1: \documentclass[12pt]{ronbun}

2: 

3: \usepackage{amsmath,amssymb}

4: \usepackage{graphicx,color}

5: \usepackage{cite}

6: \usepackage{bm}

7: \usepackage{dcolumn}

8: %\usepackage{showkeys}

9:  

10: \newcommand{\hx}{\hat{x}}

11: \newcommand{\hxm}{\hat{x}^{\mu}}

12: \newcommand{\hxn}{\hat{x}^{\nu}}

13: \newcommand{\tmn}{\theta^{\mu\nu}}

14: \newcommand{\hD}{\hat{\Delta}}

15: \newcommand{\mo}{\mathcal{O}}

16: \newcommand{\mL}{\mathcal{L}}

17: \newcommand{\pd}{\partial}

18: \newcommand{\vp}{\varphi}

19: \newcommand{\bR}{\mathbb{R}}

20: \newcommand{\nn}{\nonumber}

21: \newcommand{\e}{{\rm e}}

22: \newcommand{\tr}{\rm Tr}

23: \newcommand{\del}{\delta}

24: \newcommand{\ra}{\rangle}

25: \newcommand{\la}{\langle}

26: \newcommand{\rar}{\rightarrow}

27: \newcommand{\lar}{\leftarrow}

28: \newcommand{\lra}{\longrightarrow}

29: \newcommand{\lla}{\longleftarrow}

30: \newcommand{\pdm}{\partial_{\mu}}

31: \newcommand{\pdn}{\partial_{\nu}}

32: \newcommand{\fmn}{F_{\mu\nu}}

33: \newcommand{\am}{A_{\mu}}

34: \newcommand{\an}{A_{\nu}}

35: \newcommand{\dg}{\dagger}

36: \newcommand{\db}{\delta_{\rm B}}

37: \newcommand{\bdb}{\bar{\delta}_{\rm B}}

38: \newcommand{\bx}{\bar{x}}

39: \newcommand{\fx}{x_{f}}

40: \newcommand{\ix}{x_{i}}

41: \newcommand{\vph}{\vec{\phi}}

42: \newcommand{\ph}{\phi}

43: \newcommand{\hphi}{\hat{\phi}}

44: \newcommand{\al}{\alpha}

45: \newcommand{\be}{\beta}

46: \newcommand{\fr}{\frac}

47: \newcommand{\bfx}{{\bf x}}

48: \newcommand{\lam}{\lambda}

49: \newcommand{\lag}{{\mathcal L}}

50: \newcommand{\cpn}{\mathbb{C}P^N}

51: \newcommand{\bfn}{{\bf n}}

52: \newcommand{\Dm}{D_{\mu}}

53: \newcommand{\Dn}{D_{\nu}}

54: \newcommand{\ep}{\epsilon}

55: \newcommand{\bfw}{{\bf w}}

56: \newcommand{\epmn}{\epsilon_{\mu\nu}}

57: \newcommand{\cp}{\mathbb{C}P}

58: \newcommand{\bz}{\bar{z}}

59: \newcommand{\bpd}{\bar{\pd}}

60: \newcommand{\sig}{\sigma}

61: \newcommand{\emn}{\eta_{\mu\nu}}

62: \newcommand{\iemn}{\eta^{\mu\nu}}

63: \newcommand{\mN}{\mathcal{N}}

64: \newcommand{\slD}{\not{\! \mbox{D}}}

65: \newcommand{\da}{a^{\dagger}}

66: \renewcommand{\th}{\theta}

67: \newcommand{\xmu}{x^{\mu}}

68: \newcommand{\xnu}{x^{\nu}}

69: \newcommand{\bth}{\bar{\theta}}

70: \newcommand{\delb}{\delta_{\rm B}}

71: \newcommand{\bdelb}{\bar{\delta}_{\rm B}}

72: \newcommand{\red}{\textcolor{red}}

73: \newcommand{\green}{\textcolor{green}}

74: \newcommand{\magenta}{\textcolor{magenta}}

75: \newcommand{\cyan}{\textcolor{cyan}}

76: \newcommand{\blue}{\textcolor{blue}}

77: \newcommand{\black}{\textcolor{black}}

78: \newcommand{\tg}{\tilde{\gamma}}

79: \newcommand{\s}{\scriptscriptstyle}

80: \newcommand{\hu}{\hat{u}}

81: \newcommand{\hv}{\hat{v}}

82: \newcommand{\I}{\bf 1}

83: \newcommand{\up}{\uparrow}

84: \newcommand{\down}{\downarrow}

85: \newcommand{\Rb}{\mathbb{R}}

86: \newcommand{\Cb}{\mathbb{C}}

87: \newcommand{\non}{\nonumber\\}

88: \newcommand{\dis}{\displaystyle}

89: \newcommand{\1}{\mathbb I}

90: \newcommand{\CC}{\mathcal C}

91: \newcommand{\CM}{\mathcal M}

92: \newcommand{\CO}{\mathcal O}

93: \newcommand{\CP}{\mathcal P}

94: \renewcommand{\baselinestretch}{1.4}

95: \newcommand\Tr{{\rm Tr}}

96: 

97: 

98: \setlength{\jot}{3pt}

99: \setlength{\voffset}{-2.4cm}

100: \setlength{\textwidth}{17cm}

101: \setlength{\textheight}{23cm}

102: \setlength{\hoffset}{-1.9cm}

103: 

104: \numberwithin{equation}{section}

105: %\renewcommand{\thefootnote}{\fnsymbol{footnote}}

106: \setcounter{figure}{0}

107: 

108: 

109: \begin{document}

110: 

111: \begin{flushright}

112: \parbox{4.2cm}

113: {KEK-TH-990 \hfill \\

114: UTHEP-499 \\

115: {\tt hep-th/0410139}

116:  }

117: \end{flushright}

118: 

119: \vspace*{0.7cm}

120: 

121: \begin{center}

122:  \Large\bf 

123: Open Semiclassical Strings and Long Defect Operators 

124: in AdS/dCFT Correspondence 

125: \end{center}

126: \vspace*{0.7cm}

127: \centerline{\large Yoshiaki Susaki$^{\ast\dagger a}$, 

128: Yastoshi Takayama$^{\ddagger b}$ 

129: and Kentaroh Yoshida$^{\ast c}$}

130: \begin{center}

131: $^{\ast}$\emph{Theory Division,   	

132: Institute of Particle and Nuclear Studies, \\

133: High Energy Accelerator Research 

134: Organization (KEK),\\

135: Tsukuba, Ibaraki 305-0801, Japan.} 

136: \\

137: $^{\dagger}$\emph{Institute of Physics, University of Tsukuba, \\

138: Tsukuba, Ibaraki 305-8571, Japan.} 

139: \\

140: $^{\ddagger}$\emph{

141: Department of Particle and Nuclear Physics, \\

142: The Graduate University for Advanced Studies, \\

143: Tsukuba, Ibaraki 305-0801, Japan.} 

144: 

145: \vspace*{0.3cm}

146: %E-mail:~~

147: $^a${\tt susaki@post.kek.jp} \quad  

148: $^b${\tt takaya@post.kek.jp} \quad 

149: $^c${\tt kyoshida@post.kek.jp}

150: \end{center}

151: 

152: \vspace*{0.7cm}

153: 

154: \centerline{\bf Abstract} 

155: 

156: We consider defect composite operators in a defect superconformal field

157: theory obtained by inserting an AdS$_4\times$S$^2$-brane in the

158: AdS$_5\times$S$^5$ background. The one-loop dilatation operator for the

159: scalar sector is represented by an integrable open spin chain.  We give

160: a description to construct coherent states for the open spin

161: chain. Then, by evaluating the expectation value of the Hamiltonian with

162: the coherent states in a long operator limit, a Landau-Lifshitz type of

163: sigma model action is obtained. This action is also derived from the

164: string action and hence we find a complete agreement in both SYM and

165: string sides. We see that an $SO(3)_{\rm H}$ pulsating string solution

166: is included in the action and its energy completely agrees with the

167: result calculated in a different method. In addition, we argue that our

168: procedure would be applicable to other AdS-brane cases.

169: 

170: 

171:   

172: \vspace*{0.5cm}

173: 

174: \vfill \noindent {\bf Keywords:}~~{\footnotesize AdS/CFT, spin chain,

175: integrability, semiclassical string, defect CFT}

176: 

177: \thispagestyle{empty}

178: \setcounter{page}{0}

179: 

180: \newpage 

181: 

182: 

183: \section{Introduction}

184: 

185: One of the most important subjects in string theory is to test the

186: AdS/CFT correspondence \cite{M} beyond BPS sectors.  In the analysis at

187: almost BPS region, a great development was made by

188: Berenstein-Maldacena-Nastase (BMN) \cite{BMN}. They presented the

189: AdS/CFT correspondence at string theoretical level by using the Penrose

190: limit \cite{P,BFHP1} and the exact solvability of pp-wave string theory

191: \cite{MT}. The pp-wave string states and their energies correspond to

192: the BMN operators and the full conformal dimensions (including the

193: anomalous dimensions) in $\mathcal{N}$=4 super Yang-Mills (SYM),

194: respectively.  

195: 

196: The BMN analysis was generalized by considering the semiclassical

197: quantization \cite{GKP2,FT} around classical rotating string solutions,

198: instead of pp-wave strings. The semiclassical string states and their

199: energies correspond to certain single-trace local operators and their

200: conformal dimensions in SYM in specific regimes.\footnote{The classical

201: integrability of type IIB string on the AdS$_5\times$S$^5$ background

202: \cite{BPR} may be deeply related to the correspondence between string

203: and SYM sides \cite{DNW}. The matching of the spectra and the

204: equivalence of the integrable structures between the spin chain and

205: string theories are confirmed up to and including the two-loop order on

206: a specific example \cite{AS}. } The energy of the semiclassical string

207: state can be expanded in positive power of new effective coupling

208: $\tilde{\lambda}\equiv \lambda/J^2$ ($\lambda$ is 't Hooft coupling)

209: when the classical spinning string solutions have, at least, one of the

210: S$^5$-spins $J_1,~J_2$ and $J_3$\,. It is therefore possible to compare

211: the energy with the conformal dimension perturbatively calculated in

212: $\mathcal{N}$=4 SYM. Several types of classical solutions which give

213: regular expansions of energies have been discovered by some

214: works\cite{GKP2,FT,FT2,ART} (For a review, see \cite{review}). On the

215: other hand, the anomalous dimensions of the single-trace composite

216: operators can be computed by using the Bethe ansatz techniques. This

217: fact was firstly shown by Minahan and Zarembo in a study of the $SO(6)$

218: scalar sector \cite{MZ}. This kind of analysis is applicable to other

219: sectors \cite{BKS} such as $SU(2)$ \cite{BFST,BMSZ,Kazakov}, $SU(3)$

220: \cite{SU(3)}, $SL(2)$ \cite{BFST,SL2,QCD,BS,KZ}\footnote{The

221: integrability related to the $SL(2)$ Heisenberg spin chain was

222: previously found in non-supersymmetric QCD. For example, see

223: \cite{QCD}.}, $SU(2|3)$ \cite{su(2|3)} and the full $PSU(2,2|4)$ sectors

224: \cite{BS}.  In particular, the $SU(2)$ sector is well studied.  An

225: approach to the analysis with the Inozemtsev long range spin chain at

226: three-loop level \cite{SS} and, after that, the asymptotic 

227: all-loop Bethe ansatz\footnote{This is indeed asymptotic in

228: a sense that it is unsuitable for describing the operators of small

229: dimensions at sufficiently high orders of perturbation theory

230: (e.g.\,certain interactions of wrapping type are missed). } 

231: is proposed \cite{BDS}. 

232: 

233: For the Bethe ansatz, in parallel to studies of quantum strings on the

234: AdS$_5\times$S$^5$\,, new types of Bethe ansatz for them were developed

235: \cite{quantum}. This important development should be a promising

236: procedure to investigate the quantum aspects of the AdS-string.

237: \vspace*{0.3cm}

238: 

239: In this paper we study the correspondence between semiclassical strings

240: and long scalar operators in open string cases.  By inserting spatial

241: defects (called AdS-branes) in the bulk AdS$_5\times$S$^5$ background

242: \cite{KR}, one may consider open strings in the context of the AdS/CFT

243: correspondence where the CFT side has defects and so becomes a

244: defect conformal field theory (dCFT).  We focus on the case that an

245: AdS$_4\times$S$^2$-brane is inserted.  The dCFT was particularly

246: investigated by DeWolfe-Freedman-Ooguri \cite{DFO} and it was shown to

247: be superconformal \cite{EGK}. The BMN operator correspondence for this

248: system was also studied by Lee and Park \cite{LP}. Furthermore, the

249: matrix of one-loop anomalous dimension for defect operators (which have

250: the defect fields in the end-points, instead of the trace) in the scalar

251: sector was calculated by DeWolfe and Mann \cite{DM}\footnote{For open

252: strings in other setup, see \cite{BGMNN,BHLN} and

253: \cite{Open-sol,CWW}.}. The resulting dilatation operator is an

254: integrable {\it open} spin chain Hamiltonian while a {\it closed} spin

255: chain appears in the case of single trace operators. By using the Bethe

256: ansatz techniques, the exact BMN operators \cite{B} for the open strings

257: were derived as in \cite{MZ}. In the long operator limit, those are

258: consistently reduced to the result of \cite{LP}. On the other hand,

259: except the BMN operators, the string side is not uncovered yet. 

260: 

261: Motivated by this fact, we will reveal the open semiclassical string

262: corresponding to the open spin chain by using the method 

263: of \cite{Kruczenski,KRT} with coherent state \cite{coherent}. 

264: We give a description to construct coherent

265: states for the defect operators and then the expectation value of the

266: open spin chain Hamiltonian is taken with them. In a continuum limit

267: (long operator limit), the Landau-Lifshitz (LL) type sigma model action

268: is obtained as in \cite{ST-coh,KT-coh,review-coh}. Namely, the first

269: time-derivative appears instead of the usual quadratic one. On the other

270: hand, we derive the sigma model from the string action with a suitable

271: gauge-fixing \cite{KT-coh,review-coh}. In conclusion, we find a complete

272: agreement of the sigma model actions as in

273: \cite{Kruczenski,KRT,HL,ST-coh,KT-coh,review-coh}. We also see that the

274: resulting sigma model action includes a pulsating solution and its

275: energy is recovered by putting the solution ansatz into the action.

276: 

277: \vspace*{0.5cm}

278: This paper is organized as follows: In section 2, the dCFT we consider

279: is briefly introduced. Section 3 is devoted to a short review of the

280: one-loop result in the scalar sector \cite{DM}. In section 4, 

281: we give a description to construct

282: coherent states for the open spin chain Hamiltonian. Then we take the

283: expectation value of the Hamiltonian, and the LL type sigma

284: model is obtained in a continuum limit. We find that an $SO(3)_{\rm H}$

285: pulsating string solution is included in this action. In section 5, we

286: rederive the sigma model obtained in section 4 from the string

287: action. In section 6, we discuss coherent states for other AdS-brane

288: cases. One may see that our consideration for an AdS$_4\times$S$^2$

289: brane would be extended to other AdS-branes.  Section 7 is devoted to a

290: conclusion and discussions.

291: 

292: 

293: \section{Setup of Defect Conformal Field Theory} 

294: 

295: From now on, we will consider a dCFT by inserting a supersymmetric

296: AdS$_4\times$S$^2$ brane into the AdS$_5\times$S$^5$ background

297: \cite{DFO}. The AdS$_4$ brane inside AdS$_5$ could be naturally realized

298: in string theory by considering a supersymmetric D5-brane intersecting a

299: stack of $N$ D3-branes as depicted in Fig.\,\ref{D3-D5:fig}. The

300: near-horizon limit of the $N$ D3-branes produces an AdS$_5\times$S$^5$

301: background where the D5-brane is realized as an AdS$_4\times$S$^2$

302: submanifold. That is, in this system, closed strings propagating

303: throughout the spacetime provide a holographic description of an

304: $\mathcal{N}$=4 SYM$_4$ on the boundary of AdS$_5$\,. In addition, the

305: fluctuations on the AdS$_4$-brane should be dual to additional physics

306: confined to the boundary of the AdS$_4$\,, and the dual field theory

307: contains new fields living on a $(1+2)$-dimensional defect, obtained

308: from the low-energy limit of the 3-5 open strings interacting with the

309: 3-3 strings of the original brane setup. The 5-5 strings are realized as

310: open strings on the AdS$_4\times$S$^2$ brane in the gravity side.

311: 

312: \vspace*{0.2cm}

313: \begin{figure}[htbp]

314:  \begin{center}

315:   \includegraphics[scale=.7]{D3-D5.eps}

316:  \end{center}

317: \vspace*{-0.5cm}

318: \caption{An intersecting D5-brane with a stack of $N$ D3-branes.}

319: \label{D3-D5:fig}

320: \end{figure}

321: 

322: In the CFT side, $\mathcal{N}$=4 SYM has a three-dimensional defect

323: due to the presence of the AdS-brane. One may see the usual

324: $\mathcal{N}$=4 SYM in the long distance from the defect, while

325: $\mathcal{N}$=4 SYM couples to the defect fields in the neighborhood of

326: the defect. Then the vector multiplet of $\mathcal{N}$=4 in four

327: dimensions decomposes into a vector multiplet and a hypermultiplet in

328: three dimensions. 

329: 

330: Now let us introduce the field contents of $\mathcal{N}$=4 SYM and

331: dCFT. The $\mathcal{N}$=4 vector multiplet is composed of a gauge field

332: $A_{\mu}~(\mu=0,1,2,3)$, adjoint Majorana spinors

333: $\lambda^{\al}~(\al=1,2,3,4)$ and six real scalars

334: $X^i~(i=1,\ldots,6)$\,. The $\lambda^{\al}$ and $X^i$ are in the 

335: ${\bf 4}$ and ${\bf 6}$ of the $SO(6)$ R-symmetry, respectively. Our

336: convention basically follows that of \cite{DM}.  Then the defect

337: introduces an additional 3D hypermultiplet propagating on the

338: hypersurface $x^3 = 0$\,. This multiplet is composed of a complex scalar

339: $q^m$ and a complex 2-component fermion $\Psi^a$\,. These new fields

340: break the total symmetry, and the breaking pattern is as follows:

341: \begin{eqnarray}

342: && \mbox{Conformal symmetry}: \quad SO(2,4) \rightarrow SO(2,3) \cong 

343: Sp(4)\,, \nn \\

344: && \mbox{R-symmetry}: \quad 

345: SO(6)_{\rm R} \rightarrow SO(3)_{\rm H} \times SO(3)_{\rm V}\cong 

346: O(4)\,, \nn \\

347: && \mbox{Superconformal symmetry}: \quad PSU(2,2|4) \rightarrow

348:  OSp(4|4)\,. \nn 

349: \end{eqnarray}

350: The defect fields 

351: $q^m$ and $\Psi^a$ transform 

352: in the $({\bf 2}, {\bf 1})$ and in the $({\bf 1}, {\bf 2})$ 

353: of $SO(3)_{\rm H}\times SO(3)_{\rm V}$\,, respectively. 

354: The $\mathcal{N}$=4 vector multiplet is also decomposed into 

355: \begin{eqnarray}

356: && \mbox{3D vector multiplet}:~~

357: \{A_k, P_+\lambda^{\al}, X_{\rm V}^A, D_3X_{\rm H}^I\}\,, \nn \\ 

358: && \mbox{3D hypermultiplet}:~~

359: \{A_3, P_-\lambda^{\al}, X_{\rm H}^I, D_3 X_{\rm V}^A\}\,, \nn 

360: \end{eqnarray}

361: where $k=0,1,2$ and $i=1,\ldots,6$ of $SO(6)$ is decomposed into

362: $A=1,2,3$ of ${\bf 3}$ in the $SO(3)_{\rm V}$ and $I = 4,5,6$ of ${\bf

363: 3}$ in the $SO(3)_{\rm H}$\,. That is, $X_{\rm H}^I$ and $X_{\rm V}^A$

364: are in the $({\bf 3}, {\bf 1})$ and $({\bf 1}, {\bf 3})$ of the

365: $SO(3)_{\rm H}\times SO(3)_{\rm V}$\,, respectively. 

366: 

367: These defects couple directly only to the bulk vector multiplet, and

368: their dynamics is described by the following action\footnote{We will

369: consider the case that a single D5-brane intersecting with a stack of

370: $N$ D3-branes in this paper.  For $M \ll N$\,, we may consider the case

371: that $M$ multiple D5-branes intersect with $N$ D3-branes, and flavor

372: indices appear on $q$ and $\Psi$ in an obvious way, while the $M\sim N$

373: case is more involved and difficult because the backreaction of the

374: D5-branes should be considered.}

375: \begin{eqnarray}

376: S_3 &=& S_{\rm kin} + S_{\rm yuk} + S_{\rm pot}\,, \\

377: S_{\rm kin} &=& \frac{1}{g^2}\int\!\!d^3x\Bigl[

378: (D^k q^m)^{\dagger}D_k q^m -i\bar{\Psi}^a\rho^kD_k\Psi^a

379: \Bigr]

380: \,, \nn \\

381: S_{\rm yuk} &=& \frac{1}{g^2}\int\!\!d^3x\Bigl[

382: i\bar{\Psi}^aP_+\lambda_{am}q^m -i\bar{q}^m\bar{\lambda}_{ma}P_+\Psi^a 

383: + \bar{\Psi}^a\sigma^A_{ab}X_{\rm V}^{A}\Psi^a

384: \Bigr]\,, \nn \\

385: S_{\rm pot} &=& \frac{1}{g^2}\int\!\!d^3x\Bigl[

386: \bar{q}^m X_{\rm V}^AX_{\rm V}^Aq^m 

387: + i\epsilon_{IJK}\bar{q}^m\sigma_{mn}^{I}X_{\rm H}^JX_{\rm H}^Kq^n

388: \Bigr] \nn \\

389: && + \frac{1}{g^2}\int\!\!d^3x\left[

390: \bar{q}^m\sigma^I_{mn}(D_3 X_{\rm H}^I)q^n 

391: + \frac{1}{4}\delta(0){\rm Tr}(\bar{q}^m\sigma^I_{mn}q^n)^2

392: \right]\,, \nn

393: \end{eqnarray}

394: where the covariant derivative is defined as $D_k \ast \equiv \partial_k

395: \ast -i A_k \ast$\,. The total action of the theory is given by the

396: action of $\mathcal{N}$=4 SYM$_4$ and the above action for the defect

397: fields. Notably, the total theory has only one parameter, coupling

398: $g_{\rm YM}$\,, and so it is exactly marginal \cite{EGK}.

399: 

400: 

401: \section{The One-Loop Anomalous Dimensions of Defect Operators}

402: 

403: Here we shall briefly review the one-loop result obtained by DeWolfe and

404: Mann \cite{DM}. We are interested in open strings on the

405: AdS$_4\times$S$^2$-brane. Their states are described by local,

406: gauge-invariant defect operators composed of Lorentz scalar fields:

407: \begin{eqnarray}

408: \mathcal{O} = \psi_{m,j_1,\ldots,j_L,n}\,\bar{q}_mX^{j_1}\cdots

409:  X^{j_L}q^n\,. 

410: \label{cop}

411: \end{eqnarray}

412: Here two defect scalar fields appear at both ends of a sequence of

413: scalar fields,   

414: instead of the trace operation in closed string cases.  

415: By evaluating the correlation function,

416: \begin{eqnarray}

417: \la \bar{q}_{n'}(z_{\beta})X^{i_L}(z_L)\cdots X^{i_1}(z_1)

418: q^{m'}(z_{\al})\mathcal{O}(0)\ra\,, 

419: \end{eqnarray}

420: one can compute the matrix of anomalous dimension for (\ref{cop}).  In

421: the one-loop level analysis, the bulk interactions lead to the same

422: result in closed string cases, but the defect interactions give

423: additional terms to the matrix of anomalous dimension.

424: 

425: The matrix of anomalous dimension at one-loop level is 

426: computed by evaluating the one-loop planar diagrams. 

427: It is represented by the Hamiltonian of an integrable

428: open spin chain and it is composed of the bulk part and the defect

429: part as follows:

430: \begin{eqnarray}

431: && D_{SO(6)} = \Gamma^{\rm bulk}_{\mathcal{O}} + 

432: \Gamma^{\rm defect}_{\mathcal{O}} =   \frac{\lambda}{16\pi^2}

433: \sum_{l=1}^{L-1}H_{l,l+1} + \frac{\lambda}{16\pi^2}\Bigl[

434: (2I_{\bar{\al}1} +2\bar{S}_{\bar{\al}1}) 

435: + (2I_{L\beta} +2 S_{L\beta})

436: \Bigr]\,, 

437: \label{open-spin} \\

438: && H_{l,l+1} = K_{l,l+1} + 2I_{l,l+1} -2 P_{l,l+1}\,. 

439: \end{eqnarray}

440: The bulk part (the first term in (\ref{open-spin})) is the same 

441: as the result of Minahan and Zarembo \cite{MZ}, and it is written by 

442: the trace $I$, permutation $P$ and trace $K$ operators: 

443: \begin{eqnarray}

444: I_{j_{l}j_{l+1}}^{i_{l}i_{l+1}} 

445: = \delta^{i_l}_{j_l}\delta^{i_{l+1}}_{j_{l+1}}\,, \quad 

446: P_{j_{l}j_{l+1}}^{i_li_{l+1}} = \delta^{i_l}_{j_{l+1}}

447: \delta_{j_l}^{i_{l+1}}\,, \quad 

448: K_{j_lj_{l+1}}^{i_li_{l+1}} = \delta^{i_li_{l+1}}\delta_{j_lj_{l+1}}\,.  

449: \end{eqnarray}

450: The defect part (the second term in (\ref{open-spin})) 

451: is written in terms of 

452: \begin{eqnarray}

453: && I^{\bar{m} I}_{\bar{n} J} = \del^{\bar{m}}_{\bar{n}} 

454: \del_{J}^{I}\,, \quad 

455: I^{I m}_{J n} = \del_{J}^{I}\del^{m}_{n}\,, 

456: \quad I^{\bar{m} A}_{\bar{n} B} 

457: = I^{\bar{m} I}_{\bar{n} B} = I^{\bar{m} A}_{\bar{n} J} = 0\,, \quad 

458: I_{A m}^{B n} = I_{A m}^{J n} = I_{I m}^{B n} = 0\,, \nn \\  

459: && S_{Jn}^{Im} = -i\epsilon_{IJK}\sigma^K_{mn}\,, \quad 

460: S_{Bn}^{Am} = \del^m_n\del^A_B\,, \quad S^{Im}_{Bn} = S^{Am}_{Jn} = 0\,, 

461:  \\ && 

462: \bar{S}_{\bar{n}J}^{\bar{m}I} = i\epsilon_{IJK}\sigma^K_{\bar{m}\bar{n}}\,, 

463: \quad~~

464: \bar{S}_{\bar{n}B}^{\bar{m}A} = \del_{\bar{n}}^{\bar{m}}\del^A_B\,, 

465: \quad \bar{S}^{\bar{m}I}_{\bar{n}B} = \bar{S}^{\bar{m}A}_{\bar{n}J} = 0\,. 

466: \nn 

467: \end{eqnarray} 

468: Here we should note that the bulk part has no periodicity. In the

469: present case, the integrable boundary arising from the defect

470: contribution ensures the integrability of the spin chain, instead of

471: periodicity.

472: 

473: It may be helpful for a check of calculations to use the chiral primary

474: operators \cite{DFO}, 

475: \begin{eqnarray}

476: \bar{q}^m \sigma_{mn}^{(I_1}X_{\rm H}^{I_2}X_{\rm H}^{I_3} \cdots 

477: X_{\rm H}^{I_{L+1})}q^n\,,

478: \label{chiral}

479: \end{eqnarray}

480: where parentheses denote total symmetrization and traceless with respect

481: to the index of $SO(3)_{\rm H}$\,. One can indeed verify that 

482: each $H_{l,l+1}$, as well as $2I_{\bar{\al}1} + 2\bar{S}_{\bar{\al}1}$

483: and $2I_{L\beta} + 2S_{L\beta}$\,, separately gives zero by acting them

484: to (\ref{chiral}). 

485: 

486: Finally, by introducing the $SO(6)$ generators $M_{ab}^{ij} \equiv

487: \del^{i}_a \del^j_b - \del_a^j\del_b^i$\,, a piece of the spin chain

488: Hamiltonian, $H_{l,l+1}$ is rewritten as 

489: \begin{eqnarray}

490: \label{int}

491: H_{l,l+1} = M_l^{ij}M_{l+1}^{ij} - \frac{1}{16}(M_l^{ij}M_{l+1}^{ij})^2 + \frac{9}{4}\,. 

492: \end{eqnarray}

493: In this form the spin-spin interaction is manifest. The expression

494: (\ref{int}) is useful to evaluate the expectation value of the spin

495: chain Hamiltonian with coherent states. 

496: 

497: \section{Coherent States and Sigma Model for Defect Operators} 

498: 

499: 

500: \subsection{Construction of Coherent States for Open Spin Chain}

501: 

502: We will consider how to construct the coherent states for the open spin

503: chain. The open spin chain has boundaries which break the $SO(6)$

504: symmetry to $SO(3)_{\rm H} \times SO(3)_{\rm V}$\,. This fact reflects

505: that a D5-brane, whose shape is an AdS$_4\times$S$^2$ geometry, is

506: inserted into the AdS$_5\times$S$^5$\,.  We may consider various open

507: semiclassical strings attaching to the S$^2$ part, but we concentrate on

508: open strings sticking to the S$^2$ described in the three-dimensional

509: space spanned by $X_{\rm H}^I~(I=1,2,3)$ with the condition

510: $\sum_{I=1}^3(X_{\rm H}^I)^2 = 1$\,. It would be difficult to obtain a

511: regular semiclassical string energy (i.e., a regular BMN limit) in the

512: other cases. In fact, we have not succeeded so far.

513: 

514: Hence let us consider the following ansatz for the whole coherent state:

515: \begin{eqnarray}

516:  | m \ra = |\bar{q} \ra \otimes 

517: \prod_{l=1}^{L}|m_l\ra \otimes  |q \ra\,,  

518: \label{whole}

519: \end{eqnarray}

520: where $|m_l\ra $ are $SO(3)_{\rm H}$ coherent states, and $|\bar{q}\ra$,

521: $|q\ra$ are $SU(2)_{\rm H}$ ones. We need to construct and adjust the

522: these coherent states so that the well-defined continuum limit can be

523: taken.

524: 

525: For the $SO(3)_{\rm H}$ sector, the coherent states are described by

526: using a coset $SO(3)_{\rm H}/SO(2)$\,. Then, according to the choices of

527: the $SO(6)$ coherent states, there are two possibilities:

528: \begin{enumerate}

529: \item the vacuum $|0\ra = (0,1,i)$ and $H= SO(2)$\, where the vacuum is

530: invariant up to a phase factor. The Cartan generator is $M_{56}$ and the

531: coherent state is generated by $M_{45}$ and $M_{64}$\,. 

532: \item the vacuum $|0\ra =(0,0,1)$ and $H = SO(2)$\,. The Cartan

533: generator is $M_{45}$ and the coherent state is generated by $M_{56}$

534: and $M_{64}$\,.

535: \end{enumerate} 

536: Here we have extracted the vacuum and the generators for $SO(3)_{\rm

537: H}/SO(2)$ from the full $SO(6)$ case. As we will see below, the

538: ``local'' BPS property (case 1) and the large ``extensive'' one-loop

539: shift of the dimension (case 2) are also inherited even after the

540: $SO(6)$ is spontaneously broken. We will concentrate on the case 1. In

541: the end of this section we will comment on the case 2. We will explain

542: the construction of $SO(3)_{\rm H}/SO(2)$ and $SU(2)_{\rm H}/U(1)$

543: coherent states below.

544: 

545: \subsubsection*{Construction of $SO(3)_{\rm H}/SO(2)$ coherent states}

546: 

547: Let us introduce $M^{IJ}$ as $SO(3)_{\rm H}$ generators in the

548: fundamental representation $(M^{IJ})_{ab} = \delta^I_a \delta^J_b -

549: \delta^I_b \delta^J_a ~(a,b=1,2,3,~I,J=1,2,3)$\footnote{We have

550: redefined the expressions of the $SO(3)_{\rm H}$ generators as $M_{45}$,

551: $M_{56}$, $M_{64}$ $\rightarrow$ $M_{31}$, $M_{12}$, $M_{23}$.}.

552: Choosing a vector $(1,i,0)$ as a vacuum $|0\ra$ (i.e., the highest

553: weight state for $M^+= M^{23}+iM^{31}$)\,, the $SO(3)_{\rm H}/SO(2)$

554: coherent state is given by

555: \begin{eqnarray}

556:  | m \rangle = \mbox{\boldmath{$m$}}

557: \equiv

558:  \frac{1}{\sqrt{2}}\exp[aM^{23}+bM^{31}]

559: 	\left[

560:    \begin{array}{@{\,}c@{\,}}

561:     1\\

562: 	 i\\

563:     0

564: 	\end{array}

565:    \right]

566: = \frac{\e^{-i\phi}}{\sqrt{2}}

567:  \left[

568: 	\begin{array}{@{\,}c@{\,}}

569: 	\cos\theta \cos\phi + i \sin\phi\\

570: 	-\cos\theta \sin\phi + i \cos\phi\\

571: 	\sin\theta

572: 	\end{array}

573:  \right]\,, \label{SO3m}

574: \end{eqnarray} 

575: where $\theta=\sqrt{a^2+b^2}$, $a=\theta\sin\phi$ and

576: $b=\theta\cos\phi$\,. One can explicitly check that $\langle m|m \rangle

577: = |\mbox{\boldmath{$m$}}|^2= 1\,, \quad \mbox{\boldmath{$m$}}^2 =

578: (\mbox{\boldmath{$m$}}^*)^2 = 0$\,. By putting this coherent state at

579: each $l$-th site as $|m_l\ra$\,,

580: \begin{eqnarray}

581:  |m_l \rangle = \mbox{\boldmath $m$}_l =

582:  \frac{\e^{-i\phi_l}}{\sqrt{2}}

583:  \left[

584: 	\begin{array}{@{\,}c@{\,}}

585: 	\cos\theta_l \cos\phi_l + i \sin\phi_l\\

586: 	-\cos\theta_l \sin\phi_l + i \cos\phi_l\\

587: 	\sin\theta_l

588: 	\end{array}

589: \right]

590: \qquad (l=1,\ldots, L)\,, \label{m-coh}

591: \end{eqnarray}

592: the $SO(3)_{\rm H}$ coherent states for the

593: bulk part of the spin chain are constructed.  

594: 

595: 

596: \subsubsection*{Construction of $SU(2)_{\rm H}/U(1)$ coherent states }

597: 

598: We have to determine two more coherent states $|\bar{q}\rangle$ and

599: $|q\rangle$\,. The $SU(2)_{\rm H}/U(1)$ coherent states can be written,

600: up to a $U(1)$ factor, as

601: \begin{eqnarray}

602: && \hspace*{-0.5cm}

603: \bigg| \frac{SU(2)_{\rm H}}{U(1)}\bigg \rangle

604: =  \e^{i\alpha \sigma_1 + i\beta \sigma_2}

605:  \left[

606: 	\begin{array}{@{\,}c@{\,}}

607: 	1\\

608: 	0

609: 	\end{array}

610:  \right]

611: =

612:  \left[

613: 	\begin{array}{@{\,}c@{\,}}

614: 	\cos(\vartheta/2)\\

615: 	\e^{-i \varphi} \sin(\vartheta/2)

616: 	\end{array}

617:  \right] 

618: \simeq

619:  \left[

620: 	\begin{array}{@{\,}c@{\,}}

621: 	\e^{i \varphi /2}\cos(\vartheta/2)\\

622: 	\e^{-i \varphi /2} \sin(\vartheta/2)

623: 	\end{array}

624:  \right]\,, 

625: \end{eqnarray}

626: where $\vartheta = 2\sqrt{\alpha^2+\beta^2}\,,~

627: \alpha = \frac{\vartheta}{2}\sin\varphi\,,~\beta 

628: = \frac{\vartheta}{2}\cos\varphi$\,. 

629: The vacuum $(1,0)$ is the highest weight state

630: for $\sigma_+=\sigma_1+i\sigma_2$\,. We take 

631: $|\bar{q}\rangle$ and $|q\rangle$ as

632: \begin{eqnarray}

633:  |\bar{q}\ra = \left[

634: 	\begin{array}{@{\,}c@{\,}}

635: 	\e^{-i \varphi_{\bar q} /2}\cos(\vartheta_{\bar q}/2) \\

636: 	\e^{i \varphi_{\bar q} /2} \sin(\vartheta_{\bar q}/2)

637: 	\end{array}

638:  \right]\,, \qquad

639:  |q \rangle = \left[

640: 	\begin{array}{@{\,}c@{\,}}

641: 	\e^{i \varphi_q /2}\cos(\vartheta_q/2)\\

642: 	\e^{-i \varphi_q /2} \sin(\vartheta_q/2)

643: 	\end{array}

644:  \right]\,.

645: \label{q-coh}

646: \end{eqnarray} 

647: These coherent states satisfy 

648: \begin{eqnarray}

649: \la \bar{q} | \mbox{\boldmath{$\sigma$}}  | \bar{q} \ra =

650: \la q | \mbox{\boldmath{$\sigma$}}^T  | q \ra 

651: = \bigl(

652: \sin\vartheta \cos\varphi, \sin\vartheta \sin \varphi, \cos\vartheta \bigr) 

653: \equiv \mbox{\boldmath $n$}\,.

654: \end{eqnarray} 

655: 

656: We next discuss the expectation values of $\Gamma_{\cal O}^{\rm

657: defect}$\,. The boundary parts are described as

658: \begin{eqnarray}

659:  |{\cal L} \ra = | m_1 \ra \otimes |\bar{q} \ra\,, \qquad |{\cal R} \ra

660:  = | m_L \ra \otimes |q \ra\,.

661: \end{eqnarray} 

662: For the left end-point, the expectation value is\footnote{For a moment,

663: we omit the factor $\frac{\lambda}{8\pi^2}$ for convenience since it is

664: irrelevant here.}

665: \begin{eqnarray}

666:  \langle {\cal L} | \Gamma^{\rm defect-L}_{\cal O} | {\cal L} \ra &\simeq&

667:  \langle {\cal L} | I_{\bar{q}1}+S_{\bar{q}1} | {\cal L} \ra 

668: = 1 + i\mbox{\boldmath $n$}_{\bar{q}}\cdot(\mbox{\boldmath $m$}_1^{\ast}\times 

669: \mbox{\boldmath $m$}_1) \nn \\

670: &=& 1 -(\sin\vartheta_{\bar{q}} \cos \varphi_{\bar q}, 

671: \sin\vartheta_{\bar{q}} \sin \varphi_{\bar q}, \cos\vartheta_{\bar q})

672: \left[

673: \begin{array}{@{\,}c@{\,}}

674: \sin \theta_1 \cos(\pi-\phi_1)\\

675: \sin \theta_1 \sin(\pi-\phi_1)\\

676: \cos\theta_1

677: \end{array}

678: \right]\,.

679: \end{eqnarray} 

680: In a similar way, for the right end-point, we obtain 

681: \begin{eqnarray}

682:  \langle {\cal R} | \Gamma^{\rm defect-R}_{\cal O} | {\cal R} \ra 

683:  &\simeq& \langle {\cal R} | I_{Lq}+S_{Lq} | {\cal R} \ra 

684: =  1 - i\mbox{\boldmath $n$}_{q}\cdot(\mbox{\boldmath $m$}_L^{\ast}\times 

685: \mbox{\boldmath $m$}_L) \nn \\ 

686: &=& 1 + (\sin\vartheta_{q} \cos \varphi_{q},

687:  \sin\vartheta_{q} \sin \varphi_{q}, \cos\vartheta_{q}) \left[

688: 	\begin{array}{@{\,}c@{\,}}

689: 	\sin \theta_L \cos(\pi-\phi_L)\\

690: 	\sin \theta_L \sin(\pi-\phi_L)\\

691: 	\cos\theta_L

692: 	\end{array}

693: \right]\,.

694: \end{eqnarray} 

695: 

696: Now let us consider the following conditions:

697: \begin{eqnarray}

698:  \left\{

699:  \begin{array}{c} 

700:   \displaystyle 

701: \vartheta_{\bar q} =  \theta_1 + \frac{1}{L}\delta\theta_1 

702: + \frac{1}{L^2}\Delta\theta_1 

703: \vspace*{0.2cm} \\

704: 	\displaystyle 

705: \varphi_{\bar q} = \pi - \phi_1 

706: - \frac{1}{L}\delta\phi_1  - \frac{1}{L^2}\Delta\phi_L

707:  \end{array}

708: \right. 

709: \quad {\rm and} \quad 

710:  \left\{

711:  \begin{array}{c}

712:   \displaystyle 

713: \vartheta_q =  \pi + \theta_L + \frac{1}{L}\delta\theta_L 

714: + \frac{1}{L^2}\Delta\theta_L 

715: \vspace*{0.2cm}\\

716: 	\displaystyle 

717: \varphi_q = \pi - \phi_L 

718: - \frac{1}{L}\delta\phi_L - \frac{1}{L^2}\Delta\phi_L

719:  \end{array}

720: \right. 

721: \,. \label{relax}

722: \end{eqnarray} 

723: In general $\la {\cal L} | \Gamma^{\rm defect-L} | {\cal L} \ra$ is

724: expanded in terms of $1/L$ as

725: \begin{eqnarray}

726: \la {\cal L} | \Gamma^{\rm defect-L} | {\cal L}\ra 

727: = \lambda\left\{a_0 + a_1\frac{1}{L} + a_2 \frac{1}{L^2} 

728: \right\} + \mathcal{O}\left(\frac{1}{L^3}\right)\,,  

729: \end{eqnarray}

730: but under the conditions (\ref{relax}) the constants $a_0$ and $a_1$

731: vanish. In other words, the leading parts of (\ref{relax}) are 

732: determined so that we can take a regular BMN limit. 

733: Then the terms with higher order than $1/L^3$ also vanish

734: in the $L\rightarrow\infty$ limit. 

735: The second-order term gives only a non-vanishing contribution 

736: \begin{eqnarray}

737:   \langle {\cal L} | \Gamma^{\rm defect-L} | {\cal L} \rangle

738: &=&

739:  \frac{\lambda}{8 \pi^2} \times \frac{1}{2L^2} \bigg\{ (\delta\theta_1)^2 

740: + \sin^2 \theta_1 (\delta\phi_1)^2 \bigg\} \\

741: &=& \frac{1}{4 \pi^2} \frac{\tilde{\lambda}}{2} 

742: \bigg\{ |\delta \mbox{\boldmath{$m$}}_1|^2 - |\mbox{\boldmath{$m$}}^*_1 \cdot

743: \delta \mbox{\boldmath{$m$}}_1 |^2 \bigg\}\,.  

744: \label{HamiltLeft}

745: \end{eqnarray} 

746: This will be absorbed into the action obtained from the bulk

747: contribution. The result that the non-diagonal integrable boundary

748: terms give the same action as the bulk part would be rather

749: non-trivial, though it is physically quite natural.  

750: The same argument can be applied for $\langle {\cal R} |

751: \Gamma^{\rm defect-R} | {\cal R} \rangle $\,.

752: 

753: The Dirichlet and Neumann boundary conditions for open strings 

754: may be encoded into $\delta m_1$ in

755: (\ref{HamiltLeft})\,.  Those cannot be however determined before taking

756: the continuum limit because the $\sigma$-dependence of 

757: \mbox{\boldmath{$m$}} should be explicitly specified. 

758: It is necessary to take a variation of the continuum action 

759: in order to determine the boundary conditions. 

760: 

761: Finally we shall give a simple geometrical interpretation of the

762: conditions $a_0 = a_1=0$\,. 

763: %(\ref{vanish}). 

764: We focus on the boundary on the left hand

765: side since the analysis on the other side is similar. The key point is

766: that $n_0^K \equiv \langle\bar{q} | \sigma^K | \bar{q}\rangle $ should

767: be interpreted as the coordinates for the end-point of the open string

768: on S$^2$ \cite{Open-sol}. Since $(L^K)_{I_1 J_1} \equiv i

769: \epsilon_{I_1KJ_1}$ are the $SO(3)_{\rm H}$ generators, $n_1^K \equiv

770: \langle m_1 | L^K | m_1 \rangle $ should denote the nearest point from

771: the end-point. Then we find that $a_0=0$ leads to 

772: $1 - \mbox{\boldmath $n$}_0 \cdot \mbox{\boldmath $n$}_1 =0\,.$  

773: In the continuum limit, $\mbox{\boldmath $n$}_0$ and $\mbox{\boldmath

774: $n$}_1$ are very close and hence we can expand as $\mbox{\boldmath

775: $n$}_1 \simeq \mbox{\boldmath $n$}_0 + d\mbox{\boldmath $n$}_0$\,. Then

776: the condition $a_1=0$ leads to 

777: $ \mbox{\boldmath $n$}_0 \cdot d\mbox{\boldmath $n$}_0 = 0\,.$ 

778: Thus the conditions $a_0=a_1=0$ imply that the endpoints of the open

779: string should lie on the S$^2$-brane. 

780: 

781: 

782: \subsection{Derivation of Landau-Lifshitz type Sigma Model from the dCFT}

783: 

784: 

785: Now let us consider the expectation value of the open spin chain

786: Hamiltonian in terms of coherent states (\ref{m-coh}), (\ref{q-coh}). 

787: 

788: The expectation value of $D_{SO(6)}$ in the whole coherent states 

789: $|m \ra$ consists of three parts as 

790: \begin{eqnarray}

791: \la m | D_{SO(6)} | m \ra &=& \la m |\Gamma^{\rm bulk}_{\cal O} | m \ra 

792: + \langle{\cal L}|\Gamma_{\cal O}^{\rm defect-L} | {\cal L} \rangle

793: + \langle{\cal R}|\Gamma_{\cal O}^{\rm defect-R} | {\cal R} \rangle\,. 

794: \label{EVD6}

795: \end{eqnarray} 

796: Here the defect contributions have been already evaluated and the next

797: task is to evaluate the bulk contribution. Following the procedure in

798: \cite{ST-coh}, we shall introduce an antisymmetric $3 \times 3$ matrix

799: $m^{IJ}$ defined as 

800: \begin{eqnarray}

801:  m^{IJ}

802: \equiv

803:  \langle m | M^{IJ} | m \rangle = m^a{}^* M^{IJ}_{ab} m^b =  m^I{}^* {m^J} 

804: - m^I m^J{}^* \quad (I,J=1,2,3)\,. \label{mij}

805: \end{eqnarray} 

806: The expectation value of $H_{l,l+1}$ is evaluated as

807: \begin{eqnarray}

808:  \la m | H_{l,l+1} | m \ra &=&

809:  \frac{15}{16}  \sum_{I,J=1}^3 m_l^{IJ} m_{l+1}^{IJ} 

810: - \frac{1}{16}\sum_{I,J,K,L=1}^3 m_l^{IJ} m_l^{JK} m_{l+1}^{KL} m_{l+1}^{LI}

811: \nn \\

812: &=& \frac{15}{32} \mbox{Tr}(m_l - m_{l+1} )^2

813:  + \frac{1}{32} \mbox{Tr}[(m_l^2 - m_{l+1}^2)^2]\,.

814: \end{eqnarray} 

815: 

816: Then we consider a continuum limit $L\to \infty$ with $\tilde{\lambda}

817: \equiv \frac{\lambda}{L^2}$ fixed. 

818: We expand in powers of $\pi/L$ as

819: \begin{eqnarray}

820:  m^{IJ}(\sigma_{l+1}) = m^{IJ}(\sigma_l) + \frac{\pi}{L}\partial_\sigma 

821: m^{IJ} + \cdots\,, 

822: \end{eqnarray}

823: and keep the leading terms. The continuum limit is taken as 

824: \begin{eqnarray}

825:  \langle m | D_{SO(6)} | m \rangle

826: &\to& 

827:  \frac{\lambda}{16 \pi^2} L \int_0^\pi\!\!\frac{d\sigma}{\pi} 

828:  \left(\frac{\pi}{L}\right)^2 

829:  \mbox{Tr}\left[

830:  \frac{15}{32}(\partial_\sigma m)^2 

831:  + \frac{1}{16} m^2 (\partial_\sigma m)^2 + \frac{1}{16} 

832: (m \partial_\sigma m)^2\right] \nn \\

833: &\simeq& 

834:  L \int_0^\pi\!\!\frac{d \sigma}{\pi}\, \frac{\tilde{\lambda}}{32} 

835: \mbox{Tr}\left[ 

836: (\partial_\sigma m)^2 + \frac{1}{16}(m \partial_\sigma m)^2\right]\,,   

837: \label{EV6-2}

838: \end{eqnarray} 

839: where we have used $\mbox{Tr}(\partial_\sigma m)^2 = 2

840: \mbox{Tr}\left[m^2(\partial_\sigma m)^2\right] 

841: + \mbox{Tr}\left(m \partial_\sigma m\right)^2$\,. 

842: By substituting (\ref{mij}) into (\ref{EV6-2}),

843: \begin{eqnarray} 

844:  \langle m | D_{SO(6)} | m \rangle 

845: &\simeq& 

846:  \frac{1}{4} L \int_0^\pi\!\! \frac{d\sigma}{\pi}\,

847:  \frac{\tilde{\lambda}}{2} \Bigl\{

848: |\partial_\sigma \mbox{\boldmath $m$}|^2 -  

849: |\mbox{\boldmath $m$}{^*} \cdot \partial_\sigma \mbox{\boldmath $m$}|^2

850: \Bigr\} \nn \\

851: &=& - L \int_0^\pi\!\!\frac{d\sigma}{\pi}

852: \Bigl\{- \frac{1}{2}\cdot\frac{\tilde{\lambda}}{4} | D_\sigma \mbox{\boldmath $m$}|^2 \Bigr\} 

853: \equiv -L \int_0^\pi\!\!\frac{d\sigma}{\pi} 

854: \langle m |{\cal H}| m\rangle\,. \label{EV6-3}

855: \end{eqnarray}

856: The phase-space Lagrangian $\mathcal{L}$ is defined as 

857: \begin{eqnarray}

858:  \mathcal{L} \equiv - \la m | i\frac{d}{d t}|m \ra + \la m|{\cal H}|m\ra\,,  

859: \end{eqnarray}

860: and the resulting sigma model action is given by 

861: \begin{eqnarray}

862: I = L \int\!\! dt\! \int_0^\pi\!\!

863: \frac{d\sigma}{\pi}\, {\cal L}\,, \qquad {\cal L} = 

864: -i \mbox{\boldmath $m$}^* \cdot \partial_t \mbox{\boldmath $m$}

865:  - \frac{1}{2}\cdot\frac{\tilde{\lambda}}{4} 

866: |D_\sigma \mbox{\boldmath $m$}|^2

867: % + O(\tilde{\lambda}^2)

868: \,. \label{sigmaSYM}

869: \label{sym}

870: \end{eqnarray} 

871: In the next section we will reproduce the same result as (\ref{sigmaSYM}) 

872: from the string side.

873: 

874: \subsubsection*{Comment on the $SO(6)/SO(5)$ case} 

875: 

876: We may consider the case 2 related to $H=SO(5)$ in the unbroken $SO(6)$

877: case. When we follow the above analysis in this case, we cannot

878: completely delete the contribution of the additional anomalous dimension

879: because the $\epsilon_{IJK}$ terms vanish due to the reality of the

880: $SO(3)_{\rm H}$ coherent state (i.e. $\la m|_{I}\epsilon_{IJK} |m\ra_{K}

881: = 0$). Thus the Kronecker delta terms still remain even after taking the

882: expectation value of the spin chain Hamiltonian, and lead to the term

883: proportional to $\lambda$\,. This contribution is absorbed into the

884: large ``extensive'' one-loop shift that comes from the bulk

885: contribution, but the problem of the consistent continuum limit still

886: remains. After all, the problem of the defect interactions reduces to

887: that of the continuum limit.

888: 

889: 

890: \section{Landau-Lifshitz Type Sigma Model from the String Action}

891: 

892: \subsection{Derivation of Landau-Lifshitz Type Sigma Model}

893: 

894: We shall derive a Landau-Lifshitz type sigma model action as

895: an effective action of the open string moving almost at the speed of

896: light, following \cite{KT-coh}. In our case, open strings can move only

897: on the S$^2$ spanned by $X_{\rm H}^I\ (I=1,2,3)$ with

898: $\sum_{I=1}^3(X_{\rm H}^I)^2=1$. Hence  

899: we start from the bosonic part of the phase-space Lagrangian for an open

900: string on $R\times$S$^2$\,,  

901: %written in terms of $t$ (AdS time) and $X_{\rm H}^I$ is 

902: \begin{equation} \label{lak} 

903: {\cal L} = -\frac{1}{2}\kappa^2 + P_{\rm H}^I \dot{X_{\rm H}^{I}} 

904: - \frac{1}{2}P_{\rm H}^I P_{\rm H}^I 

905: - \frac{1}{2} {X_{\rm H}^I}{}' {X_{\rm H}^I}{}' - \frac{1}{2}

906: \Lambda (X_{\rm H}^I X_{\rm H}^I -1 )\,, 

907: \end{equation}

908: where the symbols ``$\cdot$'' and ``{\footnotesize $\prime$}'' mean the

909: derivatives with respect to the world-sheet coordinates $\tau$ and

910: $\sigma$, respectively. Here we have chosen the conformal gauge and the 

911: additional gauge-fixing condition $t=\kappa \tau$\,. 

912: 

913: Next we consider an effective action where the slowly-changing

914: coordinates only survive, and so we first isolate a fast coordinate. In

915: this process, it is inevitable to use the phase-space description.  Then

916: we take a limit in which the velocity of the fast coordinate is put to

917: infinity. Finally we get the following Landau-Lifshitz type effective

918: action

919: \begin{eqnarray} 

920: \label{vbvb} && I = L \int\! dt

921: \!\int_0^{\pi}\!\frac{d\sigma}{\pi}\,{\cal L}\,, \quad {\cal L} = - i

922: V^* \dot{V} - \frac{1}{2} |D_\sigma V|^2\,, 

923: \end{eqnarray} 

924: where $L= \sqrt{\lambda}\,l~(l\simeq \kappa)$ and the symbol ``$\cdot$''

925: implies the

926: derivative with respect to the time $t$, instead of $\tau$\,. The time

927: coordinate is rescaled as $t \rightarrow \kappa^2 t$ and the Lagrange

928: multiplier terms are omitted. In addition, when we rescale the time $t$

929: as $t \rightarrow \frac{1}{4}\tilde{\lambda}\,t$\,,\footnote{The factor

930: $1/4$ arises since $\tilde{\lambda}\equiv \lambda/L^2$ and 

931: the $L$ for closed strings is replaced by $2L$ for open

932: strings. } the Wess-Zumino term is kept invariant but the remaining part

933: is shifted as 

934: \begin{eqnarray}

935: I = L \int\! dt \!\int_0^{\pi}\!\frac{d\sigma}{\pi}\,{\cal L}\,, \quad

936: {\cal L} = - i V^* \dot{V} - \frac{1}{2}\cdot\frac{\tilde{\lambda}}{4}

937: |D_\sigma V|^2\,.

938: \label{5.10}

939: \end{eqnarray} 

940: We thus find a complete agreement of the sigma model actions (\ref{sym})

941: in SYM and (\ref{5.10}) in the string side, under the identification

942: of {\boldmath{$m$}} with $V$\,. 

943: 

944: In the closed string case, the Landau-Lifshitz type sigma model can also 

945: be derived from the approach of Mikhailov \cite{Mikhailov}.  It would

946: be an interesting practice to rederive the action in our open string

947: case, by following \cite{Mikhailov}.

948: 

949: 

950: \subsection{$SO(3)_{\rm H}$ Pulsating String Solution}

951: 

952: We can find a pulsating solution under a solution ansatz. Let us

953: parametrize the complex vector $V^I$ as 

954: \begin{equation}

955: V^I = \frac{a^I - ib^I}{\sqrt{2}}\,, \quad 

956: V^I {V^I}^* =1\,, \quad (V^I)^2 =0\,. 

957: \end{equation}

958: By setting the real vectors $a$ and $b$ as

959: \begin{eqnarray}

960: \label{ansatz} 

961: a = (\cos\th \cos\phi, -\cos\th\sin\phi,\sin\th)\,, \quad 

962: b = (- \sin\phi, -\cos\phi,0)\,,

963: \end{eqnarray}

964: the Lagrangian becomes 

965: \begin{eqnarray}

966: \mathcal{L} = \cos\th\, \dot{\phi} - \frac{\tilde{\lambda}}{16}

967: \left({\th'}^2 + \sin^2\th\, {\phi'}^2\right)\,,

968: \end{eqnarray}

969: in terms of two coordinates $\th$ and $\phi$ of S$^2$\,. The

970: pulsating solution is now described by taking a special solution $\th =

971: \pi/2$ and $\phi = m\sigma$ where $m$ is an integer\footnote{For more general pulsating string solutions, see \cite{EMZ,pulsating3}.}. The energy of the

972: pulsating string solution is obtained as

973: \begin{eqnarray}

974: E = \frac{1}{16}\tilde{\lambda}L\sin^2\th\,{\phi'}^2 

975: = \frac{\lambda}{16L}m^2\,. 

976: \end{eqnarray}

977: It completely agrees with the result of \cite{pulsating1} in the open

978: string case. Here we have utilized the doubling trick formula that is

979: discussed by Stefanski \cite{Open-sol},

980: \begin{eqnarray}

981: E_{\rm open}(L) = \frac{1}{2}E_{\rm closed}(2L)\,. 

982: \label{doubling}

983: \end{eqnarray}

984: Then the energy for the pulsating closed string solution is 

985: \begin{eqnarray}

986: E_{\rm closed} =  \frac{\lambda}{4L}m^2\,,

987: \end{eqnarray} 

988: and hence the doubling trick formula (\ref{doubling}) is surely

989: satisfied. 

990: 

991: We should remark that the pulsating closed string solution is also a

992: solution in our open string case by imposing Neumann and Dirichlet

993: conditions at the end-points: 

994: \begin{eqnarray}

995: \mbox{N}: \quad \partial_{\sigma}V(t,\sigma=0,\pi) = 0\,, \qquad 

996: \mbox{D}: \quad V(t,\sigma=0,\pi) = 0\,. 

997: \label{bound} 

998: \end{eqnarray}  

999: In the parametrization of (\ref{ansatz}) and the solution ansatz:

1000: $\th=\pi/2$ and $\phi = m\sigma$\,, $V^2$ and $V^3$ satisfy Neumann

1001: conditions, and $V^1$ does a Dirichlet one.  

1002: 

1003: 

1004: \subsection{Construction of Coherent Defect Operators}

1005: 

1006: Finally we shall comment on the forms of defect operators corresponding

1007: to open string solutions. The defect operators naturally associated to

1008: open semiclassical string solutions should be ``locally BPS'' coherent

1009: defect operators. Slightly generalizing the factorization ansatz in the closed

1010: string case \cite{KT-coh}, we suppose that the coefficient of

1011: (\ref{cop}) is decomposed as

1012: \begin{eqnarray}

1013:  \psi_{m I_1 I_2 \cdots I_L n} = \eta_m' m_1^{I_1} m_2^{I_2} \cdots 

1014: m_L^{I_L} \eta_n\,.

1015: \end{eqnarray} 

1016: Here $m_l$'s are the coherent states (\ref{m-coh}) and correspond to

1017: open string solutions under the identification of $\mbox{\boldmath $m$}$

1018: with $V$ in the continuum limit. 

1019: By assuming that $\eta'_m$ and $\eta_n$ are given as 

1020: $ (\eta_m') = |\bar{q} \rangle \quad \mbox{and} \quad (\eta_n) =

1021: |q\ra$\,, the corresponding operators are 

1022: \begin{eqnarray}

1023:  {\cal O} = [\eta_m' \bar{q}_m] \left[\prod_{l=1}^L m_l^{I_l} X_{\rm H}^{I_l}

1024: \right] [\eta_n q_n]\,. \label{localBPS}

1025: \end{eqnarray} 

1026: In particular, when we set $\theta_l=\phi_l=0$ for all $l$\,, 

1027: \begin{eqnarray}

1028:  (\eta_m') = \left[

1029: 	\begin{array}{@{\,}c@{\,}}

1030: 	 1 \\

1031: 	 0

1032:   \end{array} 

1033:  \right]\,,

1034: \quad 

1035:  (\eta_n) =

1036:  \left[

1037: 	\begin{array}{@{\,}c@{\,}}

1038: 	 0 \\

1039: 	 1

1040:   \end{array} 

1041:  \right]\,,

1042: \qquad 

1043:  (m_l^{I_l})

1044: =

1045:  \left[

1046:  \begin{array}{@{\,}c@{\,}}

1047:   1 \\

1048:   i \\

1049:   0

1050:  \end{array} 

1051:  \right]

1052: \quad (l=1,\cdots, L)\,, 

1053: \end{eqnarray} 

1054: we can consistently reproduce the open string BPS vacuum \cite{LP} 

1055: \begin{eqnarray}

1056:  {\cal O} = \bar{q}_1 Z^L q_2 

1057: \qquad ( Z \equiv X_{\rm H}^1 + i X_{\rm H}^2)\,. 

1058: \end{eqnarray}

1059: 

1060: In addition we may construct the operator corresponding to the pulsating

1061: open string solution discussed above by putting the following data into

1062: (\ref{localBPS})\,,

1063: \begin{eqnarray}

1064:  (\eta_m') =

1065:  \frac{i}{\sqrt{2}}

1066:  \left[

1067: 	\begin{array}{@{\,}c@{\,}}

1068: 	 1 \\

1069: 	 -1

1070:   \end{array} 

1071:  \right]\,,

1072: \quad 

1073:  (\eta_n)

1074: =

1075:   \frac{i^{m+1}}{\sqrt{2}}

1076:  \left[

1077: 	\begin{array}{@{\,}c@{\,}}

1078: 	 1 \\

1079: 	 (-1)^m

1080:   \end{array}

1081:  \right]\,,

1082: \quad 

1083:  (m_l^{I_l})

1084: =

1085:  \frac{e^{-im\sigma_l}}{\sqrt{2}}

1086:  \left[

1087:  \begin{array}{@{\,}c@{\,}}

1088:    i \sin(m \sigma_l) \\

1089:    i \cos(m \sigma_l)\\

1090:    1

1091:  \end{array}

1092:  \right]\,.

1093: \end{eqnarray} 

1094: Here a few examples have been discussed, but we need more efforts to

1095: confirm the ansatz (\ref{localBPS})\,.

1096: 

1097: \section{Other AdS-brane Cases}

1098: \label{general:app}

1099: 

1100: We have discussed the case that an AdS$_4\times$S$^2$-brane is inserted

1101: in the AdS$_5\times$S$^5$ background. By considering other AdS-branes,

1102: one may obtain other defect conformal field theories (For the

1103: classification of possible AdS-branes, see the work of Skenderis and

1104: Taylor \cite{ST} or an approach from $\kappa$-symmetry

1105: \cite{SY:string}). The possible 1/2 supersymmetric AdS-branes are

1106: summarized in Tab.\,\ref{brane:tab}. For all cases, the holographic

1107: duals are described by dCFTs. The feature of the defect fields is

1108: different for the number of Neumann-Dirichlet directions ($\sharp$ND).

1109: In the case of $\sharp$ND=4 intersections, the ``quark'' field $q$

1110: supplied by the defect is a complex scalar field, 

1111: while $q$ is a fermion field in the case of $\sharp$ND=8

1112: intersections.

1113: 

1114: \begin{table}[htbp]

1115: \begin{center} 

1116: \begin{tabular}{cc}

1117: \begin{tabular}{|c|c|c|}

1118: \hline 

1119: Brane & $\sharp$ND=4 & Embedding \\

1120: \hline\hline 

1121: D1 & (0$|$ D1 $\bot$ D3) & AdS$_{2}$ \\

1122: \hline 

1123: D3 & (1$|$ D3 $\bot$ D3) & AdS$_3\times$S$^1$ \\

1124: \hline  

1125: D5 & (2$|$ D5 $\bot$ D3) & AdS$_{4}\times$S$^{2}$ \\

1126: \hline 

1127: D7 & (3$|$ D7 $\bot$ D3) & AdS$_{5}\times$S$^{3}$ \\

1128: \hline

1129: \end{tabular} 

1130: \quad & \quad  

1131: \begin{tabular}{|c|c|c|} 

1132: \hline 

1133: Brane & $\sharp$ND=8 &  Embedding \\

1134: \hline\hline 

1135: D5 & (0$|$ D5 $\bot$ D3) & AdS$_2\times$S$^4$ \\

1136: \hline 

1137: D7 & (1$|$ D7 $\bot$ D3) & AdS$_3\times$S$^5$ \\

1138: \hline

1139:   \end{tabular} 

1140: \end{tabular} 

1141: \end{center}

1142: \caption{The possible configurations of 1/2 supersymmetric AdS-branes. }

1143: \label{brane:tab}

1144: \end{table}

1145: 

1146: When we focus upon the analysis of the S$^5$ part, open strings can live

1147: on the S$^n$ part $(n=2,3,4,5)$\footnote{The $n=0$ case has no part in

1148: S$^5$\,. For the $n=1$ case, the group $SO(2)_{\rm H}$ is abelian and the $H$

1149: becomes trivial. Then the coherent state leads to the BPS vacuum.  Hence

1150: we will not consider them here.} that is a part of the above-mentioned

1151: AdS-branes.  In the case of the S$^n$\,, we would need the $SO(n+1)_{\rm

1152: H}$ coherent states to evaluate the classical sigma model action by

1153: taking a continuum limit. We argue that the appropriate choice of the

1154: maximal stability subgroup $H$ and the vacuum $|0\ra$ would be

1155: \begin{eqnarray}

1156: H = SO(n-2) \times SO(2)\,, \qquad |0\ra = 

1157: (\underbrace{0,\ldots,0}_{n-1},1,i)\,. 

1158: \end{eqnarray}

1159: That is, the coherent states for general AdS-branes are constructed for

1160: the coset:

1161: \begin{eqnarray}

1162: \label{coset}

1163: SO(n+1)_{\rm H}/[SO(n-2) \times SO(2)]\,. 

1164: \end{eqnarray}

1165: By using the coherent state for the coset (\ref{coset}),  

1166: \begin{eqnarray}

1167: |m\ra = \exp\left[\sum_{i=1}^{n-1}(a_iM_{i5} + a_{i+4}M_{i6})\right]|0\ra\,,

1168: \end{eqnarray}

1169: with $|0\ra = (0,\ldots,0,1,i)$\,, we can evaluate the expectation value

1170: of the integrable open spin chain Hamiltonian $D_{SO(6)}$\,. Following

1171: the work of Stefanski and Tseytlin \cite{ST-coh}, let us introduce an

1172: antisymmetric imaginary $(n+1) \times (n+1)$ matrix $m^{ij}\equiv \la m

1173: | M^{ij}|m \ra~$\,. The only difference from the work \cite{ST-coh} is

1174: the range of the index $i$\,.  

1175: %Even in our case, the following formulae are satisfied:

1176: %\begin{eqnarray}

1177: %\sum_{i,j=1}^{n+1} m^{ij}m^{ji} = 2\,, \qquad \sum_{k,l=1}^{n+1}

1178: %m^{ik}m^{kl}m^{lj} = m^{ij}\,, 

1179: %\end{eqnarray}

1180: %and on the basis of the symmetry, 

1181: %\begin{eqnarray}

1182: %\la m | M^{ij} M^{kl}|m \ra &=& \frac{1}{2} \left(\del^{il}m^{jk} 

1183: %- \del^{ik}m^{jl} - \del^{jl}m^{ik} + \del^{jk}m^{il}\right) \nn \\

1184: %&& - \frac{1}{2}\left(\del^{il}w^{jk} - \del^{ik}w^{jl} -

1185: %		 \del^{jl}w^{ik} + \del^{jk}w^{il}\right)\,, 

1186: %\end{eqnarray}

1187: %where $w^{ij} \equiv \sum_{k=1}^{n+1}m^{ik}m^{kj}$\,. 

1188: The integrable

1189: boundary terms may possibly vanish according to our description and then

1190: the expectation value of $H_{l,l+1}$ is evaluated as follows:

1191: \begin{eqnarray}

1192: \la m|H_{l,l+1}|m \ra &=& \frac{1}{2}\left\{1 -

1193: \frac{1}{16}(n-1)\right\}{\rm Tr}(m_l - m_{l+1})^2 +

1194: \frac{1}{32}(n-1){\rm Tr}(m_l^{2} - m_{l+1}^2)^2\,.

1195: \end{eqnarray}

1196: After taking a continuum limit, we obtain the following

1197: classical sigma model action:

1198: \begin{eqnarray}

1199: \la m| D_{SO(6)}|m\ra \rightarrow 

1200: L\int^{\pi}_0\!\!\frac{d\sigma}{\pi}\,\frac{\tilde{\lambda}}{32}

1201: \left[

1202: {\rm Tr}(\partial_{\sigma}m)^2 + \frac{1}{16}(n-1)

1203: {\rm Tr}(m\partial_{\sigma}m)^2

1204: \right]\,, 

1205: \end{eqnarray}

1206: where we have used the formula: 

1207: \begin{eqnarray}

1208: {\rm Tr}[m^2(\partial_\sigma m)^2] = 

1209: \frac{1}{2}{\rm Tr}(\partial_{\sigma}m)^2 - \frac{1}{2}{\rm Tr}

1210: (m\partial_{\sigma}m)^2\,. 

1211: \end{eqnarray}

1212: Here we should notice the identity ${\rm Tr}(m\partial_{\sigma}m)^2 =0$

1213: and thus the resulting sigma model action is independent of $n$\,.  That

1214: is, the $n$-dependence appears at the quantum spin chain level but it

1215: disappears after taking the continuum limit. As the result, by

1216: introducing the complex unit vector $V^i$, we obtain the sigma model

1217: Lagrangian, 

1218: \begin{eqnarray}

1219: \mathcal{L} &=& -iV^{i\ast}\partial_{t}V^i 

1220: - \frac{1}{2}|D_{\sigma}V^i|^2\,. \label{sigma}

1221: \end{eqnarray}

1222: This expression is identical for each of AdS-branes,

1223: except the range of the index $i$\,. 

1224: 

1225: The sigma model action (\ref{sigma}) can be rederived from the string

1226: action in the same way as in section 5. Thus it may be argued that the

1227: complete agreement of the sigma model action would be valid for general

1228: AdS-branes as well as AdS$_4\times$S$^2$\,.  This fact would ensure that

1229: the supersymmetric AdS-branes may be included in studies of the

1230: correspondence between semiclassical strings and SYM operators.

1231: 

1232: Moreover, our consideration for coherent states may be generalized

1233: for some cases other than AdS-branes, for example, orbifold

1234: backgrounds \cite{BGMNN,Open-sol,CWW,Ideguchi}. A study in this direction

1235: is favorable and interesting.

1236: 

1237: 

1238: 

1239: \section{Conclusion and Discussion}

1240: 

1241: We have discussed the correspondence between long defect operators in

1242: the dCFT and open semiclassical strings. The coherent states for the

1243: integral open spin chain have been constructed. In particular, we have

1244: given a prescription to treat the boundary terms in the open spin chain

1245: Hamiltonian. By evaluating the expectation value of the Hamiltonian with

1246: the coherent states and taking the continuum limit, a Landau-Lifshitz

1247: (LL) type of sigma model has been obtained. This action has been also

1248: derived by starting from the string action with the appropriate gauge

1249: fixing and $\kappa\rightarrow\infty$\,.  In conclusion, for an open

1250: string case, we have found a complete agreement of the sigma model

1251: actions in both sides as in closed string cases. It has been also found

1252: that an $SO(3)_{\rm H}$ pulsating solution is included as a solution of

1253: the derived LL-type sigma model and its energy is completely identical

1254: to the result obtained in a different method \cite{pulsating1}. In

1255: addition, we have discussed the sigma model action in other general

1256: AdS-brane cases. When we consider other AdS-branes, the index range of

1257: the sigma model variable is different in each of cases according to

1258: S$^n$ $\subset$ S$^5$ ($n=2,\ldots,5$). However, we have seen that this

1259: difference may not affect on the resulting sigma model action and thus

1260: we argue that the similar LL-type sigma model would be obtained in other

1261: AdS-brane cases. 

1262: 

1263: We have discussed the one-loop dilatation operator for defect operators

1264: composed of the $SO(6)$ scalar fields. It is an interesting future work

1265: to investigate at two-loop level. The $SO(6)$ sector is closed at

1266: one-loop level, but it is not the case at two-loop level. That is, the

1267: fields belonging to other sectors may mix into the $SO(6)$ scalar

1268: sector. The mixing of fermions occurs at two-loop level, but recently an

1269: interesting possibility has been proposed by Minahan

1270: \cite{Minahan:higher}. Following the idea of \cite{Minahan:higher}, it

1271: may be suppressed in taking $L\rightarrow\infty$ limit. Thus, the

1272: two-loop integrability would possibly hold in this limit.  It is

1273: important to calculate the two-loop defect interaction in the system we

1274: considered and check the consistency to the doubling trick in the string

1275: side at two-loop level. We will work in this direction and report the

1276: result in the near future \cite{future}.

1277: 

1278: It is also an interesting subject to clarify the connection between the

1279: integrability and supersymmetric D-branes (i.e., BPS conditions). The

1280: relation between BPS conditions and rotating string solutions is

1281: discussed in \cite{MMT,BCM}. A study in this direction would be helpful

1282: to investigate such an issue. In addition, it is interesting to clarify

1283: the relationship of the integrabilities (Yangian symmetries) of the

1284: Landau-Lifshitz type sigma models in closed and open cases (For a closed

1285: case, see \cite{AR}).

1286: 

1287: 

1288: 

1289: \section*{Acknowledgments}

1290: 

1291: We would like to thank H.~Fuji, M.~Hatsuda, H.~Kawamura, D.~Tomino and

1292: A.~Yamaguchi for useful discussion.  The work of K.~Y.\ is supported in

1293: part by JSPS Research Fellowships for Young Scientists.

1294: 

1295: \begin{thebibliography}{99}

1296: 

1297: \bibitem{M}

1298: J.~M.~Maldacena, 

1299: ``The large N limit of superconformal field theories and supergravity,''

1300: Adv.\ Theor.\ Math.\ Phys.\  {\bf 2} (1998) 231

1301: [Int.\ J.\ Theor.\ Phys.\  {\bf 38} (1999) 1113] [arXiv:hep-th/9711200]; 

1302: S.~S.~Gubser, I.~R.~Klebanov and A.~M.~Polyakov,

1303: ``Gauge theory correlators from non-critical string theory,''

1304: Phys.\ Lett.\ B {\bf 428} (1998) 105 [arXiv:hep-th/9802109]; 

1305: E.~Witten, ``Anti-de Sitter space and holography,''

1306: Adv.\ Theor.\ Math.\ Phys.\  {\bf 2} (1998) 253 [arXiv:hep-th/9802150].

1307: 

1308: \bibitem{BMN}

1309: D.~Berenstein, J.~M.~Maldacena and H.~Nastase,

1310: ``Strings in flat space and pp waves from N = 4 super Yang Mills,''

1311: JHEP {\bf 0204} (2002) 013 [arXiv:hep-th/0202021].

1312: 

1313: \bibitem{P} R.~Penrose, ``Any spacetime has a plane wave as a limit,''

1314: Differential geometry and relativity, Reidel, Dordrecht, 1976,

1315: pp.~271-275. \\ R.~Gueven, ``Plane wave limits and T-duality,'' Phys.\

1316: Lett.\ B {\bf 482} (2000) 255 [arXiv:hep-th/0005061].

1317: 

1318: \bibitem{BFHP1}

1319: M.~Blau, J.~Figueroa-O'Farrill, C.~Hull and G.~Papadopoulos,

1320: ``A new maximally supersymmetric background of IIB superstring theory,''

1321: JHEP {\bf 0201} (2002) 047 [arXiv:hep-th/0110242].

1322: 

1323: \bibitem{MT}

1324: R.~R.~Metsaev,

1325: ``Type IIB Green-Schwarz superstring in plane wave Ramond-Ramond  background,''

1326: Nucl.\ Phys.\ B {\bf 625} (2002) 70 [arXiv:hep-th/0112044]; 

1327: R.~R.~Metsaev and A.~A.~Tseytlin,

1328: ``Exactly solvable model of superstring in plane wave Ramond-Ramond

1329: background,'' Phys.\ Rev.\ D {\bf 65} (2002) 126004 [arXiv:hep-th/0202109].

1330: 

1331: \bibitem{GKP2}

1332: S.~S.~Gubser, I.~R.~Klebanov and A.~M.~Polyakov,

1333: ``A semi-classical limit of the gauge/string correspondence,''

1334: Nucl.\ Phys.\ B {\bf 636} (2002) 99 [arXiv:hep-th/0204051].

1335: 

1336: \bibitem{FT}

1337: S.~Frolov and A.~A.~Tseytlin,

1338: ``Semiclassical quantization of rotating superstring in AdS$_5\times$S$^5$,''

1339: JHEP {\bf 0206} (2002) 007 [arXiv:hep-th/0204226]. 

1340: 

1341: \bibitem{BPR} 

1342: G.~Mandal, N.~V.~Suryanarayana and S.~R.~Wadia,

1343: ``Aspects of semiclassical strings in AdS$_5$,'' Phys.\ Lett.\ B {\bf

1344: 543} (2002) 81 [arXiv:hep-th/0206103], \\

1345: I.~Bena, J.~Polchinski and

1346: R.~Roiban, ``Hidden symmetries of the AdS$_5\times$S$^5$ superstring,''

1347: Phys.\ Rev.\ D {\bf 69} (2004) 046002 [arXiv:hep-th/0305116], \\ 

1348: L.~F.~Alday, ``Nonlocal charges on AdS$_5\times$S$^5$ and pp-waves,''

1349: JHEP {\bf 0312} (2003) 033 [arXiv:hep-th/0310146], \\

1350: M.~Hatsuda and K.~Yoshida, ``Classical integrability and super Yangian

1351: of superstring on AdS$_5\times$S$^5$,'' arXiv:hep-th/0407044. 

1352: 

1353: \bibitem{DNW}

1354: L.~Dolan, C.~R.~Nappi and E.~Witten, ``A relation between approaches to

1355: integrability in superconformal Yang-Mills theory,'' JHEP {\bf 0310}

1356: (2003) 017 [arXiv:hep-th/0308089]; ``Yangian symmetry in D = 4

1357: superconformal Yang-Mills theory,'' arXiv:hep-th/0401243. 

1358: 

1359: \bibitem{AS}

1360: G.~Arutyunov and M.~Staudacher,

1361: ``Matching higher conserved charges for strings and spins,''

1362: JHEP {\bf 0403} (2004) 004 [arXiv:hep-th/0310182]; 

1363: ``Two-loop commuting charges and the string / gauge duality,''

1364: arXiv:hep-th/0403077. 

1365: 

1366: \bibitem{FT2} S.~Frolov and A.~A.~Tseytlin, ``Multi-spin string

1367: solutions in AdS$_5\times$S$^5$,'' Nucl.\ Phys.\ B {\bf 668} (2003) 77

1368: [arXiv:hep-th/0304255]; ``Quantizing three-spin string solution in

1369: AdS$_5\times$S$^5$,'' JHEP {\bf 0307} (2003) 016 [arXiv:hep-th/0306130];

1370: ``Rotating string solutions: AdS/CFT duality in non-supersymmetric

1371: sectors,'' Phys.\ Lett.\ B {\bf 570} (2003) 96 [arXiv:hep-th/0306143].

1372: 

1373: \bibitem{ART}

1374: G.~Arutyunov, S.~Frolov, J.~Russo and A.~A.~Tseytlin,

1375: ``Spinning strings in AdS$_5$$\times$S$^5$ and integrable systems,''

1376: Nucl.\ Phys.\ B {\bf 671} (2003) 3 [arXiv:hep-th/0307191]; 

1377: G.~Arutyunov, J.~Russo and A.~A.~Tseytlin,

1378: ``Spinning strings in AdS$_5$$\times$S$^5$: 

1379: New integrable system relations,''

1380: Phys.\ Rev.\ D {\bf 69} (2004) 086009 [arXiv:hep-th/0311004].

1381: 

1382: \bibitem{review}

1383: A.~A.~Tseytlin,``Spinning strings and AdS/CFT duality,'' 

1384: arXiv:hep-th/0311139. 

1385: 

1386: \bibitem{MZ}

1387: J.~A.~Minahan and K.~Zarembo,

1388: ``The Bethe-ansatz for N = 4 super Yang-Mills,''

1389: JHEP {\bf 0303} (2003) 013 [arXiv:hep-th/0212208]. 

1390: 

1391: \bibitem{BKS}

1392: N.~Beisert, C.~Kristjansen and M.~Staudacher,

1393: ``The dilatation operator of N = 4 super Yang-Mills theory,''

1394: Nucl.\ Phys.\ B {\bf 664} (2003) 131 [arXiv:hep-th/0303060].

1395: 

1396: \bibitem{BFST}

1397: N.~Beisert, S.~Frolov, M.~Staudacher and A.~A.~Tseytlin,

1398: ``Precision spectroscopy of AdS/CFT,''

1399: JHEP {\bf 0310} (2003) 037 [arXiv:hep-th/0308117].

1400: 

1401: \bibitem{BMSZ}

1402: N.~Beisert, J.~A.~Minahan, M.~Staudacher and K.~Zarembo,

1403: ``Stringing spins and spinning strings,''

1404: JHEP {\bf 0309} (2003) 010 [arXiv:hep-th/0306139].

1405: 

1406: \bibitem{Kazakov}

1407: V.~A.~Kazakov, A.~Marshakov, J.~A.~Minahan and K.~Zarembo,

1408: ``Classical / quantum integrability in AdS/CFT,''

1409: JHEP {\bf 0405} (2004) 024 [arXiv:hep-th/0402207]. 

1410: 

1411: \bibitem{SU(3)}

1412: C.~Kristjansen, ``Three-spin strings on AdS$_5\times$S$^5$ from N = 4 SYM,''

1413: Phys.\ Lett.\ B {\bf 586} (2004) 106 [arXiv:hep-th/0402033]; 

1414: L.~Freyhult, ``Bethe ansatz and fluctuations in SU(3) Yang-Mills operators,''

1415: JHEP {\bf 0406} (2004) 010 [arXiv:hep-th/0405167]; 

1416: C.~Kristjansen and T.~Mansson,

1417: ``The circular, elliptic three-spin string from the SU(3) spin chain,''

1418: Phys.\ Lett.\ B {\bf 596} (2004) 265 [arXiv:hep-th/0406176].

1419: 

1420: \bibitem{SL2}

1421: A.~V.~Belitsky, A.~S.~Gorsky and G.~P.~Korchemsky,

1422: ``Gauge / string duality for QCD conformal operators,''

1423: Nucl.\ Phys.\ B {\bf 667} (2003) 3

1424: [arXiv:hep-th/0304028]; 

1425: A.~V.~Belitsky, S.~E.~Derkachov, G.~P.~Korchemsky and A.~N.~Manashov,

1426: ``Quantum integrability in (super) Yang-Mills theory on the light-cone,''

1427: Phys.\ Lett.\ B {\bf 594} (2004) 385

1428: [arXiv:hep-th/0403085], 

1429: %A.~V.~Belitsky, S.~E.~Derkachov, G.~P.~Korchemsky and A.~N.~Manashov,

1430: ``Dilatation operator in (super-)Yang-Mills theories on the light-cone,''

1431: arXiv:hep-th/0409120.

1432: 

1433: \bibitem{QCD}

1434: V.~M.~Braun, S.~E.~Derkachov and A.~N.~Manashov,

1435: ``Integrability of three-particle evolution equations in {QCD},''

1436: Phys.\ Rev.\ Lett.\  {\bf 81} (1998) 2020

1437: [arXiv:hep-ph/9805225]; 

1438: S.~E.~Derkachov, G.~P.~Korchemsky and A.~N.~Manashov,

1439: ``Evolution equations for quark gluon distributions in multi-color QCD 

1440: and open spin chains,''

1441: Nucl.\ Phys.\ B {\bf 566} (2000) 203

1442: [arXiv:hep-ph/9909539] and references therein. 

1443: 

1444: \bibitem{BS}

1445: N.~Beisert and M.~Staudacher, ``The N = 4 SYM integrable super spin chain,''

1446: Nucl.\ Phys.\ B {\bf 670} (2003) 439 [arXiv:hep-th/0307042]. 

1447: 

1448: \bibitem{KZ}

1449: V.~A.~Kazakov and K.~Zarembo,

1450: ``Classical/quantum integrability in non-compact sector of AdS/CFT,''

1451: arXiv:hep-th/0410105.

1452: 

1453: \bibitem{su(2|3)}

1454: N.~Beisert, ``The su(2$|$3) dynamic spin chain,''

1455: Nucl.\ Phys.\ B {\bf 682} (2004) 487 [arXiv:hep-th/0310252].

1456: 

1457: \bibitem{SS}

1458: D.~Serban and M.~Staudacher,

1459: ``Planar N = 4 gauge theory and the Inozemtsev long range spin chain,''

1460: JHEP {\bf 0406} (2004) 001 [arXiv:hep-th/0401057].

1461: 

1462: \bibitem{BDS}

1463: N.~Beisert, V.~Dippel and M.~Staudacher,

1464: ``A novel long range spin chain and planar N = 4 super Yang-Mills,''

1465: JHEP {\bf 0407} (2004) 075 [arXiv:hep-th/0405001].

1466: 

1467: \bibitem{quantum} 

1468: G.~Arutyunov, S.~Frolov and M.~Staudacher, 

1469: ``Bethe ansatz for quantum strings,'' 

1470: JHEP {\bf 0410} (2004) 016 [arXiv:hep-th/0406256]; \\ 

1471: N.~Beisert, ``Spin Chain for Quantum Strings,'' 

1472: arXiv:hep-th/0409054; \\

1473: M.~Staudacher, ``The factorized S-matrix of CFT/AdS,''

1474:   arXiv:hep-th/0412188.

1475: 

1476: \bibitem{KR}

1477: A.~Karch and L.~Randall, ``Localized gravity in string theory,''

1478: Phys.\ Rev.\ Lett.\  {\bf 87} (2001) 061601 [arXiv:hep-th/0105108]; 

1479: ``Open and closed string interpretation of SUSY CFT's on branes 

1480: with boundaries,'' JHEP {\bf 0106} (2001) 063 [arXiv:hep-th/0105132].

1481: 

1482: \bibitem{DFO}

1483: O.~DeWolfe, D.~Z.~Freedman and H.~Ooguri,

1484: ``Holography and defect conformal field theories,''

1485: Phys.\ Rev.\ D {\bf 66} (2002) 025009 [arXiv:hep-th/0111135]. 

1486: 

1487: \bibitem{EGK}

1488: J.~Erdmenger, Z.~Guralnik and I.~Kirsch,

1489: ``Four-dimensional superconformal theories with interacting boundaries 

1490: or defects,'' Phys.\ Rev.\ D {\bf 66} (2002) 025020 [arXiv:hep-th/0203020].

1491: 

1492: \bibitem{LP}

1493: P.~Lee and J.~Park,

1494: ``Open strings in PP-wave background from defect conformal field theory,''

1495: Phys.\ Rev.\ D {\bf 67} (2003) 026002 [arXiv:hep-th/0203257].

1496: 

1497: \bibitem{DM}

1498: O.~DeWolfe and N.~Mann,

1499: ``Integrable open spin chains in defect conformal field theory,''

1500: JHEP {\bf 0404} (2004) 035 [arXiv:hep-th/0401041]. 

1501: 

1502: \bibitem{BGMNN}

1503: D.~Berenstein, E.~Gava, J.~M.~Maldacena, K.~S.~Narain and H.~Nastase,

1504: ``Open strings on plane waves and their Yang-Mills duals,''

1505: arXiv:hep-th/0203249.

1506: 

1507: \bibitem{BHLN}

1508: V.~Balasubramanian, M.~Huang, T.~S.~Levi and A.~Naqvi,

1509: ``Open strings from N = 4 super Yang-Mills,''

1510: JHEP {\bf 0208} (2002) 037 [arXiv:hep-th/0204196]. 

1511: 

1512: \bibitem{Open-sol}

1513: B.~Stefanski, Jr., ``Open spinning strings,''

1514: JHEP {\bf 0403} (2004) 057 [arXiv:hep-th/0312091].

1515: 

1516: \bibitem{CWW}

1517: B.~Chen, X.~J.~Wang and Y.~S.~Wu,

1518: ``Integrable open spin chain in super Yang-Mills 

1519: and the plane-wave / SYM duality,''

1520: JHEP {\bf 0402} (2004) 029 [arXiv:hep-th/0401016]; 

1521: ``Open spin chain and open spinning string,''

1522: Phys.\ Lett.\ B {\bf 591} (2004) 170 [arXiv:hep-th/0403004]. 

1523: 

1524: \bibitem{B}

1525: N.~Beisert,

1526: ``The complete one-loop dilatation operator of N = 4 super Yang-Mills theory,''

1527: Nucl.\ Phys.\ B {\bf 676} (2004) 3 [arXiv:hep-th/0307015]. 

1528: 

1529: \bibitem{Kruczenski}

1530: M.~Kruczenski,``Spin chains and string theory,'' arXiv:hep-th/0311203. 

1531: 

1532: \bibitem{KRT} M.~Kruczenski, A.~V.~Ryzhov and A.~A.~Tseytlin, ``Large

1533: spin limit of AdS$_5$$\times$S$^5$ string theory and low energy

1534: expansion of ferromagnetic spin chains,'' Nucl.\ Phys.\ B {\bf 692}

1535: (2004) 3 [arXiv:hep-th/0403120].

1536: 

1537: \bibitem{coherent} 

1538: A.~Perelomov, ``Generalized Coherent States and Their Applications,'' 

1539: Berlin, Germany: Springer (1986) 320 p.; 

1540: W.~M.~Zhang, D.~H.~Feng and R.~Gilmore, ``Coherent States: Theory 

1541: And Some Applications,'' Rev.\ Mod.\ Phys.\ {\bf 62} (1990) 867. 

1542: 

1543: 

1544: \bibitem{ST-coh}

1545: B.~Stefanski, Jr., and A.~A.~Tseytlin,

1546: ``Large spin limits of AdS/CFT and generalized Landau-Lifshitz equations,''

1547: JHEP {\bf 0405} (2004) 042 [arXiv:hep-th/0404133]. 

1548: 

1549: \bibitem{KT-coh}

1550: M.~Kruczenski and A.~A.~Tseytlin, 

1551: ``Semiclassical relativistic strings in S$^5$ and long coherent 

1552: operators in N = 4 SYM theory,'' arXiv:hep-th/0406189. 

1553: 

1554: \bibitem{review-coh}

1555: A.~A.~Tseytlin,

1556: ``Semiclassical strings in AdS$_5\times$S$^5$ and scalar operators 

1557: in N = 4 SYM theory,'' arXiv:hep-th/0407218; 

1558: ``Semiclassical strings and AdS/CFT,'' arXiv:hep-th/0409296. 

1559: 

1560: \bibitem{HL}

1561: R.~Hernandez and E.~Lopez,

1562: ``The SU(3) spin chain sigma model and string theory,''

1563: JHEP {\bf 0404} (2004) 052 [arXiv:hep-th/0403139].

1564: 

1565: \bibitem{pulsating1}

1566: J.~A.~Minahan, 

1567: ``Circular semiclassical string solutions on AdS$_5\times$S$^5$,''

1568: Nucl.\ Phys.\ B {\bf 648} (2003) 203 [arXiv:hep-th/0209047]. 

1569: 

1570: \bibitem{EMZ}

1571: J.~Engquist, J.~A.~Minahan and K.~Zarembo,

1572: ``Yang-Mills duals for semiclassical strings on AdS$_5\times$S$^5$,''

1573: JHEP {\bf 0311} (2003) 063 [arXiv:hep-th/0310188]. 

1574: 

1575: \bibitem{pulsating3}

1576: H.~Dimov and R.~C.~Rashkov, ``Generalized pulsating strings,''

1577: JHEP {\bf 0405} (2004) 068 [arXiv:hep-th/0404012]; 

1578: M.~Smedback, ``Pulsating strings on AdS$_5\times$S$^5$,''

1579: JHEP {\bf 0407} (2004) 004 [arXiv:hep-th/0405102].

1580: 

1581: \bibitem{Mikhailov} A.~Mikhailov,``Speeding strings,'' JHEP {\bf 0312}

1582: (2003) 058 [arXiv:hep-th/0311019]; ``Slow evolution of nearly-degenerate

1583: extremal surfaces,'' arXiv:hep-th/0402067; ``Supersymmetric

1584: null-surfaces,'' arXiv:hep-th/0404173; ``Notes on fast moving strings,''

1585: arXiv:hep-th/0409040.

1586: 

1587: \bibitem{ST}

1588: K.~Skenderis and M.~Taylor, ``Branes in AdS and pp-wave spacetimes,''

1589: JHEP {\bf 0206} (2002) 025 [arXiv:hep-th/0204054].

1590: 

1591: \bibitem{SY:string}

1592: M.~Sakaguchi and K.~Yoshida, ``D-branes of covariant AdS superstrings,''

1593: Nucl.\ Phys.\ B {\bf 684} (2004) 100 [arXiv:hep-th/0310228]; 

1594: ``Notes on D-branes of type IIB string on AdS$_5\times$S$^5$,''

1595: Phys.\ Lett.\ B {\bf 591} (2004) 318 [arXiv:hep-th/0403243].

1596: 

1597: \bibitem{Ideguchi}

1598: K.~Ideguchi, ``Semiclassical strings on AdS$_5\times$S$^5$/Z$_M$ 

1599: and operators in orbifold field theories,''

1600: JHEP {\bf 0409} (2004) 008 [arXiv:hep-th/0408014]. 

1601: 

1602: \bibitem{Minahan:higher}

1603: J.~A.~Minahan,``Higher loops beyond the SU(2) sector,''

1604: arXiv:hep-th/0405243. 

1605: 

1606: \bibitem{future}

1607: Y.~Susaki, Y.~Takayama and K.~Yoshida, to appear. 

1608: 

1609: \bibitem{MMT}

1610: D.~Mateos, T.~Mateos and P.~K.~Townsend,

1611: ``Supersymmetry of tensionless rotating strings in AdS$_5\times$S$^5$, and

1612: nearly-BPS operators,'' JHEP {\bf 0312} (2003) 017 [arXiv:hep-th/0309114]; 

1613: ``More on supersymmetric tensionless rotating strings in AdS$_5\times$S$^5$,''

1614: arXiv:hep-th/0401058. 

1615: 

1616: \bibitem{BCM}

1617: F.~Bigazzi, A.~L.~Cotrone and L.~Martucci,

1618: ``Semiclassical spinning strings and confining gauge theories,''

1619: Nucl.\ Phys.\ B {\bf 694}, 3 (2004) [arXiv:hep-th/0403261]. 

1620: 

1621: \bibitem{AR}

1622: A.~Agarwal and S.~G.~Rajeev,

1623: ``Yangian symmetries of matrix models and spin chains: 

1624: The dilatation operator of N = 4 SYM,'' arXiv:hep-th/0409180.

1625: 

1626: 

1627: \end{thebibliography}

1628: 

1629: \end{document}

1630: