : 13.85.ni , 24.85.+p , 11.80.la + propagation of a highly relativistic particle through a medium is of interest in several areas of physics . high energy proton - nucleus scattering has been studied for many decades by both the nuclear and particle physics communities @xcite .
such studies are particularly relevant for the relativistic heavy ion collider ( rhic ) , which will collide beams of gold nuclei at an energy of 100 gev per nucleon , and for the large hadron collider ( lhc ) , which will collide beams of lead nuclei at 1500 gev per nucleon @xcite . in a recent paper
@xcite we modeled the production of high mass , drell - yan , pairs of leptons as measured in collisions of 800 gev / c protons incident on a variety of nuclear targets by the fermilab collaboration e772 @xcite .
our modeling was done on the basis of hadronic degrees of freedom .
we took into account the energy loss of the beam proton as it traversed the nucleus as well as the landau - pomeranchuk - migdal effect @xcite .
the latter basically acknowledges that there is a time ordering in the appearance of produced particles ; the hard particles ( drell - yan pairs ) appear before the soft particles ( typically pions ) .
this is a quantum mechanical effect , essentially the uncertainty principle . by fitting to the atomic mass number dependence of the drell - yan cross section at high feynman @xmath1 we were able to infer a value for the proper coherence or formation time of [email protected] fm / c .
this value is about what should be expected _ a priori_. in the center of mass frame of the colliding nucleons at the energies of interest a typical pion
is produced with an energy of e@xmath3 mev . by the uncertainty principle
this takes a time of order @xmath4 fm / c .
a related process is the production of @xmath0 in high energy proton - nucleus collisions which we shall address in this paper .
this is also a relatively hard process and so both energy loss of the beam proton and the landau - pomeranchuk - migdal effect must be taken into account .
however , there is an additional effect which plays a role , and that is the occasional absorption or breakup of the @xmath0 in encounters with target nucleons .
( the inelastic interaction of one of the leptons in drell - yan production with target nucleons is ignorably small . )
the absorption cross section , @xmath5 , has been estimated in a straightforward glauber analysis without energy loss and with an infinite coherence / formation time to be about 6 - 7 mb @xcite .
this has formed the basis for many analyses of @xmath0 suppression in heavy ion collisions .
any anomalous suppression may be an indication of the formation of quark - gluon plasma @xcite , hence the importance of obtaining the most accurate value of @xmath5 possible .
this cross section has also been inferred from photoproduction experiments of @xmath0 on nuclei from which a value much less than that has been obtained @xcite .
this has been a puzzle .
one attempt to resolve this apparent discrepancy consists of modeling the produced @xmath6 state as a pre - resonant color dipole state with two octet charges @xcite ; however , the results are only semi - quantitative . for a basic description of high energy proton - nucleus scattering we prefer to work with hadronic variables rather than partonic ones .
we make a straightforward linear extrapolation from proton - proton scattering .
this extrapolation , referred to as lexus , was detailed and applied to nucleus - nucleus collisions at beam energies of several hundred gev per nucleon in ref .
@xcite , and to drell - yan production in 800 gev proton - nucleus collisions in ref .
@xcite . in order to compute the production cross section of @xmath0 in proton - nucleus collisions
we need a parametrization of it in the more elementary nucleon - nucleon collisions . for this
we call upon the parametrization of a compilation of data by loureno @xcite .
@xmath7 here @xmath8 is the branching ratio into dimuons and @xmath9 is the ratio of the momentum carried by @xmath0 to the beam momentum in the center of mass frame @xmath10 . due to the degradation in momentum of the proton as it traverses the nucleus it is important to know the @xmath9 dependence of the production .
the fermilab experiment e789 has measured this dependence at 800 gev / c @xcite to be proportional to @xmath11 .
assuming that this holds at lower energy too we use the joint @xmath12 and @xmath9 functional dependence and magnitude : @xmath13 the cross section in proton - nucleus collisions can now be computed in lexus with no ambiguity .
with @xmath14 in proton - nucleus collisions at 200 gev / c .
the data is from na38 @xcite .
the dashed line is @xmath15 times the nucleon - nucleon production cross section .
the solid curve represents full energy loss with zero coherence / formation time , while the banded region represents partial energy loss with a coherence / formation time within the limits set by drell - yan production .
( computations were done for c , al , cu , w and u and the points connected by straight lines to guide the eye.),width=377 ] figures 1 and 2 show the results of our calculation in comparison to data taken by na38 @xcite .
the dashed curves are @xmath15 times the nucleon - nucleon production cross section ; they obviously overestimate the data .
the solid curves show the result of lexus with full energy degradation of the beam proton without account taken of the landau - pomeranchuk - migdal effect ; they obviously underestimate the data .
the hatched regions represent the inclusion of the latter effect with a proper formation / coherence time @xmath16 in the range of 0.3 to 0.5 fm / c consistent with drell - yan production @xcite .
the time delay is implemented as follows @xcite : the energy available for the production of @xmath0 is that which the proton had @xmath17 collisions prior ; that is , the previous @xmath17 collisions are ignored for the purpose of determining the proton s energy .
this is an approximate treatment of the landau - pomeranchuk - migdal effect .
the @xmath17 is related to the beam energy and to the coherence time @xmath16 in the center of mass frame of the colliding nucleons .
the proper coherence time is essentially the same as the proper formation time of a pion since most pions are produced with rapidities near zero in that frame .
the first proton - nucleon collision is the most important , so boosting this time into the rest frame of the target nucleus and converting it to a path length ( proton moves essentially at the speed of light ) gives @xmath18 .
this path length may then be equated with @xmath17 times the mean free path @xmath19 . using a total cross section of 40 mb , a nuclear matter density of 0.155 nucleons/@xmath20 , and @xmath21
fm / c we obtain @xmath22 at 200 gev / c and @xmath23 at 450 gev / c .
as may be seen from the figures , the data is overestimated , indicating the necessity for nuclear absorption .
absorption cross section of 3.6 mb.,width=377 ] absorption cross section of 3.6 mb.,width=377 ] we now introduce a @xmath0 absorption cross section on nucleons and compute its effect within lexus in the canonical way @xcite .
when the @xmath0 is created there will in general be a nonzero number of nucleons blocking its exit from the nucleus .
knowing where the @xmath0 is created allows one to calculate how many nucleons lie in its path , and hence , to compute the probability that it will be dissociated into open charm .
we choose a value of @xmath16 allowed by drell - yan measurements , mentioned above , and then vary @xmath5 , assuming that it is energy independent .
the lowest value of chi - squared for the 200 and 450 gev / c data set taken together is obtained with @xmath24 fm / c and @xmath25 mb .
the results are shown in figures 3 and 4 .
the fitted values all lie within one standard deviation of the data points .
this is quite a satisfactory representation of the data .
it means that both drell - yan and @xmath0 production in high energy proton - nucleus collisions can be understood in terms of a conventional hadronic analysis when account is taken of the energy loss of the beam proton , the landau - pomeranchuk - migdal effect , and nuclear absorption of the @xmath0 in the final state .
it also means that the absorption cross section for @xmath0 inferred from high energy proton - nucleus collisions is consistent with the value inferred from photoproduction experiments on nuclei
. it will be very instructive to repeat this analysis in the language of partonic variables .
actually , the analysis with parton energy loss alone was reported by gavin and milana @xcite with satisfactory results obtained for drell - yan and @xmath0 if the partons lose about 1.5 gev / fm .
nuclear shadowing @xcite needs to be taken into account too .
the relationship among all these effects is not well - understood , nor is the relationship between these effects in partonic and hadronic variables .
finally , the implications for nucleus - nucleus collisions @xcite will undoubtedly be important ; they are under investigation . |
confined polymers play an important role in many industrial processes and biological systems .
examples range from membrane filtration and oil recovery to gel electrophoresis and protein translocation @xcite .
importantly , recent technological development in nano-/micro - fluidics has made it possible to manipulate and trap biomolecules such as double - stranded ( ds ) dna in confined environments with a characteristic lengthscale that is much smaller than the radius of gyration of the polymers @xcite .
also under biological conditions , dna is often strongly confined , e.g. packed into a viral capsid @xcite , bacteria @xcite or the eukaryotic cell nucleus @xcite . in this article , we report molecular dynamics simulations that allow us to determine the typical speed of the segregation of initially mixed polymers in cylindrical confinement .
this problem has particular relevance for the understanding of chromosome segregation in bacteria , where the nature of its underlying mechanism is currently under debate . here , the basic issue is whether the major driving force for segregation of duplicating chromosomes in strong confinement is physical ( driven by entropy or mechanical `` pushing '' ) @xcite or biological ( such as cytoskeletal and motor proteins ) @xcite .
our results show that the effective repulsion between two chains in a cylindrical geometry of confinement can be very strong .
typically , the segregation requires a time proportional to @xmath0 , which is much faster than the @xmath1 timescale of chain diffusion , where @xmath2 is the chain length .
this suggests that for filamentous bacteria such as _ streptomyces coelicolor _
@xcite or cyanobacterium _ anabaena _
@xcite , the main driving force of chromosome segregation might be entropic and any additional mechanisms are for `` optimization . '' as we shall discuss later , our proposal is fully consistent with the recent results that chromosome segregation in some filamentous bacteria is a _ random _ process @xcite .
the chains consist of @xmath2 beads of size @xmath3 each .
@xmath4 denotes the chain extension , and @xmath5 the distance of the centers of mass . ]
consider two linear chains with excluded - volume interactions , which are initially intermingled and confined in an infinitely long cylinder with a diameter @xmath6 that is much smaller than the radius of gyration @xmath7 of the unconfined chains ( fig .
[ fig : blobs ] ) . as the two chains can gain conformational entropy by demixing
, they effectively repel each other .
note that the free energy cost for simultaneous overlap of @xmath8 chains scales as @xmath9 in the dilute regime and then increases faster as @xmath10 in the more concentrated regime , independent of the chain length @xcite .
thus , for two intermingling chains illustrated in fig .
[ fig : blobs ] , each blob - blob overlap contributes @xmath11 ( @xmath12 ) to the free - energy cost .
the potential of mean force for segregation between the two chains then is proportional to the total number of overlapping blobs , i.e. , @xmath13 , where @xmath14 is the equilibrium length of an isolated individual chain in the pore , and @xmath5 the center - to - center distance between the two chains .
the effective repulsive force is then obtained as @xmath15 and , thus , the equation of motion for the center of mass is @xmath16 where @xmath17 and @xmath18 are the total mass and the effective friction of the chain , respectively ( @xmath19 is the mass of a single monomer ) . ignoring hydrodynamic interactions between monomers , one can assume that the frictions @xmath20 on the individual monomers are additive , i.e. , also @xmath21 . then , the solution of eq .
[ eq : eqmotion ] with an initial condition @xmath22 at @xmath23 can be obtained as @xmath24 where @xmath25 is the `` inertial '' timescale . in practice , @xmath26 and
hence the characteristic segregation speed is constant and given by @xmath27 the equilibrium length of confined chains , @xmath14 , is proportional to @xmath2 .
therefore , the time for reaching complete segregation , @xmath28 , scales as @xmath29 this is time is much shorter than @xmath30 , the typical time it takes a single chain to diffuse over a distance equal to its own length : @xmath31 however , the above considerations do apply for the initial situation of complete overlap , @xmath32 . in this case
the system is in a state of unstable equilibrium , since the effective segregation force is @xmath33 .
hence , the system will initially show purely diffusive behaviour until a certain separation , typically @xmath34 , is reached .
we refer to the time until segregation sets in as the `` induction time '' @xmath35 that should scale as @xmath1 . with increasing @xmath6 , diffusion becomes easier , because the monomer concentration decreases , and @xmath35 decreases , while @xmath36 increases with @xmath6 .
below , we show that for all practically relevant diameters , the segregation process is rate limiting . for real bacteria , entropic segregation already sets in during replication .
therefore , segregation always takes place for comparatively short pieces of dna , so that the induction time does not play a role .
in the molecular dynamics simulations , we model the polymers using a bead spring model in a cylindrical compartment of diameter @xmath6 ; each chain consists of @xmath2 beads of diameter @xmath3 . the bead
bead and bead compartment interactions were modeled by a weeks - chandler - andersen potential ( wca ) @xcite , which corresponds to the repulsive part of the lennard - jones potential : @xmath37 \label{eq : lj}\ ] ] for @xmath38{2}\sigma$ ] and 0 elsewhere .
@xmath39 denotes the distance between two bead centers for the bead bead interactions , and the distance between the bead center and the compartment minus @xmath3 for the bead compartment interactions . at @xmath40 ,
the interaction energy is @xmath41 ; since the potential is quite steep , @xmath39 will typically stay above @xmath42 .
this models soft beads of diameter @xmath3 , whose centers can not come much closer than @xmath3 to each other , and can not penetrate the wall ( i. e. the wall imposes a constraint on the sphere centers , as depicted in fig .
[ fig : blobs ] ) . in the simulation
, @xmath3 defines the basic length scale and @xmath43 the energy scale .
our unit of mass is given by @xmath19 , the mass of a bead .
we choose the temperature such that @xmath44 .
having specified our basic units , the time unit is given by @xmath45 . in the following , we will omit these units .
the springs between the beads in a chain were formed by the fene ( finite extensible nonlinear elastic ) potential @xmath46\ , , \label{eq : fene}\ ] ] where @xmath39 is the distance of the bead centers , @xmath47 is the radius at which the potential becomes singular , and @xmath48 is the interaction strength . in the present simulations
, we chose @xmath49 and @xmath50 . in combination with the wca potential this results in a typical bond length of @xmath51 .
n=100 n=200 n=300 + .the simulation parameters for the different runs .
the first three tables give for different chain lengths @xmath2 the simulated pore diameters @xmath6 and the corresponding equilibrium end to end distances @xmath14 of a single chain .
the last table contains the number @xmath52 of timesteps used for equilibration of the interconnected chains , the number of timesteps @xmath53 between recorded configurations , and the number @xmath54 of independent simulations runs with different random seeds . [ cols="<,<,<,<,<,<,<,<,<,<,<",options="header " , ] we simulate this system using the simulation package espresso @xcite . to propagate the system , we employ a velocity - verlet md integrator with a fixed time step of @xmath55 ; the system is kept at constant temperature by means of a langevin thermostat with a fixed friction of @xmath56 , so that @xmath57 .
other parameters vary for the different simulation runs , see table [ tab : simparams ] .
our simulation procedure contains four steps : the system is initially prepared in a `` ladder '' configuration formed by two interconnected zig
zag strands , i.e. , the system consists of two linear chains where the @xmath58th bead of one chain is bonded to the @xmath58th bead of the other chain , in addition to the bonds to its neighbors within the same chain . to equilibrate the system , we simulate for @xmath52 steps with a `` soft '' wca potential , i.e. a wca potential that has been modified such that the potential is linear for distances smaller than a radius @xmath59 .
we reduce @xmath59 gradually during the equilibration phase , so that the potential converges to the plain wca interaction .
this procedure allows more overlap between beads during the initial equilibration phase , which helps the bonds to quickly relax to the equilibrium length .
after the equilibration of the interconnected chains , we remove the interconnecting bonds to obtain two separate chains whose centers of mass very nearly coincide .
the chains are stretched by about @xmath60 compared to a single chain in confinement due to the cross linking ; however , their length relaxes quickly to almost the same length as a single chain once the cross linking is released .
the timescale for this relaxation is negligible compared to the segregation time .
we continue to simulate the system until the two chains have segregated , i.e. , until the chains do not overlap and their centers of mass are separated by at least the equilibrium length @xmath14 of a single chain , which had been determined beforehand by separate simulations . during this run ,
we record configurations every @xmath53 simulation steps . , @xmath61 , starting from the removal of the interconnecting bonds .
the two gray bands give the total extents along the tube axis of the two chains for a single run , the black lines the positions of their centers of masses .
the positions are relative to the total center of mass @xmath62 of the system and rescaled by the equilibrium length @xmath14 of a single confined chain . ]
this procedure is repeated @xmath54 times ( see table [ tab : simparams ] ) , resulting in @xmath54 independent data sets similar to fig .
[ fig : segrun ] . for each of these data sets ,
we calculate the distance @xmath63 of the centers of mass of the two chains parallel to the cylindrical compartment as a function of time .
initially , @xmath5 is zero due to the preparation of the system , and stays close to zero during the induction time .
eventually , segregation sets in , and @xmath5 grows rapidly until @xmath28 is reached , at which time the chains do not overlap anymore . further increase in @xmath5
is only due to diffusion and is therefore much slower . in the range from @xmath6 to @xmath64
. from this fit
, the induction time @xmath35 is determined as intersection with @xmath65 , and the segregation speed @xmath66 as its slope .
the segregation time is then @xmath67 . ]
[ fig : segscheme ] displays schematically how we extract the induction and segregation times from each run : we fit a linear function @xmath68 to the range in which @xmath63 is between @xmath6 and @xmath64 . here
, @xmath35 is the extrapolated onset time of segregation and @xmath66 is the speed with which the two centers of mass separate in the linear regime .
we always find linear segregation behavior for @xmath69 .
the lower limit implies that the chains are separated by at least one blob diameter , the upper limit guarantees that there is at least one blob - size overlap left .
rescaled by @xmath70 . for better visualization , we actually plot its inverse , @xmath71 .
the dashed line demonstrates the linear scaling of the segregation time with @xmath6 .
the inset shows the segregation time @xmath36 rescaled by @xmath0 , as a function of @xmath6 . ] as can be seen in fig .
[ fig : segtime ] , our simulation clearly support the scaling prediction @xmath72 ( eqn . ) .
the prediction that the segregation time , scales as @xmath0 is only recovered for small tube diameters .
this is not unexpected , because when @xmath6 approaches @xmath14 , the simple blob prediction for @xmath14 breaks down @xcite and the segregation time levels off at the relaxation of a free chain .
the segregation speed relation eqn . however seems to be quite robust even for finite systems .
rescaled by @xmath1 .
the lines are just guides to the eye . ]
the measured average induction time @xmath35 is shown in fig .
[ fig : conftime ] .
the distribution of the induction times has a long tail , which makes it difficult to sample @xmath35 accurately . keeping this caveat in mind
, we find that for small tube diameters the @xmath35 s computed for different chain lengths can be made to collapse if we assume @xmath1-scaling , as expected for a diffusive process .
moreover , we do observe the expected decrease of @xmath35 with increasing @xmath6 .
for larger tube diameters , the induction time increases again ; this is probably due to the fact that for larger @xmath6 the segregation and induction times can not be clearly separated ( @xmath73 for larger diameters .
in fact , in our simulations , the induction time seems to converge to about one quarter of the segregation time for all @xmath2 and @xmath74 .
we stress that for highly confined chains , the diffusive process is only responsible for the segregation over the tiny initial separation necessary to obtain a significant effective entropic force @xmath75 .
the overwhelming part of the chain `` demixing '' is due to directed segregation ( see fig . [
fig : segrun ] ) .
in other words : it may take a while for the system to _ start _ segregating , but the segregation process itself is always governed by the effective entropic force . along the cylinder
axis for @xmath61 , @xmath76 , averaged over all configurations with a center of mass distance @xmath65 ( a ) , @xmath77 ( b ) , @xmath78 ( c ) and @xmath79 ( d ) .
the graphs are centered around the systems center of mass @xmath62 . ]
[ fig : densityseries ] shows the average monomer densities of the two polymers for different center - of - mass distances @xmath5 .
as predicted , the monomer densities of each chain in the overlap region are almost unaffected by the presence of a second polymer in the same space .
hence , the initial monomer density is almost twice as large as for a single chain . during segregation ,
the monomer densities of the individual polymers increase somewhat .
in fact , the snapshots show that the polymers have very nearly separated at @xmath78 , which is significantly less than @xmath80 ; this demonstrates that the polymers deform during segregation : the entropic driving force is strong enough to compress the polymers .
after demixing , the chains expand to their equilibrium length .
our simulations support the scaling prediction that the entropically driven segregation of two confined chains requires a time proportional to @xmath0 . for long chains ,
this time is much shorter than the diffusive time that scales as @xmath1 .
we stress that this speed up of entropically driven segregation does not involve any active ( energy - consuming ) process .
considering the geometry of confinement and the length scales of ( filamenteous ) bacteria , our results strongly suggest that the partitioning of duplicated chromosomes in these organisms is , at least partly , entropy - driven .
since the segregation sets already in during replication , there is no initial `` induction '' regime for bacteria .
indeed , the recent data obtained by hu _
_ on the cyanobacterium _ anabaena _ sp .
pcc 7120 suggests that mreb , a bacterial actin homologue that is speculated by some as a `` track '' for transporting chromosome by putative motor proteins , is important for cell shape but not for chromosome segregation @xcite .
moreover , they also have shown that the ratios of dna content in two daughter cells have a much wider distribution than in the case that the two cells were identical .
this suggests that chromosome partitioning is a random process , in good agreement with our entropy - driven segregation process we presented in this article .
we thank daan frenkel and bae - yeun ha for many helpful comments and discussions . in addition , we are grateful to bae - yeun ha for providing us computational resources that made the simulations performed in this work possible .
this work is part of the research program of the stichting voor fundamenteel onderzoek der materie ( fom ) , which is supported by the nederlandse organisatie voor wetenschappelijk onderzoek ( nwo ) .
aa acknowledges support from the marie - curie program under the european community s sixth framework programme , and sj the post - doctoral fellowships from nserc ( canada ) and marie - curie ( eu ) . | we report molecular dynamics simulations of the segregation of two overlapping chains in cylindrical confinement . we find that the entropic repulsion between the chains can be sufficiently strong to cause segregation on a time scale that is short compared to the one for diffusion .
this result implies that entropic driving forces are sufficiently strong to cause rapid bacterial chromosome segregation . |
one - dimensional ( 1d ) electronic systems are realized in carbon nanotubes and individual polymer molecules and provide an approximate description of strongly anisotropic 3d materials .
it has been known for many years that 1d systems can support unusual correlated electron phenomena such as luttinger liquid physics .
however , electrical and thermal transport in real materials are usually not governed by the free low - energy luttinger liquid fixed point but by an interplay between dangerously irrelevant operators scattering the currents and conserved quantities protecting them.@xcite in order to connect to actual experiments , it is thus essential to study generic microscopic models . over the last decades
a significant number of works @xcite investigated equilibrium charge ( or spin ) transport properties . in particular , the question whether or not so - called integrable models , which possess a complete set of local conserved quantities , can support dissipationless currents at finite temperature was addressed extensively .
less is known about the quantitative effects of integrability - breaking perturbations which are naturally present in any experimental system , and even the qualitative question whether the linear - response conductivity of a nonintegrable model can still be infinite is not answered conclusively.@xcite while experimental measurements of thermal transport driven by a temperature gradient in quasi-1d spin systems already exist,@xcite only a few works investigate this theoretically.@xcite studying nonequilibrium thermal ( or charge ) transport is complicated in general one reason being that is not even clear whether the long - time dynamics can be described by a low - energy theory and constitutes one of the most active areas of research in strongly correlated condensed matter physics.@xcite the primary goal of our work is to obtain quantitative results on steady - state energy flow both near and far from equilibrium and to understand the effects of integrability and correlations .
this is motivated by the experiments listed above and by recent technical advances in dynamical simulations.@xcite as prototypical models we consider a xxz spin-@xmath6 chain in the presence of two perturbations ( dimerization and a staggered magnetic field ) which break integrability@xcite as well the quantum ising model .
using density matrix renormalization group methods we demonstrate that the nonequilibrium energy current driven by a temperature gradient @xmath0 relaxes fast to a finite steady - state value if the linear - response thermal conductivity is infinite,@xcite i.e. if the drude weight @xmath1 is nonzero .
our data indicate that the dimerized chain might support such dissipationless transport ( @xmath2 ) despite the fact that it is nonintegrable ( @xmath1 can be extracted from the asymptote of the equilibrium energy current correlation function,@xcite and we can not exclude that the latter decays on a hidden large temperature - independent time scale ) .
one of our main results is that for a large class of problems the steady - state current takes , within numerical accuracy , the functional form @xmath7 in words , its dependence on the two temperatures is tightly constrained : the steady - state current is the difference between the total radiated power from the left and right leads .
the function @xmath5 is thus a generalization of the stefan - boltzmann law for photons , for which @xmath8 in @xmath9 spatial dimensions .
moreover , eq . ( [ ffunction ] ) implies that nonequilibrium thermal transport is entirely determined by linear response
@xmath5 can simply be obtained by integration of the equilibrium conductance @xmath10 .
we give an intuitive argument for the existence of a stefan - boltzmann function @xmath5 and also shows that for the xxx ferromagnet , @xmath5 can be estimated via thermodynamic bethe ansatz in good agreement with the numerics at low temperatures .
we demonstrate that at low temperatures the gapless integrable xxz chain as well as the quantum ising model exhibit universal nonequilibrium behavior conjectured by conformal field theory,@xcite which provides a check on the accuracy of the numerical calculations .
we finally study inhomogeneous systems featuring different bulk interactions as well as the long - studied luttinger liquid physics@xcite induced by an impurity at the interface .
we aim at investigating the real - time dynamics of the energy current @xmath11 through a one - dimensional infinite lattice system driven out of equilibrium by an initial sharp temperature gradient @xmath0 .
our main focus is to study the long - time behavior of @xmath11 and specifically the question how it relates to linear - response thermal transport properties . as a prototypical model
, we consider a chain of interacting spin-@xmath6 degrees of freedom @xmath12 governed by local hamiltonians @xmath13 or equivalently spinless fermions through a jordan - wigner transformation . by choosing the couplings @xmath14 , @xmath15 , and
@xmath16 appropriately : @xmath17 we can study systems which are gapless or gapped and as a key aspect of this work investigate the role of integrability . for @xmath18 and @xmath19 , eq .
( [ eq : h ] ) can be diagonalized via bethe ansatz;@xcite the model is nonintegrable otherwise .
the spectrum is gapless for @xmath20 and gapped for @xmath21 .
a gap opens for @xmath22 or @xmath23 , where @xmath24 and @xmath25 only if @xmath26.@xcite in addition , we study the quantum ising model @xmath27 thermal nonequilibrium is introduced via the following protocol : we initially consider two seperate semi - infinite chains ( @xmath28 ) @xmath29 each being in thermal ( grand - canonical ) equilibrium at temperatures @xmath30 and @xmath31 .
the corresponding density matrix factorizes , @xmath32 at time @xmath33 , the chains are coupled through @xmath34 , and the time evolution of @xmath35 is computed w.r.t .
the energy current is defined by a continuity equation,@xcite @xmath37~,\ ] ] and its time evolution is simply given by @xmath38~,\ ] ] which can be computed efficiently using the real - time @xcite finite - temperature @xcite density matrix renormalization group @xcite ( dmrg ) algorithm introduced in ref . .
dmrg is essentially controlled by the so - called discarded weight @xmath39 .
we ensure that @xmath39 is chosen small enough and that @xmath40 is chosen large enough to obtain numerically - exact results ( i.e. , @xmath41 to an accuracy of one percent ) in the thermodynamic limit .
we stop our simulation once the dmrg ` block hilbert space dimension ' has reached values of about 1000 .
chains which are initially in thermal equilibrium at different temperatures @xmath42 and coupled at time @xmath33 and position @xmath43 .
( a ) integrable xxz chain with @xmath44-anisotropy @xmath45 .
the behavior in the gapped phase @xmath21 is similar .
( b , lower inset ) nonintegrable dimerized xxz chain where the coupling on every second bond is reduced by @xmath46 .
the latter is irrelevant at @xmath47 but opens a gap for @xmath48 .
( c ) xxz chain in presence of a staggered field @xmath49 rendering the model nonintegrable .
a gap opens around @xmath50 . despite the fact that the local energy density @xmath51 does not relax ( upper inset ) , the current saturates fast to a unique finite value except for @xmath52 .
we attribute this to a finite linear - response thermal drude weight of both the pure xxz chain and the nonintegrable dimerized chain ( see figure [ fig : drude ] ) . ]
we start by studying a xxz chain with two additional perturbations ( dimerization @xmath53 and a staggered field @xmath52 ) which both render the system nonintegrable.@xcite at time @xmath33 , two semi - infinite chains each being prepared in thermal equilibrium at temperatures @xmath42 are coupled by @xmath34 to an overall translationally - invariant chain .
exemplary results for @xmath11 are shown in figure [ fig : current ] .
the current at the interface @xmath43 saturates on a scale @xmath54 [ note the definition of units via eq .
( [ eq : parahom ] ) ] irrespective of the temperature difference @xmath55 or the absolute values of @xmath42 and regardless of the fact whether or not the system is gapped .
the only exception is @xmath52 where @xmath11 does not reach a finite steady - state value within the time scales accessible by our numerics [ figure [ fig : current](c ) ] , again irrespective of the fact whether or not @xmath49 opens a gap .
we will now try to understand this in more detail .
the time evolution of the local energy density @xmath56 of the xxz chain ( which for a homogeneous system might be a measure for an effective temperature ) is shown in the inset to figure [ fig : current](a ) .
it does not reach a steady - state value but becomes increasingly smooth .
this is not suprising since we are simulating a closed quantum system but gives rise to the questions : ( 1 ) why does the current saturate except for @xmath52 , implying that it is _ not determined by local temperature gradients _ ? ( 2 )
would we obtain the same steady - state current if we kept the ` reservoirs ' at a fixed temperature?@xcite both are reasonable if it does not matter over which length scale @xmath57 the temperature difference @xmath55 is applied ; qualitatively , this should be the case if thermal transport properties of the chain are length - independent , i.e. if the thermal conductance @xmath58 of a finite system does not decrease with its length @xmath57 , or equivalently , if the conductivity @xmath59 of an infinite chain @xmath60 is infinite .
more quantitatively , we conjecture a relation between nonequilibrium and linear response : the nonequilibrium energy current relaxes to a finite steady - state value if the linear - response thermal conductivity is infinite , i.e. if the drude weight @xmath1 is nonzero . before we proceed with calculating @xmath1
, we note that ( 2 ) can be shown explictly for the xx chain @xmath61 , @xmath18 by carrying out the so - called wide - band limit ( which strictly pins the temperatures ) and by computing the current analytically using keldysh green functions ; moreover , the nonequilibrium steady state current was recently obtained from a generalized landauer - buettiker formula.@xcite both currents agree with the one in our setup at long times [ see , e.g. , figure [ fig : steady](b ) ] . ) . for the integrable xxz chain ( @xmath18 , @xmath19 ) ,
the global energy current @xmath62 is conserved ; thus , @xmath63 , and dmrg can be compared with exact bethe ansatz results ( symbol ) .
( a , b ) nonintegrable dimerized chain .
the current correlation function seems to saturate at a finite value ( or decays on a hidden large temperature - independent time scale ) , indicating a finite drude weight @xmath2 .
( c ) the data in presence of a staggered field @xmath52 is consistent with @xmath64 . ]
to support our conjecture we now extract @xmath1 from the long - time behavior of the energy current correlation function,@xcite @xmath65 which can be readily computed using dmrg .
results are shown in figure [ fig : drude ] . for @xmath18 and @xmath19 ,
@xmath3 is conserved , thus @xmath63 ; the drude weight can alternatively be obtained via bethe ansatz.@xcite the energy current correlation functions of the dimerized chain seem to saturate at a finite value for any @xmath66 ; alternatively , they decay on a hidden large time scale which is temperature - independent and becomes _ larger _ as the dimerization is increased from @xmath18 to @xmath67 [ see figure [ fig : drude](a ) and ( b ) ] .
our data thus indicate a nonzero drude weight .
this is interesting on general grounds because the model is nonintegrable.@xcite most previous numerical works on the dimerized chain @xcite yield @xmath64 @xcite but focus on @xmath68 where also our results are less conclusive .
the deeper reason for a potentially finite drude weight the protection of the energy current by an unknown nonlocal conserved operator @xcite will be left as a subject for future work .
in contrast , our data for @xmath52 is consistent with @xmath64 [ see figure [ fig : drude](c ) ] .
note that in both cases it does not seem to play a role whether @xmath53 and @xmath52 are irrelevant [ main panel of figure [ fig : drude](c ) ; inset to ( b ) ] or open a gap [ main panels of ( a ) and ( b ) ; insets to ( a ) and ( c ) ] : if @xmath66 is decreased in a regime where @xmath49 is relevant [ e.g. at @xmath48 ; see the inset to figure [ fig : drude](c ) for @xmath69 , the scale on which @xmath70 decays becomes _ successively _ larger ; for temperatures smaller than the gap it can no longer be reached by our numerics . the behavior of the dimerized chain for parameters where @xmath46 is relevant is completely different : if @xmath70 decays on a hidden large scale , the latter is temperature - independent and does not manifest even at @xmath71 [ compare figure [ fig : drude](a ) and ( b ) ] . recalling that the nonequilibrium current relaxes to a nonzero steady - state value in all cases where @xmath19 , the observation of a finite ( vanishing ) drude weight for @xmath72 ( @xmath52 ) supports our above conjecture .
we now turn to study the temperature - dependence of the steady - state ( position - independent ) current .
the result for the xxz chain both in the gapless and gapped regime is illustrated in figure [ fig : steady ] .
the asymptotic current seems to be of a strikingly simple functional form : @xmath73 indicating a second relation between nonequilibrium and linear response : the linear thermal conductance @xmath74 determines the steady - state nonequilibrium current at any @xmath55 . equation ( [ eq : f ] ) can be established by varying @xmath31 at fixed @xmath30 ; the corresponding curves collapse if shifted vertically [ see figure [ fig : steady](b ) ] . the limiting behavior ( both in the gapless and gapped regime ) of @xmath75 is given by @xmath76 other details of @xmath5 such as prefactors or the crossover scale ( which for @xmath77 and @xmath78 is determined by the bandwidth or the size of the gap , respectively ) in general depend on the model parameters .
however , a recent conformal field theory approach@xcite conjectures that the low - temperature behavior of a gapless system is universally given by @xmath79 with @xmath80 being the cft central charge , which follows intuitively from the version of the stefan - boltzmann law satisfied by a cft.@xcite we confirm this prediction for the xxz chain ( @xmath81 ) as well as the quantum ising model ( @xmath82 ) ; this is illustrated in the insets to figure [ fig : steady](b ) .
this is a nontrivial result because : ( 1 ) it is unclear why for a microscopic model whose equilibrium physics is governed by a certain low - energy field theory the very same field theory should describe the long - time behavior of the microscopic model in nonequilibrium ( note that the behavior for @xmath83 is not captured by the cft ! ) , and ( 2 ) even linear - response transport properties ( such as the drude weight ) are not determined by the low - energy theory alone but by a delicate interplay of conserved quantities protecting the current and dangerously irrelevant operators scattering it.@xcite our results for the dimerized chain at @xmath84 are still consistent with eq .
( [ eq : f ] ) , indicating that it might be a universal property of any system with a thermal drude weight , if indeed that system has a drude weight . at smaller @xmath46 and low @xmath66
, we can not reach time scales where oscillations of the current have died out completely .
we expect that models which are strongly nonintegrable and have zero drude weight will not show a steady state even in the homogeneous case .
@xcite the free fermion case @xmath77 can be solved exactly;@xcite eq .
( [ eq : f ] ) reflects a noninteracting thermal landauer - bttiker formula .
this analytic result can be used to test our dmrg numerics at any temperature [ see the comparison in figure [ fig : steady](b ) as well as in the inset to figure [ fig : inhom ] ] .
two ( effectively semi - infinite ) reservoirs featuring different temperatures @xmath30 and @xmath31 are connected , and eventually a steady - state energy current is established .
( b ) gedanken experiment with three reservoirs .
as described in the text , conservation of the energy current implies that there is a relationship between the currents flowing at the three interfaces , until such times as the different interfaces begin to interact with each other . ]
( b ) . at @xmath33 , three xxz chains ( @xmath48 ) of lengths @xmath85 and temperatures @xmath86 , @xmath87 , @xmath88
are coupled to a ring .
the energy currents saturate to their unique steady - state values at the interfaces ( @xmath89 ) but remain zero far away from them ( e.g. , for @xmath90 ) .
the sum of the current over the regions where it is nonzero vanishes [ see eq .
[ eq : currentsum ] ] due to the conservation of the total current .
as explained in the main text , this motivates the `` cyclic sum rule '' of eq .
( [ eq : cyclicform ] ) , which is equivalent to the existence of a stefan - boltzmann function @xmath5 . ] eq .
( [ eq : f ] ) can alternately be understood as a cyclic relation for the 3-reservoir geometry in figure [ fig : geometry](b ) : the existence of a stefan - boltzmann function @xmath5 is equivalent to the statement that the steady - state currents between three reservoirs @xmath91 satisfy @xmath92 we now show that integrable models with a conserved total energy current , such as the xxz model , have a cyclic `` sum rule '' structure which is loosely similar to but not ( at first glance ) equivalent to eq .
( [ eq : cyclicform ] ) .
further physically motivated assumptions then lead to the existence of the stefan - boltzmann function . to understand why there is any relationship between the three pairs of temperatures in the cyclic formula eq .
( [ eq : cyclicform ] ) , consider the initial condition shown in fig .
[ fig : geometry](b ) .
three segments of equal lengths @xmath40 of a ring are prepared at three different temperatures .
this system can be studied straightforwardly via dmrg , and examplary results are shown in figure [ fig : ring ] .
they can be interpreted as follows .
suppose that the typical velocity of the system is @xmath93 , and consider energy currents at a time @xmath94 .
for any such time , the middle of each reservoir is essentially unperturbed from its initial state , so the local energy current is zero ( see the curve at @xmath90 ) .
the local energy currents rise in the vicinity of the junctions ( @xmath89 ) ; let @xmath95 be the _ spatially integrated _ energy current at the @xmath96 to @xmath97 boundary at time @xmath98 , and similarly for the other boundaries .
of course once @xmath99 the energy currents are no longer localized in the region of the boundaries and @xmath100 is not defined .
the initial energy current around the ring is zero and is a conserved quantity of the connected system , so the total energy current must still be zero .
in other words , for all times @xmath101 we have ( see figure [ fig : ring ] ) @xmath102 this is not the same as eq .
( [ eq : cyclicform ] ) , however , because @xmath3 is the current at a point , while @xmath103 is spatially integrated and has some complicated profile in general . for the conformal field theory limit ,
all excitations have a single velocity @xmath104 and @xmath103 is expected to be dominated by @xmath105 . to summarize
, general principles suggest that there is a sum rule ( at all times less than an upper cutoff determined by the reservoir size ) relating the three pairs of temperatures ; however , the detailed form of this sum rule is not the same as the observed relationship between steady - state currents , although they are related in the conformal limit .
now we proceed to give a definition of the stefan - boltzmann function @xmath5 and then argue that the steady - state current is determined by the difference in the stefan - boltzmann functions of the two reservoirs . consider a large reservoir of size @xmath106 prepared at initial temperature @xmath66 . at @xmath33 , it is connected to semi - infinite leads at each end , and these leads are prepared at some reference temperature , say 0 ( ignoring any subtleties from possible symmetry breaking ) . at some very long time
so that no excitations remain in the reservoir , let @xmath107 be the total energy current integrated over all sites to the right of the reservoir .
this `` right - moving energy current '' will be extensive in @xmath106 , so we define @xmath108 as the radiated right - moving energy current per site . in order to give some intuition for the existence of a steady state and stefan - boltzmann law even when there is not a single velocity as in the conformal case , we present a free - particle example in appendix a. this model has a gaussian distribution of velocities and a steady - state energy current at the reservoir boundary that is determined solely by the total radiated power as described above .
the variation in particle velocity does not otherwise affect the answer .
the steady state lives for a time that is arbitrarily long as the reservoir size becomes infinite , but is short compared to the amount of time it takes for the reservoir to radiate all its energy , as at that time there is no current left at the reservoir boundaries .
this notion of the steady state as actually a long - time phenomenon compared to transients ( and diverging in the limit that the reservoir is infinite ) , but a short - time phenomenon compared to the time on which @xmath107 is defined , is used in our argument for the interacting case below . now comes a subtle and surprising point .
the cyclic relation at finite times in ( [ eq : currentsum ] ) means that the behavior of the energy current at any of the junctions is remarkably constrained : any change that only affects one junction but not the other two will not affect the spatially integrated current around the junction .
for example , consider the two different initial conditions shown in fig .
[ fig : spacer]a : one is the abrupt junction simulated in our numerics , where a single bond is restored at @xmath33 ; the other contains a `` spacer '' of @xmath109 bonds , all of which are turned on at @xmath33 .
the total energy current integrated over bonds between these two reservoirs must be exactly the same in these two cases , up to some long time ( we assume that the reservoir size @xmath110 . )
but for a long spacer @xmath111 , the spatial profile of the current will look different at short times than for @xmath112 : for the long spacer , the current consists of one right - moving patch from the left lead and one left - moving patch from the right lead , as shown schematically in fig .
[ fig : spacer]b . for @xmath112 , the current is expected to be spatially monotonic , or at least not to decouple into these separated regions .
is different between the geometries with and without a `` spacer '' region prepared at some reference temperature , but the spatial integrals are the same .
( b ) two large reservoirs of size @xmath57 can be viewed as made up of translated pairs of smaller sub - reservoirs of size @xmath106 .
the numbers indicate the pairings of the sub - reservoirs .
the sum of many translated copies of a function approaches a constant , as long as the function is smoothly varying on the scale of the translations .
if the energy current can indeed be viewed as such a sum , then the existence of a stefan - boltzmann function is recovered , independent of the precise shape of the energy current distribution . ]
this argument indicates that the nontrivial time evolution of spatially integrated current is exactly the same whether the right - movers and left - movers pass through each other or are spatially separated , i.e. , treating the problem as two separate pulses and adding them together gives exactly the right total current .
we now give a ( non - rigorous ) picture for how a steady state described by a stefan - boltzmann function can exist independent of the `` pulse shape '' , i.e. , the spatial distribution of energy current , which is certainly sensitive to interactions .
consider each large reservoir , now of size @xmath57 , as made up of many contiguous sub - reservoirs of size @xmath106 ( fig .
[ fig : spacer]b ) .
we can pair sub - reservoirs as shown so that the total problem is a combination of copies translated by @xmath106 .
the point of the small sub - reservoirs is that the observation time for the steady state is long enough that the sub - reservoirs are in the long - time limit where @xmath107 is defined , although the total reservoir is not .
the shape of these contributions could be modified by being in a non - trivial background , but their total weight is not , and as we are at long times , we expect the pulse shape to be broad compared to @xmath106 . now whatever the shape of the contribution from a pair of sub - reservoirs is , its total weight is @xmath113 . when we add together many translated copies of the same shape , we obtain a constant , since by the poisson summation formula @xmath114 for functions @xmath115 that do not vary strongly on the scale of @xmath116 .
so this constant is exactly @xmath117 , the desired result .
the conformal limit is a case where the subreservoir size can be taken to be arbitrarily small . in words ,
the steady state exists in a time window where all that matters is the total energy per length of right - movers emitted from the left , less that of left - movers emitted from the right . ) to absorb half of the excitations from each initial reservoir ; which half are absorbed ( right - movers or left - movers ) depends on the reservoir s temperature .
the absorbers are then removed and further evolution is unitary . energy current is conserved and , if the dispersion of velocities leads to a translation - invariant final state , that final state must have energy current given by the stefan - boltzmann prediction . ]
we can give a specific example for which this picture is correct by constructing a geometry , different from the original geometry of two semi - infinite leads , for which the time evolution can be shown to give a steady - state described by the stefan - boltzmann relation , even in the presence of interactions .
this sidesteps the difficulty of solving the time evolution from the non - translation - invariant initial state
it is difficult to prove in general even that a translation - invariant steady state exists .
we would like to make a translation - invariant system of interacting particles whose final state has left - movers originating at a different temperature from right - movers . to do so , consider the process in fig .
[ fig : protocol]a .
suppose for simplicity that the model has a finite range of particle velocities and that there is a `` vacuum '' state with no particles .
two regions of different temperatures are spaced by a large enough region of vacuum that the left - movers from the right region fail to interact with the left - movers from the other until both types of particles have moved out of their original reservoir regions .
particle absorbers , which could be extra lengths of spin chain for example , are placed at the left and right ends of the system and absorb left - movers from the left region and right - movers from the left region .
after this absorption , the absorbers are removed and the remaining evolution is unitary .
repeat this arrangement in a discretely translation - invariant way , as in fig .
[ fig : protocol]b .
the further time evolution will preserve the total energy current , even though the particles coming from different reservoirs will now interact . assuming that a homogeneous energy current is reached in the final state because of the dispersion of particle velocities , the energy current in this final state has to be @xmath118 $ ] , where @xmath80 is the volume fraction occupied by the reservoirs and @xmath5 is the directed radiated energy current per unit length .
in other words , the conservation of energy current means that , in this specific example where left - movers and right - movers are drawn from different temperatures , the final state is described by a stefan - boltzmann function .
however , the existence of the same energy current in the actual two - reservoir geometry is so far more difficult to establish .
while particle velocities are certainly modified by the density of particles from the other reservoir in an interacting system , this modification need not alter the total energy current at the boundary , as in the example above .
a quantum field theory approach to thermal steady states leads in some cases to a factorization of the density matrix from which the existence of a stefan - boltzmann function follows , although integrability and conserved energy current seem to play less of a role in this approach @xcite .
it should be noted that applying this field - theory approach to the steady - state energy current in the massive sine - gordon model gives behavior that slightly violates the existence of a stefan - boltzmann function @xcite , at the level of a few percent .
it is not clear whether this discrepancy indicates a fundamental difference in the type of steady state ( as , analytically , solving the time evolution for all times to see the unique steady state emerge is not possible in either approach ) .
finally , we note that the same logic presented in this section would apply to charge currents in a homogeneous system with a conserved charge current , but the xxz model s charge current is not conserved ( does not commute with the hamiltonian ) , although there is a drude weight .
within the accuracy of our numerics , the steady state current does not depend jointly on @xmath30 and @xmath31 but rather is just the difference of the function @xmath5 evaluated at the two temperatures .
it is known that at low @xmath66 this property should hold as a result of conformal invariance,@xcite with @xmath119 . earlier in this paper
we gave a somewhat lengthy analytic argument suggesting this result ; in order to make it less mysterious , we now show that a relatively simple approximation for the isotropic xxx ferromagnet ( @xmath120 ) gives a reasonable description of @xmath75 at all temperatures and is exact in the low - temperature limit . ) of two estimates for the function @xmath5 whose differences give the steady - state energy current @xmath3 via @xmath121 between reservoirs @xmath30 and @xmath31 .
one estimate is obtained from the dmrg results , using reference temperatures .
the other is obtained from the numerical solution of the thermodynamic bethe ansatz equations and the formula eq .
( [ eq : tbaresult ] ) . ] for a spatially uniform gas of free particles with a particle distribution function @xmath122 in velocity space , the stefan - boltzmann function @xmath123 can be obtained as @xmath124 where @xmath125 is the energy of a particle with velocity @xmath104 and the integral is over only right - moving particles .
this has the units of an energy current ( an energy per time ) if @xmath126 has units of particle number per length per velocity .
( [ eq : f_free ] ) motivates the following approximation : the thermodynamic bethe ansatz solution of the xxx model by takahashi@xcite is in terms of particles named `` magnons '' and `` strings '' . in general
, these particles are not independent because of the bethe conditions on phase shifts . however , at low temperature and for ferromagnetic interactions , the thermodynamic state is a dilute gas of magnons and strings , and we might hope that a formula similar to eq .
( [ eq : f_free ] ) , generalized to multiple kinds of particles , is a good starting point . , @xmath19 but different bulk interactions @xmath127 and different couplings @xmath128 to tune the backscattering at the interface to zero.@xcite the magnitude of the current is the same for @xmath129 and @xmath130 .
the curves at different @xmath97 collapse ( left inset ) , indicating that eq .
( [ eq : f ] ) still holds .
right inset : linear thermal conductance of two identical xxz chains connected by a strong barrier @xmath131 .
the low-@xmath66 behavior is consistent with @xmath132 where @xmath133 is the luttinger liquid parameter.@xcite at @xmath77 we repoduce the exact result of ref .
for any @xmath66 . ] in the notation of refs . , which introduces a variable @xmath134 that parametrizes the momentum @xmath135 via @xmath136 the needed quantity is @xmath137 where @xmath109 ranges over the different types of excitations ( magnons and strings ) and @xmath138 are the density , velocity , and energy .
note that there is an ambiguous additive constant in @xmath5 .
we evaluate only for right - moving excitations based on a landauer - type picture where the steady - state can be viewed as the combination of right - movers from the left lead and left - movers from the right lead .
the velocity @xmath139 was obtained as @xmath140 and the energy for an excitation of total momentum @xmath133 is @xmath141 we have solved the standard tba equations for @xmath142 using the numerical method introduced of schlottman.@xcite the result of evaluating this form for @xmath5 is surprisingly good for the ferromagnetic case : it is correct at low temperature and underestimates the correct value by about 10% at high temperature ( see fig .
[ fig : bethe ] ) . for the xxx antiferromagnet ,
the result is much worse and fails to reproduce the cft result at low temperature , which is natural as the low - temperature state is now not dilute in terms of these particles .
however , the agreement is improved at low temperature by using a group velocity derived from the dressed excitation energy .
a more complete comparison to this approach , including other values of the anisotropy parameter , is currently underway .
a full explanation of our numerical results from the bethe ansatz is an open problem , and just taking the right - movers ( as done here ) does not satisfy the bethe conditions on the phase shifts of the particles . nevertheless , the quantitative agreement between the estimate from eq .
( [ eq : tbaresult ] ) and the dmrg result means that the steady - state current is close in the ferromagnet to a free - particle interpretation although the particles and their densities are rather complicated .
we finally investigate systems which are not translationally invariant .
tunneling across a barrier between luttinger liquids has been well studied by bosonization and other field - theoretic methods in the low - energy limit , and beyond just verifying these predictions , numerics allow a determination of when the asymptotic properties accessible by bosonization become apparent .
we start by studying the effects of different bulk parameters @xmath143 , @xmath144 , @xmath145 , @xmath146 ( and @xmath147 ) .
if these parameters are chosen such that the renormalized fermi velocities in the left and right halves coincide , backscattering due to the barrier ( which is naturally present at the interface @xmath43 @xcite ) can be tuned to zero @xcite , and luttinger liquid ( ll ) boundary physics @xcite is absent .
results are displayed in figure [ fig : inhom ] , indicating that eq .
( [ eq : f ] ) still holds . if two homogeneous xxz chains @xmath148 are connected through a barrier @xmath149 , the linear thermal conductance @xmath150 is expected to feature a low-@xmath66 power law @xmath132 with @xmath133 being the ll parameter.@xcite for @xmath77 , eq .
( [ eq : f ] ) still holds even if @xmath151,@xcite and @xmath152 just reflects the asymptotic @xmath153 behavior of @xmath5 .
our data for @xmath58 in presence of interactions is consistent with @xmath132 .
this is illustrated in the inset to figure [ fig : inhom ] where @xmath149 is chosen small so that the scale on which ll boundary effects manifests becomes large.@xcite eq .
( [ eq : f ] ) still holds approximately above this scale .
there are obviously many possible bulk and tunneling parameters that can be studied , and we reserve a comprehensive study for future work ; the main point is to note that the luttinger liquid tunneling physics and other subtle properties can be accessed via our approach .
in this paper we provided evidence for a connection between nonequilibrium and linear - response thermal transport properties of isolated infinite spin chains ( or equivalently , interacting spinless fermions ) : ( 1 ) the energy current of a system which initially features a temperature gradient @xmath0 saturates to a finite value if the equilibrium thermal drude weight @xmath1 is finite , and ( 2 ) the value of the steady - state current at arbitrary @xmath42 is of the functional form @xmath4 , i.e. it is completely determined by the linear thermal conductance .
this can be viewed as a generalized stefan - boltzmann law describing freely moving quasiparticles ; for the xxx ferromagnet , @xmath5 can be computed via thermodynamic bethe ansatz in good agreement with the numerics .
our data suggests that @xmath2 for a nonintegrable dimerized chain ( or that the current correlation function decays on a hidden large temperature - independent time scale ) .
_ acknowledgments _ we are indebted to e. altman , b. doyon , f. essler , f. heidrich - meisner , v. meden , k. schnhammer , and d. schuricht for fruitful discussions and comments and acknowledge support by the deutsche forschungsgemeinschaft via ka3360 - 1/1 ( c.k . ) as well as by the afosr muri on `` control of thermal and electrical transport '' ( r.i . ) , the nanostructured thermoelectrics program of lbnl ( j.e.m . and c.k . ) , and the simons foundation ( j.e.m . ) .
we would like to understand in a simple example how the steady - state current arises outside the conformal limit , i.e. , when particle velocities are variable . consider a system of classical non - interacting particles that at time @xmath33 has the maxwellian distribution @xmath154 , \\ 0\ { \rm otherwise}. \end{cases}\ ] ] the boltzmann equation contains only the streaming term , with the result that the function @xmath155 is simply equal to @xmath156 : the number of particles with a given velocity @xmath104 at the point @xmath134 and time @xmath98 is given by the number of particles with that velocity at spatial point @xmath157 at time 0 .
we would like to compute the energy current @xmath158 the integral over the initial distribution will contribute if @xmath159 $ ] , which means @xmath160.\ ] ] assuming @xmath161 , the energy current is @xmath162_{(-l/2 + x ) / t}^{(l/2 + x ) / t}. \label{energycurrent}\end{aligned}\ ] ] the stefan - boltzmann function @xmath5 as defined above is given by the total right - moving energy current per length , or @xmath163 now the question is whether ( [ energycurrent ] ) , evaluated at the right edge of the reservoir , is equal to @xmath5 for some period .
we have @xmath164.\ ] ] we see that this is indeed equal to @xmath5 while @xmath165 and that the steady state described by @xmath5 persists forever if @xmath166 .
b. s. shastry and b. sutherland , phys .
lett . * 65 * , 243 ( 1990 ) ; x. zotos , phys .
lett . * 82 * , 1764 ( 1999 ) ; j. benz , t. fukui , a. klmper , and c. scheeren , j. phys .
. suppl . * 74 * , 181 ( 2005 ) . j. v. alvarez and c. gros , phys .
* 89 * , 156603 ( 2002 ) ; f. heidrich - meisner , a. honecker , d. c. cabra , and w. brenig , phys .
* 92 * , 069703 ( 2004 ) ; j. v. alvarez and c. gros , phys .
lett . * 92 * , 069704 ( 2004 ) .
g. vidal , phys .
lett . * 93 * , 040502 ( 2004 ) ; s. r. white and a. e. feiguin , phys .
lett . * 93 * , 076401 ( 2004 ) ; a. daley , c. kollath , u. schollwck , and g. vidal , j. stat .
mech . : theory exp .
p04005 ( 2004 ) ; p. schmitteckert , phys .
b * 70 * , 121302(r ) ( 2004 ) . | we study the real - time dynamics of spin chains driven out of thermal equilibrium by an initial temperature gradient @xmath0 using density matrix renormalization group methods .
we demonstrate that the nonequilibrium energy current saturates fast to a finite value if the linear - response thermal conductivity is infinite , i.e. if the drude weight @xmath1 is nonzero .
our data suggests that a nonintegrable dimerized chain might support such dissipationless transport ( @xmath2 ) .
we show that the steady - state value @xmath3 of the current for arbitrary @xmath0 is of the functional form @xmath4 , i.e. it is completely determined by the linear conductance .
we argue for this functional form , which is essentially a stefan - boltzmann law in this integrable model ; for the xxx ferromagnet , @xmath5 can be computed via thermodynamic bethe ansatz in good agreement with the numerics .
inhomogeneous systems exhibiting different bulk parameters as well as luttinger liquid boundary physics induced by single impurities are discussed briefly . |
homoclinic tangencies , i.e. tangencies between stable and unstable invariant manifolds of the same saddle periodic orbit , play a very important role in the theory of dynamical chaos . starting with the works by smale and shilnikov @xcite ( and in the ideological sense as far back as with the memoirs by poincar and birkhoff @xcite ) ,
the existence of transversal homoclinic intersections is considered as the universal criterium of the complexity of dynamical systems . at the same time , the presence of non - transversal homoclinic orbits ( homoclinic tangencies ) indicates an extraordinary richness of bifurcations of such systems and , what is very important , the principal impossibility of providing of a complete description of bifurcations of such systems within framework of finite parameter families , @xcite .
therefore , when studying homoclinic bifurcations , the main problems are related to the analysis of their principal bifurcations and characteristic properties of dynamics as a whole .
one of the important properties of homoclinic tangencies is that they can freely produce homoclinic and heteroclinic tangencies of higher orders @xcite . in particular , in a two parameter family which unfolds generally a quadratic homoclinic tangency , cubic homoclinic tangencies appear unavoidably .
this means that the study of bifurcations of cubic homoclinic tangencies itself becomes an important problem .
moreover , cubic homoclinic tangencies can play an independent role when studying global bifurcations in many dynamical models . for example , such tangencies appear naturally in the problem on periodically perturbed two - dimensional flows with a homoclinic figure - eight of a saddle equilibrium @xcite , where the corresponding bifurcation points of codimension two give rise to the curves of quadratic homoclinic tangencies forming ( non - smooth ) boundaries of homoclinic zones .
moreover , bifurcations of cubic tangencies lead to the appearance of specific windows of stability ( i.e. parameter domains corresponding to the existence of stable periodic orbits ) which are well observable during numerical explorations , @xcite , unlike the windows produced by quadratic homoclinic tangencies .
in fact , bifurcations of cubic homoclinic tangencies were studied first in @xcite for the sectionally dissipative case ( when a saddle periodic orbit has eigenvalues @xmath0 such that @xmath1 and @xmath2 ) , see also @xcite .
note that in @xcite the two main cases were considered when the leading stable eigenvalue is real ( i.e. @xmath3 ) and the two leading eigenvalues are complex conjugate ( i.e. @xmath4 , @xmath5 , and @xmath6 ) .
the main attention in @xcite was focused on the study of bifurcations of _ single - round periodic orbits _ ,
i.e. periodic orbits which pass a neighborhood of the homoclinic orbit only once .
we note that the study of such orbits can be reduced to an analysis of fixed points of the corresponding first return maps defined near some point of the homoclinic tangency . in the case of cubic tangencies
, such maps can be brought to the maps close to the two - dimensional cubic hnon maps of the form @xmath7 being @xmath8 and @xmath9 parameters and @xmath10 the jacobian
. moreover , map @xmath11 appears in the case of cubic tangency incoming from below ( figure [ fig2type](a ) ) and map @xmath12 arises in the case of cubic tangency incoming from above ( figure [ fig2type](b ) ) . in the dissipative case @xmath13 , maps ( [ cubhm ] ) with @xmath14 were derived ( using the rescaling method , @xcite ) in @xcite , where the corresponding formulas for the rescaled first return maps near homoclinic tangencies of arbitrary finite orders were obtained .
it is worth mentioning that maps @xmath12 and @xmath11 were studied in @xcite for the dissipative case @xmath15 , where bifurcation diagrams for fixed points and 2-periodic orbits were constructed analytically as well as strange attractors were analyzed numerically .
also maps ( [ cubhm ] ) are good approximations of the corresponding rescaling first return maps @xmath16 when the return time @xmath17 is sufficiently large ( in fact , @xmath17 is the period of a single - round periodic orbit under consideration ) .
then @xmath18 and @xmath19 are rescaled coordinates and parameters , respectively , which are defined in domains @xmath20 and @xmath21 , where @xmath22 as @xmath23 , the jacobian @xmath10 is proportional to @xmath24 ( i.e. @xmath14 in the dissipative case and @xmath25 in the conservative case ) , and the error terms are of order @xmath26 . in the present paper
we study bifurcations of single - round periodic orbits in the case of two - dimensional symplectic maps with cubic homoclinic tangencies .
we put emphasis on bifurcations leading to the birth of elliptic periodic orbits .
the main results can be briefly formulated as follows let @xmath27 , @xmath28 , be a two parameter family of symplectic maps unfolding generally the cubic homoclinic tangency to the saddle fixed point @xmath29 with eigenvalues @xmath30 and @xmath31 , where @xmath32 .
let @xmath33 be a small neighborhood of the origin in the @xmath34-plane .
then the following holds 1 . in @xmath33 , there exists a bifurcation curve ( semicubical parabola ) @xmath35 with the cusp - point @xmath36 such that for @xmath37 , map @xmath27 has a quadratic homoclinic tangency branched from the initial cubic tangency .
2 . in @xmath33
, there exists an infinite sequence of elliptic zones @xmath38 accumulated to @xmath35 as @xmath23 such that for @xmath39 , map @xmath27 has an elliptic single - round orbit of period @xmath17 .
the boundary of @xmath38 consists of curves @xmath40 and @xmath41 corresponding to the appearance of single - round periodic orbits with double eigenvalues @xmath42 and @xmath43 , respectively .
the elliptic orbit is generic for all values of @xmath44 , except for those belonging to the strong 1:3 and 1:4 resonance curves @xmath45 and @xmath46 ( the orbit has eigenvalues @xmath47 and @xmath48 ) and the non - twist curve @xmath49 ( the first birkhoff coefficient equals zero ) .
[ thm : main ] an illustration to theorem [ thm : main ] is shown in figure [ fig : bdiagr2 ] .
we see that the bifurcation diagrams for single - round periodic orbits are different in the cases ( a ) @xmath50 and the tangency is incoming from below ; ( b ) @xmath50 and the tangency is incoming from above ; ( c ) @xmath51 .
correspondingly , we label zones @xmath38 as @xmath52 ( squids ) in the case ( a ) and @xmath53 ( cockroaches ) in the case ( b ) .
notice that in the case @xmath51 zones @xmath38 of different types alternate depending on the parity of @xmath17 , and , therefore , for more definiteness , we enumerate these zones as @xmath54 .
-plane , where curve @xmath35 and elliptic zones @xmath52 ( a ) and @xmath55 ( b ) are shown .
these zones alternate for @xmath51 ( c ) .
curves @xmath40 correspond to nondegenerate conservative fold bifurcations except for cusp - points @xmath56 of conservative pitch - fork bifurcations .
curves @xmath41 correspond to nondegenerate conservative period doubling bifurcations except for points @xmath57 and @xmath58 of curve @xmath41 of zone @xmath53 where the first lyapunov value @xmath59 vanishes . points @xmath57 and @xmath58 are the endpoints of bifurcation curves @xmath60 and @xmath61 that correspond to conservative fold bifurcations of double - round periodic orbits ( 2-periodic orbits in the first return maps).,height=264 ] we prove theorem [ thm : main ] at the end of section [ sec : stateprob ] ( item 1 ) and section [ sect : bifcub ] ( item 2 ) .
the proof of item 2 of theorem [ thm : main ] is mainly based on the rescaling results of lemma [ henmain - c ] which show that the corresponding first return map @xmath16 can be brought , by linear changes of coordinates and parameters , to a regular map asymptotically close as @xmath23 to the conservative cubic hnon map of the form ( [ cubhm ] ) with @xmath25 . moreover , we obtain map @xmath62 in the case of incoming from below tangency and map @xmath63 in the case of incoming from above tangency ( in the case @xmath51 , maps @xmath62 and @xmath63 appear , respectively , for even and odd numbers @xmath17 )
. bifurcations of fixed points and accompanied bifurcations of 2-periodic orbits for these conservative cubic hnon maps are described in section [ sect : cubichenon ] .
the corresponding bifurcation diagrams are shown in figures [ gk88-+2 ] and [ gk882 ] for the maps @xmath62 and @xmath63 , respectively .
we also give the analytical expressions for the corresponding bifurcation curves , see formulas ( [ eq : criv ] ) .
after this , the proof of theorem [ thm : main ] becomes quite straightforward , see section [ sec : endofproof ] .
note that in the symplectic case , one of the main questions when studying elliptic points is related to their generic properties . in the case of the conservative hnon map @xmath64
, the problem of genericity ( stability ) of elliptic fixed points was completely solved in @xcite , see also @xcite .
concerning the conservative cubic hnon maps , the problem of genericity ( stability ) of fixed elliptic points has been ( almost completely ) solved in @xcite , where conditions for the kam - stability of these points were obtained .
note also that bifurcations of lower - periodic orbits ( of period 1 , 2 , 3 ) were also studied in @xcite . concerning 4-periodic orbits , in section [ sec:14rez ] we carry out the corresponding bifurcation analysis paying main attention to those phenomena which are connected with the peculiarities of bifurcations of fixed points with eigenvalues @xmath65 , i.e. 1:4 resonance .
we perform this study separately for map @xmath62 in section [ sec:14rez+ ] and for map @xmath63 in section [ sec:14rez- ] .
note that the main complex local normal form of a map having a fixed point with eigenvalues @xmath48 can be written as follows , @xcite , @xmath66 where @xmath67 is a parameter characterizing deviation of the angle argument @xmath68 of eigenvalues of the fixed point from @xmath69 ( @xmath70 ) , the coefficients @xmath71 and @xmath72 are real and smoothly depend on @xmath67 .
then the following conditions , @xcite , @xmath73 imply that bifurcations of the fixed point in ( [ hecomplnew13 ] ) are nondegenerate .
we show that conditions ( [ d21d03 ] ) can be violated for both maps @xmath62 and @xmath63 .
namely , the case @xmath74 occurs for map @xmath62 and the equality @xmath75 can hold for map @xmath63 .
this means that the structure of 1:4 resonance is different in these maps and we study the accompanied bifurcation phenomena in the present paper .
as far as we know , the case @xmath76 , when the fixed point in the normal form ( [ hecomplnew13 ] ) is a nonlinear center ( map ( [ hecomplnew13 ] ) without @xmath77 terms is a nonlinear rotation ) was not considered before .
nevertheless , this case is very interesting from various points of view . in particular
, the problem of the existence of quasi - central elliptic points in reversible analytical maps is very important . here , bifurcations of such points ( when the map is not conservative ) lead to the birth of pairs of periodic orbits sink - source and , as a consequence , to the break - down of conservativity and appearance of the so - called mixed dynamics @xcite even inside symmetric elliptic islands @xcite .
[ rem01 ] the case @xmath75 of 1:4 resonance was considered in @xcite for the conservative hnon map , see also @xcite . however , for the cubic hnon map @xmath63 , the case @xmath75 is of another ( more complicated ) structure .
in particular , proper bifurcations of a fixed point with eigenvalues @xmath48 are replaced here by extrinsic bifurcations of some 4-periodic orbits ( thus , 1:4 resonance demonstrates certain non - twist properties ) .
[ rem02 ] in the present paper we only study the most interesting and principal peculiarities of 1:4 resonance in maps @xmath62 and @xmath63 .
in addition to the analytical considerations , we also display certain numerical results in figures [ pi2plusres ] and [ fig_pi2-a1 ] .
we plan to continue to study this in more detail in the nearest future .
the paper is organized as follows . in section [ sec : stateprob ] we start with the statement of the problem and also present the results on the normal forms of symplectic saddle maps ( lemmas [ lem : nfgen ] and [ lem : tknfgen ] ) and the normal form of global map @xmath78 defined near a homoclinic point ( lemma [ nft1 ] ) .
we also prove item 1 of theorem [ thm : main ] in section [ sec : stateprob ] , see proposition [ th1g85 ] .
in section [ sect : bifcub ] we complete the proof of theorem [ thm : main ] .
the main technical result there is the rescaling lemma , lemma [ henmain - c ] , which shows that the first return maps can be written , in some rescaling coordinates , as maps asymptotically close to the conservative cubic hnon maps @xmath62 and @xmath63 .
once we know the structure of bifurcations of fixed points in these maps we can easily recover bifurcations of single - round periodic orbits and , hence , prove the theorem .
in section [ sec:14rez ] we consider the problem on the structure of 1:4 resonance in maps @xmath62 and @xmath63 .
let @xmath79 be a @xmath80-smooth , @xmath81 , two - dimensional symplectic diffeomorphism which satisfies the following conditions : 1 .
@xmath79 has a saddle fixed point @xmath29 with eigenvalues @xmath30 and @xmath31 , where @xmath32 ; 2 . the invariant manifolds @xmath82 and
@xmath83 have a cubic homoclinic tangency at the points of some homoclinic orbit @xmath84 .
we distinguish two types of cubic tangencies of @xmath82 and @xmath83 at a homoclinic point : the tangency of first type ( incoming from above ) and the tangency of the second type ( incoming from below ) .
both these types are shown in figure [ fig2type ] .
when the eigenvalue @xmath30 is positive the type of tangency remains the same for all points of the homoclinic orbit .
however , if @xmath30 is negative the incoming from above and incoming from below tangencies alternate from point to point .
let @xmath85 be a ( codimension two ) bifurcation surface composed of symplectic @xmath80-maps close to @xmath79 and such that every map of @xmath85 has a close to @xmath84 homoclinic orbit at whose points the manifolds @xmath82 and @xmath83 have a cubic tangency .
let @xmath86 be a family of symplectic @xmath80-maps that contains map @xmath79 at @xmath87 .
we suppose that the family depends smoothly on parameter @xmath88 and satisfies the following condition : * the family @xmath86 is transverse to @xmath85 at @xmath89 .
it is natural to assume that condition * c * holds for a two parameter family @xmath90 , where @xmath91 and @xmath92 and @xmath93 are some parameters which split generally the initial cubic tangency of the manifolds @xmath82 and @xmath83 at some homoclinic point .
we consider this question in more detail when proving lemma [ nft1 ] and we deduce that , without loss of generality , one can assume that @xmath94 .
let @xmath95 be a sufficiently small fixed neighborhood of @xmath96 .
it consists of a small disk @xmath97 containing @xmath29 and a number of small disks @xmath98 surrounding those points of @xmath84 that do not lie in @xmath97 ( see figure [ fg : fabhab](a ) ) .
we call a periodic or homoclinic to @xmath29 orbit entirely lying in @xmath95 to be _ @xmath99-round _ if it has exactly @xmath99 intersection points with every disk @xmath100 . for @xmath101 and @xmath102
, we use the terms _ single - round _ and _ double - round _ orbits , respectively . .
some of these homoclinic points are shown as grey circles .
also a small neighborhood of the set @xmath96 is shown as the union of the squares ; ( b ) construction of first return map @xmath103 . , title="fig:",scaledwidth=40.0% ] . some of these homoclinic points are shown as grey circles .
also a small neighborhood of the set @xmath96 is shown as the union of the squares ; ( b ) construction of first return map @xmath103 . , title="fig:",scaledwidth=30.0% ] in the present paper we study bifurcations of single - round periodic orbits in the families @xmath90 ( we keep in mind that @xmath91 ) .
note that every point of such an orbit can be considered as a fixed point of the corresponding _ first return map_. such a map is usually constructed as a superposition @xmath104 of two maps @xmath105 and @xmath106 , see figure [ fg : fabhab](b ) .
map @xmath107 is called _ local map _ and it is the restriction of @xmath90 onto @xmath97 , i.e. @xmath108 . map @xmath78 is called _ global map _ , it is defined as @xmath109 with an integer @xmath110 and maps a small neighborhood @xmath111 of some point @xmath112 of orbit @xmath84 into a neighborhood @xmath113 of another point @xmath114 of @xmath84 such that @xmath115 .
thus , any fixed point of @xmath16 is a point of a single - round periodic orbit of period @xmath116 for @xmath90 . in order to study maps @xmath16 for all sufficiently large integer @xmath17
, it is very important to have good coordinate representations for both maps @xmath107 and @xmath78 , especially it concerns local map @xmath107 and its iterations @xmath117 for large @xmath17 .
we use in @xmath97 the canonical coordinates @xmath18 given by the following lemma .
@xcite[lem : nfgen ] for any given integer @xmath118 ( such that @xmath119 or @xmath118 is arbitrary for @xmath120 or @xmath121 ( the real analytic case ) ) , there is a canonical change of coordinates , of class @xmath80 for @xmath122 or @xmath123 for @xmath124 , that brings @xmath107 to the following form for @xmath122 or @xmath125 for @xmath124 , respectively , see more details in @xcite . ]
@xmath126 where @xmath127 and @xmath128 are the birkhoff coefficients . the normal forms ( [ eq : fng1 ] ) are very suitable for effective calculations of maps @xmath129 with sufficiently large integer @xmath17 .
thus , the following result is valid .
@xcite [ lem : tknfgen ] let @xmath107 be given by ( [ eq : fng1 ] ) , then map @xmath117 can be written , for any integer @xmath17 , as follows @xmath130 where @xmath131 @xmath132 , @xmath133 are some @xmath134-th degree polynomials of @xmath17 with coefficients depending on @xmath135 and the functions @xmath136 are uniformly bounded in @xmath17 along with all their derivatives by coordinates up to order either @xmath137 for @xmath122 or @xmath138 for @xmath124 .
the main case @xmath122 of lemmas [ lem : nfgen ] and [ lem : tknfgen ] , related to the so - called main normal form of a saddle map , was proved in the papers @xcite .
analogues of lemmas [ lem : nfgen ] and [ lem : tknfgen ] in the case of area - preserving and orientation - reversing maps were proved in @xcite .
[ rem1 ] normal form ( [ eq : fng1 ] ) can be considered as a certain finitely smooth approximation of the following analytical moser normal form @xmath139 which takes place for @xmath50 @xcite , where @xmath140 .
accordingly , relation ( [ eq : tkgen ] ) looks as a very good approximation for the corresponding formulas in the analytical case , @xcite , @xmath141 where @xmath142 and @xmath143 etc .
[ rem2 ] in the coordinates of lemma [ lem : nfgen ] , we have that @xmath144 . without loss of generality
, we assume that @xmath145 and @xmath146 .
let neighborhoods @xmath147 and @xmath148 of the homoclinic points @xmath149 and @xmath150 , respectively , be sufficiently small such that @xmath151 and @xmath152 . then , as usually ( see e.g. @xcite ) , the successor map from @xmath153 into @xmath154 by orbits of @xmath107 is defined , for all sufficiently small @xmath88 , on the set consisting of infinitely many strips @xmath155 , @xmath156 . the image of @xmath157 under @xmath117 is the strip @xmath158 . as @xmath23 , the strips @xmath157 and @xmath159 accumulate to @xmath160 and @xmath161 , respectively .
we can write global map @xmath162 as follows ( in the coordinates of lemma [ lem : nfgen ] ) @xmath163 where @xmath164 and @xmath165 these relations follow from condition * ( b ) * , which means that for @xmath89 curve @xmath166 has a cubic tangency with @xmath167 at @xmath168 .
when the parameters vary , this tangency can split and , by condition * ( c ) * , family ( [ eq : t1 ] ) unfolds generally the initial cubic tangency . in this case
, global map @xmath78 can be written in a certain normal form that the following lemma shows .
if condition * ( c ) * holds , then map @xmath169 can be brought to the following form @xmath170 where @xmath171 and the coefficients @xmath172 as well as @xmath173 depend smoothly ( with the same smoothness as for the initial map ( [ eq : t1 ] ) ) on new parameters @xmath88 such that @xmath174 and @xmath175 ( the other parameters are not essential ) .
[ nft1 ] by virtue of ( [ eq : t1-cubt ] ) , when @xmath176 and @xmath88 are small enough , equation @xmath177 can be solved for @xmath178 .
the solution has the form @xmath179 , where @xmath180 .
then , we can write the following taylor expansion for function @xmath181 near curve @xmath179 : @xmath182 since @xmath183 , we obtain that @xmath184 where @xmath185 .
hence , map @xmath169 can be written in the following form @xmath186 then the equation of the piece of @xmath187 in @xmath153 has the following form @xmath188 condition * ( c ) * implies that @xmath189 and the coefficients @xmath190 and @xmath191 can take any values from the ball @xmath192 , where @xmath193 is a small constant .
thus , the system @xmath194 has always a solution and we can consider @xmath92 and @xmath93 as new parameters .
due to lemma [ nft1 ] , we can consider directly two parameter families @xmath195 of symplectic maps .
it is easy to see from ( [ eq : t1par ] ) that @xmath92 and @xmath93 are the splitting parameters of the manifolds @xmath196 and @xmath197 with respect to the homoclinic point @xmath168 .
indeed , by ( [ eq : t1wu ] ) , curve @xmath198 has the equation @xmath199 thus , family @xmath195 is a general two parameter unfolding of the initial cubic tangency which occurs for @xmath200 . for any such unfolding , curves @xmath160 and @xmath187 must have a quadratic tangency for certain values of @xmath92 and @xmath93 .
it is true that for our family the following result , which implies item 1 of theorem [ thm : main ] , holds . in ( a ) incoming from above ( @xmath201 ) and ( b ) incoming from below ( @xmath202 ) cases . for @xmath203 ,
map @xmath27 has only one transverse homoclinic orbit close to @xmath84 , whereas , for @xmath204 , @xmath27 has exactly 3 transverse homoclinic orbits.,width=453 ] in the @xmath34-parameter plane , there exists a bifurcation curve @xmath35 : @xmath205^{3/2}\ ] ] ( see figure [ fig18 - 01 ] ) such that for @xmath206 , map @xmath27 has a close to @xmath84 homoclinic orbit with a quadratic tangency of the manifolds @xmath207 and @xmath208 .
[ th1g85 ] consider curve @xmath209 given by equation ( [ eq : t1wuloc ] ) .
if the curve has a tangency with the line @xmath210 , the following system has solutions @xmath211 we solve the second equation for @xmath212 , put it into the first equation , and we obtain the equation of curve @xmath35 .
note that the tangency is always quadratic except for the case @xmath213 which corresponds to a cubic tangency at @xmath200 .
the main goal in the present paper is to study bifurcations of single - round periodic orbits of family @xmath195 . as said above
, it is equivalent to the study of bifurcations of fixed points in the first return maps @xmath214 for every sufficiently large integer @xmath17 ( @xmath215 ) .
we apply the rescaling method in order to find the rescaled normal forms for these maps in the following lemma .
+ for every sufficiently large @xmath17 , the first return map @xmath16 can be brought , by a linear transformation of coordinates and parameters , to the following form @xmath216 where @xmath217 @xmath218 and @xmath219 .
[ henmain - c ] by lemmas [ lem : tknfgen ] and [ nft1 ] , the first return map @xmath214 can be written as follows @xmath220 we shift the coordinates , @xmath221 , where @xmath222 , @xmath223 and @xmath224
. then system ( [ tk ] ) is rewritten as follows @xmath225 where @xmath226 are small coefficients . here
the first equation of ( [ tk+ ] ) does not contain constant terms and the quadratic term in @xmath227 vanishes in the second equation . finally , by means of the coordinate rescaling @xmath228 we bring map ( [ tk+ ] ) to the required form ( [ henon0 - 3 ] ) .
by the rescaling lemma [ henmain - c ] , the following conservative cubic hnon maps @xmath229 and @xmath230 have to be considered as certain normal forms for the first return maps in the cases @xmath231 and @xmath232 , respectively .
thus , if @xmath50 , map @xmath62 corresponds to the cubic homoclinic tangency with @xmath202 ( the incoming from below tangency , see figure [ fig2type](b ) ) , while map @xmath63 is related to the cubic homoclinic tangency with @xmath201 ( the incoming from above tangency , see figure [ fig2type](a ) ) . in figures
[ gk88-+2 ] and [ gk882 ] , the main elements of the bifurcation diagrams for these cubic maps are presented .
the bifurcation curves in figures [ gk88-+2 ] and [ gk882 ] are found analytically ( see e.g. @xcite ) and their equations are as follows , where @xmath233 and @xmath234 correspond to the maps @xmath62 and @xmath63 , respectively : @xmath235 ^ 2\left[-5m_2 -12 + ( -1)^i \sqrt{9m_2 ^ 2 + 24m_2}\right ] ; \end{array } \label{eq : criv}\ ] ] where if @xmath236 , then there exists a fixed point with double eigenvalue @xmath42 ; if @xmath237 , there exists a fixed point with double eigenvalue @xmath43 ; if @xmath238 , there exists a 2-periodic orbit with double eigenvalue @xmath42 ; if @xmath239 , there exists a 2-periodic orbit with double eigenvalue @xmath43 .
the main bifurcations related to fixed points are as follows .
@xmath240 + * bifurcation scenario in map @xmath241 , see figure [ gk88-+2 ] .
* bifurcation curves @xmath242 and @xmath243 from ( [ eq : criv ] ) with @xmath244 divide the @xmath19-parameter plane into 3 regions symmetric with respect to the @xmath9-axis . when parameters @xmath8 and @xmath9 are in region @xmath245 , map @xmath241 has a unique fixed point @xmath246 which is a saddle - plus ( with eigenvalues @xmath30 and @xmath31 , where @xmath247 ) .
the transition @xmath248 corresponds to the birth of two ( saddle and elliptic ) fixed points . at @xmath249 ,
the fixed point @xmath246 is a non - hyperbolic saddle with double eigenvalue @xmath42 , and then this point falls into three ( two saddle and one elliptic ) fixed points in domain @xmath250 under a conservative pitch - fork bifurcation .
the transition @xmath251 corresponds to a nondegenerate period doubling bifurcation of the elliptic fixed point .
thus , for region @xmath252 , map @xmath241 has three saddle ( two saddle - plus and one saddle - minus ) fixed points and one elliptic 2-periodic orbit .
further primary bifurcations , when crossing curves @xmath253 and @xmath254 , are related to conservative fold and period doubling bifurcations of 2-periodic orbits , respectively , and we do not discuss them here ( see more details in @xcite ) . .
the equations of bifurcation curves @xmath242 , @xmath243 are given by ( [ eq : criv ] ) with @xmath255 .
the equation of @xmath256 is given by ( [ eq : twistless ] ) with `` + '' .
the codimension 2 bifurcation points @xmath257 are as follows : @xmath258 a nonhyperbolic saddle fixed point with double eigenvalue @xmath42 exists ; @xmath259 and @xmath260 two 2-periodic cycles with double eigenvalue @xmath43 coexist ; @xmath261 two 2-periodic cycles , one with double eigenvalues @xmath43 and the other with double eigenvalues @xmath42 , coexist .
examples of ( b ) symmetric ( @xmath262 ) and ( c ) asymmetric @xmath263 bifurcations are also shown .
, width=604 ] .
the equations of bifurcation curves @xmath242 and @xmath243 are given by ( [ eq : criv ] ) with @xmath234 .
the equation of @xmath256 is given by ( [ eq : twistless ] ) with `` @xmath264 '' .
the codimension 2 bifurcation points @xmath257 are as follows : @xmath258 and @xmath265 there exists a fixed point with eigenvalues @xmath266 whose first lyapunov value is zero but second is nonzero , these points divide curve @xmath243 into the segments where the corresponding period doubling bifurcation is either subcritical ( when crossing the piece @xmath267 of @xmath243 ) or supercritical ( crossing outside @xmath267 ) ; @xmath259 a triple fixed point of elliptic type exists ; @xmath268 and @xmath269 two fixed points with eigenvalues @xmath270 and @xmath271 coexist ; @xmath260 two fixed points with eigenvalues @xmath270 coexist ; @xmath272 two 2-periodic orbits with eigenvalues @xmath270 coexist .
, width=604 ] @xmath240 + * bifurcation scenario in map @xmath273 , see figure [ gk882 ] . * bifurcation curves ( [ eq : criv ] ) with
@xmath274 divide the @xmath19-parameter plane into 15 regions including the regions @xmath245 , @xmath275 , @xmath276 , @xmath277 and @xmath278 which are symmetric with respect to the @xmath9-axis , while the other regions are pairwise symmetric ( e.g @xmath279 and @xmath280 ) . for @xmath19
belonging to region @xmath245 , map @xmath273 has a unique fixed point @xmath281 which is a saddle - minus ( with eigenvalues @xmath30 and @xmath31 , where @xmath282 ) .
the transition @xmath248 through the segment @xmath267 of curve @xmath243 corresponds to a subcritical period doubling bifurcation with the point @xmath281 : the point becomes elliptic and a saddle 2-periodic orbit emerges in its neighborhood .
the transition @xmath283 ( and symmetrically @xmath284 ) corresponds to a nondegenerate conservative fold bifurcation under which saddle and elliptic 2-periodic orbits are born .
the transition @xmath285 ( as well as @xmath286 ) corresponds to a supercritical period doubling bifurcation under which the elliptic 2-periodic orbit merges with the saddle fixed point and becomes an elliptic fixed point .
thus , in domain @xmath250 , map @xmath273 has an elliptic fixed point and a saddle 2-periodic orbit .
the transitions @xmath287 and @xmath288 ( and , symmetrically , @xmath289 and @xmath290 ) correspond to a period doubling of the elliptic 2-periodic orbit ( it becomes a saddle 2-periodic orbit and an elliptic 4-periodic orbit is born in its neighborhood ) .
we also illustrate bifurcations occurring at a passage around the point @xmath258 in figure [ gm2 ] ( and , symmetrically , for the point @xmath265 ) .
transitions through curve @xmath242 at increasing @xmath9 correspond to the appearance of two new ( saddle and elliptic ) fixed points under a conservative fold bifurcation .
thus , in region @xmath291 , map @xmath273 has three fixed points , two elliptic and one saddle .
the elliptic fixed points undergo a period doubling bifurcation at transitions @xmath292 and @xmath293 for one of the points , or @xmath294 and @xmath295 for the other point .
we note that for @xmath296 , there exists a triple ( of elliptic type ) fixed point which splits into three ( two elliptic and one saddle ) fixed points in region @xmath291 . [ tb ] of the bifurcation diagram in figure [ gk882 ] .
region @xmath297 ( homo / heteroclinic zone ) corresponds to the values of @xmath19 where the invariant manifolds of all saddles intersect . in a rough approximation , these bifurcations are similar to those of a two - dimensional hamiltonian system whose potential function is symmetric and changes as shown.,width=453 ] in the @xmath19-parameter plane , bifurcation curves @xmath242 and @xmath243 bound those open regions of parameter values where elliptic fixed points of maps @xmath241 and @xmath273 exist . such a fixed point is called _ generic _ if , first , @xmath298 , i.e. no strong resonances occurs , and , second , in the main complex normal form @xmath299 the first birkhoff coefficient @xmath300 is nonzero .
note that if an elliptic point is generic , then it is stable . however , non - generic elliptic points can be both stable ( of elliptic type ) and unstable ( of saddle type , for example , a saddle with 6 or 8 separatrices if @xmath301 or @xmath302 ) .
the following curve @xmath303 in the @xmath19-parameter plane defines those parameter values where map @xmath304 has a fixed point @xmath305 with eigenvalues @xmath306 with @xmath307 .
accordingly , curves @xmath308 and @xmath309 of strong 1:3 and 1:4 resonances are resulted from ( [ eq : ellfp ] ) for @xmath310 and @xmath311 , respectively . to find the conditions when the point @xmath305 is non - generic , i.e. @xmath312 , we use the formula for @xmath300 from @xcite , see formula ( 35 ) there .
then we obtain that @xmath313 for @xmath314 this formula is true for both maps @xmath315 and @xmath273 .
equations ( [ eq : ellfp ] ) and ( [ eq : ellb0 ] ) define in the @xmath19-parameter plane the so - called non - twist curve @xmath256 such that for @xmath316 , map @xmath304 has a degenerate elliptic fixed point ( with @xmath313 ) .
these curves are shown in figures [ gk88-+2 ] and [ gk882 ] .
for both maps , the curves consist of two connected pieces , the infinite branches of @xmath256 asymptotically tend to the curve @xmath317 with @xmath318 . in the case of map @xmath63 , one of the branches of the curve @xmath256 has a self - intersection point @xmath319 , where map @xmath320 has simultaneously two degenerate elliptic fixed points ( both with @xmath321 ) . for map @xmath62 , there also exists a self - intersection point ( @xmath322 ) of curve @xmath256 .
this point corresponds to the existence of a fixed point with eigenvalues @xmath47 ( 1:3 resonance ) , which is kam - stable in this case , @xcite .
note that , in the case of map @xmath63 , the fixed point @xmath323 at @xmath322 has eigenvalues @xmath47 and is also kam - stable ( here @xmath256 and @xmath324 are tangent ) .
note that the fixed point @xmath325 is always nondegenerate ( generic ) if @xmath326 in the case of map @xmath315 and if @xmath327 in the case of map @xmath320 .
we also note that curves @xmath256 can be written in the explicit form @xcite as follows @xmath328 now we can complete the proof of theorem [ thm : main ] .
indeed , we only need to translate the results on the bifurcation diagrams obtained in the @xmath19-plane for the rescaling maps @xmath241 and @xmath273 into the initial @xmath34-parameter plane .
we can easily make this using relations ( [ mui - c ] ) between the initial and rescaling parameters .
then , using equations ( [ eq : criv ] ) for curves @xmath242 , @xmath243 and @xmath329 , we obtain the following equations for the curves in the @xmath34-plane , where @xmath330 , @xmath331 analogous formulas can be obtained for the other bifurcation curves , given in ( [ eq : criv ] ) , ( [ eq : ellfp ] ) for @xmath332 , and ( [ eq : twistless ] ) .
this completes the proof of theorem [ thm : main ] .
the structure of the strong 1:3 resonance , i.e. the structure of bifurcations related to the existence of fixed points with eigenvalues @xmath47 , in the case of cubic maps @xmath241 and @xmath273 was studied in @xcite . in this section
we analyze bifurcations related to the existence of fixed points with eigenvalues @xmath65 ( the main 1:4 resonance ) in maps @xmath241 and @xmath273 .
we find conditions of nondegeneracy of the corresponding resonances and give a description of accompanying bifurcations including among the latter bifurcations we pay main attention to bifurcations of symmetric 4-periodic points , i.e. such 4-periodic orbits which has two points on the symmetry line @xmath333 .
we will call such orbits @xmath334-symmetric . note that both maps @xmath241 and @xmath273 are reversible with respect to the involution @xmath335 , i.e. the maps preserve the form if one makes the coordinate change @xmath336 and consider the inverse map .
the line @xmath334 is the line of fixed points of involution @xmath336 and , thus , @xmath334-symmetric periodic orbits can undergo nondegenerate bifurcations of two types : symmetry preserving parabolic ( hold ) bifurcations and symmetry breaking pitch - fork bifurcations . in the conservative case , there are two types of pitch - fork bifurcations : ( i ) when a symmetric elliptic fixed ( periodic ) point becomes a symmetric saddle fixed point and a pair of elliptic fixed points ( symmetric each other with respect to @xmath334 ) is born , and ( ii ) when a symmetric saddle fixed ( periodic ) point becomes a symmetric elliptic fixed point and a pair of symmetric each other saddle fixed points is born .
we note that , in the case under consideration , there are nonsymmetric 4-periodic orbits whose structure and bifurcations can be important to understand some thin details of 1:4 resonance . in this case
, we study these orbits only numerically . the map @xmath241 has a unique fixed point with eigenvalues @xmath65 at values of parameters @xmath8 and @xmath9 belonging to the curve @xmath337 its equation is obtained from ( [ eq : ellfp ] ) with the sign `` @xmath338 '' and @xmath339 .
it was found in @xcite that the complex local normal form ( [ hecomplnew13 ] ) near such a point has the following coefficients @xmath340 and , thus , in ( [ d21d03 ] ) @xmath341 since @xmath342 in the case of curve @xmath343 , we always have @xmath344 .
thus , in ( [ d21d03 ] ) only condition @xmath345 is violated at @xmath346 .
the main local bifurcations which occur here at transition of the parameters through curve @xmath343 ( except for the two points with @xmath346 ) are illustrated in figure [ fig : res1_4chm1 ] .
curve @xmath343 divides the @xmath19-parameter plane into two regions i and ii . in both regions ,
the fixed point is generic elliptic , however , it is not generic for @xmath347 although the point is of elliptic type .
the transition from i into ii is accompanied by the appearance of a resonant 1:4 chain containing the saddle and elliptic 4-periodic cycles around the fixed point .
we note that in curve @xmath343 there are two points @xmath348 and @xmath349 with @xmath346 , where @xmath350 .
now we consider in more detail a neighborhood of the point @xmath349 ( point @xmath348 can be studied in the same way ) .
we find that @xmath349 is an endpoint of two bifurcation curves @xmath351 and @xmath352 .
these two curves divide domain ii into 3 domains , ii@xmath353 , ii@xmath354 and ii@xmath355 . in domain
ii@xmath353 , the 1:4 chain looks as in figure [ pi2plusres ] for @xmath356 , @xmath357 .
thus , the chain contains the @xmath334-symmetric elliptic 4-periodic orbit ( whose two points belong to the bisecting line @xmath358 ) . at crossing the line @xmath351 ( from ii@xmath353 to ii@xmath354 ) , the elliptic orbit undergoes a pitch - fork bifurcation : it becomes a symmetric saddle 4-periodic orbit and a pair of symmetric each other elliptic 4-periodic orbits emerges .
when the parameters change in zone ii@xmath354 , some bifurcations connected with the reconstruction of homo / heteroclinic structures take place , and , finally , the saddle 4-periodic orbit of the chain which is not @xmath334-symmetric undergoes a pitch - fork bifurcation at crossing curve @xmath352 : two elliptic 4-periodic orbits merge into it and the orbit becomes elliptic .
thus , in domain ii@xmath355 , the 1:4 chain looks as in figure [ pi2plusres ] for @xmath359 . and
@xmath352 when @xmath8 changes for a fixed @xmath357 .
effect of thinning of the 1:4 resonance chain is observed and some accompanied pitch - fork bifurcations with 4-periodic orbits are shown . ,
width=529 ] in figure [ res14plus ] we interpret these bifurcations by means of the following flow normal form ( which is invariant under rotation by angle @xmath69 ) @xmath360 where @xmath67 and @xmath361 are small real parameters , coefficients @xmath362 and @xmath363 are real and @xmath364 ( in this case the divergence of ( [ nf1b30 ] ) is zero , @xcite ) .
note that system ( [ nf1b30 ] ) is also reversible , it is invariant under the time reversal @xmath365 and the following involutions : @xmath366 , @xmath367 , @xmath368 and @xmath369 .
accordingly , if @xmath370 , the system has 8 nontrivial equilibria lying in pairs in the 4 lines @xmath371 of fixed points of the corresponding involutions . )
when @xmath361 changes and @xmath370 is fixed , related to the cases ( a ) @xmath372 , ( b ) @xmath373 , ( c ) a critical value @xmath374 corresponding to the homo / heteroclinic zone , ( d ) @xmath375 , ( e ) @xmath376 . here , @xmath333 and @xmath377 are the lines of fixed points of the corresponding involutions . in the lower plots for the potential function
, we need to identify the left and rights hand sides ( @xmath378 and @xmath379 ) of the bands .
, width=604 ] we see that the theoretical results , figure [ res14plus ] , are in good accordance with the numerical evidence for a parameter path @xmath380 in figure [ pi2plusres ] .
note that in figure [ res14plus ] the case ( c ) approximates the homo / heteroclinic zone of domain ii@xmath354 , where the invariant manifolds of the saddles are reconstructed from the ( b)-configuration to the ( d)-configuration .
simultaneously when the parameters change in domain ii@xmath354 , we observe the effect of thinning of the 1:4 resonance chain .
this is connected with the fact that the normal form ( [ nf1b30 ] ) with @xmath381 and @xmath382 corresponds to a nonlinear center .
the cubic map @xmath273 has a fixed point with eigenvalues @xmath65 for @xmath383 , where curve @xmath384 has equation @xmath385 this equation is obtained from ( [ eq : ellfp ] ) with the sign `` @xmath264 '' and @xmath339 .
note that curve @xmath384 has a self - intersection point ( @xmath386 ) , and only in this moment the map has simultaneously two fixed points with eigenvalues @xmath65 .
the coefficients of the local complex normal form ( [ hecomplnew13 ] ) are , @xcite , @xmath387 since @xmath388 in curve @xmath384 , in ( [ d21d03 ] ) we have @xmath389 and @xmath390 thus , both cases @xmath344 ( when @xmath391 ) and @xmath392 ( when @xmath393 ) are possible here
. in @xmath384 , there are two points @xmath394 and @xmath395 where the case @xmath75 takes place .
note that if @xmath392 , the fixed point @xmath29 is a saddle with 8 separatrices for @xmath396 , and the main bifurcations are connected with the reconstruction of a saddle 4-periodic orbit ( as at the transition between the domains ii and iii from the bifurcation diagram of figure [ fig_pi2-a1 ] ) . in the degenerate case
@xmath75 , the main bifurcations can be described by means of a two parameter family of the hamiltonian system @xmath397 where @xmath67 and @xmath361 are parameters , @xmath398 and @xmath399 are real coefficients , see more details in @xcite .
the typical bifurcation diagram , @xcite . ] for family ( [ 2p5_2 ] ) is shown in figure [ fig_pi2-a1 ] .
it contains three bifurcation curves which divide the @xmath400-parameter plane into 3 domains . crossing the curves @xmath401 and @xmath402 corresponds to the reconstruction of nonzero saddle equilibria .
curve @xmath403 corresponds to the appearance of 4 nonzero parabolic equilibria ( this bifurcation is nondegenerate for the family ( [ 2p5_2 ] ) since it is invariant under rotation by angle @xmath69 ) .
note that the trivial equilibrium is a nondegenerate conservative center for @xmath404 ; it is a degenerate conservative center for @xmath405 and a saddle with 8 separatrices for @xmath406 . concerning map @xmath273 ,
the main elements of the bifurcation diagram related to 1:4 resonance are shown in figure [ fig : res1_4_chm2 ] . here
the 4 main bifurcation curves , @xmath384 , @xmath407 and @xmath408 , are shown in the @xmath19-parameter plane .
the equation for @xmath384 is given by ( [ eq : res- ] ) , the remaining curves have the following equations @xmath409 curves ( [ p4 - 0 ] ) correspond to nondegenerate parabolic bifurcations for @xmath334-symmetric 4-periodic orbits .
the birth of a pair of saddle and elliptic 4-periodic orbits occurs at upward crossing these curves .
curves @xmath410 and @xmath411 are ( quadratically ) tangent to curve @xmath384 at the points @xmath412 , see figure [ fig : res1_4_chm2 ] . in principle , bifurcations of map @xmath273 , when @xmath8 and @xmath9 vary near @xmath412 , look similar to the bifurcations in the flow normal form ( [ 2p5_2 ] ) , compare figures [ fig_pi2-a1 ] and [ fig : res1_4_chm2 ] ( in the latter figure , such bifurcations are displayed for the case of point @xmath413 ) .
the presence of curve @xmath408 shows that the structure of 1:4 resonance in the case of map @xmath273 is not trivial that can be imagined from figure [ fig : res1_4_chm2 ] , where the main elements of the corresponding bifurcation diagram are shown .
v. s. biragov .
bifurcations in a two - parameter family of conservative mappings that are close to the hnon mapping .
( russian ) translated in selecta math .
soviet . 9(3):273282 , 1990 .
_ methods of the qualitative theory of differential equations ( russian ) _ , 1024 , 1987 .
s.v.gonchenko . on a two parameter family of systems close to a system with a nontransversal poincar homoclinic curve .
i. _ methods of qualitative theory of differential equations _ ; editor : e.a.leontovich ; gorky st .
, 55 - 72 , 1985 // [ english transl . in : selecta math .
sovietica,10 , 1990 ] .
gonchenko , l.p .
shilnikov , d.v . turaev . on newhouse regions of two - dimensional diffeomorphisms close to a diffeomorphism with a nontransversal heteroclinic cycle , _ proc .
steklov inst .
_ , 216:7 - 118 , 1997 .
gonchenko , l.p .
shilnikov , d.v .
homoclinic tangencies of an arbitrary order in newhouse domains , _ itogi nauki tekh .
_ , 67:69 - 128 , 1999 [ english translation in _
_ , 105:1738 - 1778 , 2001 ] .
bifurcations of codimension two near homoclinic tangencies of second order . _ in proceedings of the european conference on iteration theory _ , world sci .
publishing , teaneck , nj . 306318 , 1992 . | we study bifurcations of cubic homoclinic tangencies in two - dimensional symplectic maps . we distinguish two types of cubic homoclinic tangencies , and each type gives different first return maps derived to diverse conservative cubic hnon maps with quite different bifurcation diagrams . in this way
, we establish the structure of bifurcations of periodic orbits in two parameter general unfoldings generalizing to the conservative case the results previously obtained for the dissipative case .
we also consider the problem of 1:4 resonance for the conservative cubic hnon maps . |
type ia supernovae ( sne ia ) have been used as standardized candles to discover the acceleration of the universe @xcite .
programs are now underway to characterize the dark energy driving this expansion by measuring its average equation of state , @xmath4 .
essence aims to discover and follow - up @xmath5 200 sne ia for this purpose @xcite , while the goal of the snls ( supernova legacy survey ) is to obtain 700 well observed sne ia in the redshift range 0.2 to 0.9 @xcite .
the snls uses the canada - france - hawaii telescope legacy survey ( cfhtls ) imaging data for sn discoveries and lightcurves . over the course of a year ,
four fields are imaged every four days ( 23 days rest frame ) during dark and gray time , in four filters ( griz ) .
since the same fields are continuously imaged , this is a `` rolling search , '' where new sn candidates are discovered throughout the month as lightcurves are being built .
the snls uses the very large telescope ( vlt ) , keck and gemini telescopes to determine sne types and redshifts .
the vlt has the shortest setup time , so it handles the bulk of the lower redshift ( @xmath6 ) candidates , as well as some higher redshift targets .
gemini - n and gemini - s generally observe the faintest , highest redshift targets ( @xmath7 ) , where the nod - and - shuffle mode provides a reduction of sky line residuals in the red part of the spectrum .
finally , keck provides essential observations when the northernmost field can not be observed by the vlt . here
we present the methods and spectroscopy from the gemini telescopes for the first year of operation of the snls .
separate science analyses on certain spectroscopic features , as well as observations from other telescopes , will be presented in upcoming papers . a note on our candidate names : the classical international astronomical union ( iau ) naming convention for supernovae does not work well for projects of the scale of the snls , where there are more candidates than confirmed sne . under the traditional naming system ,
the candidates from high - redshift searches have an internal ( unofficial ) name that they were given at the time of discovery , then if they are suspected sne they are given one type of official iau name , and if they are confirmed sne they are given a second type of official iau name .
to avoid these problems , we use a unified naming scheme for all snls discoveries , regardless of whether they are confirmed to be sne .
an example name is snls-04d3aa the first two digits are the year of discovery , the second two are the field name ( d1 , d2 , d3 , d4 ) , and the final two are identifying letters .
we start with `` aa '' for the first candidate discovered in a field in a given year .
the second is `` ab , ''
then `` ac '' and so on , until the 27th candidate , which is named `` ba . ''
historically , only @xmath8 of high - redshift sn candidates observed spectroscopically in the range @xmath9 are confirmed as type ia sne . the rest of the candidates are either identified or unidentified core collapse sne or agn , or are sne ia which have such high host contamination or low signal - to - noise ( s / n ) that they can not be identified .
the exact percentage confirmed depends on the design of the search and the redshift range studied ( see section [ photoz ] for a detailed discussion ) . also found that the sn ia confirmation rate was higher for classical searches than it was for `` rolling '' searches .
a classical high - z sn search technique @xcite is optimized for the discovery of sne ia .
it involves taking a reference image after new moon to be subtracted from a discovery image taken @xmath103 weeks later ( 2 weeks restframe at z=0.5 ) .
since the rise time of sne ia is about 18 days , this guarantees that the sne discovered will be either rising or near maximum light .
cuts are made to retain only candidates that have increased in brightness by an amount consistent with the rapid rise of an sn ia . in `` rolling searches , '' such as the snls and essence
, the field is continuously imaged every few days in several filters .
sne are discovered very early , so there is no built - in bias toward the discovery of sne ia ( except that sne ia are usually brighter than core collapse sne at a given redshift ) .
rolling searches are typically limited not by the number of candidates discovered , but by the amount of spectoscopic confirmation time available , so techniques must be developed to increase the percentage of targets confirmed as sne ia by spectroscopy . in principle
, this should be easier in a rolling search than in a classical search . in the snls , sne
are almost almost always discovered well before maximum light , so multicolor information on the rising part of the lightcurve is available before the candidates are observed spectroscopically . here
we demonstrate the use of a technique to prioritize candidate spectroscopy . after each photometric data point
is taken , a candidate s lightcurve is fit under the assumption that it is an sn ia .
the lightcurve shape , phase , and redshift of the candidate are allowed to vary in the fit , while the milky way reddening is fixed to the values of @xcite .
host reddening can be solved for or held fixed , depending on the number of data points fit .
we do not use a fit until it has detections in at least two epochs in at least two filters , as well as at least one prior epoch of nondetections . in practice ,
most candidates have detections on several epochs in at least three filters , and dozens of nondetections .
significantly deviant objects are lowered in priority ( for example , type ii sne and agn are often too blue at early times ( see * ? ? ? * ; * ? ? ?
* ) ) , while rising candidates that are consistent with being an sn ia are given top priority for spectroscopic observation .
because the inferred color of the object depends on the redshift , one product of the fitting process is an estimated redshift .
we call this a `` supernova photometric redshift , '' or `` sn photo - z , '' since the redshift is estimated from prior assumptions about the intrinsic brightness of sne ia ( 0.2 mag of dispersion is allowed for in the assumed sn ia properties ) .
the sn photo - z is independent from galaxy photometric redshifts , and is useful for determing the telescope and instrument setup necessary to observe key sn ia spectroscopic features .
note that galaxy photometric redshifts were not used in assigning the follow - up priority for the targets considered in this paper .
figures [ iapz ] and [ iipz ] show examples of sn photo - z fits to objects that were eventually confirmed to be type ia and type ii sne , respectively .
while we show the entire lightcurve fit here for clarity , it is apparent even before maximum light that the candidate in figure [ iapz ] is likely to be an sn ia , while the candidate in figure [ iipz ] is too blue at early times to be an sn ia .
the candidates are usually spectroscopically observed near maximum light , so most sn photo - z fits are done using only the rising part of the lightcurve .
since we observe many of our highest redshift ( and therefore spectroscopically expensive ) targets at gemini , we often apply the most strict selection critera on what to observe there .
therefore the data from this telescope make an excellent testbed for our selection method .
however , at other telescopes we observe a broader range of candidates to ensure that we are not missing unusual sne ia .
it is also important to note that the sn photo - z is only used to prioritize follow - up observations for all scientific studies we use the measured spectroscopic redshift .
the sn photo - z will be described in more detail in @xcite , and the results of its implementation on our sn ia confirmation rate are given in section [ photoz ] .
the observational setup and conditions for each candidate are given in table [ obs - table ] . in this section
we discuss the observational modes used in greater detail .
all supernovae were observed with the gemini multi object spectrographs ( gmos ) @xcite .
longslit spectra of the sn candidates were taken with the gmos r400 grating ( 400 lines per mm ) , using the 0.75 slit , giving a resolution of 6.9 .
the detector binning was set at @xmath11 , giving a spatial scale of 0.145 per pixel and a spectral dispersion of 1.34 per binned pixel .
the slit angle was chosen to include the host galaxy wherever possible , as the galaxy features provide the most accurate determination of the redshift .
the central wavelength used was based on our estimated redshift of each candidate . for candidates at estimated redshifts
@xmath12 we used a central wavelength of 680 nm ( with the order blocking filter gg455 ) , covering 4650 to 8900 . for higher redshifts
the 720 nm setting ( with the og515 order blocker ) was used , giving coverage from 5100 to 9300 .
the observations were executed in queue mode which allows us to specify the desired observing conditions .
we require better than 0.75 image quality ( corrected to zenith ) and clear sky and generally those requirements were met .
observations were made in either nod - and - shuffle ( n&s ) , electronic nod - and - shuffle , or classical observing modes . during a classical observation , the object and the nearby sky
are simultaneously imaged on adjacent portions of the ccd .
the sky is fit column - by - column and subtracted , but this can leave systematic residuals near bright and spatially variable sky lines .
since this technique requires no additional overhead we use it on the brightest targets ( @xmath13 ) .
( all snls magnitudes reported in this paper are in the ab system . )
the nod - and - shuffle mode @xcite enables more accurate sky subtraction by nodding frequently ( every 60 seconds ) between two positions along the slit .
the charge on the ccd detectors is simultaneously shuffled between the illuminated science region of the ccd and the unilluminated ( storage ) region .
thus , both the object and its immediate sky background are imaged on both the same pixels and at the same slit position .
systematics due to pixel response , fringing in the sky lines , slit irregularities , and any temporal sky variations can be removed by subtracting the n&s acquired sky spectrum from the object spectrum .
this method works well towards the red end of the spectrum ( where sky lines are more problematic ) and we found it particularly useful for candidates at magnitudes fainter than @xmath14 .
typical nod distances for this method are a few arcseconds .
there are a few drawbacks to nod - and - shuffle .
first , the noise in sky - subtracted n&s images is higher by a factor of @xmath15 compared to classically - reduced long - slit spectra , because one image is subtracted from another .
second , there are increased overheads on any observation , as each nod cycle adds approximately 24 seconds of nod time . for a typical 1800s observation , with 15 ( 60 second )
nod cycles , these overheads add an extra 360s . the extra overhead time can be minimized by choosing a small nod distance , or by employing the electronic n&s mode .
the gmos instrument requires the use of an on - instrument wavefront sensor ( oiwfs ) , which provides `` fast guiding '' ( image motion compensation ) and higher order corrections . during a normal n&s observation ,
this sensor physically moves during each nod cycle .
the electronic n&s mode avoids this by electronically changing the position that the oiwfs guides around instead of physically moving the sensor .
this decreases the overheads for a n&s observation by nearly 200s ( for an 1800s exposure ) , but is only available for small nod distances ( up to 2 ) .
these small nod distances are sometimes not possible when the candidate resides in an extended host galaxy .
data reduction was done by collaborators at both university of toronto and oxford university using independent pipelines primarily written in iraf using the gemini software package version 1.6 . here
we describe the toronto pipeline , although the reduction methods are similar for both pipelines .
the independent reductions have been used to check the consistency of the final output spectra . in gmos the spectra
are spread over 3 ccds .
first , bias subtraction is performed using a master bias for each chip generated from bias frames covering an entire gmos queue run .
next , flat - fielding is done on a chip - by - chip basis using flat - fields taken before and after each observation .
the pipeline then branches depending on whether a nod - and - shuffle or classical observing setup was used .
a side effect of the n&s observing is that charge - traps , local defects in the gmos detectors , appear as low - level horizontal stripes in the science observations .
these are removed using a special n&s dark - frame observation .
the dark image reveals the charge - traps , and is used to construct a bad pixel mask ( bpm ) that screens out these areas during image combination .
additionally , between exposures the detector is translated by a few pixels to dilute the effect of the charge - traps on any one part of the ccd .
the next stage ( for both observing modes ) is to locate cosmic rays in the science frames and to add them to the bpm generated from the n&s dark frame .
we use the lacosmic package @xcite , which locates @xmath1699% of cosmic - rays in an image .
next , n&s spectra are sky - subtracted using the gnsskysub iraf routine , while classical spectra are sky - subtracted using a spline function fit along the spatial direction with the science object pixels excluded from the fit .
the resulting frames for each object are combined using an average combine , rejecting charge - traps and cosmic rays identified in the bpms . at this stage a sky frame median - filtered in the spatial direction is added back to the data to ensure the correct variance weighting can be used in the extraction stage .
this leaves cosmetically clean 2-d spectral frames .
the spectra are then extracted using the apall task with variance weighting ( using an appropriate effective gain and readnoise ) , subtracting off the median sky added back in the previous step .
this produces an error spectrum and a science spectrum essentially free from systematic sky subtraction errors , ensuring that careful @xmath2 fits can be made to the extracted spectra ( discussed in section [ chi2 ] ) .
wavelength calibration is then done on the extracted 1-d spectrum using a solution derived from arc lamps taken once per run , and tweaked using the night - sky lines for each observation . the wavelength solution is then applied to the sky - subtracted 2-d data so that galaxy lines may be overplotted ( fig .
[ gallines ] ) .
the final steps are flux calibration on the extracted 1-d spectrum , combined with a telluric correction ( based on standard star spectra ) , and atmospheric extinction correction derived from the effective airmass of observation and the mauna kea extinction curve of @xcite .
the spectra from the 3 chips are then combined into one spectrum , and an error spectrum is simultaneously created .
the chip gaps are effectively weighted to zero so that they have no influence on the @xmath2 fitting .
when possible , the slit was placed through both the sn and the center of the host galaxy . in these cases
the redshift was determined from lines in the host galaxy spectrum . even if the host and the object are completely blended , it is still possible to identify the host galaxy lines as they are narrower than sn features .
sky - subtracted , wavelength calibrated , combined two - dimensional spectra were created for each candidate and the positions of common galaxy spectral features were overplotted , as shown in figure [ gallines ] . the lines used to identify the redshift of each host galaxy are identified in the last column of table [ der - table ] . in some cases the host galaxy was too faint and
the redshift was determined from the sn spectrum .
type ia sne can be identified by a lack of hydrogen in their spectra , combined with broad ( several thousand km s@xmath17 ) p - cyngi lines of elements such as , , , , blends of fe - peak lines , and sometimes , , or .
see for a review of sn classification , and @xcite , , @xcite , and @xcite for particular issues and techniques associated with classification of high - redshift sne .
high - redshift sn spectra are often blended with their host galaxies , so determining the sn type can be a challenge ( at high redshift the galaxies have a smaller angular size , so a greater fraction of the host light is covered by the sn point - spread function ) .
some studies do not attempt to separate sn and host galaxy light , but use a cross - correlation technique to identify sne @xcite .
other authors attempt to separate sn and host galaxy light using point source deconvolution . here
we use a @xmath2 fitting technique to separate sn from host galaxy light and determine the sn type .
it was first developed by @xcite and subsequently used by and @xcite .
each sn spectrum was fit using a @xmath2 matching program which compares the observed spectrum to a library of template sne of all types covering a range of epochs .
the redshift , amount of host galaxy contamination , and reddening are varied to find the best fit . at a given redshift
, the code computes : @xmath19 ^ 2}{\sigma(\lambda)^2}},\ ] ] where @xmath20 is the observed spectrum , @xmath21 is the sn template spectrum , @xmath22 is the host galaxy template spectrum , @xmath23 is the redding law , @xmath24 is the error on the spectrum , and @xmath25 , @xmath26 , and @xmath27 are constants that are varied to find the best fit in host galaxy , template sn , and reddening space .
if the redshift was not fixed , this equation is then reevaluated over a range of redshifts to find the minimum @xmath2 in redshift space .
we use the reddening law of @xcite and @xmath28 .
if the host galaxy spectrum free of sn light could be extracted , then it was used as the galaxy spectrum subtracted in the fitting process .
if no galaxy contamination was evident , then no host was subtracted in the fit . for other cases , template galaxies from @xcite and were subtracted . in cases where the hubble type could be estimated from imaging , the galaxy spectral energy distribution , or narrow galaxy lines ,
the galaxy type was restricted in the fitting procedure .
the sn template library includes 184 sn ia spectra , 75 sn ib / c spectra , and 47 sn ii spectra .
the spectra were chosen to have good s / n , to span the widest possible range of epochs , to have the widest possible wavelength coverage , and to represent all known sn subtypes .
the @xmath2 fitting program produces a list of the best matching sn templates , host galaxy templates , redshifts , and reddening . as in sn detection
, it is not yet possible to fully automate this process a human must still inspect the results and make the ultimate determination of the sn type . the type of the sn is estimated and placed into one of six categories reflecting the type and uncertainty in the classification ( section [ ci ] ; fig . [ specfig ] ) .
the spectroscopic epoch is determined from the weighted mean of the best 5 epochs ( weighted by the @xmath2 of each fit ) .
the average of several epochs was chosen to dilute the effect of outliers , and to help smooth out the diversity in sn ia spectra .
the exact number of epochs averaged has little effect on the result , since the average is weighted by the @xmath2 of each fit the best few fits will be the dominant contributors to the average .
we find @xmath29 for an error of @xmath30 days on the spectroscopic date determination . "
two factors make it harder to classify sne at high redshift as compared to their counterparts at low redshift : they generally have lower signal - to - noise spectra and there is a greater degree of contamination from the host galaxy .
while many classifications are obvious , others have a degree of uncertainty associated with them . to quantify this , after examination of its spectrum we give each sn a sn ia confidence index : 1 .
* certain ia : * the spectrum shows distinctive features of an sn ia such as or . often 6150 and 5400 are redshifted out of the observed spectral range , so 4000 is used as the key indictor for sne ia
. however , at some phases , or for some sn ia subtypes ( e.g. sn 1991 t ) , 4000 would not be expected . in these cases
the candidate can be classified as a category 5 if the spectrum is an exact match to the overall spectral energy distribution ( sed ) of an sn ia ( at the phase indicated by the lightcurve ) and no other type of sn matches .
* highly probable ia : * the spectrum is a match to a ia , but lacks an unambiguous detection of one of the features that is unique to sne ia ( or ) .
other sn types do not match the spectrum well .
these candidates usually have another piece of confirming evidence , such as a lightcurve consistent with a sn ia at the measured redshift , or they are found in an e or s0 galaxy .
* probable ia : * the spectrum matches an sn ia better than any other sn type , but another sn type ( usually sne ic ) is not ruled out from the spectrum alone .
this is either because the spectrum has low s / n or because other sne look similar at the same phase .
these candidates have a lightcurve consistent with an sn ia at the measured redshift , and the spectrum has a phase consistent with the lightcurve .
these sne are denoted as ia * , following the notation of .
* unknown : * the type can not be determined from the spectrum . often the spectrum has low signal - to - noise , or there is too much host galaxy contamination to make a reliable determination .
probably not a ia : * the spectrum has features marginally inconsistent with an sn ia , but the type can not be unambiguously determined .
* not a ia : * the spectral features are inconsistent with a sn ia . in this case the spectrum can usually be identified as an sn ii , sn ib / c , or an agn ( though note that there are no clear cases of agn from gemini spectra these were screened out in advance ) .
the primary means of classification ibfinpstws the sn spectrum , although we use all information available to supplement this including the lightcurves , colors , and the agreement between lightcurve phase and the spectroscopic phase determined from the fitting process .
for example , if the type can not be determined from the spectrum it would normally be classified as index 2 .
but if the lightcurves are also inconsistent with the lightcurve of an sn ia , then it would be moved into index 1 or 0 .
the host galaxy spectrum is never used on its own to classify a candidate , but if the host is clearly an elliptical galaxy , then this information may be used in conjunction with the candidate spectrum and the lightcurve to solidify the status of a candidate as a probable sn ia .
we emphasize that the lightcurve alone is never used to classify an sn ia for a candidate to have a classification of sn ia or sn ia * it must have a spectrum matching with the expected spectral energy distribution of an sn ia at the phase determined from the lightcurve .
one representative spectrum from each category is shown in figure [ specfig ] .
all spectra discussed in this paper are available in the online version of this article .
the snls officially started in june 2003 ( there was a presurvey ramp - up to full operations ) , and we began using gemini - n and s to observe sn candidates in august 2003 ( semester 2003b ) .
we usually obtain the gemini data within a day or two of the observations and produce `` real - time '' reductions . after the full calibration data is released at the end of a run the data are rereduced . here
we report on the final spectroscopic reductions through november 2004 .
all of the spectra used in this study are available in the online version of the paper .
table [ der - table ] lists properties derived from the observations of each candidate , such as type , redshift , and epoch relative to maximum light .
table [ class - host - table ] shows the distribution of sn types with respect to host galaxy type . while it is not always possible to extract the host galaxy separately and examine its sed , galaxy lines are apparent because they are narrower than sn features .
therefore we group galaxies into absorption - line galaxies and emission - line galaxies ( if a galaxy has any emission lines it is considered an emission - line galaxy ) .
just as at low redshift , in our sample core - collapse sne are never seen in absorption - line ( early type ) galaxies .
figure [ hist ] is a histogram of the number of candidates of each different type with redshift . for clarity
we group index 0 and 1 sne together as `` not ia '' and index 4 and 5 sne together as `` ia . ''
sne with weaker classifications ( index 3 ) are plotted separately as `` ia*. '' candidates that could not be identified ( index 2 ) , but whose redshifts could be determined , are also shown .
it is apparent that we targeted the highest redshift observations at gemini the median redshift of sne ia / ia * is 0.81 . as the redshift increases the fraction of less secure identifications rises , because for faint sne it is hard to achieve a high signal - to - noise ratio in a reasonable integration time .
furthermore , at the higher redshifts one must rely more on the rest - frame uv light to classify the sne , and template uv observations of all sn types are scarce . core collapse sne cluster at lower redshifts on the histogram because they are usually intrinsically fainter .
for example , a candidate with a peak magnitude of @xmath31 could be an sn ia at @xmath32 or could be a type ii sn at a lower redshift .
the technique of the sn photo - z is introduced in section [ photozintro ] and described in much greater detail in @xcite .
here we report the result of its application on our sn ia confirmation rate .
the snls implemented the sn photo - z and improved real - time photometry in march 2004 , although in two cases after this ( 04d1dr and 04d4 ft ) , candidates had to be observed spectroscopically before a photo - z could be obtained .
when no photo - z information was available before selecting candidates for spectroscopy , 14/26 ( 54% ) of the candidates were confirmed as sne ia close to previosly published rates .
however , when we did have an sn photo - z to guide the decision making , the ia confirmation rate jumped to 27/38 ( 71% ) . there will be always be some unidentified candidates for which it is difficult to estimate a type from spectroscopy these may still be sne ia , but are too buried in a host , or have a spectrum with too low s / n , to be identified . however , where the photo - z excels is in its power to reject candidates that are not sne ia . with no prior photo - z information , 7/26 ( 27% ) of candidates with gemini spectroscopy
were found to be certainly not or probably not sne ia .
after implementation of the technique , only 3/38 ( 8% ) of the observations were certainly not or probably not sne ia .
another benefit of the sn photo - z is that it provides a prediction of the time of maximum light for an sn ia , allowing observations to be scheduled to within a few days of this date for maximum efficiency and minimal host galaxy contamination .
figure [ hiepoch ] shows the distribution of the sne ia observed by gemini with respect to maximum light .
ninety percent of sne ia were observed within 0.5 mag of maximum light , and over half of the sne ia were observed within 0.1 mag of maximum light .
it is clear that the flexibility provided by queue observing plays a large role in optimizing the efficiency of the spectroscopic classification of targets in the snls .
note that it is not always possible to schedule observations at maximum light because we require dark time to observe such faint targets , and gmos is not always on the telescope and available in queue mode .
figure [ hiepoch ] also shows that the least solid classifications ( index 3 ) , labeled ia * , were usually observed after maximum light .
sne before maximum were almost always classified with more certainty .
this is partially because after maximum light , especially around + 7 to + 10 days after max , it is often difficult to distinguish between sne ia and sne ic . by this time
certain distinguishing sn ia features , such as 4000 , may no longer be apparent , and the sn ia line velocities have decreased to the range more typical of those found in sne ic . the more uncertain classifications ( sne ia * ) often occur near maximum light , because the most difficult sne ( those at the highest redshift or with the most host galaxy contamination ) can only be observed near maximum light . before or after maximum light
they are so faint that they would not be placed in the spectroscopic observing queue .
an added benefit to obtaining early spectroscopy is that sne ia show the greatest diversity at early times @xcite , when the spectroscopy is probing the outer layers of supernova .
figure [ hiepoch ] shows that one compromise would be to target observations at about one week before maximum light . at @xmath33d ,
a typical ia is only about 0.25 fainter than at peak , near enough to peak brightness to make the observations feasible . at the same time it would provide an opportunity to observe sne ia when they show the greatest diversity and
also are the most distinct from sne ic .
this window of opportunity is very narrow , however . at @xmath34d
a typical sn ia is 0.75 mag fainter than at peak in the restframe b - band still too faint compared to its host galaxy , and by @xmath35d sn ia spectra have lost some of their diversity .
figure [ imagpi ] shows the i magnitude at the time of spectroscopy versus the percentage increase in i brightness in a 6 pixel ( 1.12 ) diameter at the time of spectroscopy .
the percentage increase is measured relative to the reference image , where there is no supernova light .
the i magnitude and percentage increase were measured from cfht images , interpolated to the time of spectroscopy .
typically candidates were sent to gemini only if they were in the magnitude range @xmath36 .
( there are some exceptions , especially in the d3 field , which can not be seen by the vlt . )
it is apparent from figure [ imagpi ] that when the sn signal is greater than that of the host galaxy , candidate identification is relatively easy when the percentage increase is greater than 100% the candidates were not identified only 7% ( 2/30 ) of the time .
clearly , image quality ( iq ) plays a role in whether or not sne can be successfully identified in the presence of significant host galaxy light , as shown in figure [ iqpi ] .
we plot the image quality at the time of spectroscopy ( determined from the gemini acquisition image ) against the percent increase ( determined from cfht images as described above ) .
if the iq was better than 0.55@xmath37 , candidates were identified 88% of the time . in good seeing the sn light
is more concentrated and can be extracted in a narrow aperture , even in the presence of host contamination .
the @xmath2 matching of the sn spectrum against a host and sn spectral template library produces more identifications , more robust identifications , and greater coverage of parameter space than traditional methods ( unaided expert matching by eye ) .
this method works optimally with spectra that are free from systematic deviations , and whose errors are well characterized .
it works exceptionally well with gmos nod and shuffle data , where systematic effects associated with sky subtraction are almost completely removed . correcting the spectra for telluric features also helps , as does a carefully generated error spectrum .
while both the spectra in this paper and in were classified by the same person ( dah ) , here the @xmath2 matching software is upgraded with more template supernova and galaxy spectra . were not able to identify any candidates with a percentage increase below 25% ,
but in this work we have successfully identified several candidates with percent increases between 15%25% .
it is unclear if this is due to software improvements , observing conditions , differences in the spectrographs used , or a better selection of candidates .
figure [ date ] also shows that the spectroscopic epoch determined by the fitting program matches well with the epoch determined from the lightcurve . unlike the spectral feature age technique @xcite , it is important to note that this program is not specifically tuned to determine the epoch of an sn ia , and it makes no assumptions that the input spectrum is a ia ( it is possible for it to determine the epoch of an sn ib / c for example , although this has not been extensively tested ) .
we find the remarkable result that our spectroscopic fitting technique can determine the epoch to within 2.5 days , despite the low s / n and significant host contamination in this data set .
before comparing these confirmation rates to previous work , a few caveats are necessary . first , not all spectroscopic follow - up programs have as their goal the confirmation of candidates as sne ia .
one may wish to determine the types of all transients to have a better determination of sn rates . or in some cases the goal may be to spectroscopically identify sne ii , as they also have cosmological utility .
while we have pursued these goals at other telescopes , the role of gemini has been largely to confirm sne ia .
second , different programs may have different criteria for what is called an sn ia / ia*. for example , we have lightcurve infomation available at the time of classification , which was not always the case in previous studies . still , our criteria are as close as possible to the criteria used in other published works .
we adopt similar criteria to that of @xcite , and we are using the the same human classifier ( dah ) and classification program ( albeit slightly upgraded ) as . despite these caveats ,
it is important to have some comparison to previous work .
a comparison to is problematic because of the inhomogeneous nature of the sample and techniques presented there .
they present results from six separate searches , some classical , and some rolling , some targeting @xmath38 , and some targeting @xmath39 .
the most appropriate comparison to this work is to consider all searches in lidman et al .
that targeted @xmath40 . in this case
the sn ia / ia * confirmation rate was 62% ( median sn ia @xmath41 ) . a similar sn ia confirmation rate is found if the lidman et al .
search 4 ( the rolling search ) is excluded , but sne followed by the scp at all telescopes for @xmath40 are included . since rolling searches
find all sne above some magnitude cutoff , the percentage of sne ia in the sample is lower than in a classical search . in their first two years of operation , essence @xcite , had a 43% sn ia / ia ?
confirmation rate with a median redshift of 0.43 [ though in the third year the sn ia / ia ?
confirmation rate rose to @xmath1060% ( matheson , private communication ) ] .
again , a direct comparison to the results presented here may not be appropriate if the two searches have different goals or different criteria for counting sne ia , but does make the point that if the goal is to maximize the yield of sne ia , a method like the sn photo - z is helpful . despite the difficulties in comparing this study to previous ones
as noted above , we can draw some broad conclusions from the results .
first , to make the most efficient use of 810 m time it is helpful to have as much information as possible before sending a target to the telescope . in a rolling search ,
more information is available at follow - up time than in a classical search , so new techniques are called for . here
we have demonstrated that by fitting the available data before spectroscopy we can improve the sn ia / ia * confirmation rate over our rates when the technique was not applied .
our sn ia / ia * confirmation rate ( 71% ) is also an increase over comparable previously published rates ( 43%-62% ) , despite being at a much higher median redshift ( z=0.81 vs. z=0.4 - 0.5 ) . by making it possible to separate likely sne ia from likely core collapse sne before a large amount of telescope time
is invested in spectroscopy , the techniques demonstrated here improve the success rate of all follow - up programs , no matter their goal .
over the five - year lifetime of the snls , the these techniques will result in @xmath3 more confirmed sne ia .
we aim to spectroscopically observe @xmath42 sn candidates during the survey . in the absence of the photo - z
this would result in @xmath10 540 sne ia , but with the photo - z we could achieve 710 sne ia .
this is equivalent to adding another year to the survey .
we have demonstrated several techniques for improving the yield of spectroscopically identified sne ia at high redshift ( @xmath43 ) .
the sn photo - z is shown to effectively screen out non - ia candidates before they are observed spectroscopically . when no photometric redshifts were available , 35% of candidates turned out to be probably or certainly not sne ia after spectroscopy ( ci 0 or 1 ) .
with photo - z information , the non - ia `` contamination '' rate dropped to 8% . using the photo - z
we show that we can schedule sn observations to within a few tenths of a magnitude of maximum light , and that the optimal phase for sn ia identifcation and diversity is @xmath33 days .
after spectroscopy , we show that by @xmath2 fitting of template sne we can effectively subtract host light and determine the type of an sn when the sn is only 15% as bright as the host in some cases . for targets where the sn is at least as bright as the underlying host , or
when the image quality is exceptional ( better than 0.55 ) , the candidate is identified more than 90% of the time .
using @xmath2 fitting we can also obtain an independent measurement of the spectroscopic epoch that agrees well with the phase determined from the lightcurve .
these techniques have been developed using the first year s observations of the highest redshift candidates of the snls at gemini north and south .
of the candidates observed at gemini , 41/64 are certain or probable sne ia .
this is roughly one - third of the spectroscopic follow - up program of the snls .
the techniques outlined here will add @xmath3 more confirmed sne ia over the five - year project to discover , confirm , and follow @xmath44 sne ia to measure the equation of state of dark energy .
the snls collaboration gratefully acknowledges the assistance of pierre martin and the cfht queued service observations team .
jean - charles cuillandre and kanoa withington were also indispensable in making possible real - time data reduction at cfht .
we also thank gemini queue observers and support staff , especially inger jrgensen , kathy roth , percy gomez , and marcel bergmann , for both taking the data presented in this paper and making observations available quickly .
canadian collaboration members acknowledge support from nserc and ciar ; french collaboration members from cnrs / in2p3 , cnrs / insu and cea ; portuguese collaboration members acknowledge support from fct - fundao para a cincia e tecnologia .
snls relies on observations with megacam , a joint project of cfht and cea / dapnia , at the canada - france - hawaii telescope ( cfht ) which is operated by the national research council ( nrc ) of canada , the institut national des science de lunivers of the centre national de la recherche scientifique ( cnrs ) of france , and the university of hawaii .
this work is based in part on data products produced at the canadian astronomy data centre as part of the canada - france - hawaii telescope legacy survey , a collaborative project of the national research council of canada and the french centre national de la recherche scientifique .
this work is also based on observations obtained at the gemini observatory , which is operated by the association of universities for research in astronomy , inc .
, under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the particle physics and astronomy research council ( united kingdom ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , cnpq ( brazil ) and conicet ( argentina ) .
this research used observations from gemini program numbers : gn-2004b - q-16 , gs-2004b - q-31 , gn-2004a - q-19 , gs-2004a - q-11 , gn-2003b - q-9 , and gs-2003b - q-8 .
, a. l. , matheson , t. , filippenko , a. v. , leonard , d. c. , tonry , j. , riess , a. g. , challis , p. , clocchiatti , a. , garnavich , p. m. , hogan , c. j. , jha , s. , kirshner , r. p. , leibundgut , b. , phillips , m. m. , schmidt , b. p. , schommer , r. a. , smith , r. c. , soderberg , a. m. , spyromilio , j. , stubbs , c. , suntzeff , n. b. , & woudt , p. 2000 , , 544
, l111 , c. , howell , d. a. , folatelli , g. , garavini , g. , nobili , s. , aldering , g. , amanullah , r. , antilogus , p. , astier , p. , blanc , g. , burns , m. s. , conley , a. , deustua , s. e. , doi , m. , ellis , r. , fabbro , s. , fadeyev , v. , gibbons , r. , goldhaber , g. , goobar , a. , groom , d. e. , hook , i. , kashikawa , n. , kim , a. g. , knop , r. a. , lee , b. c. , mendez , j. , morokuma , t. , motohara , k. , nugent , p. e. , pain , r. , perlmutter , s. , prasad , v. , quimby , r. , raux , j. , regnault , n. , ruiz - lapuente , p. , sainton , g. , schaefer , b. e. , schahmaneche , k. , smith , e. , spadafora , a. l. , stanishev , v. , walton , n. a. , wang , l. , wood - vasey , w. m. , & yasuda ( the supernova cosmology project ) , n. 2005 , , 430 , 843 , t. , blondin , s. , foley , r. j. , chornock , r. , filippenko , a. v. , leibundgut , b. , smith , r. c. , sollerman , j. , spyromilio , j. , kirshner , r. p. , clocchiatti , a. , aguilera , c. , barris , b. , becker , a. c. , challis , p. , covarrubias , r. , garnavich , p. , hicken , m. , jha , s. , krisciunas , k. , li , w. , miceli , a. , miknaitis , g. , prieto , j. l. , rest , a. , riess , a. g. , salvo , m. e. , schmidt , b. p. , stubbs , c. w. , suntzeff , n. b. , & tonry , j. l. 2005 , , 129 , 2352 , s. , aldering , g. , goldhaber , g. , knop , r. a. , nugent , p. , castro , p. g. , deustua , s. , fabbro , s. , goobar , a. , groom , d. e. , hook , i. m. , kim , a. g. , kim , m. y. , lee , j. c. , nunes , n. j. , pain , r. , pennypacker , c. r. , quimby , r. , lidman , c. , ellis , r. s. , irwin , m. , mcmahon , r. g. , ruiz - lapuente , p. , walton , n. , schaefer , b. , boyle , b. j. , filippenko , a. v. , matheson , t. , fruchter , a. s. , panagia , n. , newberg , h. j. m. , couch , w. j. , & the supernova cosmology project . 1999 , , 517 , 565 , s. , pennypacker , c. r. , goldhaber , g. , goobar , a. , muller , r. a. , newberg , h. j. m. , desai , j. , kim , a. g. , kim , m. y. , small , i. a. , boyle , b. j. , crawford , c. s. , mcmahon , r. g. , bunclark , p. s. , carter , d. , irwin , m. j. , terlevich , r. j. , ellis , r. s. , glazebrook , k. , couch , w. j. , mould , j. r. , small , t. a. , & abraham , r. g. 1995 , , 440 , l41 , a. g. , filippenko , a. v. , challis , p. , clocchiatti , a. , diercks , a. , garnavich , p. m. , gilliland , r. l. , hogan , c. j. , jha , s. , kirshner , r. p. , leibundgut , b. , phillips , m. m. , reiss , d. , schmidt , b. p. , schommer , r. a. , smith , r. c. , spyromilio , j. , stubbs , c. , suntzeff , n. b. , & tonry , j. 1998 , , 116 , 1009 , a. g. , filippenko , a. v. , leonard , d. c. , schmidt , b. p. , suntzeff , n. , phillips , m. m. , schommer , r. , clocchiatti , a. , kirshner , r. p. , garnavich , p. , challis , p. , leibundgut , b. , spyromilio , j. , & smith , r. c. 1997 , , 114 , 722 , a. g. , strolger , l. , tonry , j. , tsvetanov , z. , casertano , s. , ferguson , h. c. , mobasher , b. , challis , p. , panagia , n. , filippenko , a. v. , li , w. , chornock , r. , kirshner , r. p. , leibundgut , b. , dickinson , m. , koekemoer , a. , grogin , n. a. , & giavalisco , m. 2004 , , 600 , l163 llllrclrrr 03d1as & 02:24:24.520 & -04:21:40.19 & 2003 - 09 - 27 & 6000 & n+s & 720 & 0.41 & 23.96 & 1475 + 03d1ax & 02:24:23.320 & -04:43:14.41 & 2003 - 09 - 29 & 2400 & c & 720 & 0.61 & 23.16 & 65 + 03d1bk & 02:26:27.410 & -04:32:11.99 & 2003 - 09 - 28 & 4800 & n+s & 720 & 0.47 & 23.95 & 83 + 03d1cj & 02:26:25.081 & -04:12:39.89 & 2003 - 10 - 26 & 5400 & n+s & 720 & 0.57 & 24.11 & 76 + 03d1 cm & 02:24:55.288 & -04:23:03.68 & 2003 - 10 - 27 & 5400 & n+s & 720 & 0.72 & 23.77 & 970 + 03d1co & 02:26:16.238 & -04:56:05.76 & 2003 - 11 - 01 & 7200 & n+s & 720 & 0.84 & 23.64 & 188 + 03d1ew & 02:24:14.088 & -04:39:56.98 & 2003 - 12 - 21 & 7200 & n+s & 720 & 0.71 & 23.76 & 1661 + 03d1fp & 02:26:03.073 & -04:08:02.02 & 2003 - 12 - 26 & 7200 & n+s & 720 & 0.55 & 23.86 & 30 + 03d1fq & 02:26:55.683 & -04:18:08.10 & 2003 - 12 - 24 & 5400 & n+s & 720 & 0.48 & 23.59 & 41 + 03d4cj & 22:16:06.660 & -17:42:16.72 & 2003 - 08 - 26 & 2700 & c & 680 & 0.83 & 21.85 & 10000 + 03d4ck & 22:15:08.910 & -17:56:02.17 & 2003 - 08 - 27 & 2400 & c & 680 & 0.46 & 22.78 & 1905 + 03d4cn & 22:16:34.600 & -17:16:13.55 & 2003 - 08 - 27 & 4800 & c & 720 & 0.46 & 23.81 & 49 + 03d4cy & 22:13:40.460 & -17:40:53.90 & 2003 - 09 - 26 & 5400 & n+s & 720 & 0.59 & 24.19 & 149 + 03d4cz & 22:16:41.870 & -17:55:34.54 & 2003 - 09 - 27 & 3600 & c & 720 & 0.41 & 24.41 & 15 + 03d4fd & 22:16:14.471 & -17:23:44.37 & 2003 - 10 - 24 & 3600 & n+s & 720 & 0.67 & 23.61 & 321 + 03d4fe & 22:16:08.844 & -17:55:19.21 & 2003 - 10 - 24 & 3600 & n+s & 720 & 0.48 & 23.58 & 66 + 03d4gl & 22:14:44.177 & -17:31:44.47 & 2003 - 10 - 29 & 3600 & n+s & 720 & 0.63 & 23.43 & 121 + 04d1de & 02:26:35.925 & -04:25:21.65 & 2004 - 08 - 17 & 7200 & n+s & 720 & 0.51 & 23.63 & 56 + 04d1dr & 02:27:23.905 & -04:51:27.43 & 2004 - 08 - 14 & 5400 & n+s & 720 & 0.57 & 24.28 & 72 + 04d1hd & 02:26:08.850 & -04:06:35.22 & 2004 - 09 - 13 & 2400 & c & 680 & 0.85 & 22.16 & 1254 + 04d1ho & 02:24:44.856 & -04:39:15.55 & 2004 - 09 - 16 & 3600 & c & 720 & 0.59 & 23.13 & 24 + 04d1hy & 02:24:08.678 & -04:49:52.22 & 2004 - 09 - 11 & 5400 & n+s & 720 & 0.73 & 23.53 & 996 + 04d1jf & 02:25:18.914 & -04:49:09.05 & 2004 - 10 - 13 & 2400 & c & 680 & 0.92 & 22.94 & 19 + 04d1ln & 02:25:53.482 & -04:27:03.75 & 2004 - 10 - 17 & 2400 & c & 680 & 0.62 & 22.80 & 29 + 04d1ow & 02:26:42.708 & -04:18:22.55 & 2004 - 11 - 08 & 5400 & n+s & 720 & 0.69 & 23.87 & 4201 + 04d2aa & 10:02:02.100 & + 02:40:51.76 & 2004 - 01 - 23 & 7320 & n+s & 720 & 0.58 & 24.19 & + 04d2ad & 10:00:08.093 & + 02:39:01.40 & 2004 - 01 - 22 & 7320 & n+s & 720 & 0.54 & 24.17 & + 04d2ae & 10:01:52.414 & + 02:13:21.11 & 2004 - 01 - 21 & 5490 & n+s & 720 & 0.75 & 23.64 & + 04d3aa & 14:16:49.935 & + 52:45:31.12 & 2004 - 01 - 30 & 5400 & n+s & 720 & 1.02 & 24.05 & 12 + 04d3ae & 14:22:21.569 & + 52:21:39.21 & 2004 - 01 - 25 & 2400 & c & 680 & 0.94 & 23.10 & 196 + 04d3ax & 14:22:39.072 & + 52:51:52.57 & 2004 - 01 - 28 & 5400 & n+s & 720 & 1.30 & 24.95 & 246 + 04d3bf & 14:17:45.096 & + 52:28:04.31 & 2004 - 02 - 17 & 2700 & c & 680 & & 23.16 & + 04d3dd & 14:17:48.431 & + 52:28:14.72 & 2004 - 04 - 25 & 5400 & n+s & 720 & 0.59 & 24.06 & 375 + 04d3de & 14:22:13.503 & + 52:17:09.71 & 2004 - 04 - 27 & 7200 & n+s & 720 & 0.57 & 24.01 & 651 + 04d3fj & 14:19:50.703 & + 52:41:31.84 & 2004 - 04 - 28 & 7200 & n+s & 720 & 0.72 & 24.19 & 108 + 04d3fq & 14:16:57.906 & + 52:22:46.53 & 2004 - 04 - 26 & 5400 & n+s & 720 & 0.97 & 23.04 & 412 + 04d3gu & 14:22:07.359 & + 52:38:54.60 & 2004 - 05 - 22 & 4800 & c & 720 & 1.03 & 22.60 & 6 + 04d3gx & 14:20:13.678 & + 52:16:58.60 & 2004 - 05 - 21 & 7200 & n+s & 720 & 0.40 & 24.38 & 2439 + 04d3hn & 14:22:06.878 & + 52:13:43.46 & 2004 - 05 - 22 & 4800 & c & 720 & 0.47 & 22.94 & 14 + 04d3kr & 14:16:35.937 & + 52:28:44.20 & 2004 - 06 - 16 & 2400 & c & 680 & 0.82 & 21.60 & 156 + 04d3lp & 14:19:50.927 & + 52:30:11.85 & 2004 - 05 - 27 & 5400 & n+s & 720 & 0.52 & 24.45 & 78 + 04d3lu & 14:21:08.009 & + 52:58:29.74 & 2004 - 06 - 23 & 3600 & n+s & 720 & 0.84 & 23.49 & 17 + 04d3mk & 14:19:25.830 & + 53:09:49.56 & 2004 - 06 - 19 & 4320 & n+s & 720 & 1.25 & 23.27 & 86 + 04d3ml & 14:16:39.107 & + 53:05:35.66 & 2004 - 06 - 20 & 3600 & n+s & 720 & 0.41 & 23.90 & 2252 + 04d3nc & 14:16:18.224 & + 52:16:26.09 & 2004 - 07 - 13 & 2400 & c & 720 & 0.66 & 23.39 & 149 + 04d3nh & 14:22:26.729 & + 52:20:00.92 & 2004 - 06 - 23 & 1800 & c & 680 & 0.80 & 21.66 & 199 + 04d3nq & 14:20:19.193 & + 53:09:15.90 & 2004 - 07 - 14 & 1500 & c & 680 & 0.61 & 21.03 & 10000 + 04d3nr & 14:22:38.526 & + 52:38:55.89 & 2004 - 07 - 15 & 7200 & n+s & 720 & 0.69 & 24.40 & 2300 + 04d3ny & 14:18:56.332 & + 52:11:15.06 & 2004 - 07 - 10 & 5400 & n+s & 720 & 0.79 & 23.23 & 203 + 04d3oe & 14:19:39.381 & + 52:33:14.21 & 2004 - 07 - 11 & 3600 & n+s & 720 & 0.62 & 23.21 & 28 + 04d3og & 14:20:39.748 & + 53:01:15.02 & 2004 - 07 - 19 & 2700 & c & 720 & 1.10 & 21.81 & 151 + 04d3pd & 14:22:33.506 & + 52:13:47.77 & 2004 - 07 - 18 & 3600 & n+s & 720 & 0.83 & 23.34 & 72 + 04d4 dm & 22:15:25.470 & -17:14:42.71 & 2004 - 07 - 18 & 3600 & n+s & 720 & 0.96 & 23.55 & 123 + 04d4ec & 22:16:29.286 & -18:11:04.13 & 2004 - 07 - 19 & 3600 & n+s & 720 & 0.70 & 23.08 & 37 + 04d4 ft & 22:14:31.097 & -17:40:19.74 & 2004 - 08 - 12 & 3600 &
c & 720 & 0.67 & 23.21 & 78 + 04d4gg & 22:16:09.268 & -17:17:39.98 & 2004 - 08 - 16 & 3600 & c & 720 & 0.81 & 23.03 & 30 + 04d4hu & 22:15:36.193 & -17:50:19.81 & 2004 - 09 - 18 & 5400 & n+s & 720 & 0.52 & 23.45 & 84 + 04d4hx & 22:13:40.587 & -17:23:03.35 & 2004 - 09 - 16 & 5400 & n+s & 720 & 0.73 & 24.15 & 25 + 04d4ic & 22:14:21.841 & -17:56:36.43 & 2004 - 09 - 12 & 5160 & n+s & 720 & 0.69 & 23.34 & 231 + 04d4ih & 22:17:17.041 & -17:40:38.74 & 2004 - 10 - 07 & 5400 & n+s & 720 & 0.52 & 23.95 & 41 + 04d4ii & 22:15:55.645 & -17:39:27.09 & 2004 - 09 - 15 & 7200 & n+s & 720 & 0.45 & 23.86 & 57 + 04d4im & 22:15:00.885 & -17:23:45.84 & 2004 - 10 - 10 & 7200 & n+s & 720 & 0.44 & 23.23 & 18 + 04d4jy & 22:13:51.605 & -17:24:18.13 & 2004 - 10 - 14 & 8881 & n+s & 720 & 0.77 & 24.08 & 1779 + 04d4kn & 22:15:04.324 & -17:19:45.05 & 2004 - 10 - 19 & 5400 & n+s & 720 & 0.60 & 23.74 & 97 + llllcrrrll 03d1as & 0.872 & 0.001 & sn : & 2 & & & & : + 03d1ax & 0.496 & 0.001 & sn ia & 5 & -3.0 & 0.1 & -3.3 & h&k + 03d1bk & 0.8650 & 0.0005 & sn ia & 5 & -6.3 & 0.1 & -2.1 & h&k , h@xmath45 , h@xmath46 + 03d1cj & 0.364 & 0.001 & sn ib / c : & 1 & & & & h@xmath47 , + 03d1 cm & 0.87 & 0.02 & sn ia & 4 & -5.0 & 0.3 & 2.4 & sn + 03d1co & 0.68 & 0.01 & sn ia & 5 & -4.8 & 0.3 & -1.2 & sn + 03d1ew & 0.868 & 0.001 & sn ia & 5 & 0.9 & 0.4 & -0.2 & + 03d1fp & 0.270 & 0.001 & sn iib & 0 & & & & h@xmath47 , h@xmath45 , , , + 03d1fq & 0.80 & 0.02 & sn ia & 4 & -1.6 & 0.3 & 2.2 & sn + 03d4cj & 0.27 & 0.01 & sn ia & 5 & -9.7 & 0.1 & -6.9 & sn + 03d4ck & 0.189 & 0.001 & sn iin & 0 & & & & sn + 03d4cn & 0.818 & 0.001 & sn ia & 4 & -0.5 & 0.6 & 2.8 & , , h@xmath45 + 03d4cy & 0.9271 & 0.0005 & sn ia & 4 & 5.3 & 0.3 & 7.7 & + 03d4cz & 0.695 & 0.001 & sn ia & 4 & 9.4 & 0.1 & 8.7 & h&k , g - band + 03d4fd & 0.791 & 0.003 & sn ia & 5 & -1.6 & 0.3 & -1.5 & + 03d4fe & ? & & sn : & 1 & & & & + 03d4gl & 0.56 & 0.01 & sn ia & 4 & -9.0 & 0.2 & -6.0 & sn + 04d1de & 0.7677 & 0.0002 & sn ia * & 3 & -6.8 & 0.1 & -4.5 & , , h@xmath45 , poss h&k + 04d1dr & 0.6414 & 0.0003 & sn : & 2 & & & & , , h@xmath45 , h@xmath46 + 04d1hd & 0.3685 & 0.0005 & sn ia & 5 & -4.8 & 0.1 & -4.0 & , weak + 04d1ho & 0.7012 & 0.0004 & sn & 2 & & & & , , h&k , h@xmath45 , h@xmath46 + 04d1hy & 0.85 & 0.02 & sn ia & 5 & -3.5 & 0.3 & -3.2 & sn + 04d1jf & 0.3800 & 0.0002 & sn : & 2 & & & & , , h@xmath45 + 04d1ln & 0.2072 & 0.0002 & sn ii - p & 0 & & & & h@xmath47 , , + 04d1ow & 0.93 & 0.02 & sn ia & 4 & 2.6 & 0.2 & -0.8 & sn + 04d2aa & ? & & sn : & 2 & & & & + 04d2ad & 0.6802 & 0.0002 & sn : & 2 & & & & , , h@xmath45 + 04d2ae & 0.843 & 0.001 & sn ia & 4 & 0.0 & 0.1 & -1.5 & h&k + 04d3aa & 0.2045 & 0.0002 & sn ii & 0 & & & & h@xmath47 , h@xmath45 , , + 04d3ae & 0.217 & 0.001 & sn ii & 0 & & & & h@xmath47 , + 04d3ax & 0.3558 & 0.0002 & sn ii : & 1 & & & & h@xmath47 , h@xmath45 , + 04d3bf & 0.1560 & 0.0005 & sn ia & 5 & & & 15.1 & , , , h@xmath47 , h@xmath45 + 04d3dd & 1.01 & 0.02 & sn ia & 4 & 2.9 & 0.3 & -1.4 & sn + 04d3de & ? & & sn ii - p & 0 & & & & + 04d3fj & ? & & sn : & 1 & & & & + 04d3fq & 0.73 & 0.01 & sn ia & 4 & 0.8 & 0.3 & 3.6 & sn + 04d3gu & 0.748 & 0.001 & sn : & 2 & & & & h&k , balmer , weak + 04d3gx & 0.91 & 0.02 & sn ia * & 3 & 10.8 & 0.2 & 8.3 & sn + 04d3hn & 0.5516 & 0.0003 & sn ia & 5 & 7.2 & 0.1 & 4.9 & h&k , balmer + 04d3kr & 0.3373 & 0.0002 & sn ia & 5 & 4.4 & 0.1 & 1.7 & , , h@xmath45 + 04d3lp & 0.983 & 0.001 & sn ia * & 3 & 1.0 & 0.2 & -1.2 & + 04d3lu & 0.8218 & 0.0002 & sn ia & 4 & 5.5 & 0.1 & 5.6 & h&k + 04d3mk & 0.813 & 0.001 & sn ia & 5 & -2.4 & 0.1 & 1.2 & , h&k + 04d3ml & 0.95 & 0.02 & sn ia & 4 & -1.8 & 0.3 & 1.1 & sn + 04d3nc & 0.817 & 0.001 & sn ia * & 3 & 7.2 & 0.2 & & poss + 04d3nh & 0.3402 & 0.0002 & sn ia & 5 & 3.5 & 0.1 & 3.7 & h@xmath47 , h@xmath45 , , poss h&k + 04d3nq & 0.22 & 0.01 & sn ia & 5 & 8.1 & 0.1 & 8.8 & sn + 04d3nr & 0.96 & 0.02 & sn ia * & 3 & 8.1 & 0.3 & 3.9 & sn + 04d3ny & 0.81 & 0.02 & sn ia & 5 & 1.6 & 0.2 & 2.5 & sn + 04d3oe & 0.756 & 0.001 & sn ia & 4 & 1.4 & 0.1 & 1.8 & h&k + 04d3og & 0.352 & 0.001 & sn & 2 & & & & h@xmath47 , h@xmath45 , , , + 04d3pd & 0.760 & 0.001 & sn : & 2 & & & & h@xmath45 , , + 04d4 dm & 0.811 & 0.001 & sn ia & 4 & 2.7 & 0.2 & -0.7 & , + 04d4ec & 0.593 & 0.001 & sn : & 2 & & & & , , h@xmath45 , h@xmath46 + 04d4 ft & 0.2666 & 0.0002 & sn & 2 & & & & , h@xmath47 , h@xmath45 + 04d4gg & 0.4238 & 0.0004 & sn ia & 5 & -9.7 & 0.1 & -10.0 & , , h@xmath47 , h@xmath45 , h@xmath46 + 04d4hu & 0.7027 & 0.0003 & sn ia & 5 & 5.2 & 0.5 & 5.0 & , , h@xmath45 + 04d4hx & 0.545 & 0.005 & sn : & 2 & & & & 4000 break , poss h&k + 04d4ic & 0.68 & 0.02 & sn ia & 4 & 2.5 & 0.3 & 4.8 & sn + 04d4ih & 0.934 & 0.001 & sn ia * & 3 & 8.7 & 0.2 & 5.6 & , h&k , some balmer + 04d4ii & 0.866 & 0.001 & sn ia & 5 & -4.6 & 0.2 & -3.5 & + 04d4im & 0.7510 & 0.0005 & sn ia & 4 & 0.7 & 0.2 & 0.2 & h&k , h@xmath48 + 04d4jy & 0.93 & 0.02 & sn ia * & 3 & 2.0 & 0.4 & 0.0 & sn + 04d4kn & 0.9095 & 0.0005 & sn : & 2 & & & & , h&k , balmer + | we present new techiques for improving the efficiency of supernova ( sn ) classification at high redshift using 64 candidates observed at gemini north and south during the first year of the supernova legacy survey ( snls ) .
the snls is an ongoing five year project with the goal of measuring the equation of state of dark energy by discovering and following over 700 high - redshift sne ia using data from the canada - france - hawaii telescope legacy survey .
we achieve an improvement in the sn ia spectroscopic confirmation rate : at gemini 71% of candidates are now confirmed as sne ia , compared to 54% using the methods of previous surveys .
this is despite the comparatively high redshift of this sample , where the median sn ia redshift is @xmath0 ( @xmath1 ) .
these improvements were realized because we use the unprecedented color coverage and lightcurve sampling of the snls to predict whether a candidate is an sn ia and estimate its redshift , before obtaining a spectrum , using a new technique called the `` sn photo - z . '' in addition , we have improved techniques for galaxy subtraction and sn template @xmath2 fitting , allowing us to identify candidates even when they are only 15% as bright as the host galaxy . the largest impediment to sn identification
is found to be host galaxy contamination of the spectrum when the sn was at least as bright as the underlying host galaxy the target was identified more than 90% of the time . however
, even sne on bright host galaxies can be easily identified in good seeing conditions .
when the image quality was better than 0.55 , the candidate was identified 88% of the time . over the five - year course of the survey , using the selection techniques presented here we will be able to add @xmath3 more confirmed sne ia than would be possible using previous methods . |
initiated by work of wegner @xcite , schfer & wegner @xcite , and pruisken & schfer @xcite , non - compact non - linear sigma models as well as their supersymmetric generalizations due to efetov @xcite have long been a standard tool in the field of disordered electron physics .
areas of application include single electron motion in disordered and chaotic mesoscopic systems @xcite , chaotic scattering @xcite , localization and delocalization in systems exhibiting the integer quantum hall effect @xcite , statistical properties of the dirac spectrum in non - abelian gauge field backgrounds @xcite , to mention only a few .
the utility of the non - linear sigma model derives from the fact that it exposes the long - wavelength degrees of freedom of the disordered system , which are hidden in the original microscopic formulation by a hamiltonian with quenched , i.e. , time - independent , random parameters .
these degrees of freedom correspond to interacting diffusion modes , known to be responsible for the universal behavior of spectral and eigenfunction statistics in a broad class of disordered systems @xcite . while the final results are expected to exhibit a high degree of universality , the mathematical tools employed to derive the non - linear sigma model may vary depending on the type of microscopic model under consideration .
a concise introduction to this issue can be found in @xcite ; see also the recent papers @xcite .
for some types of microscopic model the following mathematical identity plays a key role in the derivation : @xmath2 where @xmath3 is a constant and @xmath4 denotes the lebesgue measure of a normed vector space of matrices @xmath5 . in a situation with compact orthogonal symmetry , which arises when the fermionic sector of the theory is considered , one has license to take the integration domain @xmath6 to be the real symmetric matrices @xmath5 , and it is then trivial to do the gaussian integral by completing the square and shifting variables
however , for reasons reviewed in sect.[sect : background ] the bosonic sector ( say , of time - reversal invariant systems without spin ) calls for @xmath7 to be a _ non_-symmetric matrix composed of elements @xmath8 with @xmath9 real numbers @xmath10 and @xmath11 .
the diagonal matrix @xmath12 with entries @xmath13 determines a non - compact variant @xmath14 of the real orthogonal group by the condition @xmath15 , where @xmath16 means the transpose of the matrix @xmath17 .
first discovered in the present context by wegner @xcite , the group @xmath14 is sometimes referred to as the ` hyperbolic symmetry ' of the problem at hand . in the situation with hyperbolic symmetry , choosing a good domain of integration for @xmath18 and making rigorous sense of the integral ( [ eq : hs ] ) are non - trivial tasks .
the difficulty arises from the general context of the so - called hubbard - stratonovich method : the exponential @xmath19 must be kept bounded or else the next step of the method , which is to integrate over the microscopic fields @xmath20 , would be invalid .
since the real matrix @xmath7 satisfies the symmetry relation @xmath21 one might think that one should try to integrate over the domain of all real matrices @xmath18 subject to the symmetry @xmath22 .
unfortunately , the resulting quadratic form @xmath23 is of indefinite sign and therefore such a choice of integration domain for @xmath18 makes the @xmath18-integral divergent .
nonetheless , a valid solution of the problem posed by ( [ eq : hs ] ) , i.e. , a choice of integration domain for @xmath18 making the @xmath18-integral converge while keeping the integrand bounded as a function of @xmath7 , was offered in the paper by schfer and wegner @xcite ; by deforming some of the real freedoms in @xmath18 into the complex numbers they parameterize @xmath18 as @xmath24 where @xmath25 runs through the real symmetric matrices that commute with @xmath26 , the constant @xmath27 is any positive real number , and @xmath28 .
the reader is referred to the review @xcite for a detailed discussion .
now the schfer - wegner choice of domain has , besides its many merits , a certain drawback : it lacks invariance under the action of the symmetry group @xmath29
. this may be the reason why the schfer - wegner choice was never much in use by the disordered electron physics community .
another choice of integration domain for ( [ eq : hs ] ) , which meets the afore - mentioned requirements and also has the desirable property of @xmath14-invariance , first appeared in a paper by pruisken and schfer @xcite . beginning with @xcite
this choice came to be used in numerous later works . in the present paper we are going to establish this so - called pruisken - schfer domain as a rigorous alternative to that of schfer - wegner .
there exists , however , a mathematical subtlety : although our choice of domain agrees with pruisken - schfer as a union @xmath30 of _ sets _ @xmath31 , it differs as an element @xmath32 of the vector space of integration _ chains _ ( for the details , see below ) .
some low - dimensional special cases of it were worked out in @xcite .
the pruisken - schfer domain , @xmath6 , is an open subspace of the @xmath33-dimensional space of real matrices @xmath18 obeying the mixed symmetry relation @xmath34 .
because such matrices may in general have _ complex _ eigenvalues and eigenvectors , one defines @xmath6 as the subspace of matrices @xmath35 that can be diagonalized by conjugation @xmath36 with an element @xmath37 of the _ real _ ( non - compact ) orthogonal group .
this set @xmath6 turns out to be the union @xmath38 of @xmath39 sub - domains @xmath40 .
the domains @xmath41 are enumerated as follows . with the exception of cases forming a set of measure zero
, an @xmath1-diagonalizable matrix @xmath34 has @xmath42 ` space - like ' and @xmath43 ` time - like ' eigenvalues @xmath44 here we use the language of relativity theory to communicate that the corresponding eigenvector @xmath45 has the property @xmath46 or @xmath47 , respectively . in the generic situation without degeneracies ,
the eigenvalues of @xmath18 can be arranged in decreasing order and this ordered sequence translates into a binary sequence @xmath48 by writing , say , the symbols @xmath49 for space - like and @xmath50 for time - like eigenvalues .
the binary sequence @xmath48 encoding the relative order of space - like and time - like eigenvalues is an @xmath1-invariant .
moreover , collisions between eigenvalues of space - like and time - like type generically lead to the birth of complex eigenvalues , thereby taking us outside the integration domain for @xmath5 .
this means that the @xmath14-diagonalizable matrices @xmath18 of binary sequence @xmath48 form an @xmath1-invariant domain , which we denote by @xmath31 , and any two such domains @xmath41 and @xmath51 with @xmath52 can only touch each other in subspaces of lower dimension ( more precisely , of codimension two ) .
the integration measure @xmath4 is defined to be the flat one for all domains @xmath31 : @xmath53 now the startling feature of the following statement is that the integral on the right - hand side of ( [ eq : hs ] ) is proposed to be a sum of integrals @xmath54 with _ alternating _ sign ! [ thm:1 ] there exists some choice of cutoff function @xmath55 ( converging pointwise to unity as @xmath56 ) , and a unique choice of sign function @xmath57 and a constant @xmath58 such that @xmath59 holds true for all matrices @xmath60 with the positivity property @xmath61 .
it will be shown that @xmath62 is the parity of the number of transpositions @xmath63 needed to reduce the binary sequence @xmath48 to the extremal form @xmath64 .
it should be emphasized that ( [ eq : thm1 ] ) is _ not _ a standard gaussian integral ; in fact , we have not succeeded in finding a proof of this formula by completing the square and shifting .
it is informative to give an alternative expression for the integral ( [ eq : thm1 ] ) in terms of the eigenvalues of @xmath5 .
let @xmath65 with @xmath66 and @xmath67 .
the volume element transforms as @xmath68 where @xmath69 and @xmath70 is a positive haar measure for the connected group @xmath71 .
the jacobian is @xmath72 [ cor:1 ] let a function @xmath73 with alternating sign on @xmath74 be defined by @xmath75 then , assuming the positivity @xmath76 of the matrix @xmath77 , we have @xmath78 the first @xmath42 eigenvalues @xmath79 are space - like , the last @xmath43 time - like . to deduce cor .
[ cor:1 ] from thm .
[ thm:1 ] , we need only observe that the function @xmath73 differs from the jacobian @xmath80 precisely by our sign function @xmath81 : @xmath82 the presence of this alternating sign factor was conjectured in @xcite ; see also @xcite . in previous work ,
beginning with @xcite , the identity of cor.[cor:1 ] was assumed to hold without the alternating sign .
it became apparent in 1995 that this assumption was unfounded @xcite though it remained unclear at that time how to correct the mistake .
problems with the pruisken - schfer domain were mentioned in @xcite , but for various reasons the issue was never much emphasized in the published literature .
( for one thing , all results derived using @xmath80 instead of @xmath73 stand correct after saddle - point approximation for large @xmath83 , as the integral in the large-@xmath83 limit is dominated by contributions from a single domain @xmath31 . for another , the schfer - wegner domain was available @xcite as a rigorous alternative in order to do all intermediate steps correctly . )
the present paper solves this long standing problem in a satisfactory if perhaps surprising manner .
let it be pointed out that the matrix @xmath84 appearing in applications is not strictly positive but positive _ semi_-definite .
this is , however , a minor issue as an easy variant of the formula ( [ eq : thm1 ] ) takes care of the semi - definite case ; see @xcite for the details . for the related case of non - compact unitary symmetry @xmath85 a formula analogous to that of cor .
[ cor:1 ] had been established in @xcite , using the special feature of semiclassical exactness by the duistermaat - heckman theorem @xcite .
although it will not be shown here , the methods of the present paper are robust and can be adapted to handle the case of @xmath86 as well , without taking recourse to semiclassical exactness .
the paper is organized as follows . in sect .
[ sect : background ] we present some background material concerning the hubbard - stratonovich method .
basic results needed from integral calculus are collected in sect .
[ sect : basics ] .
then , in sect
. [ sect:1,1 ] , we work out the simple but instructive case of @xmath87 in detail .
the case of general @xmath42 and @xmath43 is handled in sect .
[ sect : p , q ] by reduction to @xmath87 .
to make the present paper self - contained , we now describe the steps prior to ( [ eq : thm1 ] ) , called the hubbard - stratonovich transformation in the present context . although the true power of this transformation is to allow the treatment of non - trivial models with _ local _
@xmath88 gauge symmetry @xcite or even without such symmetry @xcite , we will restrict ourselves here to reviewing the basic steps at the example of the simplest random matrix model of this universality class : the gaussian orthogonal ensemble ( goe ) . in the present work we are concerned with time - reversal invariant systems without spin . for such systems ,
the quantum mechanics of stationary states can be done over the field @xmath89 of real numbers .
assuming the real hilbert space to be of finite dimension , @xmath83 , we may express the hamiltonian operators @xmath90 by real symmetric @xmath91 matrices .
the gaussian orthogonal ensemble by definition is an @xmath88-invariant probability measure @xmath92 on the space of such matrices . for our purposes ,
the goe measure @xmath92 is best characterized by its fourier transform : @xmath93 where the fourier variable @xmath94 is any symmetric matrix , and @xmath95 is a parameter
. an important object of the theory ( see , e.g. , @xcite ) is the expectation value of the reciprocal of a product of ( square roots of ) characteristic polynomials of @xmath96 @xmath97 we assume that @xmath98 for @xmath99 and @xmath100 for @xmath101 . to compute such an expectation value
, we use the trick of representing the reciprocal square root of each determinant as a gaussian integral over vector variables @xmath102 @xmath103 where @xmath104 and the factor @xmath105 ensures the convergence of the integral .
@xmath106 means the euclidean scalar product of @xmath107 . substituting such integrals into ( [ 2 ] )
, one immediately performs the goe average by employing the characteristic property ( [ eq : goe ] ) .
the result can then be expressed in terms of a matrix @xmath108 of size @xmath109 with entries @xmath110 @xmath111 here is where the identity ( [ eq : hs ] ) comes in : it enables us to linearize the quadratic term @xmath112 in the exponent , thereby reducing the integral over the @xmath83-component vectors @xmath113 to @xmath83 decoupled gaussian integrals , one for each component @xmath114 . when the latter integrals are carried out , ( [ 4 ] ) becomes an integral over the collective variables @xmath18 only . using the rigorous form ( [ eq : thm1 ] ) of the relation ( [ eq : hs ] ) we obtain @xmath115 where @xmath116 and , with the assumption that @xmath83 is large enough , the cutoff function @xmath117 has been removed by sending the regularization parameter @xmath118 to zero . the integral representation ( [ eq : intrep ] ) for @xmath119 is well suited for saddle - point analysis in the limit of large @xmath83 .
we can now appreciate why the identity ( [ eq : hs ] ) is useful here : it serves the purpose of exposing the good variables in which to perform the large-@xmath83 limit .
let this fact suffice as a motivation for our labors in the body of this paper .
recalling from sect . [ sect : result ] the definition of the connected domains @xmath41 ( cf . also below )
we consider the following alternating sum of integrals : @xmath120 where @xmath117 is some smooth cutoff function which regularizes the integral and converges pointwise to unity in the limit @xmath121 .
the real matrix @xmath7 is subject to the conditions @xmath60 and @xmath61 . choosing the lebesgue measure @xmath4 as in ( [ eq : flat ] )
we shall prove that the integral @xmath122 does not depend on the matrix @xmath7 , or equivalently , that all of its directional derivatives vanish : @xmath123 once ( [ eq : indep ] ) has been established , the hubbard - stratonovich transformation ( [ eq : thm1 ] ) follows by multiplying both sides of ( [ eq : sum1 ] ) with a constant @xmath124 . to prepare our treatment , we recall a few basic facts from calculus . given a vector field @xmath45 on an @xmath125-dimensional differentiable manifold @xmath126 ,
let @xmath127 \ni t \mapsto \phi_t$ ] be the flow of @xmath45 , i.e. , the one - parameter family of mappings @xmath128 determined by @xmath129 and @xmath130 for all @xmath131 . in local coordinates @xmath132 , we have @xmath133 we are using the summation convention : an index appearing twice ( once as a covariant and once as a contravariant index ) is understood to be summed over .
for a differential form @xmath134 on @xmath126 let @xmath135 denote the pullback of @xmath136 by @xmath137 .
the lie derivative of @xmath136 w.r.t .
the vector field @xmath45 is then defined by differentiation at @xmath138 @xmath139 by the relation @xmath140 , the lie derivative satisfies the leibniz rule @xmath141 .
cartan s formula expresses it as @xmath142 where @xmath143 is the exterior derivative , @xmath144 and @xmath145 denotes the operation of contraction with the vector field @xmath146
@xmath147 after this summary of basic facts , let @xmath148 be an @xmath125-form on @xmath126 and consider the integral @xmath149 where @xmath150 is the derivative of some function @xmath151 in the direction of the vector field @xmath45 .
assuming that @xmath148 is invariant by the flow of @xmath45 , the leibniz rule gives @xmath152 by stokes theorem for the integral of the exact @xmath125-form @xmath153 it follows that @xmath154 where @xmath155 is the @xmath156-dimensional boundary of @xmath126 .
in what follows , we will employ ( [ eq : ibp ] ) in the closely related case where one is integrating against a density @xmath157 instead of an @xmath125-form @xmath148 .
formula ( [ eq : ibp ] ) still holds true in that case assuming of course that @xmath158 .
note also that in order to integrate the twisted @xmath156-form @xmath159 , the boundary @xmath155 must be equipped with an _
orientation , i.e. , a transverse vector field pointing , say , from the inside to the outside of @xmath155 .
( technically speaking , a density @xmath157 on @xmath126 is a top - degree differential form @xmath148 tensored by a section @xmath12 of the orientation line bundle of @xmath126 , and a choice of splitting @xmath160 determines an isomorphism between outer and inner orientations for @xmath155 . ) to apply the formula ( [ eq : ibp ] ) to the situation at hand , let @xmath161 denote the vector field generating translations @xmath162 .
then @xmath163 where @xmath164 . here
, recognizing the fact that our integrand depends on @xmath18 and @xmath7 only through the combination @xmath165 , we have used the identity @xmath166 now the lebesgue measure ( or positive density ) @xmath4 is invariant under translations @xmath167 . therefore @xmath168 , and we may apply formula ( [ eq : ibp ] ) to obtain @xmath169 the main achievement of this paper is a proof that the alternating sum of integrals on the right - hand side vanishes in the limit @xmath121 . for that purpose
we need to develop a good understanding of the boundary components @xmath170 and the twisted form @xmath171 restricted to them . as a first step ,
we illustrate the essential features of the argument at the example of the symmetry group being @xmath172 .
this example plays a very important role as the general case will be handled by reduction to it .
here we are going to deal with the special case of real @xmath174 matrices @xmath18 subject to the linear symmetry relation @xmath175 where @xmath176 means the operation of taking the matrix transpose .
such matrices can be parameterized by three real variables @xmath177 , @xmath178 and @xmath179 as @xmath180 thus our space of matrices @xmath18 is isomorphic as a vector space to @xmath181 .
the measure of integration is the positive density @xmath182 .
the matrix @xmath7 is of the same form as @xmath18 but its matrix elements are constrained by @xmath183 another way of describing the setting is to equip the real vector space @xmath184 with the sign - indefinite bilinear form @xmath185 determined by @xmath186 @xmath187 the matrices @xmath18 are then symmetric with respect to @xmath185 in the sense that @xmath188 for all @xmath189 .
we denote the linear space of real matrices @xmath18 with this symmetry property by @xmath190 .
since @xmath185 has signature @xmath191 , the symmetry group of @xmath185 is the non - compact real orthogonal group @xmath192 .
elements @xmath193 satisfy the equation @xmath194 which is equivalent to saying that @xmath195 for all @xmath196 . for reasons indicated in the introduction ( see sect.[sect : intro ] ) let us recall that in order for the integral ( [ eq : hs ] ) to exist one needs the inequality @xmath197 , at least asymptotically we want our matrices @xmath198 to be diagonalizable by the action @xmath199 of the real group @xmath173 by conjugation . the condition of diagonalizability is formulated most clearly by expressing @xmath18 in a basis of light - like vectors of @xmath184 : @xmath200 which obey @xmath201 and @xmath202 . applying @xmath18 to this basis one has @xmath203 where @xmath204 , @xmath205 , and @xmath206 .
since the diagonal piece of @xmath18 in this basis is a multiple of unity , the condition for diagonalizability is that the product of off - diagonal matrix elements be positive : @xmath207 .
the domain @xmath208 of @xmath173-diagonalizability of our matrices @xmath209 consists of two connected components : @xmath210 the motivation for the notation ( @xmath211 ) will become clear shortly . here
we simply remark that if @xmath212 denotes a pair of @xmath18-eigenvectors with @xmath213 and @xmath214 are the associated eigenvalues , then @xmath215 for @xmath216 and @xmath217 for @xmath218 .
next , let @xmath219 denote the regularized gaussian integrand @xmath220 where we choose the cutoff function to be @xmath221 using the coordinates @xmath222 we have the expression @xmath223 with @xmath224 let us observe that the integral @xmath122 defined in ( [ eq : sum1 ] ) can now be written as @xmath225 by their definition via inequalities , the domains @xmath41 ( for @xmath226 ) are open in the three - dimensional space @xmath227 and have boundaries @xmath170 which are two - dimensional . each of the boundaries @xmath170 is a union of 2 half - planes . to describe this in detail , let
4 half - planes @xmath228 and @xmath229 be defined by @xmath230 @xmath228 and @xmath229 make up the boundary of @xmath41 ( for both @xmath231 and @xmath232 ) .
being part of @xmath233 , the half - plane @xmath234 inherits from @xmath235 an outer orientation by the transverse vector field @xmath236 .
( indeed , starting from any point @xmath237 very close to the boundary @xmath234 and making a small step in the direction of @xmath238 , one crosses @xmath239 . ) in the same sense , @xmath240 is oriented by @xmath241 while @xmath242 and @xmath243 for @xmath231 are oriented by the opposite vector fields @xmath244 and @xmath245 respectively ; see fig.[fig:2dom ] . with these orientations being understood , we have the relations @xmath246 which are equalities in the sense of chains , or distributions . ( to achieve equality in the sense of sets
, we would have to include the real line @xmath247 , which is part of both @xmath248 and @xmath233 .
however , our aim is to integrate bounded differential forms , and for that purpose the lower - dimensional parts of the boundary are of no concern . ) with @xmath170 as described above , formula ( [ eq : dida ] ) now reads @xmath249 the key idea is to reorganize the boundary half - planes into consistently oriented closed manifolds .
define the full planes @xmath250 so that we have @xmath251 in the sense of chains . since @xmath252 ( still as chains )
, we obtain the following relation between integrals : @xmath253 for any twisted 2-form @xmath148 which is bounded ( if not continuous or smooth ) . in particular
, this relation holds for @xmath254 . thus turns into @xmath255 a graphical sketch of the situation
is shown in fig .
[ fig : or ] .
we are now going to show that the integral @xmath256 vanishes for @xmath257 vanishes in this limit . to handle the case of @xmath258 ,
let @xmath259 be the identical embedding and introduce coordinates @xmath260 on @xmath258 so that @xmath261 recalling @xmath262 we pull back the functions @xmath263 and @xmath264 to @xmath258 : @xmath265 next we compute the pullback of the twisted 2-form @xmath266 . for this , we write @xmath267 where @xmath268 when @xmath269 is oriented by the ordered set of linear coordinates @xmath27 , @xmath270 , @xmath271 ( or any even permutation thereof ) , and @xmath272 when the ordering @xmath27 , @xmath271 , @xmath270 ( or any other odd permutation of @xmath27 , @xmath270 , @xmath271 ) is chosen . using the coordinate expression of the vector field @xmath273 , @xmath274 we then obtain @xmath275 now the presence of the orientation factor @xmath276 tells us that an outer orientation by @xmath277 translates into an inner orientation by @xmath278 . since our @xmath279-plane @xmath258 is oriented by @xmath280 , we must assign to it the orientation given by @xmath281 . in this way , assembling everything we arrive at the expression @xmath282 where the right - hand side has been reduced to a product of riemann integrals by fubini s theorem and the definition of what it means to integrate a differential form .
the crucial point now is that the integral over @xmath271 vanishes in the limit @xmath121 . indeed , writing @xmath283 for short , the @xmath271-integral gives @xmath284 and this does go to zero for @xmath285 as long as the real number @xmath286 is non - zero .
the latter condition is always satisfied , as the constraints on the matrix elements of @xmath7 ensure that @xmath287 thus we have shown that @xmath288 .
the situation is no different for the integral over the 2-plane @xmath289 .
we therefore conclude that @xmath290 .
hence @xmath291 is a constant independent of @xmath7 for @xmath61 .
this proves the validity of the hubbard - stratonovich transformation in the following precise form with cutoff function @xmath292 : @xmath293 where @xmath294 and @xmath295 , and we have replaced @xmath296 by @xmath297 .
let us emphasize once again that the right - hand side is _ not _ a standard gaussian integral , and that our proof of this formula was not by completing the square and shifting .
we will show in the next subsection that @xmath298 .
we now spell out another line of approach for the case of @xmath299 , which is a variant of that given in @xcite . this calculation will be much quicker , but does not extend ( at least not in any way known to us ) to higher values of @xmath300 .
recall that in light - cone coordinates @xmath270 , @xmath271 , and @xmath27 , we have the expression @xmath301 where @xmath302 and @xmath303 .
it is clear that this can be integrated against @xmath304 resulting in a factor @xmath305 .
it then remains to integrate the other factor @xmath306 against @xmath307 over the two quadrants @xmath308 and @xmath309 .
for this we make the substitution of variables @xmath310 and @xmath311 which transforms the measure @xmath312 to @xmath313 . noticing that @xmath314 on @xmath315 , @xmath316 on @xmath317 , and our chain of integration is @xmath318
, we drop the absolute value on @xmath319 and integrate against the sign - alternating distribution @xmath320 .
more precisely , we take the @xmath321 integration to be the inner one ( this obviates the need for regularization albeit at the expense of fubini s theorem becoming inapplicable , so that the order of integrations is now fixed and can no longer be changed ) .
the resulting @xmath321 integral is @xmath322 we now observe the relation @xmath323 which shows that the remaining @xmath324 integral is another standard gaussian integral in the variable @xmath325 , yielding the value @xmath326 .
altogether we then obtain @xmath327 for @xmath328 , which is the desired result but for scaling @xmath329 and @xmath330 .
( of course it must be understood here that the @xmath321 integration has to be done first . )
this computation , while direct and short , does not seem to carry over to the case of higher values of @xmath42 and @xmath331 , whereas our conceptual proof of sects .
[ sect:4.1]-[sect:4.3 ] does .
we now turn to the general case of real matrices @xmath18 of size @xmath332 , @xmath333 where the blocks @xmath334 and @xmath335 are symmetric . such matrices @xmath18 as a whole satisfy the linear symmetry relation @xmath336 equivalently , if we equip the real vector space @xmath74 with the indefinite bilinear form @xmath337 then our matrices @xmath18 form the linear space @xmath338 of @xmath185-symmetric matrices , i.e. , @xmath339 the symmetry group of the bilinear form @xmath185 with signature @xmath340 is the non - compact real orthogonal group @xmath14 of matrices @xmath17 satisfying the equation @xmath341 or equivalently , @xmath195 for all @xmath342 . we still want our matrices @xmath343 to be diagonalizable by the @xmath14-action @xmath344 . in the simple case @xmath87 we saw that diagonalizability fails when the off - diagonal block @xmath345 is larger than the difference of the diagonal blocks @xmath334 and @xmath346 . a similar phenomenon is expected to occur for general @xmath347 . roughly speaking , @xmath18 will fail to have real eigenvectors and eigenvalues when the degree of non - hermiticity coming from the off - diagonal blocks becomes too large . in the hubbard - stratonovich integral ( [ eq : sum1 ] ) we integrate over the set , @xmath6 , of @xmath343 that can be brought to diagonal form by the @xmath348-action , i.e. , matrices of the form @xmath349 with @xmath350 and @xmath351 .
we know from our considerations for @xmath299 that @xmath6 decomposes into connected sets @xmath41 whose boundaries intersect in lower - dimensional sets ( of codimension 2 ) . here
we describe these connected components @xmath31 .
let @xmath349 be @xmath14-diagonalizable in the sense just described , so that the @xmath352 column vectors which make up @xmath17 form a basis of @xmath74 consisting of eigenvectors of the matrix @xmath5 .
let @xmath353 denote the @xmath354 column of @xmath17 .
then by the equation @xmath355 the vector @xmath356 has positive norm square @xmath357 for @xmath358 and negative norm square @xmath359 for @xmath360 . to distinguish between these two cases , we use the language of relativity theory and call the corresponding eigenvalue @xmath361 of ` space - like ' type in the former case and time - like in the latter . assuming that we are in the generic situation , where all eigenvalues @xmath361 of @xmath18 differ from each other , we can arrange them as a descending sequence .
given any such sequence , we assign to it a motif , @xmath48 : if an eigenvalue of the ordered sequence is space - like , then we represent it as a full dot , otherwise as an empty dot .
thus our motifs @xmath48 are sequences of @xmath42 full and @xmath43 empty dots ; e.g. , @xmath362 is a motif for @xmath363 . with each motif
@xmath48 we now associate a domain @xmath41 as follows . while insisting that eigenvalues of opposite type ( space - like vs. time - like ) must never be equal , we lift the condition of non - degeneracy of eigenvalues partially : eigenvalues of the same type are now allowed to collide and exchange positions , but eigenvalues of opposite type still are not . this rule of eigenvalue motion defines a connected domain @xmath41 for each @xmath364 .
we define a sign function @xmath365 by setting @xmath366 where @xmath125 is the number of transpositions @xmath367 which are needed to reduce @xmath48 to the reference motif @xmath368 made of @xmath43 empty dots following @xmath42 full dots .
the parity of this number @xmath125 is well - defined although @xmath125 itself is not .
indeed , another expression for @xmath369 is @xmath370 where the @xmath27 s are the eigenvalues of any @xmath371 .
as was explained above , a point in the interior of @xmath41 is a matrix @xmath372 which is diagonalizable by the @xmath14-action and has eigenvalues that order according to the motif @xmath48 in particular , no two eigenvalues of opposite type are equal .
this means that if @xmath373 is a continuous curve which starts in @xmath41 but leads to a collision of two or more eigenvalues of opposite type at @xmath374 , then @xmath375 is a point in the boundary of @xmath31 .
the generic boundary situation is that ( i ) exactly two eigenvalues of opposite type collide , ( ii ) the norms of the two corresponding eigenvectors go to zero , and ( iii ) the boundary point @xmath375 is a matrix which fails to be diagonalizable by any real transformation @xmath376 . to carry out the computations leading to the desired result ( [ eq : indep ] ) , we need a good description of the interior of @xmath377 as well as the boundary @xmath378 .
this is achieved as follows .
let us first make a slight change of perspective : abandoning the viewpoint of @xmath18 as a matrix , we switch to regarding @xmath18 as an invariantly defined linear operator ( giving rise to a matrix when expressed w.r.t . some fixed basis )
; i.e. , taking the matrix elements @xmath379 we let @xmath380 ( summation convention ! ) with @xmath381 the standard basis of @xmath382 defined in terms of the standard basis @xmath383 of @xmath74 by @xmath384 .
consider now the grassmannian @xmath385 of @xmath185-orthogonal decompositions @xmath386 where @xmath387 is any lorentzian plane , i.e. , a @xmath388-dimensional subspace @xmath389 spanned by a pair of null lines or light - like directions @xmath390 such that @xmath391 .
the restriction of @xmath185 to @xmath392 is then non - degenerate , and the @xmath185-orthogonal complement @xmath393 of @xmath392 is spanned by @xmath394 space - like and @xmath395 time - like vectors .
note that @xmath396 is acted upon transitively by the group @xmath397 , which is to say that every lorentzian plane @xmath398 can be regarded as the image @xmath399 of the standard lorentzian plane @xmath400 by some transformation @xmath401 .
since @xmath402 has isotropy @xmath403 , our manifold @xmath396 is the base of a principal fibre bundle @xmath404 moreover , since the lorentzian planes @xmath387 are in one - to - one correspondence with pairs of real lines , one in @xmath405 and @xmath406 each , @xmath396 is a product @xmath407 of real projective spaces @xmath408 for @xmath409 and @xmath410 . from this identification
it follows immediately that the manifold @xmath385 is compact , non - simply connected , and non - orientable unless both @xmath42 and @xmath43 are even .
it is also seen that ( [ eq : pfb ] ) decomposes as a direct product of two principal fibre bundles @xmath411 for @xmath409 , @xmath331 .
we note that these principal fibre bundles are non - trivial .
choosing a point @xmath412 is equivalent to choosing an orthogonal projector @xmath413 by orthogonality , @xmath185 restricts to a bilinear form of signature @xmath191 on @xmath414 and a bilinear form of signature @xmath415 on the orthogonal complement @xmath416 .
we denote by @xmath417 resp .
@xmath418 the symmetry groups of these symmetric bilinear forms .
let @xmath419 denote the subspace of linear operators @xmath420 which are symmetric with respect to @xmath185 restricted to @xmath421 , and let @xmath422 denote the subspace of @xmath423-diagonalizable linear transformations @xmath424 .
now consider the set , @xmath425 , of triples @xmath426 consisting of any point @xmath412 and two linear operators @xmath427 and @xmath428 .
this set @xmath425 has the structure of a fibre bundle @xmath429 where the fibre over @xmath430 is a direct product @xmath431 one may also view @xmath425 as an associated bundle @xmath432 with @xmath433 , @xmath434 where the action of @xmath94 on @xmath435 is by conjugation . since the principal bundle @xmath436 is non - trivial
, so is the associated bundle @xmath437 .
@xmath425 can be viewed as a subset of our space @xmath438 of operators @xmath5 .
indeed , for @xmath439 we may use the projector @xmath440 to define @xmath441 for any vector @xmath442 by @xmath443 in other words , @xmath18 acts on @xmath414 as @xmath444 and on @xmath445 as @xmath446 .
since both @xmath444 and @xmath447 are symmetric with respect to @xmath185 , so is @xmath448 .
more formally , this correspondence @xmath449 defines a mapping @xmath450 .
note that the mapping @xmath451 is not surjective , as the operators @xmath18 with more than two non - real eigenvalues can not be presented in this form .
however , the image @xmath452 is large enough to contain our integration domain @xmath6 and also ( in the measure - theoretic sense ) the boundary @xmath378 .
now for any point @xmath430 of the grassmann manifold @xmath396 let @xmath453 denote the two null lines of the lorentzian vector space @xmath421 .
in other words @xmath454 for @xmath455 or @xmath456 . the three - dimensional space
@xmath457 then decomposes as @xmath458 let us choose a basis @xmath459 of null vectors of @xmath414 so that @xmath460 and @xmath461 . in such a basis @xmath440 is expressed as @xmath462 .
linear coordinates @xmath27 , @xmath270 , @xmath271 for @xmath463 are then introduced by decomposing @xmath444 according to ( [ eq : inv - decomp ] ) : @xmath464 extending the notation used before , let @xmath465 and @xmath466 denote the linear subspaces which are defined by the equations @xmath467 and @xmath468 respectively .
these 2-planes @xmath469 and @xmath470 will assume the roles played by @xmath258 and @xmath289 in sect .
[ sect:4.2 ] .
it should be emphasized , however , that @xmath270 and @xmath271 do not extend to globally defined coordinate functions for @xmath471 .
the point here is that transporting the pair of null lines @xmath453 around a closed path in @xmath472 one may find that @xmath473 and @xmath474 get interchanged . in other words , although @xmath453 are invariantly defined as a _ pair _ of lines , there is no invariant way of telling which line is @xmath473 and which is @xmath474 . as a result , the linear coordinate functions @xmath270 and @xmath271 projecting from @xmath475 to @xmath476 and @xmath477 are only defined locally in the variable @xmath478 . by the same token , the union of planes
@xmath479 is invariantly defined for all @xmath480 , whereas the planes @xmath469 and @xmath470 individually are not .
the smooth assignment @xmath481 gives a bundle @xmath482 with fibre @xmath483 over @xmath484 . by applying the map
@xmath485 we then obtain a submanifold @xmath486 .
as a matter of fact , since the generic situation at the boundary of @xmath487 is that exactly two eigenvalues of opposite type collide , the set @xmath488 agrees with the set @xmath378 up to lower - dimensional pieces .
we now introduce an outer orientation on @xmath489 to achieve the stronger property that @xmath488 agrees with the boundary @xmath378 as a chain ( or distribution ) . for that , fix any @xmath490 and consider some @xmath491 .
arrange the single eigenvalue @xmath492 of @xmath444 and the eigenvalues @xmath493 of @xmath447 as a decreasing sequence .
( this ordering is uniquely defined except for the set of measure zero where some eigenvalues are degenerate . )
following the prescription of sect .
[ sect:2.3.1 ] , draw the space - like eigenvalues of @xmath447 as full dots and the time - like eigenvalues as empty dots . in view of fig .
[ fig : or ] , if @xmath494 then represent @xmath492 by @xmath495 , else by @xmath496 .
this defines a motif @xmath497 for almost every pair @xmath498 . for @xmath499 for example , if @xmath500 , @xmath501 is space - like , @xmath502 is time - like , and @xmath503 , then @xmath504 .
the point @xmath505 is a point in the boundary of @xmath506 . according to the alternating sign of the sum @xmath507 the outer orientation of @xmath378 at @xmath505
is directed from the inside to the outside if @xmath508 and from the outside to the inside if @xmath509 .
consulting again fig .
[ fig : or ] we assign to the boundary point @xmath505 the transverse vector @xmath510 where @xmath511 means the differential @xmath512 of the map @xmath451 at the point @xmath513 .
the motif @xmath497 switches sign as @xmath514 passes through zero ; therefore our assignment is smooth along the @xmath271-axis of @xmath515 , although discontinuities may occur when the eigenvalue @xmath492 hits eigenvalues of @xmath447 or when eigenvalues of @xmath447 of opposite type collide . in the case of the other component , @xmath516 , we proceed in the same way , albeit with the two boundary coordinate functions @xmath270 and @xmath271 exchanging roles .
thus we assign to a point @xmath505 for @xmath517 the transverse vector @xmath518 which is again smooth as a function of @xmath270 .
now while the distinction between @xmath270 and @xmath271 ( or @xmath469 and @xmath470 ) is only defined locally in @xmath430 , the outcome of our discussion is free of the ambiguity of exchanging @xmath519 .
indeed , a glance at fig . [ fig : or ] shows that the picture is invariant under the reflection ( @xmath519 ) at the axis dividing the first quadrant .
hence our procedure gives a _ globally _ defined piecewise smooth transverse vector field and hence a piecewise smooth outer orientation of the boundary @xmath488 . adopting this outer orientation
we have @xmath520 as an equality between chains , i.e. , in the weak sense : @xmath521 for any bounded twisted differential form @xmath522 of top degree .
this result holds true , in particular , for the integrand @xmath523 of equation ( [ eq : dida ] ) with the gaussian @xmath524 regularized by the cutoff function @xmath525 . in the next step the left - hand side of the integral ( [ eq : bound - final ] ) for @xmath527
is pulled back to @xmath528 .
this results in an iterated integral where the outer integral is over the compact manifold @xmath529 and the inner integral is over @xmath530 .
it will now be shown that the inner integral over @xmath531 is of exactly the same form as in the case of @xmath87 and hence , for the reasons explained in sect .
[ sect:4.3 ] , vanishes in the limit @xmath121 .
we will do this by inspection of @xmath532 .
first we compute the pull back @xmath533 the differential of the inverse map @xmath534 we have @xmath535 with this relation in mind we now compute @xmath536 .
for this we notice that the map @xmath537 , @xmath538 schematically has the variation @xmath539 the first two terms contribute factors of unity to the jacobian when one pulls back the translation - invariant measure @xmath4 of @xmath540 . now
the third term is purely off - diagonal w.r.t .
the decomposition @xmath541 @xmath542 and may be expressed as a commutator @xmath543 $ ] with a generator @xmath544 .
the two components of @xmath545 in @xmath546 and @xmath547 are related to each other by the condition of @xmath185-symmetry . identifying @xmath414 with its dual @xmath548 by means of the non - degenerate form @xmath549 we have an isomorphism @xmath550 . by this isomorphism
we may view the multiplication operator @xmath551 as a linear transformation @xmath552 on @xmath553 .
we thus arrive at the formula @xmath554 with the jacobian of the transformation being @xmath555 here @xmath556 denotes a suitably normalized @xmath557-invariant measure for the compact manifold @xmath396 , while @xmath558 and @xmath559 are translation - invariant measures for @xmath475 and @xmath560 . as an observation to be used presently
, we note that the jacobian @xmath561 becomes independent of @xmath270 and @xmath271 for @xmath562 .
we now have to contract the density @xmath536 with the vector field @xmath563 and make the restriction to @xmath489 .
since @xmath564 on @xmath469 and @xmath565 on @xmath566 , the contraction after restriction depends only on the @xmath270- and @xmath271-components of @xmath567 : @xmath568 while these components vary with @xmath480 , @xmath569 they are constant along @xmath469 and @xmath566 .
hence , recalling the expression @xmath570 and assembling factors , we see that @xmath571 for fixed @xmath430 is proportional to the form @xmath572 ( with constant coefficient ) along @xmath469 and @xmath573 ( still with constant coefficient ) along @xmath566 . in other words ,
the inner dependence ( along @xmath469 and @xmath470 ) of the integration form @xmath574 is the same as in sect.[sect:4.3 ] .
our final step is to compute the restriction to @xmath489 of the pull back @xmath575 of the gaussian @xmath576 .
for this we decompose @xmath577 fixing any @xmath478 and inserting @xmath578 we obtain @xmath579 next we fix @xmath580 and restrict the range of the variable @xmath427 to @xmath581 . with this restriction
we get @xmath582 the coefficients @xmath583 never vanish .
indeed , from @xmath584 we have @xmath585 by the positivity assumption @xmath76 and the non - degeneracy of @xmath185 . as a consequence of the fact that @xmath586
, it will suffice for our purpose of taking the limit @xmath587 to examine the cutoff function @xmath588 at @xmath589 . by the relations @xmath590 and @xmath591 the operator @xmath12 commutes with the projector @xmath592 .
therefore the pull back @xmath593 of the cutoff function @xmath594 separates into two factors , one for @xmath414 and @xmath445 each .
we again keep @xmath428 fixed and investigate the dependence on @xmath595 . using the easily verified relation @xmath596
we obtain the expression @xmath597 * acknowledgment*. yvf acknowledges the financial support of this research project by the alexander von humboldt foundation via a bessel research award .
the research in nottingham was supported by grant ep / c515056/1 from epsrc ( uk ) .
the research of mrz is supported by the deutsche forschungsgemeinschaft , sfb / tr 12 .
99 f. wegner , _ the mobility edge problem : continuous symmetry and a conjecture _
, z. phys .
b * 36 * ( 1979 ) 207 - 210 l. schfer and f. wegner , _ disordered systems with @xmath125 orbitals per site : lagrange formulation , hyperbolic symmetry , and goldstone modes _
, z. phys .
b * 38 * ( 1980 ) 113 - 126 a.m.m .
pruisken and l. schfer , _ the anderson model for electron localization non - linear sigma model , asymptotic gauge invariance _ , nucl .
b * 200 * ( 1982 ) 20 - 44 k.b .
efetov , _ statistics of levels in small metal particles _ , zh . ksp .
* 83 * ( 1982 ) 833 - 847 ; k.b .
efetov , _ supersymmetry and the theory of disordered metals _ , adv .
* 32 * ( 1983 ) 53 - 127 ; k.b .
efetov , _ supersymmetry in disorder and chaos _
( cambridge university press : cambridge , 1997 ) a.d .
mirlin , _ statistics of energy levels and eigenfunctions in disordered systems _ , phys
. rep .
* 326 * ( 2000 ) 260 - 382 j.j.m .
verbaarschot , h.a .
weidenmller , and m.r .
zirnbauer , _ grassmann integration in stochastic quantum physics the case of compound nucleus scattering _ , phys
( 1985 ) 367 - 438 y.v .
fyodorov and h .- j .
sommers , _ statistics of resonance poles , phase shifts and time delays in quantum chaotic scattering : random matrix approach for systems with broken time - reversal invariance _ , j. math .
phys . * 38 * ( 1997 ) 1918 - 1981 a.m.m .
pruisken , in _ the quantum hall effect _ , edited by r.e .
prange and s.m .
girvin , graduate texts in contemporary physics ( springer - verlag : berlin , 1987 ) j.j.m .
verbaarschot and t. wettig , _ random matrix theory and chiral symmetry in qcd _ ,
nucl . part .
* 50 * ( 2000 ) 343 - 410 m.r .
zirnbauer , _ supersymmetry methods in random matrix theory _ , encylop . of math .
vol . 5 , p. 151
- 160 ( elsevier : amsterdam , 2006 ) ; arxiv : math - ph/0404057 p. littelmann , h .- j .
sommers , and m.r .
zirnbauer , _ superbosonization of invariant random matrix ensembles _ , commun .
( in press ) ; arxiv:0707.2929 [ math - ph ] j.e .
bunder , k.b .
efetov , v.e .
kravtsov , o.m .
yevtushenko , and m.r .
zirnbauer , _ superbosonization formula and its applications to random matrix theory _ , j. stat .
* 129 * ( 2007 ) 809 - 832 y.v .
fyodorov , _ on the hubbard - stratonovich transformations over hyperbolic domains _ , j. phys . :
condensed matter * 17 * ( 2005 ) s1915-s1928 y. wei and y.v .
fyodorov , _ a conjecture on hubbard - stratonovich transformations for the pruisken - schfer parameterisations of real hyperbolic domains _ , j. phys . a * 40 * ( 2007 ) 13587 - 13605 p .- b .
gossiaux , private communication to m.r .
zirnbauer ( july 1995 ) m.r .
zirnbauer , _ riemannian symmetric superspaces and their origin in random matrix theory _ , j. math .
* 37 * ( 1996 ) 4986 - 5018 y.v .
fyodorov and e. strahov , _ characteristic polynomials of random hermitian matrices and duistermaat - heckman localisation on non - compact kaehler manifolds _ , nucl . phys .
b * 630 * ( 2002 ) 453 - 491 | we revisit a long standing issue in the theory of disordered electron systems and their effective description by a non - linear sigma model : the hyperbolic hubbard - stratonovich ( hs ) transformation in the bosonic sector . for time - reversal invariant systems without spin
this sector is known to have a non - compact orthogonal symmetry @xmath0 .
there exists an old proposal by pruisken and schfer how to perform the hs transformation in an @xmath1-invariant way . giving a precise formulation of this proposal
we show that the hs integral is a sign - alternating sum of integrals over disjoint domains . |
99 j. marro and r. dickman , _ nonequilibrium phase transitions in lattice models _ ( cambridge university press , cambridge , 1999 ) . h. hinrichsen , advances in physics * 49 * , 815 ( 2000 ) ; m. henkel , h. hinrichsen , and s. lbeck , _ non - equilibrium phase transitions _ , berlin : springer ( 2008 ) .
p. grassberger , z. phys .
b * 47 * , 365 ( 1982 ) ; h. k. janssen , z. phys .
b * 42 * , 151 ( 1981 )
. j. w. essam , j. phys .
a * 22 * , 4927 ( 1989 ) .
i. jensen , j. phys .
a * 26 * , 3921 ( 1993 ) .
h. hinrichsen , phys .
e * 55 * , 219 ( 1997 ) .
a. g. moreira and r. dickman , phys .
e * 54 * , r3090(1996 ) .
p. grassberger , h. chate , and g. rousseau , phys .
e * 55 * , 2488 ( 1997 ) ; r. caero , a. gabrielli , and m. a. muoz , phys . rev .
e * 57 * , 5060 ( 1998 ) .
i. jensen , phys .
lett . * 70 * , 1465 , ( 1993 ) ; i. jensen and r. dickman , phys .
e * 48 * , 1710 ( 1993 ) ; j. f. f. mendes , r. dickman , m. henkel , and m. c. marques , j. phys .
a * 27 * , 3019 ( 1994 ) ; d. ben- avraham and j. kohler , j. stat . phys . * 65 * , 839 ( 1992 ) ; e. v. albano , j. phys .
a * 25 * , 2557 ( 1992 ) .
a. lipowski and m. lopata , phys .
e * 60 * , 1516(1999 ) .
m. basu , u. gayen , and p. k. mohanty , _ arxiv:1102.1631 _ ; a. ghosh , u. basu , a. chakraborti , and b. k. chakrabarti , _ arxiv:1102.1968_. m. rossi , r. pastor - satorras , and a. vespignani , phys .
* 85 * , 1803 ( 2000 ) ; s. lubeck and p. c. heger , phys .
e * 68 * , 56102 ( 2003 ) .
t. nagatani , jnl .
japan * 61 * , 3196 ( 1992 ) ; j. d. noh and h. park , phys .
lett . * 94 * , 145702 ( 2005 ) ; s. kwon and y. kim , phys .
e * 72 * , 066122 ( 2006 ) .
s. cornell , m. droz , r. dickman and m. c. marques , j. phys .
* 24 * 5605 ( 1991 ) .
h. k. janssen , phys .
lett . * 78 * , 2890 ( 1997 ) ; j. stat .
phys.*103 * , 801 ( 2001 ) .
u. c. tuber , m. j. howard , and h. hinrichsen , phys .
lett . * 80 * , 2165 ( 1998 ) . j. l. cardy and u. c. tuber , phys .
lett . * 77 * , 4780 ( 1996 ) ; j. stat . phys . *
90 * , 1 ( 1998 ) .
e. arashiro and t. tom , j. phys . a : math
. theor . * 40 * , 887 ( 2007 ) .
s. kwon and h. park , phys .
e * 69 * , 066125 ( 2004 ) ; h. park and s. kwon , braz .
, * 30 * , 133 ( 2000 ) .
m. j. washenberger , m. mobilia and u. c. tuber , j. phys .
: cond . mat .
* 19 * , 065139 ( 2007 ) .
a. lipowski and d. lipowska , physica a * 276 * , 456 ( 2000 ) .
r. h. schonmann , j. stat . phys . * 44 * , 505 ( 1986 ) .
ronald dickman and guilherme j. m. garcia , phys
e * 71 * , 066113 ( 2005 ) . | the model of competition between densities of two different species , called predator and prey , is studied on a one dimensional periodic lattice , where each site can be in one of the four states say , empty , or occupied by a single predator , or occupied by a single prey , or by both .
along with the pairwise death of predators and growth of preys , we introduce an interaction where the predators can eat one of the neighboring prey and reproduce a new predator there instantly .
the model shows a non - equilibrium phase transition into a unusual absorbing state where predators are absent and the lattice is fully occupied by preys .
the critical exponents of the system are found to be different from that of the directed percolation universality class and they are robust against addition of explicit diffusion .
^s__a absorbing configurations do not have any outgoing rates @xcite .
once reached there , the system can not escape from these configurations .
presence of absorbing configurations in a phase space raises a possibility that the concerned system may undergo a non - equilibrium phase transition into absorbing states .
the critical behavior of these absorbing state phase transitions ( apt)s @xcite depends on the conservation in dynamics and the symmetry between absorbing states .
it has been conjectured @xcite that in absence of any special symmetry the apt belong to the directed percolation ( dp ) universality class as long as the system has a single absorbing state .
additional symmetries , like particle - hole symmetry @xcite , conservation of parity @xcite , and symmetry between different absorbing states @xcite lead to different universalities . spreading process with spatially quenched randomness @xcite or with long - term memory @xcite
are known to destroy the critical behavior completely , whereas the presence of infinitely many absorbing states may @xcite or may not @xcite belong to dp universality class . again a different critical behavior is observed when the activity field does not have any special symmetry , but it is coupled to a conserved density @xcite .
recent studies have indicated that dp - critical behavior is possible , even in presence of an additional conserved field @xcite .
it is not quite clear , what microscopic ingredients can make an apt belong to the dp class . the models of directed percolation has been extended to more than one species @xcite . along with the simple dp behavior , a line of first order transition @xcite has been observed in @xmath0 dimension when two species compete for survival .
janssen @xcite studied coupled dp processes with bilinear and bidirectional interspecies couplings in the framework of bosonic field theory , where no other critical phenomena were found other than the dp .
hierarchy of unidirectionally coupled dp processes with many species show multicritical behavior @xcite .
coupled percolation processes have been also studied @xcite , where the absorbing phase become unstable with respect to an arbitrarily small branching rate even in one dimension .
predator - prey cellular automaton models @xcite in two dimension show dp universality class . coupled directed percolation ( dp ) processes with more than two species of particles ( in one dimension ) with different kind of interspecies coupling have shown dp - type @xcite transitions .
lotka - volterra like models in one dimension always show coexistence @xcite , either in form of well mixed states or as irregular bursts of the predator and prey population .
a four state predator prey model @xcite in one dimension with a restriction that a site can have at best one particle of each kind , shows an apt to an absorbing ( extinct ) state which belongs to dp - class . in this article we study a model of two species , say @xmath1 ( prey ) and @xmath2 ( predator ) , on a @xmath3-dimensional lattice .
each lattice site is either vacant @xmath4 or occupied by at best one particle of each kind .
the preys grow independently as @xmath5 and the predators die as @xmath6 , whereas they interact through a process @xmath7 , where birth of a new predator occurs instantly along with the death of the prey .
the system show a line of continuous absorbing state transition different from dp as the rates of these processes are tuned .
unlike other multispecies models , in the absorbing state both preys and predators do not separately extinct , rather predators extinct and preys proliferate to fill up the whole prey branch .
the model is defined on a one dimensional periodic lattice with lattice sites labelled by @xmath8 .
each site @xmath9 can be either vacant or occupied by a single particle @xmath1 ( prey ) , or occupied by a single particle @xmath2 ( predator ) or by both particles ( co - existing @xmath1 and @xmath2 ) , thus the model can be treated as 4-state predator - prey ( 4spp ) model .
more than one particle of any kind is not allowed . these hardcore restrictions on individual particles , where co - existence is allowed , can be realized alternatively by considering two separate branches , one for @xmath1 and the other for @xmath2 particles .
particles living in one branch can not move to the other branch , as schematically shown in the fig .
[ fig : model ] .
correspondingly , each site @xmath9 is associated with four states ; @xmath10 and @xmath11 , where @xmath12 ( @xmath13 ) denotes the presence ( absence ) of a particle at site @xmath9 . on a periodic lattice ,
these given below .
the prey ( @xmath1 ) can grow on their own branch with rate @xmath14 independent of the predators ( @xmath2 ) .
again , two neighboring predators ( @xmath2 ) die simultaneously with rate @xmath15 independent of @xmath1 due to their own crowding or overpopulation . here
@xmath16 in @xmath1-branch ( @xmath2-branch ) corresponds to an arbitrary state of @xmath1 ( @xmath2 ) particles i.e. presence or absence .
these two species interact with rate @xmath17 as follows ; when a predator @xmath2 at site @xmath9 meets a prey @xmath1 as the right neighbor @xmath18 , with site @xmath18 is not already occupied by a predator @xmath2 , then the predator @xmath2 eats the prey @xmath1 and simultaneously reproduces another predator @xmath2 . effect
on addition of symmetry and diffusion are briefly discussed towards the end .
let the steady state density of @xmath1 and @xmath2 particles be @xmath19 and @xmath20 respectively . clearly in absence of predators ( @xmath20 ) the prey density @xmath21 as the preys do not have an independent death process .
again , note that the dynamics do not allow @xmath1 particles to be eaten when they co - exist with @xmath2 on the same site .
this indicates that these @xmath1s can only die after their coexisting @xmath2s die , which can occur with rate @xmath15 if there is a @xmath2 particle present or created at the immediate left neighbor .
thus , for reasonably small death rate of predators , @xmath19 is expected to have a small value ( @xmath22 ) when @xmath23 and then it increases along with @xmath24 so , the prey density @xmath19 can never vanish .
the predator density @xmath20 can , however , become zero by repetitive death process .
the isolated @xmath2s wait until the prey invades their neighboring site and then they subsequently eat and reproduce with rate @xmath17 and die with rate @xmath15 .
so , along with the _ coexisting phase _ where both @xmath19 and @xmath20 are non - zero , we have another phase where @xmath21 and @xmath25 clearly , the later phase @xmath26 is absorbing as once all the predators die , even then the single surviving prey can lead to proliferation of the prey population in the whole lattice . of course , as argued earlier , @xmath19 can not vanish , and the other possible absorbing state @xmath27 which requires simultaneous death of all predators and preys , is not dynamically accessible . thus the 4spp model can undergo an absorbing state phase transition by tuning the different rates with @xmath20 as the order parameter .
our aim here is to study this critical behavior in details .
we have used the standard monte carlo methods to study the critical behavior of this model . from an initial arbitrary configuration , where each branch @xmath1 and @xmath2
are filled by arbitrary number of respective particles , the system is allowed to evolve according to the random sequential update following the monte carlo dynamics of 4spp model .
we have studied the system with different values of the rates @xmath14 , @xmath15 and @xmath17 with system size @xmath28 . for illustration ,
we have fixed two of the reaction rates , say @xmath29 , @xmath30 , and vary @xmath14 as the the control parameter in the following simulations .
for @xmath31 the average @xmath2 density @xmath20 decreases continuously until the system reaches a state with no predators ( @xmath32 ) and ultimately the whole prey lattice branch is filled with preys .
once reached in this absorbing state @xmath26 , the system remains there forever . while , for @xmath33 the average density of @xmath2 particles @xmath20 saturates to a nonzero value along with @xmath19 which also takes a value smaller than unity . in fig .
[ fig : phase](a ) we have plotted @xmath19 and @xmath20 as a function of @xmath14 for a system size @xmath34 as expected , the density of the preys @xmath19 never vanish and they proliferates in the whole lattice ( with @xmath21 ) in the absorbing state where predators are absent .
thus the system undergoes an absorbing state phase transition as the birth rate of preys @xmath14 crosses a critical threshold @xmath35 is plotted against the control parameter @xmath14 for fixed value of @xmath36 and @xmath28 , critical point is indicated at @xmath37 showing absorbing to active phase transition .
both densities are again plotted against @xmath15 keeping @xmath38 , @xmath30 fixed ; critical point is @xmath39 showing active to absorbing phase transition ( see inset ) .
( b ) , title="fig:",width=188,height=154 ] is plotted against the control parameter @xmath14 for fixed value of @xmath36 and @xmath28 , critical point is indicated at @xmath37 showing absorbing to active phase transition .
both densities are again plotted against @xmath15 keeping @xmath38 , @xmath30 fixed ; critical point is @xmath39 showing active to absorbing phase transition ( see inset ) .
( b ) , title="fig:",width=188,height=151 ] ) is plotted with fixed value of @xmath29,@xmath30 ; with different choices of @xmath40 shown from top to bottom .
the correct choice of @xmath37 and corresponding slope in logarithmic plot gives critical exponent @xmath41 .
( b ) time evolution of @xmath42 with @xmath29,@xmath30 . below the critical point for @xmath43 density @xmath42 eventually extincts and above the critical point @xmath44 density @xmath42 saturates . at the critical point @xmath45
, @xmath46 gives the critical exponent @xmath47 and the data collapse gives @xmath48.,title="fig:",width=188,height=154 ] ) is plotted with fixed value of @xmath29,@xmath30 ; with different choices of @xmath40 shown from top to bottom .
the correct choice of @xmath37 and corresponding slope in logarithmic plot gives critical exponent @xmath41 .
( b ) time evolution of @xmath42 with @xmath29,@xmath30 . below the critical point for @xmath43 density @xmath42
eventually extincts and above the critical point @xmath44 density @xmath42 saturates . at the critical point @xmath45
, @xmath46 gives the critical exponent @xmath47 and the data collapse gives @xmath48.,title="fig:",width=188,height=154 ] for @xmath33 , the order parameter @xmath20 shows power law behavior with the distance from criticality , @xmath49 when @xmath14 approaches @xmath50 this eq .
( [ eq : beta ] ) can be used to estimate @xmath51 and @xmath52 . as shown in fig .
[ fig : beta_nu](a ) , @xmath20 versus @xmath53 is linear in log scale , for the correct choice of @xmath37 ; the corresponding slope @xmath41 gives the estimated value of the order parameter exponent .
one can obtain few other critical exponents from the decay of the order parameter @xmath54 from an initial configuration with large number of predators . clearly , in the vicinity of critical point @xmath42 is a function of time @xmath55 and the temporal correlation length @xmath56 which vary as @xmath57 .
again , after an initial decay @xmath58 the predator density in the active phase approaches the steady state value @xmath20 in the @xmath59 limit .
so @xmath42 must scale as , @xmath60 thus , one expects that @xmath54 for different values of @xmath14 ( shown in the fig .
[ fig : beta_nu](b ) ) collapsed into a single scaling function @xmath61 , when @xmath62 is plotted against @xmath63 .
the main figure here shows the data collapse when we choose @xmath47 , and @xmath64 since at the critical point @xmath65 , one can obtain both @xmath51 and @xmath66 directly from the log scale plot of @xmath42 versus @xmath55 which is linear ( as shown in the inset of fig .
[ fig : beta_nu](b ) ) .
the resulting @xmath51 and @xmath66 are consistent with those obtained from the data collapse .
again , in the upper critical regime , @xmath20 vanishes as @xmath67 , in the @xmath68 limit .
this can happen only when the off - critical scaling function @xmath69 ; thus @xmath70 since all three exponents @xmath52 , @xmath66 and @xmath71 are calculated independently , one can check if the above scaling relation holds . in this case
it holds to a great accuracy for the values of @xmath52 , @xmath66 and @xmath71 calculated here .
plotted against the scaled variable @xmath72 for @xmath73 ( bottom to top in figures ) with @xmath45 , estimates @xmath74 .
( b ) the same data could be collapsed according to eq .
( [ eq : rhot_l1 ] ) by choosing @xmath75 which gives an estimate @xmath76 , title="fig:",width=188,height=154 ] plotted against the scaled variable @xmath72 for @xmath73 ( bottom to top in figures ) with @xmath45 , estimates @xmath74 .
( b ) the same data could be collapsed according to eq .
( [ eq : rhot_l1 ] ) by choosing @xmath75 which gives an estimate @xmath76 , title="fig:",width=188,height=154 ] now we turn our attention to the finite size scaling of @xmath42 at the critical point .
again , the system of length @xmath77 with a high density of predators @xmath78 decays as @xmath79 , indicating a scaling form @xmath80 where @xmath81 is the dynamic critical exponent .
thus , one expects @xmath42 for different values of @xmath77 to be collapsed to a single function when plotted against @xmath82 .
this is described in fig .
[ fig : zed_nu](a ) .
the inset there shows variation of @xmath54 for different system size @xmath73 , which were made to collapse to a single function using @xmath47 and @xmath74 . from the scaling relation @xmath83 one
expects that @xmath84 this can be verified from the modified scaling relation eq .
( [ eq : rhot_l ] ) . since @xmath85 , we have @xmath86 where @xmath87 in fig . [
fig : zed_nu](b ) we have plotted @xmath88 as a function of @xmath82 and found that the data for system size @xmath89 and @xmath90 could be collapsed into a single curve by choosing @xmath91 .
this gives us @xmath92 which is consistent with the earlier estimation .
the critical exponents of the model are summarized in table-[tab1 ] along with the the critical exponents of directed percolation universality class .
clearly the exponents , particularly @xmath52 and @xmath66 , are very different from those of dp - class , indicating that the 4spp model belongs to a different universality class .
we have studied the model by varying other rates also .
.[tab1 ] comparison of the critical exponents between the 4spp model with the dp universality class .
[ cols="^,^,^,^,^,^",options="header " , ] for example , one can take @xmath15 as the control parameter , keeping @xmath14 and @xmath17 fixed . the inset of fig .
[ fig : phase](a ) shows the variation of @xmath19 and @xmath20 as a function of @xmath15 for fixed @xmath38 and @xmath30 .
evidently , the order parameter @xmath20 vanishes continuously as @xmath15 crosses the threshold value @xmath93 the critical exponents obtained in this case was found to be consistent with table-[tab1 ] .
thus , for any fixed values of @xmath17 , one expects a line of criticality in the @xmath94-plane , which is shown in fig . [
fig : phase](b ) .
that , the critical behavior of the absorbing phase transition observed in the 4spp model is different from dp , can be visualized from the growth of clusters .
the space - time diagram , starting from an arbitrary initial configuration is shown in fig .
[ fig : space_time](a ) , where the occupancy of species @xmath1 and @xmath2 are represented separately in the upper and lower part respectively .
the evolution of clusters are visibly different from that of @xmath0 dimensional directed percolation model .
since the prey species can grow independently , the space is always filled locally by preys where predators are absent . again
, regular _ striped _ structures appear in these figures as the species do not diffuse . a natural question would be whether diffusion can drive the system to have an absorbing state transition belongs to dp universality class or not .
time steps with prey @xmath1 ( green ) and predator @xmath2 ( blue ) with system size @xmath28 for the regime @xmath31 and at @xmath95 : ( a ) with asymmetric rules ( no diffusion ) , ( b ) with asymmetric rules and _ explicit diffusion_.,title="fig:",width=170,height=170 ] time steps with prey @xmath1 ( green ) and predator @xmath2 ( blue ) with system size @xmath28 for the regime @xmath31 and at @xmath95 : ( a ) with asymmetric rules ( no diffusion ) , ( b ) with asymmetric rules and _ explicit diffusion_.,title="fig:",width=4,height=170 ] in the following we introduce diffusion of both the species _ explicitly _ in this model .
along with the usual dynamics of the 4spp model described earlier , both @xmath1 and @xmath2 particles are now allowed to move to the neighboring available vacant space in their respective branches . in this case
, it turns out that the clusters evolve more like the dp model ( see the space - time diagram in fig .
[ fig : space_time](b ) ) .
however , the detailed study of the absorbing phase transition reveals that the critical exponents are same as given in table-[tab1 ] . in presence of
_ explicit diffusion _ , we choose to study the system with fixed rates @xmath96,@xmath97 monte - carlo simulations show that the predator density @xmath20 vanishes continuously as @xmath14 is decreases below a critical threshold @xmath98 as described in fig .
[ fig : beta_nu_dfsm](a ) , near the critical point , @xmath99 with @xmath100 again , starting from a large number of predators , the density @xmath54 decays to its stationary value @xmath20 which is nonzero only in the upper - critical region @xmath101 we find that @xmath54 for different values of @xmath14 could be merged to an unique scaling function which satisfy eq .
( [ eq : rhot_eps ] ) by choosing @xmath102 and @xmath103 this data collapse is described in fig .
[ fig : beta_nu_dfsm](b ) .
the critical exponents @xmath104 and @xmath71 are more or less consistent with those listed in table-[tab1 ] .
it is not surprising that addition of explicit diffusion did not alter the universal behavior .
in fact , though slow , effective diffusion of predators was already occurring in the 4spp model through the rates @xmath15 and @xmath17 . ) with @xmath96 , @xmath30 with _ explicit diffusion _ : absorbing to active state phase transition at @xmath105 and the critical exponent @xmath106 ( see inset ) .
( b ) above the critical point for @xmath107 the density @xmath20 saturates and below the critical point for @xmath108 density @xmath20 eventually extincts . at the critical point @xmath105
, @xmath46 gives the critical exponent @xmath109 and @xmath110 .
, title="fig:",width=188,height=154 ] ) with @xmath96 , @xmath30 with _ explicit diffusion _ : absorbing to active state phase transition at @xmath105 and the critical exponent @xmath106 ( see inset ) .
( b ) above the critical point for @xmath107 the density @xmath20 saturates and below the critical point for @xmath108 density @xmath20 eventually extincts . at the critical point @xmath105
, @xmath46 gives the critical exponent @xmath109 and @xmath110 .
, title="fig:",width=188,height=154 ] in summary , we introduce a two species ( predator @xmath2 and prey @xmath1 ) model in one dimension where each lattice site is either vacant or occupied by , a single predator , a single prey or both .
more than one predator or prey are not allowed at any site .
the preys are allowed to grow independent of the predators , whereas two predators , if present at the neighboring sites , die simultaneously .
the species interact through a dynamics where the predator produces an offspring by eating a prey from its neighbor .
when the predator density @xmath32 , even a single prey can invade the whole lattice by its independent birth process .
thus , the system has an unusual absorbing state @xmath111 the other absorbing state @xmath112 is not dynamically accessible . using dynamical monte - carlo simulation
we show that the system shows an absorbing state phase transition , as the birth rate of prey @xmath14 is increased beyond a critical value @xmath51 keeping death rate of predator @xmath15 fixed . for a fixed @xmath17 , the line of criticality ( @xmath51 as a function of @xmath15 )
is governed by a universality class different from the most generic one , namely directed percolation .
this critical behavior is found to be robust against addition of explicit diffusion .
note that the dynamical rules of the model is different from other four species predator prey models studied earlier @xcite in a way that the predators in 4spp model can not eat the prey at the same site .
we believe that , non - equilibrium phase transition to a unusual and unique absorbing state , may result in different universality class .
further study in this direction could clarify this issue . |
cosmic rays ( cr ) pervade the galaxy . in the dense interstellar medium ( ism )
, they play a crucial role as they are the primary source of ionization , starting a complex chemistry which leads to hundreds molecules , and influencing the star and planet formation processes by creating ions , which in turn are coupled with the magnetic fields , this regulating gravitational collapse .
there is little doubt that cr are accelerated in the expanding shocks of supernova remnants ( snrs , e.g. hillas 2005 ; caprioli et al .
high energy ( @xmath9 1 gev ) cr ( mainly protons ) interact with hydrogen atoms and produce @xmath1 rays via @xmath2 decay ( hayakawa 1952 ; stecker 1971 ) . in this case
, the predicted @xmath1-ray luminosity l@xmath10 is proportional to the @xmath2-decay @xmath1-ray emissivity , which depends on the local cr density , and to the irradiated molecular cloud mass ( e.g. aharonian & atoyan 1996 ) .
hence , giant molecular clouds close to , or even better , penetrated by snrs can be bright @xmath1-ray sources .
conversely , molecular clouds associated with bright @xmath1-ray sources can probe enhanced cr densities .
several cr interaction sites have now been tentatively identifed with the latest generation of gev / tev observatories ( e.g. montmerle 2010 ) .
the main difficulty remains to distinguish between emission from proton and/or bremsstrahlung or inverse compton from electrons . even in well - documented cases such as ic443 ( albert et al .
2007 ) and w28 ( aharonian et al . 2008 ) snrs , the gev - tev emission mechanism remains unclear .
in other cases ( in particular for w51c ; abdo et al .
2009 ; feinstein et al .
2009 ) @xmath2-decay appears to be the dominant @xmath1-ray emission mechanism . in these cases ,
the derived local relativistic proton density is very high , typically one to two orders of magnitude higher than the average galactic cr density .
such high densities should also have visible effects at lower cr energies ( @xmath11 1 gev ) , in the regime where cr ionize molecular clouds ( e.g. kamae et al .
2006 ; gabici et al .
2009 ; padovani et al . 2009 ; fatuzzo et al .
2010 ) .
we propose that the physical interaction between sn - accelerated energetic particles and molecular gas , where an association is suggested by the presence of a @xmath1-ray source , can be demonstrated by using the impact of low - energy cr on the cloud chemistry .
the method is based on the determination of the ionization degree of the molecular cloud , which in turn gives a measure of the cr ionization rate @xmath12 ( to be compared with the `` standard '' value @xmath13 s@xmath5 for dense clouds ; glassgold & langer 1974 ) . in dense gas , the ionization can be obtained by measurements of the dco@xmath14/hco@xmath14 abundance ratio ( e.g. gulin et al . 1977 ) .
briefly , hco@xmath14 and dco@xmath14 are formed by the reaction of co with h@xmath15 and h@xmath16d@xmath14 respectively : co + h@xmath15 @xmath17 hco@xmath14 + h@xmath16 and co + h@xmath16d@xmath14 @xmath17 dco@xmath14 + h@xmath16 .
consequently , the dco@xmath14/hco@xmath14 ratio directly depends on the h@xmath16d@xmath14/h@xmath15 only . in molecular clouds ,
cr ionize the gas by ionizing h and h@xmath16 at a rate @xmath12 and forming the molecular ion h@xmath15 , whereas h@xmath16d@xmath14 is formed by the reaction of h@xmath15 with hd . since both molecules are destroyed by the reaction with the most abundant neutral species , co , and by the recombination with electrons , the dco@xmath14/hco@xmath14 ratio is an almost direct measure of the gas ionization degree .
the method has been extensively applied to derive the ionization degree of molecular clouds and dense cores .
the measured values range between @xmath18 and @xmath19 , depending on the gas density , leading to estimates of @xmath12 between @xmath20 and @xmath21 s@xmath5 ( caselli et al .
1998 : hereinafter cwth98 ) . finally , indriolo et al ( 2010 ) recently reported line absorption observations of h@xmath15 , measuring the ionization in the _ diffuse gas _ of the outer parts of the cloud associated with ic443 .
their measurements indicate values of @xmath12 only five times larger than the average in diffuse clouds . in this letter , we report observations of the dco@xmath14/hco@xmath14 ratio towards the _ dense gas _ of a cloud interacting with the snr w51c .
w51c ( also known as j1923 + 141 and g49.2 - 0.7 in the literature ) is a well known snr at a distance of @xmath22 kpc ( kundu & velusamy 1967 ) .
the snr extends about 30@xmath23 ( equivalent to @xmath2460 pc ) , is about @xmath25 yrs old ( koo et al . 1995 ) and is associated with a molecular cloud , engulfed by the blast wave , whose mass is @xmath26 m@xmath27 and average density @xmath28 @xmath29 ( koo & moon 1997 ) .
five hii regions lie at the border of the snr , the w51a and b star forming regions .
observations of hi and co line emission have shown the presence of shocked material north - west , at about a distance of 10 pc from the center of the snr ( koo et al .
1995 ; 1997 ) , where oh maser emission is also detected ( hewitt et al . 2008 ) . in this work
, we targeted five positions of the molecular cloud ( table [ tab : obs - tel ] ) .
four of them lie a few arcmin away from the shocked region , while one ( point a ) lies at the northern edge of it .
an extended ( @xmath30 pc ) gev tev source has been detected by hess ( feinstein et al . 2009 ) and fermi - lat ( abdo et al . 2009 ) , with a total @xmath1-ray luminosity of @xmath31 erg s@xmath32@xmath33 ( abdo et al .
2009 ) , making w51c one of the most luminous @xmath1-ray sources of our galaxy .
based on the observed gev
tev @xmath1-ray spectrum , abdo et al .
( 2010 ) claim that almost 99@xmath34 of the observed @xmath1-rays is due to the decay of @xmath2 mesons produced in inelastic collisions between accelerated protons and target gas .
we observed five positions roughly sampling the cloud overlapping the hess emission ( 2 ; tab .
[ tab : obs - tel ] ) . to constrain the gas physical conditions , we observed the @xmath35co and c@xmath36o 10 and 21 transitions , while to measure the ionization degree , we observed the h@xmath35co@xmath14 10 and dco@xmath14 21 transitions .
the observations were obtained using the apex and iram-30 m telescopes ( tab . [
tab : obs - tel ] ) .
the amplitude calibration was done typically every 15 min , and pointing and focus were checked every 1 and 2 hrs respectively ensuring @xmath37 accuracy .
the apex spectra were obtained in the classical on - off mode , where the off at eq 2000 coordinates 19:22:44.3 , 14:05:50.0 shows no signal .
the fast fourier transform spectrometer facility was used as a backend . for the iram observations
, we used the vespa autocorrelator , and the frequency - switching mode .
all spectra were reduced using the class package ( hily - blant et al 2005 ) of the gildas software .
residual bandpass effects were subtracted using low - order ( @xmath38 3 ) polynomials .
@llcccccccc species & transition & frequency & main beam & hpbw & telescope & journal & res .
& t@xmath39 & + & & ( ghz ) & efficiency & ( arcsec ) & & & ( mhz ) & ( k ) & + & 10 & 109.782 & 0.79 & 22 & iram & may 2009 & 0.08 & @xmath40 & + & 21 & 219.560 & 0.75 & 28 & apex & dec 2008 & 0.5 & @xmath41 & + & 10 & 110.201 & 0.79 & 22 & iram & may 2009 & 0.08 & @xmath40 & + & 21 & 220.399 & 0.55 & 12 & iram & may 2009 & 0.08 & @xmath40 & + & 21 & 220.399 & 0.75 & 28 & apex & dec 2008 & 0.5 & @xmath41 & + & 10 & 86.754 & 0.79 & 28 & iram & may 2009 & 0.08 & @xmath40 & + & 21 & 144.077 & 0.79 & 17 & iram & may 2009 & 0.08 & @xmath40 & + position & coordinates ( j2000 ) & @xmath42 & & & h@xmath35co@xmath14 & dco@xmath14 + & ( b1950 ) & & 10 & 21@xmath43 & 21@xmath44 & 10 & 21 & 10 & 21 + a & 19:22:53.8 + 14:15:44 & 71.2 & 4.30(0.03 ) & 6.20(0.05 ) & 9.00(0.1 ) & 0.30(0.03 ) & 0.45(0.04 ) & 0.040(0.006 ) & ( 0.01 ) + & 19:20:35.9 + 14:09:54 & & & & & & & & + b & 19:22:31.0 + 14:15:45 & 69.0 & 7.60(0.04 ) & 9.20(0.05 ) & 17.85(0.2 ) & 0.50(0.05 ) & 1.10(0.04 ) & ( 0.01 ) & ( 0.02 ) + & 19:20:13.1 + 14:09:57 & & & & & & & & + c1 & 19:22:21.3 + 14:05:12 & 69.0 & 14.00(0.04 ) & 19.40(0.10 ) & 18.80(0.1 ) & 2.00(0.03 ) & 4.00(0.06 ) & 0.23(0.01 ) & ( 0.02 ) + c2 & 19:22:21.3 + 14:05:12 & 73.0 & 7.70(0.04 ) & 14.80(0.10 ) & - & 1.00(0.03 ) & 2.70(0.06 ) & 0.08(0.01 ) & ( 0.02 ) + & 19:20:03.2 + 13:59:24 & & & & & & & & + d & 19:23:27.3 + 14:16:24 & 65.5 & 2.80(0.07 ) & 4.20(0.07 ) & 6.30(0.1 ) & 0.15(0.02 ) & 0.30(0.06 ) & ( 0.007 ) & ( 0.01 ) + & 19:21:09.4 + 14:10:32 & & & & & & & & + e & 19:23:08.0 + 14:20:00 & 67.0 & 11.70(0.10 ) & 15.00(0.10 ) & 15.6(0.1 ) & 2.20(0.06 ) & 4.20(0.06 ) & 0.60(0.01 ) & 0.040(0.006 ) + & 19:20:50.1 + 14:14:09 & & & & & & & & + table [ tab : obs - tel ] summarises the observed signals at each point and each observed transition .
@ccccccccc & [ @xmath35co]/ & temp & density & n(c@xmath36o ) & av & extent & n(h@xmath35co@xmath14 ) & [ dco@xmath14]/ + & [ c@xmath36o ] & ( k ) & ( @xmath46@xmath29 ) & ( @xmath47@xmath33 ) & ( mag ) & ( arcsec ) & ( @xmath48@xmath33 ) & [ hco@xmath14 ] + a & 15 & @xmath49 & @xmath50 & 0.40.6 & 26 & 19 & 34 & @xmath51 + b & 15 & @xmath52 & @xmath53 & 1.52.5 & 714 & 27 & @xmath54 & - + c1 & 15 & 2328 & 620 & 3.94.1 & 1624 & @xmath5528 & 79 & @xmath56 + c2 & 10 & @xmath57 & 415 & 1.8 - 2.5 & 814 & @xmath5528 & 68 & @xmath58 + d & 10 & @xmath49 & @xmath50 & 0.20.9 & 15 & 20 & @xmath59 & - + e & 10 & 2124 & 820 & 3.94.1 & 1624 & @xmath5528 & 1618 & 0.0012 - 0.0016 + co emission is detected in the four observed transitions in all points .
emission is , on the contrary , detected towards three positions only : a , c and e. finally , dco@xmath14 10 emission is detected only towards the position e. the spectra towards the position e are shown in fig.[fig : pose ] .
in general , there are two clouds in the line of sight , at different velocities v@xmath60 around 69 and around 73 km / s respectively .
the line widths are similar in all points and transitions , about 3 km / s .
the comparison of the @xmath35co 21 observed with apex and iram gives a direct measurement of the emitting sizes .
points b , c and e show extended ( @xmath61 ) emission , while in points a and d the sizes are about 20@xmath62 .
co 21 , @xmath35co 10 , c@xmath36o 21 , c@xmath36o 10 , h@xmath35co@xmath14 10 and dco@xmath14 21 .
the signals are in main beam temperature k. the vertical line shows 67 km / s.,width=302 ]
we used the non - lte lvg code described in ceccarelli et al .
( 2003 ) , with the co - h@xmath16 collisional coefficients by wernli et al .
( 2006 ) for the first 6 co levels and the temperature range 5 - 70 k. we run a grid of models with temperatures between 10 and 50 k , densities between @xmath46 and 10@xmath63 @xmath29 , c@xmath36o column densities between @xmath64 and @xmath65 @xmath33 and @xmath35co / c@xmath36o ratio between 5 and 15 .
for each point , we used the extent derived by the comparison of the @xmath35co 21 line intensity obtained at apex and iram telescopes , respectively .
the values of the parameters minimising the @xmath66 are reported in tab . [
tab : signals ] ( the range corresponds to @xmath66=0.2 contours ) .
the h@xmath35co@xmath14 column density has been derived by comparing the observed signals with non - lte lvg computations using the collisional coefficients by flower ( 1999 ) and the density and temperature derived from the previous step .
the h@xmath67co@xmath14 column density is then obtained by multiplying the n(h@xmath35co@xmath14 ) by 50 ( koo & moon 1997 ) .
note that the [ dco@xmath14]/[hco@xmath14 ] ratio very little depends on the actual gas temperature , for the range of temperatures considered here .
the clouds in the direction of points c1 and e have temperatures somewhat larger than 20 k , while the temperature is unconstrained in the other points .
the densities are around @xmath26 @xmath29 in points c1 , c2 and e , and unconstrained in the other points .
the visual extinction a@xmath68 is few mag in points a and d , and it is larger than 10 mag in the other points . it is interesting and intriguing that points c1 and e have similar physical conditions but different column densities of hco@xmath14 and dco@xmath14 . using the analytical method mentioned in 1 ( eqs . 1 and 2 in cwth98 ) , from the n(dco@xmath14)/n(h@xmath35co@xmath14 )
ratio we derive an extremely high ionization degree toward the point e , @xmath69 .
this value is one or two orders of magnitude larger than those derived in other dark clouds by cwth98 .
that point e is `` special '' with respect to the dark clouds of the cwth sample is indeed already clear from the observed signals : in those clouds the dco@xmath14 1 - 0 line intensity is of the same order of magnitude than the h@xmath35co@xmath14 1 - 0 line intensity ( butner et al .
1995 ) , whereas we observed more than ten times lower dco@xmath14 1 - 0 than the h@xmath35co@xmath14 1 - 0 line intensity .
this is a hallmark that the ionization degree is much higher , and does not leave much of uncertainty on this conclusion .
however , the analytical method is an oversimplification , especially for high ionization degrees , as more routes of hco@xmath14 formation become important in addition to the h@xmath15+co one . in order to correctly evaluate the chemical structure of the cloud towards the point e , including the penetration of the interstellar ( is ) uv photons and a ( more ) complete chemical network , we used the photo - dissociation region ( pdr ) models by le petit et al .
( 2006 ) . to help facilitate
the discussion , figure [ fig : family ] shows the structure of a 20 mag cloud whose density is @xmath70 @xmath29 , like point e , illuminated on the two sides by the is uv field and with different cr ionization rates .
as found by several previous studies , for a low value of @xmath12 ( @xmath71 s@xmath5 ) and a low abundance of metals , the ionization through the cloud is governed by the uv photons penetration and depends on the abundance of carbon ( a@xmath722 mag ) and sulphur ( 2@xmath38a@xmath724 mag ) , and on the crs ionization deeper inside .
the ionization at the interior of the cloud is low , @xmath73@xmath74 , in a phase called lip ( low ionization phase ) in the literature ( pineau des forts et al .
this situation is represented by the case @xmath12=@xmath75 s@xmath5 in fig . [
fig : family ] . in general , as @xmath12 increases the ionization in the cloud interior ( a@xmath76 mag ) increases roughly quadratically until it reaches a critical point where it abruptly jumps to high values , in a phase called hip ( high ionization phase ) .
the lip / hip transition is a result of the non - linear nature of the chemical network of molecular gas ( e.g. boger & sternberg 2006 ; rollig et al .
2007 ) , which have been claimed to present a bistable nature ( le bourlot et al . 1993 ) .
the spatial location of this transition depends upon various control parameters , amongs which the density , the abundance of he and metals , and the cr ionization rate ( e.g. wakelam et al .
2006 ) . in practice , for a cloud of a given density , the lip / hip jump is governed by the metal abundances , specifically s , si , fe and mg going inwards in the cloud , as metals are responsible for the ( non - linear ) charge exchange .
as shown by fig .
[ fig : family ] , the zones closer to the cloud edge flip from lip to hip first , and the lip / hip jump moves further into the cloud with increasing @xmath12 , until the whole cloud is in the hip , regardless the metal abundances
. the hco@xmath14 abundance is affected by @xmath12 ( the larger the @xmath12 the larger the hco@xmath14 formation rate ) and by the presence of the hip region ( the larger the ionization degree the larger the hco@xmath14 destruction rate ) . in general
, the hco@xmath14 abundance is roughly divided into three zones : in the skin of the cloud , it is very low because of the photo - dissociation by the uv photons , deeper in the cloud it increases but the high ionization in the hip region keeps the hco@xmath14 abundance low , and at still larger av it increases , when eventually the gas jumps into the lip .
the ratio n(co)/n(hco@xmath14 ) of column densities , integrated over the 20 mag model cloud , is shown in fig .
[ fig : cr - constrain ] .
it decreases with @xmath12 ( as the hco@xmath14 abundance increases ) , until a critical point ( at @xmath77 s@xmath5 depending on the metals abundance ) where the whole cloud is in the hip state and the hco@xmath14 abundance becomes very low all across the whole cloud .
note that for high metal abundances the cloud is always in the hip state . as noted by wakelam et al .
( 2006 ) , the ionization distribution is bimodal with the metal abundances .
in fact , the theoretical n(co)/n(hco@xmath14 ) ratio either follows the high or the intermediate metal abundance curves of our fig . [ fig : cr - constrain ] . @xmath29 density illuminated by the is uv field .
electrons ( red curves ) , hco@xmath14 ( blue curves ) and 100xdco@xmath14 ( green curves ) abundances ( with respect to h nuclei ) for four different values of the cr ionization rate : 3 ( dotted line ) , 100 ( dashed line ) , 126 ( solid line ) and 150 ( dashed - dotted line ) @xmath78 s@xmath5 .
the lip and hip regions in the four cases are also marked ( see text ) .
the computations refer to the metal low abundances case . only the 10 mag from the edge of the cloud
are shown .
note that the @xmath12=126@xmath78 s@xmath5 case likely describes the situation towards point e.,width=321 ] the dco@xmath14 abundance follows qualitatively the behaviour of hco@xmath14 and the hco@xmath14/dco@xmath14 ratio across the cloud depends on the ionisation degree ( @xmath40 in the lip and @xmath79 in the hip zones , respectively ) .
the dependence of the n(hco@xmath14)/n(dco@xmath14 ) ratio as function of @xmath12 is shown in fig .
[ fig : cr - constrain ] . in practice ,
the values of n(co)/n(hco@xmath14 ) and n(hco@xmath14)/n(dco@xmath14 ) depend on the relative weight of the lip - to - hip zones in the integral over the cloud .
the comparison between the predicted n(co)/n(hco@xmath14 ) and n(hco@xmath14)/n(dco@xmath14 ) ratios with the values observed towards the point e provides a value of @xmath80 1.01.3 @xmath81 s@xmath5 .
the likely structure of the cloud in the point e is shown in fig .
[ fig : family ] . in practice ,
a large fraction of the cloud is the hip state whereas the rest is in the lip state .
we stress , however , that , given the high non - linearity of the equations , the exact value of @xmath12 which fits the observations depends on the details of the adopted model ( wakelam et al .
2006 ; boger & sternberg 2006 ) .
regardless of these details , the modeling indicates a situation where high ( hip ) and low ( lip ) ionisation must be both present , to give the observed hco@xmath14/dco@xmath14value which is an intermediate value of the two regions .
the relative amounts of hip / lip may change with other control parameter values , but the result is that the line of sight we probe must contain both lip and hip . finally , cr heat the cloud as they deposit energy so that the relatively large value of the gas temperature ( 2124 k ) toward point
e ( tab . [ tab : signals ] ) also favours the high @xmath12 .
( blue ) and n(hco@xmath14)/n(dco@xmath14 ) ( green ) integrated over the 20 mag cloud .
three families of theoretical curves are reported with different metal abundances , following wakelam et al ( 2006 ; see also cwth98 ) : high- ( dotted ) , low- ( solid ) and intermediate- ( dashed ) abundances .
the boxes show the observed values towards the point e and , for comparison , in the cwth98 sample .
the right panel shows a zoom around the solution .
the red stars show the computed models .
, width=321 ] finally , the lower limit to the n(hco@xmath14)/n(dco@xmath14 ) towards point c1 suggests a value for @xmath12 larger than in point e.
our observations lead to two important conclusions : + 1 ) the cloud towards the point e has a cr ionization rate @xmath82 s@xmath5 , i.e. enhanced by about a factor 100 with respect to the standard value ( glassgold & dalgarno 1974 ) .
this is the first time that such an high ionization degree is measured in a dense molecular cloud .
the fact that this cloud coincides with a tev @xmath1-rays source close to a snr conforts the idea that it is irradiated by an enhanced flux of freshly formed low - energy cr .
we have identified a location close to a cr accelerator with both enhanced low - energy cr ionization and high cr @xmath1-ray emission , suggesting novel ways to study the acceleration and the diffusion of cr close to sn shocks .
+ 2 ) the high low - energy cr flux induces the simultaneous presence of a region of high- and low- ionization , called hip and lip in the literature . to our best knowledge , this is the first time that such a situation has been observed and offers the possibility to study in more detail the chemistry of the two states simultaneously .
therefore , a more general implication of this work is that tev sources identified with snrs interacting with molecular clouds are promising sites not only to study cr acceleration , but also to study unusual chemical conditions .
abdo a.a . , ackermann m. , ajello m.et al .
2009 , apj 706 , l1 aharonian f.a . &
atoyan a.m. 1996 , a&a 309 , 917 aharonian f.a .
, akhperjanian a.g .
, bazer - bachi a.r .
2008 , a&a 481 , 401 albert j. , aliu e. , anderhub h. et al .
2007 , apj 664 , l87 boger g.i .
& sternberg a. 2006 , apj 645 , 314 butner h.m . ,
, loren r.b .
1995 , apj 448 , 207 caprioli d. , blasi p. , amato e. 2011 , aph 34 , 447 caselli p. , walmsley c.m .
, terzieva r. , herbst e. 1998 , apj 499 , 234 ( cwth98 ) ceccarelli c. , maret s. , tielens a.g.g.m .
, castets a. , caux e. 2003 , a&a 410 , 587 fatuzzo m. , melia f. , todd e. , adams f.c .
2010 , apj 725 , 515 feinstein et al .
2009 , aipc 1112 54f flower d.r . 1999 ,
mnras 305 , 651 gabici s. , aharonian f.a . , casanova s. 2009 , mnras 396 , 1629 glassgold a.e . & langer w.d .
1974 , apj 193 , 73 gulin m. , langer w.d . , snell r.l .
, wootten h.a .
1977 , apj 217 , l165 hayakawa s. 1952 , pthph 8 , 571 hewitt j.w . , yusef - zadeh f. , wardle m. 2008 , apj 683 , 189 a.m. hillas 2005 , j. phys .
g : nucl . part .
phys . , 31 hily - blant p. , pety j. & guilloteau s. 2005 , class manual , http://www.iram-institute.org/medias/uploads/class-evol1.pdf le bourlot j. , pineau des forts g. , roueff e. , schilke p. 1993
, apj 416 , l87 le petit f. , nehm c. , le bourlot j. , roueff e. 2006 , apjs 164 , 506 kamae t. , karlsson n. , mizuno t. , abe t. , koi t. 2006 , apj 647 , 692 koo b - c .
, kim k - t .
, seward f.d .
1995 , apj 447 , 211 koo b - c . &
moon d - s .
1997 , apj 485 , 263 kundu m.r .
& velusamy t. 1967 , anap 30 , 59 montmerle t. 2010 , aspc 422 , 85 padovani m. , walmsley c.m . , tafalla m. , galli d. , mller h.s.p .
2009 , a&a 505 , 1199 pineau des forts g. , roueff e. , flower d.r .
1992 , mnras 258 , 45 rollig m. , abel n.p . , bell t. , bensch f. , black j. et al .
2007 , a&a 467 , 187 stecker f.w . 1971 , nature 234 , 28 wakelam v. , herbst e. , selsis f. , massacrier g. 2006 , a&a 459 , 813 wernli m. , valiron p. , faure a. , wiesenfeld l. , jankowski p. , szalewicz k. 2006 a&a 446 , 367 | cosmic rays pervade the galaxy and are thought to be accelerated in supernova shocks .
the interaction of cosmic rays with dense interstellar matter has two important effects : 1 ) high energy ( @xmath0 1 gev ) protons produce @xmath1-rays by @xmath2-meson decay ; 2 ) low energy ( @xmath3 1 gev ) cosmic rays ( protons and electrons ) ionize the gas .
we present here new observations towards a molecular cloud close to the w51c supernova remnant and associated with a recently discovered tev @xmath1-ray source .
our observations show that the cloud ionization degree is highly enhanced , implying a cosmic ray ionization rate @xmath4 s@xmath5 , i.e. 100 times larger than the standard value in molecular clouds .
this is consistent with the idea that the cloud is irradiated by an enhanced flux of freshly accelerated low - energy cosmic rays . in addition , the observed high cosmic ray ionization rate leads to an instability in the chemistry of the cloud , which keeps the electron fraction high , @xmath6 , in a large fraction ( av@xmath7mag ) of the cloud and low , @xmath8 , in the interior .
the two states have been predicted in the literature as high- and low - ionization phases ( hip and lip ) .
this is the observational evidence of their simultaneous presence in a cloud . |
the present paper introduces a consistent notation for the weierstrass elliptic function @xmath5 , the roots @xmath6 , and the half - periods @xmath4 , where each half - period @xmath7 and its associated root @xmath8 satisfy the definition @xmath9 .
the notation for the half - periods @xmath10 depends on the signs of @xmath2 and the modular discriminant defined as @xmath11 . the standard notation for the weierstrass half - periods @xmath12 when @xmath13 follows refs .
@xcite . in the case
@xmath14 , the half - period @xmath15 is real and the half - period @xmath16 is imaginary .
the half - periods @xmath17 , therefore , form a rectangular lattice with the fundamental period parallelogram defined with corners located at @xmath18 . when @xmath2 changes sign , i.e. , @xmath19 , the fundamental period parallelogram is rotated clockwise by @xmath20 to a new parallelogram with corners located at @xmath21 , where we used @xmath22 .
the notation for the half - periods @xmath23 adopted here will also follow this standard notation when @xmath13 . when @xmath24 changes sign , i.e. , @xmath25 , the standard notation @xcite states that the half - period @xmath26 is real and the half - period @xmath27 is complex - valued ( where @xmath28 is purely imaginary ) . the half - periods @xmath29 form a rhombic lattice with the fundamental period parallelogram defined with corners located at @xmath30 .
when @xmath2 changes sign , i.e. , @xmath31 , the fundamental period parallelogram is rotated clockwise by @xmath20 to the new fundamental period parallelogram with corners located at @xmath32 , i.e. , @xmath33 is now defined as @xmath34 when @xmath31 . the new notation for the half - periods @xmath10 adopted here when @xmath35 will instead assume that @xmath36 , as in the standard case @xmath13 .
the purpose of the present notes is to introduce a consistent classification of the roots and half - periods based on the orbit solutions derived for all energy levels associated with unbounded and bounded motion in a cubic potential .
additional notes on the analysis of periodic solutions in classical mechanics involving doubly - periodic elliptic functions are presented in refs .
@xcite .
the remainder of this paper is organized as follows . in sec .
[ sec : cubic ] , we introduce a parameterization of the turning points @xmath37 for particle motion in a cubic potential based on the definition @xmath38 , where @xmath39 and @xmath40 denotes the particle s total energy . in sec .
[ sec : weierstrass ] , we introduce a consistent formulation of the weierstrass elliptic function @xmath0 associated with the parameterization introduced in sec .
[ sec : cubic ] . in sec .
[ sec : w_orbits ] , the weierstrass orbit solutions are presented for all energy - level classes for the motion in a cubic potential .
lastly , the jacobi orbit solutions are presented in sec .
[ sec : j_orbits ] based on connection relations between the weierstrass and jacobi elliptic functions .
we consider the motion of a ( unit - mass ) particle in the cubic potential @xmath41 which has an inflection point at @xmath42 , a minimum at @xmath43 , with @xmath44 and @xmath45 ( i.e. , @xmath43 is a stable equilibrium point ) , and a maximum at @xmath46 , with @xmath47 and @xmath48 ( i.e. , @xmath46 is an unstable equilibrium point ) .
the period of small oscillations near the stable equilibrium point at @xmath43 is @xmath49 .
the equation of motion for a particle of unit mass and total energy @xmath40 is @xmath50 where @xmath51 denotes the particle velocity . the right side of eq .
is expressed in the weierstrass standard form with the invariants @xmath39 and @xmath52 , where bounded and unbounded orbit solutions are classified according to four energy levels summarized in table [ tab : energy_levels ] .
the total energy @xmath53 is defined in terms of the weierstrass phase @xmath54 , which is defined along a continuous path in the complex-@xmath54 plane : @xmath55 which allows us to go continuously from @xmath56 ( lowest energy level in fig .
[ fig : orbits ] ) to @xmath57 ( highest energy level in fig . [ fig : orbits ] ) .
.[tab : energy_levels]energy levels @xmath58 defined in terms of the weierstrass phase . [ cols="<,^,^",options="header " , ]
we now express the orbit solutions for each of the four energy - level regions shown in table [ tab : energy_levels ] .
these orbit solutions are shown in fig .
[ fig : weierstrass_cubic_orbits ] . for region ( i )
, with energy levels @xmath56 and @xmath25 , only unbounded ( u ) motion is possible since @xmath59 is real and @xmath60 are complex - conjugate . using the initial condition @xmath61 and @xmath62 from table [ tab : periods ]
, we find @xmath63 since @xmath64 is real and @xmath65 , then the complete unbounded orbit @xmath66 can be mapped out for @xmath67 . for region ( ii ) , with energy levels @xmath68 and @xmath14 ,
unbounded ( u ) and bounded ( b ) motions are possible for @xmath69 and between @xmath70 . for the bounded - motion solution , using the initial condition @xmath71 and @xmath72 from table [ tab : periods ] , we find the periodic solution @xmath73 with a period @xmath74 , so the bounded orbit @xmath75 can be mapped out for @xmath76 . at the half - period @xmath77
, we find @xmath78 we note that as the energy level approaches @xmath79 ( from above ) , the period approaches @xmath80 , which is the expected period of small oscillations in the vicinity of the stable equilibrium point at @xmath43 , where @xmath45 .
the bounded - motion solution approaches @xmath81 as @xmath82 .
for the unbounded - motion solution , using the initial condition @xmath83 and @xmath84 from table [ tab : periods ] , we find @xmath85 once again , since @xmath86 is real and @xmath87 , then the complete unbounded orbit @xmath88 can be mapped out for @xmath89 . for region ( iii ) , with energy levels @xmath90 and @xmath19 , unbounded ( u ) and bounded ( b ) motions are possible for @xmath91 and between @xmath92 , where @xmath93 and @xmath94 .
for the bounded - motion solution , using the initial condition @xmath95 and @xmath96 from table [ tab : periods ] , we find the periodic solution @xmath97 where we used the inversion formula .
this bounded solution has the period @xmath98 , so the bounded orbit @xmath99 can be mapped out for @xmath100 . at at the half - period @xmath101 , we find @xmath102 we note that as the energy level approaches @xmath103 ( from below ) , the period approaches @xmath104 while the bounded - motion solution approaches the separatrix solution ( with @xmath105 and @xmath106 ) @xmath107 where @xmath108 . here , we recover @xmath109 and , as @xmath110 , we find @xmath111 . for the unbounded - motion solution , using the initial condition @xmath112 and @xmath113 where we used the inversion formula .
once again , since @xmath114 is real and @xmath115 , then the complete unbounded orbit @xmath116 can be mapped out for @xmath117 .
the unbounded separatrix solution @xmath118 may be obtained from a slight modification of the unbounded solution .
first , we define the new unbounded solution in region ( iii ) to be @xmath119 which assumes that @xmath120 and @xmath121 the unbounded separatrix solution may be written as @xmath122 where @xmath123 and @xmath124 .
lastly , for region ( iv ) , with energy levels @xmath57 and @xmath31 , only unbounded ( u ) motion is possible since @xmath125 is real and @xmath126 are complex - conjugate . using the initial condition @xmath127 and @xmath128 from table [ tab : periods ]
, we find @xmath129 where we used the inversion formula .
we note that at @xmath130 , or @xmath131 , we find @xmath132 so that the complete unbounded orbit @xmath133 can be mapped out for @xmath134 . = 2.5 in the weierstrass orbit solutions are shown in fig .
[ fig : weierstrass_cubic_orbits ] , with the bounded and unbounded separatrix solutions ( for @xmath105 and @xmath106 ) shown as dashed curves .
solutions found inside the bounded separatrix solution are the bounded solutions and , while solutions found outside are the unbounded solutions , , , and .
the bounded solution in region ( ii ) is periodic between @xmath135 , while the bounded solution in region ( iii ) is periodic between @xmath136 .
the unbounded solutions [ , , ] in regions ( i)-(iii ) have a turning points at @xmath137 , while the unbounded solution in region ( iv ) has a turning point at @xmath138 .
lastly , the potential has stable equilibrium point at @xmath43 ( shown as a square dot in fig .
[ fig : weierstrass_cubic_orbits ] ) and an unstable equilibrium point @xmath46 , shown in fig .
[ fig : weierstrass_cubic_orbits ] as the point where the bounded and unbounded separatrix solutions cross .
lastly , by using the @xmath2-inversion formula , we observe the following relations between bounded and unbounded orbit solutions @xmath139 hence , bounded motion can be viewed in terms of unbounded motion in imaginary time .
in this last section , we proceed with the transformation of our weierstrass orbit solutions into expressions involving the jacobi elliptic functions @xcite @xmath140 , @xmath141 , and @xmath142 , where @xmath143 and the modulus @xmath144 is defined as @xmath145}. \label{eq : m_phi_def}\end{aligned}\ ] ] here , for @xmath68 ( region ii ) , we have @xmath146 and @xmath147 , while for @xmath90 ( region iii ) , we have @xmath148 and @xmath149 . for @xmath150 ( regions
i and iv ) , @xmath151 is a complex - valued number : @xmath152 & & ( e < -1/2 ) \\
\exp\left[i\frac{}{}\chi(\psi)\right ] & & ( e > 1/2 ) \end{array } \right .
\label{eq : m_psi}\ ] ] where the phase @xmath153 $ ] vanishes at @xmath154 and reaches @xmath155 at @xmath156 .
hence , the relation between @xmath157 ( for @xmath56 ) and @xmath158 ( for @xmath57 ) is @xmath159 , where @xmath160}{\sin[(\pi + \phi)/3]}. \label{eq : mprime_phi_def}\end{aligned}\ ] ] for energy levels @xmath150 , we easily verify that @xmath161 . the jacobi elliptic functions @xmath140 , @xmath141 , and @xmath142 have real and imaginary periods determined by the complete elliptic integrals of the first kind @xmath162 and @xmath163 , respectively : @xmath164 for all integers @xmath165 and @xmath166 .
in addition , these functions have zeroes at @xmath167 , @xmath168 , and @xmath169 , as well as singularities ( poles ) at @xmath170 .
lastly , they satisfy the following identities @xmath171 , @xmath172 , and have the following limits : @xmath173 , @xmath174 , @xmath175 , and @xmath176 , @xmath177 the function @xmath178 satisfies the differential equation @xmath179 the function @xmath180 satisfies the differential equation @xmath181 and the function @xmath182 satisfies the differential equation @xmath183 other jacobi elliptic functions that will be useful below are @xmath184 , which satisfies the differential equation @xmath185 and @xmath186 , which satisfies the differential equation @xmath187 by comparing eqs . and , for example , we obtain the following identity @xmath188 while eq
. yields the identity @xmath189 .
we begin with bounded orbit solutions in region ( ii ) and ( iii ) , where @xmath190 .
first , we substitute @xmath191 into eq . to obtain @xmath192\;\left[(x_{2 } - x_{3 } )
- \alpha y^{2}\right].\ ] ] next , using the definition , we define @xmath193 and @xmath194 , so that we obtain the differential equation , whose solution is @xmath195 . hence , in regions ( ii ) and ( iii ) , the bounded orbit solution is @xmath196 the bounded separatrix solution is obtained from eq . in the limit @xmath197 : @xmath198 which is identical to eq . .
since the jacobi elliptic function @xmath199 has a period @xmath200 , where @xmath162 is the complete elliptic integral of the first kind , the period of the orbit solution is @xmath201 for @xmath79 , we find @xmath202 , with @xmath203 and @xmath204 , so that the period is @xmath205 , where we used @xmath206 . in the limit @xmath197 @xmath207 , @xmath208 since @xmath209 .
when @xmath210 ( @xmath211 ) , we find @xmath212 and the period is @xmath213 .
we now consider the unbounded orbit solutions in regions ( i)-(iii ) , where @xmath214 .
first , we substitute @xmath215 into eq . to obtain @xmath216\;\left[y^{2}(x_{1 } - x_{3 } ) + \alpha\right],\ ] ] where the solution @xmath217 must satisfy the initial condition @xmath218 .
next , using the definition , we define @xmath193 and @xmath219 , so that we obtain the differential equation , whose solution is @xmath220 .
hence , in regions ( i)-(iii ) , the unbounded orbit solution is @xmath221 in regions ( ii ) and ( iii ) , all roots @xmath222 are real and @xmath223 .
the unbounded orbit solution reaches infinity after a time @xmath224 when @xmath225 . in region
( i ) , where @xmath226 is real and @xmath227 are complex - conjugate roots , and the definition for @xmath151 implies that it is complex - valued .
we note that the unbounded separatrix solution can not be obtained from eq .
since @xmath228 when @xmath103 ( or @xmath229 ) .
if we use the initial condition @xmath230 , i.e. , @xmath231 , and the definitions @xmath232 , we obtain the differential equation , whose solution is @xmath233 .
hence , the unbounded orbit solution in region ( iii ) can also be expressed as @xmath234 so that the unbounded separatrix solution becomes @xmath235 which is identical to eq . . in region ( iv ) , where @xmath57 , @xmath236 , and @xmath237 , the parameter @xmath151 is a complex - valued number with unit magnitude @xmath238 .
the unbounded orbit solution @xmath239 given by eq .
is still applicable in this region .
using the fact that @xmath199 is infinite at @xmath240 for @xmath241 .
hence , at a finite ( complex ) time @xmath242 , the particle reaches infinity .
we now wish to verify that the jacobi orbit solutions derived in this section satisfy the relations between the bounded and unbounded solutions .
if we use the relation , eq . becomes @xmath243 where we used the fact that @xmath244 and we recover the relation .
next , we look at the bounded solution in region ( iii ) @xmath245 and the bounded solution in region ( ii ) @xmath246 hence , the relations , which were originally obtained through the @xmath2-inversion relation , are also obtained through the transformation @xmath247 , even when @xmath151 is complex - valued .
a consistent notation for the half - periods @xmath4 of the weierstrass elliptic function @xmath248 was constructed from the analysis of the motion of a particle in a cubic potential . a guide to the accurate interpretation of the mathematica output for the function
weierstrasshalfperiods@xmath249 was also provided . | a consistent notation for the weierstrass elliptic function @xmath0 , for @xmath1 and arbitrary values of @xmath2 and @xmath3 , is introduced based on the parametric solution for the motion of a particle in a cubic potential .
these notes provide a roadmap for the use of mathematica to calculate the half - periods @xmath4 of the weierstrass elliptic function .
october 13 , 2015 |
the recent work of ercolani , mclaughlin , and pierce @xcite analyzed the fine structure of families of probability measures on the space of @xmath1 hermitian matrices .
the principle measures analyzed in @xcite are of the form @xmath2 \right ) dm,\ ] ] where @xmath3 is the product measure @xmath4 the potential is @xmath5 and where @xmath6 is the normalization of the measure . the question of interest in that paper was the asymptotic structure of @xmath7 for large values of @xmath8 , and its analytic dependence on @xmath9 . more generally , the partition function of random matrices , introduced above as the normalization factor of the probability measure ( [ mu ] ) , is defined as @xmath10 where the integral is taken over the space of @xmath1 hermitian matrices .
the partition function has been used extensively as a model for the partition function of 2d - quantum gravity @xcite , @xcite , @xcite .
derivatives of @xmath11 evaluated at @xmath12 capture the gaussian moments @xmath13 .
these moments are evaluated with the wick lemma in terms of pair correlators of the matrix entries . in analogy to feynman diagrams , there is a correspondence between labeled maps and the terms in the wick expansion .
this fact led to the conjecture @xcite that the logarithm of the partition function ( [ part ] ) has an asymptotic expansion of the form @xmath14 = n^2 e_0(t_3 , t_4 , \dots , t_{2\nu } ) + e_1(t_3 , t_4 , \dots , t_{2\nu } ) + \dots\ ] ] where the coefficients @xmath15 are generating functions for counting labeled , connected , oriented maps embedded on a riemann surface of genus @xmath16 .
a map @xmath17 on a compact , oriented connected surface @xmath18 is a pair @xmath19)$ ] where 1 .
@xmath20 is a connected 1-complex ; 2 .
@xmath21 $ ] is an isotopical class of inclusions @xmath22 ; 3 .
the complement of @xmath20 in @xmath18 is a disjoint union of open cells ( faces ) ; 4 .
the complement of @xmath23 ( vertices ) in @xmath20 is a disjoint union of open segments ( edges ) .
the number of edges connected to a vertex @xmath24 is called the degree of @xmath25 .
we may represent maps as fat or ribbon graphs embedded on an oriented connected surface @xmath18 .
see also the theory of dessins denfants @xcite .
ercolani and mclaughlin @xcite rigorously showed that @xmath7 possesses an asymptotic expansion of the form @xmath26 in a non - trivial @xmath27 domain .
the functions @xmath15 are analytic in a neighborhood of @xmath28 and are counting functions for labeled maps .
the function @xmath29 where @xmath30 is the number of labeled , genus @xmath16 , maps with @xmath23 containing : @xmath31 vertices of degree @xmath32 , @xmath33 vertices of degree @xmath34 , @xmath35 , @xmath36 vertices of degree @xmath0 .
harer and zagier @xcite solved this enumeration problem in the special case of 1-vertex maps ( or monopoles ) .
let @xmath37 be the generating function for oriented maps with a single vertex , that is @xmath38 where @xmath39 is the number of oriented maps with 1 vertex , and @xmath40 edges , which are embedded into a genus @xmath16 surface .
this generating function is a finite series in @xmath8 and is given by the matrix integral @xmath41 dm,\ ] ] where the integral is taken over the space of @xmath42 hermitian matrices , and @xmath43 dm \right]^{-1}.\ ] ] harer and zagier explicitly evaluated ( [ g_n(n)_int ] ) , to find that @xmath44 bauer and itzykson @xcite show that the data of a labeled map is equivalent to a pair of permutations satisfying some compatibility conditions .
this result gives an efficient method to determine the genus of a map given by its pair of permutations .
what emerges is an algorithm for counting the number of genus @xmath16 maps with a particular structure on @xmath23 .
the idea is to count the number of pairs of permutations satisfying the characterization which produce a genus @xmath16 map .
zvonkin @xcite expanded on this idea .
the particular partition function @xmath45 has been the subject of an extensive calculation @xcite .
the conclusion of this calculation is a prescription for calculating @xmath46 .
explicit formulas for these functions are found for low values of @xmath16 ( @xmath47 ) in terms of an auxiliary function . in the case of @xmath48
this result produces a closed form expression for the taylor coefficients @xmath49 . in
the other cases worked out explicitly this result is used to find the first few taylor coefficients of each of the functions .
a purely combinatoric method for producing an expression for @xmath50 in terms of generating functions of fundamental objects is given in @xcite .
the algorithm detailed in this paper has been developed as an independent method for checking the detailed work on the fine structure of @xmath46 carried out in @xcite and of deriving the constants of integration for the formulas derived in that work , and in the author s doctoral thesis .
as of the writing of this paper the calculation has been done for the taylor coefficients satisfying @xmath51 and we have found exact agreement for all of these terms .
we will conclude this paper with a modification of our method for counting the number of unoriented maps or mobius graphs .
these are maps which are embedded into an unoriented surface . to be more precise the objects we will count are mobius graphs , ribbon graphs where the edges are allowed to twist .
this is the setting of interest if the partition function ( [ part ] ) is replaced with @xmath52 dm , \ ] ] where the integral is over @xmath1 symmetric matrices .
this function is called the partition function of the gaussian orthogonal ensemble ( goe ) . in this case
the hypothesis is that @xmath53 where ( at least formally ) @xmath54 where @xmath55 is the number of labeled , euler characteristic @xmath56 , unoriented maps with @xmath23 containing : @xmath31 vertices of degree 3 , @xmath33 vertices of degree 4 , @xmath35 , @xmath36 vertices of degree @xmath0 .
the justification of this enumeration problem is given in @xcite and @xcite .
the existence of the expansion ( [ goe_expansion ] ) is not rigorously shown to date .
in particular a proof is needed of the analyticity of @xmath57 .
we will give a heuristic argument for the counting property of @xmath57 .
goulden and jackson @xcite generalized the calculation of harer and zagier @xcite for 1-vertex maps ( monopoles ) .
let @xmath58 be the generating function for unoriented maps with a single vertex , that is @xmath59 where @xmath60 is the number of unoriented maps with 1 vertex , and @xmath40 edges , which are embedded into a genus @xmath16 surface .
this generating function is a finite series in @xmath8 and is given by the matrix integral @xmath61 dm,\ ] ] where the integral is taken over the space of @xmath42 symmetric matrices , and @xmath62 dm \right]^{-1}.\ ] ] [ gj_thm ] goulden and jackson explicitly evaluated ( [ f_n(n)_int ] ) , to find that @xmath63 our algorithm agrees with the numbers given by @xmath58 for all cases checked .
what we do not have is a method for generalizing theorem [ gj_thm ] to larger families of unoriented maps . in section [ wick_sec ]
we outline the heuristic argument which motivates using the terms of the asymptotic expansion of the partition function ( [ part ] ) as the generating functions of the number of maps of genus @xmath16 . in section [ group_sec
] we illustrate the result of @xcite which gives a one - to - one correspondence between maps and pairs of permutations , of a restricted type .
so the group structure provides an algorithm for counting the number of maps of a specified type .
section [ results_sec ] is a collection of tables summarizing some of the calculations we have done with this algorithm .
it is provided as a means of checking with other known results such as the 1-face maps @xcite , @xcite and those in @xcite and @xcite .
in addition the tables contain all the numbers which were needed to evaluate the constants of integration found in @xcite . in section [ sym_sec ]
we redo the heuristic argument of section [ wick_sec ] in the setting of an integral over the symmetric matrices rather than the hermitian matrices .
we show that the equivalent counting problem we are concerned with is that of counting maps embedded into an unoriented surface . in section [ unoriented ]
we generalize the results of section [ group_sec ] and associate to each unoriented map a triple of permutations .
this gives a one - to - one correspondence between unoriented maps and triples of permutations of the given type .
we will first outline the heuristic argument that the @xmath64 are counting functions of labeled genus @xmath16 maps ( see @xcite and @xcite ) .
this calculation centers around the wick lemma for gaussian expectation s : the wick lemma states [ wicklemma ] if @xmath65 is a gaussian expectation , and @xmath66 are homogeneous linear functions then @xmath67 where the sum is taken over the set @xmath68 and @xmath69 .
we are working with the gaussian expectation @xmath70 dm , \ ] ] where the integral is taken over the space of @xmath1 hermitian matrices , and @xmath71 is a normalizing constant .
the quadratic expectations of this probability are @xmath72 consider the partition function @xmath73 dm .\ ] ] we may expand out the @xmath74 expression as a taylor series and ( formally at least ) commute this sum with the integration to write @xmath75^n \rangle .\ ] ] write ( [ eq1 ] ) as a function of the entries of @xmath76 : @xmath77 if we use wick s lemma on the @xmath78 term of ( [ eq2 ] ) we find that the contribution is given by @xmath79 each of these terms is non - zero if @xmath80 and @xmath81 satisfy the conditions for the corresponding quadratic expectations to be non - zero . in this example
we find the conditions are respectively @xmath82 , @xmath83 , and @xmath84 . in an analogy to feynman
diagrams these reductions are cataloged by families of maps ( see for example figure [ fig1 ] ) .
moreover the number of relations between the indices is the number of faces of the corresponding map , while the contribution to the expectation @xmath85 \rangle$ ] is given by the euler characteristic of the map .
the fact that the maps organizing the contributions to the hermitian matrix integral ( [ hermitian_matrix ] ) are oriented is because of ( [ gue_exp ] ) .
the final step is to take a logarithm of @xmath86 to restrict our consideration to connected maps .
this argument easily generalizes to include other times .
we will now outline the group data associated to a map , the data comes in the form of a pair of permutations .
these two permutations provide a mechanism for counting the number of maps by computing the number of permutations satisfying some conditions .
conveniently this permutation structure also provides an efficient method of computing the structure of the faces of the map and therefore of calculating the euler characteristic of the map .
let @xmath88 be the number of edges of @xmath20 .
we will view the permutations of @xmath89 as acting on the set of darts of @xmath20 , which is represented by @xmath90 where @xmath91 and @xmath92 the element @xmath93 represents the dart based at @xmath94 and going along @xmath95 .
the element @xmath96 represents the dart based at @xmath94 going along @xmath95 in the counterclockwise ( resp .
clockwise ) orientation . for each edge
there are two darts therefore @xmath97 and we can think of @xmath89 as acting by permutations on the set @xmath98 . given a map , @xmath99)$ ] ,
we define a subgroup of @xmath89 generated by two permutations @xmath100 .
the orientation on @xmath18 induces ( via @xmath21 $ ] ) a cyclic ordering on the darts attached to each vertex ; the first permutation @xmath101 is given by this action .
explicitly , @xmath101 maps the element @xmath102 to the element @xmath103 or @xmath104 where @xmath105 is the edge counter clockwise in the orientation at @xmath94 from @xmath95 .
likewise , @xmath101 maps the element @xmath96 to the element @xmath103 or @xmath106 where @xmath105 is the edge counter clockwise in the orientation at @xmath94 from @xmath95 .
the second permutation @xmath107 , is given explicitly as the permutation which acts on @xmath108 by sending @xmath109 to @xmath110 where @xmath111 is the other endpoint of @xmath95 ; and @xmath107 acts on @xmath112 by sending @xmath113 to @xmath114 .
one sees that @xmath101 is a product of independent cycles ; the length of the cycles in @xmath101 correspond to the degrees of the vertices in @xmath23 .
the permutation @xmath107 is a product of @xmath88 disjoint 2-cycles .
the orbit of @xmath115 ( the group generated by @xmath101 and @xmath107 ) of any element in @xmath98 , is all of @xmath98 because @xmath20 is connected . to compute the genus of a map from @xmath116
we compute the genus of the associated graph from the euler characteristic @xmath117 where @xmath118 , the number of vertices , is the number of cycles in @xmath101 ; where @xmath88 , the number of edges , is @xmath119 the number of elements acted upon by @xmath101 and @xmath107 , or is the number of 2-cycles in @xmath107 ; and where @xmath120 , the number of faces , is the number of cycles of the permutation @xmath121 ( this is merely computing each face by finding all the edges which border it ) .
this leads to the theorem : there is a one - to - one correspondence between connected maps @xmath122)$ ] and pairs of permutations ( @xmath101 , @xmath107 ) , where @xmath101 is a product of disjoint cycles , and @xmath107 a product of @xmath88 disjoint 2-cycles , which satisfy connectedness . to complete the proof
we show that given ( @xmath101 , @xmath107 ) as above one may determine @xmath17 .
the condition of connectedness is that the orbit of the group @xmath123 is all @xmath124 letters .
the following steps produce a map : 1 .
let @xmath23 be the set of cycles of @xmath101 ; 2 .
let @xmath125 be the set of 2-cycles of @xmath107 ; 3 .
each 2-cycle in @xmath107 ties together two points of @xmath23 identifying the two endpoints of that element of @xmath125 ; 4 .
the euler characteristic ( [ euler ] ) gives the genus of @xmath18 ; 5 .
each cycle in @xmath101 induces the orientation on @xmath18 in a neighborhood of @xmath126 , this is sufficient to determine both @xmath18 and @xmath127 up to isotopy ( see section [ unoriented ] ) .
the punchline is that what we now have is an algorithm for computing @xmath30 for finite @xmath16 , @xmath31 , @xmath33 , @xmath35 , @xmath36 : let @xmath128 fix @xmath101 to be a permutation formed by @xmath31 cycles of length 3 , @xmath33 cycles of length 4 , @xmath35 , @xmath36 cycles of length @xmath0 formed from a permutation of @xmath40 letters .
then we choose each product of @xmath88 disjoint 2-cycles in @xmath89 .
check if @xmath116 is connected ( by verifying that the orbit of @xmath129 is all @xmath124 letters ) .
if @xmath116 is connected compute the genus as above ( [ euler ] ) with @xmath130 and bin the result . as an example we will show how this algorithm works by computing @xmath131 .
fix @xmath132 . then choose each @xmath107 which is a product of 2-cycles of @xmath133 .
for example , we might choose @xmath134 ; for this choice of @xmath107 the orbit of @xmath135 forms two disjoint sets @xmath136 and @xmath137 . choose @xmath138 ; for this choice of @xmath107 the orbit of @xmath135 is the set @xmath139 , therefore @xmath116 represent a map .
we compute the genus of this map by computing the number of distinct orbits of @xmath140 .
we find that @xmath141 , therefore the euler characteristic is @xmath142 and the genus is @xmath28 .
choose @xmath143 ; for this choice of @xmath107 the orbit of @xmath135 is the set @xmath139 , therefore @xmath116 represent a map .
we find that @xmath144 , therefore the euler characteristic is @xmath145 and the genus is @xmath146 .
in tables [ table1]-[table3 ] we present a selection of the results of this calculation , included are the numbers which were needed for the calculations in @xcite .
the algorithm can also be used to compute the number of maps of mixed type . for the example from section [ example_section ] of 2 vertices of degree 3 and one of degree 4
we find 432 genus 0 maps and 468 genus 1 maps .
for the number of maps with one vertex of degree 3 , one of degree 4 , and one of degree 5 : 2160 ( genus 0 ) , 6480 ( genus 1 ) , and 1440 ( genus 2 ) . .
[ table1 ] first the number of one vertex maps found by @xcite ( whose duals are the well studied 1-face maps @xcite ) [ cols=">,>,<,<,<,<,<,<",options="header " , ]
this section is motivated by the partition function @xmath147 given by ( [ zngoe ] ) .
we will now show , using the wick lemma , that the asymptotic expansion of @xmath148 enumerates unoriented maps .
this will be a formal calculation only , the necessary analyticity of the terms of this expansion has not been rigorously shown .
we present the calculation here as a motivation for the counting problem we are computing .
we are working with the gaussian expectation @xmath149 dm,\ ] ] where the integral is taken over the space of @xmath1 symmetric matrices , and @xmath71 is a normalizing constant .
the quadratic expectations of this probability are @xmath150.\ ] ] consider the partition function @xmath151 dm.\ ] ] we may expand out the @xmath74 expression as a taylor series and ( formally at least ) commute this sum with the integration to write @xmath152 write this expression as a function of the entries of @xmath76 : @xmath153 if we use wick s lemma on the @xmath78 term we find that the contribution is given by @xmath154 each of these terms is non - zero if @xmath155 and @xmath81 satisfy the conditions for the corresponding quadratic expectations to be non - zero . as in the case of oriented maps these reductions may be cataloged by families of maps . in this case
the maps are allowed to have edges that reverse orientation because of formula ( [ quad ] ) .
focus on the middle term of ( [ wick ] ) @xmath156 there are four cases which give that this contribution is non - zero : @xmath157 , @xmath158 , @xmath159 , or @xmath160 , these correspond to the four unoriented maps represented by the mobius graphs in figure [ fig2 ] .
the final step is to take a logarithm of @xmath161 to restrict to connected unoriented maps .
we will lift the algorithm for counting labeled connected oriented maps to count labeled connected unoriented maps .
we will first go over the modifications which must be made to the group data to account for the independence that the local orientations of each vertex have . an unoriented map on a compact unoriented connected surface @xmath18 is a pair @xmath162 ) $ ] where 1 .
@xmath20 is a connected 1-complex ; 2 .
@xmath21 $ ] is an isotopical class of inclusions @xmath163 ; 3 .
the complement of @xmath20 in @xmath18 is a disjoint union of open cells ( faces ) ; 4 .
the complement of @xmath23 ( vertices ) in @xmath20 is a disjoint union of open segments ( edges ) . in the same way that ribbon graphs are associated to an oriented map we associate a mobius graph ( ribbon graph with orientation reversing and orientation preserving edges ) to an unoriented map . given an unoriented map @xmath17 , there are three permutations associated to the data contained in @xmath17 . to begin , for each vertex , @xmath164 choose a local orientation , @xmath165 induced by the embedding @xmath127 .
we will view the four permutations as acting on the set of darts @xmath166 , given explicitly as : @xmath167 where @xmath168 and @xmath169 the element @xmath170 represents the dart based at @xmath94 pointing along @xmath95 together with the local orientation @xmath165 ( resp .
@xmath171 ) .
the element @xmath172 represents the dart based at @xmath94 which traverses @xmath95 in the counter clockwise direction with respect to the local orientation ( resp .
clockwise direction with respect to the local orientation ) together with the local orientation @xmath165 . likewise
the element @xmath173 represents the dart based at @xmath94 which traverses @xmath95 in the counter clockwise direction with respect to the local orientation ( resp .
clockwise direction with respect to the local orientation ) together with the normal vector @xmath171 .
an edge is said to be orientation reversing if @xmath165 at @xmath94 on the edge is @xmath174 after translating to @xmath111 along the edge ; otherwise we say that the edge is orientation preserving .
this is a property of both the map and the choice of local orientations .
we find three permutations @xmath175 : on @xmath108 we define the action of @xmath176 by : * @xmath177 acts by sending @xmath178 to @xmath179 . * @xmath101 acts by sending @xmath180 ( resp .
@xmath181 ) to the dart which is counter clockwise ( resp . clockwise ) from @xmath109 at @xmath94 , maintaining the orientation @xmath165 ( resp .
@xmath171 ) .
* @xmath107 acts by sending @xmath178 to @xmath182 when @xmath95 is orientation preserving or @xmath183 when @xmath95 is orientation reversing . on @xmath112
we define the action of @xmath175 by : * @xmath177 acts by sending @xmath184 to @xmath185 . * @xmath101 acts by sending @xmath186 ( resp .
@xmath187 ) to the dart which is counter clockwise ( resp . clockwise ) from @xmath113 at @xmath94 , maintaining the orientation @xmath165 ( resp .
@xmath171 ) .
* @xmath107 acts by sending @xmath186 to @xmath188 ( resp . sending @xmath187 to @xmath189 ) depending on whether @xmath107 is orientation preserving or reversing from this definition the following properties are apparent : 1 .
@xmath190 and @xmath191 2 .
@xmath177 and @xmath107 are fixed point free permutations , each is a disjoint product of 2-cycles .
no 2-cycle of @xmath107 is a 2-cycle of @xmath177 .
3 . @xmath101 is a disjoint product of @xmath192 3-cycles , @xmath193 4-cycles , @xmath35 , @xmath194 @xmath0-cycles .
4 . a connected map has the property that @xmath195 has a single orbit .
@xmath107 contains the data of how the darts of @xmath20 are attached together to form the map ; together with @xmath177 , @xmath107 determines how the local orientations relate globally .
the cycles of @xmath121 correspond to faces of the map .
there are two cycles of equal length for each face , representing the two directions the face can be traced along its edges .
this leads to the theorem : there is a one - to - one correspondence between unoriented maps and triples @xmath176 satisfying conditions ( 1 - 6 ) .
we will complete the proof by showing that given three permutations @xmath175 satisfying conditions ( 1 - 6 ) , we can reconstruct ( up to the choice of local orientation ) the map @xmath196)$ ] and the surface @xmath18 .
a. let @xmath23 be the set of cycle pairs @xmath197 in @xmath101 such that @xmath198 .
b. from each pair @xmath197 choose @xmath199 to represent the counter clockwise or up orientation .
c. let @xmath125 be the set of 2-cycle pairs @xmath200 in @xmath107 such that @xmath201 .
each pair @xmath200 connects two ( possibly the same ) elements of @xmath23 .
if @xmath200 connects the up orientation to the up orientation this edge is orientation preserving , otherwise it is orientation reversing .
d. to construct the surface @xmath18 and the embedding class @xmath21 $ ] do the following : a. for
each vertex @xmath197 draw a polygon with sides labeled and oriented by the elements that @xmath199 acts on .
b. for an orientation preserving edge connecting two vertices , glue together the corresponding sides of their polygons so that the relative orientations agree at the corresponding side .
for an orientation reversing edge connecting two vertices , glue together the corresponding sides of their polygons so that the relative orientations are opposite at the side . in conclusion : up to the choice of local orientations at each vertex , the data given by @xmath175 produces a map . likewise , a map , up to the choice of local orientations at each vertex , produces a triple @xmath176 satisfying conditions 1 - 6 .
we now have all the tools necessary to lift our algorithm to the enumeration of unoriented maps .
fix @xmath177 to be a fixed point free disjoint product of 2-cycles .
fix @xmath101 to be a product of @xmath192 3-cycles , @xmath193 4-cycles , @xmath35 , and @xmath194 @xmath0-cycles , such that @xmath177 connects cycles in @xmath101 into pairs @xmath197 of the same cycle type , choose @xmath199 to represent the @xmath202 orientation .
cycle through all fixed point free products of disjoint 2-cycles @xmath107 such that conditions 1 - 6 are satisfied by @xmath175 , in particular no 2-cycle in @xmath177 appears in @xmath107 .
if , for this choice of @xmath107 , the corresponding map is connected , then compute the euler characteristic : the number of vertices is half the number of cycles in @xmath101 , the number of edges is half the number of 2-cycles in @xmath107 , and the number of faces is half the number of cycles in @xmath140 .
bin the results sorted by the euler characteristic .
the algorithm presented here for counting oriented maps was used in @xcite and the authors dissertation to evaluate constants of integration appearing in the explicit calculation of @xmath203 for low genus ( @xmath204 ) . additionally the algorithm provided a useful method for verifying the validity of the results in @xcite . in looking beyond the calculations carried out in @xcite it became apparent that it would be useful to have a similar algorithm to compute the number of unoriented maps .
the permutation structure associated to oriented maps has a precise analogue in unoriented maps .
this method may prove useful in further studies of the goe and gse matrix integrals , and associated problems .
9 m. bauer and c. itzykson , _ triangulations _ , discrete mathematics * 156 * , ( 1996 ) 29 - 81 .
d. bessis , c. itzykson , and j.b .
zuber , _ quantum field theory techniques in graphical enumeration _ , advances in applied mathematics , * 1 * , ( 1980 ) , no . 2 , 109 - 157 . j. bouttier , p. di francesco , and e. guitter , _ census of planar maps : from the 1-matrix model solution to a combinatorial proof _ , nucl .
b. * 645 * ( 2002 ) , 477 - 499 .
p. difrancesco , p. ginsparg , and j. zinn - justin , _ 2d gravity and random matrices _ , physics reports , * 254 * , ( 1995 ) , 1 - 133 . n. m. ercolani and k. d. t .-
mclaughlin , _ asymptotics of the partition function for random matrices via riemann - hilbert techniques and application to graphical enumeration _ , international mathematics research notes * 14 * ( 2003 ) , 755 - 820 .
n. m. ercolani , k. d. t .-
mclaughlin , and v. pierce , _
random matrices , graphical enumeration and the continuum limit of the toda lattices _ , math - ph/0606010 ( 2006 ) .
goulden , and d.m .
jackson , _ maps in locally orientable surfaces and integrals over real symmetric surfaces _ , canadian journal of mathematics , * 49 * , 865 - 882 , ( 1997 ) .
j. harer , and d. zagier , _ the euler characteristic of the moduli space of curves .
_ inventiones mathematica , * 85 * ( 1986 ) , 457 - 485 . m. mulase , a. waldron , _ duality of orthogonal and symplectic matrix integrals and quaternionic feynman graphs _ , communications in mathematical physics , * 240 * , 553 - 586 ( 2003 ) .
l. schneps , editor , _ the grothendieck theory of dessins denfants _ , london mathematical society lecture notes series , * 200 * , cambridge university press ( 1994 ) .
e. witten , _ two - dimensional gravity and intersection theory on moduli space .
_ surveys in differential geometry * 1 * , ( 1991 ) , 243 - 310 .
a. zvonkin , _ matrix integrals and map enumeration an accessible introduction _ ,
mathematics and computational modeling , * 26 * ( 1997 ) , no .
8 - 10 , 281 - 304 .
a. zvonkin , _ how to draw a group ?
_ discrete mathematics , * 180 * , 403 - 413 , ( 1998 ) . | bauer and itzykson showed that associated to each labeled map embedded on an oriented riemann surface there was a group generated by a pair of permutations . from this result an algorithm may be constructed for enumerating labeled maps , and this construction is easily augmented to bin the numbers by the genus of the surface the map is embedded in .
the results agree with the calculations of harer and zagier of 1-vertex maps ; with those of bessis , itzykson , and zuber of 4-valent maps ; and with those of ercolani , mclaughlin , and pierce for @xmath0-valent maps .
we then modify this algorithm to one which counts unoriented maps or mobius graphs .
the results in this case agree with the calculation of goulden and jackson on 1-vertex unoriented maps . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.