1: \documentclass{jetpl} \twocolumn
2: \usepackage{amsmath}
3: \usepackage{amsthm}
4: \usepackage{amsfonts}
5: \usepackage{amssymb}
6: \usepackage{graphicx}
7: \newcommand{\lb}[1]{\label{#1}}
8: \newcommand{\nn}{\nonumber}
9: \newcommand{\sbr}[1]{{\langle #1 \rangle}} %single bracket
10: \newcommand{\dbr}[1]{{\langle\!\langle #1 \rangle\!\rangle}} %double bracket
11: \newcommand{\beq}{\begin{equation}}
12: \newcommand{\ee}{\end{equation}}
13: \newcommand{\bea}{\begin{eqnarray*}}
14: \newcommand{\eea}{\end{eqnarray*}}
15: \newcommand{\EP}{\epsilon}
16: \input{epsf.tex}
17: %%% article in English
18: \lat
19:
20: %%% declaration of a new mathematical operator
21: %\DeclareMathOperator{\sign}{sign}
22: \def\eqn{\begin{equation}}
23: \def\[{\begin{equation}}
24: \def\]{\end{equation}}
25: \def\refs{\cite}
26: \def\exp{{\rm exp}}
27: \def\sh{{\rm sinh}}
28: \def\ch{{\rm cosh}}
29: \def\sin{{\rm sin}}
30: \def\cos{{\rm cos}}
31: \def\cth{\rm coth}
32: %\def\frac#1#2{{#1\over#2}}
33:
34: %%% article title
35: \title{On scaling fields in $Z_N$ Ising models}
36:
37: %%% article title - for colontitle (at the top of the page)
38: \rtitle{On scaling fields in $Z_N$ [Ising] models}
39:
40: %%% article title - for table of contents (usualy identical with \title)
41: \sodtitle{On scaling fields in $Z_N$ models}
42:
43: %%% author(s) ( + e-mail)
44: \author{V.~A.~Fateev$^{a,b}$, V.~V.~Postnikov$^c$, Y.~P.~Pugai$^a$
45: %\thanks{e-mail: slava@landau.ac.ru}
46: }
47:
48: %%% author(s) - for colontitle (at the top of the page)
49: \rauthor{V.~A.~Fateev, V.~V.~Postnikov, Y.~P.~Pugai}
50:
51: %%% author(s) - for table of contents
52: \sodauthor{V.~A.~Fateev, V.~V.~Postnikov, Y.~P.~Pugai}
53:
54: %%% author's address(es)
55: \address{$^a$L.~D.~Landau Institute for Theoretical Physics RAS,
56: 117940 Moscow, Russia}
57:
58: \address{$^b$ Laboratoire de Physique
59: Th\'eorique
60: et Astroparticules, Universit\'e Montpellier II, 34095 Montpellier, France}
61:
62:
63: \address{$^c$ Sochi's branch of People Friendship University, Sochi, Russia}
64: %%% dates of submition & resubmition (if submitted once, second argument is *)
65: \dates{\today}{*}
66: %%% abstract
67: \abstract{We study the space of scaling fields in the $Z_N$
68: symmetric models with the factorized scattering and propose simplest
69: algebraic relations between form factors induced by the action of
70: deformed parafermionic currents. The construction gives a new free
71: field representation for form factors of perturbed
72: Virasoro algebra primary
73: fields, which are parafermionic algebra descendants.
74: We find exact vacuum expectation values of
75: physically important fields and study correlation functions of order
76: and disorder fields in the form factor and CFT perturbation
77: approaches.}
78:
79: %%% PACS numbers
80: \PACS{74.50.+r, 74.80.Fp}
81:
82: \begin{document}
83: \maketitle
84:
85: %\section{Introduction}
86:
87: {\it 1. Introduction}
88: %
89:
90: \par\noindent
91: The calculation of exact correlation functions is one of the most complicated
92: and interesting problems in the two dimensional integrable quantum
93: field theory.
94:
95: Remarkable solutions of this problem were found for the free field
96: models including the Ising model without a magnetic field and for
97: the conformal field theories (CFT) \cite{BPZ,DotsFat,ZaFa85}. One of
98: the difficulties in investigating massive integrable field theories
99: with a non-trivial interaction is related with the fact that the
100: space of scaling fields and, in particular, the space of different
101: solutions of form factor equations, has no clear and simple
102: description in terms of dynamical symmetry algebras like the free
103: fermion algebra in the Ising model or Virasoro algebra in CFT.
104: Existence of such infinite-dimensional algebras drastically
105: simplifies the problem and allows applications of powerful algebraic
106: methods. For instance in CFTs the Ward identities generated by the
107: holomorphic stress energy tensor $T(z)$ permit to express the
108: correlation functions of Virasoro descendant fields in terms of the
109: correlators of the primary fields and to develop the free field
110: approach for their calculation in the minimal models of CFT
111: \cite{DotsFat}.
112:
113:
114: In massive integrable theories the situation with descendant fields is
115: more difficult. Perturbing away from the conformal invariant point
116: \cite{Zam} one expects that the space of scaling fields is in a
117: one-to-one correspondence with that of conformal fields. This
118: statement was supported by several results on counting solutions of
119: form factor equations \cite{KaWe78,smirnovbook} in massive
120: integrable field theories which demonstrate appearance of conformal
121: algebra characters even off criticality \cite{Jim}. This phenomena
122: may indicate existence of some deformed conformal algebra action in
123: the space of form factors which would allow to extract useful
124: information on descendants from knowledge on primaries (see also
125: recent works in this direction \cite{Del}).
126:
127: In this work we study the structure of the space of form factors of
128: scaling fields in the parafermionic CFT with the central charge
129: %
130: \[
131: \label{Charge} c={2(N-1)}/({N+2})\,,\quad N=2,3,\cdots \,,\]
132: %
133: perturbed by the first thermal operator \cite{KoSw,Fat91}. In the
134: free field approach \cite{Luk95,ZF} to form factors we develop
135: further the idea of the paper \cite{Smir} on relations between
136: different form factors of the theory. Namely, we introduce actions
137: of simplest parafermionic modes in the space of form factors by
138: exploiting the deformed parafermionic symmetry \cite{ABFII} of the
139: underlying lattice model \cite{ABF}. We check consistency of our
140: construction by numerical studying of correlation functions of the
141: theory by using a combination of form factor and conformal
142: perturbation theory approaches \cite{Zam91}. To do this we also
143: compute new exact VEVs of physically important fields.
144: %\section{Space of states in Parafermionic CFT}
145:
146: \vspace{0.2cm} {\it 2. Space of states of parafermionic CFT
147: \cite{ZaFa85}}
148: %
149:
150: \par\noindent
151: Conformal field theories with the parafermionic symmetry describe
152: self-dual critical points of $Z_N$ ($N=2,3,\ldots$) generalizations
153: \cite{ZaFa85,ABF} of the Ising model, the last one corresponds to
154: the $N=2$ case. In a continuum limit order parameters $\sigma_k(x)$
155: ($k=1,\ldots,N-1$) determining long range correlations of spins have
156: anomalous dimensions
157: %
158: \[\label{SpinDim}
159: 2d_k=\frac{k(N-k)}{N(N+2)}\,.\]
160: %
161: Their $Z_N$ charge is equal to $k$, i.e., under the action of the
162: $Z_N$ symmetry they transform as
163: %
164: \[ \sigma_{k}\rightarrow\omega^{kn}\sigma_{k}, \quad n\in Z\,.
165: \label{SigmZ}
166: \]
167: %
168: $Z_N$ models allow Kramers-Wannier symmetry and the theory also
169: contains the disorder parameters $\mu_{k}$ ($k=1,\ldots,N-1$) with
170: the same dimensions $2d_k$ at the self-dual point. Operators
171: $\mu_{k}$ transform under the action of the dual $\tilde{Z}_N$
172: symmetry as
173: %
174: \[
175: \quad\mu_{k}\rightarrow \omega^{kn}\mu_{k},\quad n\in Z\,. \]
176: %
177: All other fields naturally separating into families with the fixed
178: $Z_{N}\times\widetilde{Z}_{N}$ charges $(k,l)$ behave as
179: $$
180: \phi\rightarrow\omega^{kn+ln'}\phi, \quad n,n'\in Z\,.
181: $$
182:
183:
184: The parafermions $\psi_{k}$ and $\overline{\psi}_{k}$ generalizing
185: usual fermions appear in the OPE of the order and disorder fields
186: %
187: \begin{eqnarray}
188: &&\sigma_{k}(z,\overline{z})\mu_{k}(0,0) =|z|^{-4d_{k}}z^{\Delta_{k}}%
189: [\ \psi_{k}(0)+\cdots]\,, \cr &&
190: \sigma_{k}(z,\overline{z})\mu_{k}^{+}(0,0) =|z|^{-4d_{k}}\overline
191: {z}^{\Delta_{k}}[\ \overline{\psi}_{k}(0)+\cdots]\,. \end{eqnarray}
192: %
193: These currents are holomorphic and generate the infinitely
194: dimensional symmetry due to the conservation laws
195: %
196: \[
197: \partial_{\overline{z}}\psi_{k}=0\,,\qquad
198: \partial_{z}\bar{\psi}_{k}=0\,.
199: \label{ConsLaw}
200: \]
201: %
202: We concentrate on the simplest solution of the associativity
203: condition for the operator algebra of currents
204: %
205: \[
206: \psi_{k}(z)\psi_{l}(0)=z^{\Delta_{k+l}-\Delta_{k}-\Delta_{l}}\left[
207: \psi_{k+l}+...\right]\,, \]
208: %
209: which corresponds to the conformal
210: field theory with the central charge (\ref{Charge}) and the
211: conformal dimensions
212: %
213: \[ \Delta_{k}=\frac{k(N-k)}{N}\,. \]
214: %
215: The fields $\psi_{1}\equiv\psi$
216: ($\overline{\psi}_{1}\equiv\overline{\psi}$) \ are the basic ones in the
217: pa\-ra\-fer\-mi\-onic algebra. It will be convenient for us to
218: consider as well conjugate currents $\psi^\dagger\equiv \psi_{N-1}$
219: ($\bar{\psi}^\dagger\equiv \bar{\psi}_{N-1}$).
220:
221: In the conformal models the space of states splits naturally into a
222: direct sum of subspaces with the specified
223: $Z_{N}\times\widetilde{Z}_{N}$ charge $(k,l)$
224: %
225: \[ {\{F\}}=\oplus \ {\{F\}}_{\left[ m,\overline{m}\right] } ,\quad
226: N\geq m,\overline{m}\geq1-N_{{}}\,,%
227: \]
228: %
229: where $\left[ m,\overline{m} \right] =\left[k+l,k-l\right]
230: ,m+\overline{m} \in2Z$.
231: %The mutual locality exponent of two fields
232: %from the sectors ${\{F\}}_{\left[ k,\overline{k}\right]}$ and
233: %${\{F\}}_{\left[q,\overline{q}\right]}$ equals to
234: %
235: %\[
236: %\gamma^{q,\overline{q}}_{k,\overline{k}} =({\bar{q}\bar{k}-kq
237: %})/{2N}\,.
238: %\]
239: %
240: In these notations parafermionic currents and order-disorder fields
241: belong to the following subspaces
242: %
243: \begin{eqnarray} &&\ \psi\in {\{F\}}_{\left[ 2,0\right] }\,, \quad \
244: \psi^\dagger\in {\{F\}}_{\left[ -2,0\right] }\,,\label{Spaces}\\ &&\
245: \overline{\psi}\in {\{F\}}_{\left[ 0,2\right] }\,, \quad \
246: \overline{\psi}^\dagger\in {\{F\}}_{\left[ 0,-2\right] }\,, \cr &&
247: \sigma_{k}\in {\{F\}}_{\left[ k,k\right] }\,, \ \quad \mu_{k}\in
248: {\{F\}}_{\left[ k,-k\right] }\nn \,.\end{eqnarray}
249: %
250:
251: Conformal fields are conveniently classified according to the
252: representations of the pa\-ra\-fer\-mi\-o\-nic algebra. The action
253: of the parafermionic generators $A_{\nu}$ ($A_{\nu}^\dagger$) is
254: defined by the OPE
255: %
256: \begin{eqnarray} && \psi(z)\phi_{[m,\bar{m}]}=\sum
257: z^{-\frac{m}{N}+n-1}A_{\frac{1+m}{N}-n}\ \phi_{[m,\bar{m}]}\,, \label{PFactionCFT}\\
258: \vspace{0.2cm} && \psi^\dagger(z)\phi_{[m,\bar{m}]}=\sum
259: z^{\frac{m}{N}+n-1}A^\dagger_{\frac{1-m}{N}-n}\
260: \phi_{[m,\bar{m}]}\,.\nn
261: \end{eqnarray}
262: %
263: Notice that, if $\phi_{[m,\bar{m}]}\in F_{[m,\bar{m}]}$ has the
264: conformal dimensions $(d,\bar{d})$ then the conformal dimensions of
265: fields
266: %
267: \begin{eqnarray}
268: && A_{\nu}\phi_{[m,\bar{m}]}\in \{F\}_{[m+2,\bar{m}]} \,,\\ &&
269: A^\dagger_{\nu}\phi_{[m,\bar{m}]}\in \{F\}_{[m-2,\bar{m}]} \nn\,,
270: \end{eqnarray}
271: %
272: are $(d-\nu,\bar{d})$.
273:
274: Basic fields $\mu_{k}$ $\left(\sigma_{k} \right) $ are the primaries
275: of the pa\-ra\-fer\-mi\-onic algebra. For instance, the following
276: equations hold for $n\geq 0$
277: %
278: \begin{eqnarray} && A_{\frac{1+k}{N}+n}\ \mu_{k}=A^\dagger_{\frac{1-k}{N}+n+1}\
279: \mu_{k}=0\,,\label{HighestW}\\ && \bar{A}_{\frac{1-k}{N}+n+1}\
280: \mu_{k}=\bar{A}^\dagger_{\frac{1+k}{N}+n}\ \mu_{k}=0\,.\nn
281: \end{eqnarray}
282: %
283: %
284: %They have dimensions
285: %\[
286: %\label{SpinDim} d_{k}=\frac{k(N-k)}{2N(N+2)}.\]
287: %
288: All other fields of the model are obtained by the action of the
289: currents $\psi,\bar{\psi}$ on $\mu_k \ (\sigma_k)$. Thus, the space
290: of states of the conformal field theory decomposes into a direct sum
291: of irreducible representations of the parafermionic algebra
292: \[
293: \{F\}=\oplus_{k=0}^{N-1}\ [\mu_k]_A\,,
294: \]
295: with the disorder operators $\mu_k$ playing the role of the highest
296: weight fields.
297:
298: However, from the viewpoint of Virasoro algebra, each of the spaces
299: $[\mu_k]_A$ expands into a direct sum of representations
300: with the highest weight fields $\phi_{m,\bar{m}}^{k}$. These Virasoro
301: primaries are, in general, descendants in the parafermionic algebra
302: representation. Indeed, let us denote as $(\psi^\dagger)^{l}\mu_{k}$
303: the field with the minimal conformal dimension that can be obtained
304: by $l-$times application of the parafermionic generators
305: $\psi^\dagger$ to $\mu_{k}$ (and respectively for $\bar{\psi}$)
306: %
307: \begin{eqnarray}
308: (\psi^\dagger)^{l}\mu_{k}=
309: A^\dagger_{\frac{2l-1-k}{N}}A^\dagger_{\frac{2l-3-k}{N}}\cdots
310: A^\dagger_{\frac{1-k}{N}}\mu _{k}\,. \end{eqnarray}
311: %
312: Its conformal dimensions are easily computed from (\ref{SpinDim}).
313: Then, up to a normalization, the following relations take place
314: %
315: \begin{eqnarray}
316: \label{Prim}
317: &&\phi_{k-2l,-k+2\bar{l}}^{k}=(\psi^{+})^{l}(\overline{\psi})^{\bar{l}%
318: }\mu_{k}\,,\quad l,\bar{l}=0,1,...,k\,,\\ &&
319: \phi_{k+2l,-k-2\bar{l}}^{k}=(\psi)^{l}(\overline{\psi^\dagger})^{\bar{l}}
320: \mu_{k}\,, \quad l,\bar{l}=0,1,...,N-k\,.\nn\end{eqnarray}
321: %
322: %We will generalize these relations for the massive theory.
323: For example, the physically important thermal operators
324: $\varepsilon_{k}=\phi_{0,0}^{2k}$ are among these primaries. Recall
325: that $\varepsilon_{k}$ are local with respect to all fields and have
326: the conformal dimensions
327: %
328: \[ D_{k}={k(k+1)}/{(N+2)}.\]
329:
330: %Recall that the duality transformation
331: %$\sigma_{k}\leftrightarrow\mu_{k}$ and $\varepsilon
332: %_{j}\rightarrow(-)^{j}\varepsilon_{j}$.
333:
334:
335:
336:
337: %\section{Form factors for perturbed primaries}
338: %\vspace{0.5cm}
339: %\newpage
340: \vspace{0.2cm} {\it 3. Form factors for perturbed primary fields}
341: \par\noindent
342: %
343: We consider the perturbation of the parafermionic conformal field
344: theory by the first energy operator
345: \[
346: \mathcal{A}=\mathcal{A}_{CFT}+\lambda\int d^{2}x\
347: \varepsilon_{1}(x)\,.
348: \]
349: The resulting massive theory is again integrable and $Z_N$
350: symmetric. Depending on the sign of $\lambda$ the system is in the
351: ordered or disordered phase. We fix $\lambda>0$ phase where the
352: $\tilde{Z}_N$ symmetry is destroyed and vacuum expectation values of
353: disorder operators are non-zero.
354:
355: Alternatively, this theory can be described also as a model with
356: factorized scattering of $Z_N$ charged particles \cite{KoSw}. The
357: particles $a\in \{1,\ldots, N-1\}$ in the $Z_N$ $(N=2,3,4,\ldots)$
358: symmetric models have masses
359: %
360: \[ M_a=M\frac{\sin(\pi a /N)}{\sin(\pi /N)}\,. \]
361: %
362: Note, that the antiparticle ${a}^\dagger$ is identified with the
363: particle $N-a$. The scattering matrix of the lightest particles
364: $a=1$ has the form
365: \begin{eqnarray}
366: S_{11}(\beta)=
367: \frac{\sinh(\frac{\beta}{2}+\frac{i\pi}{N})}
368: {\sinh(\frac{\beta}{2}-\frac{i\pi}{N})}\,.
369: \end{eqnarray}
370: The S matrices for higher particles are also diagonal and can be
371: extracted from $S_{11}$ according to a standard bootstrap
372: prescription \cite{KoSw}.
373: %For example, the scattering matrix between
374: %the particle $1$ and the antiparticle $1^\dagger$ will be
375: %$S_{1{1}^\dagger}(\beta)=S_{11}(i\pi -\beta)$.
376:
377:
378: We follow the algebraic approach \cite{Luk95,ZF,ABFII,VEV} to the
379: form factors originated from the analysis of lattice ABF model
380: \cite{ABF}. This lattice model falls into the same universality
381: class as the $Z_N$ Ising model. In the corner transfer matrix
382: approach its dynamical symmetry is a deformation of a parafermionic
383: symmetry. The deformed currents $\Psi(\alpha)$, as well as deformed
384: primaries $\Phi_{m}^{k}(\alpha)$ (which are primary for deformed
385: Virasoro algebra \cite{ZF}) act in the corner transfer matrix space
386: of states. Note, that there is an essential difference between the
387: symmetries in off-critical situation and in the CFT. Thus, there
388: exists only one chiral algebra, the parameter $\alpha$ is related
389: with the spectral parameter but not with the distance, etc. However,
390: from the mathematical viewpoint, the analysis of ABF models,
391: including fusions rules, $Z_N$ symmetry, the structure of
392: irreducible representations etc., is very similar to the conformal
393: situation (\ref{SigmZ})-(\ref{Spaces}). Respectively, as in CFT, the
394: problem of the calculation of physical quantities reduces to
395: studying of operators acting in the direct sum of representations of
396: the deformed parafermionic algebra.
397:
398: In particular, the form factors are matrix elements of local
399: operators in the basis of asymptotic states formed by operators
400: diagonalizing the Hamiltonian. A peculiar feature of the $Z_N$
401: models is that the Hamiltonian is diagonalized by the symmetry
402: generators $\Psi$ ($\Psi^{\dagger}$) themselves. At the scaling
403: limit the currents become particle creation operators
404: ${{\mathcal{B}}}$ (${{\mathcal{B}}}^{\dagger}$) which in the free
405: fields approach \cite{Luk95} can be described as following.
406: %
407: \begin{eqnarray} &&\mathcal{B}(\beta) =\frac{e^{-\frac{\beta}{N}P_2}}{
408: \sqrt{2\sin \frac{\pi}{N} } } \sum_{a=\pm 1}a\,e^{\frac{ \pi
409: i}{2N}(P_1+P_2+1)a} \mathcal{Z}_{a}(\beta),
410: \lb{expo1}\\
411: &&\mathcal{B}^\dagger(\beta) =-\frac{e^{\frac{\beta}{N}P_2}}{
412: \sqrt{2\sin \frac{\pi}{N} } }\quad \sum_{a=\pm 1}a\,e^{\frac{ \pi
413: i}{2N}(-P_1+P_2-1)a} \mathcal{Z}^\dagger_{a}(\beta). \nonumber
414: \end{eqnarray}
415: %
416: These linear combinations contain the exponentials of free bosons
417: $\mathcal{Z}_{\pm}(\beta)$ and $\mathcal{Z}^\dagger_{\pm}(\beta)$
418: and the "zero modes" operators $P_{1,2}$. The operators
419: $\mathcal{Z}_{a}(\beta)$ and $\mathcal{Z}_{a}^{\dagger}(\beta)$ are
420: assumed to satisfy the Wick theorem
421: %
422: \begin{eqnarray} &&
423: \dbr{\mathcal{Z}_{a_1}(\beta_1)\cdots\mathcal{Z}_{a_n}(\beta_n)
424: \mathcal{Z}_{b_1}^\dagger(\beta_1')\cdots\mathcal{Z}_{b_m}^\dagger(\beta_m')
425: }
426: \label{Wick} \\
427: %
428: &&=
429: \prod_{i<j}\dbr{\mathcal{Z}_{a_i}(\beta_i)\mathcal{Z}_{a_j}(\beta_j)}
430: \prod_{i<j}\dbr{\mathcal{Z}_{b_i}^\dagger(\beta_i')\mathcal{Z}_{b_j}^\dagger(\beta_j')}\times
431: \nn\\
432: %
433: &&\quad
434: \prod_{
435: i,j}\dbr{\mathcal{Z}_{a_j}(\beta_j)\mathcal{Z}_{b_i}^\dagger(\beta_i')}\,,
436: \nn\end{eqnarray}
437: %
438: with the following contraction rules (where
439: $\beta=\beta_1-\beta_2$):
440: %
441: \begin{eqnarray} && \dbr{\mathcal{Z}_{a}(\beta_1)\mathcal{Z}_{b}(\beta_2)}
442: =\dbr{\mathcal{Z}^\dagger_{-a}(\beta_1)\mathcal{Z}^\dagger_{-b}(\beta_2)}
443: \\
444: && \qquad= \zeta(\beta)
445: \frac{\sinh\bigl(\frac{\beta}{2}+\frac{i\pi}{2N}(a-b)\bigr)}
446: {\sinh\frac{\beta}{2}}\,, \cr &&
447: \dbr{\mathcal{Z}_{a}(\beta_1)\mathcal{Z}^\dagger_{b}(\beta_2)}
448: =\dbr{\mathcal{Z}^\dagger_{-a}(\beta_1)\mathcal{Z}_{-b}(\beta_2)}
449: \cr && \qquad =\zeta^{\dagger}(\beta)
450: \cosh\bigl(\frac{\beta}{2}-\frac{i\pi}{2 N}(a+b)\bigr)\,.\nonumber
451: \end{eqnarray}
452: %
453: The functions $\zeta(\beta)$, $\zeta^{\dagger}(\beta)$ and their
454: basic properties are given in the appendix A.
455:
456: The perturbed primary fields $\phi^{k}_{m,-m}$ corresponds to the
457: projectors to the states $\langle\langle \Phi_{-m}^k|$ and
458: $|\Phi_{m}^k\rangle\rangle$. We treat the last ones as states
459: created by the deformed operators $\Phi^{k}_{\pm m}$ from the vacuum
460: vector \cite{VEV}
461: %
462: \begin{eqnarray}
463: \label{BrKt} |\Phi_{m}^k\rangle\rangle &&=\lim_{\alpha\rightarrow
464: \infty}\Phi^k_m(\alpha)|0\rangle\rangle
465: \,,\\
466: \langle\langle \Phi_{-m}^k|&&=\lim_{\alpha\rightarrow
467: -\infty}\langle\langle 0|\Phi_{-m}^{k}(\alpha) \,.\nn\end{eqnarray}
468: %
469: Let us note that these bra and ket vectors in the corner transfer
470: matrix picture correspond to the Virasoro algebra states created by
471: left and right chiral parts of fields $\phi_{m,-m}^k$. The action of
472: the zero modes on the vectors (\ref{BrKt}) reads as
473: %
474: \begin{eqnarray} &&P_1|\Phi_{m}^k\rangle\rangle=k|\Phi_{m}^k\rangle\rangle,\quad
475: P_2|\Phi_{m}^k\rangle\rangle=-m|\Phi_{m}^k\rangle\rangle
476: \,,\end{eqnarray}
477: %
478: and we assume that the operators $P_{j}$ commute with operators
479: $\mathcal{Z}_{a}(\beta)$, $\mathcal{Z}_{a}^{\dagger}(\beta)$. The
480: normalization of vectors $|\Phi_{m}^k\rangle\rangle$ is chosen
481: according to the prescription \cite{VEV}
482: \[
483: \dbr{\Phi_{-m}^k|\Phi_{m}^k}=\sbr{\phi^{k}_{m,-m}}\,,
484: \label{VEVLat}\]
485: %
486: which would correspond to
487: the conformal normalization of scaling fields
488: %
489: \[
490: %
491: \langle \phi(x) \phi(0) \rangle = {|x|^{-4\Delta_\phi}}, \ \ \
492: x\rightarrow 0\,. \label{ConfNorm}\]
493: %
494: Then the form factors of the Virasoro algebra primary fields $
495: {\phi}^{k}_{m,-m}$, which are local with respect to the fields
496: $\mu_k$, can be compactly written as following
497: %
498: \begin{multline}
499: \lb{energy2}
500: \langle {\phi}^{k}_{m,-m}|\{\beta\},\{\beta'\}\rangle_{(n,n)}=\\
501: =
502: \dbr{\Phi_{-m}^k|\prod^n\mathcal{B}(\beta_j)
503: \prod^n\mathcal{B}^\dagger(\beta_j')|\Phi_{m}^k}\,.
504: %
505: \end{multline}
506: %
507: We implied here a shorthand notation for the state with $n'$
508: lightest particles 1 and $n$ antiparticles $1^\dagger$
509: %
510: \bea && |\{\beta\},\{\beta'\}\rangle_{(n,n')}\equiv
511: |\beta_1,\ldots,\beta_n,\beta_1',\ldots,\beta_{n'}'\rangle_{
512: 1^\dagger\cdots 1^\dagger,1\cdots 1} \,\eea
513: %
514: The symbols of ordered products of particle creation operators stand
515: for
516: %
517: \bea
518: %&&
519: \prod^n\mathcal{B}(\beta_j)=\mathcal{B}(\beta_1)\cdots
520: \mathcal{B}(\beta_n)\,.
521: % \cr
522: %\prod^n\mathcal{B}^{\dagger}(\beta_j')\ &&=
523: %\mathcal{B}^{\dagger}(\beta'_1)\
524: %\cdots\mathcal{B}^{\dagger}(\beta'_{n}) \,.
525: \eea
526: %
527: The equation (\ref{energy2}) is a natural generalization of the
528: thermal operators form factors \cite{ABFII} to the $m\neq 0$ case.
529: With the definitions (\ref{expo1})-(\ref{Wick}) it has the
530: conventional form
531: %
532: \begin{multline}
533: \lb{energy3}
534: \langle {\phi}^{k}_{m,-m}|\{\beta\},\{\beta'\}\rangle_{(n,n)}
535: =\frac{(-1)^n}{(2\ \sin\frac{\pi}{N})^n}e^{\frac{m}{N}\sum
536: (\beta_j-\beta_j')}\times\\
537: \sum_{\{a_j,b_j\}}\prod_{j}a_jb_je^{\frac{i\pi
538: }{2N}((k+1)(a_j-b_j)-m(a_j+b_j))}\times
539: \\
540: \dbr{\mathcal{Z}_{a_1}(\beta_1)
541: \cdots
542: \mathcal{Z}_{b_n}^\dagger(\beta_n')}\dbr{\Phi_{-m}^k|\Phi_{m}^k}\,.
543: %
544: \end{multline}
545: %
546:
547: The remarkable property of our theory is that form factors of
548: perturbed Virasoro primaries $ {\phi}^{k}_{m,\bar{m}}$ including
549: (\ref{energy2}) can be obtained form factors of fields $\mu_k$ by
550: the equation analogous to (\ref{Prim}). Indeed, for large $\alpha$
551: the operators $\Phi^{k}_{\pm m}(\alpha)$ from (\ref{BrKt}) behave in
552: the conformal limit, roughly speaking, as chiral parts of primaries
553: $\phi_{m,-m}^k(e^{-\alpha},e^{-\bar{\alpha}})$. Respectively, the
554: operators $\Psi(\alpha)$ in CFT limit become the parafermionic
555: currents $\psi(e^{-\alpha})$. Basing on this correspondence, we
556: introduce the analogs of rules (\ref{PFactionCFT})-(\ref{Prim}) for
557: the action of the simplest Fourier modes of deformed currents $\Psi$
558: ($\Psi^{\dagger}$) as
559: %
560: \begin{eqnarray}
561: |\Phi_{m+2}^k\rangle\rangle
562: =&&|A_{\frac{1+m}{N}}\Phi^{k}_{m}\rangle\rangle =\lim_{\alpha\to
563: \infty}
564: e^{-\frac{1+m}{N}\alpha}\Psi(\alpha)|\Phi_{m}^k\rangle\rangle\,, \cr
565: %
566: |\Phi_{m-2}^k\rangle\rangle=&&|{A}^\dagger_{\frac{{1-m}}{N}}\Phi^k_{m}\rangle\rangle
567: =\lim_{{\alpha}\to \infty}
568: e^{-\frac{1-{m}}{N}{\alpha}}\Psi^\dagger({\alpha})
569: |\Phi_{m}^k\rangle\rangle\,, \cr
570: %
571: \langle\langle \Phi_{\bar{m}+2}^k| =&&
572: %
573: \langle\langle\bar{A}_{\frac{1+\bar{m}}{N}}\Phi^k_{\bar{m}}|
574: =\lim_{{\alpha}\to -\infty} e^{\frac{1+\bar{m}}{N}{\alpha}}
575: \langle\langle\Phi_{\bar{m}}^k|\Psi({{\alpha}}) \,, \cr
576: \langle\langle\Phi_{\bar{m}-2}^k|=&&
577: %
578: \langle\langle\bar{A}^\dagger_{\frac{1-\bar{m}}{N}}\Phi^k_{\bar{m}}|
579: =\lim_{{\alpha}\to -\infty} e^{\frac{1-\bar{m}}{N}{\alpha}}
580: \langle\langle\Phi_{\bar{m}}^k|{\Psi}^\dagger({{\alpha}})\,.
581: \nonumber\end{eqnarray}
582: %
583: Since the operators $\mathcal{B}$ ($\mathcal{B}^\dagger$) are,
584: essentially, the scaling limits of the deformed pa\-ra\-fer\-mi\-ons
585: $\Psi$ ($\Psi^{\dagger}$), we put%%%it is natural to define the following relations:
586: %
587: \begin{eqnarray}
588: \label{PFOne}&&\langle
589: A_{\frac{1+m}{N}}\phi^k_{m,\bar{m}}|\{\beta\},\{\beta'\}\rangle_{(n,n')}=\\
590: &&\quad\lim_{\alpha\to \infty}
591: e^{-\frac{1+m}{N}\alpha}\dbr{\Phi_{\bar{m}}^k|
592: \prod^{n}\mathcal{B}(\beta_j)
593: \prod^{n'}\mathcal{B}^\dagger(\beta'_j)\ \mathcal{B}(\alpha) |\Phi_{m}^k}
594: \,,\nonumber
595: \end{eqnarray}
596: \begin{eqnarray}
597: &&\langle\bar{A}_{\frac{1+\bar{m}}{N}}\phi^k_{m,\bar{m}}|\{\beta\},\{\beta\}\rangle_{(n,n')}=\\
598: &&\quad\lim_{\alpha\to -\infty}
599: e^{\frac{1+\bar{m}}{N}\alpha}\dbr{\Phi_{\bar{m}}^k|\mathcal{B}(\alpha)
600: \prod^{n}\mathcal{B}(\beta_j)
601: \prod^{n'}\mathcal{B}^\dagger(\beta'_{j})|\Phi_{m}^k}
602: \,.\nonumber
603: \end{eqnarray}
604: %
605: Starting from (\ref{energy2}) we have the $Z_N$ neutrality condition
606: $2(n'-n)=m+\bar{m}+2$. For Fourier modes $A^\dagger$, we obtain
607: %
608: \begin{eqnarray} &&\langle
609: A^\dagger_{\frac{1-m}{N}}\phi^k_{m,\bar{m}}|\{\beta\},\{\beta'\}\rangle_{(n,n')}=\\
610: &&\quad \lim_{\alpha\to \infty} e^{-\frac{1-m}{N}\alpha}\dbr{
611: \Phi_{\bar{m}}^k|
612: \prod^{n}\mathcal{B}(\beta_j)
613: \prod^{n'}\mathcal{B}^\dagger(\beta'_j)\ \mathcal{B}^\dagger(\alpha)|\Phi_{m}^k}
614: \,,\nonumber
615: \end{eqnarray}
616: \begin{eqnarray} && \label{PFTwo}
617: \langle\bar{A}^\dagger_{\frac{1-\bar{m}}{N}}\phi^k_{m,\bar{m}}|\{\beta\},\{\beta\}\rangle_{(n,n')}=\\
618: && \quad\lim_{\alpha\to -\infty}
619: e^{\frac{1-\bar{m}}{N}\alpha}\dbr{\Phi_{\bar{m}}^k|
620: \mathcal{B}^\dagger(\alpha) \prod^{n}\mathcal{B}(\beta_j)
621: \prod^{n'}\mathcal{B}^\dagger(\beta'_{j})|\Phi_{m}^k}
622: \,,\nonumber \end{eqnarray} with the selection rule $\
623: 2(n'-n)=m+\bar{m}-2$.
624:
625: These actions provide maps between solutions of form factor
626: equations for fields with the different $Z_N$ charges. Note, that
627: the equation (\ref{Prim}) holds for corresponding multi-particles
628: form factors, as it can be directly verified from (\ref{energy2})
629: and (\ref{PFOne}-\ref{PFTwo}). In this way we extend free field
630: formulae for primaries $\phi^{k}_{m,\bar{m}}$.
631:
632: %\section{Conformal Perturbation Theory}
633:
634: \vspace{0.2cm} {\it 4. Short distance expansions and VEVs}
635: \par\noindent
636: As an application for new form factor formulae (\ref{energy2}),
637: (\ref{PFOne}-\ref{PFTwo}) and for checking the consistency of the
638: proposed parafermionic actions we analyze the correlation functions
639: %
640: \[ G_{+}(x)=\langle\sigma_{1}(x)\sigma_{1}^{+}(0)\rangle\,,\ \
641: G_{-}(x)=\langle\mu_{1}(x)\mu_{1}^{+}(0)\rangle\,, \]
642: %
643: computing it in the form factor approach (long distance expansion)
644: from one side and in the
645: conformal perturbation theory method (short distance expansion)
646: \[
647: G_{\pm}(x)=r^{-4d_{1}}\sum C^{A_{l}}_\pm (r)\langle
648: A_{l}(0)\rangle\,, \ \ r=|x|\,,\label{ShortA}
649: \]
650: %
651: from the other side \cite{Zam91}. The structure functions
652: $C_{\pm}^{A_{l}}(r)$ above allow expansions into perturbation series
653: %
654: \bea C_{\pm}^{A_{l}}(r)=r^{2\Delta_{A_{l}}}\left(
655: C_{\pm,(0)}^{A_{l}}+ \lambda r^{2(1-D_1)}
656: C_{\pm,(1)}^{A_{l}}+\cdots\right)\,, \eea
657: %
658: where the coefficients $C_{\pm,(1)}^{A_{l}}$ can be expressed
659: through the integrals of correlation functions in CFT.
660:
661: In the first order approximation the main contributions to
662: (\ref{ShortA}) come from the primaries
663: $I,\varepsilon_{1},\varepsilon_{2}$ and the W-algebra descendant field
664: $E_{1}=W_{-1}^{(3)}\overline{W}_{-1} ^{(3)}\varepsilon_{1}$
665: %The upper sign corresponds to $G_{_{1}%
666: %}^{\sigma}(x)$ and the lower to $G_{_{1}}^{\mu}(x).$%
667: \begin{eqnarray}
668: &&C_{\pm}^{I}=1\pm\frac{\lambda\pi
669: r^{2(1-2u)}(\gamma(u)\gamma(3u))^{1/2}\gamma
670: (4u)}{2(1-4u)^{2}\gamma^{3}(2u)}%
671: \,, \label{ShortB}\\
672: &&C_{\pm}^{\varepsilon_{1}}=\mp\frac{r^{4u}(\gamma(u)\gamma(3u))^{1/2}}
673: {2\gamma(2u)}+\frac{\lambda\pi r^{2}u^{2}\gamma(4u)}
674: {4\gamma^{4}(2u)\gamma
675: ^{4}(1-u)}%
676: \,,\cr &&C_{\pm}^{\varepsilon_{2}}=\mp r^{12u}\frac{\lambda\pi
677: r^{2(1-2u)}u^{2}\gamma
678: ^{3}(u)(\gamma(5u))^{1/2}}{12(1+2u)^{2}\gamma^{2}(2u)(\gamma(3u))^{1/2}}%
679: \,,\cr &&C_{\pm}^{E_{1}}=-r^{2+4u}
680: \frac{u(1-4u)(\gamma(u)\gamma(3u))^{1/2}}{(1+2u)(1-2u)\gamma
681: (2u)}^{{}}\,.\nn
682: \end{eqnarray}
683: Here and later we denote
684: \[\gamma(z)=\frac{\Gamma(z)}{\Gamma(1-z)}\,, \quad
685: u=\frac{1}{N+2}\,.\]
686: %$u=\frac{1}{N+2}$ as in the equation (\ref{kap}).
687: %\section{Vacuum expectation values}
688: Let us recall that the mass $M$ of the lightest particle and the
689: coupling constant $\lambda$ can be expressed via each other. The
690: exact relation between them in the conformal normalization
691: $\langle\varepsilon_{1}(x)\varepsilon_{1}(0)\rangle$
692: $=|x|^{-4D_{1}}$ is known \cite{Fat94} to have a form
693:
694: \begin{eqnarray}
695: &&(2\pi\lambda)^{2}=\kappa^{4(1-2u)}\gamma(u)\gamma(3u)\,,\\ &&\cr
696: %
697: &&\quad \kappa =M\frac{\Gamma(2/N)\Gamma(1-1/N)}{\Gamma (1/N)} \,.
698: \label{kap}
699: %&& \gamma(z)=\frac{\Gamma(z)}{\Gamma(1-z)}\,, \quad
700: %u=\frac{1}{N+2}\,.\nonumber
701: \end{eqnarray}%
702: Finally, to compare the CFT perturbation result
703: (\ref{ShortA})-(\ref{ShortB}) with the form factor predictions we
704: have to compute vacuum expectation values of operators \cite{LZ} in
705: the conformal normalization (\ref{ConfNorm}).
706: %
707: %\bea
708: %
709: %\langle \phi(x) \phi(0) \rangle = {|x|^{-4\Delta_\phi}}, \ \ \ x
710: %\rightarrow 0\,. \eea
711: %
712: For $\mu_k$ these important quantities can be derived either from
713: the reflection relations \cite{FLZZ,Fateev2} or via the deformed
714: vertex operators \cite{VEV}
715: %Both ways produce for the s-th vacuum state
716: \begin{eqnarray}
717: \label{MuVev}
718: &&\langle\mu_{k}\rangle_s= \omega^{sk} \kappa^{2d_{k}}\times\\
719: &&\ \ \ \exp\int\frac{dt}{t}\left( \frac{\sinh
720: kut\sinh(N-k)ut}{\sinh t\tanh Nut}-2d_{k}e^{-2t})\right)\,.\nonumber
721: \end{eqnarray}
722: Here the index $s = 0,...,N-1$ enumerates the vacuum states in the
723: $\tilde{Z}_N$ broken phase. (Note, that the correlation functions
724: $G_\pm$ do not depend on it.)
725: %It is interesting that for even
726: %values of $k$ the integral above can be taken, giving%
727: %\[
728: %\langle\mu_{2k}\rangle=\kappa^{2d_{2k}}\left(
729: %\frac{\gamma(\frac{1}{N+2})} {\gamma(\frac{2k+1}{N+2})}\right)
730: %^{1/2} {\displaystyle\prod\limits_{i=1}^{k}}
731: %\frac{\gamma(\frac{2i-1}{N})}{\gamma(\frac{2i-1}{N+2})}
732: %\left(\frac{N+2}{N}\right)^{\frac{k(N-2k)}{N}} \,.
733: %\]
734:
735: For the general case,
736: $\phi_{m,-m}^{k}=(\psi^\dagger)^{l}(\overline{\psi})^{l}\mu_{k}$,
737: where $m=k-2l$, the vacuum expectation values of fields
738: are elegantly expressed in terms of $\langle\mu_k \rangle_s$ as %
739: \[
740: \frac{\langle\phi_{m,-m}^{k}\rangle_s}
741: {\langle\mu_{k}\rangle_s}
742: =\left(\frac{\kappa}{Nu}\right)^{\frac{k^{2}-m^{2}}{2N}}%
743: {\displaystyle\prod\limits_{i=0}^{l-1}}\bar{\omega}^{2s}
744: \frac{(i+1)\gamma(\frac{i+1}{N})}{(k-i)\gamma(\frac{k-i}{N})}\,.%
745: \]
746: In particular, VEVs of thermal fields $\varepsilon_{k}=(\psi)^{k}%
747: (\overline{\psi})^{k}\mu_{2k}$ correspond to the $m=0$ case in the
748: equation above:
749: \begin{eqnarray}
750: \langle\varepsilon_{k}\rangle_s=&&\frac{\kappa^{2D_k}}{(Nu)^k}
751: \frac{(k!)^2}{(2k)!} \left(\frac{\gamma((2k+1)u)}
752: {\gamma(u)}\right)^{\frac{1}{2}} \times
753: \\
754: &&{\displaystyle\prod\limits_{i=1}^{k}}
755: \frac{\gamma^2(\frac{i}{N})}{\gamma(\frac{2i}{N})\gamma((2i+1)u)}\,.\nonumber%
756: \end{eqnarray}
757:
758: Using the method developed in \cite{FFLZZ} we also obtained the
759: exact result for the expectation values of the normalized descendent
760: fields $E_{k}=W_{-1}^{(3)}\overline{W}_{-1} ^{(3)}\varepsilon_{k}$
761: \begin{eqnarray}
762: \frac{\langle E_{k}\rangle_s}{\langle\varepsilon_{k}\rangle_s}
763: \,\nonumber
764: =&& \ \kappa^{2}\frac{(N+2)^{2}}{2N}%
765: \frac{\Gamma^{2}(1+\frac{k+1}{N})\Gamma^{2}(1-\frac{k}{N})}{\Gamma^{2}%
766: (1-\frac{k+1}{N})\Gamma^{2}(1+\frac{k}{N})}\times\\
767: &&\frac{\Gamma(1+\frac{2k}{N})\Gamma(1-\frac{2k+2}{N})}{\Gamma
768: (2-\frac{2k}{N})\Gamma(2+\frac{2k+2}{N})} \,.%
769: \end{eqnarray}
770:
771:
772: Collecting these results we can analyze the short and long distance
773: behaviors of correlators. Figs. 1-2 demonstrate typical results
774: obtained by numerical computations. The lower lines correspond to
775: the short distance asymptotic from the CFT perturbation theory
776: (\ref{ShortA})-(\ref{ShortB}) and the upper lines give the form
777: factor decomposition up to two particle. We see that both expansions
778: are in a perfect agreement at the intermediate distances and
779: therefore provide good numerical data for $G_\pm$ at all scales.
780: This confirms the proposed identifications of form factors and also
781: gives one of few examples of unitary theories where the correlators
782: can be effectively studied in the scheme of Ref. \cite{Zam91}.
783: %A similar analysis was also provided for other cases \cite{FP}.
784: %
785: \centerline{\epsfxsize 8.0 truecm \epsfbox{N7paper.eps}}
786: \centerline{Fig 1. Disorder correlator $G_-(r)$. N=7 Model.}
787: %
788: \centerline{ }
789:
790: \centerline{\epsfxsize 8.0 truecm \epsfbox{N7Spin.eps}}
791: \centerline{Fig 2. Order correlator $G_+(r)$. N=7 Model.}
792: %
793: \centerline{ }
794:
795: %\vspace{0.1cm}
796: %
797:
798:
799: %\section{Conclusion}
800: \vspace{0.2cm} {\it 5. Discussions}
801: \par\noindent
802: Algebraic relations among form factors of $Z_N$ models can be
803: extended for higher Fourier modes of parafermions as well. Though,
804: in general, it is not clear how to confirm the proposals, we were
805: able to provide an additional check supporting the validity of our
806: construction. Namely, we verified the form factor counterpart of the
807: quantum equations of motion, which look like:
808: %
809: \begin{eqnarray}
810: && \bar{\partial} \psi=(1-D_1)\lambda\ (\psi)\varepsilon_1\,,\cr &&
811: \partial \bar{\psi}= (1-D_1)\lambda\ (\bar{\psi})\varepsilon_1\,.
812: %
813: \label{PFff}
814: \end{eqnarray}
815: %
816: The form factors
817: of operators $(\psi)\varepsilon_1$, $(\bar{\psi})\varepsilon_1$ are
818: defined as in Eqs. (\ref{PFOne})-(\ref{PFTwo}), while to compute
819: form factors of currents in the left hand side, we expanded
820: corresponding matrix elements from the lattice theory \cite{ABFII}
821: up to first non vanishing terms in the parameter of the lattice
822: spacing. Equations (\ref{PFff}) explain the relations between form
823: factors of $\psi$ and $\varepsilon_1$ originally found and discussed
824: by F. Smirnov \cite{Smir} for the $Z_3$ model.
825:
826: Notice, that for small number of particles the form factors obtained
827: from Eqs. (\ref{energy2}), (\ref{PFOne})-(\ref{PFTwo}), (\ref{PFff})
828: agree up to normalization factors with the results for $Z_3$ case
829: \cite{Smir} (see also Ref. \cite{Cass}) and for $Z_N$ case
830: \cite{Bab}.
831:
832: %Finally we would like to note that order and disorder operator
833: %correlation functions in the $Z_3$ case were also studied in the
834: %recent work \cite{Cass}, where the relation of these functions
835: %with some observables in the (2+1)
836: %dimensional SU(3) lattice gauge theories were discussed.
837:
838: %Our form factor mappings procedure seems to disagree, unfortunately,
839: %with the so-called "asymptotic factorization equations" \cite{Del}
840: %in its naive translation to parafermionic descendants (\ref{Prim}).
841: %This fact is probably related with non-local structure of parafermionic
842: %currents.
843:
844:
845: %\newpage
846: \vspace{0.2cm} {\it Acknowledgments}
847:
848: \vspace{0.2cm}
849:
850: \par\noindent
851: We would like to thank M. Jimbo, S. Lukyanov, M. Lashkevich and F.
852: Smirnov for useful discussions. This work was supported by grant
853: INTAS 03-51-3350, by the Programm of support of leading scientific
854: schools 2004.2003.2 and by contract EUCLID HRPN-CT-2002-00325. Y.P
855: was also supported by Russian Science Foundation and by grants
856: RFBR-04-02-16027, MD-3702.2005.2. V.A.F is grateful to Institute of
857: Strongly Correlated and Complex Systems at BNL for hospitality.
858:
859: %\appendix
860:
861: %\section{Properties of two particle form factors}
862:
863: \vspace{0.2cm} {\it A. Properties of two particle form factors}
864: \par\noindent
865: Let the contour $C$ goes from infinity above the real axe, then
866: around zero and then to infinity below the real axe. Introduce the
867: notation
868: \[
869: S_2(x)=\exp \frac{1}{2}\int_C \frac{dt}{2\pi i
870: t}\frac{\sinh(x-2\pi)t}{\sinh^2(\pi t)}\log (-t)\,.
871: \]
872: Then the functions $\zeta^{(\dagger)}$ appearing in the equation
873: (\ref{expo1}) read
874: %
875: \begin{eqnarray}
876: && \zeta(\beta)=\frac{i\sinh(\frac{\beta}{2})}
877: {2\sinh(\frac{\beta}{2}+\frac{i\pi}{N})\sinh(\frac{\beta}{2}-\frac{i\pi}{N})}\times
878: \\ && \hspace{2.5cm} \frac{S_2(i\beta+2\pi
879: +\frac{2\pi}{N})S_2(-i\beta+\frac{2\pi}{N})}{S_2^2(2\pi
880: +\frac{2\pi}{N})}\,, \cr
881: &&\zeta^\dagger(\beta)=\frac{1}{\cosh\frac{\beta}{2} }
882: \frac{S_2^2(2\pi +\frac{2\pi}{N})} {S_2(i\beta+3\pi
883: +\frac{2\pi}{N})S_2(-i\beta+\pi+\frac{2\pi}{N})}\,. \nonumber
884: \end{eqnarray}
885: %
886: These functions satisfy the following important for us properties
887: %
888: \begin{eqnarray} && \frac{\zeta(\beta)}{\zeta(-\beta)}=S(\beta)\,, \quad
889: \frac{\zeta^\dagger(\beta)}{\zeta^\dagger(-\beta)}=S^\dagger(\beta)\,,\\
890: %\eea
891: %
892: %and %
893: %\bea
894: && \zeta(\beta)=\zeta(2\pi i -\beta) \,, \quad
895: \zeta^\dagger(\beta)=\zeta^\dagger(2\pi i -\beta) \,,\cr
896: %\eea
897: %
898: %\bea
899: && \zeta(\beta)\zeta^\dagger(\beta+i\pi)=\left({ i
900: \sinh\left(\frac{\beta}{2}-\frac{i\pi }{N}\right)}\right)^{-1}\,,\cr
901: %\eea
902: %
903: %Using the explicit form for that functions we can easily find that
904: %these functions satisfy the equation
905: %
906: %
907: %\begin{multline}
908: &&\prod_{j=1}^{N}\zeta\left( \beta+\frac{2j-1}{N}i\pi \right)=\cr
909: &&\qquad \frac{S_2^{-2N}(2\pi
910: +\frac{2\pi}{N})}{4\sinh\left(\frac{\beta}{2}+\frac{i\pi}{2N}\right)
911: \sinh\left(\frac{\beta}{2}-\frac{i\pi}{2N}\right)} \,.\nonumber
912: \end{eqnarray}
913: %\end{multline}
914: %
915: %\section{Explicit formulae for disorder field}
916: %As an example we write down the explicit formulae for first form
917: %factors
918: %of disorder operator %
919: %\begin{eqnarray}
920: %&&\langle\mu_k|0\rangle=\langle\mu_k\rangle\,, \cr &&
921: %\langle\mu_k|\beta_1,\beta_1'\rangle_{ 1^\dagger
922: %1}=-\zeta^\dagger(\beta_{1}-\beta_2')\langle\mu_k\rangle \{k\}
923: %e^{\frac{N-2k}{2N}(\beta_1-\beta'_1)}
924: % \,, \cr
925: %&&\langle\mu_k|\beta_1,\beta_2,\beta_1',\beta'_2\rangle_{ 1^\dagger
926: %1^\dagger 11}=\zeta(\beta_{12}) \zeta(\beta_{12}')\prod_{ij}^{2}
927: %\zeta^\dagger(\beta_{i}-\beta_j')\cr && \qquad
928: %\times\frac{1}{4}\langle\mu_k\rangle \{k\}^2
929: %e^{\frac{k-N}{N}(\beta_1+\beta_2)-
930: %\frac{k}{N}(\beta_1'+\beta_2')}\cr && \qquad\quad \times
931: %\Bigl(\nu_1\tau_1+\frac{\{2\} \{k+1\}}{\{1\}\{k\}}\nu_2+\frac{\{2\}
932: %\{k-1\}}{\{1\}\{k\}} \tau_2\Bigr)\,,
933: %\end{eqnarray} %
934: %where $\nu_j$ and $\tau_j$ are $j$-th symmetric polynomials of
935: %variables $e^{\beta_i}$ and $e^{\beta_i'}$ respectively. Here we
936: %used the notation $\{a\}:=\sin(\frac{\pi a}{N})$.
937:
938:
939: \begin{thebibliography}{100}
940: \bibitem{BPZ}{A.~A.~Belavin, et al.}
941: %A.~M.~Polyakov, A.~B.~Zamolodchikov}.
942: {\em Nucl.\ Phys.} {\bf B241}, 333 (1984)
943:
944: \bibitem{DotsFat}{V.S. Dotsenko, V.A. Fateev.
945: {\em Nucl. Phys.} {\bf B240}[{\bf FS12}], 312(1984); {\em Nucl.
946: Phys.} {\bf B251}[{\bf FS13}], 691 (1985)}
947:
948:
949: \bibitem{ZaFa85}
950: A.~B.~Zamolodchikov, V.~A.~Fateev. {\em Sov. Phys. JETP}, {\bf
951: 62}(2), 215 (1985)
952:
953:
954: \bibitem{Zam}
955: {A.~B.~Zamolodchikov. {\em Adv. Stud. in Pure Math.} {\bf 19}, 641
956: (1989) }
957:
958: \bibitem{KaWe78}
959: M.~Karowski, P.~Weisz. {\em Nucl.~Phys.} {\bf B139}, 455 (1978)
960:
961: \bibitem{smirnovbook}
962: F.~A.~Smirnov. Form factors in completely integrable models of
963: quantum field theory. Singapore: World Scientific (1992)
964:
965: \bibitem{Jim}
966: J.~L.~Cardy, G.~Mussardo. {\em Nucl. Phys.} {\bf B340}, 387, (1990);
967: A.~Koubek. {\em Nucl.Phys.} {\bf B428}, 655 (1994); M.~Jimbo, et
968: al. [math-ph/0303059];
969:
970: \bibitem{Del} G.~Delfino, G.~Niccoli. {\em Nucl. Phys.} {\bf B707}, 381
971: (2005); [hep-th/0501173].
972:
973: %\bibitem{SmirKir}
974:
975: \bibitem{KoSw}
976: R.~Koberle, J.~A.~Swieca. {\em Phys. Lett.} {\bf B86}, 209 (1979)
977:
978: \bibitem{Fat91}
979: V.~A.~Fateev. {} {} {\em Int. J. Mod. Phys.}, {\bf A6}, 2109 (1991)
980:
981:
982: %\bibitem{FateevIII}
983: %{ C.~Ahn, et al. {\em Nucl.Phys.} {\bf B565} 611 (2000). }
984:
985: \bibitem{Luk95}
986: S.~Lukyanov. {\em Comm. Math. Phys.} {\bf 167}, 183 (1995); {\em
987: Mod. Phys. Lett.} {\bf A12}, 2543 (1997); {\em Phys.~Lett.}, {\bf
988: B408}, 192, 1997.
989:
990: %\bibitem{DFJMN}
991: %B.~Davies, et al. {} {} {\em Comm. Math. Phys.}, {\bf 151},89 1993.
992:
993: \bibitem{ZF}
994: S.~Lukyanov, Y.~Pugai. {\em J. Exp. Theor. Phys.} {\bf 82}, 1021
995: (1996); {\em Nucl. Phys.} {\bf B473}, 631, 1996.%hep-th/9602074.
996:
997:
998: \bibitem{Smir}
999: F.~A.~Smirnov. {\em Comm.~Math.~Phys.}, {\bf 132}, 415, (1990);
1000: A.~N.~Kirillov, F.~A.~Smirnov. preprint ITF-88-73R, Kiev (1988)
1001:
1002:
1003: \bibitem{ABFII}
1004: M.~Jimbo, et al. {\em J. Statist. Phys.} {\bf 102}, 883 (2001)
1005:
1006: \bibitem{ABF}
1007: G.~E.~Andrews, et al.
1008: %R.~J.~Baxter, P.~J.~Forrester.
1009: {\em J. Stat. Phys.}
1010: {\bf 35}, 193 (1984)
1011:
1012: \bibitem{Zam91}
1013: Al.~B.~Zamolodchikov. {\em Nucl.~Phys.}, {\bf B348}, 619 (1991)
1014:
1015:
1016:
1017: %\bibitem{KadFra}
1018: %E.~Fradkin, L.~P.~Kadanoff. {\em Nucl. Phys.} {\bf B170 [FS1]}, 1
1019: %(1980)
1020:
1021: \bibitem{VEV}
1022: Y.~Pugai. {\em JETP Lett.} {\bf 79}, 457 (2004)
1023:
1024: \bibitem{Fat94}{V.~A.~Fateev. {\em Phys. Lett.} {\bf B324}, 45 (1994)}
1025:
1026:
1027: \bibitem{LZ}
1028: {S.~Lukyanov, A.~Zamolodchikov. {\em Nucl. Phys.} {\bf B493}, 571
1029: (1997)};
1030: {V.~A.~Fateev, et al. {\it Nucl. Phys.} {\bf B516}, 652
1031: (1998)}
1032:
1033: \bibitem{FLZZ}
1034: V.~A.~Fateev. [hep-th/0103014]
1035:
1036: \bibitem{Fateev2}
1037: {V.~A.~Fateev. {\em Mod. Phys. Lett.} {\bf A15}, 259 (2000)}
1038:
1039:
1040: \bibitem{FFLZZ}
1041: {V.~Fateev et al. {\it Nucl.Phys.} {\bf B540}, 587 (1999)}
1042:
1043:
1044: %\bibitem{FP}
1045: %{V.A. Fateev et al. To appear.}
1046:
1047:
1048: \bibitem{Cass} M.~Casselle et al. [hep-th/051168]
1049:
1050: \bibitem{Bab} H.~Babujian et al. [hep-th/0510062]
1051:
1052:
1053:
1054:
1055: %\bibitem{AlZam}
1056: %{ Zamolodchikov Al.~B. Int.J.Mod.Phys., {\bf A10}, 1125 (1995) }
1057:
1058:
1059:
1060:
1061: %%%%%%
1062: %An models
1063: %%%%%%%
1064:
1065:
1066:
1067:
1068:
1069:
1070:
1071:
1072:
1073:
1074:
1075:
1076: \end{thebibliography}
1077:
1078: \end{document}
1079: