1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %
3: %
4: % J.Phys.Soc.Jpn. Vol.74No.3
5: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
6: \documentclass[12pt,a4paper,twoside]{article}
7: %\documentclass[a4paper,twoside]{article}
8: \usepackage{amsmath,amsfonts,amsthm,graphicx}
9: %\usepackage{amsmath,amsfonts,amsthm,graphicx,showkeys}
10: %
11: \numberwithin{equation}{section}
12: \newtheorem{theorem}{Theorem}[section]
13: \newtheorem{conjecture}[theorem]{Conjecture}
14: \def\beqa{\begin{eqnarray}}
15: \def\enqa{\end{eqnarray}}
16: \def\beq{\begin{equation}}
17: \def\enq{\end{equation}}
18: \def\nonum{\nonumber}
19: %%%%%%%%%%%%%%%%%%%%%%%%%
20: \begin{document}
21: \title{
22: Nonlinear Integral Equations for Thermodynamics
23: of the $U_{q}(\widehat{sl(r+1)})$
24: Perk-Schultz Model
25: }
26: \author{Zengo Tsuboi
27: \footnote{E-mail address: tsuboi@issp.u-tokyo.ac.jp}
28: and Minoru Takahashi
29: \footnote{E-mail address:
30: mtaka@issp.u-tokyo.ac.jp}
31: \\
32: {\it Institute for Solid State Physics, University of Tokyo,} \\
33: {\it Kashiwanoha 5-1-5, Kashiwa, Chiba, 277-8581, Japan}
34: }
35: \date{}
36: \maketitle
37: \begin{abstract}
38: We propose a system of nonlinear integral equations (NLIE)
39: which describes the thermodynamics of the
40: $U_{q}(\widehat{sl(r+1)})$ Perk-Schultz model.
41: These NLIE correspond to a trigonometric analogue of
42: our previous result, and contain only $r$ unknown functions.
43: In particular, they reduce to Takahashi's NLIE for
44: the $XXZ$ spin chain if $r=1$.
45: We also calculate the high temperature expansion of
46: the free energy.
47: In particular for $r=1$ case, we have succeeded to derive
48: the coefficients of order $O((\frac{J}{T})^{99})$.
49: \end{abstract}
50: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
51: Short title: Nonlinear integral equation\\
52: {\it MSC:} 82B23; 45G15; 82B20; 17B80 \\
53: {\it PACS2001:} 02.30.Rz; 02.30.Ik; 05.50.+q; 05.70.-a \\
54: {\it Key words:}
55: nonlinear integral equation;
56: Perk-Schultz model;
57: quantum transfer matrix;
58: thermodynamic Bethe ansatz;
59: $T$-system \\
60: {\bf to appear in Journal of the Physical Society of Japan, Vol.74 No.3 (2005)}
61: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
62: \section{Introduction}
63: Traditionally, thermodynamic Bethe ansatz (TBA)
64: equations have been used to analyze thermodynamics of
65: various kind of
66: solvable lattice models \cite{Ta99}.
67: However TBA equations contain, in general, an infinite number
68: of unknown functions, and are difficult to deal with.
69: To overcome this difficulty, alternative
70: nonlinear integral equations (NLIE),
71: which contain only a finite number of unknown functions,
72: have been considered. For example, for $XXZ$-model
73: (or $XYZ$-model),
74: Kl\"umper \cite{K93,K92},
75: Destri and de Vega \cite{DD92}
76: derived a NLIE, which contains
77: two unknown functions.
78:
79: Recently, the second author \cite{Ta01} derived another type of NLIE
80: for the $XXZ$-model. The number of the unknown functions is
81: only one, which corresponds to the rank of the
82: underlying algebra $U_{q}(\widehat{sl(2)})$.
83: Moreover the same NLIE was rederived \cite{TSK01} from
84: the $T$-system (a
85: system of functional relations among transfer matrices)
86: \cite{KR87,KNS94} of the quantum transfer matrix (QTM) \cite{S85};
87: from a fugacity expansion formula \cite{KW02}.
88: We find \cite{ShT02} that this NLIE \cite{Ta01}
89: is very useful to calculate the high temperature expansion
90: of the free energy.
91: In view of this situation, the first author
92: derived NLIE for the $osp(1|2s)$ model \cite{T02},
93: the $sl(r+1)$ Uimin-Sutherland model \cite{T03} and the
94: spin $\frac{s}{2}$ Heisenberg model \cite{T04}.
95: The numbers of the unknown functions of
96: these NLIE are $s,r,1$, respectively,
97: which correspond to the rank of the underlying algebras.
98: In particular, the NLIE in refs. \cite{T03} and \cite{T04} contain
99: the rational limit of
100: Takahashi's NLIE \cite{Ta01} as special cases $r=1$ or $s=1$
101: respectively.
102: The high temperature expansion of the free energy
103: from our NLIE \cite{T03} was applied \cite{BGOSTF03}-\cite{BGOF04}
104: to spin ladder models, and good agreements were
105: seen between experimental data and theoretical results.
106:
107: The Perk-Schultz model \cite{PS81,Sc83}
108: is a graded vertex model
109: associated with the superalgebra $U_{q}(\widehat{sl(r+1|s+1)})$.
110: In this paper, we consider a special case ($s=-1$) of it:
111: the $U_{q}(\widehat{sl(r+1)})$ Perk-Schultz model \cite{algebra}.
112: This model is a natural multicomponent generalization of the $XXZ$ model
113: which is very important in physics. In fact the hamiltonian
114: of the Perk-Schultz model (eqs. (\ref{ham0}) and (\ref{hamch})) reduces to the one
115: for the $XXZ$ model eq. (\ref{hamXXZ}) if $r=1$.
116: Moreover the rational limit of the Perk-Schultz model is
117: the Uimin-Sutherland model \cite{U70,S75}, which we considered
118: previously \cite{T03}.
119:
120: The purpose of this paper is to derive NLIE for
121: $U_{q}(\widehat{sl(r+1)})$ Perk-Schultz model \cite{PS81,Sc83}.
122: This NLIE contains only $r$ unknown functions,
123: and corresponds to a
124: trigonometric analogue of the NLIE in ref. \cite{T03}.
125: Moreover it reduces to Takahashi's NLIE \cite{Ta01} for
126: $XXZ$-model when $r=1$.
127:
128: The shape of
129: TBA equations varies with $q$
130: \cite{TS72,KSS98}
131: when $q$ is a root of unity since
132: the corresponding
133: $T$-system (or $Y$-system),
134: from which the corresponding TBA equations are derived,
135: truncates with respect to a variable $m$ (cf. eq. (\ref{T-system})).
136: In contrast, our new NLIE eq. (\ref{nlie4})
137: will be also valid as it stand even when $q$ is a root of unity
138: since we use only the first part of the $T$-system
139: (eq. (\ref{T-system}) for $m=1$).
140:
141: In 2, we introduce the $U_{q}(\widehat{sl(r+1)})$
142: Perk-Schultz model \cite{PS81,Sc83} and define the
143: QTM and the $T$-system.
144: In 3, we derive our new NLIE eq. (\ref{nlie4})
145: taking note on the periodicity of the
146: QTM. We will write only the outline of the derivation
147: since it is similar to the ones in refs. \cite{TSK01} and \cite{T03}.
148: In 4, we will calculate the high temperature
149: expansion of the free energy from our new NLIE.
150: In particular, $r=1$ case is a detailed explanation of
151: a calculation in the letter \cite{ShT02}, which has been extended to
152: the $XXZ$-model case now.
153: It will be difficult to derive the same results
154: from the traditional TBA equations.
155: Section 5 is devoted to concluding remarks.
156: %%%%%%%%%%%%%%%%%%%%%%%%
157: \section{$T$-system and QTM Method}
158: We will introduce
159: the quantum transfer matrix (QTM) method
160: \cite{S85}, \cite{SI87}-\cite{SAW90}, \cite{K92,K93}
161: and the $T$-system \cite{KNS94} for $U_{q}(\widehat{sl(r+1)})$
162: Perk-Schultz model.
163: The QTM analyses of the Perk-Schultz model
164: was done in ref. \cite{KWZ97}
165: (see also, ref. \cite{JKS98,FK99}).
166:
167: The $R$-matrix of the model
168: is given as
169: \begin{eqnarray}
170: R(v)=
171: \sum_{a_{1}=1}^{r+1}\sum_{a_{2}=1}^{r+1}
172: \sum_{b_{1}=1}^{r+1}\sum_{b_{2}=1}^{r+1}
173: R^{a_{1},b_{1}}_{a_{2},b_{2}}(v)
174: E^{a_{1},a_{2}}\otimes E^{b_{1},b_{2}},
175: \end{eqnarray}
176: where $E^{a,b}$ is a $r+1$ by $r+1$ matrix
177: whose $(i,j)$ element is given as
178: $(E^{a,b})_{i,j}=\delta_{ai}\delta_{bj}$;
179: $R^{a_{1},b_{1}}_{a_{2},b_{2}}(v)$ is defined as
180: %
181: \begin{eqnarray}
182: && R^{a,a}_{a,a}(v)=[v+1]_{q}, \\
183: && R^{a,b}_{a,b}(v)=[v]_{q} \quad (a \ne b), \\
184: && R^{b,a}_{a,b}(v)=q^{\mathrm{sign}(a-b)v}
185: \quad (a \ne b), \label{R-mat}
186: \end{eqnarray}
187: where $v \in \mathbb{C}$;
188: $a,b \in \{1,2,\dots,r+1\} $; $[v]_{q}=(q^{v}-q^{-v})/(q-q^{-1})$;
189: $q=e^{\eta}$.
190: Let $L$ be a positive integer (the number of lattice sites).
191: The row-to-row transfer matrix on ${\mathbb C}_{r+1}^{\otimes L}$
192: is defined as \cite{r-index}
193: %
194: %
195: \begin{eqnarray}
196: t(v)={\mathrm tr}_{0}(R_{0L}(v)
197: \cdots R_{02}(v)R_{01}(v)).
198: \label{rtr}
199: \end{eqnarray}
200: The hamiltonian $H=H_{0}+H_{ch}$
201: of the model has two parts. The first part $H_{0}$
202: is given by
203: \begin{eqnarray}
204: && \hspace{-20pt}
205: H_{0}=\frac{J\sinh \eta}{\eta}\frac{d}{dv}\log t(v) |_{v=0} %\nonum\\&& \hspace{-20pt}
206: = J\sum_{j=1}^{L}\biggl\{
207: \cosh \eta \sum_{a=1}^{r+1} E^{a,a}_{j}E^{a,a}_{j+1} +
208: \nonumber \\ && \hspace{20pt}
209: \sum_{a=1}^{r+1}
210: \sum_{
211: {\scriptsize \begin{array}{c}
212: b=1 \\
213: a\ne b
214: \end{array}}
215: }^{r+1}
216: \left( {\rm sign}(a-b) \sinh \eta
217: E^{a,a}_{j}E^{b,b}_{j+1} +
218: E^{b,a}_{j}E^{a,b}_{j+1}
219: \right)
220: \biggl\}, \label{ham0}
221: \end{eqnarray}
222: where we adopt periodic boundary condition
223: $E^{a,b}_{L+1}=E^{a,b}_{1}$.
224: The second part is the chemical potential term,
225: \begin{eqnarray}
226: H_{ch}=-\sum_{j=1}^{L}\sum_{a=1}^{r+1}\mu_{a}E^{a,a}_{j}.\label{hamch}
227: \end{eqnarray}
228: %%%%%%%%%%%%%%%%%%%%%%%%%%
229: Let $\widetilde{R}(v)$ be \lq 90 degree rotation\rq of
230: $R(v)$:
231: \begin{eqnarray}
232: \widetilde{R}(v)&=&
233: \sum_{a_{1}=1}^{r+1}\sum_{a_{2}=1}^{r+1}
234: \sum_{b_{1}=1}^{r+1}\sum_{b_{2}=1}^{r+1}
235: \widetilde{R}^{a_{1},b_{1}}_{a_{2},b_{2}}(v)
236: E^{a_{1},a_{2}}\otimes E^{b_{1},b_{2}} \nonumber \\
237: &=&
238: \sum_{a_{1}=1}^{r+1}\sum_{a_{2}=1}^{r+1}
239: \sum_{b_{1}=1}^{r+1}\sum_{b_{2}=1}^{r+1}
240: R^{a_{1},b_{1}}_{a_{2},b_{2}}(v)
241: ^{t} \! E^{b_{1},b_{2}}\otimes E^{a_{1},a_{2}} \nonumber \\
242: &=&
243: \sum_{a_{1}=1}^{r+1}\sum_{a_{2}=1}^{r+1}
244: \sum_{b_{1}=1}^{r+1}\sum_{b_{2}=1}^{r+1}
245: R^{a_{1},b_{1}}_{a_{2},b_{2}}(v)
246: E^{b_{2},b_{1}}\otimes E^{a_{1},a_{2}}.
247: \end{eqnarray}
248: Namely we have $\widetilde{R}^{a_{1},b_{1}}_{a_{2},b_{2}}(v)=
249: R^{b_{1},a_{2}}_{b_{2},a_{1}}(v)$.
250: %
251: Let $N$ be a positive even integer (the Trotter number).
252: We define the QTM on
253: ${\mathbb C}_{r+1}^{\otimes N}$ as
254: \begin{eqnarray}
255: && t_{\mathrm{QTM}}(v)=
256: {\mathrm tr}_{j} e^{- \frac{H_{j}}{T}}
257: R_{N,j}(u_{N}+iv)\widetilde{R}_{N-1,j}(u_{N}-iv)
258: \cdots \nonumber \\
259: && \hspace{45pt}
260: R_{4,j}(u_{N}+iv)\widetilde{R}_{3,j}(u_{N}-iv)
261: R_{2,j}(u_{N}+iv)\widetilde{R}_{1,j}(u_{N}-iv),
262: \label{QTM}
263: \end{eqnarray}
264: where $j \in \{1,2,\dots, L \}$;
265: $u_{N}=-\frac{J\sinh \eta}{\eta NT}$
266: ($T$: temperature).
267: The matrix elements of the QTM are given as
268: \begin{eqnarray}
269: && \hspace{-30pt} t_{\mathrm{QTM}}(v)=\sum_{\{\alpha_{k},\beta_{k}\}}
270: t_{\mathrm{QTM}}(v)
271: ^{\{\beta_{1},\dots, \beta_{N} \}}
272: _{\{\alpha_{1},\dots,\alpha_{N} \}}
273: E^{\beta_{1}\alpha_{1}}_{1}
274: E^{\beta_{2}\alpha_{2}}_{2}
275: \cdots
276: E^{\beta_{N}\alpha_{N}}_{N}, \\
277: && \hspace{-46pt}
278: t_{\mathrm{QTM}}(v)^{\{\beta_{1},\dots, \beta_{N} \}}
279: _{\{\alpha_{1},\dots,\alpha_{N} \}}=
280: \sum_{\{\nu_{k}\}}e^{\frac{\mu_{\nu_{1}}}{T}}
281: \prod_{k=1}^{\frac{N}{2}}
282: R^{\beta_{2k},\nu_{2k+1}}_{\alpha_{2k},\nu_{2k}}(u_{N}+iv)
283: \widetilde{R}^{\beta_{2k-1},\nu_{2k}}_{\alpha_{2k-1},\nu_{2k-1}}(u_{N}-iv),
284: \label{QTM2}
285: \end{eqnarray}
286: where $\nu_{N+1}=\nu_{1}$ and $\nu_{k},\alpha_{k},\beta_{k}
287: \in \{1,2,\dots, r+1 \}$.
288: We can express \cite{S85} the free energy per site
289: in terms of only the largest eigenvalue $\Lambda_{1}$ of
290: the QTM eq. (\ref{QTM}) at $v=0$:
291: \begin{eqnarray}
292: f=
293: -T\lim_{N\to \infty}\log \Lambda_{1},
294: \end{eqnarray}
295: where the Boltzmann constant is set to $1$.
296:
297: The eigenvalue formula of the row-to-row transfer matrix eq.
298: (\ref{rtr}) was derived \cite{BVV82,Sc83} by the algebraic Bethe ansatz.
299: On the other hand, the eigenvalue formula $T_{1}^{(1)}(v)$
300: of the QTM eq. (\ref{QTM}) is conjectured to be given by
301: replacing the vacuum part (the eigenvalue on the
302: pseudo-vacuum $|0>$)
303: of the eigenvalue formula of the row-to-row transfer matrix
304: with that of the QTM \cite{aba}.
305: We adapt the following pseudo-vacuum:
306: $|0>:=v_{r+1} \otimes v_{1} \otimes \cdots \otimes v_{r+1} \otimes v_{1}
307: \in {\mathbb C}_{r+1}^{\otimes N}$, where
308: $v_{1}:=^{t}\!(1,0,\dots, 0) \in {\mathbb C}_{r+1}$ and
309: $v_{r+1}:=^{t}\!(0,\dots, 0,1) \in {\mathbb C}_{r+1}$.
310: Indeed $|0>$ is an eigenvector for $t_{\rm QTM}(v)$:
311: \begin{eqnarray}
312: t_{\rm QTM}(v)|0>=\sum_{a=1}^{r+1} \psi_{a}(v)|0>,
313: \end{eqnarray}
314: where the eigenvalue consists of the functions:
315: \begin{eqnarray}
316: && \hspace{-30pt}
317: \psi_{a}(v)=
318: e^{\frac{\mu_{a}}{T}}
319: \phi_{+}(v+i \delta_{a,r+1})\phi_{-}(v-i\delta_{a,1})
320: \quad
321: \mbox{for} \quad a \in \{1,2,\dots,r+1\},
322: \label{vac-QTM} \\
323: && \hspace{60pt} \phi_{\pm}(v)=\left(
324: \frac{\sin \eta (v\pm iu_{N})}{\sinh \eta }\right)^{\frac{N}{2}}.
325: \nonumber
326: \end{eqnarray}
327: %
328: Then the general eigenvalue formula $T_{1}^{(1)}(v)$
329: of the QTM eq. (\ref{QTM}) will be (cf. refs. \cite{KWZ97}-\cite{FK99})
330: \begin{eqnarray}
331: T_{1}^{(1)}(v)=\sum_{d=1}^{r+1}z(d;v).
332: \label{dvf-qtm}
333: \end{eqnarray}
334: Here the functions $\{z(d;v)\}$ are defined as
335: \begin{eqnarray}
336: && z(a;v)=\psi_{a}(v)
337: \frac{Q_{a-1}(v-\frac{i}{2}(a+1))Q_{a}(v-\frac{i}{2}(a-2))}
338: {Q_{a-1}(v-\frac{i}{2}(a-1))Q_{a}(v-\frac{i}{2}a)}
339: \nonumber \\
340: && \hspace{150pt}
341: \mbox{for} \quad a \in \{1,2,\dots,r+1\},
342: \end{eqnarray}
343: %
344: where $Q_{a}(v)=\prod_{k=1}^{M_{a}}\sin \eta(v-v_{k}^{(a)})$;
345: $M_{a}\in {\mathbb Z}_{\ge 0}$; $Q_{0}(v)=Q_{r+1}(v)=1$.
346: %
347: $\{v^{(a)}_{k}\}$ is a solution of the Bethe ansatz equation
348: (BAE)
349: \begin{eqnarray}
350: && \hspace{-20pt}
351: \frac{\psi_{a}(v^{(a)}_{k}+\frac{i}{2}a)}
352: {\psi_{a+1}(v^{(a)}_{k}+\frac{i}{2}a)}=
353: -
354: \frac{Q_{a-1}(v^{(a)}_{k}+\frac{i}{2})Q_{a}(v^{(a)}_{k}-i)
355: Q_{a+1}(v^{(a)}_{k}+\frac{i}{2})}
356: {Q_{a-1}(v^{(a)}_{k}-\frac{i}{2})Q_{a}(v^{(a)}_{k}+i)
357: Q_{a+1}(v^{(a)}_{k}-\frac{i}{2})}
358: \label{BAE} \\
359: && \hspace{40pt} \mbox{for} \quad k\in \{1,2, \dots, M_{a}\} \quad
360: \mbox{and} \quad a\in \{1,2, \dots, r\}.\nonumber
361: \end{eqnarray}
362: %%%
363: Some remarks on eq. (\ref{dvf-qtm}) are in order.
364: For $r=1$ case, eq. (\ref{dvf-qtm}) can be
365: proved \cite{alge-ba} by algebraic Bethe ansatz.
366: As for $r>1$ case, we have checked that
367: eq. (\ref{dvf-qtm}) agrees with the
368: numerical diagonalization of eq. (\ref{QTM}) for $r=2$ and
369: small Trotter numbers $N$.
370: In ref. \cite{KWZ97} the algebraic Bethe ansatz was executed for one particle
371: state, from which an eigenvalue formula of the QTM for the Perk-Schultz model was
372: conjectured, although a different pseudo-vacuum was adopted.
373: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
374:
375: As an auxiliary function, we define
376: (cf. Bazhanov-Reshetikhin formula in ref. \cite{BR90}):
377: \begin{eqnarray}
378: T_{m}^{(a)}(v)=\sum_{\{d_{j,k}\}} \prod_{j=1}^{a}\prod_{k=1}^{m}
379: z(d_{j,k};v-\frac{i}{2}(a-m-2j+2k)),
380: \label{DVF}
381: \end{eqnarray}
382: %
383: where the summation is taken over
384: $d_{j,k}\in
385: \{1,2,\dots,r+1\}$
386: such that $d_{j,k} \prec d_{j+1,k}$ and $d_{j,k} \preceq d_{j,k+1}$
387: ($ 1 \prec 2 \prec \cdots \prec r+1$);
388: $m \in \mathbb{Z}_{\ge 1}$;
389: $a \in \{1,2,\dots,r \}$.
390: This function contains
391: $T_{1}^{(1)}(v)$ eq. (\ref{QTM}) as a special case
392: $(a,m)=(1,1)$.
393: It is related to a fusion \cite{KRS81} hierarchy of the QTM.
394: %
395: We can show that the poles of $T^{(a)}_{m}(v)$
396: are spurious under the BAE eq. (\ref{BAE}).
397: %%%%%%%%%%%
398: For $a \in \{1,2,\dots,r\}$ and $m \in {\mathbb Z}_{\ge 1}$, we
399: shall normalize eq. (\ref{DVF}) as
400: $ \widetilde{T}^{(a)}_{m}(v)=
401: T^{(a)}_{m}(v)/\widetilde{{\mathcal N}}^{(a)}_{m}(v)$,
402: where
403: \begin{eqnarray}
404: \hspace{-30pt} && \widetilde{{\mathcal N}}^{(a)}_{m}(v)=
405: \frac{\phi_{-}(v-\frac{a+m}{2}i)\phi_{+}(v+\frac{a+m}{2}i)}{
406: \phi_{-}(v-\frac{a-m}{2}i)\phi_{+}(v+\frac{a-m}{2}i)}
407: \nonumber \\
408: \hspace{-30pt} && \hspace{20pt} \times
409: \prod_{j=1}^{a}\prod_{k=1}^{m}
410: \phi_{-}(v-\frac{a-m-2j+2k}{2}i)\phi_{+}(v-\frac{a-m-2j+2k}{2}i).
411: \label{normal}
412: \end{eqnarray}
413: %
414: One can show that
415: $\widetilde{T}^{(a)}_{m}(v)$ satisfies
416: so called $T$-system \cite{KNS94}
417: \begin{eqnarray}
418: && \hspace{-20pt}
419: \widetilde{T}^{(a)}_{m}(v+\frac{i}{2})
420: \widetilde{T}^{(a)}_{m}(v-\frac{i}{2})
421: =\widetilde{T}^{(a)}_{m+1}(v)\widetilde{T}^{(a)}_{m-1}(v)
422: +\widetilde{T}^{(a-1)}_{m}(v)\widetilde{T}^{(a+1)}_{m}(v),
423: \label{T-system}
424: \\ && \hspace{70pt}
425: {\rm for} \quad a \in \{1,2,\dots,r\}
426: \quad {\rm and} \quad m \in {\mathbb Z}_{\ge 1},
427: \nonumber
428: \end{eqnarray}
429: %
430: where
431: \begin{eqnarray}
432: && \widetilde{T}^{(a)}_{0}(v)=1
433: \quad {\rm for} \quad a \in {\mathbb Z}_{\ge 1},\nonumber \\
434: && \widetilde{T}^{(0)}_{m}(v)=
435: \frac{\phi_{-}(v+\frac{m}{2}i)\phi_{+}(v-\frac{m}{2}i)}
436: {\phi_{-}(v-\frac{m}{2}i)\phi_{+}(v+\frac{m}{2}i)}
437: \quad {\rm for} \quad m \in {\mathbb Z}_{\ge 1}, \\
438: && \widetilde{T}^{(r+1)}_{m}(v)=
439: e^{\frac{m(\mu_{1}+\mu_{2}+\cdots +\mu_{r+1})}{T}}
440: \quad {\rm for} \quad m \in {\mathbb Z}_{\ge 1} \nonumber.
441: \end{eqnarray}
442: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
443: \section{Nonlinear Integral Equations with Only a
444: Finite Number of Unknown Functions}
445: It is known \cite{TSK01} that
446: Takahashi's NLIE \cite{Ta01},
447: which describes thermodynamics of the $XXZ$-model,
448: can be derived from
449: the $T$-system of the QTM. The number of unknown functions
450: for this NLIE
451: is only one, which corresponds to the rank of
452: the underlying algebra
453: $U_{q}(\widehat{sl(2)})$.
454: In this section, we will derive the
455: NLIE eq. (\ref{nlie4}) for $U_{q}(\widehat{sl(r+1)})$,
456: which contain only $r$ unknown functions, from
457: the $T$-system eq. (\ref{T-system}).
458: %%%%%%%%%%%%%%%
459:
460: From a numerical analysis for finite $N,u_{N},r$,
461: we conjecture that
462: a one-string solution (for every color) in the sector
463: $\frac{N}{2}=M_{1}=M_{2}=\cdots =M_{r}$
464: of the BAE eq. (\ref{BAE}) gives
465: the largest eigenvalue of the QTM eq. (\ref{QTM}) at $v=0$.
466: From now on, we will consider only this one-string solution.
467: Thus $T^{(1)}_{1}(0)$ should give the
468: largest eigenvalue $\Lambda_{1}$.
469: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
470: We expect that
471: the following conjecture is valid for
472: this one-string solution
473: (cf. Figs. \ref{roots} and \ref{zeros}).
474: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
475: \begin{figure}
476: \begin{center}
477: \includegraphics[width=0.95\textwidth]{fig1.eps}
478: \end{center}
479: \caption{Location of the roots $\{v^{(a)}_{k}\}$ ($a=1,2$)
480: of the BAE for $U_{q}(\widehat{sl(3)})$
481: ($\eta=2.5$, $N=12$, $u_{N}=-0.05$,
482: $\mu_{1}=h,\mu_{2}=0,\mu_{3}=-h$), which will give
483: the largest eigenvalue of the QTM $t_{\mathrm{QTM}}(v)$
484: at $v=0$.
485: Color 1 roots $\{v^{(1)}_{k}\}$ and
486: color 2 roots $\{v^{(2)}_{k}\}$
487: form six one-strings.
488: Only the roots in the fundamental domain
489: ${\rm Re} v \in [-\frac{\pi}{2\eta},\frac{\pi}{2\eta}]$
490: are exhibited.
491: See also Fig. 1 in ref. \cite{T03} for $\eta=0$, $h=0$ case. }
492: \label{roots}
493: \end{figure}
494: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
495: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
496: \begin{figure}
497: \begin{center}
498: \includegraphics[width=0.95\textwidth]{fig2.eps}
499: \end{center}
500: \caption{Location of the zeros of $\widetilde{T}^{(a)}_{1}(v)$
501: ($a=1,2$)
502: corresponding to the roots in Fig. \ref{roots}.
503: Only the zeros in the fundamental domain
504: ${\rm Re} v \in [-\frac{\pi}{2\eta},\frac{\pi}{2\eta}]$
505: are exhibited.
506: See also Figs. 2 and 3 in ref. \cite{T03} for $\eta=0$, $h=0$ case.}
507: \label{zeros}
508: \end{figure}
509: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
510: \begin{conjecture}\label{conj}
511: For small $u_{N}$ ($|u_{N}|\ll 1$),
512: all the zeros $\{\tilde{z}^{(a)}_{1} \}$
513: of $\widetilde{T}^{(a)}_{1}(v)$ ($a \in \{1,2,\dots,r\}$)
514: are located outside of a domain (the physical strip)
515: $\mathrm{Im} v \in [-\frac{1}{2},\frac{1}{2}]$.
516: To be precise, they are located near the lines
517: $\mathrm{ Im} v= \pm \frac{1+a}{2}$.
518: \end{conjecture}
519: We expect that this conjecture is also valid in the
520: Trotter limit $N \to \infty$.
521: %%%%%%%%%%
522: $\widetilde{T}^{(a)}_{m}(v)$ has poles only at
523: $\pm \tilde{\beta}^{(a)}_{m,n}$:
524: $\tilde{\beta}^{(a)}_{m,n}=\frac{m+a}{2}i +iu_{N}+\frac{n\pi}{\eta}$
525: ($n \in \mathbb{Z}$) whose
526: order is at most $N/2$. It is a periodic function with respect
527: to $v$ whose
528: period is $\frac{\pi }{\eta}$.
529: Moreover,
530: \begin{eqnarray}
531: Q^{(a)}_{1} :=
532: \lim_{v\to i \eta \infty}\widetilde{T}^{(a)}_{1}(v)
533: =\sum_{1 \le i_{1} < i_{2} < \cdots < i_{a} \le r+1}
534: e^{\frac{\mu_{i_{1}}}{T}}
535: e^{\frac{\mu_{i_{2}}}{T}}
536: \cdots
537: e^{\frac{\mu_{i_{a}}}{T}}
538: \label{Q-sl(r+1)}
539: \end{eqnarray}
540: is a finite number \cite{q-sys}.
541: Thus, we must put
542: \begin{eqnarray}
543: \widetilde{T}^{(a)}_{1}(v)=Q^{(a)}_{1} +
544: \sum_{n \in \mathbb{Z}}
545: \sum_{j=1}^{\frac{N}{2}}
546: \left\{
547: \frac{b^{(a)}_{j}}{(v-\tilde{\beta}^{(a)}_{1,n})^{j}}
548: +\frac{\overline{b}^{(a)}_{j}}{(v+\tilde{\beta}^{(a)}_{1,n})^{j}}
549: \right\},
550: \label{expan}
551: \end{eqnarray}
552: %
553: where the coefficients $b^{(a)}_{j},
554: \overline{b}^{(a)}_{j} \in {\mathbb C}$ are
555: independent of $v$.
556: %
557: Taking note on the relation
558: \begin{eqnarray}
559: \lim_{m \to \infty}
560: \sum_{n=-m}^{m}\frac{1}{v-\frac{\pi n}{\eta}}
561: =\frac{\eta}{\tan \eta v},
562: \end{eqnarray}
563: we can derive the following NLIE for finite Trotter
564: number $N$ from eqs.
565: (\ref{expan}) and (\ref{T-system}) by the similar procedures
566: in refs. \cite{TSK01} and \cite{T03}.
567: %%%%%%%%
568: \begin{eqnarray}
569: && \hspace{-20pt}
570: \widetilde{T}^{(a)}_{1}(v)=Q^{(a)}_{1} \nonumber \\
571: && +
572: \oint_{C^{(a)}} \frac{{\mathrm d} y}{2\pi i}
573: \frac{\eta
574: \widetilde{T}^{(a-1)}_{1}(y+\tilde{\beta}^{(a)}_{1}-\frac{i}{2})
575: \widetilde{T}^{(a+1)}_{1}(y+\tilde{\beta}^{(a)}_{1}-\frac{i}{2})}
576: {\tan \eta (v-y-\tilde{\beta}^{(a)}_{1})
577: \widetilde{T}^{(a)}_{1}(y+\tilde{\beta}^{(a)}_{1}-i)}
578: \nonumber \\
579: && +
580: \oint_{\overline{C}^{(a)}} \frac{{\mathrm d} y}{2\pi i}
581: \frac{
582: \eta
583: \widetilde{T}^{(a-1)}_{1}(y-\tilde{\beta}^{(a)}_{1}+\frac{i}{2})
584: \widetilde{T}^{(a+1)}_{1}(y-\tilde{\beta}^{(a)}_{1}+\frac{i}{2})}
585: {\tan \eta (v-y+\tilde{\beta}^{(a)}_{1})
586: \widetilde{T}^{(a)}_{1}(y-\tilde{\beta}^{(a)}_{1}+i)}
587: \nonumber \\
588: && \hspace{180pt}
589: {\rm for} \quad a \in \{1,2,\dots,r\}, \label{nlie3}
590: \end{eqnarray}
591: where we set $\tilde{\beta}^{(a)}_{1}:=\tilde{\beta}^{(a)}_{1,0}$;
592: the contour $C^{(a)}$ (resp. $\overline{C}^{(a)}$)
593: is a counterclockwise closed loop around $0$
594: which satisfies the condition
595: $y \ne v-\tilde{\beta}^{(a)}_{1,n}$
596: (resp. $y \ne v+\tilde{\beta}^{(a)}_{1,n}$) and
597: does not surround
598: $\tilde{z}^{(a)}_{1}-\tilde{\beta}^{(a)}_{1,n}+i$,
599: $-2\tilde{\beta}^{(a)}_{1}+\frac{\pi n}{\eta}$, $\frac{\pi k}{\eta}$
600: (resp.
601: $\tilde{z}^{(a)}_{1}+\tilde{\beta}^{(a)}_{1,n}-i$,
602: $2 \tilde{\beta}^{(a)}_{1}+\frac{\pi n}{\eta}$, $\frac{\pi k}{\eta}$)
603: ($n \in \mathbb{Z}$, $k \in \mathbb{Z}-\{0\}$).
604: %%%
605:
606: Next we will take the Trotter limit $N \to \infty$.
607: %
608: We put $\mathcal{T}^{(a)}_{1}(v):=
609: \lim_{N \to \infty} \widetilde{T}^{(a)}_{1}(v)$.
610: In particular, we have
611: \begin{eqnarray}
612: \mathcal{T}^{(0)}_{1}(v)=
613: \lim_{N \to \infty} \widetilde{T}^{(0)}_{1}(v)
614: =\exp \left(\frac{2J (\sinh \eta)^{2} }
615: {T(\cosh \eta -\cos (2\eta v))}\right).
616: \end{eqnarray}
617: Then, we arrive at a system of NLIE, which
618: contains only a {\em finite} number of unknown
619: functions $\{{\mathcal T}^{(a)}_{1}(v) \}_{1\le a \le r}$:
620: \begin{eqnarray}
621: {\mathcal T}^{(a)}_{1}(v)=Q^{(a)}_{1}
622: &+&
623: \oint_{C^{(a)}} \frac{{\mathrm d} y}{2\pi i}
624: \frac{\eta
625: \mathcal{T}^{(a-1)}_{1}(y+\frac{i a}{2})
626: \mathcal{T}^{(a+1)}_{1}(y+\frac{i a}{2})}
627: {\tan \eta (v-y-\frac{i(a+1)}{2})
628: \mathcal{T}^{(a)}_{1}(y+\frac{i(a-1)}{2})}
629: \nonumber \\
630: &+&
631: \oint_{\overline{C}^{(a)}} \frac{{\mathrm d} y}{2\pi i}
632: \frac{\eta
633: \mathcal{T}^{(a-1)}_{1}(y-\frac{i a}{2})
634: \mathcal{T}^{(a+1)}_{1}(y-\frac{i a}{2})}
635: {\tan \eta (v-y+\frac{i(a+1)}{2})
636: \mathcal{T}^{(a)}_{1}(y-\frac{i(a-1)}{2})}
637: \nonumber \\
638: && \hspace{120pt}
639: {\rm for} \quad a \in \{1,2,\dots,r\},
640: \label{nlie4}
641: \end{eqnarray}
642: where $\mathcal{T}^{(r+1)}_{1}(v)=Q^{(r+1)}_{1}=
643: e^{\frac{\mu_{1}+\cdots +\mu_{r+1}}{T}}$;
644: %
645: the contour $C^{(a)}$ (resp. $\overline{C}^{(a)}$)
646: is a counterclockwise closed loop around $0$
647: which satisfies the condition
648: $y \ne v-\beta^{(a)}_{1,n}$
649: (resp. $y \ne v+\beta^{(a)}_{1,n}$) and
650: does not surround
651: $z^{(a)}_{1}-\beta^{(a)}_{1,n}+i$,
652: $-(a+1)i%-2\beta^{(a)}_{1}
653: +\frac{\pi n}{\eta}$, $\frac{\pi k}{\eta}$
654: (resp.
655: $z^{(a)}_{1}+\beta^{(a)}_{1,n}-i$,
656: $(a+1)i%2 \beta^{(a)}_{1}
657: +\frac{\pi n}{\eta}$, $\frac{\pi k}{\eta}$);
658: %$\beta^{(a)}_{1}=\lim_{N\to \infty}\tilde{\beta}^{(a)}_{1}$;
659: $\beta^{(a)}_{1,n}=\lim_{N\to \infty}\tilde{\beta}^{(a)}_{1,n}
660: =\frac{a+1}{2}i+\frac{\pi n}{\eta}$;
661: $z^{(a)}_{1}=\lim_{N\to \infty }{\tilde z}^{(a)}_{1}$
662: ($n \in \mathbb{Z}$, $k \in \mathbb{Z}-\{0\}$).
663: %%%%%%%%%%%%%%%%%%
664: Note that eq. (\ref{nlie4}) reduces
665: to Takahashi's NLIE \cite{Ta01} if $r=1$.
666: %%%%%%%%%%%%%%%%
667: Although we assumed that
668: $q=e^{\eta}$ is not a root of unity,
669: we think that eq. (\ref{nlie4}) will be also valid as it stand
670: even when $q$ is a root of unity.
671: One can calculate the free energy per site $f$
672: by using eq. (\ref{nlie4})
673: and the relation
674: \begin{eqnarray}
675: f=J \cosh \eta -T\log \mathcal{T}^{(1)}_{1}(0).
676: \label{free-en}
677: \end{eqnarray}
678: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
679: \section{High Temperature Expansion}
680: In this section, we will calculate the high temperature
681: expansion of the free energy eq. (\ref{free-en}) for
682: $U_{q}(\widehat{sl(r+1)})$
683: by our new NLIE eq. (\ref{nlie4}).
684: It is known that
685: the high temperature expansion of the free energy
686: for the $XXX$-model
687: was calculated \cite{ShT02} by Takahashi's NLIE
688: up to order 100.
689:
690: First of all, we assume the following expansion for large $T/|J|$:
691: \begin{eqnarray}
692: \mathcal{T}^{(a)}_{1}(v)=
693: \exp \left(\sum_{n=0}^{\infty}b_{n}^{(a)}(v)
694: (\frac{J}{T})^{n} \right),
695: \quad a \in \{1,2,\dots, r\},
696: \label{t-expan}
697: \end{eqnarray}
698: where $b_{0}^{(a)}(v)=\log Q^{(a)}_{1}$.
699: Substituting eq. (\ref{t-expan}) into eq. (\ref{nlie4}), we
700: can obtain the coefficients $\{b_{n}^{(a)}(v)\}$.
701: We treat $r=1$ case separately. \\
702: $\bullet $ $r=1$ case:\\
703: In this case, the Hamiltonian eqs. (\ref{ham0}) and (\ref{hamch}) becomes XXZ chain
704: in magnetic field:
705: \beqa
706: &&H_0+H_{ch}= 2J\sum_{j=1}^L \Bigl\{ S_j^xS_{j+1}^x+S_j^yS_{j+1}^y+
707: \Delta (S_j^zS_{j+1}^z+\frac{1}{4}) \Bigr\} \nonum\\
708: && \hspace{20pt} -L(\mu_1+\mu_2)-2
709: (\mu_1-\mu_2)\sum_{j=1}^L S_j^z, \label{hamXXZ}
710: \enqa
711: where $\Delta=\cosh \eta$.
712: The NLIE has especially simple form.
713: Equation (\ref{nlie4}) is
714: \beqa
715: && \mathcal{T}^{(1)}_{1} (v)= Q^{(1)}_{1} +Q^{(2)}_{1} \oint_C
716: \Bigl(\frac{\mathcal{T}^{(0)}_{1}(y+i/2)}{\tan{\eta}(v-y-i)}
717: +\frac{\mathcal{T}^{(0)}_{1}(y-i/2)}{\tan{\eta}(v-y+i)}
718: \Bigr){\eta\over \mathcal{T}^{(1)}_{1}(y)}{{\rm d}y\over 2\pi i }\nonumber
719: \\ \label{NLIEsl2}
720: \enqa
721: Contour $C$ is a closed loop counterclockwise around $0$.
722: \noindent
723: To remove trigonometric functions, we use transformation
724: \beq
725: v=F(X), \quad y=F(Y):
726: \quad F(X)=\frac{1}{\eta}\tan^{-1}\Bigl({iX\sinh\eta\over \sqrt{1+(\cosh\eta)^{2} X^{2}}}\Bigr).
727: \enq
728: %For formula manipulation we use the following transformation
729: %\[v=\frac{1}{\eta}\tan^{-1}({iX\sinh\eta\over \sqrt{1+(\cosh\eta X)^2}}),\quad
730: %y=\frac{1}{\eta}\tan^{-1}({iY\sinh\eta\over \sqrt{1+(\cosh\eta Y)^2}}).\]
731: %\noindent
732: By this transformation, points $(0,\pm i, \infty)$ on $v$ plane move to $(0, \infty, \pm i)$
733: on $X$ plane, respectively.% After a lengthy calculation we have
734: Then eq. (\ref{NLIEsl2}) becomes
735: \beqa
736: &&\mathcal{T}^{(1)}_{1} (F(X))=Q^{(1)}_{1} +Q^{(2)}_{1} (1+X^2) \oint_C A(X,Y)
737: \frac{1}{\mathcal{T}^{(1)}_{1} (F(Y))}\frac{dY}{2\pi i},\nonum\\
738: &&A(X,Y)=\sum_{\epsilon=\pm 1}\frac{\frac{Y(1-X^2Y^2+2\Delta^2X^2(1+Y^2))}{1+Y^2}
739: +\epsilon\frac{\Delta(1+(1-2\Delta^2)X^2Y^2)}{\sqrt{1+\Delta^2 Y^2}}}
740: {(1-X^2Y^2)^2-4\Delta^2 X^2 Y^2(1+X^2)(1+Y^2)}\nonum\\
741: &&~~~~~~~~~~\times\exp\bigl\{\frac{J}{T}(1+Y^2)(\Delta+\epsilon\frac{\sqrt{1+\Delta^2Y^2}}{Y})\bigr\},
742: \enqa
743: where we assume that $\mathcal{T}^{(1)}_{1} (F(Y))$ is an even function of $Y$.
744: New integration kernel $A$ is expanded by power series of $J/T$. Coefficient of
745: $l$-th term is
746: \beqa
747: &&
748: \frac{1}{ l!}\frac{(1+Y^2)^l}
749: {(1-X^2Y^2)^2-4\Delta^2 X^2 Y^2(1+X^2)(1+Y^2)}\nonum\\
750: &&\Bigl(\frac{Y(1-X^2Y^2+2\Delta^2X^2(1+Y^2))}{1+Y^2}
751: ((\Delta+\frac{\sqrt{1+Y^2\Delta^2}}{Y})^l+(\Delta-\frac{\sqrt{1+Y^2\Delta^2}}{Y})^l)\nonum\\
752: &&+\frac{\Delta(1+(1-2\Delta^2)X^2Y^2)}{\sqrt{1+\Delta^2 Y^2}}
753: ((\Delta+\frac{\sqrt{1+Y^2\Delta^2}}{Y})^l-(\Delta-\frac{\sqrt{1+Y^2\Delta^2}}{Y})^l)
754: \Bigr).\label{exA}
755: \enqa
756: Analytic part of function $A$ at $Y=0$ is not relevant in this integral. Then we can put
757: \beqa
758: &&A=({\rm analytic~~ part~~at~~}Y=0)+ \sum_{l=1}^\infty{J^l\over l!T^l}A_l,
759: \enqa
760: We get $A_l$ from eq. (\ref{exA}) as polynomials of $X^2, \Delta^2$ and $1/Y$;
761: {\beqa
762: &&A_1=\frac{\Delta }{Y},\quad
763: A_2=\frac{1 + 2\,{\Delta }^2 + 2{\Delta }^2X^2}{Y},\nonum\\
764: &&A_3=\frac{\Delta }{Y^3} + \frac{6\Delta +
765: 4{\Delta }^3 + (3\Delta + 8{\Delta }^3)X^2 + 4 {\Delta }^3X^4}{Y},\quad
766: A_4=\frac{1 + 4{\Delta }^2 + 2 {\Delta }^2X^2}{Y^3}\nonum\\
767: && + \frac{3 + 24{\Delta }^2 + 8{\Delta }^4 + (1+ 24{\Delta }^2 + 24{\Delta }^4)X^2
768: + 8({\Delta }^2 + 24{\Delta }^4) X^4+ 8{\Delta }^4X^6}{Y},\nonum\\
769: &&....
770: \enqa
771: }
772: We can determine $b^{(1)}_{l}(F(X))$ successively from the integral equation.
773: These are polynomials of $X^2, \Delta$ and ${Q^{(2)}_{1}}/{{Q^{(1)}_{1}}^2}$
774: ($=1/(2 \cosh \frac{h}{T})^{2}$, $h=(\mu_{1}-\mu_{2})/2$ ).
775: \begin{eqnarray}
776: && \hspace{-30pt} b^{(1)}_{1}(0)=
777: \frac{2\,\Delta\,Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2},
778: \label{hte-1} \\
779: %
780: && \hspace{-30pt} b^{(1)}_{2}(0)=
781: \frac{\left( 1 + 2\,\Delta^2 \right) \,
782: Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2} -
783: \frac{6\,\Delta^2\,{Q^{(2)}_{1}}^2}{{Q^{(1)}_{1}}^4},
784: \label{hte-2} \\
785: %
786: && \hspace{-30pt} b^{(1)}_{3}(0)=
787: \frac{\left( 2\,\Delta + \frac{4\,\Delta^3}{3} \right) \,
788: Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2} + \frac{\left( -6\,\Delta - 12\,\Delta^3 \right)
789: \,{Q^{(2)}_{1}}^2}{{Q^{(1)}_{1}}^4} +
790: \frac{80\,\Delta^3\,{Q^{(2)}_{1}}^3}{3\,{Q^{(1)}_{1}}^6},
791: \label{hte-3} \\
792: %
793: && \hspace{-30pt} b^{(1)}_{4}(0)=
794: \frac{\left( \frac{1}{4} + 2\,\Delta^2 + \frac{2\,\Delta^4}{3} \right) \,
795: Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2}
796: + \frac{\left( -\frac{3}{2} - \
797: \frac{56\,\Delta^2}{3} - 14\,\Delta^4 \right)
798: \,{Q^{(2)}_{1}}^2}{{Q^{(1)}_{1}}^4}
799: \nonumber \\ &&
800: \hspace{50pt} + \frac{\left(40\,\Delta^2 + 80\,\Delta^4 \right) \,
801: {Q^{(2)}_{1}}^3}{{Q^{(1)}_{1}}^6} -
802: \frac{140\,\Delta^4\,{Q^{(2)}_{1}}^4}{{Q^{(1)}_{1}}^8}.
803: \label{hte-4}
804: \end{eqnarray}
805: In Table \ref{coet}, we give $ j!b^{(1)}_{j}(0)$ up to $j=10$.
806: %If we write ${Q^{(2)}_{1}}/{{Q^{(1)}_{1}}^2}=M$, $ j!b^{(1)}_{j}(0)$ are given in Table (\ref{coet}).
807: %The number of terms in $b_j^{(1)}(X)$ is proportional to $j^3$ because there exists three variables
808: %$\Delta, X, M$. Using Mathematica we can expand up to 16-th order.
809: %If we put $\Delta$ a constant, numer of terms is proportional to $j^2$.
810: %So we can expand up to 100-th order.
811: \begin{table}
812: \caption{ $ j!\cdot b^{(1)}_{j}(0)$ for $j=1,...,10$.
813: We put $M={Q^{(2)}_{1}}/{{Q^{(1)}_{1}}^2}$
814: ($=1/(2 \cosh \frac{h}{T})^{2}$, $h=(\mu_{1}-\mu_{2})/2$ ).}
815: {\tiny
816: \begin{eqnarray}
817: 1 &&2M\Delta ,\quad\nonum\\
818: 2 &&2M + 4M{\Delta }^2 - 12M^2{\Delta }^2, \nonum\\
819: 3 &&
820: 12\,M\,\Delta - 36\,M^2\,\Delta + 8\,M\,{\Delta }^3 -
821: 72\,M^2\,{\Delta }^3 + 160\,M^3\,{\Delta }^3, \nonum\\
822: 4 &&
823: 6\,M - 36\,M^2 + 48\,M\,{\Delta }^2 - 448\,M^2\,{\Delta }^2 +
824: 960\,M^3\,{\Delta }^2 + 16\,M\,{\Delta }^4 - 336\,M^2\,{\Delta }^4\nonum\\
825: &&
826: +
827: 1920\,M^3\,{\Delta }^4 - 3360\,M^4\,{\Delta }^4, \nonum\\
828: 5 &&
829: 60\,M\,\Delta - 960\,M^2\,\Delta + 2400\,M^3\,\Delta +
830: 160\,M\,{\Delta }^3 - 3600\,M^2\,{\Delta }^3 + 20000\,M^3\,{\Delta }^3 -
831: 33600\,M^4\,{\Delta }^3 \nonum\\
832: &&
833: + 32\,M\,{\Delta }^5 - 1440\,M^2\,{\Delta }^5 +
834: 16000\,M^3\,{\Delta }^5 - 67200\,M^4\,{\Delta }^5 + 96768\,M^5\,{\Delta }^5, \nonum\\
835: 6 &&
836: 20\,M - 600\,M^2 + 2400\,M^3 + 360\,M\,{\Delta }^2 -
837: 13320\,M^2\,{\Delta }^2 + 85200\,M^3\,{\Delta }^2 -
838: 151200\,M^4\,{\Delta }^2\nonum\\
839: &&
840: + 480\,M\,{\Delta }^4 - 23808\,M^2\,{\Delta }^4 +
841: 259200\,M^3\,{\Delta }^4 - 1048320\,M^4\,{\Delta }^4 +
842: 1451520\,M^5\,{\Delta }^4 \nonum\\
843: &&
844: + 64\,M\,{\Delta }^6 - 5952\,M^2\,{\Delta }^6 +
845: 115200\,M^3\,{\Delta }^6 - 873600\,M^4\,{\Delta }^6 +
846: 2903040\,M^5\,{\Delta }^6\nonum\\
847: &&
848: - 3548160\,M^6\,{\Delta }^6, \nonum\\
849: 7 &&
850: 280\,M\,\Delta - 18760\,M^2\,\Delta + 168000\,M^3\,\Delta -
851: 352800\,M^4\,\Delta + 1680\,M\,{\Delta }^3\nonum\\
852: &&
853: - 135408\,M^2\,{\Delta }^3 +
854: 1689856\,M^3\,{\Delta }^3 - 7197120\,M^4\,{\Delta }^3 +
855: 10160640\,M^5\,{\Delta }^3 \nonum\\
856: &&
857: + 1344\,M\,{\Delta }^5 -
858: 141120\,M^2\,{\Delta }^5 + 2696960\,M^3\,{\Delta }^5 -
859: 19756800\,M^4\,{\Delta }^5 \nonum\\
860: &&
861: + 63221760\,M^5\,{\Delta }^5 -
862: 74511360\,M^6\,{\Delta }^5 + 128\,M\,{\Delta }^7 -
863: 24192\,M^2\,{\Delta }^7 \nonum\\
864: &&
865: + 770560\,M^3\,{\Delta }^7 -
866: 9408000\,M^4\,{\Delta }^7 + 54190080\,M^5\,{\Delta }^7 -
867: 149022720\,M^6\,{\Delta }^7 \nonum\\
868: &&
869: + 158146560\,M^7\,{\Delta }^7, \nonum\\
870: %\enqa
871: %\beqa
872: 8 &&
873: 70\,M - 8820\,M^2 + 117600\,M^3 - 352800\,M^4 + 2240\,M\,{\Delta }^2 -
874: 322560\,M^2\,{\Delta }^2 \nonum\\
875: &&
876: + 5388096\,M^3\,{\Delta }^2 -
877: 26530560\,M^4\,{\Delta }^2 + 40642560\,M^5\,{\Delta }^2 +
878: 6720\,M\,{\Delta }^4 \nonum\\
879: &&
880: - 1145088\,M^2\,{\Delta }^4 +
881: 25079040\,M^3\,{\Delta }^4 - 193277952\,M^4\,{\Delta }^4 +
882: 629959680\,M^5\,{\Delta }^4 \nonum\\
883: &&
884: - 745113600\,M^6\,{\Delta }^4 +
885: 3584\,M\,{\Delta }^6 - 780288\,M^2\,{\Delta }^6 +
886: 24729600\,M^3\,{\Delta }^6 \nonum\\
887: &&
888: - 292626432\,M^4\,{\Delta }^6 +
889: 1625702400\,M^5\,{\Delta }^6 - 4314562560\,M^6\,{\Delta }^6 \nonum\\
890: &&
891: +
892: 4428103680\,M^7\,{\Delta }^6 + 256\,M\,{\Delta }^8 -
893: 97536\,M^2\,{\Delta }^8 + 4945920\,M^3\,{\Delta }^8 \nonum\\
894: &&
895: -
896: 91445760\,M^4\,{\Delta }^8 + 812851200\,M^5\,{\Delta }^8 -
897: 3775242240\,M^6\,{\Delta }^8 + 8856207360\,M^7\,{\Delta }^8 \nonum\\
898: &&
899: -
900: 8302694400\,M^8\,{\Delta }^8, \nonum\\
901: 9 &&
902: 1260\,M\,\Delta - 337680\,M^2\,\Delta + 8023680\,M^3\,\Delta -
903: 50803200\,M^4\,\Delta + 91445760\,M^5\,\Delta \nonum\\
904: &&
905: + 13440\,M\,{\Delta }^3 -
906: 4072320\,M^2\,{\Delta }^3 + 116998560\,M^3\,{\Delta }^3 -
907: 1030458240\,M^4\,{\Delta }^3 \nonum\\
908: &&
909: + 3617187840\,M^5\,{\Delta }^3 -
910: 4470681600\,M^6\,{\Delta }^3 + 24192\,M\,{\Delta }^5 -
911: 8570880\,M^2\,{\Delta }^5 \nonum\\
912: &&
913: + 311921280\,M^3\,{\Delta }^5 -
914: 3884186880\,M^4\,{\Delta }^5 + 21973690368\,M^5\,{\Delta }^5 -
915: 58502062080\,M^6\,{\Delta }^5 \nonum\\
916: &&
917: + 59779399680\,M^7\,{\Delta }^5 +
918: 9216\,M\,{\Delta }^7 - 4112640\,M^2\,{\Delta }^7 +
919: 209088000\,M^3\,{\Delta }^7 \nonum\\
920: &&
921: - 3759436800\,M^4\,{\Delta }^7 +
922: 32286449664\,M^5\,{\Delta }^7 - 144850083840\,M^6\,{\Delta }^7 +
923: 328786698240\,M^7\,{\Delta }^7\nonum\\
924: &&
925: - 298896998400\,M^8\,{\Delta }^7 +
926: 512\,M\,{\Delta }^9 - 391680\,M^2\,{\Delta }^9 +
927: 30976000\,M^3\,{\Delta }^9 - 835430400\,M^4\,{\Delta }^9 \nonum\\
928: &&
929: +
930: 10762149888\,M^5\,{\Delta }^9 - 75107450880\,M^6\,{\Delta }^9 +
931: 292254842880\,M^7\,{\Delta }^9 \nonum\\
932: &&
933: - 597793996800\,M^8\,{\Delta }^9 +
934: 501851750400\,M^9\,{\Delta }^9, \nonum\\
935: 10 &&
936: 252\,M - 128520\,M^2 + 4445280\,M^3 - 38102400\,M^4 + 91445760\,M^5 +
937: 12600\,M\,{\Delta }^2\nonum\\
938: &&
939: - 7072800\,M^2\,{\Delta }^2 +
940: 282582720\,M^3\,{\Delta }^2 - 3083330880\,M^4\,{\Delta }^2 +
941: 12395980800\,M^5\,{\Delta }^2\nonum\\
942: &&
943: - 16765056000\,M^6\,{\Delta }^2 +
944: 67200\,M\,{\Delta }^4 - 42297600\,M^2\,{\Delta }^4 +
945: 1999721280\,M^3\,{\Delta }^4 \nonum\\
946: &&
947: - 28282381440\,M^4\,{\Delta }^4 +
948: 171577405440\,M^5\,{\Delta }^4 - 475808256000\,M^6\,{\Delta }^4 +
949: 498161664000\,M^7\,{\Delta }^4\nonum\\
950: &&
951: + 80640\,M\,{\Delta }^6 -
952: 58803840\,M^2\,{\Delta }^6 + 3444222720\,M^3\,{\Delta }^6 -
953: 65241899520\,M^4\,{\Delta }^6 \nonum\\
954: &&
955: + 570726051840\,M^5\,{\Delta }^6 -
956: 2568513024000\,M^6\,{\Delta }^6 + 5803583385600\,M^7\,{\Delta }^6 \nonum\\
957: &&
958: -
959: 5230697472000\,M^8\,{\Delta }^6 + 23040\,M\,{\Delta }^8 -
960: 20930560\,M^2\,{\Delta }^8 + 1671936000\,M^3\,{\Delta }^8 \nonum\\
961: &&
962: -
963: 44003635200\,M^4\,{\Delta }^8 + 548674560000\,M^5\,{\Delta }^8 -
964: 3702575923200\,M^6\,{\Delta }^8\nonum\\
965: &&
966: + 13948526592000\,M^7\,{\Delta }^8 -
967: 27675648000000\,M^8\,{\Delta }^8 + 22583328768000\,M^9\,{\Delta }^8 \nonum\\
968: &&
969: +
970: 1024\,M\,{\Delta }^{10} - 1569792\,M^2\,{\Delta }^{10}+
971: 191078400\,M^3\,{\Delta }^{10} - 7333939200\,M^4\,{\Delta }^{10} +
972: 131681894400\,M^5\,{\Delta }^{10} \nonum\\
973: &&
974: - 1295901573120\,M^6\,{\Delta }^{10} +
975: 7439214182400\,M^7\,{\Delta }^{10} - 24908083200000\,M^8\,{\Delta }^{10}\nonum\\
976: &&
977: +
978: 45166657536000\,M^9\,{\Delta }^{10} -
979: 34326659727360\,M^{10}\,{\Delta }^{10} \nonum
980: \end{eqnarray}
981: }
982: \label{coet}
983: \end{table}
984:
985: %
986: There are a lot of works on high temperature expansions for the XXX model;
987: only a few works exist for the XXZ model.
988: Destri and de Vega 's \cite{DV95} result contains some misprint.
989: Rojas et al \cite{RST02} also calculated the coefficients
990: up to order $O((\frac{J}{T})^{4})$. Unfortunately, their formula
991: (eqs. (46)-(48) in ref. \cite{RST02})
992: contains a misprint of a sign.
993: We note that eqs.
994: (\ref{hte-1})-(\ref{hte-3}) agree with eq. (46) in ref. \cite{RST02}.
995: In addition, eq. (\ref{hte-4}) also agrees with eq. (46) in ref.
996: \cite{RST02} {\em if}
997: we replace the term $-\frac{t^4}{32}\beta^3$ in eq. (48)
998: with $+\frac{t^4}{32}\beta^3$.
999: Moreover, these coefficients eqs. (\ref{hte-1})-(\ref{hte-4}) reduce to
1000: known results (see for example, refs. \cite{ShT02} and \cite{BEU00})
1001: in the rational limit $\Delta \to 1$.
1002: We have plotted specific heat by using our high temperature
1003: expansion formula of order 99 (see Fig. \ref{specific-h=2}). \\
1004: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1005: $\bullet $ $r \ge 2$ case: \\
1006: This case is more involved than $r=1$ case since
1007: we have to take account of
1008: the poles in the coefficients $\{b_{n}^{(a)}(v)\}$. Thus
1009: we need a further non-trivial assumption
1010: \begin{eqnarray}
1011: b_{n}^{(a)}(v)=\left(\frac{2 (\sinh \eta)^{2}}
1012: {\cosh (a+1)\eta -\cos 2\eta v}\right)^n c^{(a)}_{n}(v),
1013: \end{eqnarray}
1014: where $c_{n}^{(a)}(v)$ is a polynomial
1015: with respect to some trigonometric functions.
1016: We will present $b^{(1)}_{n}(0)$ for $U_{q}(\widehat{sl(r+1)})$ up
1017: to order $n=3$.
1018: \begin{eqnarray}
1019: && b^{(1)}_{1}(0)=
1020: \frac{2\,\Delta\,Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2},
1021: \label{hte-hr1} \\
1022: && b^{(1)}_{2}(0)=
1023: \frac{\left( 1 + 2\,\Delta^2 \right) \,Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2} -
1024: \frac{6\,\Delta^2\,{Q^{(2)}_{1}}^2}{{Q^{(1)}_{1}}^4} +
1025: \frac{\left( -1 + 4\,\Delta^2 \right) \,Q^{(3)}_{1}}{{Q^{(1)}_{1}}^3},
1026: \label{hte-hr2}
1027: \end{eqnarray}
1028: %\\
1029: \begin{eqnarray}
1030: && b^{(1)}_{3}(0)=
1031: \frac{\left( 2\,\Delta + \frac{4\,\Delta^3}{3} \right)
1032: \,Q^{(2)}_{1}}{{Q^{(1)}_{1}}^2} +
1033: \frac{\left( -6\,\Delta - 12\,\Delta^3 \right)
1034: \,{Q^{(2)}_{1}}^2}{{Q^{(1)}_{1}}^4} +
1035: \frac{80\,\Delta^3\,{Q^{(2)}_{1}}^3}{3\,{Q^{(1)}_{1}}^6}
1036: \nonumber \\ && \hspace{10pt} +
1037: \frac{8\,\Delta^3\,Q^{(3)}_{1}}{{Q^{(1)}_{1}}^3} +
1038: \frac{\left( 8\,\Delta - 32\,\Delta^3 \right)
1039: \,Q^{(2)}_{1}\,Q^{(3)}_{1}}{{Q^{(1)}_{1}}^5} +
1040: \frac{\left( -4\,\Delta + 8\,\Delta^3 \right)
1041: \,Q^{(4)}_{1}}{{Q^{(1)}_{1}}^4}. \label{hte-hr3}
1042: \end{eqnarray}
1043: %
1044: We note the following relation:
1045: $b_{n}^{(1)}(0)=c_{n}^{(1)}(0)$.
1046: We can obtain coefficients for $r=2$ if we formally
1047: set $Q^{(4)}_{1}=0$ in eq. (\ref{hte-hr3}).
1048: Note that eqs. (\ref{hte-hr1}), (\ref{hte-hr2}), (\ref{hte-hr3})
1049: reduce to eqs.
1050: (\ref{hte-1}), (\ref{hte-2}), (\ref{hte-3})
1051: if $Q^{(3)}_{1}=Q^{(4)}_{1}=0$.
1052: Note also that eqs. (\ref{hte-hr1})-(\ref{hte-hr3}) reduce to
1053: eq. (5.6) in ref. \cite{T03} in the rational limit $\Delta \to 1$.
1054: We have plotted the specific heat by using our high temperature
1055: expansion formula eqs. (\ref{hte-hr1})-(\ref{hte-hr3}) of order 3
1056: (see Fig. \ref{specific-sl4}).
1057: \begin{figure}
1058: \begin{center}
1059: \includegraphics[width=0.75\textwidth]
1060: %{specific-h=2.eps}
1061: {fig3.eps}
1062: \end{center}
1063: \caption{Temperature dependence of the specific heat $C$
1064: for $U_{q}(\widehat{sl(2)})$ ($\mu_{1}=-\mu_{2}=0$) calculated
1065: from Pade approximation
1066: (numerator: a polynomial of degree 49,
1067: denominator: a polynomial of degree 50)
1068: for $\Delta=$2 (thin), 1 (medium), 0.5 (thick),
1069: -0.5 (dashed thick), -1 (dashed medium), -2 (dashed thin).
1070: We have calculated high-temperature expansion of free energy up to 99-th order.
1071: The case for $\Delta=\pm 1$ coincides with ref. \cite{ShT02}.}
1072: \label{specific-h=2}
1073: \end{figure}
1074: %%%%%%%%%%%%%%%%%%%%%%%%%%%%
1075: %%%
1076: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1077: \begin{figure}
1078: \begin{center}
1079: \includegraphics[width=0.75\textwidth]
1080: {fig4.eps}
1081: \end{center}
1082: \caption{Temperature dependence of the high temperature
1083: expansion of the specific heat $C$
1084: for $U_{q}(\widehat{sl(4)})$ ($J=0.48$, $\mu_{1}=6.75$,
1085: $\mu_{2}=\mu_{3}=\mu_{4}=0$):
1086: each line denotes $C$ for
1087: $\eta=\frac{3}{2}$ (thin),$1$ (dashed thin), $0$ (thick),
1088: $\frac{\pi i}{3}$ (dashed thick) respectively.
1089: The case for $\eta=0$
1090: was calculated in ref. \cite{BGOSTF03} in relation
1091: with the $su(4)$ ladder model.}
1092: \label{specific-sl4}
1093: \end{figure}
1094: %%%%%%%%%%%%%%%%%%%%%%%%%%%%
1095: \section{Concluding Remarks}
1096: We have derived NLIE with only $r$ unknown functions
1097: for the $U_{q}(\widehat{sl(r+1)})$ Perk-Schultz model.
1098: In contrast with traditional TBA equations,
1099: our new NLIE will be also valid even when $q$ is a root of unity.
1100: We also calculated the high temperature expansion
1101: of the free energy from our new NLIE.
1102:
1103: It will be an important problem to apply our result
1104: to physical problem such as the thermodynamics of
1105: the spin ladder models as in ref. \cite{BGOSTF03}.
1106: In Ref. \cite{YRFC04}, they treated a spin ladder model whose
1107: leg part has isotropic interaction
1108: while rung part has anisotropic interaction.
1109: In contrast, our new NLIE has a potential
1110: applicability to spin ladder models
1111: whose leg parts have anisotropic interaction.
1112:
1113: We note that our results in this paper reduce to the ones in
1114: ref. \cite{T03} in the rational limit $q=e^{\eta} \to 1$.
1115: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1116: \section*{Acknowledgments}
1117: \noindent
1118: This work is partially supported by
1119: Grant-in-Aid for Scientific Research
1120: from JSPS (no. 16914018).
1121: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1122: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1123: %\eqreset
1124: %\renewcommand{\theequation}{A.1.\arabic{equation}}
1125: %\section*{Appendix:}
1126: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1127: %\begin{small}
1128: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1129: \begin{thebibliography}{80}
1130: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1131: \bibitem{Ta99}
1132: M. Takahashi: Thermodynamics of One-Dimensional Solvable
1133: models, (Cambridge University Press, Cambridge, 1999).
1134:
1135: \bibitem{K93}
1136: A. Kl\"umper: Z. Phys. {\bf B91} (1993) 507-519.
1137:
1138: \bibitem{K92}
1139: A. Kl\"umper:
1140: Ann. Physik {\bf 1} (1992) 540-553.
1141:
1142: \bibitem{DD92}
1143: C. Destri and H. J. de Vega:
1144: Phys. Rev. Lett. {\bf 69} (1992) 2313-2317.
1145:
1146: \bibitem{Ta01}
1147: M. Takahashi: in Physics and Combinatorics,
1148: eds. A. N. Kirillov and N. Liskova, (2001) 299--304
1149: (World Scientific, Singapore);
1150: cond-mat/0010486.
1151:
1152: \bibitem{TSK01}
1153: M. Takahashi, M. Shiroishi and A. Kl\"umper:
1154: J. Phys. A: Math. Gen. {\bf 34} (2001) L187--L194;
1155: cond-mat/0102027.
1156:
1157: \bibitem{KR87}
1158: A. N. Kirillov and N. Yu. Reshetikhin:
1159: J. Phys. A: Math. Gen. {\bf 20} (1987) 1565--1585.
1160:
1161: \bibitem{KNS94}
1162: A. Kuniba, T. Nakanishi and J. Suzuki:
1163: Int. J. Mod. Phys. {\bf A9} (1994) 5215-5266;
1164: hep-th/9309137.
1165:
1166: \bibitem{S85}
1167: M. Suzuki: Phys. Rev. {\bf B31} (1985) 2957--2965.
1168:
1169: \bibitem{KW02}
1170: G. Kato and M. Wadati:
1171: J. Math. Phys. {\bf 43} (2002) 5060--5078;
1172: cond-mat/0212325.
1173:
1174: \bibitem{ShT02}
1175: M. Shiroishi and M. Takahashi:
1176: Phys. Rev. Lett. 89 (2002) 117201; cond-mat/0205180.
1177:
1178: \bibitem{T02}
1179: Z. Tsuboi:
1180: Phys. Lett. {\bf B544} (2002) 222--230;
1181: math-ph/0209024.
1182:
1183: \bibitem{T03}
1184: Z. Tsuboi:
1185: J. Phys. A: Math. Gen. {\bf 36} (2003) 1493--1507;
1186: cond-mat/0212280.
1187:
1188: \bibitem{T04}
1189: Z. Tsuboi:
1190: J. Phys. A: Math. Gen.
1191: {\bf 37} (2004) 1747--1758; cond-mat/0308333.
1192:
1193: \bibitem{BGOSTF03}
1194: M. T. Batchelor, X. W. Guan, N. Oelkers,
1195: K. Sakai, Z. Tsuboi, A. Foerster:
1196: Phys. Rev. Lett. {\bf 91} (2003) 217202;
1197: cond-mat/0309244.
1198:
1199: \bibitem{YRFC04}
1200: Zu-Jian Ying, I. Roditi, A. Foerster,
1201: B. Chen:
1202: Euro. Phys. J. {\bf B41} (2004) 67-74;
1203: cond-mat/0403520.
1204:
1205: \bibitem{YRZ04}
1206: Zu-Jian Ying, I. Roditi, Huan-Qiang Zhou:
1207: cond-mat/0405274.
1208:
1209: \bibitem{BGO04}
1210: M. T. Batchelor, X. W. Guan, N. Oelkers:
1211: Phys. Rev. {\bf B70} (2004) 184408;
1212: cond-mat/0409310.
1213:
1214: \bibitem{BGOF04}
1215: M. T. Batchelor, X. W. Guan, N. Oelkers, A. Foerster:
1216: JSTAT (2004) P10017;
1217: cond-mat/0409311.
1218:
1219: \bibitem{PS81}
1220: J. H. H. Perk and C. L. Schultz:
1221: Phys. Lett. {\bf 84A} (1981) 407-410.
1222:
1223: \bibitem{Sc83}
1224: C. L. Schultz: Physica {\bf A122} (1983) 71-88.
1225:
1226: \bibitem{algebra}
1227: $U_{q}(\widehat{sl(r+1)})$ is an algebra for
1228: a symmetry of the $R$-matrix of the model.
1229: See for example, M. Jimbo: Commun. Math. Phys. {\bf 102} 537--547.
1230: We assume that $q=e^{\eta}$ is not a root of unity.
1231:
1232: \bibitem{U70}
1233: G.V. Uimin: JETP Lett. {\bf 12} (1970) 225.
1234:
1235: \bibitem{S75}
1236: B. Sutherland: Phys. Rev. {\bf B12} (1975) 3795-3805.
1237:
1238: \bibitem{TS72}
1239: M. Takahashi and M. Suzuki:
1240: Prog. Theor. Phys. {\bf 48} (1972) 2187.
1241:
1242: \bibitem{KSS98}
1243: A. Kuniba, K. Sakai and J. Suzuki:
1244: Nucl. Phys. {\bf B525} (1998) 597-626;
1245: math.QA/9803056.
1246:
1247: \bibitem{SI87} M. Suzuki and M. Inoue:
1248: Prog. Theor. Phys. {\bf 78} (1987) 787--799.
1249:
1250: \bibitem{K87}
1251: T. Koma:
1252: Prog. Theor. Phys. {\bf 78} (1987) 1213--1218.
1253:
1254: \bibitem{SAW90}
1255: J. Suzuki, Y. Akutsu and M. Wadati:
1256: J. Phys. Soc. Jpn. {\bf 59} (1990) 2667--2680.
1257:
1258: \bibitem{KWZ97}
1259: A. Kl\"umper, T. Wehner and J. Zittartz:
1260: J. Phys. A: Math. Gen. {\bf 30} (1997) 1897-1912.
1261:
1262: \bibitem{JKS98}
1263: G. J\"uttner, A. Kl\"umper and J. Suzuki:
1264: Nucl. Phys. {\bf B512} (1998) 581-600;
1265: hep-th/9707074.
1266:
1267: \bibitem{FK99}
1268: A. Fujii and A. Kl\"umper:
1269: Nucl. Phys. {\bf B546} (1999) 751-764;
1270: cond-mat/9811234.
1271:
1272: \bibitem{r-index}
1273: The lower index $i,j$ of $R_{ij}(v)$ is used in
1274: conventional manner. For example, suppose $E^{a,b}_{k}$
1275: is defined on ${\mathbb C}_{r+1}^{\otimes (L+1)}$:
1276: $E^{a,b}_{k}=I^{\otimes k}\otimes E^{a,b}\otimes I^{\otimes (L-k)}$
1277: where $I$ is $r+1$ by $r+1$ identity matrix;
1278: $k=0,1,\dots, L$.
1279: Then
1280: $R_{ij}(v)$ is defined as
1281: $
1282: R_{ij}(v)=\sum_{a_{1},a_{2},b_{1},b_{2}}
1283: R^{a_{1},b_{1}}_{a_{2},b_{2}}(v)
1284: E^{a_{1},a_{2}}_{i} E^{b_{1},b_{2}}_{j}
1285: $.
1286:
1287: \bibitem{BVV82}
1288: O. Babelon, H. J. de Vega and C-M. Viallet:
1289: Nucl. Phys. {\bf B200} (1982) 266-280.
1290:
1291: \bibitem{aba}
1292: This is a kind of \symbol{"60}dress universality' in the analytic Bethe ansatz:
1293: A. Kuniba and J. Suzuki: Commun. Math. Phys. {\bf 173}(1995) 225--264;
1294: hep-th/9406180; N. Yu. Reshetikhin: Sov. Phys. JETP {\bf 57} (1983) 691--696.
1295:
1296: \bibitem{alge-ba}
1297: For $r=1$ case, the algebraic Bethe ansatz for the QTM is
1298: similar to the one for the row-to-row transfer matrix.
1299: See for example,
1300: F. G\"ohmann, A. Kl\"umper and A. Seel:
1301: J. Phys. A: Math. Gen. {\bf 37} 7625-7651; hep-th/0405089.
1302:
1303:
1304: \bibitem{BR90}
1305: V. V. Bazhanov and N. Reshetikhin:
1306: J. Phys. A Math. Gen. {\bf 23} (1990) 1477--1492.
1307:
1308: \bibitem{KRS81}
1309: P. P. Kulish, N. Yu. Reshetikhin and E. K. Sklyanin:
1310: Lett. Math. Phys. {\bf 5} (1981) 393--403.
1311:
1312: \bibitem{q-sys}
1313: This quantity is related to a solution of a
1314: system of functional relations, called
1315: $Q$-system:
1316: A. N. Kirillov:
1317: J. Sov. Math. {\bf 47} (1989) 2450--2459;
1318: A. N. Kirillov and N. Yu. Reshetikhin,
1319: {\it J. Sov. Math.} {\bf 52} (1990) 3156--3164.
1320:
1321: \bibitem{DV95}
1322: C. Destri and H. J. de Vega: Nucl. Phys. {\bf B438}
1323: [FS] (1995) 413-454; hep-th/9407117.
1324:
1325: \bibitem{RST02}
1326: O. Rojas, S. M. de Souza and M. T. Thomaz:
1327: J. Math. Phys. {\bf 43} (2002) 1390-1407;
1328: hep-ph/0012368.
1329:
1330:
1331: \bibitem{BEU00}
1332: A. B\"uhler, N. Elstner and G. S. Uhrig:
1333: Eur. Phys. J {\bf B16}
1334: (2000) 475-486.
1335: \end{thebibliography}
1336: \end{document}
1337:
1338: