1:  % July 29

2: \documentclass{iopart}

3: %\usepackage{showkeys}

4: \usepackage{epsf,amssymb,psfrag,iopams,setstack}%,amsmath}

5: \usepackage{graphicx}

6: %\usepackage[totalwidth=480pt, totalheight=680pt]{geometry}

7: \catcode`\@=11

8: %-------------------------------------------------------------

9: \textwidth 155mm

10: 

11: \def \theequation {\thesection.\arabic{equation}}

12: 

13: %-------------------------------------------------------------

14: \def \be  {\begin{equation}}

15: \def \ee  {\end{equation}}

16: \def \ba  {\begin{eqnarray}}

17: \def \ea  {\end{eqnarray}}

18: \def \dd  {\partial}

19: 

20: 

21: \def \qqquad {\qquad\quad}

22: \def \qqqquad {\qquad\qquad}

23: %-------------------------------------------------------------

24: \def \matrix #1 {\left(\begin{array}{cc} #1 \end{array}\right)}

25: \def \Tr {\mathop{\rm Tr}\nolimits}

26: \def \tr {\mathop{\rm tr}\nolimits}

27: \def \Im {\mathop{\rm Im}\nolimits}

28: \def \Re {\mathop{\rm Re}\nolimits}

29: \def \res{\mathop{\rm res}\nolimits}

30: \def \e  {\mathop{\rm e}\nolimits}

31: \newcommand\lr[1]{{\left({#1}\right)}}

32: \newcommand \widebar [1] {\overline{#1}}

33: \newcommand\bin[2]{\left({#1}\atop{#2}\right)}

34: \newcommand \vev [1] {\langle{#1}\rangle}

35: \newcommand \VEV [1] {\left\langle{#1}\right\rangle}

36: \newcommand \ket [1] {|{#1}\rangle}

37: \newcommand \bra [1] {\langle {#1}|}

38: \newcommand \partder [1] {{\partial \over\partial #1}}

39: \newcommand{\as}{\ifmmode\alpha_{\rm s}\else{$\alpha_{\rm s}$}\fi}

40: \newcommand{\asbar}{\ifmmode\bar{\alpha}_{\rm s}\else{$\bar{\alpha}_{\rm s}$}\fi}

41: \def \CO {{\cal O}}

42: \def \CP {{\cal P}}

43: \def \CT {{\cal T}}

44: \def \CM {{\cal M}}

45: \def \CK {{\cal K}}

46: \def \CH {{\cal H}}

47: \def \CI {{\cal I}}

48: \def \CV {{\cal V}}

49: \def \CJ {{\cal J}}

50: \def \CL {{\cal L}}

51: \def \CR {{\mathcal R}}

52: \def \CD {{\cal D}}

53: 

54: \def \G  {\Gamma}

55: \def \D  {\Delta}

56: \def \VV {\mathbf V}

57: 

58: \newcommand{\half}{ {\textstyle\frac{1}{2}} }

59: 

60: %\font\cmss=cmss12 \font\cmsss=cmss10 at 11pt

61: \def\inbar{\,\vrule height1.5ex width.4pt depth0pt}

62: \def\IC{\relax\hbox{$\inbar\kern-.3em{\rm C}$}}

63: \def\IZ{\relax{\hbox{\cmss Z\kern-.4em Z}}}

64: \def\IR{{\hbox{{\rm I}\kern-.2em\hbox{\rm R}}}}

65: \def\R{{\tiny \IR}}

66: \def\IP{{\hbox{{\rm I}\kern-.2em\hbox{\rm P}}}}

67: \def\II{\hbox{{1}\kern-.25em\hbox{l}}}

68: 

69: \def\numberbysection{\@addtoreset{equation}{section}

70:                      \def\theequation{\thesection\arabic{equation}}}

71: %\numberbysection

72: 

73: \newcommand \mybf[1] {\mbox{\boldmath$\scriptstyle {#1} $}}

74: \newcommand \Mybf[1] {\mbox{\boldmath$ {#1} $}}

75: \newbox\lett\newdimen\lheight\newdimen\lwidth

76: \def\ontop#1#2{\setbox\lett=\hbox{#2}\lheight\ht\lett

77: \multiply\lheight by 12 \divide\lheight by 10\relax%

78: \lwidth\wd\lett \multiply\lwidth by 8 \divide\lwidth by 10\relax #2\kern-\lwidth%

79: \raise\lheight\hbox{{$\scriptstyle #1$}}\kern.1ex}

80: 

81: %%%%%%%%%%%% Figures labels %%%%%%%%%%%%%%%%

82: 

83: 

84: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

85: 

86: 

87: \begin{document}

88: \title[Factorization of the transfer matrices]{Factorization

89: of the transfer matrices for the quantum $s\ell(2)$ spin

90: chains and Baxter equation}

91: \author{S {\'E} Derkachov~$^1$,~~ A N Manashov~$^2$\footnote{ Permanent

92: address:\ Department of Theoretical Physics,  Saint-Petersburg State University,

93: \mbox{St.-Petersburg,} Russia}}

94: 

95: \address{$^1$ St.Petersburg Department of Steklov

96: Mathematical Institute of Russian Academy of Sciences,

97: Fontanka 27, 191023 St.Petersburg, Russia.

98:  }

99: 

100: \address{$^2$ Institute for Theoretical Physics, University of  Regensburg,\\

101: D-93040 Regensburg, Germany}

102: 

103: 

104: 

105: \begin{abstract}

106: It is shown that the transfer matrices of homogeneous $s\ell(2)$ invariant spin chains with

107: generic spin, both closed and open, are factorized

108: into the product of two operators. The latter satisfy the Baxter equation that follows

109: from the structure of the reducible representations of the $s\ell(2)$ algebra.

110: \end{abstract}

111: \maketitle

112: 

113: \setcounter{footnote}{0}

114: \section{Introduction}

115: The recent interest to the analysis of  noncompact spin magnets (spin chains with

116: the infinite dimensional Hilbert space at each site) is motivated by the advances in gauge

117: field theories (see for a review~\cite{BBGK,AT}). These models (spin magnets) can be solved

118: with the help of the

119: Algebraic Bethe Ansatz (ABA) method~\cite{FST,LF}.

120: Alternatively, the solution

121: is provided by the method of the Baxter $Q-$operators~\cite{Baxter}.

122: 

123: The Baxter $Q-$operator is known for a large number of 

124: integrable

125: models~\cite{PG,V95,BLZ,SD,Pronko,KSS,RW,open,KM,Korff1}.

126: Nevertheless, a universal method for obtaining of the Baxter

127: operator is absent so far and each model (or class of

128: models) needs a special consideration. The derivation of

129: the Baxter $Q-$operator for the $s\ell(2)$ spin chain

130: models is based on the Pasquier-Gaudin trick, see

131: Refs.~\cite{PG,SD,DKM01}. The generalization of the latter

132: to the spin chains with the higher rank symmetry groups is

133: not quite obvious.

134: 

135: In the present paper we give the alternative derivation of

136: the Baxter equation for the noncompact $XXX$ spin chain

137: models. We shall show that the transfer matrices for the

138: homogeneous spin chain models  factorize into a product of

139: two operators. The factorization holds for all closed

140: $s\ell(2)$ spin chains studied so far~\cite{DKM01,KM,SD2}

141: and can be traced to the factorization of the $\mathcal

142: R-$operator obtained in Ref.~\cite{SD1}. We prove that this

143: property is true for the open spin chain models as well. As

144: the factorization property is established, the Baxter

145: equation for these operators can be deduced in a simple way

146: from the structure of the reducible representations of

147: the $s\ell(2)$ algebra. (See  Ref.~\cite{AF} where similar

148: arguments were applied to the analysis of $q-$deformed spin

149: chain models.) We shall consider the spin chains with the

150: quantum space being the generic lowest weight

151: representation of the $s\ell(2)$ algebra, but the method

152: works  for the principal series representations of the

153: $SL(2,{\mathbb R})$  (\,$SL(2,{\mathbb C})$\,) group as well.

154: Taking into account that, as it was shown in Ref.~\cite{SD1},

155: factorization holds for  the $s\ell(3)$ and

156: $s\ell(2|1)$ invariant $\mathcal R-$operators one can hope

157: that the  approach presented here admits a generalization

158: for the spin chains with the symmetry group of  higher

159: rank.

160: 

161: The paper is organized as follows: In 

162: Section~\ref{prelim} we introduce notations and describe

163: the model. In  Section~\ref{fact} we prove the

164: factorization property for the transfer matrices for both

165: the  closed and open $s\ell(2)$ noncompact spin chain

166: models. In  Section~\ref{BAX} the derivation of the

167: Baxter equation based on the structure of the reducible

168: $s\ell(2)$ representations is given. 

169: Section~\ref{Concl} contains concluding remarks.

170: 

171: 

172: 

173: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

174: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

175: \section{Preliminaries}\label{prelim}

176: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

177: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

178: The basic object in the theory of the lattice integrable

179: systems is a $\mathcal R-$operator. The $\mathcal R-$ operator is a linear operator which

180: depends on a spectral parameter $u$ and  acts on the tensor product of

181: two $s\ell(2)$ modules (representations of the $s\ell(2)$

182: algebra). It satisfies the Yang-Baxter relation (YBR)

183: \begin{equation}\label{YB}

184: \CR_{12}(u)\CR_{13}(u+v)\CR_{23}(v)=\CR_{23}(v)\CR_{13}(u+v)\CR_{12}(u)\,.

185: \end{equation}

186: The operators acts on the tensor product

187: $\VV_1\otimes\VV_2\otimes\VV_3$, and, as usual, indices $ik$

188: indicate that the operator $\CR_{ik}$ acts nontrivially on

189: the tensor product $\VV_i\otimes\VV_k$. We shall consider

190: the $s\ell(2)$ invariant solutions of the YBR.

191: 

192: The $s\ell(2)$ algebra has three generators $S_+,\,S_-$ and

193: $S_0$ which satisfy the well known commutation relations

194: \begin{equation}\label{com}

195: [S_0,S_\pm]=\pm S_\pm,\ \ \ \ \ [S_+,S_-]=2S_0\,.

196: \end{equation}

197: The lowest weight representation of $s\ell(2)$ algebra,

198: $D_s$, is uniquely determined by the complex number ( spin

199: ) $s$. The generators can be realized as the differential

200: operators

201: \begin{equation}\label{gen}

202: S_-=-\partial_z,\ \ \ S_+=z^2\partial_z+2sz,\ \ \ S_0=s\partial_z+s

203: \end{equation}

204: acting on the linear space ${\mathbf V}_s={\mathbb C}[z]$

205: (the space of polynomials of arbitrary degree of a

206: complex variable $z$). For a given $s$ the

207: representation~\eref{gen} is irreducible unless $s$ is a

208: negative (half)integer. If $s=-n$, $n=0,1/2,1,...$ the

209: space ${\mathbf V}_s$ contains a finite dimensional

210: invariant subspace, ${V}_{n}$, the space of polynomials of

211: degree less or equal to $2n$, ($\dim V_{n}=2n+1$). The

212: representation induced on the factor space $\VV_{-n}/V_{n}$

213: is equivalent to the representation $D_{s'}$ with spin

214: $s'=1+n$.  The operator $\mathrm{A}$ which intertwines the

215: representations $D_{-n}$ and $D_{n+1}$,

216: ($\mathrm{A}\,D_{-n}=D_{1+n} \, \mathrm{A}$) ,  is defined

217: by the commutation relations

218: $$

219: \mathrm{A}\,S_\alpha^{(s=-n)} = S_\alpha^{(s=n+1)} \, \mathrm{A}\

220: $$

221: and has the form $\mathrm{A} =\partial_z^{2n+1}$.

222: 

223: For the real $s>1/2$ there exists the invariant scalar product

224: $(\cdot,\cdot)_s$ on the space $V_s$,

225: \begin{equation}\label{sc}

226: (\psi_1,\psi_2)_s=\int {\mathcal D}_sz\,\overline{\psi_1(z)}\,{\psi_2}(z)\,,

227: \end{equation}

228: where

229: \begin{equation}\label{DS}

230: \int{\mathcal D}_s z \,\varphi(z,\bar z)\equiv \frac{2s-1}{\pi}\,\ \int

231: _{|z|<1}\,d^2z \,(1-|z|^2)^{2s-2}\, \varphi(z,\bar z) \,.

232: \end{equation}

233: 

234: The operator $S_0$ is  hermitian with respect to the

235: scalar product~(\ref{sc}), while $S_-^\dagger=-S_+$. For

236: complex $s$ the integral~\eref{sc} defines the invariant

237: bilinear form on the  tensor product $\VV_{s^*}\otimes

238: \VV_s$. The unit operator (reproducing kernel) has the form

239: \begin{equation}

240: {\mathbb K}_s(z,w)=(1-z\bar w)^{-2s}\,.

241: \end{equation}

242: The identity

243: \begin{equation}\label{KK}

244: \psi(z)=\int {\mathcal D}_s w\, {\mathbb K}_s(z,w)\,\psi(w)\,,

245: \end{equation}

246: where $\psi(w)$ is the function analytic in the unit circle

247: holds for  complex $s$ such that $\Re s>1/2$; for all

248: other spins it should be understood as an analytic continuation

249: in $s$. \vskip 0.5cm

250: 

251: The $s\ell(2)$ invariant $\mathcal R-$operator acting on the tensor product of two spaces

252: $\VV_{s_1}\otimes \VV_{s_2}$ has the form~\cite{PKS,LF,ES91}

253: \begin{equation}\label{Rop}

254: {\mathcal R}_{12}(u) =(-1)^{\mathbb

255: J-s_1-s_2}\frac{\Gamma(s_1+s_2+iu)}{\Gamma(s_1+s_2-iu)}

256: \frac{\Gamma({\mathbb J}-iu)}{\Gamma({\mathbb J}+iu)}\,,

257: \end{equation}

258: where the operator of the conformal spin ${\mathbb J}$ is related to the two-particle Casimir

259: operator in the standard manner

260: \begin{equation}

261: {\mathbb J}({\mathbb J}-1)=(\vec{S}_1+\vec{S}_2)^2\,.

262: \end{equation}

263: It was shown in the Ref.~\cite{SD1}  that the ${\mathcal R}-$operator~\eref{Rop} can be

264: represented in the factorized form

265: \begin{equation}\label{R}

266: {\mathcal R}_{12}(u)=P_{12} \,{\mathcal

267: R}_{12}^+(\alpha)\,{\mathcal R}_{12}^-(\beta)=

268: P_{12}\,{\mathcal R}_{12}^-(\beta)\,{\mathcal

269: R}_{12}^+(\alpha)\,.

270: \end{equation}

271: Here $P_{12}$ is the permutation operator $P_{12}\psi(z_1,z_2)=\psi(z_2,z_1)$,

272: and

273: \begin{equation}\label{ab}

274: \alpha=\frac{s_2-s_1+iu}{2}\,,\   \  \       \ \ \ \ \beta=\frac{s_1-s_2+iu}{2}\,.

275: \end{equation}

276: The operator ${\mathcal R}_{12}^-(\alpha)$ is a

277: $s\ell(2)$ covariant operator, i.e. it maps

278: $$

279: \VV_{s_1}\otimes \VV_{s_2}\to \VV_{s_1-\alpha}\otimes

280: \VV_{s_2+\alpha}

281: $$

282: and has the following form

283: \begin{equation}

284: {\mathcal

285: R}_{12}^-(\alpha)=\frac{\Gamma(2s_1)}{\Gamma(2s_1-2\alpha)}

286: \frac{\Gamma(z_{12}\partial_1+2s_1-2\alpha)}

287: {\Gamma(z_{12}\partial_1+2s_1)}\,,

288: \end{equation}

289: where $z_{12}=z_1-z_2$. Such normalization implies that

290: $R_{12}^-(0)={\mathbb I}$ and $R_{12}^-(\alpha)\cdot 1=1$.

291: The second operator, ${\mathcal R}_{12}^+(\alpha)$,\ 

292: (${\mathcal R}_{12}^+(\alpha):\,\VV_{s_1}\otimes \VV_{s_2}\to \VV_{s_1+\alpha}\otimes

293: \VV_{s_2-\alpha} $) is

294: \begin{equation}

295: {\mathcal R}_{12}^+(\alpha)={\mathcal

296: R}_{21}^-(\alpha)=\frac{\Gamma(2s_2)}{\Gamma(2s_2-2\alpha)}

297: \frac{\Gamma(z_{21}\partial_2+2s_2-2\alpha)}

298: {\Gamma(z_{21}\partial_2+2s_2)}\,.

299: \end{equation}

300: The operators ${\mathcal R}^\pm_{12}(\alpha)$ depend on

301: three parameters -- the spins $s_1,s_2$ and the

302: spectral parameter $\alpha$. The spins are always fixed by

303: the tensor properties of the space $\VV_{s_1}\otimes

304: \VV_{s_2}$ the operators act on, therefore we shall

305: display the dependence of the operators on the spectral

306: parameter only. The action of the $\mathcal

307: R-$operator~\eref{R} on the space $\VV_{s_1}\otimes

308: \VV_{s_2}$ results in the following chain of 

309: transformations

310: $$

311: \VV_{s_1}\otimes \VV_{s_2}{\overset{{\mathcal

312: R_{12}^-(\beta)}}{\longrightarrow}}

313: \VV_{(s_1+s_2-iu)/2}\otimes \VV_{(s_1+s_2+iu)/2}

314: {\overset{{\mathcal

315: R_{12}^+(\alpha)}}{\longrightarrow}}\VV_{s_2}\otimes

316: \VV_{s_1}

317: {\overset{P_{12}}{\longrightarrow}}\VV_{s_1}\otimes

318: \VV_{s_2}\,.

319: $$

320:  In the next section we shall represent the operators

321: ${\mathcal R}_{12}^\pm(\alpha)$ as  integral operators

322: and prove the factorization of the transfer

323: matrices~\cite{FST,ES88}

324: \begin{eqnarray}

325: \label{Tc}

326: {\mathbf T}^{\mathrm{ cl}}_{s_0}&=&\tr_{s_0}{\mathcal R}_{10}(u)

327: \ldots {\mathcal R}_{N0}(u)\,,\\

328: {\mathbf T}^{\mathrm{ op}}_{s_0}&=&\tr_{s_0}{\mathcal

329: R}_{10}(u)\ldots {\mathcal R}_{N0}(u) {\mathcal

330: R}_{N0}^{-1}(-u) \ldots{\mathcal R}_{10}^{-1}(-u)

331: \end{eqnarray}

332: for the homogeneous closed and open $s\ell(2)$ invariant

333: spin chains. The $\mathcal R-$operator obeys the relation

334: ${\mathcal R}_{12}^{-1}(u)={\mathcal R}_{12}(-u)$ so that

335: we shall use the following expression for the ${\mathbf

336: T}^{\mathrm{ op}}_{s_0}$

337: \begin{eqnarray}

338: \label{To}{\mathbf T}^{\mathrm{

339: op}}_{s_0}&=&\tr_{s_0}{\mathcal R}_{10}(u)\ldots {\mathcal

340: R}_{N0}(u) {\mathcal R}_{N0}(u) \ldots{\mathcal R}_{10}(u).

341: \end{eqnarray}

342: 

343: 

344: 

345: 

346: 

347: 

348: 

349: 

350: \setcounter{equation}{0}

351: \section{Factorization}\label{fact}

352: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

353: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

354: \begin{figure}[t]

355: \psfrag{2j1}[cc][cc]{${2s_1-2\alpha}$}

356: \psfrag{2j2}[cc][cc]{${2s_2}$}

357: \psfrag{2a}[lc][lc]{${2\balpha}$}

358: \psfrag{z1}[cc][cc][1.2]{$\bar w_1$}

359: \psfrag{z2}[cc][cc][1.2]{$\bar w_2$}

360: \psfrag{w1}[cc][cc][1.2]{$ z_1$} \psfrag{w2}[cc][cc][1.2]{$

361: z_2$}

362: %\psfrag{s1}[cc][cc]{$2s_1$}

363: %\psfrag{s2}[cc][cc]{$2s_2$}

364: %\centerline{\epsfxsize5.0cm\epsfbox{U.eps}}

365: \centerline{\includegraphics[scale=0.7]{U.eps}}

366: %\vspace*{0.5cm}

367: \caption[]{Graphical representation of the ${\mathcal

368: R}_{12}^-(\alpha)-$operator. The arrow with the index

369: $\alpha$ directed from $\bar w$ to $z$ denotes the factor

370: $(1-z\bar w)^{-\alpha}$.} \label{U}

371: \end{figure}

372: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

373: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

374: We find convenient to represent all operators in question

375: as integral operators. Let us write the action of the

376: operator ${\mathcal A}$ on the function $\psi\in

377: \prod_{k=1}^N\otimes\VV_{s_k}$ in the following form

378: \begin{equation}\label{A}

379: [{\mathcal A}\psi](\Mybf{z})=\int \prod_{k=1}^N{\mathcal D}_{s_k} w_k\,

380: A(\Mybf{z}|\overline{\Mybf{w}})\,

381: \psi(\Mybf{w})\,,

382: \end{equation}

383: where $\Mybf{z}=(z_1,\ldots,z_N)$. It follows from the

384: definition~\eref{A} and  Eq.~\eref{KK} that the kernel of

385: the operator $\mathcal A$ can be obtained as follows

386: \begin{equation}\label{R-K}

387: A(\Mybf{z}|\overline{\Mybf{w}})={\mathcal

388: A}\cdot\prod_{k=1}^N (1-z_k\bar w_k)^{-2s_k}\,.

389: \end{equation}

390: Here  the operator ${\mathcal A}$ on the r.h.s of

391: Eq.~\eref{R-K} acts on  $\Mybf{z}$-variables.

392: % and

393: %$\overline{\Mybf{w}}$-variables are exterior parameters.}

394: 

395: It is easy to show that the kernel of the operator ${\mathcal

396: R}_{12}^-(\alpha)$ takes the following form

397: \begin{equation}\label{Rint}

398: \fl {R}_{12}^-(\alpha)(z_1,z_2|\bar w_1, \bar

399: w_2)=(1-z_1\bar w_1)^{-2s_1+2\alpha} (1-z_2\bar

400: w_1)^{-2\alpha} (1-z_2\bar w_2)^{-2s_2}\,.

401: \end{equation}

402: It is convenient to represent the kernel

403: $R_{12}^-(\alpha)(\Mybf{z}|\overline{\Mybf{w}})$ in the

404: graphical form. Namely, let us denote the reproducing

405: kernel ${\mathbb K}_{\alpha}(z,w)= (1-z\bar w)^{-2\alpha}$

406: by the arrow with the index $2\alpha$ directed from $w$ to

407: $z$. Then the kernel ${

408: R}_{12}^-(\alpha)(\Mybf{z}|\overline{\Mybf{w}})$

409:  is given by the diagram shown in the Figure~\ref{U}.

410: Similarly,  as follows from Eq.~\eref{R}, the kernel of the

411: ${\mathcal R}_{12}-$operator has the form

412: \begin{eqnarray}\fl

413: R_u(z_1,z_2|\bar w_1,\bar w_2)&=&(1-z_2\bar w_1)^{-2\gamma}

414: \int D_{(s_1+s_2+iu)/2}\zeta \nonumber\\

415: &&(1-z_1\bar\zeta)^{-2s_1}

416: (1-z_2\bar \zeta)^{-2\alpha}

417: (1-\zeta \bar w_1)^{-2\beta}

418: (1-\zeta \bar w_2)^{-2s_2}

419: \end{eqnarray}

420: where $\alpha$ and $\beta$ are defined in  Eq.~\eref{ab} and

421: $

422: \gamma=(s_1+s_2-iu)/2.

423: $

424: 

425: There exists another equivalent representation for the $\mathcal

426: R-$operator which follows from the second equality in

427: Eq.~\eref{R}. Again, it is useful to represent both of them

428: in the graphical form, see Figure~\ref{RR}.

429: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

430: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

431: \begin{figure}[t]

432: \psfrag{A1}[cc][bc]{$c$}%s_1+s_2-iu$}

433: \psfrag{A2}[rc][rc]{$a$}%s_2-s_1+iu$}

434: \psfrag{A3}[lc][lc]{$b$}%s_1-s_2+iu$}

435: \psfrag{z1}[cc][cc]{$z_1$}

436: \psfrag{z2}[cc][cc]{$z_2$}

437: \psfrag{w1}[cc][cc]{$\bar w_1$}

438: \psfrag{w2}[cc][cc]{$\bar w_2$}

439: \psfrag{s1}[cc][cc]{$2s_1$}

440: \psfrag{s2}[cc][cc]{$2s_2$}

441: \psfrag{N}[cc][cc]{$=$}

442: \centerline{\includegraphics[scale=0.66]{Rmat.eps}}

443: \caption[]{The two equivalent graphical representation of the kernel of the $\mathcal R-$

444: operator. The black dot denotes the integration vertex with the measure

445: corresponding to the spin $(s_1+s_2+iu)/2$ and the indicies

446: $a=2\alpha=s_2-s_1+iu$, $b=2\beta=s_1-s_2+iu$, $c=2\gamma=s_1+s_2-iu$.}

447: \label{RR}

448: \end{figure}

449: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

450: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

451: The identity depicted in  Figure~\ref{RR} (permutation relation) can be considered as

452: an integral identity between the reproducing kernels. It will be quite useful in the

453: subsequent analysis.

454: 

455: 

456: Let us summarize the properties of the ${\mathcal R}^--$operators.

457: One easily checks that

458: \begin{eqnarray}\label{comm}

459: {\mathcal R}^\pm_{12}(\alpha) \,{\mathcal R}^\pm_{12}(\beta)={\mathcal

460: R}^\pm_{12}(\alpha+\beta)\,,

461: \\[2mm]

462: \label{perm}

463: {\mathcal R}^+_{12}(s_2-s_1+\alpha)\,{\mathcal R}^-_{12}(\alpha)=

464: {\mathcal R}^-_{12}(\alpha)\,{\mathcal R}^+_{12}(s_2-s_1+\alpha)\,,

465: \\[2mm]

466: \label{F3}

467: {\mathcal R}^\pm_{12}(\alpha)\,{\mathcal R}^\pm_{23}(\alpha+\beta)\,{\mathcal R}^\pm_{12}(\beta)=

468: {\mathcal R}^\pm_{23}(\beta)\,{\mathcal R}^\pm_{12}(\alpha+\beta)\,

469: {\mathcal R}^\pm_{23}(\alpha)\,.

470: \end{eqnarray}

471: The first equality follows from  Eq.~\eref{Rint} and from the

472: property of the reproducing kernel. The second one is the

473: consequence of  Eq.~\eref{R}. The last one arises as the

474: selfconsistency relation of the defining equations for the

475: ${\mathcal R}^\pm$ operators~\cite{SD2} and can be 

476: checked directly by making use of the permutation relation.

477: 

478: 

479: 

480: 

481: Let us introduce operators ${\mathcal

482: L}^{\pm}_{12}(\alpha)= P_{12}{\mathcal

483: R}_{12}^{\pm}(\alpha)$. It is straightforward to check that

484: these operators satisfy the relation

485: \begin{equation}\label{LB}

486: {\mathcal L}_{12}^\pm(\alpha){\mathcal L}_{13}^\pm(\alpha+\beta)

487: {\mathcal L}_{23}^\pm(\beta)={\mathcal L}_{23}^\pm(\beta){\mathcal L}_{13}^\pm(\alpha+\beta)

488: {\mathcal L}_{12}^\pm(\alpha)\,.

489: \end{equation}

490: Equation~\eref{LB} has the form of the Yang-Baxter

491: relation, but in difference to  the $\mathcal R-$operator,

492: the operators ${\mathcal L}_{12}^\pm(\alpha)$ map the space

493: $\VV_{s_1}\otimes \VV_{s_2}\mapsto

494: \VV_{s_2\pm\alpha}\otimes \VV_{s_1\mp\alpha}$. However, for

495: the special values of the spectral parameter,

496: $\alpha_\pm=\pm(s_2-s_1)$, the operators ${\mathcal

497: L}_{12}^\pm(\alpha_\pm)$ coincide with the $\mathcal

498: R-$operator for the special values of the spectral

499: parameter,

500: \begin{equation}\label{LR}

501: {\mathcal L}_{12}^{\pm}(\pm(s_2-s_1))={\mathcal R}_{12}(\mp

502: i(s_2-s_1))\,,

503: \end{equation}

504: and play an important role in the subsequent construction.

505: 

506: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

507: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

508: \begin{figure}[t]

509: \psfrag{w2}[cc][cc]{$\bar w_2$} \psfrag{w3}[cc][cc]{$\bar

510: w_3$} \psfrag{w5}[cc][cc]{$\bar w_1$}

511: \psfrag{w4}[cc][cc]{$\bar w_{N}$}

512: \psfrag{z1}[cc][cc]{$z_1$} \psfrag{z2}[cc][cc]{$z_2$}

513: \psfrag{z5}[cc][cc]{$z_3$} \psfrag{z4}[cc][cc]{$z_1$}

514: \psfrag{z3}[cc][cc]{$z_{N}$}

515: \psfrag{s1}[cr][cc]{$\balpha_u$}

516: \psfrag{s2}[cr][cc]{$\bbeta_u$}

517: \centerline{\includegraphics[scale=0.75]{lambda.eps}}

518: \caption[]{The graphical representation of  the kernel of

519: $Q(u)$ operator for the closed spin chain; $\alpha_u=s-iu$

520: and $\beta_u=s+iu$.} \label{Qcp}

521: \end{figure}

522: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

523: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

524: \begin{figure}[t]

525: \psfrag{w1}[cc][cc]{$\bar w_1$}

526: \psfrag{w2}[cc][cc]{$\bar w_2$}

527: \psfrag{w3}[cc][cc]{$\bar w_3$}

528: \psfrag{w4}[cc][cc]{$\bar w_N$}

529: 

530: \psfrag{z1}[cc][cc]{$z_1$}

531: \psfrag{z2}[cc][cc]{$z_2$}

532: \psfrag{z3}[cc][cc]{$z_3$}

533: 

534: \psfrag{zn}[cc][cc]{$z_N $}

535: 

536: \psfrag{a}[cc][cc]{$\bgamma_u$}

537: \psfrag{s}[cc][rc]{$2s$}

538: \centerline{\includegraphics[scale=1.0]{closed-2.eps}}

539: %\vspace*{0.5cm}

540: \caption[]{The kernel of the ${\widetilde Q}(u)$ operator

541: for the closed spin chain. All horizontal lines carry the

542: index

543:  $\gamma_u=iu-s$, while the vertical ones have the index $2s$.

544:  The black dots denote the integration

545: with the measure corresponding to the  spin $s'=(s+iu)/2$. }

546: \label{Qcm}

547: \end{figure}

548: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

549: 

550: In what follows we show that the transfer matrix for the

551: closed homogeneous spin chain~\eref{To} can be represented

552: in the factorized form

553: \begin{equation}\label{Facc}\fl

554: {\mathbf T}_{s_0}^{\mathrm{cl}}(u)=Q(u+is_0)\,\widetilde Q(u-is_0)=

555: \widetilde Q(u-is_0)\,Q(u+is_0)\,,

556: \end{equation}

557: where $s\ell(2)$ invariant $Q$-operators are given by the

558: traces of ${\mathcal L}_{12}^\pm$ operators. Namely, we get

559: \begin{eqnarray}\label{QL}

560: Q(u)=\tr_{s_0} {\mathcal L}_{10}^-({s-s_0})\ldots

561: {\mathcal L}_{N0}^-(s-s_0)\bigl|_{s_0=(s-iu)/2}\,,\\[2mm]

562: \widetilde Q(u)={\mathcal P}\tr_{s_0} {\mathcal

563: L}_{10}^+(s_0-s)\ldots {\mathcal

564: L}_{N0}^+(s_0-s)\bigl|_{s_0=(s+iu)/2}\,,

565: \end{eqnarray}

566: where ${\mathcal P}$ is the cyclic permutation operator,

567: ${\mathcal P}\psi(z_1,z_2,\ldots,z_N)=\psi(z_2,z_3,\ldots,z_1)$.

568: Taking into account \eref{LR} we conclude that the operators $Q(u)$ and $\widetilde

569: Q(u)$ coincide with the transfer matrices

570: \begin{eqnarray}

571: \label{QpT}

572: Q(u)={\mathbf T}_{(s-iu)/2}\left(\frac{u-is}{2}\right)\,,\ \ \ \ \\[2mm]

573: \label{QmT}

574: \widetilde Q(u)={\mathcal P}{\mathbf T}_{(s+iu)/2}\left(\frac{u+is}{2}\right)\,.

575: \end{eqnarray}

576: The commutativity of the $Q$, $\widetilde Q$ operators,

577:  $[Q(u) , Q(v)]=[\widetilde Q(u) , \widetilde Q(v)]=

578:  [Q(u) , \widetilde Q(v)] = 0$,

579: follows immediately from the commutativity of the transfer

580: matrices, (we remind that ${\mathcal P}^{-1}={\mathbf

581: T}^{s}_{s}(0)$.)

582: 

583: Making use of Eqs.~\eref{QpT},~\eref{QmT} one can represent

584: Eq.~\eref{Facc} in the following form

585: \begin{equation}\fl

586: {\mathbf T}_{s_0}(u)={\mathbf T}_{(s+s_0-iu)/2}\left(\frac{u-i(s-s_0)}{2}\right)\,

587: {\mathcal P}\,{\mathbf T}_{(s+s_0+iu)/2}\left(\frac{u+i(s-s_0)}{2}\right)\,.

588: \end{equation}

589: 

590: 

591: To prove the factorization property  we shall show that the

592: integral kernels of the operators on the l.h.s and r.h.s of

593: Eq.~\eref{Facc} coincide. To this end let us   represent

594: the kernel of the  operators under consideration in the

595: graphical form. The diagrammatical representation of the

596: kernels for the operators $Q(u)$ and $\widetilde Q(u)$ are

597: shown in Figures~\ref{Qcp} and \ref{Qcm}, respectively. In

598: its turn, the integral kernel for the transfer matrix  is

599: shown, in two equivalent forms, in Figure~\ref{trcl}.

600: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

601: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

602: \begin{figure}[t]

603: \psfrag{w1}[cc][cc]{$\bar w_1$}

604: \psfrag{w2}[cc][cc]{$\bar w_2$}

605: \psfrag{w3}[cc][cc]{$\bar w_3$}

606: \psfrag{wn}[cc][cc]{$\bar w_N$}

607: 

608: 

609: \psfrag{z1}[cc][cc]{$z_1$}

610: \psfrag{z2}[cc][cc]{$z_2$}

611: \psfrag{z3}[cc][cc]{$z_3$}

612: 

613: \psfrag{zn}[cc][cc]{$z_N $}

614: \psfrag{eq}[cc][bc]{$=$}

615: \psfrag{a}[cc][cc]{$\alpha_u$}

616: \psfrag{b}[cc][cc]{$\beta_u$}

617: \psfrag{c}[cc][cc]{$\gamma_u$}

618: \psfrag{s}[cc][cc][0.8]{$2s$}

619: \centerline{\includegraphics[scale=0.62]{trcl.eps}}

620: %\vspace*{0.5cm}

621: \caption[]{The graphical representation of the transfer matrix for the closed spin chain.

622: The indices $\alpha_u=s-i(u+is_0)$, $\beta_u=s+i(u+is_0)$. The indices

623: of the vertical lines are equal to $2s$, and those of the horizontal are equal to

624: $\gamma_u=i(u-is_0)-s$. The black dots denotes the integration vertices

625: corresponding to the  spin $s'=(s+i(u-is_0))/2$.}

626: \label{trcl}

627: \end{figure}

628: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

629: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

630: Drawing the diagram for the product $Q(u+is_0) \widetilde Q(u-is_0)$

631: (or $\widetilde Q(u-is_0) Q(u+is_0)$) one notices that the measure of  integration

632: in the intermediate triple vertices corresponds to the spin $s$. Since the

633: vertical lines attached to this vertex correspond to the reproducing kernel with the spin $s$,

634: one can carry out the integration using

635: the property~\eref{KK} and  find that the resulting  diagram coincides with

636: the diagram for the kernel of the transfer matrix.

637: Thus the property of the factorization for  the

638: homogenous spin chain is established.

639: 

640: For completeness, we write down the analytic expressions for the kernels of the

641: $Q-$operators,

642: \begin{eqnarray}\fl\label{Qucp}

643: Q(u)(\Mybf{z}|\Mybf{w})=\prod_{k=1}^N (1-z_k\bar w_k)^{-s-iu}(1-z_k\bar w_{k+1})^{-s+iu}\,,

644: \\[2mm]

645: \fl\label{Qucm} \widetilde Q(u)(\Mybf{z}|\Mybf{w})=

646: \prod_{k=1}^{N}

647: \int {\mathcal D}_{s'}\zeta_k

648: \,(1-\zeta_k\bar\zeta_{k+1})^{s-iu}(1-z_k\bar\zeta_k)^{-2s}\,

649: (1-\zeta_k\bar w_k)^{-2s}\,,

650: \end{eqnarray}

651: where $s'=s+iu/2$, and $w_{N+1}\equiv w_1$ and so on. The expression~\eref{Qucp} coincides

652: with the expression for the Baxter operator obtained in~Ref.~\cite{SD}.

653: 

654: 

655: 

656: 

657: \vskip 0.5cm

658: 

659: Let us consider now the  homogeneous $s\ell(2)$ invariant

660: open spin chain. The transfer matrix for the open spin

661: chain,~\eref{To}, can also  be represented in the factorized

662: form, namely

663: \begin{equation}\fl\label{TQQ}

664: {\mathbf T}_{s_0}^{\mathrm{op}}(u)=g(u)

665: {\mathcal Q}(u+is_0)\,\widetilde {\mathcal Q}(u-is_0)=g(u)

666: \widetilde{\mathcal  Q}(u-is_0)\,{\mathcal Q}(u+is_0)\,,

667: \end{equation}

668: where

669: \begin{equation}\label{gu}

670: g(u)=\frac{s+s_0+iu-1}{2iu-1}\,.

671: \end{equation}

672: The operators ${\mathcal Q}(u)$ and $\widetilde{\mathcal

673: Q}(u)$ have the following form

674: \begin{eqnarray}\label{QQL} \fl

675: {\mathcal Q}(u)=\tr_{s_0} {\mathcal

676: L}_{10}^-({s-s_0})\ldots {\mathcal

677: L}_{N0}^-(s-s_0){\mathcal L}_{N0}^-(s-s_0)\ldots {\mathcal

678: L}_{10}^-({s-s_0})

679: \bigl|_{s_0=\frac{s-iu}{2}}\,,\\[2mm]

680: \fl \widetilde {\mathcal Q}(u)=\tr_{s_0} {\mathcal

681: L}_{10}^+(s_0-s)\ldots {\mathcal L}_{N0}^+(s_0-s){\mathcal

682: L}_{N0}^+(s_0-s)\ldots{\mathcal L}_{10}^+(s_0-s)

683: \bigl|_{s_0=\frac{s+iu}{2}}\,.

684: \end{eqnarray}

685: Again, taking into account  Eq.~\eref{LR} one relates $\mathcal Q-$operators to the transfer

686: matrices for the open spin chain

687: \begin{eqnarray}

688: \label{TpQ}

689: {\mathcal Q}(u)={\mathbf T}_{(s-iu)/2}\left(\frac{u-is}{2}\right)\,,\ \ \ \ \\[2mm]

690: \label{TmQ}

691: \widetilde {\mathcal Q}(u)={\mathbf T}_{(s+iu)/2}\left(\frac{u+is}{2}\right)\,.

692: \end{eqnarray}

693: Thus, similarly to the closed spin chain, one concludes that $\mathcal Q-$operators

694: commute with each other for arbitrary values of the spectral parameters.

695:  %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

696: \begin{figure}[t]

697: \psfrag{z1}[cc][cc]{$z_1$}\psfrag{z2}[cc][cc]{$z_2$}\psfrag{zn}[cc][cc]{$z_N$}

698: \psfrag{w1}[cc][cc]{$\bar w_1$}\psfrag{w2}[cc][cc]{$\bar w_2$}

699: \psfrag{wn}[cc][cc]{$\bar w_N$} \psfrag{dots}[cc][cc]{$\Mybf{\cdots}$}

700: \psfrag{a}[cc][cc]{$\beta_u$} \psfrag{b}[cc][cc]{$\alpha_u$}

701: \psfrag{c}[cc][rc]{$\alpha_u$} \psfrag{d}[cc][rc]{$\beta_u$}

702: \centerline{\includegraphics[scale=0.8]{open_b.eps}} %\vspace*{0.5cm}

703: \caption[]{Diagrammatical representation of the ${\mathcal

704: Q}(u)-$operator for the open spin chain;

705: $\alpha_u=s-iu$ and $\beta_u=s+iu$.}%

706: \label{QO+}%

707: \end{figure}%

708: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

709: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

710: \begin{figure}[t]

711: \psfrag{z1}[cc][cc]{$z_1$}

712: \psfrag{z2}[cc][cc]{$z_2$}

713: \psfrag{zn}[cc][cc]{$z_N$}

714: \psfrag{w1}[cc][cc]{$\bar w_1$}

715: \psfrag{w2}[cc][cc]{$\bar w_2$}

716: \psfrag{wn}[cc][cc]{$\bar w_N$}

717: %\psfrag{dots}[cc][cc]{$\Mybf{\cdots}$}

718: %\psfrag{a}[cc][cc]{$\beta_u$} \psfrag{b}[cc][cc]{$\alpha_u$}

719: \psfrag{c}[cc][cc]{$\gamma_u$} \psfrag{d}[cc][rc]{$\beta_u$}

720: \centerline{\includegraphics[scale=0.8]{open2.eps}} %\vspace*{0.5cm}

721: \caption[]{Diagrammatical representation of the

722: ${\widetilde{\mathcal Q}(u)}-$operator for the open spin

723: chain. Here the all horizontal lines carry the index

724:  $\gamma_u=iu-s$, all vertical lines have index $2s$.

725:  The black dots denote the integration

726: with the measure corresponding to the conformal spin $s'=(s+iu)/2$.

727: }%

728: \label{QO-}%

729: \end{figure}%

730: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

731: 

732: To prove the factorization~\eref{TQQ} we again use the

733: graphical representation for kernels.  For the ${\mathcal

734: Q(u)}$ and $\widetilde {\mathcal Q}(u)$ operators they are

735: shown in Figure~\ref{QO+} and~\ref{QO-}, respectively.  The

736: diagrammatical representation for the kernel of the

737: transfer matrix for the open spin chain is shown in

738: Figure~\ref{trop}.  In order to derive this representation one starts with the

739: definition~\eref{To} and uses the graphical representation

740: for the kernel of $\mathcal R-$operator shown on the lhs of

741: Figure~\ref{RR}. Next one should carry out the integration

742: over all intermediate ``quantum'' vertices

743: $\zeta_1,\ldots,\zeta_N$. The integration measure in each

744: vertex is given by the expression~\eref{DS} where $s$ is

745: the  ``spin'' of  the quantum space. The result of the

746: integration is the disappearance of the lines with index

747: $2s$

748: \begin{equation}\fl\label{example}

749: \int D_s'\xi \int{\mathcal  D}\zeta_s \psi(\xi,\bar \xi){\mathbb K}_s(\xi,\zeta)\phi(\zeta)=

750: \int D_s'\xi \psi(\xi,\bar \xi)\phi(\xi)\,

751: \end{equation}

752: attached to these vertices.

753: Finally one can carry out the integration over

754: ``auxiliary space'' vertices. Again, the lines attached to these vertices disappear.

755: The line with the index $2iu$ (on the right part of the diagram)

756: arises due to merging of  two lines with indices $\gamma_u$ and

757: $\beta_u$, $(1-z\bar w)^{-\gamma_u}(1-z\bar w)^{-\beta_u}=(1-z\bar w)^{-2iu}$.

758: 

759: To explain the appearance of the factor $g(u)$ and the integration vertex with spin

760: $s''=iu$, we notice that after the integration one line with index $\alpha_u$

761: becomes attached to the vertex with the

762: spin $s'=s+i(u-is_0))$ by both ends. Noticing that

763: $$

764: \int D_{s'}\xi (1-\xi\bar \xi)^{-\alpha_u}\ldots=g(u)\int D_{iu}\xi\ldots

765: $$

766: one obtains finally  the diagrammatic representation  for

767: the kernel of the transfer matrix shown in

768: Figure~\ref{trop}.

769: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

770: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

771: \begin{figure}[t]

772: \psfrag{w1}[cc][cc]{$\bar w_1$}

773: \psfrag{w2}[cc][cc]{$\bar w_2$}

774: \psfrag{w3}[cc][cc]{$\bar w_3$}

775: \psfrag{wn}[cc][cc]{$\bar w_N$}

776: 

777: 

778: \psfrag{z1}[cc][cc]{$z_1$}

779: \psfrag{z2}[cc][cc]{$z_2$}

780: \psfrag{z3}[cc][cc]{$z_3$}

781: \psfrag{gu}[cc][cc]{$g(u)\times$}

782: 

783: 

784: \psfrag{zn}[cc][cc]{$z_N $}

785: \psfrag{2u}[cc][bc][0.9]{$2iu$}

786: \psfrag{a}[cc][cc]{$\alpha_u$}

787: \psfrag{b}[cc][cc]{$\beta_u$}

788: \psfrag{c}[cc][cc]{$\gamma_u$}

789: \psfrag{2s}[cc][cc][0.8]{$2s$}

790: \centerline{\includegraphics[scale=0.68]{trop.eps}}

791: %\vspace*{0.5cm}

792: \caption[]{The graphical representation of the transfer matrix for the open spin chain.

793: Here $\alpha_u=s-i(u+is_0)$, $\beta_u=s+i(u+is_0)$, $\gamma_u=i(u-is_0)-s$,

794: The black dots denotes the integration vertices

795: corresponding to the  spin $s'=(s+i(u-is_0))/2$ and white dot denotes the integration

796: vertex corresponding to the spin $s''=iu$. The prefactor $g(u)$ is

797: given by  Eq.~\eref{gu}.

798: \label{trop}}

799: \end{figure}

800: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

801: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

802: 

803: Now we have to show that the diagram for the transfer

804: matrix can be transformed to the diagram for the product of

805: the operators ${\mathcal Q}(u+is_0)\widetilde {\mathcal

806: Q}(u-is_0)$. The first transformation is the insertion of

807: the reproducing kernels into the diagram in Fig.~\ref{trop}

808: as shown in Figure~\ref{transf-1}. This operation does not

809: change the kernel, since after the integration

810: over the new vertices one reproduces the initial

811: expression. The next transformation is the following. Let

812: us consider the subdiagram formed by the four lines (which have

813: indices $2iu, \alpha_u,\gamma_u, 2s$) attached to the right

814: (black) vertex in the middle line of the transformed

815: diagram, and the line with index $\beta_u$ which connects

816: the lines with indices $\alpha_u$ and $\gamma_u$. It can be

817: checked that the indices satisfy the conditions

818: $\alpha_u+2iu=2s'=\gamma_u+2s$ and

819: $\beta_u=2s'-\alpha_u-\gamma_u$. It allows one to use the

820: permutation relation shown in Figure~\ref{RR}. After the 

821: transformation, the line with the index $\beta_u$ changes its

822: position and will connect the endpoints of the other pair

823: of lines. In addition, the indices of the lines in the new

824: diagrams have to be changed, namely one should interchange

825: $\alpha_u$ and $2s$ (~$\alpha_u\leftrightarrow 2s$~) and

826: $\gamma_u$ and $2iu$ (~$\gamma_u\leftrightarrow 2iu$~).

827: 

828: Next one notices that the subdiagram formed by the lines

829: attached to the next vertex has exactly the same form as

830: the one considered just now. Therefore one can repeat

831: this  transformation successively. As the result all lines with index

832: $\beta_u$ in the upper part of the diagram change their

833: positions, and one has also to interchange the indices in

834: the way described above, namely, 

835: $\alpha_u\leftrightarrow 2s$. Further, since  the

836: interchange $2iu\leftrightarrow\gamma_u$ occurs twice for

837: the all lines except for the first and the last one in this

838: chain,  the whole effect will be that the line attached to

839: the leftmost vertex (white blob) in the figure will get the

840: index $2iu$, while all other ``horizontal'' lines will have

841: the same index, $\gamma_u$.

842: 

843: It is important  that after this series of 

844: transformations only three lines will be attached to the

845: leftmost vertex (white blob in the Figure~\ref{transf-1}).

846: The integration measure in this vertex corresponds to the

847: conformal spin $s''=iu$. Since the incoming arrow has 

848: index $2iu$,

849:  and two other arrows come out of the vertex, one can integrate

850: over this vertex. After the integration 

851: the line with index $2iu$ disappears (see

852: Eq.~\eref{example}) and one can easily check that the

853: resulting diagram has the form of the integral kernel for

854: the operator

855:  ${\mathcal Q}(u+is_0)\widetilde {\mathcal Q}(u-is_0)$.

856: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

857: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

858: \begin{figure}[t]

859: \psfrag{w1}[cc][cc]{$\bar w_1$}

860: \psfrag{w2}[cc][cc]{$\bar w_2$}

861: \psfrag{w3}[cc][cc]{$\bar w_3$}

862: \psfrag{wn}[cc][cc]{$\bar w_N$}

863: 

864: 

865: \psfrag{z1}[cc][cc]{$z_1$}

866: \psfrag{z2}[cc][cc]{$z_2$}

867: \psfrag{z3}[cc][cc]{$z_3$}

868: \psfrag{gu}[cc][cc]{}

869: 

870: 

871: \psfrag{zn}[cc][cc]{$z_N $}

872: \psfrag{2u}[cc][bc][0.8]{$2iu$}

873: \psfrag{a}[cc][cc][0.9]{$\alpha_u$}

874: \psfrag{b}[cc][cc][0.9]{$\beta_u$}

875: \psfrag{c}[cc][cc][0.9]{$\gamma_u$}

876: \psfrag{2s}[cc][cc][0.8]{$2s$}

877: \centerline{\includegraphics[scale=0.63]{transf-1.eps}}

878: %\vspace*{0.5cm}

879: \caption[]{The diagram for the kernel of the transfer matrix for the open spin chain.

880: All notations the same as in Figure \ref{trop}. The gray dots denote the integration

881: vertices corresponding to the  spin $s$.

882: \label{transf-1}}

883: \end{figure}

884: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

885: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

886: 

887: \setcounter{equation}{0}

888: \section{Baxter equation}\label{BAX}

889: In this section we study the relation between $Q-$operators and the transfer matrices

890: over finite dimensional auxiliary spaces.

891: For a negative (half)integer values of the spin of auxiliary space, $s_0=-n$,

892: the representation space $\VV_{s_0=-n}$ contains an invariant subspace $V_{n}$.

893: Thus, the subspace $\VV_s\otimes V_{n}$ is an invariant subspace of the

894: ${\mathcal R}_{ss_0}-$operator. It has, therefore, the triangular form

895: \begin{equation}\label{triang}

896: {\mathcal R}_{ss_0}(u)=\left(\begin{array}{cc}  r_{ss_0}(u)& *\\

897: 0&\widetilde{\mathcal R}_{ss_0}(u)

898: \end{array}

899: \right)\,,

900: \end{equation}

901: where $r_{ss_0}(u)$ is the restriction of the operator

902: ${\mathcal R}_{ss_0}$ to the subspace $\VV_s\otimes V_{n}$.

903: The operator $\widetilde{\mathcal R}_{ss_0}(u)$ acts on the

904: space $\VV_s\otimes \VV_{s_0}/V_{n}\sim\VV_s\otimes

905: \VV_{1+n/2}$ and satisfies the YB relation~\eref{YB}.

906: Therefore it has to be proportional to ${\mathcal

907: R}_{s,s'_0}(u)$ with the spin $s'_0=1-s_0=1+n$

908: \begin{equation}\label{RtoR}

909: \widetilde {\mathcal R}_{s,s_0=-n}(u)=f_n(u){\mathcal

910: R}_{s,s_0=1+n}(u)\,.

911: \end{equation}

912: The normalization coefficient

913: \begin{equation}

914: f_n(u)=(-1)^{2n+1}\frac{\Gamma(s+1+n-iu)}{\Gamma(s-n-iu)}

915: \frac{\Gamma(s-n+iu)}{\Gamma(s+1+n+iu)}\,

916: \end{equation}

917: can be found by comparing  the eigenvalues of

918: $\widetilde{\mathcal R}_{ss_0}(u)$ and ${\mathcal

919: R}_{s,s'_0}(u)$ on the eigenstates,

920: $\psi_k(z_1,z_2)=(z_1-z_2)^k$, or by using  the

921: intertwining relation for the $\mathcal{R}^\pm$-operators. The latter takes the form

922: \begin{eqnarray}

923: \frac{1}{z_{01}^{2n+1}}\cdot

924: \frac{\Gamma(z_{01}\dd_0+2(s_0-\alpha))}

925: {\Gamma(z_{01}\dd_0+2s_0)} &=&

926: \frac{\Gamma(z_{01}\dd_0+2(s'_0-\alpha'))}

927: {\Gamma(z_{01}\dd_0+2s'_0)}\dd_0^{2n+1} \ ,\

928: \\[2mm]

929: \dd_1^{2n+1}\frac{\Gamma(z_{10}\dd_1+2s_0)}

930: {\Gamma(z_{10}\dd_1+2(s_0-\alpha))} &=&

931: \frac{\Gamma(z_{10}\dd_1+2s'_0)}

932: {\Gamma(z_{10}\dd_1+2(s'_0-\alpha'))}\frac{1}{z_{10}^{2n+1}}

933: \ ,

934: \end{eqnarray}

935: where $ s_0 =-n$, $s'_0=1+n$, $\alpha'=\alpha+n+1/2$, $n$ being half-integer.

936: 

937: 

938: 

939: Equations~\eref{triang},~\eref{RtoR} imply the following

940: relation for the transfer matrices of the closed spin

941: chain

942: \begin{equation}\label{TTc}

943: {\mathbf T}_{s_0=-n}(u)=t_{n}(u)+(f_n(u))^N {\mathbf T}_{s_0=1+n}(u)\,,

944: \end{equation}

945: where $t_{n}(u)$ is the transfer matrix with finite

946: dimensional auxiliary space,  \mbox{$t_{n}(u)=\tr_{V_{n}} r_{10}(u)\ldots r_{N0}(u)$}.

947: Using the factorization property~\eref{Facc} and introducing the notation

948: \begin{equation}\label{QQt}

949: \Delta(u)\,{\widetilde Q}(u)=

950: \widebar Q(u+i)\,,

951: \end{equation}

952: where

953: \begin{equation}

954: \Delta(u)=\left(\frac{\Gamma(1-s-iu)}{\Gamma(1+s-iu)}\right)^N

955: \end{equation}

956: we rewrite Eq.~\eref{TTc} in the form

957: \begin{equation}\fl\label{tQQ}

958: \Delta(u+in)\,t_{n}(u)=Q(u-in)\, \widebar

959: Q(u+i(1+n))-Q(u+i(1+n))\, \widebar Q(u-in)\,.

960: \end{equation}

961: For $n=0$ the transfer matrix $t_{n=0}(u)=1$ and

962: Eq.~\eref{tQQ} reads

963: \begin{equation}%\fl

964: \left(\frac{\Gamma(1-s-iu)}{\Gamma(1+s-iu)}\right)^N=

965: Q(u)\,\widebar Q(u+i)-Q(u+i)\,\widebar Q(u)\,,

966: \end{equation}

967: which is the Wronskian relation between the operators $Q$ and $ \widebar Q$.

968: Further, one can exclude the operator $\widebar Q$ (or $Q$) from eq.~\eref{tQQ}.

969: Indeed, multiplying both sides by $Q(u-im)$, where $m$ is (half)integer such that

970: $m+n$ is integer and $-n\leq m\leq n-1$, after some algebra one finds

971: \begin{eqnarray}\fl\label{3tQ}

972: \tilde t_n(u)\,Q(u-im)&=&\tilde t_{(n+m)/2}\left(u+\frac{i(n-m)}{2}\right)\,Q(u-in)+\nonumber\\[2mm]

973: \fl&&\tilde t_{(n-m-1)/2}\left(u-\frac{i(n+m+1)}{2}\right)\,Q(u+i(n+1))\,,

974: \end{eqnarray}

975: where $\tilde t_n(u)=\Delta(u+in)t_n(u)$. The same equation

976: holds for the $\widebar Q$ operator as well. The first relation

977: ($n=-m=1/2$) among the ones in~\eref{3tQ} is nothing else but

978: the Baxter equation

979: \begin{equation}\label{BE}

980: \tau_N(u)\,Q(u)=(u+is)^N\,Q(u+i)+(u-is)^N\,Q(u-i)\,,

981: \end{equation}

982: where

983: $$

984: \tau_N(u)=\tr L_1(u)\ldots L_N(u)

985: $$

986: and $L(u)$ is the Lax operator

987: $$

988: L(u)=u+i\left(\begin{array}{cc}S_0&S_-\\S_+&-S_0\end{array}\right).

989: $$

990: To derive Baxter equation~\eref{BE} from \eref{3tQ} one put

991: $n=-m=1/2$ and takes into account that $t_0(u)=1$ and

992: $t_{1/2}(u-i/2)=(u-is)^{-N}\tau_N(u)$. The latter relation follows

993: from the relation $L(u)=(u-is)\, r_{s,-1/2}(u-i/2)$ which

994: can easily be  verified by comparison of the

995: eigenvalues~\cite{KRS}. It is evident that the operator

996: $\widebar Q(u)$ satisfies the same equation~\eref{BE}. Thus

997: the operators $Q(u)$ and $\widebar Q(u)$ (and, as

998: consequence, their eigenvalues) represent two independent

999: solutions of the Baxter equation~\eref{BE}. It can be shown

1000: \cite{SD} that the eigenvalues of the operator $Q(u)$ are

1001: polynomials in $u$. The eigenvalues of the second operator

1002: $\widebar Q(u)$ are meromorphic functions of $u$ with poles

1003: of  order $N$ ($N$ is the length of the chain) at the

1004: points $u_k=-i(1-s+k)$, $k=0,1,2\ldots,\infty$.

1005: 

1006: Next, using the triple relation~\eref{3tQ} one can derive

1007: the "fusion" relations for the transfer matrices. Indeed,

1008: multiplying the both sides of  Eq.~\eref{3tQ} by the

1009: operator $\widetilde Q(v)$ and using the factorized

1010: expression for the transfer matrix~\eref{Facc} one obtaines

1011: the relation which involves three $t$ and $T$ transfer matrices.

1012: After the substitution $i(v-u)/2\to s_0$ and $(u+v)/2\to u$

1013: it takes  form

1014: \begin{eqnarray}\fl

1015: \label{fusion-1}

1016: { t}_n(u+is_0)\,T_{s_0-\frac{m}2}\left(u-\frac{im}2\right)&=&

1017: { t}_{\frac{n+m}2}\left(u+is_0+\frac{i(n-m)}2\right)\,

1018: T_{s_0-\frac{n}2}\left(u-\frac{in}2\right)+\nonumber\\[2mm]

1019: &&\hskip -2.0cm f_{nm}(u+is_0)\,

1020: { t}_{\frac{n-m-1}2}\left(u+is_0-\frac{i(n+m+1)}2\right)\,

1021: T_{s_0+\frac{n+1}2}\left(u+\frac{i(n+1)}2\right)\,,

1022: \end{eqnarray}

1023: with

1024: \begin{equation}

1025:  f_{nm}(u)=\frac{\Delta(u-i(m+1))}{\Delta(u+in)}\,.

1026: \end{equation}

1027: Similarly, starting from Eq.~\eref{3tQ} involving $\widebar

1028: Q$ operator and multiplying by $Q(v)$ one gets another

1029: identity

1030: \begin{eqnarray}\fl\label{fusion-2}

1031: { t}_n(u-is_0)\, T_{s_0+\frac{m+1}2}\left(u-\frac{i(m+1)}2\right)&=&

1032: { t}_{\frac{n-m-1}2}\left(u-is_0-\frac{i(n+m+1)}2\right)\,

1033: T_{s_0-\frac{n}2}\left(u+\frac{in}2\right)+\nonumber\\[2mm]

1034: &&\hskip -2.7cm f_{-m-1,n}(u-is_0)\,{t}_{\frac{n+m}2}\left(u-is_0+\frac{i(n-m)}2\right)\,

1035: T_{s_0+\frac{n+1}2}\left(u-\frac{i(n+1)}2\right)\,.

1036: \end{eqnarray}

1037: For $n=-m=1/2$ the relations~\eref{fusion-1},~\eref{fusion-2} take the standard

1038: form~\cite{KRS}

1039: and relate the transfer matrices with adjacment spins of auxiliary space, $T_{s_0}$ and

1040: $T_{s_0\pm 1/2}$.

1041: 

1042: Next, starting from the Eqs.~\eref{tQQ} and \eref{3tQ} one can derive two quadratic

1043: relations for the finite-dimensional transfer matrices $t_n(u)$. The first one is 

1044: \begin{eqnarray}\fl\label{tt2}

1045: \tilde t_{\frac{m-n-1}2}\left(u+\frac{ik}{2}\right)\tilde

1046: t_{\frac{m+n-1}2}\left(u-\frac{ik}{2}\right)&=&

1047: \tilde t_{\frac{m-k-1}2}\left(u+\frac{in}{2}\right)\tilde

1048: t_{\frac{m+k-1}2}\left(u-\frac{in}{2}\right)\nonumber\\

1049: &&+

1050: \tilde t_{\frac{k-n-1}2}\left(u+\frac{im}{2}\right)\tilde

1051: t_{\frac{k+n-1}2}\left(u-\frac{im}{2}\right)\,.	

1052: \end{eqnarray}	

1053: Here the numbers $m,k,n$ are all integer or half-integer  and $m>k>n$. The second

1054: relation is obtained from the first one by  changing $n\to -n$.

1055: 

1056: \vskip 0.5cm

1057: 

1058: The treatment of the open spin chain goes  along the same lines.

1059: The analog of Eq.~\eref{tQQ} reads

1060: \begin{equation}\fl\label{tQQo}

1061: (2iu-1)\,\Delta_n^2(u)\,t_{n}^{\mathrm{op}}(u)={\mathcal Q}(u-in)\, \widebar{\mathcal  Q}(u+i(1+n))-

1062: {\mathcal Q}(u+i(1+n))\,

1063: \widebar{\mathcal Q}(u-in)\,,

1064: \end{equation}

1065: where

1066: \begin{equation}

1067: \widetilde {\mathcal Q}(u)=\frac{1}{s-i+iu}\left(\frac{\Gamma(1+s-iu)}{\Gamma(1-s-iu)}\right)^{2N}

1068: \widebar {\mathcal Q}(u+i)

1069: \end{equation}

1070: and

1071: $$

1072: t_{n}^{\mathrm{op}}(u)=\tr_{V_{n}} r_{10}(u)\ldots r_{N0}(u)\,r_{N0}^{-1}(-u) \ldots r_{10}^{-1}(-u).

1073: $$

1074: Obviously,  the operator ${\mathcal Q}(u)$ ($\widebar

1075: {\mathcal Q}(u)$) for the open chain satisfies the same

1076: Eq.~\eref{3tQ} with $\tilde t_n(u)=(2iu-1)\,\Delta^2_n(u)\,

1077: t_{n}^{\mathrm{op}}(u)$. The Wronskian relation and Baxter

1078: equation take the well known form~ \cite{ES88,open}

1079: \begin{eqnarray}\fl

1080: &&(2iu-1)\left(\frac{\Gamma(1-s-iu)}{\Gamma(1+s-iu)}\right)^{2N}=

1081: {\mathcal Q}(u)\,\widebar {\mathcal Q}(u+i)-{\mathcal Q}(u+i)\,\widebar{\mathcal Q}(u)\,,\\[2mm]

1082: \fl

1083: &&\tau_N^{\mathrm{op}}(u){\mathcal Q}(u)=\frac{2iu+1}{2iu}(u+is)^{2N}{\mathcal Q}(u+i)+

1084: \frac{2iu-1}{2iu}(u-is)^{2N}{\mathcal Q}(u-i)\,.

1085: \end{eqnarray}

1086: \vskip  0.2mm \noindent Here $\tau_N^{\mathrm{op}}(u)=\tr

1087: L_1(u)\ldots L_N(u)\,L_N(u)\ldots L_1(u)$. Again, the

1088: eigenvalues of the operator $Q(u)$ are polynomials in

1089: $u$~\cite{open}, while the eigenvalues of $\widebar Q$ are

1090: meromorphic functions.

1091: 

1092: In full analogy with the closed spin chain one can derive

1093: two sets of the "fusion" relations and check that the finite-dimensional transfer

1094: matrices, $\tilde t_n(u)=(2iu-1)\,\Delta^2_n(u)\,

1095: t_{n}^{\mathrm{op}}(u)$, satisfy the quadratic relation~\eref{tt2}.

1096: 

1097: 

1098: \vskip 0.3cm

1099: 

1100: 

1101: 

1102: 

1103: 

1104: \section{Conclusions}\label{Concl}

1105: In this paper we considered the quantum spin chains with

1106: $s\ell(2)$ symmetry.  The Hilbert space of the model is given

1107: by the tensor product of the $s\ell(2)$ modules. For a generic

1108: spin $s$, the latter are infinite dimensional and

1109: equivalent to the space of the polynomials of an arbitrary

1110: degree, $\mathbf{V}_s=\mathbb{C}[z]$.

1111: Using the factorization

1112: of the $\mathcal R$ operator 

1113: $$

1114: {\mathcal R}_{12}(u) = P_{12}\, {\mathcal R}^{+}_{12}\left(\frac{s_2-s_1+iu}{2}\right)\, 

1115: {\mathcal R}^{-}_{12}\left(\frac{s_1-s_2+iu}{2}\right)

1116: $$

1117: we have shown that the

1118: transfer matrices both for the closed and open homogeneous

1119: spin chains (for generic spin of the auxiliary space $s_0$)

1120: factorize into a product of two commuting operators $Q$ and

1121: $\widetilde Q$. 

1122: The latter are given by the trace of the product of operators 

1123: ${\mathcal L}^\pm_{k0}={\mathcal R}_{k0}(\mp i(s_0-s))$ over the auxiliary space

1124: (see Eqs.~\eref{QL},~\eref{QQL}).

1125: 

1126: For  negative half-integer spins,

1127: $s_0=-n$, $n=0,1/2,1,\ldots$, the module $\mathbf{V}_s$ has

1128: a finite dimensional invariant subspace, ${V}_{n}$.

1129: The representation induced on the factor space

1130: $\VV_{-n}/V_{n}$ is equivalent to the $s\ell(2)$ module with

1131: spin $s'_0=1+n$. We have shown, that the operators $Q$ and

1132: $\widetilde Q$ satisfy the finite-difference (Baxter)

1133: equation, which follows unambiguously from the structure of

1134: the reducible $s\ell(2)$ modules and factorization property of

1135: the transfer matrices. The treatment of the closed and open

1136: spin chains goes along the same lines with minor

1137: differences. We hope that similar analysis will be

1138: applicable to the spin chains with higher rank symmetry

1139: groups.

1140: 

1141: The factorization property  of the transfer matrices breaks

1142: down for the finite dimensional spin chain, therefore they

1143: require a special consideration. The problem of

1144: construction of the Baxter $Q$ operator for the finite

1145: dimensional spin chains (in particular $XXX_{1/2}$ spin

1146: magnet) was considered in the Refs.~\cite{Pronko,Korff}.

1147: 

1148: 

1149: \ack

1150: The authors are grateful to G.~P.~Korchemsky for helpful discussions.

1151: This work was supported by the RFFI grants 03-01-00837 (S.~D. and A.~M.) 

1152: and 05-01-0092 (S.~D.) 

1153: and by the Helmholtz Association, contract number VH-NG-004 (A.~M.)

1154: 

1155: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1156: \section*{References}

1157: \begin{thebibliography}{99}

1158: \bibitem{BBGK}

1159:   A.~V.~Belitsky, V.~M.~Braun, A.~S.~Gorsky and G.~P.~Korchemsky,

1160:   %``Integrability in QCD and beyond,''

1161:   Int.\ J.\ Mod.\ Phys.\ A {\bf 19} (2004) 4715

1162:   [arXiv:hep-th/0407232].

1163: %%CITATION = HEP-TH 0407232;%%

1164: 

1165: \bibitem{AT}

1166:   A.~A.~Tseytlin,

1167:   ``Spinning strings and AdS/CFT duality,''

1168:   arXiv:hep-th/0311139.

1169: %%CITATION = HEP-TH 0311139;%%

1170: \bibitem{FST}

1171:   L.~D.~Faddeev, E.~K.~Sklyanin and L.~A.~Takhtajan,

1172:   %``The Quantum Inverse Problem Method. 1,''

1173:   Theor.\ Math.\ Phys.\  {\bf 40} (1980) 688

1174:   [Teor.\ Mat.\ Fiz.\  {\bf 40} (1979) 194].

1175: %%CITATION = TMPHA,40,688;%%

1176: 

1177: \bibitem{LF}

1178:   L.~D.~Faddeev, Les-Houches Lectures, 1995,

1179: ``How Algebraic Bethe Ansatz works for integrable model,''

1180:   arXiv:hep-th/9605187.

1181: %%CITATION = HEP-TH 9605187;%%

1182: 

1183: \bibitem{Baxter}

1184: R.~J.~Baxter, {\em Exactly Solved Models in Statistical Mechanics,}

1185: Academic Press, London, 1982.

1186: 

1187: \bibitem{PG}

1188: M.~Gaudin and V.~Pasquier,

1189:   %``The periodic Toda chain and a matrix generalization of the bessel

1190:   %function's recursion relations,''

1191:   J.\ Phys.\ A {\bf 25} (1992) 5243.

1192: %%CITATION = JPAGB,A25,5243;%%

1193: 

1194: \bibitem{V95}

1195: A.~Y.~Volkov,

1196:   %``Quantum lattice KdV equation,''

1197:   Lett.\ Math.\ Phys.\  {\bf 39} (1997) 313

1198: %  [arXiv:hep-th/9509024].

1199: %%CITATION = HEP-TH 9509024;%%

1200: 

1201: \bibitem{BLZ}

1202:   V.~V.~Bazhanov, S.~L.~Lukyanov and A.~B.~Zamolodchikov,

1203:   %``Integrable Structure of Conformal Field Theory II. Q-operator and DDV

1204:   %equation,''

1205:   Commun.\ Math.\ Phys.\  {\bf 190} (1997) 247

1206: %[arXiv:hep-th/9604044].

1207: %%CITATION = HEP-TH 9604044;%%

1208: 

1209: 

1210: 

1211: \bibitem{SD}

1212: S £ Derkachov %Baxter's Q-operator for the homogeneous XXX spin chain

1213: J.\ Phys.\ A  {\bf 32}  (1999) 5299

1214: 

1215: \bibitem{Pronko}

1216:   G.~P.~Pronko,

1217:   %``On the Baxter's Q operator for the XXX spin chain,''

1218:   Commun.\ Math.\ Phys.\  {\bf 212} (2000) 687

1219: %  [arXiv:hep-th/9908179].

1220: %%CITATION = HEP-TH 9908179;%%

1221: 

1222: \bibitem{KSS}

1223: V.~B.~Kuznetsov, M.~Salerno and E.~K.~Sklyanin,

1224:   %``Quantum Backlund transformation for the integrable DST model,''

1225:   J.\ Phys.\ A {\bf 33} (2000) 171

1226: %  [arXiv:solv-int/9908002].

1227: %%CITATION = SOLV-INT 9908002;%%

1228: \bibitem{RW}

1229:  M.~Rossi, R.~Weston

1230: J.\ Phys.\  A {\bf 35} (2002) 10015

1231: 

1232: \bibitem{open}

1233: S.~E.~Derkachov, G.~P.~Korchemsky and A.~N.~Manashov,

1234:   %``Baxter Q-operator and separation of variables for the open SL(2,R) spin

1235:   %chain,''

1236:   JHEP {\bf 0310} (2003) 053

1237: %  [arXiv:hep-th/0309144].

1238: %%CITATION = HEP-TH 0309144;%

1239: 

1240: 

1241: \bibitem{KM}

1242:   M.~Kirch and A.~N.~Manashov,

1243:   %``Noncompact SL(2,R) spin chain,''

1244:   JHEP {\bf 0406} (2004) 035

1245: %  [arXiv:hep-th/0405030].

1246: %%CITATION = HEP-TH 0405030;%%

1247: 

1248: 

1249: \bibitem{Korff1} C.~Korff, {\it "Representation Theory and

1250: Baxter's $TQ$ equation for the six-vertex model.A

1251: pedagogical overview. "}, \ arXiv:cond-mat/0411758.

1252: 

1253: \bibitem{DKM01}

1254:   S.~E.~Derkachov, G.~P.~Korchemsky and A.~N.~Manashov,

1255:   %``Noncompact Heisenberg spin magnets from high-energy QCD. I: Baxter

1256:   %Q-operator and separation of variables,''

1257:   Nucl.\ Phys.\ B {\bf 617} (2001) 375

1258: %  [arXiv:hep-th/0107193].

1259: %%CITATION = HEP-TH 0107193;%%

1260: 

1261: 

1262: \bibitem{SD2}

1263:   S.~E.~Derkachov,

1264:   {\it ``Factorization of R-matrix and Baxter's Q-operator,''}

1265:   arXiv:math.qa/0507252.

1266: %%CITATION = MATH-QA 0507252;%%

1267: 

1268: \bibitem{SD1}

1269:   S.~E.~Derkachov,

1270:   {\it ``Factorization of the R-matrix.I,''}

1271:   arXiv:math.qa/0503396.

1272: %%CITATION = MATH-QA 0503396;%%

1273: 

1274: \bibitem{AF}

1275: A.~Antonov and B.~Feigin,

1276:   %``Quantum Group Representations and Baxter Equation,''

1277:   Phys.\ Lett.\ B {\bf 392} (1997) 115

1278: %  [arXiv:hep-th/9603105].

1279: %%CITATION = HEP-TH 9603105;%%

1280: 

1281: \bibitem{PKS}

1282: P.~P.~Kulish and  E.~K.~Sklyanin, {\it ``Quantum spectral transform. Recent developments,''}

1283: Lect.~Notes in Physics, {\bf 151}, (1982), 61.

1284: 

1285: 

1286: 

1287: \bibitem{ES91}

1288:   E.~K.~Sklyanin,

1289: {\it ``Quantum inverse scattering method. Selected topics,''}

1290: in ``Quantum Group and Quantum Integrable Systems'', (Nankai Lectures in Mathematical

1291: Physics) ed. Mo-Lin Ge, Singapore, World Scientific, 1992, pp.63-97;

1292:   arXiv:hep-th/9211111.

1293: %%CITATION = HEP-TH 9211111;%%

1294: 

1295: 

1296: 

1297: \bibitem{ES88}

1298:   E.~K.~Sklyanin,

1299:   %``Boundary Conditions For Integrable Quantum Systems,''

1300:   J.\ Phys.\ A {\bf 21} (1988) 2375.

1301: %%CITATION = JPAGB,A21,2375;%%

1302: 

1303: \bibitem{KRS}

1304: P.P. Kulish, N.Yu.Reshetikhin and E.K.Sklyanin, {\it

1305: "Yang-Baxter equation and representation theory"},

1306: Lett.Math.Phys. {\bf 5} (1981) 393-403

1307: 

1308: 

1309: \bibitem{Korff} C.~Korff, {\it "A $Q-$operator for the

1310: twisted $XXX$ model"},  \ arXiv:math-ph/0511022.

1311: 

1312: \end{thebibliography}

1313: \end{document}

1314: 

1315: