1: \documentclass[12pt]{article}

2: \usepackage{amssymb}

3: \usepackage{graphics}

4: \usepackage{psfrag}

5: \usepackage{epsfig}

6: 

7: \def\baselinestretch{1.2}

8: \parskip 6 pt

9: 

10: \textheight 22.3 cm

11: %%\topmargin -.5 cm

12: %\topmargin -.8 cm

13: \topmargin -1.5cm

14: \textwidth 16cm

15: \oddsidemargin 0 in

16: \evensidemargin 0 in

17: 

18: 

19: 

20: \DeclareFontFamily{U}{rsf}{}

21: \DeclareFontShape{U}{rsf}{m}{n}{

22:   <5> <6> rsfs5 <7> <8> <9> rsfs7 <10-> rsfs10}{}

23: \DeclareMathAlphabet\Scr{U}{rsf}{m}{n}

24: 

25: %

26: %

27: % Putting citations together

28: %

29: \catcode`\@=11

30: %------------

31: %NAME: cite.sty

32: %

33: %              Replacement for \@citex

34: %

35: %     Allow, but strongly discourage, line breaks within a long

36: %     series of citations.  Compress lists of successive numbers

37: %     to one number range, e.g., 5,6,7,8,9 --> 5--9.  Compatible

38: %     with versions of \@cite that use exponents.

39: %              -- Donald Arseneau  1989

40: %

41: \def\@citex[#1]#2{%

42: \if@filesw \immediate \write \@auxout {\string \citation {#2}}\fi

43: \@tempcntb\m@ne \let\@h@ld\relax \def\@citea{}%

44: \@cite{%

45:   \@for \@citeb:=#2\do {%

46:     \@ifundefined {b@\@citeb}%

47:       {\@h@ld\@citea\@tempcntb\m@ne{\bf ?}%

48:       \@warning {Citation `\@citeb ' on page \thepage \space undefined}}%

49: %   else

50:       {\@tempcnta\@tempcntb \advance\@tempcnta\@ne%

51:       \@tempcntb\number\csname b@\@citeb \endcsname \relax%

52:       \ifnum\@tempcnta=\@tempcntb %Number follows previous--hold on to it

53:         \ifx\@h@ld\relax%

54: %         % first pair of successive

55:           \edef \@h@ld{\@citea\csname b@\@citeb\endcsname}%

56:         \else%

57: %         % compressible list of successive

58:           \edef\@h@ld{\ifmmode{-}\else--\fi\csname b@\@citeb\endcsname}%

59:         \fi%

60:       \else%   %  non-successor--dump what's held and do this one

61:         \@h@ld\@citea\csname b@\@citeb \endcsname%

62:         \let\@h@ld\relax%

63:       \fi}%

64:     \def\@citea{,\penalty\@highpenalty\,}%

65:   }\@h@ld

66: }{#1}}

67: 

68: %    \@cite macro using []'s:

69: %

70: \def\@citeb#1#2{{[#1]\if@tempswa , #2\fi}}

71: %

72: %    \@cite macro using exponents:

73: %

74: \def\@citeu#1#2{{$^{#1}$\if@tempswa , #2\fi }}

75: %

76: %    \@cite macro with nothing:

77: %

78: \def\@citep#1#2{{#1\if@tempswa , #2\fi}}

79: 

80: %

81: %   Style of citations:

82: %

83: \def\bcites{         % cite with []'s

84:         \catcode`\@=11

85:         \let\@cite=\@citeb

86:         \catcode`\@=12

87: }

88: 

89: \def\upcites{         % cite with exponents

90:         \catcode`\@=11

91:         \let\@cite=\@citeu

92:         \catcode`\@=12

93: }

94: 

95: \def\plaincites{      % cite without brackets

96:         \catcode`\@=11

97:         \let\@cite=\@citep

98:         \catcode`\@=12

99: }

100: 

101: %

102: %   Draft stuff

103: %

104: 

105: \newcount\hour

106: \newcount\minute

107: \newtoks\amorpm

108: \hour=\time\divide\hour by 60

109: \minute=\time{\multiply\hour by 60 \global\advance\minute by-\hour}

110: \edef\standardtime{{\ifnum\hour<12 \global\amorpm={am}%

111:         \else\global\amorpm={pm}\advance\hour by-12 \fi

112:         \ifnum\hour=0 \hour=12 \fi

113:         \number\hour:\ifnum\minute<10 0\fi\number\minute\the\amorpm}}

114: \edef\militarytime{\number\hour:\ifnum\minute<10 0\fi\number\minute}

115: 

116: \def\draftlabel#1{{\@bsphack\if@filesw {\let\thepage\relax

117:    \xdef\@gtempa{\write\@auxout{\string

118:       \newlabel{#1}{{\@currentlabel}{\thepage}}}}}\@gtempa

119:    \if@nobreak \ifvmode\nobreak\fi\fi\fi\@esphack}

120:         \gdef\@eqnlabel{#1}}

121: \def\@eqnlabel{}

122: \def\@vacuum{}

123: \def\marginnote#1{}

124: \def\draftmarginnote#1{\marginpar{\raggedright\scriptsize\tt#1}}

125: \overfullrule=0pt

126: 

127: \def\draft{

128:         \pagestyle{plain}

129:         \overfullrule=2pt

130:         \oddsidemargin -.5truein

131:         \def\@oddhead{\sl \phantom{\today\quad\militarytime} \hfil

132:         \smash{\Large\sl DRAFT} \hfil \today\quad\militarytime}

133:         \let\@evenhead\@oddhead

134:         \let\label=\draftlabel

135:         \let\marginnote=\draftmarginnote

136:         \def\ps@empty{\let\@mkboth\@gobbletwo

137:         \def\@oddfoot{\hfil \smash{\Large\sl DRAFT} \hfil}

138:         \let\@evenfoot\@oddhead}

139:         \def\@eqnnum{(\theequation)\rlap{\kern\marginparsep\tt\@eqnlabel}%

140:         \global\let\@eqnlabel\@vacuum}  }

141: 

142: 

143: 

144: %

145: %    Personal choice here

146: %

147: \def\section{\@startsection {section}{1}{\z@}{3.ex plus 1ex minus

148:  .2ex}{2.ex plus .2ex}{\Large\bf}}

149: \def\subsection{\@startsection{subsection}{2}{\z@}{2.75ex plus 1ex minus

150:  .2ex}{1.5ex plus .2ex}{\large\bf}}        

151: \def\subsect#1{\par\penalty1000{\noindent \bf #1}\par\penalty500}

152: \def\appendix{{\newpage\section*{Appendix}}\let\appendix\section%

153:         {\setcounter{section}{0}

154:         \gdef\thesection{\Alph{section}}}\section}

155: \def\thefootnote{\arabic{footnote}}

156: \def\abstract{\if@twocolumn

157: \section*{Abstract}

158: \else %\small

159: \begin{center}

160: {\bf Abstract\vspace{-.5em}\vspace{0pt}}

161: \end{center}

162: \quotation

163: \fi}

164: 

165: 

166: %

167: % Table of contents

168: %

169: \setcounter{tocdepth}{3}

170: \renewcommand*\l@section[2]{%

171:   \ifnum \c@tocdepth >\z@

172:     \addpenalty\@secpenalty

173:     \addvspace{.0em \@plus\p@}%

174:     \setlength\@tempdima{1.5em}%

175:     \begingroup

176:       \parindent \z@ \rightskip \@pnumwidth

177:       \parfillskip -\@pnumwidth

178:       \leavevmode \bfseries

179:       \advance\leftskip\@tempdima

180:       \hskip -\leftskip

181:       #1\nobreak\hfil \nobreak\hb@xt@\@pnumwidth{\hss #2}\par

182:     \endgroup

183:   \fi}

184: \renewcommand*\l@subsection{\addvspace{-.5em \@plus\p@}\@dottedtocline{2}{1.5em}{2.3em}}

185: \renewcommand*\l@subsubsection{\addvspace{-.5em \@plus\p@}\@dottedtocline{3}{3.8em}{3.2em}}

186: %

187: %

188: %

189: 

190: 

191: \catcode`\@=12

192: 

193: 

194: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

195: \newcommand{\beq}{\begin{equation}}

196: \newcommand{\eeq}{\end{equation}}

197: \newcommand{\beqa}{\begin{eqnarray}}

198: \newcommand{\eeqa}{\end{eqnarray}}

199: \newcommand{\dd}{{\rm d}}

200: \newcommand{\A}{{\mathbb A}}

201: \newcommand{\N}{{\mathbb N}}

202: \newcommand{\NN}{{\mathbb N}}

203: \newcommand{\Z}{{\mathbb Z}}

204: \newcommand{\ZZ}{{\mathbb Z}}

205: \newcommand{\Q}{{\mathbb Q}}

206: \newcommand{\QQ}{{\mathbb Q}}

207: \newcommand{\R}{{\mathbb R}}

208: %\newcommand{\RR}{{\mathbb R}}

209: \newcommand{\C}{{\mathbb C}}

210: \newcommand{\CC}{{\mathbb C}}

211: \newcommand{\PP}{{\mathbb P}}

212: \newcommand{\e}{\,{\rm e}}

213: \newcommand{\CP}{{\CC\PP}}

214: \newcommand{\RP}{{\R\PP}}

215: \newcommand{\bra}{\langle}

216: \newcommand{\ket}{\rangle}

217: \newcommand{\no}{\nonumber}

218: \newcommand{\hsp}{\hskip5mm}

219: \newcommand{\vsp}{\vskip10mm}

220: \newcommand{\ds}{\displaystyle}

221: \newcommand{\ts}{\textstyle}

222: 

223: %  XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

224: %

225: %                Equation Definitions

226: %

227: \newcommand{\be}{\begin{equation}}

228: \newcommand{\ee}{\end{equation}}

229: \newcommand{\bea}{\begin{eqnarray}}

230: \newcommand{\eea}{\end{eqnarray}}

231: 

232: 

233: %

234: %                 Math Relations

235: %

236: \def\da{\downarrow}

237: \def\ua{\uparrow}

238: \def\upda{\updownarrow}

239: \def\to{\rightarrow}

240: \def\implies{\Rightarrow}

241: \def\To{\longrightarrow}

242: \def\longlongrightarrow{\relbar\joinrel\relbar\joinrel\rightarrow}

243: \def\ridiculousrightarrow{\relbar\joinrel\relbar\joinrel\relbar%

244: \joinrel\rightarrow}

245: 

246: \def\underarrow#1{\mathrel{\mathop{\longrightarrow}\limits_{#1}}}

247: \def\onnearrow#1{\mathrel{\mathop{\nearrow}\limits^{#1}}}

248: \def\undernearrow#1{\mathrel{\mathop{\nearrow}\limits_{#1}}}

249: \def\onarrow#1{\mathrel{\mathop{\longrightarrow}\limits^{#1}}}

250: \def\onArrow#1{\mathrel{\mathop{\longlongrightarrow}\limits^{#1}}}

251: \def\OnArrow#1{\mathrel{\mathop{\ridiculousrightarrow}\limits^{#1}}}

252: \def\lae{\mathrel{\mathop{\smash{\lower .5 ex \hbox{$\stackrel<\sim$}}}}}

253: \def\lae{\mathrel{\mathop{\smash{\lower .5 ex \hbox{$\stackrel>\sim$}}}}}

254: \def\eqq{\stackrel?=}

255: 

256: %                        Math Stuff

257: 

258: \def\ket#1{\left| #1 \right\rangle}

259: \def\bra#1{\left\langle #1 \right|}

260: \def\vev#1{\left\langle #1 \right\rangle}

261: \def\VEV#1{\left\langle #1 \right\rangle}

262: \def\f{\frac}

263: \def\pa{\partial}

264: \def\pb{\bar\pa}

265: \def\na{\nabla}

266: \def\Tr{{\rm Tr}}

267: \def\l:{\mathopen{:}\,}

268: \def\r:{\,\mathclose{:}}

269: \def\sech{\mathop{\rm sech}\nolimits}

270: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

271: 

272: %               DEFAULT SECTION SETUP

273: 

274: \catcode`\@=11

275: \def\theequation{\arabic{equation}}

276: %

277: %    for sections, subsections take from the following

278: %

279: %\def\theequation{\thesection.\arabic{equation}}

280: %\@addtoreset{equation}{section}

281: %\@addtoreset{footnote}{section}

282: %\@addtoreset{footnote}{subsection}

283: \catcode`\@=12

284: 

285: %                DEFAULT SETUP

286: 

287: 

288: \bcites

289: 

290: 

291: %                SETUP

292: 

293: %\draft

294: 

295: 

296: %             SECTION SETUP

297: 

298: \catcode`\@=11

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

300: \@addtoreset{equation}{section}

301: \@addtoreset{footnote}{section}

302: \@addtoreset{footnote}{subsection}

303: \catcode`\@=12

304: 

305: 

306: %               INFORMATION

307: 

308: \typeout{}

309: \typeout{}

310: \typeout{ }

311: \typeout{}

312: \typeout{}

313: \typeout{author}

314: \typeout{}

315: \typeout{}

316: \typeout{THIS IS A LATEX FILE: LATEX TWICE, AS USUAL. }

317: \typeout{}

318: \typeout{}

319: 

320: %

321: 

322: 

323: 

324: 

325: %

326: 

327: 

328: %

329: %           PAPER STARTS HERE

330: %

331: %load the bulking definitions etc.

332: %\input {shell}

333: 

334: 

335: \newcommand{\eps}{\epsilon}

336: \newcommand{\opsi}{\overline{\psi}}

337: \newcommand{\oQ}{\overline{Q}}

338: \newcommand{\btheta}{\overline{\theta}}

339: \newcommand{\cQ}{{\cal Q}}

340: \newcommand{\bepsilon}{\overline{\epsilon}}

341: \newcommand{\bPhi}{\overline{\Phi}}

342: \newcommand{\bP}{\overline{P}}

343: \newcommand{\bp}{\overline{p}}

344: \newcommand{\bphi}{\overline{\phi}}

345: \newcommand{\bpsi}{\overline{\psi}}

346: \newcommand{\bF}{\overline{F}}

347: \newcommand{\bJ}{\overline{J}}

348: \newcommand{\bh}{\overline{h}}

349: \newcommand{\bm}{\overline{m}}

350: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

351: \newcommand{\bi}{\overline{\imath}}

352: \newcommand{\bj}{\overline{\jmath}}

353: \newcommand{\bz}{\overline{z}}

354: \newcommand{\bchi}{\overline{\chi}}

355: \newcommand{\blambda}{\overline{\lambda}}

356: \newcommand{\bsigma}{\overline{\sigma}}

357: \newcommand{\bSigma}{\overline{\Sigma}}

358: \newcommand{\bareta}{\overline{\eta}}

359: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

360: \newcommand{\lrd}{\overleftarrow{\partial}\!\!\!\!\!

361: \overrightarrow{\vbox to 8.35pt{}}}

362: \newcommand{\lrD}{\overleftarrow{D}\!\!\!\!\!

363: \overrightarrow{\vbox to 8.2pt{}}}

364: \newcommand{\bD}{\overline{D}}

365: \newcommand{\bcQ}{\overline{\cal Q}}

366: \newcommand{\cD}{{\cal D}}

367: \newcommand{\rp}{{{\rm Re}\,p}}

368: \newcommand{\vp}{\varphi_P}

369: \newcommand{\cJ}{{\cal J}}

370: \newcommand{\rre}{{\rm Re}}

371: \newcommand{\iim}{{\rm Im}}

372: \newcommand{\ra}{\rightarrow}

373: \newcommand{\wh}{\widehat}

374: \newcommand{\bpartial}{{\bar\partial}}

375: \newcommand{\bA}{\overline{A}}

376: \newcommand{\tY}{{\tilde Y}}

377: \newcommand{\bY}{{\overline Y}}

378: \newcommand{\tL}{\tilde{L}}

379: \newcommand{\tG}{\tilde{G}}

380: \newcommand{\tJ}{\tilde{J}}

381: \newcommand{\tih}{\tilde{h}}

382: \newcommand{\tm}{\tilde{m}}

383: \newcommand{\tpsi}{\tilde{\psi}}

384: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

385: \newcommand{\nn}{\nonumber}

386: \newcommand{\kket}[1]{\vert  #1\rangle\!\rangle}

387: \newcommand{\bbra}[1]{\langle\!\langle #1 \vert}

388: \newcommand{\cket}[1]{\vert \Scr{C},#1\rangle\!\rangle}

389: \newcommand{\cbra}[1]{\langle\!\langle \Scr{C},#1 \vert}

390: 

391: 

392: \newcommand{\g}{{\it g}}

393: 

394: 

395: \newcommand{\NSNS}{{{}_{\rm NSNS}}}

396: \newcommand{\RR}{{{}_{\rm RR}}}

397: \newcommand{\NS}{{{}_{\rm NS}}}

398: \newcommand{\Ra}{{{}_{\rm R}}}

399: \newcommand{\GSO}{{{}_{\rm GSO}}}

400: \newcommand{\Pk}{{\rm P}_k}

401: \newcommand{\Mk}{{\rm M}_k}

402: \newcommand{\tilm}{\widetilde{m}}

403: \newcommand{\tilr}{\widetilde{r}}

404: \newcommand{\tilnu}{\widetilde{\nu}}

405: \newcommand{\tilGamma}{\widetilde{\Gamma}}

406: \newcommand{\tilomega}{\widetilde{\omega}}

407: \newcommand{\tilq}{\widetilde{q}}

408: 

409: \newcommand{\exact}{\longrightarrow}

410: 

411: \begin{document}

412: 

413: 

414: \begin{titlepage}

415: 

416: 

417: \begin{center}

418: 

419: \today\hfill

420: \hbox to 3cm

421: {\parbox[t]{5cm}{

422: hep-th/0401137\\

423: CERN-PH-TH/2004-006\\

424: KCL-MTH-04-01\\

425: NSF-KITP-04-01} \hss}\\

426: 

427: 

428: 

429: \vskip 2 cm

430: {\large \bf Orientifolds of Gepner Models}

431: \vskip 1 cm 

432: {Ilka Brunner$^{*}$, Kentaro Hori$^{\dagger}$, Kazuo Hosomichi$^{\dagger}$ and 

433: Johannes Walcher$^{\ddagger}$}\\

434: \vskip 0.5cm

435: {\it 

436: $^{*}$CERN Theory Division, Geneva, Switzerland}\\[0cm]

437: {\it and}\\[0cm]

438: {\it King's College, London, United Kingdom}\\[0.2cm]

439: {\it

440: $^{\dagger}$University of Toronto, Toronto, Ontario, Canada}\\[0.2cm]

441: {\it 

442: $^{\ddagger}$Kavli Institute for Theoretical Physics,

443: University of California, Santa Barbara, USA}

444: \vskip 0.2cm

445: 

446: 

447: \end{center}

448: 

449: 

450: 

451: 

452: \vskip 0.5 cm

453: \begin{abstract}

454: We systematically construct and study Type II Orientifolds

455: based on Gepner models which have

456: ${\mathcal N}=1$ supersymmetry in $3+1$ dimensions.

457: We classify the parity symmetries and construct the crosscap states.

458: We write down the conditions that a configuration of rational branes

459: must satisfy for consistency (tadpole cancellation

460: and rank constraints) and spacetime supersymmetry.

461: For certain cases, including Type IIB orientifolds of

462: the quintic and a two parameter model, one can find all solutions

463: in this class.

464: Depending on the parity, the number of vacua can be

465: large, of the order of $10^{10}-10^{13}$.

466: For other models, it is hard to find all solutions

467: but special solutions can be found --- some of them are chiral.

468: We also make comparison with the large volume regime and obtain

469: a perfect match.

470: Through this study, we find a number of new features of

471: Type II orientifolds, including the structure of moduli space and the

472: change in the type of O-planes under navigation through non-geometric

473: phases.

474: 

475: 

476: \end{abstract}

477: 

478: 

479: 

480: 

481: \end{titlepage}

482: 

483: 

484: \newpage

485: 

486: \tableofcontents

487: 

488: 

489: \newpage

490: 

491: \section{Introduction}\label{sec:intro}

492: 

493: 

494: String vacua with $\mathcal N=1$ supersymmetry in $3+1$ spacetime dimensions

495: have been re-attracting a lot of attention in recent years. One of the 

496: reasons is of course that despite a lot of efforts spent on the 

497: heterotic string, actual connections with real world particle physics

498: have proven difficult to make, and that new avenues have opened up 

499: with our growing mastery of strings, branes, and M-theory.

500: But we may also wish to turn this quest around and ask for general 

501: lessons from exploring the duality web with four supercharges, 

502: which on general grounds is expected to be quite complex. Whether 

503: or not one will be able to make contact with phenomenology, or 

504: extrapolate to a situation with broken supersymmetry, it is 

505: natural to expect that something interesting will be learned. 

506: 

507: 

508: 

509: Type II orientifolds with branes and fluxes

510: are an important class of models.

511: By a chain of duality,

512: they can be related to many other classes of models,

513: including the heterotic string

514: on Calabi-Yau 3-folds and

515: M-theory on $G_2$-holonomy manifolds, and therefore

516: may possibly provide a unifying scheme for

517: $4d$ ${\mathcal N}=1$ compactifications \cite{DougString}.

518: They provide natural set-ups for the braneworld scenario.

519: It should also be noted that the recent progress

520: in moduli stabilization is done in this framework

521: \cite{GKP,KKLT}.

522: However, most of the study in the past is done using supergravity,

523: or only toroidal orientifolds are given serious accounts.

524: This is definitely not a satisfactory state of affairs, because

525: the large volume or flat backgrounds are a

526: tiny part of the whole variety of possible theories.

527: What we need is a handle on the regime where supergravity is not

528: accessible. 

529: 

530: In this paper,

531: we study the other extreme regime where 

532: the internal space is very small but nevertheless

533: the worldsheet is extremely powerful. 

534: Namely, we construct and study Type II orientifolds

535: based on Gepner models \cite{Gepner}.

536: We will also try to

537: see how such theories are connected to the large volume regimes.

538: 

539: To avoid confusion, we emphasize that what we do here 

540: is within the framework of the perturbative NSR formalism.

541: We are obviously not able to include (RR) fluxes, and we are

542: not going to discuss the stringy quantum corrections at this stage,

543: except in the discussion of

544: the anomaly cancellation mechanism and Fayet-Iliopoulos terms.

545: In particular, the moduli including the dilaton remain unfixed.

546: However, we want to regard our work as a useful starting point 

547: for an explicit study of such models.

548: For instance, our models will have non-abelian 

549: gauge groups living on various RR tadpole canceling branes,

550: and our results may be useful also for the final step in the 

551: moduli stabilization \cite{KKLT}.

552: 

553: In fact, the roads have been

554: partially paved for us.

555: Recently, a great deal of results on D-branes in Type II string

556: compactifications were obtained.

557: They include application of Cardy's RCFT techniques

558: \cite{RS} and also the study of how they continue in the moduli space

559: to the large volume \cite{BDLR}.

560: There is also an orientifold version of Cardy, initiated by

561: Pradisi-Sagnotti-Stanev \cite{PSS} and further developed by

562: many people \cite{bantay,HSS,FSHSS,SaAn,BH1}.

563: Some of the preliminary results have been obtained in \cite{BlWi,ABPSS}

564: and more recently in \cite{BH2,GovMaj,HuiSch}.

565: In particular, we will extensively

566: use the results of \cite{BH2} on the minimal models

567: and other general properties of orientifolds of $(2,2)$ theories.

568: 

569: 

570: Our goal in this paper is threefold. Firstly, we want to adapt

571: and generalize the RCFT methods to the full string theory based on

572: the Gepner models.

573: Secondly, we want to present as unified a view as possible of the

574: various descriptions available for these worldsheets, such as the

575: Landau-Ginzburg and gauged linear sigma model pictures. In particular,

576: we want to generalize the relations between the Gepner point and 

577: large volume regimes to the situation involving unoriented strings. 

578: Thirdly, we want to

579: give rather detailed lists of explicit models that can be constructed 

580: within this framework. 

581: 

582: The ripeness of the subject and the richness of the harvest

583: have forced this paper to rather extended length. In order to

584: guide the reader towards the important results, we now give an 

585: overview over the organization of the presentation.

586: 

587: According to our global goal, we begin our discussion in section \ref{sec:CYO}

588: in the context of the gauged linear sigma model (GLSM), 

589: which provides the most global picture 

590: of Calabi-Yau compactifications on the worldsheet. The discussion

591: in subsection \ref{subsec:CYLG} is rather standard, and can safely be skipped

592: by experts. 

593: 

594: In subsection \ref{subsec:parsym}, we review the possible

595: orientifold projections, as discussed for example in \cite{BH2}.

596: As could be expected, parity symmetries of $\mathcal N=2$ supersymmetric

597: field theories come in two varieties, called A and B-type respectively.

598: The tadpoles arising from the corresponding O-planes must be canceled

599: by A and B-type D-branes, and the resulting $\mathcal N=1$ models can 

600: be thought of as Type IIA/IIB orientifolds, respectively. The associated

601: geometries are quite different, but are related to each other by mirror

602: symmetry. Of importance will be the classification of possible dressing

603: of the parity by various (classical and quantum) symmetries 

604: of the theory in such a way that the parity is involutive.

605: 

606: In subsection \ref{subsec:examples}, we make this discussion concrete

607: in the two examples which will accompany us through the rest of the

608: paper: the quintic hypersurface in $\PP^4$ and the degree $8$ hypersurface

609: in weighted projective space $\PP^4_{1,1,2,2,2}$. As we will see, many

610: interesting features arise in this two parameter model, which admits a

611: much richer set of possible orientifold projections than the quintic.

612: For example, we will see that with the appropriate dressing it is 

613: possible to project out the K\"ahler modulus corresponding to the 

614: overall size of the Calabi-Yau, or to select different sections of

615: the moduli space, corresponding to discrete fluxes in the large

616: volume regime.

617: 

618: We have organized the rest of the paper around this division into

619: A and B-type and the illustration in two examples, the quintic

620: and the two parameter model. In section \ref{sec:tad}, we discuss 

621: the construction of crosscap and boundary states in the full worldsheet 

622: theory of the Gepner model. Our approach differs slightly from the 

623: methods used in the literature \cite{RS,FS,BS,FSW,FKLLSW} in that we 

624: use a supersymmetric language throughout. Moreover, our construction

625: of B-type boundary states is new in the sense that it does not use 

626: the Greene-Plesser \cite{GreenePlesser} construction of mirror symmetry.

627: This approach also sheds new light on fixed-point resolution or

628: the appearance of so-called short-orbit branes \cite{BS,FSW,FKLLSW,MMS}.

629: 

630: We are then ready for discussing the consistency conditions that

631: constrain the possible string theory models we can build, A-

632: and B-type in sections \ref{sec:TCCA} and \ref{sec:TCCB}, respectively.

633: The discussion includes the computation of O-plane charges, the 

634: action of the parities on the D-branes, as well as the structure of

635: Chan-Paton factors. This puts us in a position to solve the consistency

636: conditions explicitly for our two examples. We also discuss the 

637: computation of the massless open string spectrum. We conclude each 

638: of the sections with lists of solutions to the tadpole cancellation 

639: conditions and open string spectra in selected cases.

640: 

641: 

642: The possibilities 

643: turn out to be extremely numerous and rich. For instance, for B-type 

644: models on the quintic, it turns out that there are 

645: $31561671503$ different supersymmetric and tadpole canceling 

646: configurations of rational branes at the Gepner point, all with the 

647: orthogonal gauge groups.

648: The number of vacua is similar in the two parameter model,

649: depending on the parity, with the the additional interesting feature

650: of allowing for configurations with unitary and symplectic

651: gauge groups.

652: 

653: 

654: For A-type models,

655: the spectrum is expected to be even richer, although we are not able

656: to solve the tadpole constraints completely in this case.

657: The number of equations and the number of branes are too many

658: for even the computer to find the solutions in a reasonable time.

659: However, special solutions can be found:

660: For any model with odd levels only, we always have a solution consisting

661: of four identical branes ---

662: four D6-branes on top of the O6-plane in the large volume limit.

663: For models including even levels, such a solution does not always exist

664: but one can use the recombination of branes in the

665: Landau-Ginzburg model to find special solutions in many cases.

666: Also, the size of the problem is much smaller when we consider

667: ``intermediate'' models whose orbifold group

668: is not minimal (single cyclic group)

669: nor maximal (the mirror of single-cyclic Gepner model).

670: 

671: 

672: Some of the theories we obtain have chiral matter contents.

673: Two out of nine special solutions for the two parameter model

674: (A-type) are chiral.

675: One of them has $U(1)^8$ gauge group with chiral quiver matters, and the

676: other is

677: $Sp(1)\times Sp(1)\times U(2)$ theory with matters in

678: $2\times ({\bf 2},{\bf 1},{\bf 2})$,

679: $({\bf 1},{\bf 2},\overline{\bf 2})$ and

680: $({\bf 2},{\bf 2},{\bf 1})$.

681: We feel that there are more chiral solutions than these two, but

682: how many and which is not clear at the moment.

683: For Type IIB orientifolds on Gepner models based on a single cyclic

684: group, such as the quintic or the two parameter model, all the solutions

685: are non-chiral.

686: However, some of the randomly chosen solutions

687: of a $\Z_5$-orbifold of quintic are chiral.

688: Thus, we obtain the first examples of chiral supersymmetric 4d theories

689: out of non-toroidal orientifolds.

690: 

691: 

692: 

693: 

694: 

695: Section \ref{sec:FI} is an interlude, in which we make remarks on chirality,

696: anomaly cancellation mechanism and Fayet-Iliopoulos terms.

697: The bilinear identity of the Witten index, where

698: the only  parity-invariant closed string

699: ground states propagate in the tree channel, 

700: plays an essential role in anomaly cancellation.

701: We make explicit the string coupling dependence of the low energy Lagrangian

702: and check that it is consistent with all of the tree level results we

703: obtained.

704: 

705: 

706: Finally, in section \ref{sec:continuation}, we compare the results on

707: consistency conditions with the geometrical expectations in the large

708: volume limit, finding complete agreement.

709: We will here make use of 

710: the results of \cite{BDLR} and \cite{DiaDou} on the connection between

711: geometry and Gepner model boundary states (see also 

712: \cite{DiaRo,KLLW,Scheidegger,Mayr}), as well as the results 

713: on the structure of the K\"ahler moduli space of the two parameter

714: model \cite{Candelas} and its real sections discussed in subsection

715: \ref{subsec:examples}.

716: We find something interesting through this study:

717: For some Type IIB orientifolds of

718: the two parameter model with two large volume regions

719: (distinguished by the B-field), the type of O-plane changes if one

720: goes from one large volume region to the other,

721: through non-geometric domains of the K\"ahler moduli space.

722: We consider an example with O5-planes at a genus 9 curve

723: and four rational curves. Here, in one region all O-planes 

724: are O5$^-$ (SO-type), whereas in the other region 

725: the O-planes at the rational curves become O5$^+$ (Sp-type).

726: For Type IIA orientifolds, we find in one example

727: an effective description of closed and open strings that matches

728: the results at the Gepner point as well as large volume.

729: An extensive study needs more technical development

730: such as an A-type analog of \cite{DiaDou,DiaRo,KLLW,Scheidegger,Mayr} 

731: (see, however \cite{Kennaway}), geometrical study of large volume branes,

732:  and methods to compute superpotential in both regimes.

733: 

734: 

735: 

736: 

737: 

738: 

739: 

740: 

741: {\it Note:}

742: A part of the present work (including

743: Section~\ref{sec:tad} and a part of Section~\ref{sec:TCCA})

744: is presented in a conference in \cite{HoriString}.

745: While the current work was under further progress

746: and was being written, we noticed these papers 

747: by Aldazabal et al \cite{Aldazabal} and by Blumenhagen \cite{Blumenhagen}, which have some overlap with our work.

748: However, in these papers, only odd level Gepner models are considered.

749: As we will see, the rich and interesting new physics arises in 

750: models including even level minimal models.

751: 

752: 

753: 

754: 

755: 

756: 

757: 

758: 

759: 

760: 

761: 

762: 

763: 

764: 

765: 

766: \section{Calabi-Yau Orientifolds}

767: \label{sec:CYO}

768: 

769: \subsection{Calabi-Yau Sigma Models and Gepner points}\label{subsec:CYLG}

770: 

771: Consider a $(2,2)$ supersymmetric gauge theory in $1+1$ dimensions

772: with $U(1)$ gauge group and

773: $r+1$ fields $X_1,\ldots,X_r,P$

774: with tree level superpotential

775: \beq

776: W=P(X_1^{k_1+2}+\cdots+X_r^{k_r+2})

777: \label{Wlsm}

778: \eeq

779: and twisted superpotential

780: $$

781: \widetilde{W}=t\Sigma.

782: $$

783: $\Sigma=\overline{D}_+D_-V$ is the superfieldstrength and $t=r-i\theta$

784: is the Fayet-Iliopoulos-Theta parameter.

785: The gauge transformations act on  the fields as

786: $$

787: P\to \e^{-iH\lambda}P,\quad

788: X_i\to \e^{iw_i\lambda}X_i,

789: $$

790: where

791: \beqa

792: &&H:={\rm lcm}\{k_i+2\},

793: \nn\\

794: &&w_i:={H\over k_i+2}.

795: \nn

796: \eeqa

797: 

798: 

799: For large values of the FI parameter,

800: the system reduces at low energies to the sigma model on the hypersurface $M

801: =\{X_1^{k_1+2}+\cdots+X_r^{k_r+2}=0\}$

802: in a weighted projective space of dimension $r-1$.

803: This gauge system, introduced in \cite{phases},

804: is called the linear sigma model for the

805: manifold $M$.

806: The condition that $M$ is Calabi-Yau is reflected by the vanishing of 

807: the sum of charges

808: $-1+\sum_{i=1}^r{1\over k_i+2}=0$. Namely

809: \beq

810: \sum_{i=1}^r{k_i\over k_i+2}=r-2=\dim M.

811: \label{CY}

812: \eeq

813: In this case, the beta function for the FI parameter vanishes and therefore

814: $t$ is a free parameter of the system.

815: 

816: 

817: 

818: At large negative ${\rm Re}(t)$, the $P$ field has a vacuum expectation value

819: and breaks the $U(1)$ gauge symmetry to the subgroup in which

820: $\e^{iH\lambda}=1$. This unbroken subgroup $\Gamma$ is generated by

821: the one with $\lambda=2\pi/H$ which acts on the fields as

822: \beq

823: \gamma:X_i\to \e^{2\pi i\over k_i+2}X_i,

824: \label{gaction}

825: \eeq

826: and is a cyclic group of order $H$.

827: The model at $t=-\infty$ is identified as the LG orbifold

828: with superpotential

829: \beq

830: W_{\it G}=X_1^{k_1+2}+\cdots+X_r^{k_r+2}

831: \eeq

832: divided by the group

833: $\Gamma\cong\Z_{H}$

834: acting on fields as (\ref{gaction}).

835: The LG model with superpotential $W=X^{k+2}$

836: flows in the infra-red limit to a $(2,2)$ superconformal field theory

837: with central charge $c={3k\over k+2}$,

838: called

839: the (A-series) level $k$ ${\mathcal N}=2$ minimal model, $M_k$.

840: The infra-red limit of the above LG orbifold is thus

841: the $\Gamma$-orbifold of the product of the minimal models;

842: $$

843: \left(\prod_{i=1}^r M_{k_i}\right)\Biggr/\Gamma

844: $$

845: This is the {\it Gepner model}.

846: The generator (\ref{gaction}) of the orbifold group $\Gamma$ is identified as

847: \beq

848: \gamma=\underbrace{(\g,\ldots,\g)}_r

849: \label{orbgen}

850: \eeq

851: in which $\g:=\e^{-2\pi i J_0}(-1)^{\widehat{F}}$

852: where $J_0$ is the $U(1)$ current of the (right-moving)

853: ${\mathcal N}=2$ superconformal algebra

854: and $(-1)^{\widehat{F}}$ is $1$ on NSNS sector but $-1$ on RR sector.

855: Note that the RR-ground states of lowest R-charge

856: $q=-{c\over 6}$

857: survives the orbifold projection, since

858: ${c\over 6}={\dim M\over 2}={r-2\over 2}$

859: and thus $\gamma=\e^{\pi i (r-2)}(-1)^r=1$.

860: This state corresponds to the holomorphic volume form of the Calabi-Yau

861: manifold. We discuss more on the ground states in

862: Section~\ref{subsub:ground}.

863: 

864: 

865: Type II string theory on $M\times \R^D$ is consistent only if

866: $2\dim M+D=10$. If we denote the complex dimension of

867: the transverse space by $d=(D-2)/2$, the criticality condition is

868: \beq

869: r+d=6.

870: \label{crit}

871: \eeq

872: In this paper we assume both the

873: Calabi-Yau condition (\ref{CY}) and the criticality condition

874: (\ref{crit}).

875: 

876: \noindent

877: {\bf Remarks.}\\{\small

878: {\bf (i)}~It is possible to have some $k_i=0$. The

879: IR limit of $W=X^2$ is empty, but

880:  can be regarded as the system with

881: a unique (ground) state in each of R/NS-sectors, with zero energy,

882: zero charge.

883: We will regard the $k_i=0$ factor as such a quantum field theory.

884: The orbifold group acts on this factor non-trivially:

885: the generator $\gamma$ acts as

886: $\g=\e^{-2\pi i J_0}(-1)^{\widehat{F}}=(-1)^{\widehat{F}}$,

887: namely, as identity on NSNS sector but

888: as $(-1)$ on the RR-sector.

889: Thus, having this factor has a non-trivial effect.

890: \\

891: {\bf (ii)}~The behaviour of the system depends very much on

892: whether there is an even $k_i$. It is useful to note that

893: when there is at least one even $k_i$ there is actually

894: an even number of $i$ with largest factors of $2$ in $k_i$, under

895: the Calabi-Yau condition,

896: $\sum_{i=1}^r{H\over k_i+2}=H$.

897: \\

898: {\bf (iii)}~Let us present some examples that satisfy the Calabi-Yau

899: and criticality conditions.\\

900: $\bullet$ $(k_i+2)=(3,3,3)$; $M=$ an elliptic curve,

901: $D=7+1$.

902: \\

903: $\bullet$ $(k_i+2)=(4,4,4,4)$; $M=$ a K3 surface.

904: $D=5+1$.

905: \\

906: We will mainly consider the case with $r=5$ and $d=1$ since this corresponds 

907: to the string compactification to $3+1$ dimensions.

908: The examples of this type are\\

909: $\bullet$ $(k_i+2)=(5,5,5,5,5)$; $M=$ a quintic hypersurface

910: in $\CP^4$.

911: \\

912: $\bullet$ $(k_i+2)=(8,8,4,4,4)$.

913: \\

914: $\bullet$ $(k_i+2)=(8,8,8,8,2)$.

915: \\

916: $\bullet$ $(k_i+2)=(12,12,6,6,2)$.

917: \\

918: The first two will be our basic examples where we

919: examine the general story in detail.

920: A complete list can be found in \cite{Gepnerclassification}.

921: \\

922: {\bf (iv)}~The non-chiral GSO projection of the minimal

923: model $M_k$ by $(-1)^F=\e^{-\pi i (J_0-\widetilde{J}_0)}$

924: is the $SU(2)_k\times U(1)_2$ mod $U(1)$ gauged WZW model,

925: or simply $SU(2)_k\times U(1)_2/U(1)_{k+2}$ coset model.

926: The latter model has primaries labeled by

927: $(l,m,s)\in \Mk$; namely $l\in P_k=\{0,1,...,k\}$,

928: $m\in \Z_{2(k+2)}$, $s\in \Z_4$, with $l+m+s$ even, $(l,m,s)\equiv

929: (k-l,m+k+2,s+2)$.\footnote{In this paper, following the convention used by

930: majority of people, the $SU(2)$ spin $j$ is labeled by

931:  $L=2j\in\Pk=\{0,1,2,\ldots,k\}$,

932: rather than $j$ itself that is used in \cite{MMS,BH2}.}

933: The product theory $M_{k_1}\times\cdots \times M_{k_r}$

934: should not be confused with the tensor product of

935: the GSO projected models of $M_{k_1},...,M_{k_r}$.

936: In the latter the space of states would have mixture of NSNS and RR

937: factors, while in the former

938:  NS/R alignment is automatically imposed, as usual in

939: ordinary supersymmetric quantum field theories.

940: \\

941: {\bf (v)} The GSO projected model

942: has global symmetries $g_{n,s}$ corresponding to

943: simple currents $(0,n,s)$ ($n\in \Z_{2(k+2)}$, $s\in \Z_4$, with

944: $n+s$ even) which act on the states in

945: $\Scr{H}_{l',m',s'}\otimes\Scr{H}_{l',-m',-s'}$ as multiplication

946: by a phase $\e^{\pi i\left({nm'\over k+2}-{ss'\over 2}\right)}$.

947: The symmetry $\g$ above induces one of them,

948: $g_{2,0}$.

949: \\

950: {\bf (vi)}~``Gepner Model'' usually refers to more general models

951: based on orbifold of the product of minimal models.

952: It doesn't have to come from linear sigma models of the above types.

953: In Appendix~\ref{app:Gepner}, we present more general models.

954: In the main text of the paper, we treat only the class of models

955: introduced above (except Sections~\ref{sub:construction} and

956: \ref{sub:Cardybranes} where

957: the discussion is general),

958: in particular the case $D=3+1$ and $r=5$.

959: We relegate the discussion on 

960: the most general models

961: to Appendix.

962: 

963: 

964: }

965: 

966: In many cases,

967: $M$ has singularities that are inherited from the orbifold singularities

968: of the ambient space,

969: and their  resolution introduces extra K\"ahler parameters.

970: This is accommodated in the linear sigma model by extending the gauge group

971: and adding charged fields.

972: In general, the gauge group will be $U(1)^k=\prod_{a=1}^kU(1)_a$ gauge theory 

973: with matter fields $P,X_1,\ldots,X_{r+k-1}$ of certain charge

974: $Q_P^a,Q_1^a,\ldots,Q_{r+k-1}^a$ and certain (twisted) superpotential.

975: For example, for $(k_i+2)=(8,8,4,4,4)$,

976: the full system after the resolution

977: has $U(1)\times U(1)$ gauge group and six matter fields

978: of the following charges \cite{MorrisonPlesser}:

979: \beq

980: \begin{array}{cccccccc}

981: &P&X_1&X_2&X_3&X_4&X_5&X_6\\

982: U(1)_1&-4&0&0&1&1&1&1\\

983: U(1)_2&0&1&1&0&0&0&-2

984: \end{array}

985: \label{2pcharge}

986: \eeq

987: The system has superpotential

988: $$

989: W=P\Bigl\{\,X_6^4(X_1^8+X_2^8)+X_3^4+X_4^4+X_5^4\,\Bigr\},

990: $$

991: and twisted superpotential

992: \beq

993: \widetilde{W}=t_1\Sigma_1+t_2\Sigma_2,

994: \label{FIth2}

995: \eeq

996: where the $t_a=r_a-i\theta_a$ and $\Sigma_a=\overline{D}_+D_-V_a$

997: are the FI-Theta parameter and the fieldstrength of the $U(1)_a$ gauge

998: group.

999: In the limit $t_2\to-\infty$ with $2t_1+t_2$ fixed, $X_6$ acquires a

1000: large absolute value and breaks the gauge group except the one

1001: generated by $(2i,i)\in \mathfrak{u}(1)_1\oplus\mathfrak{u}(1)_2$.

1002: We are then left with the original system with one $U(1)$ gauge symmetry

1003: whose FI-Theta parameter is $t=2t_1+t_2$.

1004: This corresponds to undoing the resolution.

1005: 

1006: 

1007: \subsubsection{RR Ground States and Chiral Primaries}\label{subsub:ground}

1008: 

1009: Let us present the list of supersymmetric ground states of the system.

1010: The level $k$ minimal model has

1011: $(k+1)$ supersymmetric ground states $|l\rangle_{\RR}$ ($l=0,1,...,k$)

1012: which

1013: correspond to

1014: $X^l$ and have R-charges

1015: $q=\tilq={l+1\over k+2}-{1\over 2}$. Also, on a circle twisted by

1016: $\e^{-2\pi i \nu J_0}$, there is a unique supersymmetric ground state

1017: $|0\rangle_{\nu}$ which has R-charge

1018: $q=-\tilq={l_{\nu}+1\over k+2}-{1\over 2}$ where

1019: $l_{\nu}\in \{0,1,...,k\}$ is defined by

1020: $l_{\nu}+1\equiv -\nu$ (mod $(k+2)$).

1021: The RR ground states of the Gepner model are made of these states.

1022: Since the orbifold group is generated by the tensor product of

1023: $-\e^{-2\pi i J_0}$ for the $r=5$ factors, the condition is

1024: that the sum of R-charges is an odd half-integer,

1025: $\sum_iq_i\in {1\over 2}+\Z$.

1026: Untwisted sector states are thus the products

1027: $\otimes_{i=1}^5|l_i\rangle_{\RR}$ with the condition

1028: $\sum_i({l_i+1\over k_i+2}-{1\over 2})=\sum_i{l_i\over k_i+2}-{3\over 2}\in

1029: {1\over 2}+ \Z$, or

1030: \beq

1031: \sum_{i=1}^5{l_i\over k_i+2}=0,1,2,3.

1032: \label{orbcond}

1033: \eeq

1034: They correspond to harmonic forms of degree

1035: $(3,0)$, $(2,1)$, $(1,2)$ and $(0,3)$ respectively of the relevant

1036: Calabi-Yau manifold.\footnote{

1037: For non-linear sigma model on a Calabi-Yau manifold of dimension $n$,

1038: supersymmetric ground states with R-charge

1039: $(q,\tilq)$ correspond to harmonic $(p,\bar{p})$ forms

1040: where $(q,\tilq)=({n\over 2}-p,\bar{p}-{n\over 2})$.}

1041: There are also RR ground states from the twisted sectors

1042: labeled by $\nu=1,2,...,H-1$. The orbifold condition is the same

1043: as (\ref{orbcond}) where $l_i$ is replaced by

1044: $l_{\nu}^{(i)}$ for such $i$ where the twist is non-trivial,

1045: $\nu\not\equiv 0$ mod $(k_i+2)$.

1046: For the $\nu=1$ twist, we find $l_1^{(i)}=k_i$ for all $i$ and we find a unique 

1047: ground state with $q=-\tilq={3\over 2}$. The geometrical counterpart is the

1048: $(0,0)$-form. For $\nu=(H-1)$, we also find a unique ground state that

1049: corresponds to the $(3,3)$-form.

1050: The ground states from the twisted sectors are mostly related to

1051: $(p,p)$-forms. However, there can be states corresponding to

1052: off-diagonal forms. For example, let us consider the case where $H$ is even

1053: and twist by $\nu={H\over 2}$.

1054: The twist in the $i$-th factor is non-trivial if and only if

1055: $w_i={H\over k_i+2}$ is odd.

1056: For such an $i$, $l_{H\over 2}^{(i)}$ is ${k_i\over 2}$ and

1057: the ground state is $q_i=\tilq_i=0$.

1058: For other $i$, the twist is trivial and the ground states are

1059: ordinary ones $|l_i\rangle_{\RR}$

1060: with R-charges $q_i=\tilq_i={l_i+1\over k_i+2}-{1\over 2}$.

1061: They correspond to $(2,1)$ or $(1,2)$ forms.

1062: Let us show the number of ground states in two examples.

1063: 

1064: 

1065:  

1066: \noindent

1067: \underline{$(k_i+2)=(5,5,5,5,5)$}~ Untwisted ground states

1068:  correspond to

1069: monomials of $X_i$ with degree $0$, $5$, $10$, $15$ (with relations

1070: $X_i^4=0$) and there are $1$, $101$, $101$, $1$ of them.

1071: Also there is a unique ground state $\otimes_i|0\rangle_{\nu}$

1072: from each of $\nu=1,2,3,4$ twisted sectors.

1073: These numbers are organized into

1074: the ``Hodge diamond''

1075: $$

1076: \begin{array}{ccccccc}

1077: &&&\!\! 1\!\! &&&\\[-0.2cm]

1078: &&\!\! 0\!\! &&\!\! 0\!\! &&\\[-0.2cm]

1079: &\!\! 0\!\! &&\!\! 1\!\! &&\!\! 0\!\! &\\[-0.2cm]

1080: \!\! 1\!\! &&\!\! 101\!\! &&\!\! 101\!\! &&\!\! 1\!\! \\[-0.2cm]

1081: &\!\! 0\!\! &&\!\! 1\!\! &&\!\! 0\!\! &\\[-0.2cm]

1082: &&\!\! 0\!\! &&\!\! 0\!\! &&\\[-0.2cm]

1083: &&&\!\! 1\!\! &&&

1084: \end{array}

1085: $$

1086: 

1087: \noindent

1088: \underline{$(k_i+2)=(8,8,4,4,4)$}~

1089: $X_1,X_2,X_3,X_4,X_5$ have weights $w_i=1,1,2,2,2$.

1090: Untwisted ground states correspond to

1091: monomials of $X_i$ with total weight $0$, $8$, $16$, $24$

1092: (with relations $X_1^7=X_2^7=X_3^3=X_4^3=X_5^3=0$).

1093: There are $1$, $83$, $83$, $1$ of them.

1094: There is a unique ground state $\otimes_i|0\rangle_{\nu}$

1095: from each of $\nu=1,2,3,5,6,7$ twisted sectors.

1096: They corresponds to diagonal forms.

1097: For the $\nu=4$ twisted sector, ground states are

1098: $$

1099: \bigotimes_{i=1,2}|0\rangle_{\nu}\otimes

1100: \bigotimes_{i=3,4,5}|l_i\rangle_{\RR}

1101: $$

1102: where $l_3+l_4+l_5=1$ (3 states) or $5$ (3 states).

1103: The Hodge diamond is therefore

1104: $$

1105: \begin{array}{ccccccc}

1106: &&&\!\! 1\!\! &&&\\[-0.2cm]

1107: &&\!\! 0\!\! &&\!\! 0\!\! &&\\[-0.2cm]

1108: &\!\! 0\!\! &&\!\! 2\!\! &&\!\! 0\!\! &\\[-0.2cm]

1109: \!\! 1\!\! &&\!86\! &&\! 86\! &&\!\! 1\!\! \\[-0.2cm]

1110: &\!\! 0\!\! &&\!\! 2\!\! &&\!\! 0\!\! &\\[-0.2cm]

1111: &&\!\! 0\!\! &&\!\! 0\!\! &&\\[-0.2cm]

1112: &&&\!\! 1\!\! &&&

1113: \end{array}\qquad

1114: 86=83+3

1115: $$

1116: 

1117: As usual \cite{LVW}, RR ground states are in one-to-one correspondence

1118: with chiral primaries by a spectral flow which shifts

1119: the R-charge as $q\to q\pm {c\over 6}$, $\tilq\to\tilq  \pm {c\over 6}$.

1120: The spectral flow with the sign $(++)$ maps the ground states

1121: to NSNS states corresponding to

1122: chiral fields ($(c,c)$-fields),

1123: and the $(-+)$-spectral flow maps them to NSNS states corresponding to

1124: twisted chiral  fields ($(a,c)$-fields).

1125: They are marginal operators if $q=\tilq=1$.

1126: Marginal $(c,c)$ primaries correspond to $(2,1)$-forms and

1127: marginal $(a,c)$ primaries correspond to $(1,1)$-forms.

1128: 

1129: 

1130: 

1131: 

1132: \subsubsection{The Parameter Space}

1133: 

1134: \subsubsection*{\it Worldsheet Parameter Space}

1135: 

1136: The $(c,c)$ and $(a,c)$ primaries with R-charge (1,1)

1137: are exactly marginal operators.

1138: Parameters coupled to $(c,c)$-primaries

1139: parametrize the complex structure of the target space.

1140: In the linear sigma model, they are the parameters $a_i$

1141: of the tree level superpotential $W=PG(X_i,a_i)$.

1142: If there are twisted RR ground states corresponding to $(2,1)$-forms,

1143: the corresponding parameters do not fit into the linear sigma model.

1144:  Parameters coupled to $(a,c)$-primaries

1145: parametrize

1146: the complexified K\"ahler class $[\omega -iB]$,

1147: where $\omega$ is the K\"ahler form and $B$ is the B-field.

1148: In the linear sigma model, they are the FI-Theta parameters $t^a$.

1149: In the large volume limit, the FI-Theta parameters and the complexified

1150: K\"ahler parameters are related by

1151: \beq

1152: [\omega -iB]\sim \sum_{a=1}^k (t^a+\pi iQ_P^a) \omega_a

1153: \label{kaepar}

1154: \eeq

1155: where $\omega_a\in H^2(M,\Z)$ is the first Chern class of the line bundle

1156: associated with the $U(1)_a$ gauge group and $Q_P^a$

1157: is the charge of the field $P$.

1158: 

1159: 

1160: 

1161: 

1162: The worldsheet theory is singular

1163: at certain loci of the parameter space.

1164: On the complex structure moduli space, the singularity is at

1165: the loci where $M=\{G(X_i,a_i)=0\}$ is singular 

1166: as a complex manifold.

1167: On the K\"ahler moduli space, the singularity is at the loci where

1168: the linear sigma model has an unbroken gauge symmetry

1169: and some vector multiplet

1170: is exactly massless.

1171: For example, in the case of quintic, the singularity is at

1172: $$

1173: \e^{t}=-5^5.

1174: $$

1175: In the example of $(k_i+2)=(8,8,4,4,4)$, there are two singular loci:

1176: \beq

1177: C_1=\Bigl\{e^{t_2}=4\Bigr\},

1178: \qquad

1179: C_{\rm con}=\Bigl\{\, \e^{t_2}(1-4^{-4}\e^{t_1})^2=4\,\Bigr\}.

1180: \label{sing2para}

1181: \eeq

1182: 

1183: 

1184: \subsubsection*{\it Scalar Manifold of Spacetime Theory ---

1185: Type II on Calabi-Yau}

1186: 

1187: 

1188: Let us consider Type II string theory on $\R^{3+1}$ times the internal CFT

1189: we have been discussing.

1190: The theory has ${\mathcal N}=2$ supersymmetry on $\R^{3+1}$.

1191: The moduli of the worldsheet theory give rise to

1192: massless scalar fields in $3+1$ dimensions, which

1193: are part of some ${\mathcal N}=2$ supermultiplets.

1194: Other parts in the multiplet come from the NS-R, R-NS and R-R sectors.

1195: In Type IIA string theory, the $h^{1,1}$ K\"ahler moduli

1196: are the scalar components of vector multiplets, while

1197: the $h^{2,1}$ complex structure moduli together with

1198: the periods of the RR 3-form potential

1199: constitute the scalar components of

1200: hypermultiplets

1201: For Type IIB, the complex structure moduli

1202: are in vector multiplets, while

1203: the K\"ahler moduli and

1204: the periods of the RR potentials are in

1205: hypermultiplets.

1206: The singular loci of the worldsheet theory

1207: are not singular in full string theory.

1208: It is simply that there are degrees of freedom

1209: that become massless at these loci \cite{Strominger}.

1210: 

1211: 

1212: 

1213: \subsubsection{Mirror Description}\label{subsub:mirror}

1214: 

1215: 

1216: The mirror of the Gepner model \cite{GreenePlesser}

1217: (see also \cite{VafaLGorb})

1218: is

1219: % the product of the same set of minimal models divided by a different group,

1220: % $\widetilde{\Gamma}$, which is the subgroup of $\prod_{i=1}^r\Z_{k_i+2}$ 

1221: % consisting of the transformations

1222: % $\widetilde{X}_i\to \e^{2\pi i m_i\over k_i+2}\widetilde{X}_i$

1223: % obeying $\prod_{i=1}^r\e^{2\pi i m_i\over k_i+2}=1$.

1224: % Namely, $\widetilde{\Gamma}$-orbifold of LG model with superpotential

1225: the IR limit of the LG orbifold with superpotential

1226: $$

1227: \widetilde{W}_{\it G}=\widetilde{X}_1^{k_1+2}+\cdots+\widetilde{X}_r^{k_r+2},

1228: $$

1229: and the group $\tilGamma\subset\prod_{i=1}^r\Z_{k_i+2}$

1230: acting on the fields as

1231: $$

1232: \widetilde{X}_i\to \e^{2\pi i \tilnu_i\over k_i+2}\widetilde{X}_i,\qquad

1233: \prod_{i=1}^r\e^{2\pi i \tilnu_i\over k_i+2}=1.

1234: $$

1235: The superpotential can be deformed by

1236: polynomials of the same degree as $W$ and which are invariant under

1237: the group $\widetilde{\Gamma}$.

1238: The monomial $\widetilde{X}_1\cdots\widetilde{X}_r$

1239: is an example that exists in all the cases.

1240: In fact the model with superpotential

1241: $\widetilde{W}_{\it G}

1242: +\e^{t/H}\widetilde{X}_1\cdots\widetilde{X}_r$ is the mirror of

1243: the linear sigma model with single $U(1)$ gauge group whose FI-Theta

1244: parameter is $t$ \cite{HV}.

1245: In the case $(k_i+2)=(5,5,5,5,5)$, this is

1246: $$

1247: \widetilde{W}=\widetilde{X}_1^5+\cdots+\widetilde{X}_5^5

1248: +\e^{t/5}\widetilde{X}_1\cdots\widetilde{X}_5.

1249: $$

1250: In fact $\widetilde{X}_1\cdots\widetilde{X}_5$

1251:  is the only allowed deformation for this case, which

1252:  corresponds to the fact

1253: that the quintic has only one K\"ahler modulus.

1254: In more general models, there are other

1255: $\widetilde{\Gamma}$-invariant monomials of the same degree, each

1256: corresponding to a blow up mode.

1257: For instance, in the case $(k_i+2)=(8,8,4,4,4)$,

1258: the fully deformed superpotential is

1259: \beq

1260: \widetilde{W}=\widetilde{X}_1^8+\widetilde{X}_2^8+\widetilde{X}_3^4

1261: +\widetilde{X}_4^4+\widetilde{X}_5^4+

1262: \e^{t_1/4+t_2/8}\widetilde{X}_1\cdots\widetilde{X}_5

1263: +\e^{t_2/2}\widetilde{X}_1^4\widetilde{X}_2^4,

1264: \label{dualW2para}

1265: \eeq

1266: where $t_1$ and $t_2$ are the FI-Theta parameters in (\ref{FIth2}).

1267: It indeed reduces to the one-parameter family

1268: $\widetilde{W}_G+\e^{t/8}\widetilde{X}_1\cdots\widetilde{X}_5$

1269: under the blow-down limit, $t_2\to-\infty$, $t=2t_1+t_2$ fixed.

1270: 

1271: 

1272: \subsection{Parity Symmetries}\label{subsec:parsym}

1273: 

1274: We would like to classify involutive parity symmetries of the system

1275: that preserves a half of the $(2,2)$ worldsheet supersymmetry.

1276: The superfield notation we use here is introduced in \cite{BH2}:

1277: the A and B parities on

1278: the $(2,2)$ superspace are $\Omega_A(x^{\pm},\theta^{\pm},\btheta^{\pm})

1279: =(x^{\mp},-\btheta^{\mp},-\theta^{\mp})$

1280: and $\Omega_B(x^{\pm},\theta^{\pm},\btheta^{\pm})

1281: =(x^{\mp},\theta^{\mp},\btheta^{\mp})$.

1282: 

1283: 

1284: \subsubsection{Linear Sigma Model}

1285: 

1286: We first consider the parity symmetries of the linear

1287: sigma model.

1288: 

1289: \noindent

1290: {\bf A-parities}\\

1291: A-parities of the single $U(1)$ gauge system with superpotential

1292: (\ref{Wlsm}) are

1293: $\Omega_A$ combined with $V\to V$ and

1294: \beq

1295: \tau^A_{{\bf m},\sigma}:\,\,

1296: \begin{array}{l}

1297: P\To \overline{P},\\[0.1cm]

1298: X_i\To \e^{2\pi i m_i\over k_i+2}\overline{X_{\sigma(i)}}.

1299: \end{array}

1300: \label{PAlsm}

1301: \eeq

1302: Here, ${\bf m}$ labels the elements of the global symmetry

1303: $(\prod_{i=1}^r\Z_{k_i+2})/\Z_H$.

1304: Also, $i\mapsto\sigma(i)$ is an order two permutation

1305: such that $k_{\sigma(i)}=k_i$ so that the charges are invariant.

1306: This is involutive if and only if

1307: $$

1308: m_i=m_{\sigma(i)}\quad(\mbox{mod $k_i+2$}).

1309: $$

1310: The phase rotation can sometimes be undone by a change of

1311: variables. For $X_i'=\e^{2\pi i n_i\over k_i+2}X_i$, the parity acts as

1312: $X_i'\to \e^{{2\pi i\over k_i+2}(m_i+n_i+n_{\sigma(i)})}

1313: \overline{X_{\sigma(i)}'}$. Therefore there is an equivalence relation

1314: ${\bf m}\equiv {\bf m'}$ if and only if

1315: $$

1316: m_i'=m_i+n_i+n_{\sigma(i)}\quad (\mbox{mod $k_i+2$}).

1317: $$

1318: The FI-theta parameter $t$ is unconstrained but the parameters $(a_i)$

1319: that deforms the superpotential are constrained to be essentially real,

1320: $G(\e^{2\pi im_i\over k_i+2}\overline{X_{\sigma(i)}},a_i)

1321: =\overline{G(X_i,a_i)}$.

1322: 

1323: 

1324: 

1325: 

1326: 

1327: \noindent

1328: {\bf B-parities}\\

1329: B-parities of the single $U(1)$ gauge system (\ref{Wlsm}) are

1330: $\Omega_B$ combined with $V\to V$ and

1331: \beq

1332: \tau^B_{{\bf m},\sigma}:\,\,

1333: \begin{array}{l}

1334: P\To -P,\\[0.1cm]

1335: X_i\To \e^{2\pi i m_i\over k_i+2} X_{\sigma(i)}

1336: \end{array}

1337: \label{PBlsm}

1338: \eeq

1339: where $\sigma$ is an order two permutation

1340: with $k_{\sigma(i)}=k_i$ and

1341: $$

1342: m_i+m_{\sigma(i)}=0\quad (\mbox{mod $k_i+2$})

1343: $$

1344: so that it is involutive.

1345: For the variable $X_i'=\e^{2\pi in_i\over k_i+2}X_i$,

1346: the parity acts as

1347: $X_i'\to\e^{{2\pi i\over k_i+2}(m_i+n_i-n_{\sigma(i)})}X_{\sigma(i)}'$.

1348: Thus ${\bf m}$ and ${\bf m'}$ are equivalent if and only if

1349: $$

1350: m_i'=m_i+n_i-n_{\sigma(i)}\quad (\mbox{mod $k_i+2$}).

1351: $$

1352: The FI-Theta parameter is constrained to be real

1353: $

1354: \overline{\e^t}=\e^t,

1355: $

1356: while the complex structure parameters $a_i$ are required to obey

1357: $G(\e^{2\pi i m_i\over k_i+2}X_{\sigma(i)},a_i)

1358: \equiv G(X_i,a_i)$.

1359: 

1360: 

1361: 

1362: 

1363: \subsubsection{Gepner point}

1364: 

1365: The parity symmetries we have considered above,

1366: $P^A_{{\bf m},\sigma}=\tau_{{\bf m},\sigma}^A\Omega_A$ 

1367: and

1368: $P^B_{{\bf m},\sigma}=\tau_{{\bf m},\sigma}^B\Omega_B$,

1369: are of course

1370: symmetries at the Gepner point.

1371: Since $P$ has an expectation value $\langle P\rangle$, 

1372: it is understood that a gauge transformation is

1373: used so that $\langle \tau_{{\bf m},\sigma}P\rangle=\langle P\rangle$.

1374: For A-parity, taking $\langle P\rangle$ real,

1375: the transformation of the LG fields $X_i$ is the same as in

1376: (\ref{PAlsm})

1377: $$

1378: \tau^A_{{\bf m},\sigma}:X_i\To \e^{2\pi i m_i\over k_i+2}

1379: \overline{X_{\sigma(i)}},

1380: $$

1381: while for B-parity (\ref{PBlsm}) is combined

1382: with the gauge transformation

1383: $\e^{i\lambda}=\e^{\pi i/H}$:

1384: $$

1385: \tau^B_{{\bf m},\sigma}:X_i\To\e^{2\pi i m_i\over k_i+2}\e^{\pi i\over

1386: k_i+2}X_{\sigma(i)}.

1387: $$

1388: 

1389: 

1390: 

1391: At the Gepner point, there are extra symmetries

1392: called the quantum symmetries which form a group

1393: $\widehat{\Gamma}\cong\Z_H$. 

1394: The quantum symmetry $g_{\omega}$ associated with an $H$-th root of unity

1395: $\omega$ multiplies the $\gamma^{\ell}$-twisted states

1396: by the phase $\omega^{\ell}$.

1397: It acts on

1398: the mirror variables $\widetilde{X}_1,\ldots,\widetilde{X}_r$ as

1399: \beq

1400: g_{\omega}:\widetilde{X}_i\longmapsto \omega_i\widetilde{X}_i;\qquad

1401: \omega_i^{k_i+2}=1\,(\forall i),\quad \omega_1\cdots\omega_r=\omega.

1402: \eeq

1403: The monomial $\widetilde{X}_1\cdots\widetilde{X}_r$

1404: is not invariant under $g_{\omega}$ with $\omega\ne 1$

1405: and quantum symmetry is completely broken if $\e^t\ne 0$.

1406: For other deformations it is broken to a subgroup.

1407: 

1408: 

1409: One can use this quantum symmetry to modify the

1410: parity symmetry.

1411: Thus, we have a larger set of parity symmetries at the Gepner point:

1412: \beqa

1413: &&

1414: P^A_{\omega;{\bf m},\sigma}=g_{\omega}\tau_{{\bf m},\sigma}^A\Omega_A,\\

1415: &&

1416: P^B_{\omega;{\bf m},\sigma}=g_{\omega}\tau_{{\bf m},\sigma}^B\Omega_B.

1417: \eeqa

1418: Actually, not all of them are involutive and not all of them

1419: are inequivalent.

1420: For A-type, the parity acts on the dual variables as

1421: $\Omega_A$ combined with $\widetilde{X}_i\to \omega_i \e^{\pi i\over k_i+2}

1422: \widetilde{X}_{\sigma(i)}$.

1423: This is involutive if and only if $\omega^2=1$, namely

1424: \beq

1425: \omega=\left\{

1426: \begin{array}{ll}

1427: 1&\mbox{$H$ odd},\\

1428: \pm 1&\mbox{$H$ even}.

1429: \end{array}\right.

1430: \eeq

1431: For B-type, the parity action is $\Omega_B$ combined with

1432: $\widetilde{X}_i\to\omega_i^{-1}\overline{\widetilde{X}_{\sigma(i)}}$.

1433: This is always involutive. However, some of them can be undone by a change of

1434: variables.

1435: Dressing by $g_{\omega}$ and $g_{\omega'}$ are equivalent

1436: if and only if

1437: $$

1438: \omega'=\alpha^2\omega,

1439: \qquad

1440: \alpha^H=1.

1441: $$

1442: 

1443: 

1444: If $H$ is odd, there is no non-trivial involutive

1445: dressing by quantum symmetry.

1446: For A-type, dressed parity is not involutive unless $g_{\omega}=1$.

1447: For B-type, any dressing is equivalent to no dressing.

1448: 

1449: 

1450: If $H$ is even, there is essentially

1451: a unique non-trivial involutive dressing by quantum symmetry.

1452: For A-type, it is the dressing by the order 2 element $g_{-1}$.

1453: Since $\widetilde{X}_1\cdots\widetilde{X}_r$ flips its sign

1454: under $g_{-1}$,

1455: the dressed parity is not a symmetry if $\e^t\ne 0$.

1456: Thus, the K\"ahler modulus corresponding to the overall

1457: size is frozen at $\e^t=0$ if we require this parity to be a symmetry.

1458: For B-type, it is the dressing by the primitive element

1459: $g_{\omega}$, $\omega=\e^{2\pi i/H}$.

1460: It maps the monomial $\widetilde{X}_1\cdots\widetilde{X}_r$ to

1461: $\e^{2\pi i/H}\overline{\widetilde{X}_1\cdots\widetilde{X}_r}$.

1462: Thus, the condition of parity invariance is shifted from

1463: $\e^{t/H}=\overline{\e^{t/H}}$ to

1464: $\e^{t/H}\e^{2\pi i/H}=\overline{\e^{t/H}}$.

1465: In terms of the invariant coordinate $\e^t=(\e^{t/H})^H$,

1466: the condition is $\e^t\in \R_{\geq 0}$ if not dressed by quantum symmetry

1467: while it is $\e^t\in \R_{\leq 0}$ if dressed by odd quantum symmetry.

1468: 

1469: 

1470: \subsubsection{Type II Orientifolds}

1471: 

1472: 

1473: Let us consider Type II string theory

1474: on $\R^{3+1}$ times our internal CFT, and

1475: gauge the worldsheet parity symmetry $P$ which acts trivially on

1476: the $3+1$ spacetime coordinates but acts on the

1477: internal CFT as one of the above parities (A-type or B-type).

1478: This is the Type IIA or Type IIB orientifold.

1479: (The original papers on more general orientifolds are

1480: \cite{ASI,BSII,ASII,HoI,HoII}.)

1481: To make it consistent, we need to add either D-branes or fluxes.

1482: This is one of our main themes of this paper.

1483: For now, let us discuss aspects that are independent of how it is done.

1484: 

1485: Since the left movers and right movers of the string modes are 

1486: identified by the parity, ${\mathcal N}=2$ supersymmetry will be broken

1487: to at most ${\mathcal N}=1$ supersymmetry.

1488: Use of A-parity for Type IIA string and B-parity for Type IIB string

1489: is the necessary condition for preserving

1490: an ${\mathcal N}=1$ supersymmetry.

1491: (Whether it is preserved in the full theory depends on what we add

1492: (D-branes and fluxes), and this is another main topic of the latter part

1493: of this paper.)

1494: 

1495: As in the case before orientifold, the worldsheet moduli

1496: give rise to light fields of the spacetime theory.

1497: We have seen that

1498: these moduli are constrained by the requirement that the parity is

1499: a symmetry of the worldsheet.

1500: The light fields are constrained accordingly.

1501: Together with light fields from the NS-R, R-NS and R-R sectors,

1502: they constitute ${\mathcal N}=1$ supermultiplets.

1503: The pattern at the large volume is analyzed in \cite{BH2}

1504: and is summarized in Table~\ref{clmod}.

1505: \begin{table}

1506: \begin{center}

1507: \begin{tabular}{|ll@{\quad\vrule width0.8pt\quad}c|c|}

1508: \hline

1509: &$\begin{array}{l}

1510: \\

1511: \\

1512: \end{array}$

1513: &chiral multiplets&vector multiplets

1514: \\

1515: \noalign{\hrule height 0.8pt}

1516: $\begin{array}{l}

1517: \\

1518: \\

1519: \end{array}$

1520: &IIAO(6)

1521: &$h^{1,1}_-+h^{2,1}+1$&$h^{1,1}_+$

1522: \\

1523: \hline

1524: $\begin{array}{l}

1525: \\

1526: \\

1527: \end{array}$

1528: &IIBO(9,5)~~&$h^{2,1}_++h^{1,1}+1$&$h^{2,1}_-$

1529: \\

1530: \hline

1531: $\begin{array}{l}

1532: \\

1533: \\

1534: \end{array}$

1535: &IIBO(7,3)&$h^{2,1}_-+h^{1,1}+1$&$h^{2,1}_+$

1536: \\

1537: \hline

1538: \end{tabular}

1539: \caption{Light Fields from Closed Strings}

1540: \label{clmod}

1541: \end{center}

1542: \end{table}

1543: Here IIAO(6) is for Type IIA orientifolds, where ``6'' is

1544: because we generically

1545: have orientifold 6-planes.

1546: IIBO(9,5) is for Type IIB orientifolds with O9 or O5-planes

1547: and IIBO(7,3) is for Type IIB orientifolds with O7 and/or O3-planes.

1548: Also, $h^{p,\bar{p}}_{\pm}$ are the number of harmonic $(p,\bar{p})$-forms

1549: that are invariant/anti-invariant under the involution.

1550: Note that, even when the worldsheet

1551: moduli receive antiholomorphic

1552: constraints (for example,

1553: complex structure moduli by A-parity and

1554: K\"ahler moduli by B-parity), they combine with

1555: periods of RR-potentials and form complex scalars of 

1556: ${\mathcal N}=1$ chiral multiplets.

1557: 

1558: 

1559: 

1560: 

1561: 

1562: 

1563: 

1564: 

1565: 

1566: 

1567: 

1568: \subsection{Examples}\label{subsec:examples}

1569: 

1570: Let us study the parity symmetries discussed above in

1571: typical examples with

1572: odd and even $H$'s 

1573: --- the quintic $(k_i+2)=(5,5,5,5,5)$ with $H=5$ and

1574: the two parameter model

1575: $(k_i+2)=(8,8,4,4,4)$ with $H=8$.

1576: 

1577: 

1578: 

1579: 

1580: \subsubsection{Quintic}\label{subsub:quintic}

1581: 

1582: This case is studied in detail in \cite{BH2}.

1583: As we have seen above, there is no non-trivial involutive

1584: dressing by quantum symmetry.

1585: Also, one can show there is no non-trivial involutive dressing by the 

1586: $\Z_5^4$ global symmetry: ${\bf m}$ that determines an

1587: involutive parity is equivalent to ${\bf 0}$.

1588: Thus, the parity is determined purely in terms of $\sigma\in \mathfrak{S}_5,

1589: \sigma^2=1$.

1590: Up to permutation of variables, there are only

1591: three cases: $\sigma={\rm id}$, $(12)$ and $(12)(34)$.

1592: 

1593: The table shows the projected moduli as well as

1594:  O-planes in the geometric phase,

1595: for these six orientifolds.

1596: \begin{table}[htb]

1597: \caption{Six Orientifolds of Quintic}

1598: \label{oqtable}

1599: \begin{center}

1600: \begin{tabular}{|l||l|l|}

1601: \hline

1602: parity&moduli $(K,C)$ &O-planes\\

1603: \noalign{\hrule height 0.8pt}

1604: $P^A_{\rm id}$&$(1_{\C},101_{\R})$&O6 at the real quintic $\cong\RP^3$\\

1605: \hline

1606: $P^A_{(12)}$&$(1_{\C},101_{\R})$&O6 at an $\RP^3$\\

1607: \hline

1608: $P^A_{(12)(34)}$&$(1_{\C},101_{\R})$&O6 at an $\RP^3$\\

1609: \noalign{\hrule height 0.8pt}

1610: $P^B_{\rm id}$&$(1_{\R},101_{\C})$&O9 at $M$\\

1611: \hline

1612: $P^B_{(12)}$&$(1_{\R},63_{\C})$&O3 at a point and O7 at a hypersurface\\

1613: \hline

1614: $P^B_{(12)(34)}$&$(1_{\R},53_{\C})$&O5's at a rational and a genus 6 curves

1615: \\

1616: \hline

1617: \end{tabular}

1618: \end{center}

1619: \end{table}

1620: 

1621: 

1622: For all three B-type orientifolds, the K\"ahler moduli space

1623: is the real line $\e^t=\overline{\e^t}$ as depicted in Figure~\ref{km}.

1624: It passes through the Gepner point,

1625: is broken at the conifold point and extends to the two large volume regions

1626: --- one with $B=0$ and another with $B=\pi$.

1627: The Gepner point is connected along $\e^t>0$ to the $B=\pi$ 

1628: asymptotic region, as follows from (\ref{kaepar}). The path along $e^t<0$ 

1629: is blocked by the conifold singularity.

1630: 

1631: \begin{figure}[htb]

1632: \centerline{\includegraphics{km3.eps}}

1633: \caption{K\"ahler moduli space for a B-orientifold of the quintic}

1634: \label{km}

1635: \end{figure}

1636: 

1637: 

1638: 

1639: \subsubsection{A Two parameter Model}\label{subsub:two}

1640: 

1641: 

1642: In the example $(8,8,4,4,4)$, there is a unique non-trivial and

1643: involutive dressing by quantum symmetry.

1644: Also, there are several non-trivial and involutive 

1645: dressing by the global symmetry $\Z_8\times\Z_4^3$.

1646: Here, since there are already a variety of ways to choose ${\bf m}$ for a fixed

1647: $\sigma$,

1648: we only consider the $\sigma={\rm id}$ cases.

1649: 

1650: For A-type parity, $X_i\to \e^{2\pi i m_i\over k_i+2}\overline{X_i}$,

1651: ${\bf m}$ obey the equivalence relation

1652: $m_i\equiv m_i+2n_i$ and $m_i\equiv m_i+1$, and there are

1653: six independent choices

1654: ${\bf m}=(00000)$, $(00001)$, $(00011)$, $(00111)$, $(01000)$, $(01001)$.

1655: Under the quantum symmetry $g_{-1}$, 

1656: the term $\widetilde{X}_1\cdots\widetilde{X}_5$

1657: in the dual superpotential (\ref{dualW2para}) flips its sign while

1658: $\widetilde{X}_1^4\widetilde{X}_2^4$ is invariant.

1659: Thus, if dressed by the quantum symmetry $g_{-1}$, the K\"ahler modulus

1660: $t=2t_1+t_2$ is frozen at $\e^t=0$ but $\e^{t_2}$ is unconstrained.

1661: If not dressed by quantum symmetry, the K\"ahler moduli are both unconstrained.

1662: In the regime $t_1,t_2\gg 0$, one can talk about the geometry.

1663: $\tau_{\bf m}^A$ acts as an antiholomorphic involution, and

1664: the fixed point set is

1665: $X_i=\e^{\pi i m_i\over k_i+2}x_i$, and $X_6=x_6$, where 

1666: $x_i$ are all real, obey 

1667: $$

1668: (-1)^{m_1}x_1^8x_6^4+

1669: (-1)^{m_2}x_2^8x_6^4+

1670: (-1)^{m_3}x_3^4+

1671: (-1)^{m_4}x_4^4+

1672: (-1)^{m_5}x_5^4=0,

1673: $$

1674: and are subject to the gauge conditions of the GLSM preserving the reality

1675: condition. The determination of the topology of the resulting fixed point 

1676: sets can sometimes be a little cumbersome. This problem has been studied in 

1677: \cite{RRW} and we review here the parity $P^A_{00001}$ as an example. The 

1678: topology of this O-plane can be obtained by studying the solutions of the 

1679: real equation

1680: $$

1681: x_6^4 (x_1^8+x_2^8) + x_3^4 + x_4^4 = x_5^4

1682: $$

1683: subject to the rescaling 

1684: $$

1685: (x_1,x_2,x_3,x_4,x_5,x_6)\equiv(\lambda x_1,\lambda x_2,\mu x_3,\mu x_4,

1686: \mu x_5,\lambda^2\mu x_6)

1687: $$

1688: with $\lambda,\mu\in \R^*$. Thus, we have to require $x_5\neq 0$, 

1689: whereupon we can set $x_5$ to one by rescaling with $\mu$.

1690: The second rescaling can be absorbed by noting that 

1691: $x_1^8+x_2^8>0$ in the large volume phase. After changing 

1692: variables to $x_1^8=y_1^2$, $x_2^8=y_2^2$, $x_6^4=y_6^2$, 

1693: $x_3^4=y_3^2$, $x_4^4=y_4^2$, the constraints become 

1694: $y_6^2+y_3^2+y_4^2=1$, $y_1^2+y_2^2=1$, with non-trivial

1695: $\Z_2$ identification $(y_1,y_2,y_3,y_4,y_6)\equiv

1696: (-y_1,-y_2,y_3,y_4,y_6)$ (from $\lambda=-1$). Thus, topologically,

1697: this O-plane is $S^2\times \RP^1=S^2\times S^1$.

1698: 

1699: We refer to \cite{RRW} for the remaining cases, and summarize the 

1700: results in Table \ref{o2tableA}. We do not know a simple description

1701: of the O-plane for the parity $P^A_{01001}$, except that it has Betti

1702: numbers $b_0=1$ and $b_1=2$.

1703: We have also indicated in table \ref{o2tableA} that even though the

1704: number of moduli from complex structure deformations ($86$ real parameters)

1705: is always the same, each parity selects a different real section of

1706: the moduli space. In particular, these sections can intersect in different

1707: ways with singular loci, as we will illustrate below.

1708: 

1709: 

1710: 

1711: \begin{table}[htb]

1712: \caption{A-type Orientifolds (with $\sigma=1$) of the

1713: Two Parameter Model}

1714: \label{o2tableA}

1715: \begin{center}

1716: \begin{tabular}{|l||l|l|}

1717: \hline

1718: parity&moduli $(K,C)$ &O-planes\\

1719: \noalign{\hrule height 0.8pt}

1720: $P^A_{+;00000}$&$(2_{\C},86_{\R})$&No O-plane\\

1721: \hline

1722: $P^A_{-;00000}$&$(1_{\C},86_{\R})$&non-geometric\\

1723: \hline

1724: $P^A_{+;00001}$&$(2_{\C},86_{\R}')$&O6 at an $S^2\times S^1$\\

1725: \hline

1726: $P^A_{-;00001}$&$(1_{\C},86_{\R}')$&non-geometric\\

1727: \hline

1728: $P^A_{+;00011}$&$(2_{\C},86_{\R}'')$&O6 at a $T^3$\\

1729: \hline

1730: $P^A_{-;00011}$&$(1_{\C},86_{\R}'')$&non-geometric\\

1731: \hline

1732: $P^A_{+;00111}$&$(2_{\C},86_{\R}''')$&O6 at an $S^2\times S^1$\\

1733: \hline

1734: $P^A_{-;00111}$&$(1_{\C},86_{\R}''')$&non-geometric\\

1735: \hline

1736: $P^A_{+;01000}$&$(2_{\C},86_{\R}'''')$&O6 at an $S^3$\\

1737: \hline

1738: $P^A_{-;01000}$&$(1_{\C},86_{\R}'''')$&non-geometric\\

1739: \hline

1740: $P^A_{+;01001}$&$(2_{\C},86_{\R}''''')$&O6 at a SLAG with $b_0=1$, $b_1=2$\\

1741: \hline

1742: $P^A_{-;01001}$&$(1_{\C},86_{\R}''''')$&non-geometric\\

1743: \hline

1744: \end{tabular}

1745: \end{center}

1746: \end{table}

1747: 

1748: For B-type parity,

1749: $X_i\to \e^{2\pi i (m_i+1/2)\over k_i+2}X_i$, ${\bf m}$ is constrained

1750: by $2m_i=0$ (mod $k_i+2$) and obey the equivalence relation

1751: $m_i\equiv m_i+1$. There are eight choices described by

1752: the signs $\epsilon_i=\e^{2\pi i m_i\over k_i+2}$:

1753: $(\epsilon_1\epsilon_2\epsilon_3\epsilon_4\epsilon_5)=(+++++)$,

1754: $(++-++)$,

1755: $(++--+)$,

1756: $(++---)$,

1757: $(+-+++)$,

1758: $(+--++)$,

1759: $(+---+)$,

1760: $(+----)$.

1761: If dressed with $g_{\omega}^m$ with $\omega=\e^{2\pi i /8}$,

1762: the monomials

1763: $\widetilde{X}_1\cdots\widetilde{X}_5$ and

1764: $\widetilde{X}_1^4\widetilde{X}_2^4$ of the dual variables

1765: are transformed to

1766: $\e^{2\pi i m/8}\overline{\widetilde{X}_1\cdots\widetilde{X}_5}$

1767: and $\e^{\pi i m}\overline{\widetilde{X}_1^4\widetilde{X}_2^4}$

1768: respectively.

1769: The symmetry condition $\widetilde{W}\to \overline{\widetilde{W}}$

1770: is satisfied by the dual superpotential (\ref{dualW2para})

1771: if $\e^{2\pi i m/8}\e^{t_1/4+t_2/8}=\overline{\e^{t_1/4+t_2/8}}$

1772: and $\e^{\pi i m}\e^{t_2/2}=\overline{\e^{t_2/2}}$.

1773: It follows from this that

1774: the K\"ahler moduli are constrained by

1775: \beqa

1776: &&\mbox{not dressed by quantum symmetry:}\quad

1777: \e^{t_1}\in \R,\quad\e^{t_2}\in \R_{\geq 0} 

1778: \qquad\label{undressed}\\

1779: &&\mbox{dressed by odd quantum symmetry:}

1780: \quad\e^{t_1}\in \R,\quad\e^{t_2}\in\R_{\leq 0},

1781: \qquad\label{dressed}

1782: \eeqa

1783: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1784: Each of these have two large volume regions classified by the $B$-field.

1785: By using (\ref{kaepar}) and the charge table (\ref{2pcharge}), one learns

1786: that in the case (\ref{undressed}) the $B$-field can be $B=0$ or 

1787: $B=\pi\omega_1$, while in the case (\ref{dressed}), we have 

1788: $B=\pi\omega_2$ or $B=\pi\omega_1+\pi\omega_2$.

1789: 

1790: To describe this real section of the moduli space of the two parameter 

1791: model in somewhat more detail, we recall from \cite{Candelas} that by

1792: introducing

1793: $$

1794: \xi=\e^{t_2+2t_1} \qquad \eta=\e^{t_2} \qquad \zeta=\e^{t_1+t_2} \,,

1795: $$

1796: we can embed the parameter space as the quadric

1797: \beq

1798: Q=\{ \xi \eta - \zeta^2 = 0\}

1799: \label{quadric}

1800: \eeq

1801: in $\C^3$. In these coordinates, the singularities in the parameter 

1802: space of the mirror threefold (\ref{dualW2para}) appear at the curves

1803: \beqa

1804: C_1&=& Q\cap \{ \eta= 4 \} \\

1805: C_{\rm con} &=& Q\cap \{2^{-16}\xi + \eta - 2^{-7}\zeta = 4\} \,.

1806: \eeqa

1807: The real moduli space, $Q\cap \{\xi,\eta,\zeta\in \R\}$ is an 

1808: ordinary double cone, which consist of the components 

1809: $Q_+=Q\cap \{\xi,\eta>0\}$, and $Q_-=Q\cap\{\xi,\eta<0\}$, meeting 

1810: at the tip $\xi=\eta=\zeta=0$ (Gepner point). In fact, from 

1811: (\ref{undressed}) and (\ref{dressed}), we see that the real K\"ahler 

1812: moduli space of the orientifold without (with) dressing by quantum 

1813: symmetry is given by $Q_+$ ($Q_-$). Moreover, the real versions of 

1814: $C_1$ and $C_{\rm con}$ are ordinary cone sections, and it is easy 

1815: to check that they are parabolas and lie completely in $Q_+$. Since 

1816: they intersect transversely and have co-dimension one, if we do not

1817: dress by quantum symmetry, {\it the Gepner point is completely

1818: separated from the two large volume regimes}. In that case, it is 

1819: not possible to connect the Gepner point with a geometric interpretation 

1820: of the orientifold without running into a singularity of the worldsheet

1821: theory. If dressed by odd quantum symmetry, the moduli space and 

1822: singular loci do not meet, so that Gepner point is connected to the 

1823: corresponding two large volume regimes. 

1824: 

1825: In order to capture its global structure, it is convenient to compactify 

1826: the moduli space by adding a divisor $C_\infty$ at infinity. As 

1827: explained in \cite{Candelas}, the compactification can be achieved 

1828: by embedding $Q$ in (\ref{quadric}) in the projective space 

1829: $\PP^3$, which by abuse of notation we coordinatize with 

1830: $[\xi:\eta:\zeta:\tau]$. In addition to $C_1$ and $C_{\rm con}$,

1831: we then have the distinguished locus $C_\infty= Q\cap\{\tau=0\}$,

1832: which also corresponds to a degeneration of (\ref{dualW2para}).

1833: We also have the ``orbifold locus'' $C_0= \{ \xi=\zeta=0\} \subset 

1834: Q$, which contains the Gepner point $g=[0:0:0:1]$. We show a 

1835: picture of this compactified parameter space and the location 

1836: of the singular loci $C_0$, $C_1$, $C_{\rm con}$, and $C_\infty$

1837: in Figure \ref{kmt}. It is important to emphasize that in distinction 

1838: to $C_{\rm con}$ and $C_1$, $C_0$ and $g$ do not lead to a singular 

1839: worldsheet theory. 

1840: 

1841: \begin{figure}[htb]

1842: \psfrag{Co}{$C_0$}

1843: \psfrag{Ce}{$C_1$}

1844: \psfrag{Cc}{$C_{\rm con}$}

1845: \psfrag{Ci}{$C_{\infty}$}

1846: \psfrag{a}{$a$}

1847: \psfrag{b}{$b$}

1848: \psfrag{c}{$c$}

1849: \psfrag{d}{$d$}

1850: \psfrag{e}{$e$}

1851: \psfrag{G}{$g$} 

1852: \centerline{\epsfig{width=6in,file=twoparmod.eps}}

1853: %\centerline{\includegraphics{twoparmod.eps}}

1854: \caption{Real section of the K\"ahler moduli space of the two parameter

1855: model $(k_i+2)=(8,8,4,4,4)$. The Gepner point $g$ is at the tip of

1856: a conical singularity.  The left 

1857: cone is the moduli space of the orientifold without dressing by 

1858: quantum symmetry. The lines of singularity $C_1$ and $C_{\rm con}$

1859: divide the moduli space into several perturbative regions.

1860: The right cone (shaded region), which reaches out 

1861: all the way to the large volume regime, is the moduli space of the 

1862: orientifold with dressing by quantum symmetry.}

1863: \label{kmt}

1864: \end{figure}

1865: 

1866: Also, $C_\infty$ simply corresponds to the boundary of the 

1867: uncompactified moduli space in the usual sense. In particular,

1868: the large volume limit has been hidden inside of $C_\infty$

1869: by the compactification process. To recover this (unique) large volume 

1870: limit, we need to blow up the point $b$, where the two

1871: divisors $C_1$ and $C_\infty$ intersect non transversely.

1872: Near $C_\infty$, we can work in the patch $\xi=1$ of $\PP^3$, in which 

1873: our real moduli space is given by the equation $\eta=\zeta^2$, 

1874: with $\tau$ arbitrary. In this patch, $C_1$ is given by 

1875: $\tau=\eta=\zeta^2$, while $C_\infty$ is given by $\tau=0$. A real

1876: blowup of the origin corresponds to replacing a small disc around 

1877: $\zeta=\tau=0$ with a M\"obius strip. The exceptional divisor (called 

1878: $D_{(-1,-1)}$ in \cite{Candelas}) of this blowup is simply the non-trivial 

1879: one-cycle of the M\"obius strip. In simple terms, the blowup

1880: means that when approaching the origin along some path, we keep track 

1881: of the first derivative $d\tau/d\zeta$, and we do not reach the

1882: same point depending on the value of this derivative. It is easy to

1883: see from this description that now $C_1$, $C_\infty$, and $D_{(-1,-1)}$

1884: meet at a triple intersection, and we have to perform a second blowup,

1885: replacing the origin by the exceptional divisor $D_{(0,-1)}$. The

1886: large volume point is now the unique intersection point 

1887: $D_{(0,-1)}\cap C_\infty$. In this way, we have recovered the

1888: description of the large volume limit as a cylinder $(t_1,t_2)

1889: \equiv (t_1+2\pi i,t_2)\equiv (t_1,t_2+2\pi i)$. We show this sequence of 

1890: blowups in Figure \ref{blowups}. 

1891: 

1892: \begin{figure}[htb]

1893: \psfrag{Dd}{$\scriptstyle{D_{(0,-1)}}$}

1894: \psfrag{D}{$\scriptstyle{D_{(-1,-1)}}$}

1895: \psfrag{Ce}{$\scriptstyle{C_1}$}

1896: \psfrag{Ci}{$\scriptstyle{C_{\infty}}$}

1897: \centerline{\epsfig{width=6in,file=blowups.eps}}

1898: \caption{To see the large volume limit in the compactified moduli

1899: space, we have to blowup the singular point $b=C_1\cap C_\infty$.

1900: We replace twice a small neighborhood of the origin with a 

1901: M\"obius strip, successively inserting the divisors $D_{(-1,-1)}$ 

1902: and $D_{(0,-1)}$. The multiply stroked lines are identified. 

1903: The shaded region can be reached smoothly from the Gepner

1904: point.}

1905: \label{blowups}

1906: \end{figure}

1907: 

1908: This description puts us in a position to illustrate geometrically

1909: the statements on the structure of the large volume region that

1910: we have made above. The (real) neighborhood of the large volume 

1911: point is divided by $D_{(0,-1)}$ and $C_\infty$ into four quadrants,

1912: which are geometrically distinguished by the value of the B-field.

1913: By following the sequence of blowups and the global picture

1914: in Figure \ref{kmt}, we see that starting from the Gepner point $g$, 

1915: we can reach two of these quadrants without crossing a singularity,

1916: but not the other two. 

1917: 

1918: 

1919: We will use this description of the moduli space in section 

1920: \ref{sec:continuation} when we discuss the comparison between Gepner

1921: model boundary and crosscap states and large volume. 

1922: 

1923: 

1924: 

1925: 

1926: 

1927: Let us describe the topological structure of the

1928: orientifold planes corresponding to each of the involutions

1929: $\tau^B_\epsilon$ that we have defined above.

1930: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1931: In the large volume regime,

1932: $\tau^B_{\epsilon}$ acts on the manifold as the

1933: holomorphic involution $X_i\to \epsilon_i X_i$ and $X_6\to X_6$.

1934: The fixed point set is the loci with

1935: $\epsilon_iX_i=\lambda_2X_i$ ($i=1,2$),

1936: $\epsilon_iX_i=\lambda_1X_i$ ($i=3,4,5$),

1937: $X_6=\lambda_1\lambda_2^{-2}X_6$.

1938: The solutions in the eight cases are:\\

1939: $(+++++)$:~

1940:  No condition (the whole manifold $M$).\\

1941: $(++-++)$:~ $X_3=0$ (a hypersurface).\\

1942: $(++--+)$:~ $X_3=X_4=0$ (a curve of genus 9)

1943: and $X_5=X_6=0$ (four lines)\\

1944: $(++---)$:~ $X_3=X_4=X_5=0$ (eight points)

1945: and

1946: $X_6=0$ (a hypersurface).\\

1947: $(+-+++)$:~ $X_1=0$ or $X_2=0$ (two hypersurfaces).

1948: The two are homologous since they are two fibres of the K3-fibrations

1949: (with base $\{(X_1,X_2)\}$ and fibres $\{(X_3,X_4,X_5,X_6)\}$).\\

1950: $(+--++)$:~ $X_1=X_3=0$ or $X_2=X_3=0$ (two genus 3 curves). They are

1951: homologous to each other.\\

1952: $(+---+)$:~ $X_1=X_5=X_6=0$ (four points),

1953: $X_1=X_3=X_4=0$ (four points),

1954: $X_2=X_5=X_6=0$ (four points),

1955: $X_2=X_3=X_4=0$ (four points).\\

1956: $(+----)$:~ $X_1=X_6=0$ or

1957: $X_2=X_6=0$ (two genus 3 curves). They are homologous to each other.\\

1958: These are included in Table~\ref{o2tableB}.

1959: 

1960: 

1961: 

1962: 

1963: 

1964: \newcommand{\lw}[1]{\smash{\lower2.0ex\hbox{#1}}}

1965: \begin{table}[htb]

1966: \caption{B-type Orientifolds (with $\sigma=1$) of the

1967: Two Parameter Model}

1968: \label{o2tableB}

1969: \begin{center}

1970: \renewcommand{\arraystretch}{1.2}

1971: \begin{tabular}{|l||l|l|}

1972: \hline

1973: parity&moduli $(K,C)$ &O-planes\\

1974: \noalign{\hrule height 0.8pt}

1975: \multicolumn{1}{|l||}{$P^B_{0;+++++}$}&

1976: \multicolumn{1}{l|}{$(2_{\R},86_{\C},...,83_{\C})$}&\lw{O9 at $M$}\\

1977: \cline{1-2}

1978: $P^B_{1;+++++}$&$(2'_{\R},86_{\C})$&\\

1979: \hline

1980: \multicolumn{1}{|l||}{$P^B_{0;++-++}$}&

1981: \multicolumn{1}{l|}{$(2_{\R},57_{\C},...,56_{\C})$}&

1982: \lw{O7 at a hypersurface}\\

1983: \cline{1-2}

1984: $P^B_{1;++-++}$&$(2'_{\R},57_{\C})$&\\

1985: \hline

1986: \multicolumn{1}{|l||}{$P^B_{0;++--+}$}&

1987: \multicolumn{1}{l|}{$(2_{\R},46_{\C},...,47_{\C})$}&

1988: \lw{O5's at four rational and a genus 9 curves}\\

1989: \cline{1-2}

1990: $P^B_{1;++--+}$&$(2'_{\R},46_{\C})$&\\

1991: \hline

1992: \multicolumn{1}{|l||}{$P^B_{0;++---}$}&

1993: \multicolumn{1}{l|}{$(2_{\R},41_{\C},...,44_{\C})$}&

1994: \lw{O7 at a hypersurface and O3's at eight points}\\

1995: \cline{1-2}

1996: $P^B_{1;++---}$&$(2'_{\R},41_{\C})$&\\

1997: \hline

1998: \multicolumn{1}{|l||}{$P^B_{0;+-+++}$}&

1999: \multicolumn{1}{l|}{$(2_{\R},53_{\C},...,56_{\C})$}&

2000: \lw{O7's at two homologous K3 hypersurfaces}\\

2001: \cline{1-2}

2002: $P^B_{1;+-+++}$&$(2'_{\R},53_{\C})$&\\

2003: \hline

2004: \multicolumn{1}{|l||}{$P^B_{0;+--++}$}&

2005: \multicolumn{1}{l|}{$(2_{\R},46_{\C},...,47_{\C})$}&

2006: \lw{O5's at two homologous genus 3 curves}\\

2007: \cline{1-2}

2008: $P^B_{1;+--++}$&$(2'_{\R},46_{\C})$&\\

2009: \hline

2010: \multicolumn{1}{|l||}{$P^B_{0;+---+}$}&

2011: \multicolumn{1}{l|}{$(2_{\R},45_{\C},...,44_{\C})$}&

2012: \lw{O3's at sixteen points}\\

2013: \cline{1-2}

2014: $P^B_{1;+---+}$&$(2'_{\R},45_{\C})$&\\

2015: \hline

2016: \multicolumn{1}{|l||}{$P^B_{0;+----}$}&

2017: \multicolumn{1}{l|}{$(2_{\R},46_{\C},...,43_{\C})$}&

2018: \lw{O5's at two homologous genus 3 curves}\\

2019: \cline{1-2}

2020: $P^B_{1;+----}$&$(2'_{\R},46_{\C})$&\\

2021: \hline

2022: \end{tabular}

2023: \end{center}

2024: \end{table}

2025: 

2026: 

2027: To conclude this section, we count the number of complex structure

2028: moduli in these orientifolds.

2029: This can be done by looking at the parity action on the

2030: corresponding chiral primary states.

2031: To see the action, we first consider the parities $P^B_{\omega;+++++}$

2032: that correspond to the

2033: identity of $M$ in the large volume.

2034: In such a case, we know that the complex structure moduli is unconstrained.

2035: Thus the number of moduli is full $86$. Let us now go to the Gepner point

2036: along some path in the K\"ahler moduli space.

2037: As we have seen this can be done only for the parity $P^B_{1;+++++}$

2038: dressed by an odd quantum symmetry.

2039: By continuity the number of moduli at the Gepner point is still $86$.

2040: Thus, we find that $P^B_{1;+++++}$ acts trivially

2041: on all the marginal $(c,c)$ primaries. Since other parities are obtained from

2042: $P^B_{1;+++++}$ by dressing global or quantum symmetries

2043: (whose action we know), we

2044: now know the action of all the parities $P^B_{\omega;{\bf m}}$

2045: on the marginal $(c,c)$ primaries, at the Gepner point.

2046: In this way, we find the number of complex moduli

2047: at the Gepner point.\\

2048: {\bf Remarks.}\\{\small

2049: {\bf (i)}~By continuity the number of moduli found at the Gepner point 

2050: applies everywhere in the K\"ahler moduli space

2051: for the parities $P^B_{1;\epsilon_1...\epsilon_5}$.

2052: This in particular tells us the number of moduli in the large volume limit.

2053: The numbers are listed in Table~\ref{o2tableB}.

2054: (It is an interesting exercise to check these numbers directly

2055: by analyzing geometry.)

2056: %JW: what do you mean by this comment? Doesn't this follow from the

2057: %   action on the monomials?

2058: \\

2059: {\bf (ii)}~In the large volume regime, the only difference between

2060: $P_{1;\epsilon_1...\epsilon_5}$ and $P_{0;\epsilon_1...\epsilon_5}$

2061: is the value of the $B$-field. Thus, the number found in (i) is still

2062: applicable for $P_{0;\epsilon_1...\epsilon_5}$, 

2063: in the large volume regime.

2064: \\ 

2065: {\bf (iii)}~On the other hand, one can analyze the action of

2066: $P_{0;\epsilon_1...\epsilon_5}$ at the Gepner point (as stated above).

2067: The action is the same as $P_{1;\epsilon_1...\epsilon_5}$

2068: on the untwisted sector states but differs from that by $-$ sign

2069: on the twisted sector states. Thus, if $n$ twisted ground states survive

2070: the $P_{1;\epsilon_1...\epsilon_5}$-projection,

2071: then the other $(3-n)$ survive the

2072: $P_{0;\epsilon_1...\epsilon_5}$-projection.

2073: \\

2074: {\bf (iv)}~Thus, the number of moduli is different between the Gepner point and

2075: the large volume regimes for the

2076: $P_{0,\epsilon_1...\epsilon_5}$-orientifolds.

2077: This is not a puzzle from the worldsheet point of view, because

2078: the two regions are separated by the singularity locus.

2079: There are also two other regions and the number of moduli there could

2080: be different as well.

2081: In Table~\ref{o2tableB}, we only show

2082:  the number at the large volume and at the Gepner point, and simply write

2083: dots ... for the other two regions.

2084: }

2085: 

2086: We have seen that the complex structure moduli can jump

2087: from from one component to another

2088: of the real K\"ahler moduli space.

2089: This tells us something about the full string theory.

2090: As we have discussed, the real K\"ahler moduli

2091: are combined with RR-potentials to form complex parameters

2092: (which become the lowest components of ${\mathcal N}=1$

2093: chiral superfields of the spacetime theory).

2094: An interesting problem is to find the behaviour near the singularity.

2095: One possibility is that

2096: one can go around the singular loci

2097: by turning on the RR-potential, so that

2098: the separate regions of the

2099: real K\"ahler moduli space are smoothly connected to each other in the full

2100: moduli space (Figure~\ref{singul}(a)).

2101: This happens in other situations, such as the flop transition

2102: \cite{phases,AGM}.

2103: \begin{figure}[htb]

2104: \centerline{\includegraphics{singularity.eps}}

2105: \caption{Two possibilities of complexifying the real moduli

2106: space with a codimension one singularity. (a) One can go around the

2107: singularity and the two parts are smoothly connected.

2108: (b) The moduli space consists of branches. Any path from one region to

2109: the other must go through the singular point.

2110: Third possibility (not shown in Figure) would be that

2111: the two regions are disconnected.}

2112: \label{singul}

2113: \end{figure}

2114: This possibility is, however, eliminated in the present case

2115: by the jump in the dimension of the complex structure moduli space.

2116: One picture consistent with the jump is that the moduli space consists

2117: of a number of branches (Figure~\ref{singul}(b)),

2118: and the components of the

2119: real moduli space belong to different branches so that

2120: they can have different dimensions.

2121: Another possibility is that the singularity is at infinite distance

2122: and the two components are disconnected.

2123: Of course, the jump does not necessarily occur (an example is the case of

2124: quintic), and in such a case, at this stage

2125: we do not know whether one can go around the singular

2126: loci by turning on RR-potentials.

2127: It is an interesting problem to find out what is the right picture

2128: in full string theory.

2129: 

2130: 

2131: 

2132: 

2133: 

2134: 

2135: \section{Tadpole States of the Gepner Model}

2136: \label{sec:tad}

2137: 

2138: The main purpose of the present paper is to construct

2139: consistent Type II orientifolds on Calabi-Yau manifolds and

2140: Gepner models, with and without spacetime supersymmetry.

2141: In the discussion of consistency and spacetime supersymmetry,

2142: it is useful to study the ``tadpole state'' \cite{PolCai,Callan},

2143: which is the

2144: sum of boundary and crosscap states:

2145: \beq

2146: |T\rangle=|B\rangle+|C\rangle

2147: =|B\rangle_\NSNS+i|B\rangle_\RR

2148: +i|C\rangle_\NSNS+i|C\rangle_\RR

2149: \label{kettot}

2150: \eeq

2151: and the ``bra'' version

2152: \beq

2153: \langle \theta T|=\langle \theta B_{\rm tot}|+\langle \theta C_{\rm tot}|

2154: ={}_\NSNS\!\langle B|

2155: +i{}_\RR\!\langle B|

2156: -i{}_\NSNS\!\langle C|

2157: +i{}_\RR\!\langle C|.

2158: \label{bratot}

2159: \eeq

2160: The NSNS and RR parts are

2161: \beq

2162: \begin{array}{l}

2163: |B\rangle_\NSNS=|B_+\rangle_\NSNS-|B_-\rangle_\NSNS,\\

2164: |B\rangle_\RR=|B_+\rangle_\RR-|B_-\rangle_\RR,\\

2165: |C\rangle_\NSNS=|C_{(-1)^{F_R}P}\rangle

2166: -|C_{(-1)^{F_L}P}\rangle,\\

2167: |C\rangle_\RR=|C_{P}\rangle

2168: -|C_{(-1)^FP}\rangle.

2169: \end{array}

2170: \label{bcdetail}

2171: \eeq

2172: Here $B_+$ and $B_-$ corresponds to

2173: the boundary conditions with the opposite

2174: spin structures, and $P$ is an involutive parity symmetry of the total

2175: system.

2176: Each term on the right hand sides of (\ref{bcdetail}), say

2177: $|C_{P}\rangle$, can be written as the tensor product of the

2178: spacetime part and internal part:

2179: $$

2180: |\mbox{spacetime}\rangle\otimes|\mbox{internal}\rangle.

2181: $$

2182: The spacetime part is associated with the Neumann boundary condition (for

2183: boundary state)

2184: and the standard parity $\Omega$ (for the crosscap states),

2185: and is given by the

2186: standard coherent state of the $D$ free bosons/fermions, the ghost and

2187: the superghosts.

2188: The internal part depends on the detail of D-branes and orientifold.

2189: 

2190: In this section, we construct the internal part of the crosscap

2191: states corresponding to the orientifolds introduced in the previous

2192: section. We also reconstruct the rational boundary states of Gepner 

2193: models from a perspective which is somewhat different from 

2194: the one in the literature. The Cardy-PSS construction 

2195: \cite{Cardy,PSS} and its generalizations are usually 

2196: formulated in the language of purely bosonic rational 

2197: conformal field theories. In particular, in 

2198: \cite{HSS,FS1,FSHSS}, formulas are developed for 

2199: crosscaps and boundary states of rational conformal 

2200: field theories with arbitrary simple-current modular 

2201: invariants. The Gepner model, which is based on an 

2202: $\mathcal N=2$ supersymmetric CFT, can in principle 

2203: be formulated in this language, so that the general 

2204: results are applicable. On the other hand, in 

2205: \cite{BH1}, general results were derived on boundary 

2206: and crosscap states in rational conformal field 

2207: theories directly in the supersymmetric language, and 

2208: using orbifold instead of simple-current techniques. 

2209: In following this approach, we will find that it is a 

2210: lot simpler.

2211: 

2212: 

2213: \subsection{Construction of the Crosscap States}\label{sub:construction}

2214: 

2215: The crosscap states of the Gepner model

2216: can be constructed as a straightforward application of the general

2217: method \cite{BH1}.

2218: Let ${\mathcal X}$ be a bosonic

2219: conformal field theory with a finite abelian symmetry

2220: group $G$ with which one can define an orbifold ${\mathcal X}/G$.

2221: Suppose ${\mathcal X}$ has a parity symmetry $P$ that

2222: commutes with the $G$-projection operator $\sum_{g\in G}g/|G|$.

2223: Then, a parity symmetry is induced in the orbifold theory,

2224: which is denoted again by $P$, with the crosscap state

2225: \beq

2226: |\Scr{C}_P\rangle^{\rm orb}

2227: ={1\over \sqrt{|G|}}\sum_{g\in G}|\Scr{C}_{gP}\rangle.

2228: \label{sum1}

2229: \eeq

2230: Here $|\Scr{C}_{gP}\rangle$ are the crosscap states

2231: for the parity symmetry $gP$ of ${\mathcal X}$, which are

2232: supposed to obey

2233: \beq

2234: \langle \Scr{C}_{gP}|q_t^H|\Scr{C}_{hP}

2235: \rangle

2236: =\Tr\!\!\mathop{}_{{\mathcal H}_{gh^{-1}}}\Bigl[ hPq_l^H\Bigr],

2237: \qquad \forall g,\,

2238: \forall h,

2239: \label{condition1}

2240: \eeq

2241: in which ${\mathcal H}_{gh^{-1}}$ is the space of states on the circle

2242: with $gh^{-1}$-twist.

2243: Indeed the Klein bottle amplitude

2244: \beqa

2245: {}^{\rm orb}\langle \Scr{C}_P|q_t^H|\Scr{C}_P\rangle^{\rm orb}

2246: &=&

2247: {1\over |G|}\sum_{g,h}\langle\Scr{C}_{gP}|q_t^H|

2248: \Scr{C}_{hP}\rangle={1\over |G|}

2249: \sum_{g,h}\Tr\!\!\mathop{}_{{\mathcal H}_{gh^{-1}}}\Bigl[ hPq_l^H\Bigr]\nn\\

2250: &=&\sum_{g'\in G}\Tr\!\!\mathop{}_{{\mathcal H}_{g'}}\left[

2251: \Bigl({1\over |G|}\sum_{h\in G}h\Bigr)Pq_l^H\right]

2252: \nn

2253: \eeqa

2254: is the trace of $Pq_l^H$ over the space of states of the orbifold theory,

2255: ${\mathcal H}^{\rm orb}=\oplus_{g'\in G}{\mathcal H}_{g'}^G$.

2256: The crosscap state for the parity

2257:  dressed with the quantum symmetry $g_{\omega}$

2258: associated with a character $\omega:G\to U(1)$ is

2259: \beq

2260: |\Scr{C}_{g_{\omega}P}\rangle^{\rm orb}

2261: ={1\over \sqrt{|G|}}\sum_{g\in G}\omega(g)^{-1}

2262: |\Scr{C}_{gP}\rangle.

2263: \label{sum2}

2264: \eeq

2265: If ${\mathcal X}$ has fermions, with mod 2 fermion number $(-1)^F$,

2266: the above story applies, with the condition (\ref{condition1}) modified as

2267: \beq

2268: \langle \Scr{C}_{g(\pm)^FP}|q_t^H|\Scr{C}_{hP}

2269: \rangle

2270: =\Tr\!\!\mathop{}_{{\mathcal H}_{(\mp 1)^Fgh^{-1}}}

2271: \Bigl[(-1)^FhPq_l^H\Bigr],

2272: \qquad \forall g,\,

2273: \forall h.

2274: \label{condition2}

2275: \eeq

2276: Note that ${\mathcal H}_{(-1)^F}$ and ${\mathcal H}_{\rm id}$

2277: are the NSNS and the RR sectors respectively.

2278: 

2279: In what follows, we apply this method to the Gepner model,

2280: which is the orbifold of the product of the minimal models

2281: $\prod_{i=1}^rM_{k_i}$ with respect to the group $\Gamma\cong \Z_H$.

2282: %We treat the A-type and B-type parities separately.

2283: 

2284: 

2285: \subsubsection{A-type}

2286: 

2287: We first consider A-parities of the Gepner model

2288: $P^A_{\omega;{\bf m}}=g_{\omega}\tau^A_{\bf m}\Omega_A$, where

2289: $\omega$ (an $H$-th root of unity) parametrizes the quantum symmetry

2290: and ${\bf m}=(m_1,\ldots,m_r)$ labels the $\prod_{i=1}^r\Z_{k_i+2}/\Z_H$

2291: global symmetry.

2292: They are the ones induced from the parity symmetry of the product

2293: theory $\prod_{i=1}^rM_{k_i}$

2294: $$

2295: {\bf P}^A_{\bf m}=(\g^{m_1}P_A,\ldots,\g^{m_r}P_A),

2296: $$

2297: where $P_A$ is the basic A-parity of the ${\mathcal N}=2$ minimal model

2298: associated with the transformation

2299: $X\to \overline{\Omega_A^*X}$ of the LG field.

2300: We note that $\g P_A\g^{-1}=\g^2P_A$.

2301: 

2302: 

2303: The crosscap states for the A-parities $\g^mP_A$ of the minimal model

2304: and

2305: their cousins $(-1)^F\g^mP_A$,

2306: $(\pm 1)^F\g^m\widetilde{P}_A$

2307: (where $\widetilde{P}_A=\e^{-\pi i J_0}P_A$) are

2308: obtained in \cite{BH2} as follows

2309: \beqa

2310: &&|\Scr{C}_{(\pm 1)^F\g^mP_A}\rangle

2311: =|\Scr{C}_{2m-1,-1}(\pm)\rangle,

2312: \label{idCAmm}\\

2313: &&|\Scr{C}_{(\pm 1)^{F}\g^m\widetilde{P}_A}\rangle

2314: =|\Scr{C}_{2m,0}(\mp)\rangle.

2315: \label{idCAmm2}

2316: \eeqa

2317: Here

2318: \beq

2319: |\Scr{C}_{n,s}(\pm)\rangle

2320: =\epsilon^{\pm}_s(-1)^{n-s\over 2}

2321: \left(\,{\sigma^{s}_{n,s}\over\sqrt{2}} 

2322: |\Scr{C}_{n,s}\rangle

2323: \mp {\sigma^{s}_{n,s+2}\over\sqrt{2}}

2324: |\Scr{C}_{n,s+2}\rangle\,\right),

2325: \label{defCns}

2326: \eeq

2327: where

2328: $$

2329: \sigma^{s}_{n,s'}

2330: =\e^{\pi i \left(-{n^2\over 4(k+2)}+{s^2\over 8}\right)}

2331: \e^{-\pi ih_{0ns'}},

2332: $$

2333: $$

2334: \epsilon^{\pm}_{1}=\epsilon^{\pm}_{-1}=1,\qquad

2335: \epsilon^{\pm}_0=\epsilon^{\mp}_2=

2336: \e^{\pm{\pi i\over 4}},

2337: $$

2338: and

2339: $|\Scr{C}_{n,s}\rangle$ are the PSS crosscaps

2340: of the GSO projected model

2341: ${SU(2)_k\times U(1)_2\over U(1)_{k+2}}$,

2342: $$

2343: |\Scr{C}_{n,s}\rangle=\sum_{(l'm's')\in \Mk}{P_{0ns}^{\,\,\,\,l'm's'}\over

2344: \sqrt{S_{000}^{\,\,\,\,l'm's'}}}\cket{l',m',s'}.

2345: $$

2346: In \cite{BH2}, the overall phase of these crosscaps are not determined.

2347: Here, we fix the phases as above, for the following reason.

2348: 

2349: 

2350: The $n$-dependent part of the phase is important

2351: because this will affect the sum over the orbifold group elements,

2352: as in (\ref{sum1}) or (\ref{sum2}).

2353: The above choice is motivated by

2354: the transformation property under $\g$ as well as

2355: the periodicity under $n\to n+2(k+2)$, as we now describe.

2356: One can show that the symmetry $\g$ acts on these states as

2357: \beq

2358: \g:|\Scr{C}_{n,s}(\pm)\rangle\longmapsto |\Scr{C}_{n+4,s}(\pm)\rangle.

2359: \label{ggaction}

2360: \eeq

2361: This is in accord with the fact that $\g P_A\g^{-1}=\g^2P_A$

2362: and hence that $\g|\Scr{C}_{\cdots\g^mP_A}\rangle$ is proportional to

2363: $|\Scr{C}_{\cdots\g^{m+2}P_A}\rangle$.

2364: What (\ref{ggaction}) means is that they are not just

2365: proportional but {\it equal}, under the identification

2366: (\ref{idCAmm})-(\ref{idCAmm2}) with the definition (\ref{defCns}).

2367: Also note the periodicity

2368: \beq

2369: |\Scr{C}_{n+2(k+2),s}(\pm)\rangle=(-1)^s|\Scr{C}_{n,s}(\pm)\rangle.

2370: \label{dperio}

2371: \eeq

2372: The crosscap states that lie in the RR-sector have a double periodicity

2373: $n\equiv n+4(k+2)$.

2374: Thus, with the above choice of the $n$-dependence of the phase,

2375: these crosscaps corresponds precisely

2376: to the oriented O-planes in the LG model;

2377: $|\Scr{C}_{2m-1,-1}(\pm)\rangle$ corresponds to the O-plane at

2378: $X= \e^{\pi i m\over k+2}x$, $x\in \R$, with the orientation 

2379: that goes from positive $x$ to negative $x$.

2380: See Figure~\ref{defOp}.

2381: \begin{figure}[htb]

2382: \centerline{\includegraphics{defOp.eps}}

2383: \caption{The O-plane corresponding to $|\Scr{C}_{2m-1,-1}(+)\rangle$.

2384: This is the example with $k+2=8$ and $m=2$.}

2385: \label{defOp}

2386: \end{figure}

2387: As shown in \cite{BH2}, it

2388: has the right integral RR-charge:

2389: overlap with the RR-boundary states

2390: produces the correct results

2391: for the parity-twisted open string Witten indices ---

2392: the intersection number.

2393: 

2394: 

2395: 

2396: The phase factor $\epsilon_{0}^{\pm}=\epsilon_2^{\mp}=\e^{\pm {\pi i\over 4}}$

2397:  is also added to the NSNS part of

2398: the crosscap state, in order to simplify

2399: the expression of the tension of the O-plane.

2400: In fact, we need the O-plane tension to be real in the end.

2401: Namely, we need the reality of

2402: the overlap of the crosscap with the NSNS ground state $|0\rangle_{\NSNS}$,

2403: in the Gepner model.

2404: With the above choice, the overlap in the minimal model is

2405: \beq

2406: {}_{\NSNS}\langle 0|\Scr{C}_{2m,0}(\pm)\rangle

2407: =\left\{

2408: \begin{array}{ll}

2409: \sqrt{2\over (k+2)\sin({\pi\over k+2})}

2410: \cos\Bigl({\pi\over 2(k+2)}\Bigr)&\mbox{$k$ odd},

2411: \\[0.2cm]

2412: \sqrt{2\over (k+2)\sin({\pi\over k+2})}

2413: \exp\Bigl(\pm {(-1)^m\pi i\over 2(k+2)}\Bigr)

2414: &\mbox{$k$ even}.

2415: \end{array}

2416: \right.

2417: \label{minOtens}

2418: \eeq

2419: If $k$ is odd, this is already real and is therefore the right choice.

2420: If $k$ is even, this is a non-trivial phase. However, as we will see,

2421: in the average over the orbifold group elements $\gamma^{\nu}$

2422: ($\nu\in \Z_H$),

2423: the terms from even $\nu$ and the terms from odd $\nu$

2424: have the opposite phase thanks to the $(-1)^m$ in

2425: (\ref{minOtens}), and the result of the average is real or pure imaginary

2426: depending on $\omega=1$ or $-1$.

2427: Thus, we will simply need to multiply $\omega^{1\over 2}=\pm 1$ or $\pm i$

2428: in the final expression. (See Section~\ref{subsec:total}.)

2429: 

2430: 

2431: 

2432: 

2433: Applying the general method, we find that

2434: the crosscap states

2435: for $g_{\omega}P^A_{\bf m}=P^A_{\omega;{\bf m}}$ and

2436: their cousins (including $\widetilde{P}^A_{\omega;{\bf m}}

2437: =\e^{-\pi i J_0}P^A_{\omega;{\bf m}}$)

2438: are given by

2439: \beqa

2440: &&

2441: |\Scr{C}_{(\pm 1)^FP^A_{\omega;{\bf m}}}\rangle

2442: ={1\over \sqrt{H}}\sum_{\nu=1}^H\omega^{-\nu}

2443: |\Scr{C}_{\gamma^{\nu}(\pm 1)^F{\bf P}^A_{\bf m}}\rangle^{\rm prod},

2444: \label{CARR}

2445: \\

2446: &&|\Scr{C}_{(\pm 1)^{F}\widetilde{P}^A_{\omega;{\bf m}}}\rangle

2447: ={1\over \sqrt{H}}\sum_{\nu=1}^H\omega^{-\nu}

2448: |\Scr{C}_{\gamma^{\nu}(\pm 1)^{F}

2449: \widetilde{\bf P}^A_{\bf m}}\rangle^{\rm prod}.

2450: \label{CANSNS}

2451: \eeqa

2452: in which

2453: \beqa

2454: &&|\Scr{C}_{\gamma^{\nu}(\pm 1)^F{\bf P}^A_{\bf m}}\rangle^{\rm prod}

2455: =(-1)^{\sum_i{\nu\over k_i+2}}\bigotimes_{i=1}^r

2456: |\Scr{C}_{2m_i+2\nu-1,-1}(\pm)\rangle,

2457: \label{CAAA}\\

2458: &&|\Scr{C}_{\gamma^{\nu}(\pm 1)^{F}

2459: \widetilde{\bf P}^A_{\bf m}}\rangle^{\rm prod}

2460: =\bigotimes_{i=1}^r

2461: |\Scr{C}_{2m_i+2\nu,0}(\mp)

2462: \label{CAAA2}

2463: \rangle.

2464: \eeqa

2465: The sign factor $(-1)^{\sum_i{\nu\over k_i+2}}$

2466: is introduced in the RR-crosscap,

2467: in order to maintain the periodicity under $\nu\to \nu+H$,

2468: which can be shown using (\ref{dperio}).

2469: As mentioned above, we need to multiply

2470: the NSNS-part of the crosscap state 

2471: $|\Scr{C}_{(\pm 1)^{F}\widetilde{P}^A_{\omega;{\bf m}}}\rangle$ by

2472: a phase $\omega^{1\over 2}$, in order for the O-plane tension

2473: to be real. This will be taken care of when we discuss the

2474: total crosscap state in string theory

2475: in the last subsection~\ref{subsec:total}.

2476: 

2477: 

2478: 

2479: By the property (\ref{ggaction}) of each factor,

2480: we find the relation

2481: \beq

2482: \gamma^{\nu} |\Scr{C}_{{\bf P}^A}\rangle^{\rm prod}

2483: =|\Scr{C}_{\gamma^{2\nu}{\bf P}^A}\rangle^{\rm prod}

2484: \label{property1}

2485: \eeq

2486: where ${\bf P}^A$ is one of $\gamma^{\nu'}{\bf P}^A_{\bf m}$ or any of

2487: their cousins.

2488: This is in accord with the relation

2489: $\gamma^{2\nu}{\bf P}_A=\gamma^{\nu}{\bf P}^A\gamma^{-\nu}$.

2490: %Note that $\gamma^{2\nu}{\bf P}_A=\gamma^{\nu}{\bf P}^A\gamma^{-\nu}$

2491: %and thus we knew that $|\Scr{C}_{\gamma^{2\nu}{\bf P}^A}\rangle^{\rm prod}$

2492: %is proportional to $\gamma^{\nu}|\Scr{C}_{{\bf P}^A}\rangle^{\rm prod}$.

2493: %What is non-trivial here is that they are {\it equal}

2494: %under the definitions (\ref{CAAA})-(\ref{CAAA2}).

2495: Using this property, we can rewrite

2496: the crosscap states in a useful way:\\[0.1cm]

2497: {\bf\underline{$H$ odd}}\\

2498: If $H$ is odd, the set

2499: $\{\gamma^{\nu} {\bf P}^A\}_{\nu\in \Z_H}$ is the same as

2500: $\{\gamma^{2\nu} {\bf P}^A\}_{\nu\in \Z_H}$.

2501: Then, the crosscap state for the parity $P^A$

2502: induced from ${\bf P}^A$

2503: is simply

2504: \beq

2505: |\Scr{C}_{P^A}\rangle

2506: ={1\over \sqrt{H}}\sum_{\nu=1}^H\gamma^{\nu}

2507: |\Scr{C}_{{\bf P}^A}\rangle^{\rm prod}.

2508: \label{Cexodd}

2509: \eeq

2510: This state is manifestly $\Gamma$-invariant.

2511: Note that there is no involutive dressing by quantum symmetry if $H$ is odd.

2512: \\[0.1cm]

2513: {\bf\underline{$H$ even}}\\

2514: If $H$ is even, $\{\gamma^{\nu} {\bf P}^A\}_{\nu\in \Z_H}$

2515: splits into the union of $\{\gamma^{2\nu} {\bf P}^A\}_{\nu\in \Z_{H/2}}$ and

2516: $\{\gamma^{2\nu+1} {\bf P}^A\}_{\nu\in \Z_{H/2}}$.

2517: The crosscap states for the parity $P^A$ or the one dressed by the

2518: involutive quantum symmetry $g_{-1}$ are

2519: given by

2520: \beq

2521: |\Scr{C}_{g_{\pm}P^A}\rangle

2522: ={1\over\sqrt{H}}\sum_{\nu=1}^{H/2}\Biggl\{

2523: \gamma^{\nu}|\Scr{C}_{{\bf P}^A}\rangle^{\rm prod}

2524: \pm \gamma^{\nu}|\Scr{C}_{\gamma {\bf P}^A}\rangle^{\rm prod}\Biggr\},

2525: \label{Cexeve}

2526: \eeq

2527: which is also manifestly $\Gamma$-invariant.

2528: 

2529: 

2530: 

2531: 

2532: \subsubsection{B-type}

2533: 

2534: The B-parities $P^B_{\omega;{\bf m}}$ of the Gepner model are the ones induced

2535: from the parity symmetry of the product theory

2536: $$

2537: {\bf P}_{\bf m}^B=(\g^{m_1}P_B,\ldots,

2538: \g^{m_r}P_B),

2539: $$

2540: where $P_B$ is the basic B-parity of the minimal model

2541: associated with the transformation $X\to\e^{\pi i\over k+2}\Omega_B^*X$

2542: of the LG field.

2543: 

2544: The crosscap states for the B-parities $\g^mP_B$ of the minimal model

2545: and their cousins $(-1)^F\g^mP_B$,

2546: $(\pm 1)^F\g^m\widetilde{P}_B$ (where

2547: $\widetilde{P}_B=\e^{\pi i J_0}P_B$)

2548: are

2549: obtained in \cite{BH2} as follows

2550: \beqa

2551: &&|\Scr{C}_{(\pm 1)^F\g^mP_B}\rangle

2552: =|\Scr{C}^B_{m,1}(\pm)\rangle,\\

2553: &&|\Scr{C}_{(\pm 1)^{F}\g^m\widetilde{P}_B}\rangle

2554: =|\Scr{C}^B_{m,0}(\mp)\rangle.

2555: \eeqa

2556: Here, for $r\in \Z_{k+2}$ and $p\in \{0,1\}$

2557: \beq

2558: |\Scr{C}^B_{r,p}(\pm)\rangle

2559: ={1\over \sqrt{k+2}}\sum_{\nu \in\Z_{k+2}}

2560: \e^{2\pi i{\nu (r+p/2)\over k+2}}V_M|\Scr{C}_{2\nu-p,-p}(\pm)\rangle

2561: \eeq

2562: where $|\Scr{C}_{n,s}(\pm)\rangle$ is the A-type crosscap

2563: defined in (\ref{defCns}) and $V_M$ is the mirror automorphism.

2564: More explicitly,

2565: $|\Scr{C}_{r,p}^B(\pm)\rangle

2566: =\epsilon^{B\pm}_p\e^{\pi i p{2r+p\over k+2}}

2567: \left({1\over \sqrt{2}}|\Scr{C}^B_{r0p}\rangle

2568: \pm {i\over \sqrt{2}}|\Scr{C}^B_{r1p}\rangle\right)$

2569: in which

2570: $\epsilon_1^{B\pm}=\pm\e^{\mp{\pi i\over 4}}$, $\epsilon^{B\pm}_0=1$

2571: and

2572: $$

2573: |\Scr{C}_{rqp}^B\rangle:=

2574: {1\over 2(k+2)}\sum_{n,s\atop

2575: {\rm even}}\e^{-\pi i\theta_{rq}(n,s)+\pi i \widehat{Q}_{(0pp)}(0ns)}

2576: V_M|\Scr{C}_{n+p.s+p}\rangle

2577: =\e^{-\pi i p{\widehat{2r+p}\over 2(k+2)}

2578: +\pi i p{\widehat{2q+p}\over 4}}

2579: |\Scr{C}_{rqp}\rangle',

2580: $$

2581: where

2582: $$

2583: |\Scr{C}_{rqp}\rangle'

2584: = (2(k+2))^{\frac{1}{4}}

2585: \sum_{j} \sigma_{j,2r+p,2q+p} 

2586: \ \frac{P_{\frac{k}{2}j}}{\sqrt{S_{0j}}}

2587: (-1)^{\frac{\widehat{2r+p}-p}{2} +q} 

2588: \cket{j,2r+p,2q+p}_B

2589: $$

2590: is the state denoted by $|\Scr{C}_{rqp}\rangle$ in \cite{BH2}.

2591: \footnote{Here $\theta_{rq}(n,s)=-{rn\over k+2}+{qs\over2}$ and

2592: $\widehat{Q}_{a}(b)=h_a+h_b-h_{a+b}$,

2593: and $\sigma_{j,n,s}=(-1)^{h_{j,n,s}-h_j+h_n-h_s}$

2594:  \cite{BH2}.

2595: Also, $\widehat{2r+p}$ is $2r+p$ (mod $2(k+2)$) brought into

2596:  the standard range

2597: $[-k-1,k+2]$.

2598: Same for $\widehat{2q+p}$ (mod 4).}

2599: 

2600: 

2601: 

2602: Applying the general method, we find that

2603: the crosscap states for $(\pm 1)^FP^B_{\omega;{\bf m}}$

2604: and $(\pm 1)^{F}\widetilde{P}^B_{\omega;{\bf m}}$ are

2605: \beqa

2606: &&|\Scr{C}_{(\pm 1)^FP^B_{\omega;{\bf m}}}\rangle

2607: ={1\over \sqrt{H}}\sum_{\nu=1}^H\omega^{-\nu}

2608: |\Scr{C}_{\gamma^{\nu}(\pm 1)^F{\bf P}^B_{\bf m}}\rangle^{\rm prod},

2609: \label{CGB1}

2610: \\

2611: &&|\Scr{C}_{(\pm 1)^{F}\widetilde{P}^B_{\omega;{\bf m}}}\rangle

2612: ={1\over \sqrt{H}}\sum_{\nu=1}^H\omega^{-\nu}

2613: |\Scr{C}_{\gamma^{\nu}(\pm 1)^{F}

2614: \widetilde{\bf P}^B_{\bf m}}\rangle^{\rm prod}.

2615: \label{CGB2}

2616: \eeqa

2617: in which

2618: \beqa

2619: &&|\Scr{C}_{\gamma^{\nu}(\pm 1)^F{\bf P}^B_{\bf m}}\rangle^{\rm prod}

2620: =\bigotimes_{i=1}^r

2621: |\Scr{C}^B_{m_i+\nu,1}(\pm)\rangle,\\

2622: &&|\Scr{C}_{\gamma^{\nu}(\pm 1)^{F}

2623: \widetilde{\bf P}^B_{\bf m}}\rangle^{\rm prod}

2624: =\bigotimes_{i=1}^r

2625: |\Scr{C}^B_{m_i+\nu,0}(\mp)\rangle.

2626: \eeqa

2627: 

2628: Alternatively one could start with the A-type parities in the mirror

2629: Gepner model, which is the orbifold of the product of minimal models with

2630: the group $\tilGamma\subset \prod_i\Z_{k_i+2}$,

2631: $\tilnu=(\tilnu_1,...,\tilnu_r)\in \tilGamma$ $\Leftrightarrow$

2632: $\sum_{i=1}^r{\nu_i\over k_i+2}\in\Z$

2633:  (see Section~\ref{subsub:mirror}).

2634: In the mirror system, the global symmetries are parametrized by

2635: ${\bf \tilm}=(\tilm_1,...,\tilm_r)\in\prod_i\Z_{k_i+2}$ modulo

2636: shift by $\tilnu=(\tilnu_1,...,\tilnu_r)\in \tilGamma$,

2637: and the quantum symmetries are parametrized by

2638: $\tilomega=(\tilomega_1,...,\tilomega_r)$, $\tilomega_i^{k_i+2}=1$,

2639: modulo the relation $\tilomega=\tilomega'$ $\Leftrightarrow$

2640: $\prod_i(\tilomega_i'\tilomega_i^{-1})^{\tilnu_i}=1$

2641: ($\forall\tilnu\in\tilGamma$).

2642: They correspond respectively to the quantum symmetries $\omega$

2643: and global symmetries ${\bf m}$

2644: of the original Gepner model, under the map

2645: \beqa

2646: &&\exp\left(-2\pi i {m_i\over k_i+2}\right)=\tilomega_i.

2647: \label{id1}\\

2648: &&\omega=\exp\left(2\pi i \sum_{i=1}^r{\tilm_i\over k_i+2}\right),

2649: \label{id2}

2650: \eeqa

2651: Then, the parity $P^B_{\omega;{\bf m}}$ of the Gepner model

2652: corresponds to

2653: the parity $P^A_{{\tilomega};{\bf\tilm}}$ of the mirror Gepner model,

2654: whose crosscap states are given by

2655: \beqa

2656: &&

2657: |\Scr{C}_{(\pm 1)^FP^A_{\tilomega;{\bf \tilm}}}\rangle

2658: ={1\over \sqrt{|\tilGamma|}}\sum_{\tilnu\in\tilGamma}

2659: \tilomega^{-\tilnu}

2660: (-1)^{\sum_i{\tilnu_i\over k_i+2}}\bigotimes_{i=1}^r

2661: |\Scr{C}_{2\tilm_i+2\tilnu_i-1,-1}(\pm)\rangle

2662: \\

2663: &&|\Scr{C}_{(\pm 1)^{F}\widetilde{P}^A_{\tilomega;{\bf \tilm}}}\rangle

2664: ={1\over \sqrt{|\tilGamma|}}\sum_{\tilnu\in\tilGamma}\tilomega^{-\tilnu}

2665: \bigotimes_{i=1}^r

2666: |\Scr{C}_{2\tilm_i+2\tilnu_i,0}(\mp)

2667: \rangle,

2668: \eeqa

2669: where $\tilomega^{-\tilnu}$ is the short-hand notation for

2670: $\prod_i\tilomega_i^{-\tilnu_i}$.

2671: It is straightforward to show that

2672: \beqa

2673: &&|\Scr{C}_{(\pm 1)^FP^B_{\omega;{\bf m}}}\rangle

2674: =

2675: \tilomega^{-\bf\tilm}\e^{\pi i\sum_i{\tilm_i\over k_i+2}}

2676: V_M|\Scr{C}_{(\pm 1)^FP^A_{\tilomega;{\bf \tilm}}}\rangle

2677: \\

2678: &&|\Scr{C}_{(\pm 1)^{F}\widetilde{P}^B_{\omega;{\bf m}}}\rangle

2679: =\tilomega^{-\bf\tilm}

2680: V_M|\Scr{C}_{(\pm 1)^{F}\widetilde{P}^A_{\tilomega;{\bf \tilm}}}\rangle

2681: \label{BvsAtil}

2682: \eeqa

2683: The two sets of crosscap states differ by phases.

2684: Actually, by the condition that $P^B_{\omega;{\bf m}}$ 

2685: is involutive, $m_i$ must be $0$ or $(k_i+2)/2$, or

2686: $\tilomega_i=\pm 1$. Thus the overall

2687: factor $\tilomega^{-{\bf\tilm}}$ common to all the four states

2688: is just a sign.

2689: On the other hand, the factor $\e^{\pi i\sum_i{\tilm_i\over k_i+2}}$

2690: is a non-trivial phase that makes a real difference.

2691: It turns out that the one obtained from the mirror theory

2692: is the right choice.

2693: This can be understood by noting that the parity twisted Witten indices

2694: for open strings is automatically integral in the A-type crosscaps.

2695: Thus, we modify the RR part of the crosscap in the original construction

2696: as

2697: \beq

2698: (\ref{CGB1})\longrightarrow\qquad

2699: |\Scr{C}_{(\pm 1)^FP^B_{\omega;{\bf m}}}\rangle

2700: ={1\over \sqrt{H}}\sum_{\nu=1}^H\omega^{-\nu-{1\over 2}}

2701: |\Scr{C}_{\gamma^{\nu}(\pm 1)^F{\bf P}^B_{\bf m}}\rangle^{\rm prod},\qquad

2702: \eeq

2703: where $\omega^{-{1\over 2}}$ is identified as

2704: $\e^{-\pi i\sum_i{\tilm_i\over k_i+2}}$.

2705: 

2706: 

2707: 

2708: \subsection{Boundary States}\label{sub:Cardybranes}

2709: 

2710: The construction of the boundary states is likewise a straightforward

2711: application of the general method.

2712: Let ${\mathcal X}$ and $G$ be as in the beginning of the previous subsection.

2713: We are interested in constructing a D-brane boundary state

2714: in ${\mathcal X}/G$ from a D-brane $\Scr{B}$ of ${\mathcal X}$.

2715: If $\Scr{B}$ is not invariant under any element of $G$,

2716: $\Scr{B}\mapsto g\Scr{B}\ne \Scr{B}$ ($g\ne 1$), the boundary state is

2717: simply the sum over the images

2718: \beq

2719: |\Scr{B}\rangle^{\rm orb}={1\over \sqrt{|G|}}

2720: \sum_{g\in G}g|\Scr{B}\rangle.

2721: \label{longgeneral}

2722: \eeq

2723: If $\Scr{B}$ is invariant under a subgroup $H$ of $G$, 

2724: the boundary states is the sum over images $g\Scr{B}$ ($g\in G/H$)

2725: as well as the twist $h\in H$

2726: \beq

2727: |\Scr{B}\rangle^{\rm orb}

2728: ={1\over \sqrt{|G|}}\sum_{g\in G/H\atop h\in H}

2729: g|\Scr{B}\rangle_h.

2730: \label{shortgeneral}

2731: \eeq

2732: Here $|\Scr{B}\rangle_h$ is the

2733: boundary state for $\Scr{B}$ on the $h$-twisted circle.

2734: They are assumed to obey the relation

2735: $$

2736: {}_h\!\langle\Scr{B}|q_t^Hg|\Scr{B}\rangle_h

2737: =\Tr\!\!\mathop{}_{{\mathcal H}_{\Scr{B},g\Scr{B}}}\left[ 

2738: hq_l^H\right]

2739: $$

2740: where ${\mathcal H}_{\Scr{B},g\Scr{B}}$ is the space of $\Scr{B}$-$g\Scr{B}$

2741: open string states. Indeed the cylinder amplitude

2742: \beqa

2743: {}^{\rm orb}\!\langle\Scr{B}|q_t^H|\Scr{B}\rangle^{\rm orb}

2744: &=&{1\over |G|}\sum_{g_1,g_2\in G/H\atop

2745: h\in H}

2746: {}_h\!\langle \Scr{B}|g_1^{-1}q_t^Hg_2

2747: |\Scr{B}\rangle_h

2748: ={1\over |H|}\sum_{g\in G/H\atop h\in H}

2749: {}_h\!\langle \Scr{B}|q_t^Hg|\Scr{B}\rangle_h

2750: \nn\\

2751: &=&

2752: \sum_{g\in G/H}\Tr\!\!\mathop{}_{{\mathcal H}_{\Scr{B},g\Scr{B}}} 

2753: \left[\left({1\over |H|}\sum_{h\in H}h\right)q^H\right]

2754: \nn

2755: \eeqa

2756: is the sum over all possible pairs $\Scr{B}$-$g\Scr{B}$

2757: of the trace over the $H$-invariant open string

2758: states.

2759: If we modify the action of $H$ on the Chan-Paton factor,

2760: we obtain a different brane.

2761: The brane associated with the proper $H$-action given by

2762: an $H$-character $h\mapsto \psi(h) \in U(1)$

2763: has the boundary state

2764: \beq

2765: |\Scr{B}^{\psi}\rangle^{\rm orb}

2766: ={1\over \sqrt{|G|}}\sum_{g\in G/H\atop h\in H}

2767: \psi(h) g|\Scr{B}\rangle_h.

2768: \label{shortgeneral2}

2769: \eeq

2770: 

2771: One may also want to include the effects of discrete torsion.

2772: As argued in \cite{discretetorsion}, this corresponds to having

2773: a projective representation of the orbifold group on the space

2774: of open string states. At the level of boundary states, one uses the

2775: alternating bihomomorphism $\epsilon:H\times H\longrightarrow

2776: U(1)$ obtained from the discrete torsion to define a subgroup 

2777: $K$ of $H$ for each $G$-orbit of branes as

2778: \beq

2779: K=\{k\in H; \epsilon(k,h)=1 \;\forall h\in H\}\,.

2780: \label{untwisted}

2781: \eeq

2782: One obtains one elementary brane in the orbifold theory for

2783: each character $k\mapsto\psi(k)\in U(1)$ of $K$. The corresponding

2784: boundary state is a slight modification of (\ref{shortgeneral2})

2785: \beq

2786: |\Scr{B}^{\psi}\rangle^{\rm orb}

2787: ={1\over \sqrt{|G|}}\sqrt{\frac{|H|}{|K|}}

2788: \sum_{g\in G/H\atop h\in H}

2789: \psi(h) g|\Scr{B}\rangle_h,

2790: \label{shortdiscrete}

2791: \eeq

2792: where the extension of $\psi$ to $H\setminus K$ is irrelevant.

2793: We note that by general properties of group cohomology, the

2794: factor $\sqrt{|H|/|K|}$ is always integer.

2795: 

2796: Rational boundary states of the Gepner model can be obtained

2797: as an application of the methods described above. A-branes in 

2798: the product theory are generically not invariant under any element 

2799: of the orbifold group (the case $H=\{{\rm id}\}$) and thus the

2800: boundary states in the orbifold model are simply the sum over 

2801: images. This is how these states were first written down in 

2802: \cite{RS}. B-branes of the product theory, on the other hand,

2803: are all $\Gamma$-invariant (the case $H=G$) and therefore 

2804: the boundary states are simply the sum over $\Gamma$-twists.

2805: This way of obtaining the B-type boundary states appears to

2806: be new, but we have found it to to be equivalent to the procedure

2807: developed in \cite{RS,FSHSS,FKLLSW}. In particular, as we 

2808: will see, the fixed point resolution prescription of \cite{FKLLSW} 

2809: is correctly reproduced.

2810: %

2811: 

2812: \subsubsection{A-Branes}\label{subsub:A-branes}

2813: 

2814: A-branes in the minimal model are denoted as

2815: $\Scr{B}^A_{L,M,S}$ and are labeled by $(L,M,S)\in \Mk$.

2816:  Shift of $S$-label by $2$

2817: is simply the orientation change ---

2818: the sign flip of RR boundary states.

2819: $\Scr{B}^A_{L,M,S}$ preserve the combination $\overline{G}_+-(-1)^{S}G_-$

2820: of the $(2,2)$ superconformal symmetry.

2821: The symmetry $\g$ shifts the $M$-label by $2$.

2822: The boundary states on the NSNS and RR sectors are given by

2823: \beqa

2824: &&

2825: |\Scr{B}_{L,M}^A\rangle_\NSNS={1\over \sqrt{2}}|\Scr{B}_{L,M,S}\rangle

2826: +{1\over\sqrt{2}}|\Scr{B}_{L,M,S+2}\rangle,

2827: \nn\\

2828: &&

2829: |\Scr{B}_{L,M,S}^A\rangle_\RR={1\over \sqrt{2}}|\Scr{B}_{L,M,S}\rangle

2830: -{1\over\sqrt{2}}|\Scr{B}_{L,M,S+2}\rangle,

2831: \nn

2832: \eeqa

2833: where

2834: $|\Scr{B}_{L,M,S}\rangle$ are the standard Cardy brane

2835: of the GSO projected model

2836: ${SU(2)_k\times U(1)_2\over U(1)_{k+2}}$:

2837: $$

2838: |\Scr{B}_{L,M,S}\rangle

2839: =\sum_{(lms)\in \Mk}{S_{LMS}^{\,\,\,\,lms}\over

2840: \sqrt{S_{000}^{\,\,\,\,lms}}}\kket{l,m,s}.

2841: $$

2842: In the LG model, they correspond to the D1-brane

2843: at the wedge-shaped lines cornering at $X=0$.

2844: The wedge corresponding to

2845: $\Scr{B}_{L,M,S=1}^A$ is coming in from the direction

2846: $\arg(X)={\pi(M-L-1)\over k+2}$ and going out to the direction

2847: $\arg(X)={\pi (M+L+1)\over k+2}$.

2848: Replacing $S=1$ by $S=-1$ flips the orientation.

2849: See Figure~\ref{Abr}.

2850: \begin{figure}[htb]

2851: \centerline{\includegraphics{Abr.eps}}

2852: \caption{The A-brane $\Scr{B}^A_{L,M,S}$. This is the example

2853: $(L,M,S)=(2,5,1)$ for $k+2=8$.}

2854: \label{Abr}

2855: \end{figure}

2856: 

2857: We are interested in the D-brane of the Gepner model corresponding to the

2858: product brane

2859: $$

2860: \Scr{B}^A_{{\bf L,M},S}=\Scr{B}^A_{L_1,M_1,S_1}\times\cdots

2861: \times\Scr{B}^A_{L_r,M_r,S_r}.

2862: $$

2863: We need $S_i$ all even or all odd (i.e. $L_i+M_i$ all even or all odd),

2864: so that either

2865: one of $\overline{G}_+\mp G_-$ is preserved.

2866: Orientation flip of even number of

2867: factors does not change the total orientation.

2868: Thus the brane depends only on the total orientation $S=[{\bf S}]$

2869: where $[{\bf S}]\equiv [{\bf S}']$ if the number of factors with

2870: $S_i'=S_i+2$ is even. (If $r$ is odd, $S$ can be realized as

2871: the sum $S=\sum_iS_i\in\Z_4$.)

2872: The orbifold group element $\gamma^{\nu}$ sends $\Scr{B}^A_{{\bf L,M},S}$

2873: to $\Scr{B}^A_{{\bf L,M+2\nu},S}$ (where ${\bf 2\nu}=(2\nu,2\nu,...,2\nu)$).

2874: Thus generically, $\Scr{B}^A_{{\bf L,M},S}$

2875: is not invariant under any element of the orbifold group.

2876: Then, the boundary state is simply the sum over images

2877: \beqa

2878: &&|\Scr{B}^A_{\bf L,M}\rangle_{\NSNS}

2879: ={1\over \sqrt{H}}\sum_{\nu=1}^{H}

2880: \bigotimes_{i=1}^r

2881: |\Scr{B}^A_{L_i,M_i+2\nu}\rangle_{\NSNS}

2882: \label{AbrNSNS}

2883: \\

2884: &&|\Scr{B}^A_{{\bf L,M},S}\rangle_{\RR}

2885: ={1\over \sqrt{H}}\sum_{\nu=1}^{H}

2886: \bigotimes_{i=1}^r

2887: |\Scr{B}^A_{L_i,M_i+2\nu,S_i}\rangle_{\RR}.

2888: \label{AbrRR}

2889: \eeqa

2890: It is a simple exercise to show that, adding

2891: the transverse modes of the spacetime and imposing chiral GSO projection,

2892: these lead to the boundary states obtained by Recknagel and Schomerus

2893: \cite{RS}.

2894: 

2895: The construction is different if

2896: the product brane is invariant under a

2897: non-trivial orbifold group element,

2898: $\Scr{B}^A_{{\bf L,M+2\nu},S}=\Scr{B}^A_{{\bf L,M},S}$, see \cite{BS,FSW}.

2899: Branes of different $M$ labels can be the same

2900: because of the ``Field Identification'' (FI)

2901: $(L,M,S)=(k-L,M+k+2,S+2)$.

2902: Since FI is involutive we find that $2\nu=0$ mod $H$. Thus, the stabilizer

2903: group is at most the $\Z_2$ subgroup generated by $\gamma^{H\over 2}$,

2904: which is possible only

2905: when $H$ is even. 

2906: Under the symmetry $\gamma^{H/2}$, the brane $\Scr{B}^A_{{\bf L,M},S}$

2907: transforms to $\Scr{B}^A_{{\bf L,M+{(k+2)}{{\it H}\over k+2}},S}$.

2908: For the factor $i$ such that $w_i={H\over k_i+2}$ is even,

2909: the brane remains the same because

2910: $\Scr{B}_{L_i,M_i+2(k_i+2),S_i}=\Scr{B}_{L_i,M_i,S_i}$. For the

2911: factor $i$ such that $w_i$ is odd, the brane $\Scr{B}_{L_i,M_i,S_i}$

2912: is transformed to $\Scr{B}_{L_i,M_i+k_i+2,S_i}$.

2913: In the LG picture, $M_i\to M_i+k_i+2$ is rotation by $\pi$, under which

2914: a ``straight-wedge brane'' ($L_i={k_i\over 2}$)

2915: is mapped to itself with an orientation flip. See Fig.~\ref{rotation}.

2916: \begin{figure}[htb]

2917: \centerline{\includegraphics{rotation.eps}}

2918: \caption{180${}^{\circ}$ rotation reverses the orientation of the

2919: ``straight-wedge'' branes.}

2920: \label{rotation}

2921: \end{figure}

2922: Note that there are even number of $i$'s such that $w_i$ is odd.

2923: (Remark (ii) in Section~\ref{subsec:CYLG}.)

2924: The total orientation is preserved even if the orientation is

2925: reserved for each of such $i$.

2926: Thus, the brane $\Scr{B}^A_{{\bf L,M},S}$ is invariant under

2927: $\gamma^{H/2}$ if and only if $L_i={k_i\over 2}$ for such $i$

2928: that $w_i$ is odd.

2929: The boundary state is obtained as the application of the general formula

2930: (\ref{shortgeneral2}):

2931: \beqa

2932: &&

2933: |\widehat{\Scr{B}}^{(\pm)\,A}_{{\bf L,M}}\rangle_{\NSNS}

2934: ={1\over \sqrt{H}}\sum_{\nu=1}^{H\over 2}

2935: \gamma^{\nu}|\Scr{B}^A_{{\bf L,M}}\rangle^{\rm prod}_{\NSNS}

2936: \pm{1\over\sqrt{H}}

2937: \sum_{\nu=1}^{H\over 2}\gamma^{\nu}

2938: |\Scr{B}^A_{{\bf L,M}}

2939: \rangle^{\rm prod}_{(-1)^{F}\gamma^{H/2}},

2940: \label{shANSNS}

2941: \\

2942: &&

2943: |\widehat{\Scr{B}}^{(\pm)\,A}_{{\bf L,M},S}\rangle_{\RR}

2944: ={1\over \sqrt{H}}\sum_{\nu=1}^{H\over 2}

2945: \gamma^{\nu}|\Scr{B}^A_{{\bf L,M},S}\rangle^{\rm prod}_{\RR}

2946: \pm{1\over\sqrt{H}}

2947: \sum_{\nu=1}^{H\over 2}\gamma^{\nu}

2948: |\Scr{B}^A_{{\bf L,M},S}

2949: \rangle^{\rm prod}_{\gamma^{H/2}},

2950: \label{shARR}

2951: \eeqa

2952: Note that the untwisted part is simply one half of

2953: (\ref{AbrNSNS}) or (\ref{AbrRR}).

2954: The boundary state

2955: $|\Scr{B}_{{\bf L,M},S}\rangle^{\rm prod}_{(\mp 1)^{F}\gamma^{H/2}}$

2956: is given by the product of the twisted states

2957: $|\Scr{B}_{L_i,M_i,S_i}\rangle_{(\mp 1)^{F}\tilde{\g}^{H/2}}$

2958: of the minimal model.

2959: Here we replaced $\g$ by $\tilde{\g}$ 

2960: where $\tilde{\g}=g$ if $w_i$ is even while

2961: $\tilde{g}=a^2=\e^{-2\pi i J_0}

2962: =\g(-1)^{\widehat{F}}$ if $w_i$ is odd.

2963: This is because

2964: $a^{k_i+2}$ preserves the $L_i={k_i\over 2}$

2965: branes including the orientation.

2966: (Since there are even number of $i$'s with odd $w_i$,

2967: this makes no difference.)

2968: The twist $\tilde{\g}^{H/2}$

2969: is trivial for such $i$ that $w_i$ is even,

2970: while it is non-trivial, $\tilde{\g}^{H/2}=a^{k_i+2}$,

2971: for such $i$ that $w_i$ is odd.

2972: The $a^{k+2}$-twisted boundary state

2973: in the minimal model is given by

2974: \beq

2975: |\Scr{B}_{{k\over 2},M,S}^A\rangle_{(\mp 1)^Fa^{k+2}}

2976: =\sum_{{k\over 2}+m+s\,{\rm even}\atop

2977: s\,{\rm even/odd}}

2978: \e^{\pi i \left(-{M+S^2+Ss+m\over 2}+{Mm+m\over k+2}\right)}

2979: \kket{{k\over 2},m,s}.

2980: \eeq

2981: Since the length of the sum over images is one half of

2982: the ordinary branes, these branes are called

2983: {\it short orbit branes}.

2984: ``Field Identification'' is a little different on these short orbit branes.

2985: $(L_i,M_i,S_i)\to (k_i-L_i,M_i+k_i+2,S_i+2)$ does not change the brane

2986: if $w_i$ is even, but exchanges $+$ and $-$ label if

2987: we do this for odd number of $i$'s with odd $w_i$.

2988: 

2989: 

2990: 

2991: \subsubsection{B-Branes}

2992: 

2993: 

2994: B-branes in the minimal model can be obtained as the

2995: mirror of the A-branes in the $\Z_{k+2}$-orbifold model,

2996: see \cite{RS,FKLLSW,MMS,BH1}.

2997: (They can also be obtained directly as an application of the methods

2998: of \cite{FS1} by using the results of \cite{RRW} on the $\Z_2$ 

2999: orbifold of the minimal models by mirror symmetry automorphism.)

3000: The mirror of the brane associated with $\Scr{B}^A_{L,M,S}$ is denoted 

3001: as $\Scr{B}^B_{L,M,S}$ and

3002: they preserve the combination $G_+-(-1)^SG_-$ of the worldsheet

3003: superconformal symmetry.

3004: They are invariant under the symmetry $\g$, and

3005: the boundary states on the various twisted circles

3006: are given by

3007: \beqa

3008: |\Scr{B}_{L,M,S}^B\rangle_{(-1)^{(s'+1)F}\g^{n'}}

3009: &=&{1\over \sqrt{k+2}}V_M\sum_{\tilnu\in \Z_{k+2}}

3010: \e^{2\pi i{\tilnu n'\over k+2}}

3011: |\Scr{B}^A_{L,M+2\tilnu,S}\rangle_{(-1)^{(s'+1)F}}

3012: \nn\\

3013: &=&(2k+4)^{{1\over 4}}\e^{-\pi i{Mn'\over k+2}+\pi i {Ss'\over 2}}

3014: \sum_{l'\in \Pk\atop \nu_1\in \Z_2}

3015: {S_{Ll'}\over\sqrt{S_{0l'}}}(-1)^{S\nu_1}

3016: \kket{l',n',s'+2\nu_1}_B,

3017: \nn\\

3018: \eeqa

3019: where $s'=0$ for NSNS sector and $s'=1$ for RR sector.

3020: Note that the $M$-label appears only on the overall phase

3021: for the boundary state with a non-trivial twist $g^{n'}\ne 1$.

3022: Thus, the brane themselves depend only on $(L,S)$

3023: but the $M$-label parametrizes

3024:  the action of the global symmetry $\g$ on

3025: the Chan-Paton factor. There is no short orbit branes

3026: in the na\"\i ve sense,

3027: since none of the A-branes is invariant under any

3028: non-trivial element of $\Z_{k+2}$. However, in the GSO projected model

3029: (i.e. in the coset model), B-branes are realized as the mirror of the

3030: A-branes in the $\Z_{k+2}\times \Z_2$ orbifold, and for even $k$

3031: the Cardy branes $\Scr{B}_{{k\over 2},M,S}$ are invariant under

3032: the $\Z_2$ subgroup generated by $g_{k+2,2}$.

3033: Thus, there are short-orbit branes in the coset model, and resolving the GSO

3034: projection we obtain short orbit branes

3035: $\widehat{\Scr{B}}_{{k\over 2},M,S}^B$ in the minimal model.

3036: They are invariant under the symmetry 

3037: $a^2=\e^{-2\pi i J_0}=\g(-1)^{\widehat{F}}$.

3038: The boundary states on the various twisted circles are \cite{BH2}

3039: \beqa

3040: &&\!\!\!\!

3041: |\widehat{\Scr{B}}_{{k\over 2},M,S}^B\rangle_{(-1)^{(n'+1)F}a^{2n'}}

3042: ={1\over\sqrt{2}}|\Scr{B}^B_{{k\over 2},M,S}\rangle_{(-1)^{(n'+1)F}g^{n'}}

3043: \nn\\

3044: &&\!\!\!\!

3045: |\widehat{\Scr{B}}_{{k\over 2},M,S}^B\rangle_{(-1)^{n'F}a^{2n'}}

3046: =\e^{-\pi i {Mn'+n'\over k+2}+\pi i {Sn'+n'\over 2}}

3047: \sqrt{k+2\over 2}\sum_{s=\pm 1}\e^{-\pi i{S(S-s)\over 2}}

3048: \kket{{k\over 2},{k+2\over 2}+n',s+n'}_B.

3049: \nn

3050: \eeqa

3051: 

3052: The long orbit brane $\Scr{B}^B_{L,M,1}$ is

3053: described in the LG model $W=X^{k+2}$

3054: as the one associated with the boundary

3055: superpotential

3056: $$

3057: V={\mit\Gamma}X^{L+1}

3058: $$

3059: where ${\mit\Gamma}$ is a fermionic chiral superfield on the boundary

3060: with constraint $\overline{D}{\mit\Gamma}=X^{k+1-L}$

3061: and $M$ labels the action of $\g:X\to \e^{2\pi i\over k+2}X$

3062: on the Chan-Paton ground state $|0\rangle$ (annihilated by the lowest

3063: component of ${\mit\Gamma}$) as

3064: $$

3065: \g:|0\rangle\longmapsto\e^{-{2\pi i\over k+2}{M+L+1\over 2}}|0\rangle.

3066: $$

3067: On the other hand, short orbit branes are not realized in the LG model

3068: with $W=X^{k+2}$ but in the model with $W=X^{k+2}-Y^2$ that also flows to

3069: the ${\mathcal N}=2$ minimal model.

3070: They are associated with the boundary superpotential

3071: $$

3072: V={\mit\Gamma}(X^{k+2\over 2}-Y), \qquad

3073: \overline{D}{\mit\Gamma}=X^{k+2\over 2}+Y.

3074: $$

3075: 

3076: 

3077: 

3078: 

3079: 

3080: In the open string

3081: stretched between long and short orbit branes, there are odd number of

3082: real fermionic zero modes \cite{BH2}.

3083: This imposes a strong constraint in the construction of consistent set of

3084: D-branes.

3085: 

3086: 

3087: 

3088: Let us first consider the product of long-orbit branes

3089: $$

3090: \Scr{B}^B_{{\bf L},M,S}=\Scr{B}^B_{L_1,M_1,S_1}\times\cdots\times

3091: \Scr{B}^B_{L_r,M_r,S_r}.

3092: $$

3093: $S_i$ are all even or all odd and the brane

3094: depends only on the total orientation $S=[{\bf S}]$.

3095: Also, the $\Gamma$ action on the Chan-Paton factor depends only on

3096: \beq

3097: M:=H\sum_{i=1}^r{M_i\over k_i+2}\in\Z_{2H}

3098: \eeq

3099: which is even or odd depending on whether

3100: $\sum_i L_i H/(k_i+2)+HrS_i$ is even or odd.

3101: The choice of $M$ corresponds to the choice of representation of $\Gamma$

3102: on the Chan-Paton factor.\\

3103: The brane $\Scr{B}^B_{{\bf L},M,S}$

3104: is invariant under all element of the orbifold group.

3105: Thus, the boundary state in the orbifold theory is simply the

3106: sum over the twists.

3107: \beqa

3108: \lefteqn{|\Scr{B}^B_{{\bf L},M,S}\rangle_{(-1)^{(s'+1)F}}

3109: ={1\over \sqrt{H}}\sum_{\nu\in\Z_{H}}

3110: \bigotimes_{i=1}^r|\Scr{B}^B_{L_i,M_i,S_i}\rangle_{(-1)^{(s'+1)F}g^{\nu}}

3111: }

3112: \nn\\

3113: &=&{1\over \sqrt{H}}\sum_{\nu\in\Z_{H}\atop

3114: \nu_i\in \Z_2^r,l_i'\in {\rm P}_{k_i}}

3115: \otimes_i(2k_i+2)^{1\over 4}

3116: \e^{-\pi i {M_i\nu\over k_i+2}+\pi i {S(s'+2\nu_i)\over 2}}

3117: {S_{L_il_i'}\over{\sqrt{S_{0l_i'}}}}\kket{l_i',\nu,s'+2\nu_i}_B,

3118: \nn\\

3119: \eeqa

3120: where $s'=0$ for NSNS and $s'=1$ for RR.

3121: This B-brane can be identified as the A-brane in the

3122: mirror Gepner model associated with the product

3123: $\Scr{B}_{L_1,M_1,S_1}\times\cdots\times

3124: \Scr{B}_{L_r,M_r,S_r}$.

3125: It is a simple exercise to reproduce the above boundary

3126: states from this point of view.

3127: This realization will be useful in the discussion of the

3128: tadpole cancellation.

3129: 

3130: 

3131: 

3132: 

3133: 

3134: Next let us consider the brane involving short-orbit branes of

3135: the minimal model.

3136: There is one important constraint:

3137:  the number of minimal model factors having short-orbit branes

3138: must be even. This is to avoid the open strings to

3139: have odd number of real fermionic zero

3140: modes, which would be problematic upon quantization.

3141: Thus, we will only consider product branes with even number of

3142: $\widehat{\Scr{B}}^B_{{k_i\over 2},M_i,S_i}$ such as

3143: $$

3144: \widehat{\Scr{B}}_{{\bf L},M,S}

3145: =\widehat{\Scr{B}}^B_{{k_1\over 2},M_1,S_1}\times

3146: \widehat{\Scr{B}}^B_{{k_2\over 2},M_2,S_2}\times

3147: \Scr{B}^B_{L_3,M_3,S_3}\times\cdots\times

3148: \Scr{B}^B_{L_r,M_r,S_r}.

3149: $$

3150: The global symmetry $\g$ preserves the long-orbit brane but

3151: reverses the orientation of the short-orbit brane.

3152: However, since there are even number of factors with

3153: short-orbit branes, the brane $\widehat{\Scr{B}}_{{\bf L},M,S}$ is invariant

3154: under the orbifold group.

3155:  Thus, again the boundary state is

3156: the simple sum over the twists.

3157: Note that the symmetry

3158: $\gamma=(\g,\g,\g,...,\g)$ is the same as $(a^2,a^2,\g,...,\g)$.

3159: The boundary state is therefore

3160: \beqa

3161: |\widehat{\Scr{B}}_{{\bf L},M}\rangle_\NSNS

3162: &=&{1\over 2\sqrt{H}}\sum_{\nu\,{\rm even}}

3163: \bigotimes_{i=1}^r|\Scr{B}^B_{L_i,M_i,S_i}\rangle_{(-1)^F\g^{\nu}}

3164: \nn\\

3165: &&+{1\over 2\sqrt{H}}\sum_{\nu\,{\rm odd}}

3166: |\widetilde{\Scr{B}}^B_{{k_1\over 2},M_1,S_1}

3167: \otimes

3168: \widetilde{\Scr{B}}^B_{{k_2\over 2},M_2,S_2}\rangle^{}_{\nu}

3169: \bigotimes_{i=3}^r|\Scr{B}^B_{L_i,M_i,S_i}\rangle_{(-1)^F\g^{\nu}},

3170: \label{sslllNSNS}\\

3171: |\widehat{\Scr{B}}_{{\bf L},M,S}\rangle_\RR

3172: &=&{1\over 2\sqrt{H}}\sum_{\nu\,{\rm even}}

3173: |\widetilde{\Scr{B}}^B_{{k_1\over 2},M_1,S_1}

3174: \otimes

3175: \widetilde{\Scr{B}}^B_{{k_2\over 2},M_2,S_2}\rangle^{}_{\nu}

3176: \bigotimes_{i=3}^r|\Scr{B}^B_{L_i,M_i,S_i}\rangle_{\g^{\nu}}

3177: \nn\\

3178: &&+{1\over 2\sqrt{H}}\sum_{\nu\,{\rm odd}}

3179: \bigotimes_{i=1}^r|\Scr{B}^B_{L_i,M_i,S_i}\rangle_{\g^{\nu}},

3180: \label{sslllRR}

3181: \eeqa

3182: where

3183: $$

3184: |\widetilde{\Scr{B}}^B_{{k\over 2},M,S}\rangle_{\nu}=

3185: \e^{-\pi i {M\nu+\nu\over k+2}+\pi i {S\nu+\nu\over 2}}

3186: \sqrt{k+2}\sum_{s=\pm 1}\e^{-\pi i {S(S-s)\over 2}}

3187: \kket{{k\over 2},{k+2\over 2}+\nu,s+\nu}_B.

3188: $$

3189: Let us compare this with the brane $\Scr{B}_{{\bf L},M,S}$ where

3190: the first and the second factors are the standard ones

3191: $\Scr{B}_{{k_1\over 2},M_1,S_1}$, $\Scr{B}_{{k_2\over 2},M_2,S_2}$.

3192: We note that 

3193: \beq

3194: \begin{array}{ll}

3195: |\Scr{B}_{{k\over 2},M,S}^B\rangle_{(-1)^F\g^{\nu}}=0&

3196: \mbox{for odd $\nu$ and}

3197: \\

3198: |\Scr{B}_{{k\over 2},M,S}^B\rangle_{\g^{\nu}}=0&

3199: \mbox{for even $\nu$.}

3200: \end{array}

3201: \label{rea1}

3202: \eeq

3203: Thus, it differs from $|\widehat{\Scr{B}}_{{\bf L},M,S}\rangle$

3204: by the factor of $2$ and also by the absence of 

3205: the odd $\nu$ sum in the NSNS sector

3206: (the second line of (\ref{sslllNSNS})) and 

3207: the even $\nu$ sum in the RR sector

3208: (the first line of (\ref{sslllRR})).

3209: In other words,

3210: $$

3211: |\Scr{B}_{{\bf L},M,S}\rangle

3212: =|\widehat{\Scr{B}}_{{\bf L},M,S}\rangle

3213: +|\widehat{\Scr{B}}^{(-)}_{{\bf L},M,S}\rangle

3214: $$

3215: where 

3216: $|\widehat{\Scr{B}}^{(-)}_{{\bf L},M,S}\rangle$ is

3217: obtained from 

3218: $|\widehat{\Scr{B}}_{{\bf L},M,S}\rangle$

3219: by flipping the sign of

3220: the odd $\nu$ sum in the NSNS sector

3221: and the even $\nu$ sum in the RR sector.

3222: Thus, $\Scr{B}_{{\bf L},M,S}$ cannot be thought of as an elementary brane

3223: but is a sum of two different branes.

3224: The same can be said on

3225: $\widehat{B}^B_{{\bf L},M,S}$

3226: if two or more $L_i$ from

3227: $L_3,...,L_r$ are the same as ${k_i\over 2}$.

3228: If exactly one $L_i$ from $L_3,...,L_r$ is the same as ${k_i\over 2}$,

3229: the boundary state $|\widehat{\Scr{B}}_{{\bf L},M,S}\rangle$

3230: is simply one half of the ordinary one

3231: $|\Scr{B}_{{\bf L},M,S}\rangle$ since the odd $\nu$ sum in NSNS

3232: and even $\nu$ sum in RR are killed by that $i$-th factor because of

3233: (\ref{rea1}).

3234: 

3235: By this consideration, we find that the general elementary branes are

3236: given as follows.

3237: For each $({\bf L},M,S)$,

3238: things depend on the cardinality of the set

3239: ${\bf S}\subset \{1,2,...,r\}$ of $i$ for which $L_i={k_i\over 2}$.

3240: \underline{If ${\bf S}$ is empty},

3241: that is, if $L_i\ne {k_i\over 2}$ for all $i$,

3242: the brane $\Scr{B}^B_{{\bf L},M,S}$

3243: is elementary.

3244: \underline{If $|{\bf S}|$ is even and non-zero}, the elementary branes are

3245: \beqa

3246: |\widehat{\Scr{B}}_{{\bf L},M,S}^{(\pm) B}\rangle_{\NSNS}

3247: &=&{1\over 2^{|{\bf S}|\over 2}\sqrt{H}}\left\{

3248: \sum_{\nu\,{\rm even}}

3249: |\Scr{B}^B_{{\bf L},M,S}\rangle_{(-1)^F\gamma^{\nu}}^{\rm prod}

3250: \pm 

3251: \sum_{\nu\,{\rm odd}}

3252: |\widetilde{\Scr{B}}^B_{{\bf L},M,S}\rangle^{\rm prod}_{(-1)^F\gamma^{\nu}}

3253: \right\},

3254: \label{fixed1}

3255: \\

3256: |\widehat{\Scr{B}}_{{\bf L},M,S}^{(\pm) B}\rangle_{\RR}

3257: &=&{1\over 2^{|{\bf S}|\over 2}\sqrt{H}}

3258: \left\{

3259: \pm\sum_{\nu\,{\rm even}}

3260: |\widetilde{\Scr{B}}^B_{{\bf L},M,S}\rangle^{\rm prod}_{\gamma^{\nu}}

3261: +\sum_{\nu\,{\rm odd}}

3262: |\Scr{B}^B_{{\bf L},M,S}\rangle_{\gamma^{\nu}}^{\rm prod}

3263: \right\},

3264: \label{fixed2}

3265: \eeqa

3266: where

3267: \beqa

3268: &&

3269: |\Scr{B}^B_{{\bf L},M,S}

3270: \rangle^{\rm prod}_{(\pm 1)^F\gamma^{\nu}}

3271: =\bigotimes_{i=1}^r|\Scr{B}^B_{L_i,M_i,S_i}\rangle_{(\pm 1)^F\g^{\nu}},

3272: \nn\\

3273: &&

3274: |\widetilde{\Scr{B}}^B_{{\bf L},M,S}

3275: \rangle^{\rm prod}_{(\pm 1)^F\gamma^{\nu}}

3276: =\bigotimes_{i\not\in{\bf S}}

3277: |\Scr{B}^B_{L_i,M_i,S_i}\rangle_{(\pm 1)^F\g^{\nu}}

3278: \otimes

3279: \bigotimes_{i\in {\bf S}}

3280: |\widetilde{\Scr{B}}^B_{L_i,M_i,S_i}\rangle_{\nu}.

3281: \nn

3282: \eeqa

3283: \underline{If $|{\bf S}|$ is odd}, the elementary brane is

3284: \beq

3285: |\widehat{\Scr{B}}^B_{{\bf L},M,S}\rangle_{\NSNS\atop\RR}

3286: ={1\over 2^{|{\bf S}|-1\over 2}}

3287: |\Scr{B}^B_{{\bf L},M,S}\rangle_{\NSNS\atop\RR}.

3288: \label{fixed3}

3289: \eeq

3290: 

3291: One can see that these results reproduce the fixed point

3292: resolution prescription that is obtained by constructing the

3293: B-type boundary states as A-type in the mirror \cite{FKLLSW}.

3294: In this approach, one applies the Greene-Plesser orbifold

3295: construction of mirror symmetry to the A-type brane 

3296: $$

3297: \Scr{B}^A_{{\bf L,M},S}=\Scr{B}^A_{L_1,M_1,S_1}\times\cdots

3298: \times\Scr{B}^A_{L_r,M_r,S_r}.

3299: $$

3300: The orbifold group $G$ is the subgroup of $\prod_{i=1}^r 

3301: \Z_{k_i+2}$ in the kernel of the elementary character of the 

3302: diagonal subgroup $\Z_{{\rm lcm}\{k_i+2\}}$. It is then easy to see 

3303: that the brane $\Scr{B}^A_{{\bf L,M},S}$ is invariant under the 

3304: subgroup $H=(\Z_2)^{|{\bf S}|-1}$ generated by elements of the

3305: form $f_{ij}=g_i^{(k_i+2)/2} g_j^{(k_j+2)/2}$ for all pairs $i, j\in\bf 

3306: S\neq\emptyset$. The discrete torsion on $H$ was computed in 

3307: \cite{FKLLSW}, and shown to be maximal in the sense that the size of

3308: $K$ (see eq.\ (\ref{untwisted}). $K$ is called ``untwisted stabilizer'' 

3309: in \cite{FKLLSW}) is the minimal compatible with the constraint 

3310: that $|H|/|K|$ be the square of an integer. Explicitly, one finds

3311: $$

3312: \epsilon(f_{1i},f_{1j})= (-1)^{1+\delta_{ij}} \,.

3313: $$

3314: It is easy to see that this implies $K=\{{\rm id}\}$ if $|{\bf S}|-1$

3315: is even, while $K=\Z_2$ if $|{\bf S}|-1$ is odd. Applying the general

3316: theory of \cite{FSHSS} explained around (\ref{shortdiscrete}), this 

3317: gives the same results for the structure of elementary short orbit 

3318: B-branes that we have obtained in eqs.\ (\ref{fixed1}), 

3319: (\ref{fixed2}),(\ref{fixed3}), including the normalization factor.

3320: %

3321: 

3322: 

3323: \subsection{Boundary/Crosscap States

3324: in String Theory}\label{subsec:total}

3325: 

3326: We have constructed the internal parts of the 

3327: boundary and crosscap states.

3328: We now use them to construct the ones in full string theory

3329: relevant for compactifications to $3+1$ dimensions

3330: --- we add the spacetime part ($D=3+1$ free bosons and fermions

3331: as well as ghost and superghost), and also make sure

3332: that the states obey the chiral GSO projection condition:

3333: $$

3334: \mbox{IIA}:\,

3335: \left\{

3336: \begin{array}{l}

3337: (-1)^{F_L}=-1\\

3338: (-1)^{F_R}=-(-1)^{s}

3339: \end{array}

3340: \right.

3341: \qquad

3342: \mbox{IIB}:\,

3343: \left\{

3344: \begin{array}{l}

3345: (-1)^{F_L}=-1\\

3346: (-1)^{F_R}=-1

3347: \end{array}

3348: \right.

3349: $$

3350: where $s=0$ for NS-sector and $s=1$ on R-sector.

3351: We are interested in branes filling the $D$-dimensional spacetime

3352: and the ordinary worldsheet orientation reversal $\Omega$

3353: that acts trivially on these $D$ coordinates.

3354: Thus, boundary and crosscap states in the spacetime part are independent

3355: of IIA or IIB, and are the standard coherent state

3356: $|\Scr{B}_{\pm}^{\rm \,st}\rangle$,

3357: $|\Scr{C}_{\pm}^{\rm \,st}\rangle$.

3358: They are related to each other by

3359: $$

3360: |\Scr{B}_{-}^{\rm \,st}\rangle

3361: =(-1)^{F_R^{\rm \,st}}|\Scr{B}_+^{\rm \,st}\rangle,

3362: \qquad

3363: |\Scr{C}_{-}^{\rm \,st}\rangle

3364: =(-1)^{F_R^{\rm \,st}}|\Scr{C}_+^{\rm \,st}\rangle.

3365: $$

3366: Here $(-1)^{F_R^{\rm \,st}}$ is the spacetime part of the

3367: right-moving mod 2 fermion number, which is defined so that

3368: \beq

3369: (-1)^{F_R}=(-)^{F_R^{\rm \,st}}\e^{\pi i J_0},

3370: \eeq

3371: where $J_0$ is the $U(1)$-charge of

3372: the right-moving ${\mathcal N}=2$ superconformal algebra.

3373: Finally, we also need to make sure that the O-plane tension is real.

3374: This requires us to multiply the NSNS-part of the crosscap state

3375: by a suitable phase.

3376: 

3377: In what follows in the main part of the paper,

3378: we assume $D=3+1$, $r=5$ and

3379: $$

3380: \sum_{i=1}^r{1\over k_i+2}=1.

3381: $$

3382: Since $r=5$ is odd, the $S$ label can be represented

3383: by $S_1=S_2=\cdots=S_5=:S$.

3384: More general models are treated in Appendix.

3385: 

3386: 

3387: \subsubsection{Type IIA Orientifolds}

3388: 

3389: To find the combination obeying

3390:  the chiral GSO projection condition, we need to know the action of

3391: $\e^{\pi i J_0}$ on the boundary and crosscap states we have determined.

3392: In the individual minimal model, the action is as follows:

3393: \beqa

3394: &&\e^{\pi i J_0}|\Scr{B}_{L,M,S}\rangle_{\NSNS\atop\RR}

3395: =|\Scr{B}_{L,M-1,S-1}\rangle_{\NSNS\atop\RR},

3396: \nn\\

3397: &&\e^{\pi i J_0}|\Scr{C}_{n,s}(\pm)\rangle

3398: =\left\{\begin{array}{ll}

3399: |\Scr{C}_{n-2,s}(\mp)\rangle&\mbox{$s$ even}\\

3400: \pm \e^{-\pi i {s+1\over 2}}|\Scr{C}_{n-2,s}(\mp)\rangle&\mbox{$s$ odd}

3401: \end{array}\right.

3402: \nn

3403: \eeqa

3404: %This essentially follows from the fact that $\e^{\pi i J_0}$

3405: %corresponds to the simple current symmetry $g_{-1,-1}$

3406: %of the coset model ${SU(2)_k\times U(1)_2\over U(1)_{k+2}}$.

3407: %Also, the preserved worldsheet supersymmetry is flipped

3408: %($S\to S+1$ (mod 2) and $(\pm)\to (\mp)$ corresponds to

3409: %the flip $\overline{Q}_++Q_-\leftrightarrow

3410: %\overline{Q}_+-Q_-$), because

3411: %$\e^{\pi i J_0}$ anticommutes with the right-moving supercharges

3412: %but commutes with the left-moving supercharges.

3413: Using this, we find that the boundary and crosscap states of the Gepner model

3414: are transformed as

3415: \beqa

3416: &&\e^{\pi i J_0}|\Scr{C}_{\widetilde{P}^A_{\omega;{\bf m}}}\rangle

3417: =\omega|\Scr{C}_{(-1)^F\widetilde{P}^A_{{\omega;{\bf m}}}}\rangle,

3418: \nn\\

3419: &&\e^{\pi i J_0}|\Scr{C}_{P^A_{\omega;{\bf m}}}\rangle

3420: =-\omega|\Scr{C}_{(-1)^FP^A_{{\omega;{\bf m}}}}\rangle,

3421: \nn\\

3422: &&\e^{\pi i J_0}|\Scr{B}_{{\bf L,M},S}\rangle_{\NSNS\atop\RR}

3423: =|\Scr{B}_{{\bf L,M-1},S-1}\rangle_{\NSNS\atop\RR}.

3424: \nn

3425: \eeqa

3426: The appearance of $\omega$ is because of the shift in the

3427: summation index $\nu$, and the appearance of the minus sign

3428: in the RR-part of

3429: the crosscap state is from the prefactor

3430: $(-1)^{\sum_i{\nu\over k_i+2}}$ in the summand (\ref{CAAA})

3431: of the $\nu$-summation.

3432: 

3433: 

3434: We also need to make sure that the tension of the D-branes are real positive,

3435: and the tension of the O-planes are real. We know that

3436: ${}_{\NSNS}\langle 0|\Scr{B}_{L,M,S}\rangle$ is real positive,

3437: and thus we can use the

3438: NSNS boundary state without modification.

3439: As for the crosscap states, using the formula (\ref{minOtens})

3440: for the minimal model, we find that for $H$ odd

3441: (all $k_i$ odd)

3442: $$

3443: {}_{\NSNS}\langle 0|\Scr{C}_{(\pm 1)^F\widetilde{P}_{{\bf m}}}\rangle

3444: =\sqrt{H}\prod_{i=1}^r\mbox{$\sqrt{2\over (k_i+2)\sin({\pi\over k_i+2})}$}

3445: \cos\Bigl({\pi\over

3446: 2(k_i+2)}\Bigr),

3447: $$

3448: and for $H$ even (some $k_i$ even)

3449: $$

3450: {}_{\NSNS}\langle 0|\Scr{C}_{(\pm 1)^F\widetilde{P}_{\omega;{\bf m}}}\rangle

3451: ={\sqrt{H}\over 2}\prod_{i=1}^r

3452: \mbox{$\sqrt{2\over (k_i+2)\sin({\pi\over k_i+2})}$}

3453: \prod_{k_i\,{\rm odd}}\cos\Bigl({\pi\over

3454: 2(k_i+2)}\Bigr)\cdot\Bigl(

3455: \e^{\mp i\Theta}+\omega\e^{\pm i\Theta}\Bigr),

3456: $$

3457: $$

3458: \Theta=\sum_{k_i\,{\rm even}}{(-1)^{m_i}\pi\over 2(k_i+2)}.

3459: $$

3460: We see that it is real if $H$ is odd and also for the $\omega=1$ case

3461: if $H$ is even. However, for the $\omega=-1$ case ($H$ even),

3462: it is pure imaginary. To make it real, me must multiply

3463: the state by $i$.

3464: In general, multiplication by $\omega^{1\over 2}$ will do the job.

3465: 

3466: 

3467: Collecting all these items, we find that the total crosscap and

3468: boundary states are given by

3469: \beqa

3470: &&|C_{\omega;{\bf m}}\rangle_{\NSNS}

3471: =\omega^{1\over 2}|\Scr{C}_{\widetilde{P}^A_{\omega;{\bf m}}}\rangle\otimes 

3472: |\Scr{C}_+^{\rm \,st}\rangle_{\NSNS}

3473: -\omega^{-{1\over 2}}

3474: |\Scr{C}_{(-1)^{F}\widetilde{P}^A_{\omega;{\bf m}}}\rangle\otimes 

3475: |\Scr{C}_-^{\rm \,st}\rangle_{\NSNS}\\

3476: &&|C_{\omega;{\bf m}}\rangle_{\RR}

3477: =|\Scr{C}_{P^A_{\omega;{\bf m}}}\rangle\otimes

3478: |\Scr{C}_+^{\rm \,st}\rangle_{\RR}

3479: -\omega

3480: |\Scr{C}_{(-1)^FP^A_{\omega;{\bf m}}}\rangle\otimes

3481: |\Scr{C}_-^{\rm \,st}\rangle_{\RR},

3482: \label{CARtotal}

3483: \eeqa

3484: and

3485: \beqa

3486: &&|B_{\bf L,M}\rangle_{\NSNS}=

3487: |\Scr{B}_{{\bf L,M+1},1}\otimes\Scr{B}_+^{\rm \,st}\rangle_{\NSNS}

3488: -|\Scr{B}_{{\bf L,M},0}\otimes\Scr{B}_-^{\rm \,st}\rangle_{\NSNS},

3489: \\

3490: &&|B_{{\bf L,M}}\rangle_{\RR}

3491: =|\Scr{B}_{{\bf L,M+1},1}\otimes\Scr{B}_+^{\rm \,st}\rangle_{\RR}

3492: +|\Scr{B}_{{\bf L,M},0}\otimes\Scr{B}_-^{\rm \,st}\rangle_{\RR}.

3493: \eeqa

3494: We note that there are still a freedom to flip the sign of

3495: them except the NSNS part of the boundary state.

3496: The sign flip of

3497: the RR-parts of the boundary/crosscap states corresponds to

3498: orientation flip, and the sign flip of the

3499: NSNS part of the crosscap state corresponds to the flip in the type of

3500: the orientifold.

3501: The choice of this sign for the NSNS crosscap can be made by the choice of

3502: the phase $\omega^{1\over 2}$ (that is, $1$ or $-1$ for $\omega=1$,

3503: and $i$ or $-i$ for $\omega=-1$).

3504: 

3505: 

3506: 

3507: 

3508: 

3509: 

3510: \subsubsection{Type IIB Orientifolds}

3511: 

3512: The action of $\e^{\pi i J_0}$ on B-type boundary and crosscap states

3513: can be found either directly or by using the mirror description.

3514: Here, we present the latter way.

3515: We first note that $\e^{\pi i J_0}$ and mirror automorphism obey the

3516: following relation

3517: $$

3518: \e^{\pi i J_0}V_M=V_M\e^{-\pi i J_0}

3519: =\left\{\begin{array}{ll}

3520: V_M\e^{\pi i J_0}&\mbox{on NSNS sector}\\

3521: (-1)^rV_M\e^{\pi i J_0}&\mbox{on RR sector}.

3522: \end{array}\right.

3523: $$

3524: Using this and using the mirror realization of the crosscap and boundary

3525: state we find

3526: \beqa

3527: &&\e^{\pi i J_0}|\Scr{C}_{\widetilde{P}^B_{\omega;{\bf m}}}\rangle

3528: =\e^{2\pi i\sum_i{m_i\over

3529: k_i+2}}|\Scr{C}_{(-1)^F\widetilde{P}^B_{\omega;{\bf m}}}\rangle,

3530: \nn\\

3531: &&\e^{\pi i J_0}|\Scr{C}_{P^B_{\omega;{\bf m}}}\rangle

3532: =\e^{2\pi i\sum_i{m_i\over

3533: k_i+2}}|\Scr{C}_{(-1)^FP^B_{\omega;{\bf m}}}\rangle,

3534: \nn\\

3535: &&\e^{\pi i J_0}|\Scr{B}_{{\bf L},M,S}\rangle_{\NSNS\atop\RR}

3536: =|\Scr{B}_{{\bf L},M+\sum_i{H\over k_i+2},S+1}\rangle_{\NSNS\atop\RR}

3537: \nn

3538: \eeqa

3539: We also find, by direct computation, that the B-brane including short-orbit

3540: brane factors are transformed in the same way as

3541: $|\Scr{B}_{{\bf L},M,S}\rangle$.

3542: 

3543: The next item is the reality of the overlap of the crosscap states

3544: with the NSNS ground state. Here again, the mirror description is useful.

3545: We have just experienced what to do for the A-type crosscaps.

3546: This tells us that for the reality of the overlap with

3547: $|0\rangle_{\NSNS}$ we need to multiply

3548: $|\Scr{C}_{(\pm 1)^F\widetilde{P}^A_{\tilomega;{\bf\tilm}}}\rangle$

3549: by the phase

3550: $$

3551: \tilomega^{1\over 2}=\prod_i\tilomega_i^{1\over 2}

3552: =\exp\left(-\pi i\sum_{i=1}^r{m_i\over k_i+2}\right).

3553: $$

3554: 

3555: The total crosscap and boundary states are given by

3556: \beqa

3557: &&|C^B_{\omega;{\bf m}}\rangle_{\NSNS}

3558: =\tilomega^{1\over 2}|\Scr{C}_{\widetilde{P}^B_{\omega;{\bf m}}}\rangle\otimes 

3559: |\Scr{C}_+^{\rm \,st}\rangle_{\NSNS}

3560: -\tilomega^{-{1\over 2}}

3561: |\Scr{C}_{(-1)^{F}\widetilde{P}^B_{\omega;{\bf m}}}\rangle\otimes 

3562: |\Scr{C}_-^{\rm \,st}\rangle_{\NSNS},

3563: \label{CBNStotal}\\

3564: &&|C^B_{\omega;{\bf m}}\rangle_{\RR}

3565: =|\Scr{C}_{P^B_{\omega;{\bf m}}}\rangle\otimes

3566: |\Scr{C}_+^{\rm \,st}\rangle_{\RR}

3567: -\tilomega^{-1}

3568: |\Scr{C}_{(-1)^FP^B_{\omega;{\bf m}}}\rangle\otimes

3569: |\Scr{C}_-^{\rm \,st}\rangle_{\RR},

3570: \label{CBRtotal}

3571: \eeqa

3572: and

3573: \beqa

3574: &&|B^B_{{\bf L},M}\rangle_{\NSNS}=

3575: |\Scr{B}^B_{{\bf L},M+\sum_i{H\over k_i+2},1}

3576: \otimes\Scr{B}_+^{\rm \,st}\rangle_{\NSNS}

3577: -|\Scr{B}^B_{{\bf L},M,0}\otimes\Scr{B}_-^{\rm \,st}\rangle_{\NSNS},\\

3578: &&|B^B_{{\bf L},M}\rangle_{\RR}=

3579: |\Scr{B}^B_{{\bf L},M+\sum_i{H\over k_i+2},1}

3580: \otimes\Scr{B}_+^{\rm \,st}\rangle_{\RR}

3581: +|\Scr{B}^B_{{\bf L},M,0}\otimes\Scr{B}_-^{\rm \,st}\rangle_{\RR},

3582: \eeqa

3583: The sign of the second term of the RR boundary state is $+$ because

3584: $|\Scr{B}^B_{{\bf L},M+\sum_i{2H\over k_i+2},2}\rangle_{\RR}

3585: =(-1)^r|\Scr{B}^B_{{\bf L},M,0}\rangle_{\RR}

3586: =-|\Scr{B}^B_{{\bf L},M,0}\rangle_{\RR}$, where $r=5$ is used.

3587: The choice of the sign for the NSNS crosscap can be made by the choice of

3588: the phase $\tilomega^{1\over 2}$

3589: ($1$ or $-1$ for $\tilomega=1$,

3590: and $i$ or $-i$ for $\tilomega=-1$).

3591: 

3592: 

3593: 

3594: 

3595: 

3596: 

3597: 

3598: 

3599: \section{Consistency Conditions and Supersymmetry --- A}

3600: \label{sec:TCCA}

3601: 

3602: In this and the next sections,

3603: we determine the conditions of consistency and

3604: spacetime supersymmetry of 

3605: Type II orientifolds on Gepner model with rational D-branes.

3606: We focus on compactification down to $3+1$ dimensions.

3607: 

3608: The main part of consistency conditions is the

3609: RR tadpole cancellation \cite{PolCai,Callan}

3610: $$

3611: \langle \mbox{massless RR scalar}|T\rangle =0.

3612: $$

3613: In terms of the internal CFT, this can be written as

3614: \beq

3615: {}_\RR\langle i |\Scr{C}_P\rangle\,+\,

3616: {1\over 4}\,

3617: {}_\RR\langle i |\Scr{B}_+\rangle_\RR\,=\,0,

3618: \label{internal}

3619: \eeq

3620: for any RR ground states $|i\rangle_\RR$ of the internal theory responsible

3621: for RR scalars.

3622: The factor of $1/4$ is from the $3+1$ dimensional spacetime part.

3623: The other condition is when there are D-branes invariant under

3624: the orientifold action. If that is of $Sp$-type, the number of such branes

3625: must be even.

3626: 

3627: Spacetime supersummetry is conserved by a set of branes $\Scr{B}^a$

3628: ($a=1,...,N$) when

3629: the overlaps ${}_\NSNS\langle 0|\Scr{B}^a_+\rangle_\NSNS$ and 

3630: ${}_\RR\langle 0|\Scr{B}^a_+\rangle_\RR$ differ by a phase common to all

3631: $a$.

3632: This phase determines the conserved combination of supercharges.

3633: A spacetime supersymmetry exists in the orientifold model when

3634: the supersymmetries preserved by the D-branes and the O-planes are the

3635: same;

3636: \beq

3637: {{}_\RR\langle 0|\Scr{B}^a_+\rangle_\RR

3638: \over {}_\NSNS\langle 0|\Scr{B}^a_+\rangle_\NSNS}

3639: ={{}_\RR\langle 0|\Scr{C}_P\rangle_\RR

3640: \over {}_\NSNS\langle 0|\Scr{C}_P\rangle_\NSNS}.

3641: \label{ssusy}

3642: \eeq

3643: 

3644: 

3645: In the rest of this section,

3646: we write down these conditions for Type IIA orientifolds.

3647: We will also find a very simple class of solutions to these conditions, 

3648: and compute the particle spectrum in selected examples. Finding the most 

3649: general solution is a rather hard problem, about which we will also make 

3650: some comments towards the end of this section. In section 

3651: \ref{sec:TCCB}, we will present complete solutions of the tadpole conditions 

3652: for Type IIB orientifolds of Gepner models, which are a lot simpler. 

3653: To be sure, we do not mean to say that A-type tadpole conditions are 

3654: intrinsically harder to solve than B-type. Indeed, A and B-type are 

3655: identified under mirror symmetry. The solutions in the Gepner model we 

3656: seek in this section are interpreted in the large volume limit as A-type 

3657: on the quintic or B-type on the mirror quintic (and vice-versa in section 

3658: \ref{sec:TCCB}). There are tadpole cancellation problems in Gepner models 

3659: which are of intermediate difficulty, such as in certain orbifolds of the 

3660: quintic. We discuss one of them in the appendix.

3661: %

3662: 

3663: 

3664: 

3665: \subsection{Charge and Supersymmetry of O-planes}

3666: 

3667: Let us first review the description of RR-charge of

3668: the A-type D-branes and O-planes in a general LG model

3669: (see \cite{HIV,HKKPTVVZ,BH2} for more details).

3670: Let us consider a LG model on

3671: a non-compact K\"ahler manifold $X$ of dimension $n$

3672: with superpotential $W$.

3673: A-branes are D$n$-branes wrapped on an oriented

3674: Lagrangian submanifolds of $X$ that lie in level sets of

3675: ${\rm Im}(W)$. An A-type orientifold is associated with

3676: an antiholomorphic involution $\tau$ of $X$ that maps $W$ to

3677: its complex conjugate $\overline{W}$ up to a constant shift.

3678: The O-plane $O_{\tau \Omega}$, the fixed point set of $\tau$, 

3679: is also a Lagrangian submanifold in a level set

3680: of ${\rm Im}(W)$ and we assume that an orientation is chosen.

3681: To describe their charges, we introduce

3682: the subspaces $B^{\pm}\subset X$ which are the set of points with large

3683: values of $\pm {\rm Im}(W)$, say, $B^{\pm}=\{x\in X|

3684: \pm {\rm Im}(W(x))\geq R\}$ for a sufficiently large $R>0$.

3685: For an A-brane $\gamma$, we deform its asymptotics so that their

3686: $W$-images are deformed to $\pm{\rm Im}(W)>R$. Let us denote the resulting

3687: submanifolds by $\gamma^{\pm}$.

3688: The submanifold $\gamma^{+}$ has its boundaries in $B^+$,

3689: and defines a homology class relative to $B^+$:

3690: \beq

3691: [\gamma^+]\in H_n(X,B^+).

3692: \eeq

3693: This is the one that represents the RR-charge of the A-brane.

3694: To be precise, this is the charge

3695: at the in-coming boundary

3696: preserving the supercharge $\overline{Q}_++Q_-$.

3697: The charge at the out-going boundary preserving the same

3698: supercharge (or at the

3699: in-coming boundary preserving the opposite combination $\overline{Q}_+-Q_-$)

3700: is given by the other class $[\gamma^-]\in H_n(X,B^-)$.

3701: The Witten index for the open string

3702: stretched from $\gamma_1$ and $\gamma_2$ is given by the intersection

3703: number

3704: $\#(\gamma_1^-\cap\gamma_2^+).$

3705: The RR-charge of the O-plane at the in-coming crosscap

3706: for the parity commuting with the supercharge $\overline{Q}_++Q_-$

3707: is similarly given by

3708: \beq

3709: [O_{\tau\Omega}^+]\in H_n(X,B^+).

3710: \eeq

3711: 

3712: 

3713: 

3714: Let us apply this to the LG model with $W=X^{k+2}$

3715: that flows to the ${\mathcal N}=2$ minimal model at level $k$.

3716: The $X$-plane is separated into $2(k+2)$ regions by the inverse images of

3717: the real line of the $W$-plane, and

3718: $B^+$ and $B^-$ consist of the asymptotic regions that appears alternately,

3719: as depicted in Fig.~\ref{B}.

3720: \begin{figure}[htb]

3721: \centerline{\includegraphics{B.eps}}

3722: \caption{The regions $B^{\pm}$ for the case $k=4$.}

3723: \label{B}

3724: \end{figure}

3725: 

3726: \noindent

3727: As mentioned in section \ref{subsub:A-branes}, the A-brane $\Scr{B}_{L,M,S}$

3728: corresponds to the D1-brane at the wedge-shaped line $\gamma_{L,M,S}$

3729:  coming in from the direction

3730: $\arg(X)=\pi {M-L-1\over k+2}$, cornering at $X=0$, and going out to

3731: the direction $\arg(X)=\pi {M+L+1\over 2}$

3732: if $S=0$ or $1$ ($S=2$ or $-1$ are their orientation reversals).

3733: The branes with $S=\pm 1$ preserve the supercharge

3734: $\overline{Q}_++Q_-$ while those with $S=0,2$ preserves the opposite

3735: combination $\overline{Q}_+-Q_-$.

3736: The cycle $\gamma_{L,M,1}^+$ is obtained 

3737: by slightly rotating $\gamma_{L,M,1}$, counter-clock-wise.

3738: This correspondence $\gamma_{L,M,S}\leftrightarrow |\Scr{B}_{L,M,S}\rangle$

3739: is at the in-coming boundary.

3740: At the out-going boundary, the correspondence is slightly different:

3741: $\gamma_{L,M,1}$ (resp. $\gamma_{L,M,0}$)

3742: corresponds to $\langle \Scr{B}_{L,M-1,0}|$ (resp. 

3743: $\langle \Scr{B}_{L,M-1,1}|$). This can be understood by comparing

3744: the RR-charges as well as the conserved worldsheet

3745: supersymmetries.

3746: 

3747: 

3748: 

3749: 

3750: 

3751: 

3752: 

3753: The parity $\g^mP_A$ commutes with the worldsheet supercharge

3754: $\overline{Q}_++Q_-$ which is preserved by branes $\gamma_{L,M,S}$ with

3755: odd $S$. It acts on the LG field as

3756: $X\to \e^{2\pi i m\over k+2}\overline{X}$ and

3757: the O-plane $O_{\g^mP_A}$ is the straight line

3758: at $X\in\e^{\pi i m\over k+2}\R$. We assume the orientation

3759: that goes from $\arg(X)={\pi m\over k+2}$ to the opposite direction.

3760: Note that $m\to m+(k+2)$ is the orientation flip.

3761: The cycle $O_{\g^mP_A}^+$ is obtained by deforming it so

3762: that both of the two asymptotics are in the region $B^+$.

3763: This involves bending when $k$ is odd while it is a small 

3764: rotation when $k$ is even.

3765: To see this, let us first consider the basic A-parity

3766: $P_A$ whose O-plane $O_{P_A}$ is the real line that goes from

3767: $+\infty$ to $-\infty$.

3768: If $k$ is even, $O_{P_A}^+$ is the slight counter-clockwise rotation of

3769: $\R$. In fact there is an A-brane that does the same ---

3770: $\gamma_{{k\over 2},{k+2\over 2},1}$.

3771: Thus, the O-plane $O_{P_A}$ and $\gamma_{{k\over 2},{k+2\over 2},1}$ has

3772: the same location and the same charge.

3773: If $k$ is odd, $O_{P_A}^+$ is obtained by small counter-clockwise rotation

3774: of the real-positive half and small clockwise rotation of

3775: the real-negative half.

3776: There is no A-brane at the same location, but the brane

3777: $\gamma_{{k-1\over 2},{k+1\over 2},1}$ has the same in-coming charge.

3778: (Another brane $\gamma_{{k-1\over 2},{k+1\over 2}+1,0}$

3779: may appear to have the same charge, but it preserves a different

3780: combination $\overline{Q}_+-Q_-$ of the supercharge

3781: --- $P_A$ preserves the combination $\overline{Q}_++Q_-$

3782: and thus must be compared to the branes

3783: with odd $S$.)

3784: Figure~\ref{k=3} depicts the example of $k=3$.

3785: \begin{figure}[htb]

3786: \centerline{\includegraphics{k=3.eps}}

3787: \caption{The O-plane $O=O_{P_A}$ and the brane

3788: $B=\gamma_{1,2,1}$ in the $k=3$ minimal model.

3789: They have the same in-coming RR-charge}

3790: \label{k=3}

3791: \end{figure}

3792: Repeating this consideration in the general case, we find that the

3793: O-plane $O_{\g^mP_A}$ has the same RR-charge as one of the A-branes.

3794: The result is

3795: \beqa

3796: \mbox{$k$ even}

3797: &&[O_{\g^mP_A}^+]=

3798: \left\{\begin{array}{ll}

3799: [\gamma_{{k\over 2},{k+2\over 2}+m,1}^+]&\mbox{$m$ even}\\

3800: {[}\gamma_{{k\over 2},{k+2\over 2}+m-1,1}^+]&\mbox{$m$ odd},

3801: \end{array}\right.

3802: \label{Ochevmin}

3803: \\

3804: \mbox{$k$ odd}

3805: &&[O_{\g^mP_A}^+]=

3806: \left\{\begin{array}{ll}

3807: [\gamma_{{k-1\over 2},{k+1\over 2}+m,1}^+]&\mbox{$m$ even}\\

3808: {[}\gamma_{{k+1\over 2},{k+1\over 2}+m,1}^+]&\mbox{$m$ odd}.

3809: \end{array}\right.

3810: \label{Ochodmin}

3811: \eeqa

3812: This can also be checked by showing

3813: ${}_{\RR}\langle i|\Scr{C}_{\g^mP_A}\rangle={}_{\RR}\langle

3814: i|\Scr{B}_{L,M,S}\rangle_{\RR}$ for any RR-ground state

3815: $|i\rangle_{\RR}$

3816: with $(L,M,S)$ as indicated in

3817: (\ref{Ochevmin})-(\ref{Ochodmin}).

3818: Note that $m\to m+(k+2)$ indeed corresponds to orientation flip

3819: since RR-part of the corresponding boundary states flips its sign.

3820: 

3821: 

3822: Having learned the RR-charge of the O-plane in the minimal model,

3823: we can now compute the charge in the Gepner model.

3824: For this purpose, the expressions (\ref{Cexodd}) and (\ref{Cexeve})

3825: of the crosscap states are useful. These expressions simply says that

3826: the O-plane charge in the Gepner model is given by

3827: the same type of average formula for the A-brane charge.

3828: 

3829: 

3830: If $H$ is odd, the average formula (\ref{Cexodd})

3831: is identical to the one for an A-brane.

3832: Note that we only have to consider the basic parity

3833: $P^A=P^A_{1;{\bf 0}}$ since there is no involutive dressing by quantum

3834: symmetry and dressing by global symmetry ${\bf m}$ is equivalent to

3835: no-dressing.

3836: By the relation (\ref{Ochodmin}) for each minimal model

3837: we find that the O-plane charge is the same as the charge of the

3838: D-brane associated with the product

3839: $\prod_{i=1}^r\Scr{B}_{{k_i-1\over 2},{k_i+1\over 2},1}$.

3840: Namely,

3841: \beq

3842: \Bigl[O_{P^A}\Bigr]=4\Bigl[B_{\bf {k-1\over 2},{k-1\over 2}}\Bigr],

3843: \label{Ochod}

3844: \eeq

3845: where the factor of $4$ comes from the spacetime part.

3846: 

3847: If $H$ is even, the sum splits into two parts (\ref{Cexeve})

3848: and each part is the same as the untwisted part of the sum for an 

3849: A-brane with $\Z_2$ stabilizer group.

3850: The charge for $|\Scr{C}_{{\bf P}^A_{\bf m}}\rangle$

3851: is the same as

3852: the charge for the product brane

3853: $\prod_i\Scr{B}_{{k_i\over 2},{k_i+2\over 2}+m_i-\delta_{m_i},1}$

3854: where 

3855: $$

3856: \delta_{m_i}=\left\{\begin{array}{ll}

3857: 0&\mbox{$m_i$ even}\\

3858: 1&\mbox{$m_i$ odd}.

3859: \end{array}\right.

3860: $$

3861: The charge for $|\Scr{C}_{\gamma{\bf P}^A_{\bf m}}\rangle$ is the same as

3862: the charge for

3863: $-\prod_i\Scr{B}_{{k_i\over 2},{k_i+2\over 2}+m_i+\delta_{m_i},1}$,

3864: where the minus sign is from the

3865: factor $(-1)^{\sum_i{\nu\over k_i+2}}$ in (\ref{CAAA}).

3866: Thus, the charge is

3867: \beq

3868: \Bigl[O_{P^A_{\pm;{\bf m}}}\Bigr]=

3869: 2\Bigl[B_{\bf {k\over 2},{k\over 2}+m-\delta_m}\Bigr]

3870: \mp

3871: 2\Bigl[B_{\bf {k\over 2},{k\over 2}+m+\delta_m}\Bigr],

3872: \label{Ochev}

3873: \eeq

3874: where $[B_{\bf {k\over 2},M}]$ is the sum

3875: of the two short-orbit brane charges

3876: $[\widehat{B}_{\bf {k\over 2},M}^{(+)}]

3877: +[\widehat{B}_{\bf {k\over 2},M}^{(-)}]$

3878: which has no twisted state component.

3879: 

3880: 

3881: Let us discuss the spacetime supersymmetry preserved by

3882: D-branes and the O-plane. This is to compute the ratio of the 

3883: overlap of the boundary/crosscap state with RR-ground state $|0\rangle_{\RR}$

3884: and the brane/plane tension.

3885: Here $|0\rangle_{\RR}$ is the RR ground state of the internal

3886: system with the lowest R-charge.

3887: Let us first present the RR-overlap for the A-brane in the minimal model.

3888: This has been computed in many ways.

3889: In the LG description, it is realized as the integral over

3890: $\gamma_{L,M,1}^+$ of the 1-form $\overline{c_0}\e^{-i\overline{X}^{k+2}}

3891: *\dd \overline{X}$ where $c_0$ is a certain normalization factor

3892: \cite{HIV,HKKPTVVZ,BH2}. The result is

3893: \beqa

3894: {}_{\RR}\langle0|\Scr{B}_{L,M,1}\rangle_{\RR}

3895: &=&

3896: i\sqrt{2\over (k+2)\sin({\pi\over k+2})}\e^{-\pi i {M\over k+2}}

3897: \sin\Bigl({\pi (L+1)\over k+2}\Bigr)

3898: \nn\\

3899: &=&i\e^{-\pi i {M\over k+2}}{}_{\NSNS}\langle 0|

3900: \Scr{B}_{L,M,1}\rangle_{\NSNS}.

3901: \nn

3902: \eeqa

3903: Using this, we find that the overlap in the Gepner model is

3904: $$

3905: {}_{\RR}\langle0|\Scr{B}_{{\bf L,M},1}\rangle_{\RR}

3906: =i^r\e^{-\pi i \sum_i{M_i\over k_i+2}}

3907: {}_{\NSNS}\langle 0|\Scr{B}_{{\bf L,M},1}\rangle_{\NSNS}.

3908: $$

3909: The phase determining the spacetime supersymmetry is the ratio

3910: \beq

3911: \exp\left(i\theta_{\bf L,M}\right)

3912: =-i

3913: \exp\left(-\pi i\sum_{i=1}^r{M_i\over

3914: k_i+2}

3915: \right).

3916: \label{SUSYAbrane}

3917: \eeq

3918: We find that the phase is determined by the sum over the angles

3919: ${M_i\over k_i+2}$ of the ``mean-direction'' of the wedge in the LG

3920: realization. See Figure~\ref{mean}. The result is applicable also to short

3921: orbit branes.

3922: \begin{figure}[htb]

3923: \centerline{\includegraphics{mean.eps}}

3924: \caption{The mean direction of a brane.}

3925: \label{mean}

3926: \end{figure}

3927: 

3928: 

3929: Let us next compute the RR-overlap for the crosscap states.

3930: In the minimal model,

3931: this is essentially computed in \cite{BH2}, in both using PSS

3932: crosscap and also using LG model.

3933: Here one could also use the relation of the O-plane charge and D-brane

3934: charge given in (\ref{Ochevmin}) and (\ref{Ochodmin}) and the above

3935: expression for the brane overlaps.

3936: In any way, we find

3937: \beqa

3938: {}_{\RR}\langle 0|\Scr{C}_{\g^mP_A}\rangle

3939: &=&\left\{\begin{array}{ll}

3940: i\sqrt{2\over (k+2)\sin({\pi\over k+2})}

3941: \cos\Bigl({\pi\over 2(k+2)}\Bigr)\e^{-\pi i {{k+1\over 2}+m\over k+2}}

3942: &\mbox{$k$ odd},

3943: \\[0.2cm]

3944: i\sqrt{2\over (k+2)\sin({\pi\over k+2})}

3945: \e^{-\pi i {{k+2\over 2}+m-\delta_m\over k+2}}

3946: &\mbox{$k$ even}

3947: \end{array}

3948: \right.

3949: \nn\\

3950: &=&i\e^{-\pi i {m+{k+1\over 2}\over k+2}}\times {}_{\NSNS}\langle 0|

3951: \Scr{C}_{\g^m\widetilde{P}_A}\rangle.

3952: \nn

3953: \eeqa

3954: It follows from this that in the Gepner model

3955: $$

3956: {}_{\RR}\langle 0|\Scr{C}_{P_{\omega;{\bf m}}^A}\rangle

3957: =i^r\e^{-\pi i\sum_i{m_i+{k+1\over 2}\over k_i+2}}\times

3958: {}_{\NSNS}\langle 0|\Scr{C}_{\widetilde{P}_{\omega;{\bf m}}^A}\rangle

3959: $$

3960: Since the NSNS crosscap in string theory is obtained by multiplying

3961: $\omega^{1\over 2}$ to $|\Scr{C}_{\widetilde{P}_{\omega;{\bf

3962: m}}^A}\rangle$,

3963: we find that the ratio is

3964: \beq

3965: \exp\left(i\theta_{P^A_{\omega;{\bf m}}}\right)

3966: =-i

3967: \omega^{-{1\over 2}}\exp\left(-\pi i\sum_{i=1}^r{m_i+{k_i-1\over 2}\over

3968: k_i+2}

3969: \right).

3970: \label{SUSYAOplane}

3971: \eeq

3972: The phase is essentially the sum over the slopes 

3973: ${m_i+{k_i-1\over 2}\over k_i+2}$ of the direction perpendicular

3974: to the O-planes if $\omega=1$, but it differs from that sum by

3975: right angle if $\omega=-1$.

3976: 

3977: In Table \ref{tableAOch}, we describe the

3978: RR-charge, the tension, and the phase determining the conserved

3979: supersymmetry of the twelve A-type orientifolds of the

3980: two parameter model $(k_i+2)=(8,8,4,4,4)$.

3981: \begin{table}[htb]

3982: \begin{center}

3983: \renewcommand{\arraystretch}{1.2}

3984: \begin{tabular}{|l||l|l|l|}

3985: \hline

3986: parity&RR-charge&Tension&SUSY\\

3987: \noalign{\hrule height 0.8pt}

3988: \multicolumn{1}{|l||}{$P^B_{+;00000}$}

3989: &

3990: \multicolumn{1}{l|}{$0$}

3991: &

3992: \multicolumn{1}{l|}{$0$}&\lw{$i\omega^{-{1\over 2}}$}\\

3993: \cline{1-3}

3994: $P^B_{-;00000}$

3995: &

3996: $4[B_{{\bf k\over 2},{\bf k\over 2}}]$

3997: &

3998: $-i\omega^{1\over 2}2\sqrt{2\sqrt{2}+2}$

3999: &

4000: \\

4001: \hline

4002: \multicolumn{1}{|l||}{$P^B_{+;00001}$}

4003: &

4004: \multicolumn{1}{l|}{$2[B_{{\bf k\over 2},{\bf k\over 2}}

4005: -B_{{\bf k\over 2},(33113)}]$}

4006: &

4007: \multicolumn{1}{l|}{$\omega^{1\over 2}2\sqrt{\sqrt{2}+1}$}

4008: &\lw{$i\omega^{-{1\over 2}}\e^{-{\pi i\over 4}}$}\\

4009: \cline{1-3}

4010: $P^B_{-;00001}$

4011: &$2[B_{{\bf k\over 2},{\bf k\over 2}}

4012: +B_{{\bf k\over 2},(33113)}]$

4013: &$-i\omega^{1\over 2}2\sqrt{\sqrt{2}+1}$

4014: &

4015: \\

4016: \hline

4017: \multicolumn{1}{|l||}{$P^B_{+;01000}$}

4018: &

4019: \multicolumn{1}{l|}{$2[B_{{\bf k\over 2},{\bf k\over 2}}

4020: -B_{{\bf k\over 2},(35111)}]$}

4021: &

4022: \multicolumn{1}{l|}{$2\omega^{1\over 2}$}

4023: &\lw{$i\omega^{-{1\over 2}}\e^{-{\pi i\over 8}}$}\\

4024: \cline{1-3}

4025: $P^B_{-;01000}$

4026: &$2[B_{{\bf k\over 2},{\bf k\over 2}}

4027: +B_{{\bf k\over 2},(35111)}]$

4028: &$-2i\omega^{1\over 2}(\sqrt{2}+1)$

4029: &

4030: \\

4031: \hline

4032: \multicolumn{1}{|l||}{$P^B_{+;00011}$}

4033: &

4034: \multicolumn{1}{l|}{$2[B_{{\bf k\over 2},{\bf k\over 2}}

4035: -B_{{\bf k\over 2},(33133)}]$}

4036: &

4037: \multicolumn{1}{l|}{$\omega^{1\over 2}2\sqrt{2}\sqrt{\sqrt{2}+1}$}

4038: &\lw{$\omega^{-{1\over 2}}$}\\

4039: \cline{1-3}

4040: $P^B_{-;00011}$

4041: &$2[B_{{\bf k\over 2},{\bf k\over 2}}

4042: +B_{{\bf k\over 2},(33133)}]$

4043: &$0$

4044: &

4045: \\

4046: \hline

4047: \multicolumn{1}{|l||}{$P^B_{+;01001}$}

4048: &

4049: \multicolumn{1}{l|}{$2[B_{{\bf k\over 2},{\bf k\over 2}}

4050: -B_{{\bf k\over 2},(35113)}]$}

4051: &

4052: \multicolumn{1}{l|}{$\omega^{1\over 2}2(\sqrt{2}+1)$}

4053: &\lw{$i\omega^{-{1\over 2}}\e^{-{3\pi i\over 8}}$}\\

4054: \cline{1-3}

4055: $P^B_{-;01001}$

4056: &$2[B_{{\bf k\over 2},{\bf k\over 2}}

4057: +B_{{\bf k\over 2},(35113)}]$

4058: &$-2i\omega^{1\over 2}$

4059: &

4060: \\

4061: \hline

4062: \multicolumn{1}{|l||}{$P^B_{+;11000}$}

4063: &

4064: \multicolumn{1}{l|}{$2[B_{{\bf k\over 2},{\bf k\over 2}}

4065: -B_{{\bf k\over 2},(55111)}]$}

4066: &

4067: \multicolumn{1}{l|}{$\omega^{1\over 2}2\sqrt{\sqrt{2}+1}$}

4068: &\lw{$i\omega^{-{1\over 2}}\e^{-{\pi i\over 4}}$}\\

4069: \cline{1-3}

4070: $P^B_{-;11000}$

4071: &$2[B_{{\bf k\over 2},{\bf k\over 2}}

4072: +B_{{\bf k\over 2},(55111)}]$

4073: &$-i\omega^{1\over 2}2\sqrt{\sqrt{2}+1}$

4074: &

4075: \\

4076: \hline

4077: \end{tabular}

4078: \caption{Charge and Tension of O-planes in the two parameter model

4079: ($\omega=1$)}

4080: \label{tableAOch}

4081: \end{center}

4082: \end{table}

4083: 

4084: 

4085: 

4086: 

4087: 

4088: 

4089: \subsection{Parity Action on D-branes}

4090: 

4091: The next task is to find out how the parities act on the D-branes.

4092: 

4093: 

4094: Let us first consider the action in the minimal model, which is studied

4095: \cite{BH2}.

4096: The action is encoded in the formulae

4097: \beqa

4098: &&\langle \Scr{C}_{\g^mP_A}|q_t^H|\Scr{B}_{L,M,S}\rangle_{\RR}

4099: ={}_{\RR}\langle\Scr{B}_{L,2m-M-1,-S-1}|q_t^H|

4100: \Scr{C}_{(-1)^F\g^mP_A}\rangle,

4101: \label{PonDmin1}

4102: \\

4103: &&\langle \Scr{C}_{\g^m\widetilde{P}_A}|q_t^H|\Scr{B}_{L,M}\rangle_{\NSNS}

4104: ={}_{\NSNS}\langle\Scr{B}_{L,2m-M}|q_t^H|

4105: \Scr{C}_{(-1)^F\g^m\widetilde{P}_A}\rangle.

4106: \label{PonDmin2}

4107: \eeqa

4108: They can be shown using the properties of

4109: the P-matrix $\e^{2\pi i Q_g(j)}P_{gj}^*=P_{gj}$.

4110: They can also be geometrically understood in the LG model as follows.

4111: Under the basic parity $P_A$ that acts on the LG field as

4112: $X\to\overline{X}$, the wedge $\gamma_{L,M,1}$ is mapped to its

4113: complex conjugate. The initial and final angles 

4114: $({\pi(M-L-1)\over k+2},{\pi (M+L+1)\over k+2})$ of the wedge

4115: are mapped to $({\pi(-M+L+1)\over k+2},{\pi (-M-L-1)\over k+2})$

4116: which are the initial and final angles of the wedge $\gamma_{L,-M,1}$ with

4117: the opposite orientation.

4118: Thus the brane $\gamma_{L,M,1}$ is mapped under $P_A$ to

4119: $-\gamma_{L,-M,1}=\gamma_{L,-M,-1}$.

4120: More general parity maps the brane as

4121: $$

4122: \g^mP_A:\gamma_{L,M,1}\to\gamma_{L,2m-M,-1}.

4123: $$

4124: Note that the parity exchanges in-coming and out-going

4125: boundaries. Thus, if $\gamma_{L,M,1}$ is at the in-coming boundary and

4126: corresponds to $|\Scr{B}_{L,M,1}\rangle$, then

4127: $\gamma_{L,2m-M,-1}$ is at the out-going boundary

4128: and corresponds to $\langle \Scr{B}_{L,2m-M-1,-2}|$.

4129: Thus we find that the parity acts as

4130: $$

4131: \g^mP_A:|\Scr{B}_{L,M,1}\rangle\to\langle\Scr{B}_{L,2m-M-1,-2}|.

4132: $$

4133: This is nothing but the $S=1$ case of (\ref{PonDmin1}).

4134: The one with other values of $S$ and the other

4135: relation (\ref{PonDmin2}) can also be understood in a similar way.

4136: 

4137: 

4138: 

4139: Let us now discuss the orientifold

4140: action on D-branes in the full string theory.

4141: The relations (\ref{PonDmin1}) and (\ref{PonDmin2})

4142: readily imply

4143: \beqa

4144: &&\langle \Scr{C}_{P^A_{\omega;{\bf m}}}|q_t^H|

4145: \Scr{B}_{{\bf L,M},S}\rangle_{\RR}

4146: ={}_{\RR}\langle\Scr{B}_{{\bf L,2m-M-1},-S-1}|q_t^H|

4147: \Scr{C}_{(-1)^FP^A_{\omega;{\bf m}}}\rangle,

4148: \label{PonDG1}

4149: \\

4150: &&\langle \Scr{C}_{\widetilde{P}^A_{\omega;{\bf m}}}|q_t^H

4151: |\Scr{B}_{\bf L,M}\rangle_{\NSNS}

4152: ={}_{\NSNS}\langle\Scr{B}_{\bf L,2m-M}|q_t^H|

4153: \Scr{C}_{(-1)^F\widetilde{P}^A_{\omega;{\bf m}}}\rangle.

4154: \label{PonDG2}

4155: \eeqa

4156: Applying these to the total crosscap states,

4157: we find

4158: \beqa

4159: &&{}_{\RR}\langle C_{\omega;{\bf m}}|q_t^H|B_{\bf L,M}\rangle_{\RR}

4160: =

4161: \omega\times {}_{\RR}\langle

4162: B_{\bf L,2m-M}|q_t^H|C_{\omega;{\bf m}}\rangle_{\RR},\\

4163: &&{}_{\NSNS}\langle C_{\omega;{\bf m}}|q_t^H|B_{\bf L,M}\rangle_{\NSNS}

4164: =-{}_{\NSNS}\langle

4165: B_{\bf L,2m-M}|q_t^H|C_{\omega;{\bf m}}\rangle_{\NSNS}.

4166: \label{secon}

4167: \eeqa

4168: Recall that the overlaps appears in the one-loop diagram

4169: as the combination

4170: $$

4171: i{}_{\NSNS}\langle B|q_t^H|C\rangle_{\NSNS}

4172: -i{}_{\NSNS}\langle C|q_t^H|B\rangle_{\NSNS}

4173: -{}_{\RR}\langle B|q_t^H|C\rangle_{\RR}

4174: -{}_{\RR}\langle C|q_t^H|B\rangle_{\RR}.

4175: $$

4176: Thus the equation (\ref{secon}) shows

4177: that the brane $B_{\bf L,M}$ is mapped to $B_{\bf L,2m-M}$ if the 

4178: brane orientations are ignored. The first equation includes the

4179: information on the orientations.

4180: It shows that the branes are mapped under

4181: the orientifold action as

4182: \beq

4183: P^A_{\omega;{\bf m}}:\,

4184: B_{\bf L,M}\longmapsto 

4185: \omega B_{\bf L,2m-M},

4186: \label{PAonD}

4187: \eeq

4188: where $-B$ stands for the orientation reversal of $B$.

4189: We see that dressing by quantum symmetry affects the action on orientation.

4190: 

4191: 

4192: Let us see how the short-orbit branes are transformed.

4193: We recall that if $L_i={k_i\over 2}$ for each

4194: $i$ with odd $w_i$ (possible only when $H$ is even),

4195: the brane $B_{\bf L,M}$ must be regarded as the sum of

4196: two short-orbit branes $\widehat{B}_{\bf L,M}^{(+)}$

4197: and $\widehat{B}_{\bf L,M}^{(-)}$.

4198: The boundary states are given in (\ref{shANSNS}) and (\ref{shARR}).

4199: The overlap $\langle \widehat{B}^{(\pm)}|C\rangle$

4200: is simply one-half of $\langle B|C\rangle$ for both $\pm$, since

4201: the crosscap state $|C\rangle$ does not have twisted state components.

4202: Thus we see that the 

4203: $({\bf L,M})$-label is transformed in the same way as

4204: the long-orbit branes.

4205: To see how the $\pm$ label is transformed, we need to compare with

4206: the $\langle\widehat{B}^{(\varepsilon)}|\widehat{B}^{(\varepsilon')}\rangle$

4207: overlaps.

4208: By the loop channel expansion of the latter overlaps,

4209: one can read the spectrum of open string states between two short-orbit

4210: branes: the states labeled by

4211: $\otimes_{i=1}^r(l_i,m_i,s_i)$ are subject to the projection

4212: \beq

4213:   \frac12\left(\,1+\varepsilon\varepsilon' 

4214: \prod_{w_i~{\rm odd}}(-1)^{\frac{1}{2}(l_i+m_i-s)}\,\right),

4215: \label{AshProj}

4216: \eeq

4217: where $s=0$ for NS states and $1$ for R ones.

4218: %It follows that the short-orbit branes also support $U(1)$ gauge

4219: %symmetry as the massless vector states between the branes with the

4220: %same sign are not projected out by (\ref{AshProj}).

4221: Let us compare this with the loop-channel expansion of

4222: the overlaps $\langle\widehat{\Scr{B}}|q_t^H|\Scr{C}\rangle$. 

4223: It turns out that the open string states are subject to the projection

4224: \beq

4225:   \frac12\left(\,1+\omega^{\frac{H}{2}}(-1)^{\frac{\sigma}{2}}

4226:   \prod_{w_i~{\rm odd}}(-1)^{\frac{1}{2}(l_i+m_i-s)}\,\right),

4227: \label{AshPA}

4228: \eeq

4229: where $\sigma$ is the number of $i$'s such that $w_i$ is odd.

4230: The parity action on short-orbit branes is therefore summarized as

4231: \beq

4232: P^A_{\omega;{\bf m}}~:~

4233: \widehat{B}_{\bf L,M}^{(\varepsilon)}\longmapsto

4234: \omega

4235: \widehat{B}_{\bf L,2m-M}^{(\varepsilon')},~~~

4236: \varepsilon'=\omega^{\frac{H}{2}}(-1)^{\frac{\sigma}{2}}\varepsilon.

4237: \label{PAonDshort}

4238: \eeq

4239: For the computations that leads to (\ref{AshProj})

4240: and (\ref{AshPA}), see Appendix~\ref{app:detail1}.

4241: 

4242: 

4243: \subsubsection{Invariant Branes}

4244: 

4245: Let us see which of the branes are invariant under the parity symmetries.

4246: By (\ref{PAonD}), the long-orbit brane $B_{\bf L,M}$ is invariant under

4247: $P^A_{\omega;\bf m}$ when $\omega B_{\bf L,2m-M}=B_{\bf L,M}$.

4248: This requires that, for each $i$, $\Scr{B}_{L_i,2m_i-M_i,0}$ is equal to

4249: $\Scr{B}_{L_i,M_i,0}$ up to orientation (and up to the uniform

4250: shift in $M_i$'s).

4251: One possibility is $L_i$ arbitrary and

4252: $2m_i-M_i=M_i$ (mod $2k_i+4$) which is the case with the positive

4253: orientation, and another is $L_i=k_i/2$ and

4254: $2m_i-M_i=M_i+k_i+2$ (mod $2k_i+4$) which is the case with the reversed

4255: the orientation.

4256: For $\omega=1$, we need the total orientation to be positive, and thus

4257: the case ``$L_i=k_i/2$ and

4258: $2m_i-M_i=M_i+k_i+2$'' must occur for even number of $i$'s:

4259: \beq

4260: P^A_{+;{\bf m}}\mbox{-fixed}:\left\{\begin{array}{ll}

4261: L_i={k_i\over 2},\,\,\,

4262: M_i=m_i+{k_i+2\over 2}\,\,(\mbox{mod $k_i+2$}),&

4263: \mbox{for even \# of $i$'s}\\

4264: L_i\,\,\mbox{arbitrary,}\,\,\,

4265: M_i=m_i\,\,\,(\mbox{mod $k_i+2$}),&

4266: \mbox{for other $i$}

4267: \end{array}\right.

4268: \eeq

4269: For $\omega=-1$ (which is possible only when some $k_i$ are even),

4270: we need the total orientation to be negative,

4271: and thus

4272: the case ``$L_i=k_i/2$ and

4273: $2m_i-M_i=M_i+k_i+2$'' must occur for odd number of $i$'s:

4274: \beq

4275: P^A_{-;{\bf m}}\mbox{-fixed}:\left\{\begin{array}{ll}

4276: L_i={k_i\over 2},\,\,\,

4277: M_i=m_i+{k_i+2\over 2}\,\,(\mbox{mod $k_i+2$}),&

4278: \mbox{for odd \# of $i$'s}\\

4279: L_i\,\,\mbox{arbitrary,}\,\,\,

4280: M_i=m_i\,\,\,(\mbox{mod $k_i+2$}),&

4281: \mbox{for other $i$}

4282: \end{array}\right.

4283: \eeq

4284: 

4285: Let us next consider the short orbit branes

4286: $\widehat{B}_{\bf L,M}^{(\varepsilon)}$.

4287: For this case, the ``Brane Identification'' involves the change in

4288: $\varepsilon$-label:

4289: $M_i\to M_i+k_i+2$ for $i$ with odd $w_i$

4290: does the flip of $\varepsilon$ in addition to the flip of orientation.

4291: Also, the parity acts on the $\varepsilon$ label as

4292: $$

4293: \varepsilon\to\omega^{\frac{H}{2}}(-1)^{\frac{\sigma}{2}}\varepsilon.

4294: $$

4295: Thus, the invariant branes are those with

4296: $$

4297: \left\{\begin{array}{ll}

4298: L_i={k_i\over 2},\,\,\,

4299: M_i=m_i+{k_i+2\over 2}\,\,(\mbox{mod $k_i+2$}),&

4300: i\in I\\

4301: L_i\,\,\mbox{arbitrary,}\,\,\,

4302: M_i=m_i\,\,\,(\mbox{mod $k_i+2$}),&

4303: i\not\in I

4304: \end{array}\right.

4305: $$

4306: where $I$ is a subset of $\{1,...,r=5\}$ obeying 

4307: some condition that depends on the parity and case as described in the table.

4308: \begin{table}[htb]

4309: \begin{center}

4310: \renewcommand{\arraystretch}{1.2}

4311: \begin{tabular}{|c|c|c|c|c|}

4312: \hline

4313: &$P^A_{+;{\bf m}}$, ${\sigma\over 2}$ even&

4314: $P^A_{+;{\bf m}}$, ${\sigma\over 2}$ odd&

4315: $P^A_{-;{\bf m}}$, ${\sigma+H\over 2}$ even&

4316: $P^A_{-;{\bf m}}$, ${\sigma+H\over 2}$ odd

4317: \\

4318: \noalign{\hrule height 0.8pt}

4319: $\# I$&even&even&odd&odd

4320: \\

4321: \hline

4322: $\# (I\cap\Sigma)$&even&odd&even&odd

4323: \\

4324: \hline

4325: \end{tabular}

4326: \end{center}

4327: \end{table}

4328: Here $\Sigma$ is the set of $i$'s with odd $w_i$ ($\sigma=\#\Sigma$).

4329: 

4330: 

4331: 

4332: \subsection{Structure of Chan-Paton Factor}

4333: 

4334: Let us now determine the structure of Chan-Paton factor on the

4335: D-branes. If not with orientifolds, $N$ D-branes on top of

4336: each other, i.e., $N$ copies of a D-brane, support $U(N)$ gauge group.

4337: In the orientifold model, this is modified.

4338: If the D-brane $B_a$ is not invariant under the parity,

4339: $P:B_a\to B_a'\ne B_a$,

4340: the gauge group is still $U(N)$ since $B_a$-$B_a$ string is simply related to

4341: $B_a'$-$B_a'$ string under the orientifold projection.

4342: However, for an invariant D-brane, a non-trivial projection is imposed on

4343: the open string ending on it, and the gauge group is usually either

4344: $O(N)$ or $USp(N)=Sp(N/2)$.

4345: In the latter case $N$ must be even, which

4346: is one of the consistency requirement.

4347: 

4348: Thus, we would like to find the orientifold projections on

4349: invariant D-branes.

4350: Let $\psi^{\mu}_{-{1\over 2}}|IJ\rangle$ ($1\leq I,J\leq N$) be

4351: the open string states corresponding to the massless gauge bosons.

4352: The parity action is

4353: $$

4354: \psi^{\mu}_{-{1\over 2}}|IJ\rangle\mapsto

4355: -\sum_{I'J'}\gamma_{II'}\psi^{\mu}_{-{1\over 2}}|J'I'\rangle

4356: \gamma^{-1}_{J'J}

4357: $$

4358: where $\gamma^T=\pm \gamma$ is required for the parity to be involutive.

4359: The gauge group is $O(N)$ if $\gamma^T=\gamma$ (solved by $\gamma={\bf 1}_N$)

4360: and $Sp({N\over 2})$ if

4361: $\gamma=-\gamma^T$ (solved by $\gamma={{\bf 0}\,-{\bf 1}\choose

4362: {\bf 1}\,\,\,\,{\bf 0}}$).

4363: One consequence is that

4364: $$

4365: \Tr\Bigl|_{\rm gauge\,\,boson}Pq^H

4366: =-2{\rm tr}(\gamma^T\gamma^{-1})q^{1\over 2}=-2N\sigma q^{1\over 2},

4367: $$

4368: where

4369: \beq

4370: \sigma=\left\{\begin{array}{ll}

4371: 1&\mbox{if $O(N)$}\\

4372: -1&\mbox{if $Sp({N\over 2})$}.

4373: \end{array}

4374: \right.

4375: \eeq

4376: Thus, to find out the type of the Chan-Paton factor, we want to look at

4377: the sign in front of the gauge boson part $q^0$

4378: in the twisted partition function

4379: $\Tr Pq^H=i\langle B|q_t^H| C\rangle$.

4380: 

4381: 

4382: 

4383: It is straightforward to compute the $\langle B|C\rangle$

4384: overlaps of the minimal model and their loop-channel expansions.

4385: In particular, the ground state contribution is

4386: \beq

4387: {}_{\NSNS}\langle \Scr{B}_{L,M,S}|q_t^H|\Scr{C}_{2m,0}(\mp)\rangle

4388: =\left\{

4389: \begin{array}{l}

4390: \e^{\mp{\pi i\over 4}}\delta_{M,m}\delta_{S,0}\\

4391: \e^{\pm{\pi i\over 4}}\delta_{M,m}\delta_{S,1}\\

4392: \e^{\pm{\pi i\over 4}}\delta_{L,{k\over 2}}

4393: \delta_{M,m+{k+2\over 2}}\delta_{S,0}\\

4394: \e^{\mp{\pi i\over 4}}\delta_{L,{k\over 2}}

4395: \delta_{M,m+{k+2\over 2}}\delta_{S,1}

4396: \end{array}\right\}\times\widehat{\chi}_{0,0,0}(q_l)+\cdots

4397: \label{PonGSmin}

4398: \eeq

4399: where the delta functions are mod $k+2$ for $M$-indices and 

4400: mod 2 for $S$-indices. 

4401: For the universal part, we find

4402: \beqa

4403: &&\langle\Scr{B}_+|q_t^H|\Scr{C}_+\rangle_{\NSNS}

4404: =\langle\Scr{B}_-|q_t^H|\Scr{C}_-\rangle_{\NSNS}

4405: =\e^{-{\pi i\over 4}}q^{-{1\over 2}}

4406: \prod_{n=1}^{\infty}(1-i(-1)^nq_l^{n-{1\over 2}})^2,

4407: \nn\\

4408: &&\langle\Scr{B}_+|q_t^H|\Scr{C}_-\rangle_{\NSNS}

4409: =\langle\Scr{B}_-|q_t^H|\Scr{C}_+\rangle_{\NSNS}

4410: =\e^{{\pi i\over 4}}q^{-{1\over 2}}

4411: \prod_{n=1}^{\infty}(1+i(-1)^nq_l^{n-{1\over 2}})^2,

4412: \nn

4413: \eeqa

4414: up to the factors from

4415: bosonic transverse oscillators, longitudinal modes, and ghost/super\-ghost

4416: sectors.

4417: Combining the above equations, we find

4418: \beqa

4419: \lefteqn{\langle B_{\bf L,M}|q_t^H|C_{\omega;{\bf m}}\rangle_{\NSNS}}

4420: \nn\\

4421: &=&\!\!\omega^{-{1\over 2}}\e^{{\pi i\over 4}(-1+r_1-r_2)}

4422: q^{-{1\over 2}}\prod_{n=1}^{\infty}(1-i(-1)^nq_l^{n-{1\over 2}})^2

4423: -\omega^{1\over 2}\e^{{\pi i\over 4}(1-r_1+r_2)}

4424: q^{-{1\over 2}}\prod_{n=1}^{\infty}(1+i(-1)^nq_l^{n-{1\over 2}})^2

4425: \nn\\

4426: &&+\cdots

4427: \eeqa

4428: up to the universal factor, where we have decomposed $r$ as

4429: $r_1+r_2$;

4430: $$

4431: \begin{array}{l}

4432: r_1=\#\{i|M_i=m_i\}\\

4433: r_2=\#\{i|L_i={k_i\over 2},M_i=m_i+{k_i+2\over 2}\}

4434: \end{array}

4435: $$

4436: Note that

4437: $$

4438: \omega^{1\over 2}\e^{{\pi i \over 4}(1-r_1+r_2)}

4439: =\e^{{\pi i \over 4}(1-r)}\omega^{1\over 2}\e^{{\pi i\over 2}r_2}

4440: =-\omega^{1\over 2}\e^{{\pi i\over 2}r_2}

4441: $$

4442: where $r=5$ is used. It is a sign factor since

4443: $r_2$ is even

4444: for $\omega=1$, and $r_2$ is odd if $\omega=-1$.

4445: This is the sign $\sigma$ that determines the structure of Chan-Paton factor.

4446: $N$ D-branes support $O(N)$ gauge group if it is $+1$

4447: while they support $Sp(N/2)$ gauge group if it is $-1$:

4448: $$

4449: -\omega^{1\over 2}\e^{{\pi i\over 2}r_2}=

4450: \left\{\begin{array}{lcl}

4451: 1&\Longrightarrow&O(N),\\[0.2cm]

4452: -1&\Longrightarrow&Sp({N\over 2}).

4453: \end{array}\right.

4454: \label{ACPf}

4455: $$

4456: For example, consider the case where $\omega^{1\over 2}=-1$.

4457: Then, the branes with $M_i=m_i$ for all $i$ support $O(N)$ gauge group,

4458: those with two $i$'s with $L_i={k_i\over 2},M_i=m_i+{k_i+2\over 2}$ support

4459: $Sp(N/2)$ gauge group, and those with

4460: four $i$'s with $L_i={k_i\over 2},M_i=m_i+{k_i+2\over 2}$ support

4461: $O(N)$ gauge group.

4462: 

4463: 

4464: To see the LG image of this rule, let us look at the two kinds

4465: of invariant branes in the minimal model, one with

4466: $M=m$ another with $L={k\over 2}$, $M=m+{k+2\over 2}$.

4467: \begin{figure}[htb]

4468: \centerline{\includegraphics{CPfactor.eps}}

4469: \caption{Two branes invariant under the parity with $m=0$.}

4470: \label{CPfig}

4471: \end{figure}

4472: The $M=m$ branes intersect transversely with the O-plane and the

4473: $L={k\over 2}$, $M={k+2\over 2}$ branes are parallel to the O-plane.

4474: We have seen above that replacing transverse branes

4475: by parallel brane in two factors flips the type of the CP factor

4476: from $O(N)$ to $Sp({N\over 2})$ and vice versa.

4477: Note that two factors means real four-dimensions.

4478: This is very much reminiscent of what happens in 

4479: the standard superstring in flat space.

4480: For example, consider a Type II orientifold with an O7-plane.

4481: If D7-branes parallel to O7-plane support $O(N)$,

4482: D7-branes intersecting orthogonally to O7-plane in real four-dimensions

4483: support $Sp({N\over 2})$.

4484: (This is what happens if we obtain this system by T-duality from Type I

4485: and decompactification \cite{GimPol}, but we could also consider

4486: the opposite case --- parallel branes support $Sp({N\over 2})$

4487: and orthogonal branes support $O(N)$.)

4488: 

4489: 

4490: The result (\ref{ACPf})

4491: applies also to short-orbit branes,

4492: since the overlap with the crosscap state does not

4493: receive contribution from the twisted sector.

4494: 

4495: 

4496: 

4497: 

4498: \subsection{A Class of Consistent and Supersymmetric D-brane

4499: Configurations}

4500: 

4501: We have obtained the expression of the charge and the supersymmetry

4502: of the O-plane, and we also described the orientifold action on rational

4503: D-branes and the structure of CP-factor of the invariant D-branes.

4504: Thus we have obtained the condition of consistency as well as the

4505: spacetime supersymmetry on the D-brane configurations.

4506: Now, we are interested in finding solutions.

4507: It is not an easy task to classify all solutions

4508: because the rank of the charge lattice is very large (typically 100)

4509: and also there are many D-branes preserving the unbroken supersymmetry.

4510: In this subsection, we present special solutions.

4511: In the case with odd $H$, we find one solution in each case.

4512: In the case with even $H$, we present an algorithm to find a solution.

4513: It works in most of the cases but sometimes it fails.

4514: 

4515: To simplify the notation, we consider the O-plane of the reversed

4516: orientation. Namely, we use $-|C_{\omega;{\bf m}}\rangle_{\RR}$

4517: in place of the RR-crosscap state. For this choice the

4518: RR-charge and the phase for the spacetime supersymmetry (e.g.

4519: the ones in the table of page~\pageref{tableAOch}) have extra minus sign.

4520: 

4521: 

4522: 

4523: \subsubsection{Odd $H$}

4524: 

4525: If $H$ is odd, we have seen that the RR-charge of the crosscap state

4526: is equal to the RR-charge of one of the D-branes, which is

4527: $B_{\bf {k-1\over 2},{k-1\over 2}}$.

4528: One can also see that this brane preserves the same spacetime supersymmetry

4529: unbroken by the orientifold --- the phases (\ref{SUSYAbrane}) and

4530: (\ref{SUSYAOplane}) are both $-i^r\exp(-\pi i\sum_i{k_i-1\over 2(k_i+2)})$.

4531: Furthermore, this brane is invariant under the orientifold action.

4532: This can be shown as follows.

4533: As we found above, the brane

4534: is mapped to $B_{\bf {k-1\over 2},-{k-1\over 2}}$.

4535: Here, we note that

4536: $H=(k_i+2)w_i$ where $w_i$ is an integer --- in fact $w_i$ is an odd

4537: integer in the present case where $H$ is odd.

4538: This means that $H\equiv k_i+2$ mod $2(k_i+2)$, or

4539: $$

4540: (H-3)\equiv k_i-1 \quad\mbox{mod $2(k_i+2)$}.

4541: $$

4542: Using this we find that the orientifold action is

4543: $$

4544: P^A:B_{\bf {k-1\over 2},{k-1\over 2}}\longrightarrow

4545: B_{\bf {k-1\over 2},-{k-1\over 2}}=

4546: B_{{\bf {k-1\over 2},-{k-1\over 2}}+(H-3){\bf 1}}

4547: =B_{\bf {k-1\over 2},{k-1\over 2}}.

4548: $$

4549: Namely the brane $B_{\bf {k-1\over 2},{k-1\over 2}}$

4550: is mapped to itself under the

4551: orientifold action.

4552: 

4553: Thus, we find that 

4554: a consistent and spacetime supersymmetric configuration is given by

4555: four $B_{\bf {k-1\over 2},{k-1\over 2}}$'s. 

4556: This may be regarded as the configuration of four D-branes ``on top of'' the

4557: O-plane, although we do not have a geometrical picture.

4558: 

4559: One consequence of this result is that, when continued on the

4560: K\"ahler moduli space from the Gepner point to a large volume limit,

4561: the branes $B_{\bf {k-1\over 2},{k-1\over 2}}$  becomes the

4562: D-brane wrapped on the D6-brane wrapped on the fixed point set

4563: of the involution $\tau:X_i\to \overline{X}_i$.

4564: For example, the D-brane wrapped on the real quintic in the quintic

4565: hypersurface in $\CP^4$ is the continuation of

4566:  $B_{{\bf 1,1}}$.

4567: 

4568: 

4569: \subsubsection{Even $H$}

4570: 

4571: If $H$ is even, the RR-charge of the O-plane is expressed

4572: as in (\ref{Ochev}) as the sum or the difference of the

4573: untwisted part of the RR-charge of two kinds of branes,

4574: $B_{\bf {k\over 2},{k\over 2}+m-\delta_m}$ and

4575: $B_{\bf {k\over 2},{k\over 2}+m+\delta_m}$.

4576: However, generically the two preserve different combinations of

4577: supersymmetry as one can see from their phases, and in particular, neither

4578: one of them preserve the supersymmetry unbroken by the orientifold.

4579: In order to preserve spacetime supersymmetry, one has

4580: to find the set of D-branes all with the same phase whose RR-charge

4581: in total equals that of the O-plane.

4582: We study the example $(k_i+2)=(8,8,4,4,4)$

4583: in detail.

4584: 

4585: \noindent

4586: {\underline{$\bullet$ $P^A_{\pm;00000}$}}\\[0.1cm]

4587: This is a special case in which the two D-branes are the same.

4588: For the $P^A_{+;00000}$-orientifold,

4589: the O-plane has no RR charge and hence it gives

4590: a consistent supersymmetric configuration 

4591: \underline{without adding any

4592: D-branes}.

4593: For the $P^A_{-;00000}$-orientifold,

4594: the O-plane charge

4595: is equal to $-4[B_{\bf {k\over 2},{k\over 2}}]

4596: =-4[\widehat{B}_{\bf {k\over 2},{k\over 2}}^{(+)}

4597: +\widehat{B}_{\bf {k\over 2},{k\over 2}}^{(-)}]$.

4598: Thus,

4599: \underline{four $\widehat{B}^{(+)}_{\bf {k\over 2},{k\over 2}}$ and

4600: four $\widehat{B}^{(-)}_{\bf {k\over 2},{k\over 2}}$}

4601: provide a tadpole canceling brane configuration.

4602: Note that the twisted part of the RR-charge carried

4603: by the $(+)$-brane and the $(-)$-brane cancel against each other.

4604: One can also see that

4605: they preserve the same supersymmetry unbroken by the orientifold

4606: with $\omega^{1\over 2}=-i$.

4607: The $+$ branes and $-$ branes are exchanged with each other under the

4608: orientifold. Hence

4609: the gauge group supported by the branes is $U(4)$.

4610: 

4611: \noindent

4612: {\underline{$\bullet$ $P^A_{\pm;00001}$}}\\[0.1cm]

4613: The O-plane charge is given by

4614: $$

4615:  [O_{P^A_{\pm;00001}}] ~=~

4616:     -2[\widehat{B}_{{\bf k\over 2},(33111)}]

4617: \pm 2[\widehat{B}_{{\bf k\over 2},(33113)}].

4618: $$

4619: The two D-branes indeed preserve different combinations of supersymmetry

4620: since the $M$-labels of the 5-th factor are different.

4621: Let us now focus on this factor.

4622: As one can see from Fig.~\ref{keq2},

4623: \begin{figure}[htb]

4624: \centerline{\includegraphics{k=2.eps}}

4625: \caption{The recombination of the branes for $\gamma_{1,1}+\gamma_{1,3}$

4626: (middle) and $\gamma_{1,1}-\gamma_{1,3}$ (right).}

4627: \label{keq2}

4628: \end{figure}

4629: the sum and the difference of the two charges

4630: can be recombined as follows (to simplify the notation,

4631: we denote $\gamma_{L,M+1,1}$ by $\gamma_{L,M}$):

4632: $$

4633:   [\gamma^+_{1,1}]-[\gamma^+_{1,3}] ~=~

4634:   [\gamma^+_{2,0}]+[\gamma^+_{0,0}],\quad

4635:   [\gamma^+_{1,1}]+[\gamma^+_{1,3}] ~=~

4636:   [\gamma^+_{2,2}]+[\gamma^+_{0,2}].

4637: $$

4638: In either case, the two resulting

4639: wedges have a common ``mean-direction''.

4640: This shows that the two new D-branes that result from the recombination

4641: preserve the same supersymmetry.

4642: The new D-branes are

4643: $B_{\bf L_1,M_1}$ and $B_{\bf L_2,M_1}$

4644: for $\omega=1$

4645: and $\widehat{B}_{\bf L_1,M_2}$ and $\widehat{B}_{\bf L_2,M_2}$

4646: for $\omega=-1$, where

4647: $$

4648: {\bf L_1}=(33110),\,\,{\bf L_2}=(33112),\,\,

4649: {\bf M_1}=(33110),\,\,{\bf M_2}=(33112).

4650: $$

4651: They split into the sum of the

4652: $(+)$ and the $(-)$ short orbit branes.

4653: Thus, we find

4654: supersymmetric and tadpole-canceling configurations are given by:

4655: \beqa

4656: &&\mbox{\underline{two each of $\widehat{B}^{(+)}_{\bf L_1,M_1}$,

4657: $\widehat{B}^{(-)}_{\bf L_1,M_1}$,

4658: $\widehat{B}^{(+)}_{\bf L_2,M_1}$,

4659: $\widehat{B}^{(-)}_{\bf L_2,M_1}$ for $P^A_{+;00001}$-orientifold

4660: ($\omega^{1\over 2}=-1$)}},

4661: \nn\\

4662: &&\mbox{\underline{two each of

4663: $\widehat{B}^{(+)}_{\bf L_1,M_2}$,

4664: $\widehat{B}^{(-)}_{\bf L_1,M_2}$,

4665: $\widehat{B}^{(+)}_{\bf L_2,M_2}$,

4666: $\widehat{B}^{(-)}_{\bf L_2,M_2}$

4667: for $P^A_{-;00001}$-orientifold ($\omega^{1\over 2}=-i$).}}

4668: \nn

4669: \eeqa 

4670: Again, we need the same number of $+$ branes and $-$ branes to cancel

4671: the twisted part of the RR-tadpole.

4672: In both cases,

4673: the $+$ branes and the $-$ branes are exchanged by the orientifold

4674: action (${\bf L_i}$ fixed for $P^A_{+;00001}$

4675: and exchanged for $P^A_{-;00001}$).

4676: Therefore the gauge group is $U(2)\times U(2)$.

4677: 

4678: \noindent

4679: {\underline{$\bullet$ $P^A_{\pm;01000}$}}\\[0.1cm]

4680: O-plane charge is proportional to the difference or the sum of 

4681: $B_{{\bf k\over 2},(33111)}$ and

4682: $B_{{\bf k\over 2},(35111)}$.

4683: For this case, we focus of the second factor.

4684: The recombination relevant for this is depicted in

4685: Fig.~\ref{keq6}.

4686: \begin{figure}[htb]

4687: \centerline{\includegraphics{k=6.eps}}

4688: \caption{The recombination of branes for

4689: $\gamma_{3,3}+\gamma_{3,5}$ (middle) and 

4690: $\gamma_{3,3}-\gamma_{3,5}$ (right).}

4691: \label{keq6}

4692: \end{figure}

4693: In each case, the ``mean-directions'' of the two wedges are aligned

4694: after the recombination.

4695: We also note that, in each of these cases, the two branes

4696: are mapped into each other by rotation of four steps.

4697: To be precise, the orientation is revered but that is compensated by

4698: the orientation reversal for the $L={k_1\over 2}$ brane of the first factor.

4699: Thus, they are in the same (long) orbit of the $\Z_8$ orbifold group.

4700: We therefore found a supersymmetric and tadpole-canceling

4701: configuration:

4702: \underline{four $B_{\bf L_3,M_3}$ for the $P^A_{+;01000}$-orientifold}

4703: ($\omega^{1\over 2}=-1$)

4704: and

4705: \underline{four $B_{\bf L_4,M_4}$ for the $P^A_{-;01000}$-orientifold}

4706: ($\omega^{1\over 2}=-i$),

4707: where

4708: $$

4709: {\bf L_3}=(30111),\quad

4710: {\bf M_3}=(30111),\quad

4711: {\bf L_4}=(32111),\quad

4712: {\bf M_4}=(34111).

4713: $$

4714: These branes are invariant under the respective orientifolds.

4715: The gauge group is $O(4)$ for both

4716: $P^A_{+;01000}$ ($\omega^{1\over 2}=-1$)

4717: and $P^A_{-;01000}$ ($\omega^{1\over 2}=-i$).

4718: 

4719: \noindent

4720: {\underline{$\bullet$ $P^A_{\pm;00011}$}}\\[0.1cm]

4721: O-plane charge is proportional to the difference or the sum of

4722: $B_{{\bf k\over 2},(33111)}$ and

4723: $B_{{\bf k\over 2},(33133)}$.

4724: For this case, we need to focus on the two factors, 4-th and 5-th.

4725: The recombination to align the ``mean-directions'' is not obvious, but we

4726: can use the following trick. It is to use the identity

4727: $$

4728: A_1A_2-B_1B_2

4729: ={1\over 2}(A_1+B_1)(A_2-B_2)

4730: +{1\over 2}(A_1-B_1)(A_2+B_2).

4731: $$

4732: For the $P^A_{+;00011}$-orientifold,

4733: we find

4734: \beqa

4735: (\gamma_{1,1})^2-(\gamma_{1,3})^2

4736: &=&{1\over 2}(\gamma_{1,1}+\gamma_{1,3})(\gamma_{1,1}-\gamma_{1,3})

4737: +{1\over 2}(\gamma_{1,1}-\gamma_{1,3})(\gamma_{1,1}+\gamma_{1,3})

4738: \nn\\

4739: &=&{1\over 2}(\gamma_{0,2}+\gamma_{2,2})(\gamma_{0,0}+\gamma_{2,0})

4740: +{1\over 2}(\gamma_{0,0}+\gamma_{2,0})(\gamma_{0,2}+\gamma_{2,2})

4741: \nn

4742: \eeqa

4743: where we have used the recombination used in the case of

4744: $P^A_{\pm;00001}$.

4745: This indeed aligns the sum of ``mean-directions'', and thus

4746: a supersymmetry is preserved ---

4747: it is the one preserved by the $\omega^{1\over 2}=-1$ orientifold.

4748: The resulting brane configuration is the collection of \underline

4749: {sixteen branes $\widehat{B}^{(\pm)}_{\bf L_i,M_j}$, $i=5,6,7,8$, $j=5,6$}

4750: where

4751: \beqa

4752: &{\bf L_5}=(33100),\,\,

4753: {\bf L_6}=(33102),\,\,

4754: {\bf L_7}=(33120),\,\,

4755: {\bf L_8}=(33122),

4756: \nn\\

4757: &{\bf M_5}=(33120),\,\,

4758: {\bf M_6}=(33102).

4759: \nn

4760: \eeqa

4761: The orientifold exchanges the $+$ and $-$ labels 

4762: (and acts on the ${\bf L_i}$ labels in a certain way).

4763: Thus the gauge group is $U(1)^8$.

4764: 

4765: 

4766: 

4767: For the $P^A_{-;00011}$-orientifold,

4768: the same procedure on the 4-th and 5-th factors

4769: gives

4770: $$

4771: (\gamma_{1,1})^2+(\gamma_{1,3})^2

4772: ={1\over 2}(\gamma_{0,0}+\gamma_{2,0})^2

4773: +{1\over 2}(\gamma_{0,2}+\gamma_{2,2})^2

4774: $$

4775: and the sum of the ``mean-directions'' are not aligned.

4776: (The two terms have opposite phases.)

4777: Thus, this recipe of recombination

4778: does not work to find a supersymmetric configuration.

4779: In fact, in this case, the O-plane tension is vanishing

4780: (see the table in page~\ref{tableAOch}),

4781: and there is no supersymmetric brane

4782: configuration that cancels the RR-tadpole.

4783: 

4784: 

4785: 

4786: \noindent

4787: {\underline{$\bullet$ $P^A_{\pm;01001}$}}\\[0.1cm]

4788: This case is similar to the above.

4789: For the $P^A_{+;01001}$-orientifold,

4790: the recombination successfully aligns the ``mean-direction''

4791: and we find that a supersymmetric and tadpole canceling

4792: configuration is given by \underline{two each of

4793: $B_{\bf L_9,M_7}$, $B_{\bf L_{10},M_7}$, $B_{\bf L_{11},M_8}$

4794: and $B_{\bf L_{12},M_8}$}, where

4795: \beqa

4796: &{\bf L_9}=(32110),\,\,

4797: {\bf L_{10}}=(32112),\,\,

4798: {\bf L_{11}}=(30110),\,\,

4799: {\bf L_{12}}=(30112),

4800: \nn\\

4801: &{\bf M_7}=(34110),\,\,

4802: {\bf M_8}=(30112).

4803: \nn

4804: \eeqa

4805: The preserved supersymmetry is that of $\omega^{1\over 2}=-1$.

4806: ${\bf M}={\bf M_7}$ branes are invariant under the orientifold action

4807: and are of $Sp$-type, whereas the two ${\bf M}={\bf M_8}$ branes are

4808: mapped to each other.

4809: Thus the gauge group is $Sp(1)\times Sp(1)\times U(2)$.

4810: 

4811: 

4812: For the $P^A_{-;01001}$-orientifold, recombination does not work.

4813: 

4814: 

4815: \noindent

4816: {\underline{$\bullet$ $P^A_{\pm;11000}$}}\\[0.1cm]

4817: For the $P^A_{+;11000}$-orientifold,

4818: a supersymmetric and tadpole canceling configuration is given by

4819: \underline{two each of

4820: $B_{\bf L_{13},M_9}$, $B_{\bf L_{14},M_9}$, $B_{\bf L_{15},M_{10}}$

4821: and $B_{\bf L_{16},M_{10}}$}, where

4822: \beqa

4823: &{\bf L_{13}}=(20111),\,\,

4824: {\bf L_{14}}=(26111),\,\,

4825: {\bf L_{15}}=(02111),\,\,

4826: {\bf L_{16}}=(04111),

4827: \nn\\

4828: &{\bf M_9}=(40111),\,\,

4829: {\bf M_{10}}=(04111).

4830: \nn

4831: \eeqa

4832: The preserved supersymmetry is that of $\omega^{1\over 2}=-1$.

4833: Orientifold action preserves the ${\bf M}$-label

4834: but exchanges the ${\bf L}$-labels as

4835: ${\bf L_{13}}\leftrightarrow {\bf L_{14}}$

4836: and ${\bf L_{15}}\leftrightarrow {\bf L_{16}}$.

4837: Thus the gauge group is $U(2)\times U(2)$.

4838: 

4839: For the $P^A_{-;11000}$-orientifold, 

4840: recombination does not work.

4841: 

4842: \subsection{Particle Spectra in Some Supersymmetric Models}

4843: 

4844: Let us find out the spectrum of massless particles for the

4845: configurations obtained in the previous subsection.

4846: The problem here is to count the numbers of

4847: scalar fields in various open string sectors

4848: and study the action of parity on them.

4849: They are read off from the annulus and M\"{o}bius strip amplitudes.

4850: Here are some essential facts:

4851: \begin{itemize}

4852: \item The open string states $\otimes_i(l_i,n_i,s_i)$ between

4853:   two D-branes $B_{\bf L,M}$ and $B_{\bf L',M'}$ satisfy

4854:   $n_i=M'_i-M_i+2\nu$ mod $(2k_i+4)$, and $l_i$'s are also

4855:    constrained from the $SU(2)$ fusion rule.

4856: \item Massless

4857:   scalars correspond to chiral or antichiral primary states

4858:   $\otimes_i(l_i,l_i,0)$ or $\otimes_i(l_i,-l_i,0)$

4859:   with $\sum_i\frac{l_i}{k_i+2}=1$.

4860:   They are the lowest components of four-dimensional ${\cal N}=1$

4861: %Kazuo

4862:   chiral or antichiral multiplets, and are related to each other

4863:   by the worldsheet orientation reversal.

4864:   Namely, chiral primary states on $B-B'$ string and antichiral

4865:   primary states on $B'-B$ string are related to each other.

4866: \item For the open string states on the parity invariant D-branes

4867:   we have to study the action of parity.

4868:   If the brane $B_{\bf L,M}$ is invariant under the parity

4869:   $P^A_{\omega,{\bf m}}$, the open strings $\otimes_i(l_i,n_i,s_i)$

4870:   on the M\"{o}bius strip satisfy $n_i = 2M_i-2m_i+2\nu$ mod $(2k_i+4)$,

4871:   and the constraint on $l_i$ from the $SU(2)$ fusion rule.

4872:   For chiral or antichiral states satisfying the above two conditions,

4873:   the parity eigenvalue is then given by

4874: \beq

4875:   P = -i\omega^{\nu-\frac12}(-i)^{\{\#{\rm of}\,(s_i=2)\}},

4876: \label{pareig}

4877: \eeq

4878:  where we have to put $\omega^{\frac12}=-1$ or $-i$ for $\omega=\pm1$

4879:  as before.

4880: \end{itemize}

4881: 

4882: We present here the relevant amplitudes

4883: that lead to the above conclusions. (We describe them for general $r$ and $d$

4884: but we are interested in the case $r=5$ and $d=1$.)

4885: The NS part of the annulus amplitude between the A-branes

4886: $B_{\bf L,M}$ and $B_{\bf L',M'}$ is given by

4887: \begin{equation}

4888:   \frac12\sum_{\nu=1}^H\sum_{l_i}\prod_{i=1}^rN_{L_iL'_i}^{~l_i}\times

4889:   \left\{

4890:    \chi^{\rm (st)NS+}\prod_{i=1}^r\chi^{\rm NS+}_{l_i,M'_i-M_i+2\nu}

4891:   -\chi^{\rm (st)NS-}\prod_{i=1}^r\chi^{\rm NS-}_{l_i,M'_i-M_i+2\nu}

4892:   \right\}

4893: \end{equation}

4894: where $\chi^{\rm NS\pm}_{l,n}=\chi_{l,n,0}\pm\chi_{l,n,2}$ are

4895: linear combinations of minimal model characters and

4896: $\chi^{\rm (st)NS\pm}$ represent the non-compact spacetime $\R^{2d+2}$

4897: plus ghost contribution

4898: \begin{equation}

4899:   \chi^{\rm (st)NS\pm} = q^{-\frac{d}{8}}(1\pm 2d\cdot q^{\frac12}+\cdots).

4900: \end{equation}

4901: For pairs of short-orbit branes, the sum over the open string states

4902: is subject to the projection

4903: \begin{equation}

4904:   \frac14(1+\varepsilon\bar\varepsilon\prod_{w_i~{\rm odd}}

4905:   (-1)^{\frac12(l_i+M'_i-M_i)})

4906: \end{equation}

4907: The NS part of M\"{o}bius strip amplitude between the A-brane

4908: $B_{\bf L,M}$ and its image under the parity $P^A_{\omega;\bf m}$ 

4909: is given by

4910: \begin{equation}

4911:   {\rm Re}\left\{ ie^{\frac{\pi i(r-d)}{4}}

4912:      \sum_{\nu=1}^H\sum_{l_i}

4913:      \omega^{-\frac12-\nu}

4914:      \hat{\chi}^{\rm (st)NS+}

4915:      \prod_{i=1}^rN_{L_iL_i}^{~l_i}\hat{\chi}^{\rm NS+}_{l,2M_i-2m_i+2\nu}

4916:  \right\},

4917: \end{equation}

4918: where $\hat\chi^{\rm (st)NS\pm}$ represent the spacetime and ghost

4919: contributions

4920: \begin{equation}

4921:   \hat\chi^{\rm (st)NS\pm} =

4922:   q^{-\frac{d}{8}}(1\pm 2id\cdot q^{\frac12}+\cdots)

4923: \end{equation}

4924: and $\hat{\chi}^{\rm NS\pm}_{l,n}$ are defined by

4925: \begin{equation}

4926:   \hat{\chi}^{\rm NS\pm}_{l,n} = (-1)^{\frac12(l+n)}

4927:   (\hat\chi_{l,n,0}\pm i\hat\chi_{l,n,2}),~~~

4928:    \hat\chi_{l,n,s}(\tau) =

4929:    \e^{-\pi i(\frac{l(l+2)-n^2}{4k+8}+\frac{s^2}{8}-\frac{c}{24})}

4930:    \chi_{l,n,s}(\tau+1/2).

4931: \end{equation}

4932: For short-orbit branes the amplitude gets an extra factor of $\frac12$.

4933: It is easy to read off the parity eigenvalue (\ref{pareig}) for

4934: (anti)chiral primary states from this formula.

4935: 

4936: 

4937: \subsubsection{Odd $H$}\label{subsubsec:oddH}

4938: 

4939: We have seen that the configuration of

4940: four $B_{\bf \frac{k-1}{2},\frac{k-1}{2}}$'s

4941: in the $P^A_{+,{\bf 0}}$-orientifold is

4942: supersymmetric and free of tadpoles, and the gauge symmetry

4943: is $O(4)$ in all cases.

4944: However, the spectrum of massless matters depends on the model.

4945: Let us illustrate here our analysis in some examples.

4946: 

4947: \noindent

4948: {\underline{$k_i=(33333)$}}\\[0.1cm]

4949: On the annulus and on the M\"{o}bius strip we find one

4950: chiral primary state $\otimes_{i=1}^5(2,6,2)$ which is

4951: equivalent to $\otimes_{i=1}^5(1,1,0)$.

4952: Since this has $P=1$, there is one chiral multiplet

4953: belonging to the symmetric tensor representation {\bf 10}

4954: of $O(4)$.

4955: 

4956: \noindent

4957: {\underline{$k_i=(11777)$}}\\[0.1cm]

4958: There are two chiral multiplets, which appear on the M\"{o}bius

4959: strip as chiral primary states

4960: $\otimes_{i=1,2}(0,4,2)\otimes_{i=3,4,5}(6,10,2)$ and

4961: $\otimes_{i=1,2}(0,0,0)\otimes_{i=3,4,5}(4,12,2)$ respectively.

4962: The former has $P=1$ while the latter has $P=-1$, so they belong

4963: to one symmetric and one antisymmetric tensor representations

4964: of $O(4)$.

4965: 

4966: The analysis of other models goes in much the same way.

4967: The result is summarized in the table below.

4968: 

4969: \begin{table}[htb]

4970: \begin{center}

4971: \renewcommand{\arraystretch}{1.2}

4972: \begin{tabular}{|c|c|c|c|c|c|c|}

4973: \hline

4974: ($k_i$)& (3,3,3,3,3)  &(1,1,7,7,7) &(1,3,3,3,13)

4975:        & (1,1,3,13,13)&(1,1,5,5,19)&(1,1,3,7,43) \\

4976: \noalign{\hrule height 0.8pt}

4977: $\#{\bf 10}$ & 1& 1& 2& 2& 2& 8  \\\hline

4978: $\#{\bf 6} $ & 0& 1& 1& 1& 3& 3  \\\hline

4979: \end{tabular}

4980: \label{tableAHOM}

4981: \end{center}

4982: \end{table}

4983: 

4984: \subsubsection{Even $H$ -- two parameter model $k_i=(66222)$ in detail}

4985: 

4986: \noindent

4987: {\underline{$\bullet$ $P^A_{-;00000}$}}\\[0.1cm]

4988: Four short-orbit branes $\widehat{B}^{(+)}_{\bf\frac{k}{2},\frac{k}{2}}$

4989: and their parity images support $U(4)$ gauge symmetry,

4990: and there is a single adjoint matter.

4991: 

4992: \noindent

4993: {\underline{$\bullet$ $P^A_{\pm;00001}$}}\\[0.1cm]

4994: The structure of massless spectrum are the same for the two examples

4995: \beqa

4996:   P^A_{+;00001}:&& 2\widehat{B}^{(+)}_{\bf L_1,M_1}

4997:  +2\widehat{B}^{(-)}_{\bf L_1,M_1}

4998:  +2\widehat{B}^{(+)}_{\bf L_2,M_1}

4999:  +2\widehat{B}^{(-)}_{\bf L_2,M_1},

5000: \nn\\

5001:   P^A_{-;00001}:&& 2\widehat{B}^{(+)}_{\bf L_1,M_2}

5002:  +2\widehat{B}^{(-)}_{\bf L_1,M_2}

5003:  +2\widehat{B}^{(+)}_{\bf L_2,M_2}

5004:  +2\widehat{B}^{(-)}_{\bf L_2,M_2},

5005: \nn

5006: \eeqa

5007: with ${\bf L_1}=(33110),\,\,{\bf L_2}=(33112),\,\,

5008:       {\bf M_1}=(33110),\,\,{\bf M_2}=(33112).$

5009: In both cases, the gauge group is $U(2)\times U(2)$ and

5010: there are matters in the

5011: representation $({\bf 2,\bar{2}})\oplus({\bf \bar{2},2})$.

5012: 

5013: \noindent

5014: {\underline{$\bullet$ $P^A_{\pm;01000}$}}\\[0.1cm]

5015: Here we found the configurations with four long-orbit branes

5016: which are invariant under the parity.

5017: The $P^A_{+;01000}$-orientifold with four

5018: $B_{(30111),(30111)}$

5019: gives

5020: $$

5021: \mbox{$O(4)$ pure Super-Yang-Mills.}

5022: $$

5023: The $P^A_{-;01000}$-orientifold with four

5024: $B_{(32111),(34111)}$ gives

5025: $$

5026: \mbox{$O(4)$ with one symmetric and one

5027: antisymmetric matters.}

5028: $$

5029: 

5030: \noindent

5031: {\underline{$\bullet$ $P^A_{+;00011}$}}\\[0.1cm]

5032: Here we find a very interesting situation.

5033: The tadpole canceling configuration we have found is

5034: eight short-orbit branes $B_{\rm I},\cdots,B_{\rm VIII}$

5035: and their parity images, where

5036: $$

5037: \begin{array}{rclrclrclrcl}

5038: B_{\rm I}   &=& \widehat{B}^{(+)}_{{\bf L_5, M_5}},&

5039: B_{\rm II}  &=& \widehat{B}^{(+)}_{{\bf L_6, M_5}},&

5040: B_{\rm III} &=& \widehat{B}^{(+)}_{{\bf L_7, M_5}},&

5041: B_{\rm IV}  &=& \widehat{B}^{(+)}_{{\bf L_8, M_5}},\\

5042: B_{\rm V}   &=& \widehat{B}^{(+)}_{{\bf L_5, M_6}},&

5043: B_{\rm VI}  &=& \widehat{B}^{(+)}_{{\bf L_6, M_6}},&

5044: B_{\rm VII} &=& \widehat{B}^{(+)}_{{\bf L_7, M_6}},&

5045: B_{\rm VIII}&=& \widehat{B}^{(+)}_{{\bf L_8, M_6}},

5046: \end{array}

5047: $$

5048: with

5049: ${\bf L_5}=(33100)$,

5050: ${\bf L_6}=(33102)$,

5051: ${\bf L_7}=(33120)$,

5052: ${\bf L_8}=(33122)$,

5053: ${\bf M_5}=(33120)$,

5054: and ${\bf M_6}=(33102)$.

5055: The gauge group is $U(1)^8$, and we have quite a few matter

5056: fields which are charged under two of $U(1)$'s.

5057: The spectrum are the most neatly expressed in terms of the

5058: quiver diagram, where each arrow represents a chiral multiplet

5059: charged $+1$ and $-1$ under the $U(1)$'s on its head

5060: and tail.

5061: \begin{figure}[htb]

5062: \centerline{\includegraphics{quiver1.eps}}

5063: \caption{quiver diagram representing a D-brane

5064:          configuration with $P^A_{+;00011}$}

5065: \label{quiver1}

5066: \end{figure}

5067: Note that the gauge theory is chiral.

5068: There is a mixed $U(1)_aU(1)_b^2$ anomaly for each pair

5069: $(a,b)$ of neighboring groups of the quiver

5070: (i.e. VII-I,IV-VII,VI-IV,VI-I for the first square,

5071: and similarly for the second square).

5072: Anomaly cancellation mechanism will be discussed in

5073: Section~\ref{subsec:anomaly}.

5074: 

5075: 

5076: 

5077: 

5078: 

5079: \noindent

5080: {\underline{$\bullet$ $P^A_{+;01001}$}}\\[0.1cm]

5081: For this parity we found a tadpole canceling configuration

5082: $$

5083:  2B_{\bf L_9,   M_7}

5084: +2B_{\bf L_{10},M_7}

5085: +2B_{\bf L_{11},M_8}

5086: +2B_{\bf L_{12},M_8}

5087: $$

5088: with

5089: ${\bf L_9}=(32110)$,

5090: ${\bf L_{10}}=(32112)$,

5091: ${\bf L_{11}}=(30110)$,

5092: ${\bf L_{12}}=(30112)$,

5093: ${\bf M_7}=(34110)$,

5094: and ${\bf M_8}=(30112)$.

5095: The gauge group is $Sp(1)\times Sp(1)\times U(2)$

5096: and there are matters in the representations

5097: $$

5098:  \bf {\rm 2\times}(2,2,1)\oplus(2,1,2)\oplus(1,2,\bar{2}).

5099: $$

5100: This system is also chiral.

5101: There are mixed $U(1)Sp(1)_a^2$ anomalies.

5102:  Anomaly cancellation mechanism will be discussed in

5103: Section~\ref{subsec:anomaly}.

5104: 

5105: 

5106: 

5107: \noindent

5108: {\underline{$\bullet$ $P^A_{+;11000}$}}\\[0.1cm]

5109: We found a D-brane configuration

5110: $$

5111: 2B_{\bf L_{13},M_9}+

5112: 2B_{\bf L_{14},M_9}+

5113: 2B_{\bf L_{15},M_{10}}+

5114: 2B_{\bf L_{16},M_{10}}

5115: $$

5116: with

5117: ${\bf L_{13}}=(20111)$,

5118: ${\bf L_{14}}=(26111)$,

5119: ${\bf L_{15}}=(02111)$,

5120: ${\bf L_{16}}=(04111)$,

5121: ${\bf M_9}=(40111)$, and

5122: ${\bf M_{10}}=(04111)$.

5123: The gauge group is $U(2)\times U(2)$, and the matter

5124: belongs to $\bf (2,\bar{2})\oplus(\bar{2},2)$.

5125: 

5126: 

5127: 

5128: \subsection{More general tadpole canceling configurations}

5129: \label{subsec:general}

5130: 

5131: In the previous subsection, we have seen that it is generically 

5132: rather easy to find a supersymmetric tadpole canceling brane 

5133: configuration for Type IIA orientifolds of Gepner models. When 

5134: all levels are odd, these configurations corresponds to placing 

5135: $4$ D-branes on top of the O-plane, and leads to $O(4)$ 

5136: gauge group with some matter content which depends on the 

5137: particular model. On the other hand, we have seen that when 

5138: some levels are even, we can have somewhat more interesting 

5139: configurations which support unitary gauge groups and chiral 

5140: matter.

5141: 

5142: It would be interesting to know whether these are all solutions and 

5143: if not, how to describe the set of all possibilities. Let us recall

5144: the general nature of the problem. First of all, we emphasize once

5145: again that we have only been considering certain rational boundary states,

5146: which are just a subset of all possible branes that could be used to

5147: cancel the tadpoles. It would be interesting to see if one can

5148: obtain more interesting possibilities by using for example the

5149: boundary states constructed in \cite{recknagel} (which are still

5150: rational, but more general than the ones we have considered here). 

5151: Secondly, we wish to point out that the problem of finding tadpole 

5152: canceling configurations does not actually depend on whether we 

5153: are considering Type IIA or Type IIB (the two are just exchanged 

5154: by mirror symmetry). In other words, it is sufficient to discuss 

5155: the conditions (\ref{internal}) in the internal CFT.

5156: 

5157: 

5158: \noindent

5159: With these comments in mind, we are looking for sets of rational

5160: D-branes which

5161: 

5162: (i) have the same 

5163: RR-charge as the O-plane,

5164: 

5165: (ii) are invariant under the parity, 

5166: 

5167: (iii) allow a consistent assignment of Chan-Paton factor,

5168: 

5169: \noindent

5170: and, if we are interested in spacetime supersymmetric configurations,

5171: 

5172: (iv) preserve a common ${\mathcal N}=1$ supersymmetry.

5173: 

5174: \noindent

5175: These conditions are solved in steps.\\

5176: \underline{Step 1:} Choose the parity $P$. Compute the

5177: RR-charge $[O_P]$ and the preserved spacetime supersymmetry 

5178: $M_O$ of the corresponding O-plane.\\

5179: \underline{Step 2:} Make a list of rational branes $B_i$ 

5180: preserving the same spacetime supersymmetry as the O-plane, and 

5181: compute their RR-charges $[B_i]$. We note that we 

5182: have to distinguish branes even if they have the same RR charge.\\

5183: \underline{Step 3:} Determine for each brane its image 

5184: $B_{P(i)}$ under the parity. If a brane is fixed under the parity,

5185: determine whether the gauge group is of $O$ or $Sp$-type.

5186: We will use an indicator $\sigma_i$ to concisely denote this 

5187: gauge group. If the gauge group supported on $n_i$ branes 

5188: $[B_i]$ is $O(n_i)$, we will set $\sigma_i=+1$,

5189: if it is of type $Sp(n_i/2)$ we set $\sigma_i=-1$. If

5190: the brane is not invariant under $P$ (so the gauge group is

5191: $U(n_i)$), we will set $\sigma_i=0$.\\

5192: \underline{Step 4:} Solve the equation

5193: \beq

5194: \sum_i n_i [B_i] = [O_P]

5195: \label{tadcan}

5196: \eeq

5197: for positive integers $n_i$ under the condition that

5198: $n_i=n_{P(i)}$ and that $n_i$ is even if $\sigma_i=-1$.

5199: 

5200: Steps 1-3 are of course just those that we have been taking above.

5201: The hard part is solving (\ref{tadcan}). Indeed, while this

5202: is a linear equation, there are a large number of equations

5203: to solve (on the order of $100$ RR charges) and a large number

5204: of variables (on the order of several thousands branes preserving

5205: the same supersymmetry as any given O-plane). The number of

5206: solutions to this Diophantine problem is finite when restricted

5207: to positive integers $n_i$, because there is always one

5208: equation in which all $n_i$ appear with a positive 

5209: coefficient. The simplest way to obtain this equation is 

5210: to take the overlap with the RR ground state $|0\rangle_{\rm RR}$. 

5211: By eq.\ (\ref{ssusy}), this is proportional to the overlap 

5212: with the NSNS ground state, so what we are saying is 

5213: simply that the tensions of the branes are all positive 

5214: and must cancel the tension of the O-plane. 

5215: 

5216: Knowing that the number of solutions is finite, one

5217: would like to count or even enumerate the solutions. A

5218: priori, it is not even clear that there is a single one

5219: (besides the somewhat trivial ones we have already found).

5220: To estimate the difficulty of the problem, we notice that

5221: the tension of the A-type orientifold in the quintic is about 

5222: (minus) 20 times the typical tension of A-branes.

5223: So if we try to scan all the configurations in which the tension

5224: of D-branes cancel the negative tension of orientifold, there

5225: are roughly $\left({2000\atop20}\right)\sim10^{50}$ of them,

5226: which is too large a number to look through even with the help

5227: of a computer. One has to resort to a more direct method.

5228: 

5229: The problem becomes dramatically simpler when the number of

5230: equations and the number of possible branes is smaller.

5231: For B-type on the quintic, for example, there are $2$ linearly

5232: independent equations and $32$ variables. This problem 

5233: (and its analogs for the two parameter model) can be solved 

5234: completely, as we do in section \ref{sec:TCCB}. If we consider

5235: the orbifold of the quintic by a certain $\Z_5$ phase symmetry

5236: (see \cite{GreenePlesser}), it turns out that there

5237: are $6$ equations in $96$ variables. This problem is still 

5238: tractable, and we present some solutions in appendix 

5239: \ref{orbifold}.

5240: 

5241: A purely technical difficulty is to find the right basis in

5242: which to write the equations (\ref{internal}). The simplest

5243: basis might seem to be the basis of RR ground states

5244: which are products of the minimal model ground states in 

5245: the form

5246: \beq

5247: |l_i\rangle = \prod_{i=1}^5|l_i,l_i+1,1\rangle\times|l_i,-l_i-1,-1\rangle

5248: \qquad

5249: (1\le l_i+1\le k_i+1,~\sum_{i}{\textstyle\frac{l_i+1}{k_i+2}}\in\Z)

5250: \label{ground}

5251: \eeq

5252: However, the problem is that when written in this basis,

5253: the Diophantine equations (\ref{tadcan}) are not manifestly

5254: integral --- the coefficients are certain combinations of

5255: trigonometric and exponential functions. To remedy this

5256: situation, one can use the fact that (at the level of 

5257: charges) some branes can be written as integral linear

5258: combinations of other branes. A convenient choice of

5259: reference branes --- for any Gepner model --- are the 

5260: branes with ${\bf L}={\bf 0}$ and varying ${\bf M}$.

5261: This ``basis'' has often been used in previous works on the RS 

5262: boundary states in Gepner models. We note three important facts. 

5263: \\ (i) The ${\bf L}={\bf 0}$ branes in general only generate

5264: a sublattice of the full BPS charge lattice. As we have mentioned

5265: before, the generic Gepner model has chiral ring elements from 

5266: twisted sectors, corresponding to non-toric blowups in the geometry.

5267: The corresponding RR fields do not couple to the RS branes, which

5268: preserve a diagonal chiral algebra in each minimal model.

5269: \\ (ii) The ${\bf L}={\bf 0}$ branes are not in general primitive

5270: generators of the charge lattice. It is an outstanding problem

5271: to find boundary states which are integral generators of the

5272: charge lattice.

5273: \\ (iii) The charges of the ${\bf L}={\bf 0}$ branes are not

5274: linearly independent. It can sometimes be a little cumbersome 

5275: to eliminate these relations in (iii) in order to find the linearly 

5276: independent conditions. We discuss how this can be done in appendix 

5277: \ref{integral}.

5278: 

5279: \subsubsection*{O(4) Configuration is Not Always Possible}

5280: 

5281: 

5282: 

5283: To conclude this section, we answer (in the negative) the following

5284: question: At large volume, the tadpole of the O6-plane is always

5285: exactly canceled by the four D-branes all wrapping on it. Can

5286: we always find a corresponding solution at the Gepner point?

5287: We have found such a configuration for each of odd $H$ Gepner models.

5288: However, in the model with even $H$ one cannot always

5289: find such a configuration.

5290: In the two parameter model, it turns out that the only such solutions

5291: are those for $P^A_{\pm,01000}$ found already.

5292: 

5293: 

5294: 

5295: To see this, let us first look at the tensions of the O-planes.

5296: For the

5297: $P^A_{\omega;{\bf m}}$-orientifold, it is related to

5298: the tension of the D-brane $B_{(33111),{\bf M}}$

5299: by

5300: \beqa

5301: T_{O^A_{+;{\bf m}}}&=&

5302:  -4\sin{\ts\frac{\pi n}{8}}T_{(33111),{\bf M}},\nn\\

5303: T_{O^A_{-;{\bf m}}} &=&

5304:  -4\left|\cos{\ts\frac{\pi n}{8}}\right|T_{(33111),{\bf M}}

5305:  \nn

5306: \eeqa

5307: where $n:=\sum_iw_im_i$ if we assume $m_i=0$ or $1$.

5308: One can show that there is no $O(4)$ configurations

5309: when $n\!=\!0$ or $4$, in the following way.

5310: In these cases the O-plane tension, if nonvanishing,

5311: is equal to $-4T_{(33111),{\bf M}}$,

5312: but the brane $B_{(33111),{\bf M}}$ is sum of two short orbit branes.

5313: Moreover, any elementary brane has tension less that

5314: $T_{(33111),{\bf M}}$.

5315: Thus, the O-plane tension cannot be canceled with just four identical

5316: elementary branes.

5317: For $n=1,2$ or $3$ the tensions are rewritten as

5318: \beqa

5319: T_{O^A_{+;{\bf m}}} &=&

5320:  -4T_{(3,n-1,1,1,1),{\bf M}}~=~

5321: - 4T_{(n-1,3,1,1,1),{\bf M}},\nn\\

5322: T_{O^A_{-;{\bf m}}} &=&

5323:  -4T_{(3,3-n,1,1,1),{\bf M}} ~=~

5324:  -4T_{(3-n,3,1,1,1),{\bf M}}

5325: \nn

5326: \eeqa

5327: and similarly for $n=5,6,7$.

5328: There are no other branes with the same tension.

5329: Thus the only configurations with four identical D-branes

5330: that can cancel the O-plane tensions are

5331: $$

5332: \begin{array}{c}

5333: P^A_{+;00001}\Rightarrow 4B_{(31111),{\bf M}}~,\\

5334: P^A_{+;01000}\Rightarrow 4B_{(30111),{\bf M}}~,\\

5335: P^A_{+;01001}\Rightarrow 4B_{(32111),{\bf M}}~,\\

5336: P^A_{+;11000}\Rightarrow 4B_{(31111),{\bf M}}~,

5337: \end{array}

5338: ~~

5339: \begin{array}{c}

5340: P^A_{-;00001}\Rightarrow 4B_{(31111),{\bf M}}~,\\

5341: P^A_{-;01000}\Rightarrow 4B_{(32111),{\bf M}}~,\\

5342: P^A_{-;01001}\Rightarrow 4B_{(30111),{\bf M}}~,\\

5343: P^A_{-;11000}\Rightarrow 4B_{(31111),{\bf M}}~

5344: \end{array}

5345: $$

5346: and those obtained by exchanging $L_1$ and $L_2$.

5347: 

5348: Let us then see whether any of the above relations are lifted

5349: to the full equality between the RR-charges.

5350: The possibilities in the first and the fourth rows

5351: can be easily excluded by looking at their transformation

5352: property under the permutations of the first two minimal models.

5353: Also, those in the first and the third rows are excluded because

5354: the branes cannot be parity invariant under any choice of ${\bf M}$ label.

5355: This is easily seen by noting that, for the branes $B_{\bf L,M}$

5356: in the two parameter model to be invariant under

5357: $P^A_{\omega,{\bf m}}$, $M_i-m_i$ have to be all even or all add.

5358: Thus we are left with the ones in the second row, for which

5359: the permutation of minimal models, parity invariance and $O(4)$

5360: gauge symmetry reduce the possibilities to

5361: $$

5362: P^A_{+;01000}\Rightarrow 4B_{(30111),(30111)}\,,~~

5363: P^A_{+;01000}\Rightarrow 4B_{(30111),(12111)}\,,~~

5364: P^A_{-;01000}\Rightarrow 4B_{(32111),(34111)}\,.

5365: $$

5366: Two of them are the configurations that was already found before.

5367: The remaining (second) case does not satisfy the full tadpole condition,

5368: as can be guessed from the comparison with the first one and

5369: confirmed by a more detailed analysis of the tadpole condition.

5370: 

5371: 

5372: 

5373: 

5374: 

5375: 

5376: 

5377: 

5378: \section{Chirality, Anomaly Cancellation, and Fayet-Ilio\-poulos Terms}

5379: \label{sec:FI}

5380: 

5381: In this section, we take a break and make some general remarks

5382: on chirality, anomaly cancellation mechanism, and

5383: Fayet-Iliopoulos terms.

5384: 

5385: 

5386: 

5387: 

5388: \subsection{Chirality and Witten Indices}

5389: 

5390: Chirality of the theory can be measured by

5391: the open string Witten indices of the internal CFT,

5392: \beqa

5393: &&

5394: I(B,B')=\Tr\!\!\mathop{}_{{\mathcal H}_{\Scr{B},\Scr{B}'}}(-1)^{F_{\rm int}},

5395: \nn\\

5396: &&

5397: I(B,O_P)=\Tr\!\!\mathop{}_{{\mathcal H}_{\Scr{B},P(\Scr{B})}}

5398: (-1)^{F_{\rm int}}P.

5399: \nn

5400: \eeqa

5401: To see this, note that the GSO operator can be written as the product 

5402: $(-1)^F=(-1)^{F_{\rm int}}(-1)^{F_{\rm st}}$.

5403: The spacetime part $(-1)^{F_{\rm st}}$ acts on the RR sector states,

5404: that is, on the spacetime fermions, as the gamma-five matrix,

5405: $\Gamma^5=\Gamma^0\Gamma^1\Gamma^2\Gamma^3$.

5406: Thus, by the GSO projection, the Witten index of the internal part

5407: $\Tr (-1)^{F_{\rm int}}$ is proportional to

5408: $\Tr_{\rm massless}\Gamma^5$, which measures the chirality of the theory.

5409: 

5410: 

5411: 

5412: Let $B$ and $B'$ be branes that are {\it not} the orientifold images

5413: of each other, $P(B)\ne B'$. The parity maps the $B$-$B'$ string to

5414: the $P(B')$-$P(B)$ string, and hence

5415: the orientifold projection simply relates the two

5416: string sectors. The chirality of the bifundamental representation

5417: is given by the index

5418: \beq

5419: \#(\overline{\bf n}_B,{\bf n}_{B'})-

5420: \#({\bf n}_B,\overline{\bf n}_{B'})

5421: =I(B,B').

5422: \eeq

5423: By definition, it must be antisymmetric with respect to

5424: the exchange of $B$ and $B'$. This is guaranteed by the antisymmetry

5425: of the index in the internal CFT, $I(B',B)=-I(B,B')$.

5426: 

5427: 

5428: The string stretched from a brane $B$

5429: to its parity image $P(B)$ is invariant under the parity action.

5430: Its Chan-Paton factor is 

5431: $\overline{\bf n}_B\otimes {\bf n}_{P(B)}

5432: =\overline{\bf n}_B\otimes\overline{\bf n}_B$,

5433: the second rank tensor product of $\overline{\bf n}_B$.

5434: The orientifold projection selects the symmetric tensor

5435: times the $P=1$ states

5436: as well as the antisymmetric tensor times the $P=-1$ states.

5437: Note that the ordinary $B$-$P(B)$ index

5438: is the sum of the index in the $P=1$ subspace and the one in the

5439: $P=-1$ subspace,

5440: %$I(B,P(B))=I^{P=1}(B,P(B))+I^{P=-1}(B,P(B))$,

5441: while the twisted index is the difference.

5442: %,

5443: %$I(B,O_P)=I^{P=1}(B,P(B))-I^{P=-1}(B,P(B))$.

5444: Thus, the chirality of the symmetric and antisymmetric representation

5445: is

5446: \beqa

5447: &&\#{\rm S}^2\overline{\bf n}_B

5448: -\#{\rm S}^2{\bf n}_B={1\over 2}\Bigl(I(B,P(B))+I(B,O_P)\Bigr)

5449: \\

5450: &&\#{\rm A}^2\overline{\bf n}_B

5451: -\#{\rm A}^2{\bf n}_B={1\over 2}\Bigl(I(B,P(B))-I(B,O_P)\Bigr)

5452: \eeqa

5453: They must vanish if the brane is parity invariant,

5454: $P(B)=B$, and the representation $\overline{\bf n}_B$ is real or pseudo-real.

5455: This is again guaranteed by the antisymmetry of the indices

5456: $I(P(B),B)=-I(B,P(B))$, $I(O_P,P(B))=-I(B,O_P)$.

5457: 

5458: 

5459: \subsubsection{Examples in IIA Gepner Models}

5460: 

5461: Let us study the Witten index of branes

5462: in Type IIA orientifolds of Gepner model.

5463: Long-orbit A-branes are simply the sum over images

5464: under the orbifold group $\Z_H$. Accordingly, the index is also given by the

5465: sum of the product of the minimal model indices

5466: \beq

5467: I(B_{\bf L,M},B_{\bf L',M'})

5468: =\sum_{\nu=0}^{H-1}\prod_{i=1}^5I(B_{L_i,M_i},

5469: B_{L_i',M_i'+2\nu}).

5470: \label{sumoo}

5471: \eeq

5472: Furthermore, the index for each minimal model is given by

5473: the intersection number of the wedge cycle, $I(B_{L_i,M_i},B_{L_i',M_i'})

5474: =\#(\gamma_{L_i,M_i}^-\cap \gamma_{L_i',M_i'}^+)$.

5475: In many supersymmetric brane configurations,

5476: there are two or more branes with the same ${\bf M}$-label.

5477: So let us examine such a pair, $B_{\bf L,M}$ and $B_{\bf L',M}$.

5478: In the LG picture, the two product 

5479: branes $\prod_i\gamma_{L_i,M_i}$ and $\prod_i\gamma_{L'_i,M_i}$

5480:  have the common mean-direction at each minimal model factor.

5481: After rotation by the orbifold group they no longer have the same

5482: mean-direction, but there is an interesting relation between

5483: different steps of rotation:

5484: $\gamma_{L'_i,M_i}^+$ rotated by a step $\nu$

5485: and $\gamma_{L'_i,M_i}^+$ rotated by the opposite step $-(\nu+1)$

5486: have the opposite intersection number with

5487: any $\gamma_{L_i,M_i}$:

5488: $$

5489: \#(\gamma_{L_i,M_i}^-\cap \gamma_{L_i',M_i'+2\nu}^+)

5490: =-\#(\gamma_{L_i,M_i}^-\cap \gamma_{L_i',M_i'-2\nu-2}^+).

5491: $$

5492: It follows that

5493: the $\nu$-th term in (\ref{sumoo}) is opposite to the $(H-1-\nu)$-th

5494: term, and the sum vanishes.

5495: Thus, we find the

5496: 

5497: \noindent

5498: {\bf Vanishing Theorem:}\\

5499: {\it The index between two long-orbit branes of the same ${\bf M}$-label

5500: vanishes.}

5501: 

5502: \noindent

5503: This helps us in finding chiral pairs of branes in a given model:

5504: If two D-branes have the same ${\bf M}$-labels,

5505: we find that they are non-chiral before analyzing the spectrum.

5506: We will see a similar vanishing theorem in Type IIB orientifolds.

5507: 

5508: Supersymmetry does not require that the branes to have the same

5509: ${\bf M}$-label but only that $\sum_i{M_i\over k_i+2}$ to be

5510: the same (modulo $H$). Indeed, in the examples we have studied,

5511: there are many configurations with various ${\bf M}$-labels.

5512: For example, consider the $P^A_{+;01001}$-orientifold with

5513: the branes $B_{\bf L_9,M_7}$,

5514: $B_{\bf L_{10},M_7}$,

5515: $B_{\bf L_{11},M_8}$ and

5516: $B_{\bf L_{12},M_8}$ (two each).

5517: We find

5518: $$

5519: I(B_{\bf L_9,M_7},B_{\bf L_{11},M_8})=1,\quad

5520: I(B_{\bf L_9,M_7},B_{\bf L_{12},M_8})=-1.

5521: $$

5522: Indeed, we have seen that this system is chiral by an explicit

5523: spectrum analysis.

5524: 

5525: 

5526: 

5527: 

5528: 

5529: 

5530: \subsection{Anomaly Cancellation Mechanism}\label{subsec:anomaly}

5531: 

5532: Let us consider a tadpole canceling brane configuration $\{n_aB_a\}$

5533: in a Type II orientifold with respect to a worldsheet

5534: parity symmetry $P$.

5535: The gauge group $G_a$ supported by

5536: the $n_a$ branes $B_a$ is

5537: $U(n_a)$ if $B_a$ is not invariant under the parity

5538: while it is $G_a=O(n_a)$ or $Sp(n_a/2)$ if the brane is invariant.

5539: The tadpole cancellation condition is

5540: $$

5541: \sum_an_a[B_a]=[O_P].

5542: $$

5543: The standard triangle anomalies in the low energy field theory

5544: are proportional to

5545: \beqa

5546: &&

5547: A_{U(n_a)U(n_b)^2}=I(B_a,B_b)+I(B_a,B_{P(b)})

5548: \label{Ananb}\\

5549: &&

5550: A_{U(n_a){\rm Gravi}^2}=\sum_bI(B_a,B_b)n_b=I(B_a,O_P).

5551: \eeqa

5552: where tadpole cancellation condition is used in the second equation.

5553: For the $U(n_a)G_b^2$ anomaly with $G_b=O(n_b)$ or $Sp(n_b/2)$,

5554: it is simply $I(B_a,B_b)$ (no

5555:  extra term $I(B_a,B_{P(b)})$ as in (\ref{Ananb})).

5556: Note that only the $U(1)_a$ subgroup of $U(n_a)$ is anomalous.

5557: 

5558: 

5559: This field theory anomaly is canceled by the Green-Schwarz mechanism

5560: \cite{GS,GSW,DSW,DM}. The relevant

5561:  Green-Schwarz terms are obtained from

5562: the disc diagrams with bulk insertion of a RR axion $\vartheta^{\bf i}$

5563: and one or two boundary insertion of the gauge bosons,

5564: or from the $\RP^2$ diagrams with insertion of a $\vartheta^{\bf i}$

5565: and two gravitons.

5566: They are proportional to the overlaps

5567: $$

5568: \Pi_{{\bf i}}^a=\langle\Scr{B}_a|\e^{-\pi i J_0}|{\bf i}\rangle_{\RR},

5569: \quad

5570: \widetilde{\Pi}^b_{\bf i}={}_{\RR}\langle {\bf i}|\Scr{B}_b\rangle,

5571: \quad

5572: \widetilde{\Pi}^P_{\bf i}={}_{\RR}\langle {\bf i}|\Scr{C}_P\rangle,

5573: $$

5574: and are given by

5575: \beq

5576: \Pi_{\bf i}^aA^{U(1)_a}_{\mu}\partial^{\mu}\vartheta^{\bf i}

5577: \label{axion}

5578: \eeq

5579: and

5580: $$

5581: (\widetilde{\Pi}^b_{\bf i}+\widetilde{\Pi}^{P(b)}_{\bf i})

5582: \vartheta^{\bf i}\Tr_{\bf n_b}F^b\wedge F^b,\qquad

5583: \widetilde{\Pi}^P_{\bf i}\vartheta^{\bf i}\Tr R\wedge R.

5584: $$

5585: Note that the axion $\vartheta^{\bf i}$ corresponds to the RR ground state

5586: $|{\bf i}\rangle_{\RR}$ that survives the orientifold projection.

5587: If there are $2(h+1)$ RR ground states obeying the same

5588: R-charge selection rule as the boundary states, the number of

5589: such $|{\bf i}\rangle_{\RR}$ is $(h+1)$.

5590: Also, we have chosen a basis such that the overlaps

5591: $\Pi^a_{\bf i}$, $\widetilde{\Pi}^b_{\bf i}$ are real.

5592: The coupling (\ref{axion}) induces an anomalous

5593: $U(1)_a$ gauge transformation of the axion

5594: $$

5595: \vartheta^{\bf i}\longrightarrow 

5596: \vartheta^{\bf i}+g^{\bf ij}\Pi_{\bf j}^a\lambda_a,

5597: $$

5598: where $g^{\bf ij}$ is the inverse matrix of

5599: $g_{\bf ij}={}_{\RR}\langle {\bf j}|{\bf i}\rangle_{\RR}$

5600: which determines the axion kinetic term,

5601: $g_{\bf ij}\partial^{\mu}\vartheta^{\bf i}\partial_{\mu}\vartheta^{\bf j}$.

5602: Then, the triangle anomalies are canceled

5603: as a consequence of the bilinear identity

5604: \footnote{Alternatively, one can

5605: use the $\theta^{\bf i}$-equations of motion (instead of the

5606: anomalous transformation).

5607: This again introduces $g^{\bf ij}$ here from the axion kinetic term.}

5608: $$

5609: \sum_{\bf i,j}\Pi^a_{\bf i}g^{\bf ij}(\widetilde{\Pi}^b_{\bf j}

5610: +\widetilde{\Pi}^{P(b)}_{\bf j})

5611: =I(B_a,B_b)+I(B_a,B_{P(b)}),

5612: \qquad

5613: \sum_{\bf i,j}\Pi^a_{\bf i}g^{\bf ij}\widetilde{\Pi}^P_{\bf j}=I(B_a,O_P).

5614: $$

5615: The bilinear identity holds for the sum over {\it all}

5616: RR ground states \cite{BH2}. However, in the present case, the sum

5617: can be restricted to the orientifold-invariant states

5618: $|{\bf i}\rangle_{\RR}$,

5619: because $(\widetilde{\Pi}^b_i+\widetilde{\Pi}^{P(b)}_i)$

5620: and $\widetilde{\Pi}^P_i$ are non-vanishing only for such

5621: $|{\bf i}\rangle_{\RR}$.

5622: Same is true on the overlap $\widetilde{\Pi}_i^b$ for a parity invariant

5623: brane, $P(b)=b$.

5624: 

5625: For Type IIA orientifolds of a Calabi-Yau manifold $M$,

5626: the RR-ground states $|i\rangle_{\RR}$ contributing to the overlap

5627: $\Pi_i^a$ with A-branes corresponds to middle dimensional forms,

5628: and the axions are the KK reduction of the RR 3-form

5629: on $H^3(M)$. At the Gepner point,

5630: the rational A-branes have overlap only with the untwisted states

5631: since the boundary states are sum over images.

5632: Thus the Green-Schwarz mechanism works with the untwisted

5633: RR-fields, as long as rational A-branes are concerned.

5634: Similar situations are encountered in the context of

5635: toroidal orbifold in \cite{Alda1,CSU}.

5636: 

5637: For Type IIB orientifolds of a Calabi-Yau manifold $M$,

5638: the RR-ground states $|i\rangle_{\RR}$ contributing to the overlap

5639: $\Pi_i^a$ with B-branes corresponds to diagonal forms, $H^{p,p}(M)$,

5640: and the axions are the KK reduction of the RR 0, 2, 4-forms.

5641: At the Gepner point, the rational B-branes

5642: generically have overlap with the twisted sector states

5643: since the boundary state is sum over twists.

5644: Thus the Green-Schwarz mechanism works with the twisted

5645: RR-fields, just as in Type I orbifolds studied in \cite{DM}

5646: (see also \cite{IRU1}).

5647: 

5648: 

5649: \subsection{Fayet-Iliopoulos Terms}

5650: 

5651: The coupling (\ref{axion})

5652: is extended to the full kinetic term

5653: $(\partial_{\mu}\vartheta^{\bf i}+g^{\bf ij}\Pi_{\bf j}^aA_{\mu}^{U(1)_a})^2$,

5654: and its supersymmetric completion is

5655: \beq

5656: \int\dd^4\theta\,\, K\Bigl(Y^{\bf i}+\overline{Y^{\bf i}}

5657: +g^{\bf ij}\Pi_{\bf i}^aV_a\Bigr).

5658: \label{FIterm}

5659: \eeq

5660: Here $Y^{\bf i}$ is a chiral superfield whose lowest component

5661: is a complex scalar whose imaginary part is the axion

5662: $y^{\bf i}=c^{\bf i}-i\vartheta^{\bf i}$.

5663: This means that the real part $c^{\bf i}$

5664: enters into the Fayet-Iliopoulos parameter \cite{DSW}

5665: $$

5666: \zeta^a=\sum_{\bf i}c^{\bf i}\Pi^a_{\bf i}.

5667: $$

5668: This is true as long as the gauge group includes $U(1)$ factors,

5669: independently of whether the particle spectrum is chiral.

5670: 

5671: 

5672: 

5673: 

5674: 

5675: In Type IIA orientifolds, the superpartner of RR axions are

5676: the complex structure moduli fields which are constrained

5677: to the ``real section'' by the parity invariance. 

5678: Thus, the ``real'' complex structure moduli fields

5679: can enter into the FI parameters.

5680: In the previous section, we have constructed many supersymmetric

5681: (and tadpole canceling) brane configurations at the Gepner point.

5682: As we move away from the Gepner point in the complex structure moduli space,

5683:  the phases $\Pi^a_0$ may no longer align and

5684:  the branes preserve different combinations of the

5685: spacetime supersymmetry. In such a situation, we expect either

5686: the branes recombine into other branes so that the supersymmetry is

5687: restored, or there is no such configuration and the supersymmetry is broken.

5688: This is exactly the situation

5689: described by the above low energy field theory:

5690: Under the deformation of $c^{\bf i}$ such that the FI parameter $\zeta^a$

5691: becomes non-zero,

5692: some charged scalar fields become tachyonic and condense

5693: to find a supersymmetric vacua, or supersymmetry is broken.

5694: A local model of such phenomenon was in fact constructed by

5695: Kachru and McGreevy \cite{KM}.

5696: 

5697: 

5698: 

5699: 

5700: 

5701: The $U(1)$ gauge boson with non-zero $\Pi^a_{\bf i}$

5702: acquires a mass by eating a combination of

5703: the moduli fields $Y^{\bf i}$.

5704: This must be a string loop effect to be consistent with the

5705: tree level spectrum at the Gepner point which says that the gauge bosons

5706: are all massless.

5707: On the other hand, the tachyonic mass term of some charged open string fields 

5708: after deformation of complex structure

5709: must be at string tree level.

5710: How can these be consistent?

5711: To see this, let us be careful in the factors of

5712: $g_{\it st}\propto g^2$.

5713: The term (\ref{FIterm}) is correct provided that the gauge kinetic term

5714: is normalized as ${1\over g^2}(F_{\mu\nu})^2$ and that 

5715: the $y$-field is written as $y={c\over g^2}-i\vartheta$ so that

5716: the complex structure fields $c$ have the standard NSNS kinetic term

5717: ${1\over g_{\it st}^2}(\partial_{\mu}c)^2$.

5718: Then the FI parameter behaves as

5719: $\zeta={c\over g^2}$.

5720: Therefore the relevant terms in the effective Lagrangian depend on

5721: the open string coupling $g$ as follows

5722: $$

5723: -{1\over g^2}(F_{\mu\nu})^2-(A_{\mu}+\partial_{\mu}\vartheta)^2

5724: -|D_{\mu} Q_I|^2-{g^2\over 2}\left(

5725: \pm |Q_I|^2-{c\over g^2}\right)^2,

5726: $$

5727: where $Q_I$ are open string fields charged under the $U(1)$.

5728: We indeed see that the gauge boson mass is of open string one-loop level

5729: (at the vacuum with $c=0$), which is

5730: consistent with the tree level spectrum at the Gepner point.

5731: Also, we find that the (sometimes tachyonic) mass term

5732: for the charged open string fields is

5733: $\pm c |Q_I|^2$ which is indeed tree level.

5734: 

5735: 

5736: 

5737: 

5738: 

5739: If all the branes are invariant under the parity, the gauge group

5740: has no $U(1)$ factor and there is no room for FI term.

5741: Thus, in such a case, we do not expect the brane-recombination nor

5742: supersymmetry breaking as we move away from the Gepner point, or any

5743: supersymmetric point, as long as each brane remains parity invariant.

5744: This is indeed the case.

5745: To be specific, let us show this in the large volume limit

5746: (the same can be said near the Gepner point).

5747: The supersymmetry preserved by the brane $W$ is measured by the phase of the

5748: period integral $\int_W{\mit\Omega}$ where

5749: ${\mit\Omega}$ is the holomorphic 3-form of the Calabi-Yau manifold.

5750: The supersymmetry preserved by the O-plane $O_P$ is

5751: the phase of $\int_{O_P}{\mit\Omega}$.

5752: As we change the complex structure, these phases vary.

5753: We are considering the parity $P=\tau\Omega$ associated with

5754: the antiholomorphic involution $\tau$, and we have

5755: $$

5756: \tau^*{\mit\Omega}=\e^{i\theta_{\tau}}\overline{\mit\Omega}.

5757: $$

5758: If we use the invariance $\tau W=W$, $\tau O_P=O_P$,

5759: we find that the phases for $\int_W{\mit\Omega}$ and 

5760: $\int_{O_P}{\mit\Omega}$ are both $\e^{i\theta_{\tau}/2}$

5761: up to sign. But they have the same sign since

5762: we started with the point where the phases are the same.

5763: Thus, the phases of $\int_W{\mit\Omega}$ and 

5764: $\int_{O_P}{\mit\Omega}$ are always aligned.

5765: Therefore, as long as the branes are parity invariant, 

5766: they preserve the same supersymmetry as the O-plane,

5767: under any deformation of the complex structure compatible with the

5768: parity.

5769: 

5770: 

5771: 

5772: 

5773: What is said here can be repeated for Type IIB orientifolds:

5774: This time K\"ahler moduli enter into the FI terms,

5775: corresponding to the fact that the stability of B-branes is

5776: controlled by the K\"ahler moduli \cite{DFRI,DFRII,Dougcat}.

5777: In fact, the direct computation of the FI term is done in

5778: similar contexts in \cite{DM,poppitz}.

5779: See also \cite{Cvetic,Lalak,IRU2,IbaQue} for discussions.

5780: 

5781: 

5782: 

5783: 

5784: 

5785: 

5786: \section{Consistency Conditions and Supersymmetry --- B}

5787: \label{sec:TCCB}

5788: 

5789: 

5790: In this section,

5791: we write down the conditions of consistency and supersymmetry

5792: and count the number of

5793: solutions, for Type IIB orientifolds.

5794: We will follow the general strategy outlined in subsection

5795: \ref{subsec:general} and solve the tadpole constraints

5796: completely. We find, for example, that the IIB orientifold 

5797: of the Gepner model for quintic with respect to the parity 

5798: without exchange has one the order of 30 billion supersymmetric 

5799: and exactly solvable brane configurations.

5800: 

5801: 

5802: 

5803: 

5804: 

5805: 

5806: 

5807: 

5808: \subsection{Charge and Supersymmetry of O-plane}\label{sec:chSB}

5809: 

5810: The first step is to study the charge of the O-plane.

5811: To this end, it is useful to express it

5812: in terms of the charges of the B-branes which

5813: have been studied a lot in the past.

5814: 

5815: 

5816: Here again, mirror A-type picture is convenient.

5817: We know that the B-parity $P^B_{\omega;{\bf m}}$ is

5818: the mirror of the A-parity $P^A_{\tilomega;{\bf \tilm}}$

5819: in the model with the orbifold group $\tilGamma$ of order

5820: $H^{-1}\prod_i(k_i+2)$.

5821: Dressing by global symmetry ${\bf m}$ (resp. quantum symmetry

5822: $\omega$) corresponds to

5823: dressing by quantum symmetry $(\tilomega_i)$ (resp. global symmetry

5824: ${\bf \tilm}$): %(\ref{id1})-(\ref{id2}):

5825: $$\e^{-2\pi i{m_i\over k_i+2}}=\tilomega_i,

5826: \qquad

5827: \omega=\e^{2\pi i \sum_i{\tilm_i\over k_i+2}}=:

5828: \exp\left(2\pi i {M_{\omega}\over H}\right).

5829: $$

5830: We discuss

5831: the odd $H$ and even $H$ cases separately.

5832: 

5833: 

5834: \newcommand{\tilgamma}{\widetilde{\gamma}}

5835: 

5836: If $H$ is odd, we only have to consider the basic one

5837: $P^B$ without dressing --- dressing by global symmetry

5838: is not involutive and dressing by quantum symmetry is equivalent to

5839: no dressing.

5840: The structure of the crosscap state for the mirror

5841: A-parity $P^A$ is just like

5842: (\ref{Cexodd}), where the group $\Gamma$ is replaced by

5843: the mirror orbifold group $\tilGamma$:

5844: $$

5845: |\Scr{C}_{P^A}\rangle

5846: ={1\over \sqrt{|\tilGamma|}}\sum_{\tilgamma\in \tilGamma}\tilgamma

5847: |\Scr{C}_{{\bf P}^A}\rangle^{\rm prod}.

5848: $$

5849: This has the same structure as the sum-over-image 

5850: formula for the boundary state, and we know that

5851: $|\Scr{C}_{{\bf P}^A}\rangle^{\rm prod}$

5852: has the same RR-charge as

5853: the product brane $\Scr{B}_{{k_1-1\over 2},{k_1+1\over 2},1}\times

5854: \cdots\times\Scr{B}_{{k_r-1\over 2},{k_r+1\over 2},1}$.

5855: Thus, we find that $|\Scr{C}_{P^A}\rangle$ has the same charge as the

5856: the brane $\Scr{B}_{{\bf {k-1\over 2},{k+1\over 2}},1}$.

5857: Taking the mirror, we find that

5858: $|\Scr{C}_{P^B}\rangle$ has the same RR-charge as

5859: the B-brane $\Scr{B}_{{\bf k-1\over 2},H\sum_i{k_i+1\over 2(k_i+2)},1}$.

5860: Including the spacetime part, we find the following relation of RR-charges

5861: \beq

5862: \Bigl[O_{P^B}\Bigr]=4\Bigl[

5863: B^B_{{\bf k-1\over 2},H\sum_i{k_i-1\over 2(k_i+2)}}\Bigr]

5864: \label{OBchodd}

5865: \eeq

5866: 

5867: If $H$ is even, the structure of the 

5868: crosscap state for the mirror

5869: A-parity $P^A_{\tilomega;{\bf\tilm}}$ is analogous to

5870: (\ref{Cexeve}). As in that case, we classify

5871: the orbit of parity symmetries

5872: $\{\tilgamma{\bf P}^A_{\bf m}\}_{\tilgamma\in\tilGamma}$

5873: with respect to the subgroup

5874: $$

5875: \tilGamma^2=\{\tilgamma^2|\tilgamma\in \tilGamma\}\subset\tilGamma.

5876: $$

5877: This is a proper subgroup of $\tilGamma$ if $H$ is even (if $H$ is odd,

5878: this agrees with $\tilGamma$ and hence the orbit sum has a

5879: simple structure as we have discussed above).

5880: The orbit $\{\tilgamma{\bf P}^A_{\bf m}\}_{\tilgamma\in\tilGamma}$

5881: decomposes into blocks

5882: $\{\tilgamma^2

5883: {\bf P}^A_{\bf m+\tilnu}\}_{\tilgamma^2\in \tilGamma^2}$

5884: parametrized by the coset $\tilnu\in \tilGamma/\tilGamma^2$.

5885: Thus, the crosscap state has the following structure

5886: $$

5887: |\Scr{C}_{P^A_{\tilomega;{\bf \tilm}}}\rangle

5888: ={1\over\sqrt{|\tilGamma|}}

5889: \sum_{\tilgamma\in\tilGamma}\tilomega^{-\tilgamma}

5890: |\Scr{C}_{\tilgamma{\bf P}^A_{\bf \tilm}}\rangle

5891: =

5892: {1\over\sqrt{|\tilGamma|}}

5893: \sum_{\tilnu\in\tilGamma/\tilGamma^2}\tilomega^{-\tilnu}

5894: \sum_{\tilgamma^2\in \tilGamma^2}

5895: |\Scr{C}_{\tilgamma^2{\bf P}^A_{\bf \tilm+\tilnu}}\rangle

5896: $$

5897: where we have used the fact that $\tilomega_i=\pm 1$ and

5898: hence $\tilomega^{-\tilgamma^2}=1$.

5899: At this stage we use the relation

5900: $|\Scr{C}_{\tilgamma^2{\bf P}^A_{\bf \tilm+\tilnu}}\rangle

5901: =\tilgamma|\Scr{C}_{{\bf P}^A_{\bf \tilm+\tilnu}}\rangle$,

5902: and also replace the sum over $\tilgamma^2\in \tilGamma^2$

5903: by the sum over $\tilgamma\in \tilGamma$ times the ratio of the orders

5904: $|\tilGamma^2|/|\tilGamma|$:

5905: \beq

5906: |\Scr{C}_{P^A_{\tilomega;{\bf \tilm}}}\rangle

5907: ={1\over\sqrt{|\tilGamma|}}

5908: \sum_{\tilnu\in\tilGamma/\tilGamma^2}\tilomega^{-\tilnu}

5909: {|\tilGamma^2|\over |\tilGamma|}

5910: \sum_{\tilgamma\in \tilGamma}

5911: \tilgamma|\Scr{C}_{{\bf P}^A_{\bf \tilm+\tilnu}}\rangle

5912: =

5913: {1\over |\tilGamma/\tilGamma^2|}

5914: \sum_{\tilnu\in\tilGamma/\tilGamma^2}\tilomega^{-\tilnu}

5915: \left({1\over\sqrt{|\tilGamma|}}\sum_{\tilgamma\in \tilGamma}

5916: \tilgamma|\Scr{C}_{{\bf P}^A_{\bf \tilm+\tilnu}}\rangle\right).

5917: \eeq

5918: The expression in the parenthesis of the right hand side has the same

5919: structure as the sum-over-image formula for the boundary states.

5920: If $k_i$ are all even, this has the same RR-charge as

5921: the brane $\Scr{B}_{{\bf {k\over 2},{k+2\over 2}+\tilm+\tilnu}

5922: -\delta_{\bf \tilm+\tilnu},1}$ times the possible orientation flip

5923: $(-1)^{\sum_i{\tilnu_i\over k_i+2}}$. Bringing this mirror relation

5924: back into the original side and adding the spacetime part, we find

5925: %Kazuo: $\tilm$ inserted

5926: \beq

5927: \Bigl[O_{P^B_{\omega;{\bf m}}}\Bigr]

5928: ={4\over |\tilGamma/\tilGamma^2|}\sum_{\tilnu\in\tilGamma/\tilGamma^2}

5929: \tilomega^{-\tilm-\tilnu}(-1)^{\sum_i{\tilnu_i\over k_i+2}}

5930: \Bigl[B^B_{{\bf k\over 2},M_{{\bf \tilm}+\tilnu}}\Bigr]

5931: \label{OBcheve}

5932: \eeq

5933: where

5934: \beq

5935: M_{{\bf \tilm}+\tilnu}=

5936: H\sum_{i=1}^r{{k_i\over 2}+

5937: \tilm_i+\tilnu_i-\delta_{\tilm_i+\tilnu_i}\over k_i+2}.

5938: \eeq

5939: If there are both even and odd $k_i$, the expression is the obvious

5940: mixture of (\ref{OBcheve}) and (\ref{OBchodd}).

5941: An alternative approach to find the O-plane charge

5942: directly in the B-type picture will be

5943: outlined in section \ref{subsec:Dbranecharges}.

5944: 

5945: 

5946: 

5947: The next thing to find is

5948: the phase determining the spacetime supersymmetry

5949: preserved by the branes and the orientifold.

5950: For the branes, we find

5951: $$

5952: {}_{\RR}\langle 0|\Scr{B}_{{\bf L},M,S}\rangle_{\RR}

5953: =\e^{\pi i \sum_i({M_i\over k_i+2}-{S\over 2})}

5954: {}_{\NSNS}\langle 0|\Scr{B}_{{\bf L},M,S}\rangle_{\NSNS}

5955: $$

5956: and thus the phase is

5957: \beq

5958: \exp\Bigl(i\theta^B_{{\bf L},M}\Bigr)

5959: =i

5960: \exp\left(\pi i {M \over H}\right)

5961: \eeq

5962: For the crosscap, one can see that

5963: \beq

5964: {}_{\RR}\langle 0|\Scr{C}_{P^B_{\omega;{\bf m}}}\rangle

5965: =\tilomega\e^{\pi i \sum_i{\tilm_i-{1\over 2}\over k_i+2}}

5966: {}_{\NSNS}\langle 0|\Scr{C}_{\widetilde{P}^B_{\omega;{\bf m}}}\rangle,

5967: \eeq

5968: where $\tilm_i$ parametrizes the global symmetry in the mirror which is

5969: the quantum symmetry $\omega=\e^{2\pi i \sum_i{\tilm_i\over k_i+2}}$

5970: of the original side.

5971: We note here that the NSNS part of the total crosscap state

5972: has the factor $\tilomega^{1\over 2}$,

5973: see Eqn~(\ref{CBNStotal}).

5974: Thus, the ratio is

5975: \beq

5976: \exp\Bigl(i\theta^B_{\omega;{\bf m}}\Bigr)

5977: =-i\tilomega^{{1\over 2}}

5978: \exp\left(\pi i \sum_i{\tilm_i\over k_i+2}\right)

5979: %=-i^{-r}

5980: %\omega^{1\over 2}(\pm 1)\exp\left(\pi i \sum_{i=1}^r{m_i+{k_i-1\over 2}\over

5981: %k_i+2}\right).

5982: \eeq

5983: For completeness, we record here the expression of the O-plane tension;

5984: $$

5985: 4\tilomega^{1\over 2}

5986: {}_{\NSNS}\langle 0|\Scr{C}_{\widetilde{P}^B_{\omega,{\bf m}}}

5987: \rangle

5988: ={4\over\sqrt{H}}

5989: \prod_{i=1}^r\sqrt{2\over\sin({\pi\over k_i+2})}

5990: \prod_{k_i:\,{\rm odd}}

5991: \cos\Bigl({\pi\over 2(k_i\!+\!2)}\Bigr)\cdot

5992: {1\over |\tilGamma/\tilGamma^2|}

5993: \sum_{\tilnu\in \tilGamma/\tilGamma^2}

5994: \tilomega^{-\tilnu-{1\over 2}}

5995: \e^{i\Theta_{\bf\tilm+\tilnu}},

5996: $$

5997: $$\Theta_{\bf \tilm+\tilnu}=\sum_{k_i\,{\rm even}}

5998: {\pi (-1)^{\tilm_i+\tilnu_i}\over 2(k_i+2)}.

5999: $$

6000: 

6001: 

6002: \subsubsection{Example --- Quintic}\label{subsub:qex}

6003: 

6004: For the model $(k_i+2)=(5,5,5,5,5)$ corresponding to the quintic,

6005: the charge of the O-plane for the parity $P^B=P^B_{+;{\bf 0}}$

6006: is four times that of the B-brane

6007: $B_{{\bf L},M}^B$ with ${\bf L}=(1,1,1,1,1)$ and $M=5$.

6008: \beq

6009: \Bigl[O_{P^B}\Bigr]

6010: =4\Bigl[B_{{\bf 1},5}\Bigr].

6011: \label{qcross}

6012: \eeq

6013: The tension of the O-plane is

6014: $$

6015: 4\tilomega^{1\over 2}

6016: {}_{\NSNS}\langle 0|\Scr{C}_{\widetilde{P}^B}\rangle

6017: = {4\tilomega^{1\over 2}\over \sqrt{5}}

6018: \sqrt{2\over \sin({\pi\over 5})}^5\cos^5\left({\pi\over 10}\right);\qquad

6019: \tilomega^{1\over 2}=\pm 1.

6020: $$

6021: The phase determining the spacetime supersymmetry is

6022: \beq

6023: \e^{i\theta_{\omega;{\bf 0}}^B}=-i\tilomega^{{1\over 2}}

6024: \eeq

6025: while the one for the brane

6026: $B^B_{{\bf L},M}$ is $\e^{i\theta^B_{{\bf L},M}}

6027: =i\e^{\pi i M/5}$.

6028: For the orientifold with

6029: $\tilomega^{1\over 2}=-1$, branes preserving the same

6030: supersymmetry are $B^B_{{\bf L},M=0}$

6031: and $\overline{B^B_{{\bf L},M=5}}$. More explicitly, they are

6032: $B^B_{(00000),0}$, $\overline{B^B_{(10000),5}}$ and permutations,

6033: $B^B_{(11000),0}$ and permutations,

6034: $\overline{B^B_{(11100),5}}$ and permutations,

6035: $B^B_{(11110),5}$ and permutations, and

6036: $\overline{B^B_{(11111),5}}$. In total, there are

6037: $1+5+10+10+5+1=32$ of them.

6038: 

6039: 

6040: 

6041: \subsubsection{Example --- The Two Parameter Model}\label{sec:oplanec}

6042: 

6043: Let us consider the model $(k_i+2)=(8,8,4,4,4)$.

6044: The parity symmetries $P^B_{\omega;{\bf m}}$ are denoted as in

6045: Section~\ref{subsub:two} as

6046: $P^B_{\mu;\epsilon_1...\epsilon_5}$ where $\omega=\e^{2\pi i \mu/8}$ and

6047: $\epsilon_i=\e^{2\pi i{m_i\over k_i+2}}=\pm$.

6048: We only have to consider $P^B_{0;\epsilon_1...\epsilon_5}$

6049: and $P^B_{1;\epsilon_1...\epsilon_5}$ since

6050: others are related to these by symmetry conjugations.

6051: Also, there are eight inequivalent choices for $(\epsilon_1...\epsilon_5)$:

6052: $(+++++)$, $(++-++)$, $(++--+)$, $(++---)$,

6053: $(+-+++)$, $(+--++)$, $(+---+)$, $(+----)$.

6054: 

6055: The mirror orbifold group $\tilGamma$ is the set of

6056: $\tilnu=(\nu_1,\nu_2,\nu_3,\nu_4,\nu_5)\in

6057: \Z_8\times \Z_8\times\Z_4\times\Z_4\times\Z_4$ with

6058: ${\nu_1+\nu_2\over 8}+{\nu_3+\nu_4+\nu_5\over 4}\in \Z$.

6059: One may solve for $\nu_1$ as

6060: $\nu_1=-\nu_2-2(\nu_3+\nu_4+\nu_5)$, and thus

6061: the group is isomorphic to $\Z_8\times (\Z_4)^3$.

6062: This also shows that the element of $\tilGamma/\tilGamma^2$ is labeled

6063: by the mod 2 reduction of $(\nu_2,\nu_3,\nu_4,\nu_5)$ and hence

6064: $\tilGamma/\tilGamma^2\cong (\Z_2)^4$.

6065: 

6066: For parities $P^B_{0;\epsilon_1...\epsilon_5}$

6067: without dressing by quantum symmetry,

6068: we have ${\bf\tilm}={\bf 0}$.

6069: We find

6070:  $M_{\bf\tilm+\tilnu}=M_{\tilnu}=12-2(\nu_2+\nu_3+\nu_4+\nu_5)$

6071: if $\nu_2,...,\nu_5$ are assumed to take values in $\{0,1\}$.

6072: The charge and the tension of the O-plane

6073: for the eight cases are summarized in the table \ref{Otwopar1}.

6074: \begin{table}[htb]

6075: \begin{center}

6076: \renewcommand{\arraystretch}{1.2}

6077: \begin{tabular}{|l||l|l|}

6078: \hline

6079: parity&RR-charge&Tension\\

6080: \noalign{\hrule height 0.8pt}

6081: $P^B_{0;+++++}$

6082: &

6083: ${1\over 4}\left([B_{{\bf k\over 2},12}]

6084: +4[B_{{\bf k\over 2},10}]

6085: +6[B_{{\bf k\over 2},8}]

6086: +4[B_{{\bf k\over 2},6}]

6087: +[B_{{\bf k\over 2},4}]\right)$

6088: &$\begin{array}{l}

6089: \\

6090: \\

6091: \end{array}

6092: \!\!\!\!

6093: 4\tilomega^{-{1\over 2}}{3+2\sqrt{2}\over \sqrt{2\sqrt{2}-2}}$

6094: \\

6095: \hline

6096: \multicolumn{1}{|l||}{$P^B_{0;++-++}$}&

6097: \lw{$

6098: {1\over 4}\left([B_{{\bf k\over 2},12}]

6099: +2[B_{{\bf k\over 2},10}]

6100: -2[B_{{\bf k\over 2},6}]

6101: -[B_{{\bf k\over 2},4}]\right)

6102: $}&\lw{$4i\tilomega^{-{1\over 2}}{1+\sqrt{2}\over \sqrt{2\sqrt{2}-2}}

6103: $}\\

6104: \cline{1-1}

6105: $P^B_{0;+-+++}$&&\\

6106: \hline

6107: \multicolumn{1}{|l||}{$P^B_{0;++--+}$}&

6108: \lw{$

6109: {1\over 4}\left([B_{{\bf k\over 2},12}]

6110: -2[B_{{\bf k\over 2},8}]

6111: +[B_{{\bf k\over 2},4}]\right)

6112: $}&\lw{$

6113: -4\tilomega^{-{1\over 2}}{1\over \sqrt{2\sqrt{2}-2}}

6114: $}\\

6115: \cline{1-1}

6116: $P^B_{0;+--++}$&&\\

6117: \hline

6118: \multicolumn{1}{|l||}{$P^B_{0;++---}$}&

6119: \lw{$

6120: {1\over 4}\left([B_{{\bf k\over 2},12}]

6121: -2[B_{{\bf k\over 2},10}]

6122: +2[B_{{\bf k\over 2},6}]

6123: -[B_{{\bf k\over 2},4}]\right)

6124: $}&\lw{$

6125: -4i\tilomega^{-{1\over 2}}{\sqrt{2}-1\over \sqrt{2\sqrt{2}-2}}

6126: $}\\

6127: \cline{1-1}

6128: $P^B_{0;+---+}$&&\\

6129: \hline

6130: $P^B_{0;+----}$&

6131: $

6132: {1\over 4}\left([B_{{\bf k\over 2},12}]

6133: -4[B_{{\bf k\over 2},10}]

6134: +6[B_{{\bf k\over 2},8}]

6135: -4[B_{{\bf k\over 2},6}]

6136: +[B_{{\bf k\over 2},4}]\right)

6137: $

6138: &$\begin{array}{l}

6139: \\

6140: \\

6141: \end{array}

6142: \!\!\!\!

6143: 4\tilomega^{-{1\over 2}}{3-2\sqrt{2}\over \sqrt{2\sqrt{2}-2}}

6144: $

6145: \\

6146: \hline

6147: \end{tabular}

6148: \caption{Charge and Tension of the O-plane ($\omega=1$)}

6149: \label{Otwopar1}

6150: \end{center}

6151: \end{table}

6152: The spacetime supersymmetry preserved by the orientifold is

6153: \beq

6154: \e^{i\theta_O}=-i\tilomega^{1\over 2},

6155: \eeq

6156: where we note that $\tilomega=\epsilon_1\cdots\epsilon_5$.

6157: Branes preserving the same supersymmetries are 

6158: $B_{{\bf L},8}$,$\overline{B_{{\bf L},0}}$ for

6159: $\tilomega^{1\over 2}=1$;

6160: $B_{{\bf L},0}$,$\overline{B_{{\bf L},8}}$ for

6161: $\tilomega^{1\over 2}=-1$;

6162: $B_{{\bf L},12}$,$\overline{B_{{\bf L},4}}$ for

6163: $\tilomega^{1\over 2}=i$; and

6164: $B_{{\bf L},4}$,$\overline{B_{{\bf L},12}}$ for

6165: $\tilomega^{1\over 2}=-i$.

6166: 

6167: 

6168: For parities $P^B_{1;\epsilon_1...\epsilon_5}$

6169: dressed by the quantum symmetry $\omega=\e^{2\pi i/8}$,

6170: we have ${\bf\tilm}=(1,0,0,0,0)$.

6171: We find

6172:  $M_{\bf\tilm+\tilnu}=12-2(\nu_3+\nu_4+\nu_5)$

6173: if $\nu_2,...,\nu_5$ are assumed to take values in $\{0,1\}$.

6174: The charge and the tension of the O-plane

6175: for the eight cases are summarized in the table \ref{Otwopar2}.

6176: \begin{table}[htb]

6177: \begin{center}

6178: \renewcommand{\arraystretch}{1.2}

6179: \begin{tabular}{|l||l|l|}

6180: \hline

6181: parity&RR-charge&Tension

6182: \\

6183: \noalign{\hrule height 0.8pt}

6184: $P^B_{1;+-***}$

6185: &$0$

6186: &$\begin{array}{l}

6187: \\

6188: \\

6189: \end{array}

6190: \!\!\!\!

6191: 0$

6192: \\

6193: \hline

6194: $P^B_{1;+++++}$

6195: &$

6196: {1\over 2}\left([B_{{\bf k\over 2},12}]

6197: +3[B_{{\bf k\over 2},10}]

6198: +3[B_{{\bf k\over 2},8}]

6199: +[B_{{\bf k\over 2},6}]\right)

6200: $

6201: &$\begin{array}{l}

6202: \\

6203: \\

6204: \end{array}

6205: \!\!\!\!

6206: 4\tilomega^{-{1\over 2}}\sqrt{10+7\sqrt{2}\over \sqrt{2}-1}

6207: $

6208: \\

6209: \hline

6210: $P^B_{1;++-++}$

6211: &$

6212: {1\over 2}\left([B_{{\bf k\over 2},12}]

6213: +[B_{{\bf k\over 2},10}]

6214: -[B_{{\bf k\over 2},8}]

6215: -[B_{{\bf k\over 2},6}]\right)

6216: $

6217: &$\begin{array}{l}

6218: \\

6219: \\

6220: \end{array}

6221: \!\!\!\!

6222: 4i\tilomega^{-{1\over 2}}\sqrt{2+\sqrt{2}\over \sqrt{2}-1}

6223: $

6224: \\

6225: \hline

6226: $P^B_{1;++--+}$

6227: &$

6228: {1\over 2}\left([B_{{\bf k\over 2},12}]

6229: -[B_{{\bf k\over 2},10}]

6230: -[B_{{\bf k\over 2},8}]

6231: +[B_{{\bf k\over 2},6}]\right)

6232: $

6233: &$\begin{array}{l}

6234: \\

6235: \\

6236: \end{array}

6237: \!\!\!\!\!\!

6238: -4\tilomega^{-{1\over 2}}\sqrt{2-\sqrt{2}\over \sqrt{2}-1}

6239: $

6240: \\

6241: \hline

6242: $P^B_{1;++---}$

6243: &$

6244: {1\over 2}\left([B_{{\bf k\over 2},12}]

6245: -3[B_{{\bf k\over 2},10}]

6246: +3[B_{{\bf k\over 2},8}]

6247: -[B_{{\bf k\over 2},6}]\right)

6248: $

6249: &$\begin{array}{l}

6250: \\

6251: \\

6252: \end{array}

6253: \!\!\!\!\!\!

6254: -4i\tilomega^{-{1\over 2}}\sqrt{10-7\sqrt{2}\over \sqrt{2}-1}

6255: $

6256: \\

6257: \hline

6258: \end{tabular}

6259: \caption{Charge and Tension of the O-plane ($\omega=\e^{2\pi i/8}$)}

6260: \label{Otwopar2}

6261: \end{center}

6262: \end{table}

6263: The spacetime supersymmetry preserved by the orientifold is

6264: \beq

6265: \e^{i\theta_O}=-i\tilomega^{1\over 2}\exp\left({\pi i\over 8}\right).

6266: \eeq

6267: Branes preserving the same supersymmetries are 

6268: $B_{{\bf L},9}$,$\overline{B_{{\bf L},1}}$ for

6269: $\tilomega^{1\over 2}=1$;

6270: $B_{{\bf L},1}$,$\overline{B_{{\bf L},9}}$ for

6271: $\tilomega^{1\over 2}=-1$;

6272: $B_{{\bf L},13}$,$\overline{B_{{\bf L},5}}$ for

6273: $\tilomega^{1\over 2}=i$; and

6274: $B_{{\bf L},5}$,$\overline{B_{{\bf L},13}}$ for

6275: $\tilomega^{1\over 2}=-i$.

6276: 

6277: 

6278: 

6279: \subsection{D-branes in the Orientifold Models}

6280: 

6281: \subsubsection{Parity Action on D-branes}

6282: 

6283: 

6284: \newcommand{\bfS}{{\bf S}}

6285: 

6286: Let us now find how the B-type orientifold acts on the B-branes.

6287: We first consider long-orbit branes. 

6288: To see the action, we compare the $\langle B|C\rangle$

6289: and $\langle C|B\rangle$ M\"obius strips.

6290: We find

6291: \beqa

6292: &&{}_{\RR}\langle C^B_{\omega;{\bf m}}|q_t^H|B^B_{{\bf L},M}\rangle_{\RR}

6293: =\tilomega^{-1}\times {}_{\RR}\langle

6294: B^B_{{\bf L},2M_{\omega}-M}|q_t^H

6295: |C^B_{\omega;{\bf m}}\rangle_{\RR},\\

6296: &&{}_{\NSNS}\langle C^B_{\omega;{\bf m}}|q_t^H|

6297: B^B_{{\bf L},M}\rangle_{\NSNS}

6298: =-{}_{\NSNS}\langle

6299: B^B_{{\bf L},2M_{\omega}-M}|q_t^H

6300: |C^B_{\omega;{\bf m}}\rangle_{\NSNS}.

6301: \label{seconB}

6302: \eeqa

6303: This can again be shown 

6304: using the mirror description.

6305: Thus, the parity acts on the branes as

6306: \beq

6307: P^B_{\omega;{\bf m}}:\,B^B_{{\bf L},M}

6308: \longmapsto

6309: \tilomega^{-1}

6310: B^B_{{\bf L},2M_{\omega}-M},

6311: \label{PBonBb}

6312: \eeq

6313: where we recall that

6314: $\tilomega^{-1}=\e^{2\pi i \sum_i{m_i\over k_i+2}}$

6315: and $\e^{2\pi i M_{\omega}/H}=\omega$.

6316:  

6317: Let us now consider short-orbit branes.

6318: We denote by $\bfS$ the set of $i$ such that $L_i={k_i\over 2}$

6319: If the number of elements $|\bfS|$ is odd, there is no difference

6320: from the above result. Thus we focus on the branes

6321:  $\widehat{B}^{(\varepsilon)}$ with even $|\bfS|$.

6322: The action on the $({\bf L},M)$-label is the same as above, and

6323: the difference appears in the action on

6324: the $\varepsilon$-label.

6325: We find that the result is

6326: \beq

6327: P^B_{\omega;{\bf m}}:\varepsilon\longmapsto \varepsilon'

6328: =(-1)^{|\bfS|\over 2}\prod_{i\in\bfS}\widetilde{\omega}_i^{\frac{k_i+2}{2}}

6329: \cdot \varepsilon.

6330: \label{PBonBshort}

6331: \eeq

6332: 

6333: 

6334: 

6335: \subsubsection{Invariant Branes}

6336: 

6337: Let us find out which of the B-branes are invariant under the

6338: orientifold action.

6339: By (\ref{PBonBb}), the condition is

6340: $B^B_{{\bf L},2M_{\omega}-M}=B^B_{{\bf L},M}$.

6341: Here it is useful to note the ``brane identification'':

6342: $B^B_{{\bf L}',M'}=B^B_{{\bf L},M}$ if and only if 

6343: $M'=M$ and $L_i'=L_i$ except for even number of

6344: $i$'s with $L_i'=k_i-L_i$.

6345: Also, $B^B_{{\bf L}',M'}=\overline{B^B_{{\bf L},M}}$

6346: if and only if $M'=M+H$ and $L_i'=L_i$ except for odd number of $i$'s with

6347: $L_i'=k_i-L_i$. Using this we find that invariant branes are

6348: \beqa

6349: \tilomega=1:&B^B_{{\bf L},M_{\omega}},B^B_{{\bf L},M_{\omega}+H},

6350: \,\,\mbox{${\bf L}$ arbitrary},

6351: \\

6352: \tilomega=-1:&B^B_{{\bf L},M_{\omega}\pm {H\over 2}},\,\,\,L_i={k_i\over 2}\,\,

6353: \mbox{for a single $i$}.

6354: \eeqa

6355: This applies also to short orbit branes with odd $|\bfS|$.

6356: 

6357: For short-orbit branes $\widehat{B}^{(\varepsilon)}_{{\bf L},M}$

6358: with even $|\bfS|$, this is modified because of the

6359: new type of ``Brane identification'' where

6360: $M\to M+H$ does the flip of $\varepsilon$ as well as the orientation.

6361: The invariant branes are those with

6362: $M=M_{\omega}$ (mod $H$) if $\tilomega=1$

6363: and $M=M_{\omega}+{H\over 2}$ (mod $H$) if $\tilomega=-1$, just as above

6364: but there is an extra condition

6365: on the number $|\bfS|$:

6366: \beq

6367: (-1)^{|\bfS|\over 2}

6368: =\tilomega

6369:   \prod_{i\in\bfS}\tilomega_i^{\frac{k_i+2}{2}}.

6370: \eeq

6371: 

6372: 

6373: 

6374: 

6375: 

6376: 

6377: \subsubsection{Structure of Chan-Paton Factor}

6378: 

6379: 

6380: 

6381: Let us find the gauge group supported by $N$ of the

6382: invariant D-branes by computing

6383:  the $\langle B|C\rangle$ overlap in the NSNS sector.

6384: The computation can be done most easily in the mirror picture,

6385: but we have to be careful for the factor $\tilomega^{-\tilm}$

6386: appeared in (\ref{BvsAtil}).

6387: Using the formula

6388: (\ref{PonGSmin}) for the minimal model and the ones for the universal sector,

6389: we find (up to the standard factor)

6390: $$

6391: \langle B^B_{{\bf L},M}|q_t^H|C^B_{\omega;{\bf m}}\rangle_{\NSNS}

6392: =\overline{\epsilon_{{\bf L},M}^{\omega;{\bf m}}}

6393: \prod_{n=1}^{\infty}(1-i(-1)^nq_l^{n-{1\over 2}})^2

6394: -\epsilon_{{\bf L},M}^{\omega;{\bf m}}

6395: \prod_{n=1}^{\infty}(1+i(-1)^nq_l^{n-{1\over 2}})^2

6396: +\cdots

6397: $$

6398: where $\epsilon_{{\bf L},M}^{\omega;{\bf m}}$

6399: is given as follows;

6400: \beqa

6401: \epsilon_{{\bf L},M}^{\omega;{\bf m}}

6402: &=&\e^{\pi i\over 4}

6403: \sum_{\tilnu\in \tilGamma}

6404: \tilomega^{\tilm+\tilnu+{1\over 2}}

6405: \prod_i\left(

6406: \e^{-{\pi i\over 4}}\delta_{M_i,\tilm_i+\tilnu_i}

6407: +\e^{\pi i\over 4}\delta_{L_i,{k_i\over 2}}

6408: \delta_{M_i,\tilm_i+\tilnu_i+{k_i+2\over 2}}\right)

6409: \nn\\

6410: &=&

6411: -\tilomega^{{\bf M}+{1\over 2}}

6412: \sum_{\tilnu\in \tilGamma}

6413: \prod_{L_i\ne {k_i\over 2}}\delta_{M_i,\tilm_i+\tilnu_i}

6414: \prod_{L_i={k_i\over 2}}\left(\delta_{M_i,\tilm_i+\tilnu_i}

6415: +i\tilomega_i^{k_i+2\over 2}

6416: \delta_{M_i,\tilm_i+\tilnu_i+{k_i+2\over 2}}\right)

6417: \nn\\

6418: &=&-\tilomega^{{\bf L}+{1\over 2}}

6419: \sum_{p_i\in\{0,1\}}\delta^{(H)}_{M,M_{\omega}+{\sum_ip_i\over 2}H}

6420: \prod_{L_i={k_i\over 2}}\left(i\tilomega_i^{k_i+2\over 2}\right)^{p_i}.

6421: \eeqa

6422: This is indeed a sign factor for the long-orbit branes

6423: for which $L_i={k_i\over 2}$ at most for one $i$, and

6424: $M\equiv M_{\omega}$ (mod $H$) if $\tilomega=1$

6425: and $M\equiv M_{\omega}+{H\over 2}$ (mod $H$) if $\tilomega=-1$.

6426: More concretely,

6427: \beq

6428: \epsilon_{{\bf L},M}^{\omega;{\bf m}}

6429: =\left\{

6430: \begin{array}{ll}

6431: -\tilomega^{{\bf L}+{1\over 2}}&\tilomega=1

6432: \\

6433: -i\tilomega^{{\bf L}+{1\over 2}}\tilomega_{i_*}^{k_{i_*}+2\over 2}

6434: &\tilomega=-1.

6435: \end{array}\right.

6436: \label{Ggroup}

6437: \eeq

6438: where $i_*$ is the one that has $L_{i_*}={k_{i_*}\over 2}$.

6439: The invariant branes with $\epsilon_{{\bf L},M}^{\omega;{\bf m}}=1$

6440: or $-1$ support the $O$ or $Sp$-type gauge symmetries.

6441: 

6442: One can do the same computation for short-orbit branes

6443: satisfying $L_i=\frac{k_i}{2}$ for $i\in \bfS~ (|\bfS|\ge 2)$.

6444: Taking into account the correct normalization factor, one finds

6445: \beqa

6446: \epsilon_{{\bf L},M}^{\omega;{\bf m}}

6447: &=&

6448:  -2^{-\left[|\bfS|/2\right]}\tilomega^{{\bf L}+{1\over 2}}

6449: \sum_{p_i\in\{0,1\}}\delta^{(H)}_{M,M_{\omega}+{\sum_ip_i\over 2}H}

6450: \prod_{i\in\bfS}\left(i\tilomega_i^{k_i+2\over 2}\right)^{p_i}.

6451: \nn \\ &=&

6452:  -{\rm Re}\left(2^{-\left[|\bfS|/2\right]}\tilomega^{{\bf L}+{1\over 2}}

6453:  \prod_{i\in\bfS}(1+i\tilomega_i^{k_i+2\over 2})\right).

6454: \label{Ggroup2}

6455: \eeqa

6456: This is indeed a sign factor again: for odd $|\bfS|$ the quantity in the large

6457: parenthesis is of the form $(\pm1\!\pm'\!i)$, and for even $|\bfS|$

6458: it squares to

6459: \beq

6460:   \tilomega(-1)^{\frac{|\bfS|}{2}}\prod_{i\in\bfS}\tilomega_i^{k_i+2\over 2}

6461: \eeq

6462: which is unity for parity-invariant short-orbit branes.

6463: 

6464: 

6465: 

6466: \subsubsection{Examples}

6467: 

6468: {\bf Quintic}

6469: 

6470: The branes in the $(5,5,5,5,5)$-model are transformed by the B-parity $P^B$

6471: as $B^B_{{\bf L},M}\longmapsto B^B_{{\bf L},-M}$. Invariant branes are those with

6472: $M=0$ and $M=5$. All of them support $O(N)$ (resp. $Sp(N/2)$)

6473: gauge group for the choice

6474: $\tilomega^{1\over 2}=-1$ (resp. $\tilomega^{1\over 2}=1$).

6475: 

6476: 

6477: \bigskip

6478: \noindent

6479: {\bf The Two Parameter Model}

6480: 

6481: The branes in the $(8,8,4,4,4)$-model are transformed by B-parities as

6482: \beqa

6483: &&P^B_{0;\epsilon_1...\epsilon_5}:

6484: B_{{\bf L},M}\longmapsto \epsilon_1\cdots\epsilon_5 B_{{\bf L},-M},\nn\\

6485: &&P^B_{1;\epsilon_1...\epsilon_5}:

6486: B_{{\bf L},M}\longmapsto \epsilon_1\cdots\epsilon_5 B_{{\bf L},2-M}.\nn

6487: \eeqa

6488: Invariant branes are

6489: \beqa

6490: &&P^B_{0;\epsilon_1...\epsilon_5},\,\epsilon_1\cdots\epsilon_5=1:

6491: B^B_{{\bf L},0},B^B_{{\bf L},8};

6492: \nn\\

6493: &&P^B_{0;\epsilon_1...\epsilon_5},\,\epsilon_1\cdots\epsilon_5=-1:

6494: B^B_{{\bf L^*},4},B^B_{{\bf L^*},12};

6495: \nn\\

6496: &&P^B_{1;\epsilon_1...\epsilon_5},\,\epsilon_1\cdots\epsilon_5=1:

6497: B^B_{{\bf L},1},B^B_{{\bf L},9};

6498: \nn\\

6499: &&P^B_{1;\epsilon_1...\epsilon_5},\,\epsilon_1\cdots\epsilon_5=-1:

6500: B^B_{{\bf L^*},5},B^B_{{\bf L^*},13}.

6501: \nn

6502: \eeqa

6503: Here, ${\bf L^*}$ is such that $L_i={k_i\over 2}$ for a single $i$.

6504: Invariant short-orbit branes with even $|\bfS|$ also satisfy

6505: $(-1)^{\frac{|\bfS|}{2}}=\epsilon_1\cdots\epsilon_5$.

6506: The gauge group depends on $\prod_i\epsilon_i^{L_i}$ as in

6507: (\ref{Ggroup}) and (\ref{Ggroup2}).

6508: 

6509: 

6510: 

6511: 

6512: 

6513: 

6514: 

6515: 

6516: 

6517: 

6518: \subsection{D-brane charges}\label{subsec:Dbranecharges}

6519: 

6520: 

6521: \def\II{\bf S}

6522: Recall: Rational branes in the Gepner model 

6523: $(k_1,k_2,\ldots,k_r)$ with $H={\rm lcm}\{k_i+2\}$

6524: are in bijective correspondence with the following labeling system.

6525: We need:\\

6526: (i) A label ${\bf L}=(L_1,\ldots,L_r)$ with $0\leq L_i \leq k_i/2$.

6527: We denote by $\II$ the set of $i$ for which $L_i=k_i/2$. \\

6528: (ii) A label $M\in\ZZ_{2H}$ with $M=\sum w_i L_i\bmod 2$.

6529: ($w_i=H/(k_i+2)$)\\

6530: (iii) If $d+r$ is even AND all $L_i<k_i/2$ (ie, $\II=\emptyset$),

6531: a label $S=0,2$.\\

6532: (iv) If $\II\neq\emptyset$ AND $d+r+|\II|$ is even,

6533: a label $\psi=\pm$.\\

6534: In this paper, we denote such a brane by $B^B_{{\bf L},M,S,\psi}$,

6535: where it is understood that $S$ and $\psi$ can be omitted or

6536: neglected if they are unnecessary. Again, we emphasize that any 

6537: brane has a unique label of this type and that any label of this 

6538: type uniquely specifies a brane. This labeling system assigns

6539: different labels to a brane and its antibrane, but we will 

6540: sometimes take the freedom to indicate the antibrane with a

6541: minus-sign. For branes with an $S=0$ label, the antibrane has

6542: $S=2$, while for branes without an $S$ label, the antibrane

6543: is obtained by sending $M\mapsto M+H$. The spacetime supersymmetry 

6544: preserved by such a brane is a phase given by $\e^{\pi i (M/H+S/2)}$.

6545: 

6546: To summarize the RR charge of these branes, we find it convenient

6547: to introduce generators for the charge lattice. For B-type branes,

6548: such a generating set is conveniently obtained from the charges of 

6549: the $H$ branes with ${\bf L}= {\bf 0}$, and $M=0,2,\ldots, 2H-2$. 

6550: The relations they satisfy can be understood quite easily from 

6551: divisibility properties of the weights and we will make this more 

6552: explicit below. We will denote the linear operator mapping the $H$ 

6553: ${\bf L}={\bf 0}$ branes onto a linearly independent set generically 

6554: by $T$. To expand the charges of the other branes in terms of the 

6555: ${\bf L}={\bf 0}$ ones, we shall use as before brackets $[\Scr B]$ 

6556: to denote the RR charge vector of a brane $\Scr B$. Neglecting for 

6557: a very short moment the $S$ and $\psi$ labels, we denote by 

6558: $([B^B_{{\bf L},M}])$ the $\ZZ_H$ orbit of rational branes with 

6559: definite ${\bf L}$ label and $M$ label running over $M,M+2,\ldots, 

6560: (M+2H-2)(\bmod 2H)$. We can then write

6561: \beq

6562: \left(\bigl[ B^B_{{\bf L},M}\bigr]\right) =

6563: \left(\bigl[B^B_{{\bf 0},0}\bigr]\right)

6564: Q_{{\bf L},M}(g)

6565: \label{run}

6566: \eeq

6567: where the $H\times H$-dimensional matrix $Q_{{\bf L},M}(g)$ is a simple

6568: polynomial expression in the ``shift generator'' $g$, which is 

6569: the matrix with entry $1$ on the first lower diagonal and in the 

6570: upper right corner, and zeros elsewhere. Explicitly

6571: \beq

6572: Q_{{\bf L},M}(g) = g^{M/2} \prod_{i=1}^r \Bigl( \sum_{k_i=0}^{L_i} 

6573: \bigl(g_i\bigr)^{L_i/2-k_i}\Bigr) \,,

6574: \label{branec}

6575: \eeq

6576: where $g_i=g^{w_i}$ and $w_i=H/(k_i+2)$. Even more explicitly,

6577: the components of the charge vector of the $M$-th brane on the

6578: orbit (\ref{run}) are given by the $M$-th column of the matrix

6579: (\ref{branec}).

6580: 

6581: If the brane carries an $S$ label, the formula (\ref{branec}) is 

6582: simply multiplied by $(-1)^{S/2}$. If the brane has ${\bf S}\neq

6583: \emptyset$, the formula (\ref{branec}) gets corrected by a fixed 

6584: point resolution factor $f=1/2^{[\nu/2]}$, where $\nu={\bf S}+1$ if 

6585: $d+r$ is odd and $\nu=|{\bf S}|$ if $d+r$ is even (see subsection

6586: \ref{sub:Cardybranes}). We realize that this factor gets some time 

6587: to get accustomed to, so we write it out explicitly

6588: for the canonical case $r=5$, $d=1$, all levels even, such as

6589: our two parameter model. Then if the number of $L_i$ which 

6590: are equal to $k_i/2$ is $1,2,3,4,5$, we have $f=1,2,2,4,4$, 

6591: respectively. 

6592: Let us also note here that in can happen under exceptional 

6593: circumstance that the brane charge depends on the $\psi$

6594: label, because the non-toric K\"ahler parameters sit in the

6595: $\Z_{H/2}$ twisted sector. (An example of this is the Gepner 

6596: model $(k_i)=(2,2,4,4,4)$.)

6597: 

6598: 

6599: 

6600: 

6601: 

6602: In Section~\ref{sec:chSB}, we obtained the RR charge of

6603: the O-plane using the mirror picture and expressed it in terms of

6604: the D-brane charge.

6605: Here we comment on an alternative

6606: way, directly in the B-type picture, to find and express it

6607: in terms of the generators of the charge lattice introduced in this

6608: section.

6609: We look for a stack of ${\bf L=0}$ branes that have

6610: the same intersection numbers with

6611: any other set of branes as the orientifold. Since the ${\bf L=0}$ branes

6612: form a basis of the charge lattice, it is sufficient to check this

6613: for a general configuration of ${\bf L=0}$ branes. So all we have to

6614: do is to solve the linear system ``intersection matrix of the

6615: ${\bf L}={\bf 0}$ branes times a charge vector equals 

6616: the twisted Witten indices of the ${\bf L=0}$ branes with the orientifold

6617: plane''.

6618: The charge vector of this linear system yields the orientifold

6619: charge in terms of the ${\bf L}={\bf 0}$ branes. 

6620: 

6621: 

6622: 

6623: We are now ready to explicitly write down the tadpole cancellation

6624: conditions and find supersymmetric solutions at the Gepner point.

6625: 

6626: 

6627: 

6628: \subsection{Solutions of the Tadpole Conditions --- Quintic Case}

6629: 

6630: The problem simplifies somewhat for the case of the quintic because

6631: there is only a single possible parity and because all branes

6632: preserving the same spacetime supersymmetry as the O-plane are

6633: invariant under the parity. As explained above, we will study

6634: tadpole cancellation using the ${\bf L}={\bf 0}$ RS branes as a

6635: ``basis'' for the charge lattice. The charges of these branes

6636: satisfy one linear relation

6637: \beq

6638: \bigl[B^B_{{\bf 0},0}\bigr]+

6639: \bigl[B^B_{{\bf 0},2}\bigr]+

6640: \bigl[B^B_{{\bf 0},4}\bigr]+

6641: \bigl[B^B_{{\bf 0},6}\bigr]+

6642: \bigl[B^B_{{\bf 0},8}\bigr] =0  \,.

6643: \label{quinrel}

6644: \eeq

6645: In conjunction with the invariance of the tadpole canceling brane 

6646: configuration under the parity this linear

6647: relation implies that the equation (\ref{tadcan}) will reduce to

6648: two linearly independent equations on the $n_i$.

6649: 

6650: \subsubsection{O-plane charge}

6651: 

6652: The charge of the O-plane is given in (\ref{qcross}) as

6653: $\bigl[O_{P^B}\bigr] = 4\bigl[B_{{\bf 1},5}\bigr]$

6654: and thus is expressed in terms of ${\bf L=0}$ branes as

6655: \beq

6656: \bigl[O_{P^B}\bigr]

6657: =\left(

6658: \bigl[B^B_{{\bf 0},0}\bigr],

6659: \bigl[B^B_{{\bf 0},2}\bigr],

6660: \bigl[B^B_{{\bf 0},4}\bigr],

6661: \bigl[B^B_{{\bf 0},6}\bigr],

6662: \bigl[B^B_{{\bf 0},8}\bigr]

6663: \right) \cdot

6664: \left(

6665: \begin{array}{c}

6666: 8 \\

6667: 20 \\

6668: 40\\

6669: 40\\

6670: 20

6671: \end{array}

6672: \right)

6673: \eeq

6674: 

6675: \subsubsection{Supersymmetry preserving branes}

6676:  

6677: As studied in Section~\ref{subsub:qex},

6678: for any given parity, the set 

6679: of rational branes contains $32$ branes preserving the same 

6680: supersymmetry as the O-planes, and these $32$ branes have $6$ 

6681: different charges. Representatives of these $6$ charges are

6682: the branes

6683: \beq

6684: \bigl[B^B_{(00000),0} \Bigr],

6685: \bigl[\overline{B^B_{(10000),5}} \bigr],

6686: \bigl[B^B_{(11000),0} \bigr],

6687: \bigl[\overline{B^B_{(11100),5}} \bigr],

6688: \bigl[B^B_{(11110),5} \bigr] ,

6689: \bigl[\overline{B^B_{(11111),5}} \bigr] .

6690: \label{qlist}

6691: \eeq

6692: The other branes are obtained by permuting the ${\bf L}$ label,

6693: leading to multiplicities $m_i=1,5,10,10,5,1$ for the $i$-th

6694: charge, respectively.  To simplify the enumeration of solutions,

6695: we will then consider tadpole canceling brane configurations 

6696: containing $n_i$ branes with charge of each type and subsequently 

6697: multiply by the combinatorial factor $\left(\!\!\begin{array}{c} 

6698: n_i+m_i-1\\[-.2cm] n_i\end{array}\!\!\right)$ counting the number 

6699: of ways of distributing the charge.

6700: 

6701: Using the formulas of the previous subsection, we obtain the following 

6702: expression for these $6$ charges in terms of those of the

6703: ${\bf L}={\bf 0}$ branes.

6704: \beqa

6705: &&\left(

6706: \bigl[B^B_{(00000),0} \Bigr],

6707: \bigl[\overline{B^B_{(10000),5}} \bigr],

6708: \bigl[B^B_{(11000),0} \bigr],

6709: \bigl[\overline{B^B_{(11100),5}} \bigr],

6710: \bigl[B^B_{(11110),5} \bigr] ,

6711: \bigl[\overline{B^B_{(11111),5}} \bigr] 

6712: \right) \qquad\qquad

6713: \nn\\

6714: &&\,\,\,\,\,\,\qquad=\left(

6715: \bigl[B^B_{{\bf 0},0}\bigr],

6716: \bigl[B^B_{{\bf 0},2}\bigr],

6717: \bigl[B^B_{{\bf 0},4}\bigr],

6718: \bigl[B^B_{{\bf 0},6}\bigr],

6719: \bigl[B^B_{{\bf 0},8}\bigr] 

6720: \right) 

6721: Q

6722: \eeqa

6723: where $Q$ is the matrix

6724: \beq

6725: Q= \left(

6726: \begin{array}{cccccc}

6727: 1&0&2&0&6&-2\\

6728: 0&0&1&-1&4&-5\\

6729: 0&-1&0&-3&1&-10\\

6730: 0&-1&0&-3&1&-10\\

6731: 0&0&1&-1&4&-5

6732: \end{array}

6733: \right)\,.

6734: \eeq

6735: We can take the linear relation (\ref{quinrel}) into account by

6736: multiplying from the left with the matrix

6737: \beq

6738: T=\left(

6739: \begin{array}{ccccc}

6740: 1&0&0&0&-1\\

6741: 0&1&0&0&-1\\

6742: 0&0&1&0&-1\\

6743: 0&0&0&1&-1

6744: \end{array}

6745: \right)

6746: \label{inter}

6747: \eeq

6748: 

6749: \subsubsection{Action of parity on D-branes}

6750: 

6751: As shown above, all branes are invariant under the parity and

6752: support an orthogonal gauge group, ie, we have $\sigma=+1$ for

6753: all branes in the list (\ref{qlist}).

6754: 

6755: \subsubsection{Solutions}

6756: 

6757: The positive integers $n_i$ must then satisfy the tadpole condition

6758: (\ref{tadcan}) in the explicit form

6759: \beq

6760: T\; Q \;(n_1,n_2,n_3,n_4,n_5,n_5)^t = T\;(8,20,40,40,20)^t \,,

6761: \eeq

6762: which indeed reduces to two linearly independent equations,

6763: \beq

6764: \begin{array}{rcl}

6765: n_1+ n_3+n_4+2n_5+3n_6&=&12 \\

6766: n_2+n_3+2 n_4+ 3n_5 + 5n_6 &=&20 \,.

6767: \end{array}

6768: \label{condition}

6769: \eeq

6770: Obviously, there is only a finite number of solutions to the equations 

6771: (\ref{condition}) with positive $n_i$ (negative $n_i$ means using 

6772: the antibrane and this breaks supersymmetry). A simple computer 

6773: aided count shows that there are in fact $417$ solutions. For

6774: each such solution, the number  of ways of distributing the charge 

6775: among the $32$ branes with the same supersymmetry is then given

6776: by

6777: \beq

6778: \# (n_1,n_2,n_3,n_4,n_5,n_6) = 

6779: \prod_{i=1}^6 \left(\!\!

6780: \begin{array}{c} 

6781: n_i+m_i-1 \\[-.2cm] n_i

6782: \end{array}\!\!\right)

6783: \label{distribute}

6784: \eeq

6785: where the multiplicities $m_i$ are given by $1$, $5$, $10$, $10$, 

6786: $5$, and $1$. Again using a computer, one then finds that the 

6787: grand total number of tadpole canceling brane configurations 

6788: using only the rational branes at the Gepner point is equal to 

6789: $31561671503$, as advertised.

6790: 

6791: \noindent

6792: {\bf Remarks.}\\{\small

6793: {\bf (i)}~All these solutions of the tadpole conditions we have

6794: constructed above have a worldsheet description in terms of 

6795: rational conformal field theories based on orbifolds of 

6796: ${\mathcal N}=2$ minimal models, and are ``in principle 

6797: exactly solvable'' perturbative string vacua with 

6798: ${\mathcal N}=1$ spacetime supersymmetry in $3+1$ dimensions. 

6799: This impressive number is much larger than a comparable number 

6800: in heterotic string constructions (see, e.g., \cite{Gepnerclassification}).

6801: Of course, all these vacua and their moduli spaces are potentially 

6802: unstable to non-perturbative effects.

6803: \\ 

6804: {\bf (ii)}~We emphasize that in counting the number of solutions,

6805: we have {\it not} divided out by the symmetry group $\mathfrak{S}^5$

6806: which exchanges the various minimal model factors. It might

6807: seem that this is overcounting, since the solutions mapped onto

6808: each other under such a symmetry must lead to the same 

6809: perturbative low energy physics. However, one also has to 

6810: take into account that the Gepner point is a special point in

6811: the (closed string) K\"ahler moduli space. There are perturbation

6812: away from that point which break the exchange symmetries, also

6813: after orientifold projection. Once such a perturbation has

6814: been turned on (or if the corresponding moduli are fixed away

6815: from the Gepner point by some mechanism), the various solutions

6816: will no longer lead to the same physics at low energies, so

6817: we count them as distinct ``vacua''. (But, of course, they 

6818: are still exchanged if we act on all fields (closed and open 

6819: strings) simultaneously.)

6820: \\ 

6821: {\bf (iii)}~The solutions we have constructed are valid right at the

6822: Gepner point.

6823: One can ask what happens to these solutions when one moves

6824: away from the Gepner point. 

6825: On general grounds, the branes we

6826: have used to cancel the tadpoles might disappear at lines

6827: of marginal stability. When this happens, then as discussed 

6828: in section \ref{sec:FI}, we expect that either there is

6829: a new supersymmetric brane configuration obtained by

6830: condensing some open string tachyon, or there is none, in

6831: which case the K\"ahler moduli is lifted at string loop level.

6832: On top of this there could also be string non-perturbative effect

6833: that may fix some of the K\"ahler moduli.

6834: }

6835: 

6836: %JW: new subsubsection

6837: \subsubsection{Distribution of gauge group rank}

6838: 

6839: The solutions to the tadpole conditions we have found are certainly

6840: too numerous to make a complete list. But we can gain a qualitative 

6841: overview over the possibilities by looking at the distribution of a 

6842: certain property over the set of all models, for example the rank of

6843: the unbroken gauge group or the number of massless chiral fields. 

6844: Such a statistical approach to exploring the string theory vacua

6845: has recently been advocated in particular in 

6846: \cite{DougString,statistics}. Let us here present the result of 

6847: this type of counting for a very simple quantity, the total rank 

6848: of the gauge group.

6849: 

6850: In all type IIB orientifolds of the quintic we have found, the gauge 

6851: group is a product of orthogonal groups $G=\prod_{j} O(N_j)$, where 

6852: $N_j$ is a positive integer. By  slight abuse of terminology, we will 

6853: call the maximal number of $U(1)$ subgroups of $G$ the rank of $G$. In 

6854: particular, the ranks of $O(1)$ and $O(2)$ are defined to be $0$ and 

6855: $1$, respectively. Then, for each solution of the equations

6856: (\ref{condition}), we have to distribute the $n_i$ (the number of 

6857: times a given charge appears) among the various branes with that 

6858: given charge. This is similar to what we did in (\ref{distribute})

6859: to count the total number of solutions, but we have to take into

6860: account that the ranks depends on how we distribute the $n_i$'s.

6861: In any event, the rank of a given solution is then computed as

6862: $$

6863: \sum_j \biggl[\frac{N_j}{2}\biggr]  \,.

6864: $$

6865: The maximal possible rank is $16$ \cite{Blumenhagen}. It is 

6866: obtained from the solution $n_1=12$, $n_2=20$ of (\ref{condition})

6867: by choosing $12$ times the brane $B_{(0,0,0,0,0),0}$ and

6868: $20$ branes $\overline{B_{{\bf L},5}}$ with one $L_i=1$, an 

6869: even number of each. The number of possibilities is 

6870: $\left({24\atop4}\right)=1001$. The results for the

6871: other ranks between $0$ and $16$ are shown in Figure \ref{fig:ranks}.

6872: 

6873: \begin{figure}[ht]

6874: \centerline{\scriptsize $\begin{array}[b]{|l|r|}

6875: \hline

6876: {\rm Rank} & {\rm Number\; of\; solutions}\\

6877: \hline\hline

6878: 0&   41100850    \\

6879: 1&   410137435    \\ 

6880: 2&   1767975754   \\ 

6881: 3&   4320652050   \\ 

6882: 4&   6758910800   \\ 

6883: 5&   7251800650   \\ 

6884: 6&   5593308703   \\ 

6885: 7&   3227024877   \\ 

6886: 8&   1450260204   \\ 

6887: 9&   527957402    \\ 

6888: 10&  161242450    \\ 

6889: 11&  41130702     \\ 

6890: 12&  8534850      \\ 

6891: 13&  1460250      \\ 

6892: 14&  159225       \\ 

6893: 15&  14300        \\ 

6894: 16&  1001         \\

6895: \hline

6896: \end{array}$

6897: \qquad\qquad

6898: \psfrag{r}{rank}

6899: \psfrag{n}{}

6900: \psfrag{0}{$0$}

6901: \psfrag{1}{$1$}

6902: \psfrag{2}{$2$}

6903: \psfrag{3}{$3$}

6904: \psfrag{4}{$4$}

6905: \psfrag{5}{$5$}

6906: \psfrag{6}{$6$}

6907: \psfrag{7}{$7$}

6908: \psfrag{8}{$8$}

6909: \psfrag{9}{$9$}

6910: \psfrag{10}{$10$}

6911: \psfrag{11}{$11$}

6912: \psfrag{12}{$12$}

6913: \psfrag{13}{$13$}

6914: \psfrag{14}{$14$}

6915: \psfrag{15}{$15$}

6916: \psfrag{16}{$16$}

6917: \psfrag{100}{$\!\!\!\!\!\! 1\times 10^9$}

6918: \psfrag{200}{$\!\!\!\!\!\! 2\times 10^9$}

6919: \psfrag{300}{$\!\!\!\!\!\! 3\times 10^9$}

6920: \psfrag{400}{$\!\!\!\!\!\! 4\times 10^9$}

6921: \psfrag{500}{$\!\!\!\!\!\! 5\times 10^9$}

6922: \psfrag{600}{$\!\!\!\!\!\! 6\times 10^9$}

6923: \psfrag{700}{$\!\!\!\!\!\! 7\times 10^9$}

6924: \epsfig{width=4in,file=ranks.eps}}

6925: %\includegraphics{ranks.eps}

6926: \caption{Distribution of total rank of gauge group over the 

6927: solutions of tadpole conditions for Type IIB orientifold of 

6928: quintic Gepner model.}

6929: \label{fig:ranks}

6930: \end{figure}

6931: 

6932: It is interesting that the peak of the distribution 

6933: lies around rank $5$, which is rather close to the

6934: value of the Standard Model. Let us also note that the 

6935: distribution goes to zero very rapidly for large ranks, 

6936: and --- somewhat surprisingly --- has a rather large 

6937: support at small ranks. In particular, rank $0$, which 

6938: corresponds in our language to having only distinct 

6939: parity-invariant branes supporting $O(1)$ gauge groups, 

6940: appears with appreciable frequency.

6941: 

6942: 

6943: \subsection{Solutions of the tadpole conditions --- Two parameter model}

6944: 

6945: Similarly to the quintic, we use the $8$ ${\bf L}={\bf 0}$ branes to express 

6946: the charges of the other branes and the O-planes. The charges of these $8$

6947: branes satisfy the two relations

6948: \beqa

6949: \bigl[B^B_{(00000),0} \bigr] + 

6950: \bigl[B^B_{(00000),4} \bigr] + 

6951: \bigl[B^B_{(00000),8} \bigr] + 

6952: \bigl[B^B_{(00000),12} \bigr] &=&0\nn\\

6953: \bigl[B^B_{(00000),2} \bigr] + 

6954: \bigl[B^B_{(00000),6} \bigr] + 

6955: \bigl[B^B_{(00000),10} \bigr] + 

6956: \bigl[B^B_{(00000),14} \bigr] &=&0

6957: \label{tprels}

6958: \eeqa

6959: For the practical computations, we 

6960: will use the relations (\ref{tprels}) to project to a linearly 

6961: independent set of $6$ charges with the matrix

6962: \beq

6963: T=

6964: \left(

6965: \begin{array}{cccccccc}

6966: 1&0&0&0&0&0&-1&0\\

6967: 0&1&0&0&0&0&0&-1\\

6968: 0&0&1&0&0&0&-1&0\\

6969: 0&0&0&1&0&0&0&-1\\

6970: 0&0&0&0&1&0&-1&0\\

6971: 0&0&0&0&0&1&0&-1\\

6972: \end{array}

6973: \right)

6974: \label{Ttwo}

6975: \eeq

6976: 

6977: \subsubsection{Parities and O-planes}

6978: 

6979: In the two parameter model, we have various parities $P^B_{(\omega;

6980: \epsilon_1,\ldots,\epsilon_5)}$ to consider. Here $\omega=0,1$ 

6981: and $(\epsilon_1,\ldots,\epsilon_5)$, $\epsilon_i=\pm$, denote the 

6982: twisting by quantum and classical symmetries, respectively. This is 

6983: a slightly redundant labeling because as we recall  $(\epsilon_1,

6984: \epsilon_2,\epsilon_3,\epsilon_4,\epsilon_5)$ is the same parity 

6985: as $(-\epsilon_1,-\epsilon_2,\epsilon_3,\epsilon_4,\epsilon_5)$.

6986: The charges of the corresponding O-planes are given in table

6987: \ref{Otwopar1} and table \ref{Otwopar2} in subsection 

6988: \ref{sec:oplanec}. 

6989: 

6990: \subsubsection{Supersymmetry preserving branes}

6991: 

6992: \subsubsection*{Even quantum symmetry}

6993: 

6994: According to the discussion in subsection \ref{sec:oplanec}, O-planes

6995: corresponding to parities with even quantum symmetry dressing

6996: preserve a spacetime supersymmetry with even $M$-label, $M_O=0,8,4,12$,

6997: depending on the $\epsilon_i$'s. Branes preserving the same

6998: supersymmetry must have $M=M_O$ or $M=M_O+8$, and this restricts

6999: the possible ${\bf L}$ labels of the branes to $L_1+L-2={\rm even}$.

7000: In order to get familiar with the use of the formula (\ref{branec}), 

7001: we give here a list of branes preserving the same supersymmetry as 

7002: any given O-plane as well as their charges. We only list ${\bf L}$ 

7003: labels up to permutation of factors with equal levels. In the last

7004: column, we give the expansion of the brane charge (for $M_O=0$) in 

7005: terms of the ${\bf L}={\bf 0}$ branes, modulo the relations

7006: (\ref{tprels}). As one can see, there are $15$ different charges.

7007: The eight-component vectors give, for example, the following

7008: equation:

7009: \beq

7010: [\widehat{B}^B_{{\bf\frac{k}{2}},M}]~=~

7011: \frac14[B^B_{{\bf\frac{k}{2}},M}]~=~

7012:   [B^B_{{\bf 0},M+4}]+2[B^B_{{\bf 0},M+2}]+2[B^B_{{\bf 0},M  }]

7013: +2[B^B_{{\bf 0},M-2}] +[B^B_{{\bf 0},M-4}].

7014: \label{chBk/2}

7015: \eeq

7016: It is also a useful exercise to check the ``multiplicity'' or number 

7017: of inequivalent branes with the same charge.

7018: 

7019: {\scriptsize

7020: \begin{center}

7021: \begin{tabular}{|l||lll|l|l|}

7022: \hline

7023: charge \# & ${\bf L}$ & $M$    &  $S$ & $g^{-M_O/2} Q$ & multiplicity $m_i$ \\

7024: \hline

7025: $1$     & $(00000)$ &  $M_O$   &  $0$ & $(1, 0, 0, 0, 0, 0, 0, 0)$        & $1$\\  

7026: $2$     & $(00000)$ &  $M_O+8$ &  $2$ & $(1, 0, 1, 0, 0, 0, 1, 0)$       & $1$\\

7027: $3$     & $(11000)$ &  $M_O$   &  $0$ & $(2, 1, 0, 0, 0, 0, 0, 1)$        & $1$\\

7028: $4$     & $(11000)$ &  $M_O+8$ &  $2$ & $(2, 1, 2, 0, 0, 0, 2, 1)$       & $1$\\  

7029: $5$     & $(20000)$ &  $M_O$   &  $0$ & $(1, 1, 0, 0, 0, 0, 0, 1)$        & $2$\\

7030: $6$     & $(20000)$ &  $M_O+8$ &  $2$ & $(1, 1, 1, 0, 0, 0, 1, 1)$       & $2$\\

7031: $7$     & $(22000)$ &  $M_O$   &  $0$ & $(3, 2, 1, 0, 0, 0, 1, 2)$        & $1$\\

7032: $8$     & $(22000)$ &  $M_O+8$ &  $2$ & $(3, 2, 2, 0, 0, 0, 2, 2)$       & $1$\\

7033: $9$     & $(00100)$, $(00111)$ 

7034:                     &  $M_O$   &      & $(0, 1, 0, 0, 0, 0, 0, 1)$        & $4$\\

7035: $10$    & $(00110)$ &  $M_O$   &      & $(1, 0, \frac12, 0, 0, 0, \frac12, 0)$ & $6$\\

7036: $11$    & \parbox[t]{2.5cm}{$(11100)$,  $(11111)$, \\[-.1cm] 

7037: $(31000)$, $(31110)$,\\[-.1cm]

7038: $(33100)$, $(33111)$} 

7039:                     &  $M_O$   &      & $(2, 2, 1, 0, 0, 0, 1, 2)$        & $16$\\

7040: $12$    & \parbox[t]{2.5cm}{$(11110)$, $(20100)$,\\[-.1cm]

7041:  $(20111)$, $(31100)$,\\[-.1cm]

7042: $(31111)$, $(33000)$, \\[-.1cm]

7043: $(33110)$} 

7044:                     &  $M_O$   &      & $(2, 1, 1, 0, 0, 0, 1, 1)$        & $38$\\        

7045: $13$    & $(20110)$ &  $M_O$   &      & $(1, 1, \frac12, 0, 0, 0, \frac12, 1)$ & $12$\\

7046: $14$    & $(22100)$, $(22111)$ 

7047:                     &  $M_O$   &      & $(4, 3, 2, 0, 0, 0, 2, 3)$        & $4$ \\

7048: $15$    & $(22110)$ &  $M_O$   &      & $(3, 2, \frac32, 0, 0, 0, \frac32, 2)$ &$6$\\

7049: \hline

7050: \end{tabular}

7051: \end{center}

7052: }

7053: 

7054: \subsubsection*{Odd quantum symmetry}

7055: 

7056: O-planes corresponding to dressing with odd quantum symmetry

7057: preserve a spacetime supersymmetry with odd $M$-label $M=1,5,9,13$,

7058: depending on the $\epsilon_i$'s. We are then restricted to 

7059: branes with $L_1+L_2$ odd. The list of branes and charges

7060: is

7061: 

7062: {\scriptsize

7063: \begin{center}

7064: \begin{tabular}{|l||lll|l|l|}

7065: \hline

7066: charge \# & ${\bf L}$ & $M$    &  $S$ & $g^{-(M_O-1)/2} Q$ & multiplicity $m_i$ \\

7067: \hline

7068: $1$     & $(10000)$            &  $M_O$   &  $0$ & $(1, 1, 0, 0, 0, 0, 0, 0)$    &$2$\\

7069: $2$     & $(10000)$            &  $M_O+8$ &  $2$ & $(1, 1, 1, 1, 0, 0, 1, 1)$   &$2$\\

7070: $3$     & $(21000)$            &  $M_O$   &  $0$ & $(2, 2, 1, 0, 0, 0, 0, 1)$    &$2$\\

7071: $4$     & $(21000)$            & $M_O+8$  &  $2$ & $(2, 2, 2, 1, 0, 0, 1, 2)$   &$2$\\

7072: $5$     & \parbox[t]{2.5cm}{$(10100)$, $(10111)$\\[-.1cm]

7073:          $(30000)$, $(30110)$} & $M_O$    &      & $(1, 1, 1, 0, 0, 0, 0, 1)$    &$16$\\

7074: $6$     & \parbox[t]{2.5cm}{$(10110)$, $(30100)$\\[-.1cm]

7075:         $(30111)$}             & $M_O$    &      & $(1, 1, \frac12, \frac12, 0, 0, \frac12, \frac12)$&$28$\\

7076: $7$     & \parbox[t]{2.5cm}{$(21100)$, $(21111)$\\[-.1cm]  

7077:          $(32000)$, $(32110)$} & $M_O$    &      & $(3, 3, 2, 1, 0, 0, 1, 2)$   &$16$\\

7078: $8$     & \parbox[t]{2.5cm}{$(21110)$, $(32100)$\\[-.1cm]

7079:          $(32111)$}            & $M_O$    &      & $(2, 2, \frac32,

7080:  \frac12, 0, 0, \frac12, \frac32)$ &$28$\\

7081: \hline

7082: \end{tabular}

7083: \end{center}

7084: }

7085: 

7086: 

7087: \subsubsection{Action of parities on D-branes}

7088: 

7089: We summarize the action of the B-type parities $P^B_{\omega;

7090: \epsilon_1,\ldots,\epsilon_5}$ on the branes $B^B_{{\bf L},M,S,\psi}$

7091: for the two parameter model $(6,6,2,2,2)$.

7092: 

7093: Neglecting again for a very short moment the $S$ and $\psi$ labels,

7094: the branes are mapped under parity as follows:

7095: \beqa

7096: &&P^B_{0;\epsilon_1...\epsilon_5}:

7097: B_{{\bf L},M}\longmapsto \epsilon_1\cdots\epsilon_5 

7098: B_{{\bf L},-M},\nn\\

7099: &&P^B_{1;\epsilon_1...\epsilon_5}:

7100: B_{{\bf L},M}\longmapsto \epsilon_1\cdots\epsilon_5 

7101: B_{{\bf L},2-M}.\nn

7102: \eeqa

7103: For branes carrying a $\psi$ label, this $\psi$ label is

7104: transformed according to

7105: \beq

7106: \psi\longmapsto (-1)^{|\II|/2}\;\psi.

7107: \eeq

7108: If the sign $\epsilon_1\cdots\epsilon_5$ is negative, this means

7109: that the branes are mapped to antibranes. This minus signs can be

7110: absorbed according to the rule

7111: \beq

7112: -B_{{\bf L},M,S}= B_{{\bf L},M,S+2},~~

7113: -B_{{\bf L},M}  = B_{{\bf L},M+H},~~

7114: -\widehat{B}_{{\bf L},M}^\psi  = \widehat{B}_{{\bf L},M+H}^{-\psi}.

7115: \eeq

7116: Restricting ourselves to the supersymmetry-preserving branes,

7117: we find that the branes with $|{\bf S}|=$ odd are all parity invariant.

7118: As for the branes with $|{\bf S}|=$ even, those with $|{\bf S}|=0$ or 

7119: $4$ are invariant under parities with $\epsilon_1\cdots\epsilon_5=1$

7120: and not under parities with $\epsilon_1\cdots\epsilon_5=-1$.

7121: The branes with $|{\bf S}|=2$ behave in an opposite way.

7122: 

7123: 

7124: The gauge group on parity invariant branes is either $O$ or

7125: $Sp$ according to the sign (\ref{Ggroup}) or (\ref{Ggroup2}).

7126: Other branes support the unitary gauge groups.

7127: A complete list of supersymmetry-preserving branes together

7128: with the gauge groups is given in tables \ref{cp-2pb1},

7129: \ref{cp-2pb2} in subsection \ref{spectrum}.

7130: 

7131: 

7132: We have presented the list of D-branes preserving the same

7133: supersymmetry as any given orientifold, and they formed 15 or 8

7134: groups according to the RR-charge.

7135: The number of groups depends on the dressings with quantum symmetries.

7136: Each group contains branes having different signature $\sigma$ and

7137: therefore supporting different gauge groups,

7138: and it is necessary for counting the supersymmetric vacua

7139: to know the numbers $m_i^{\sigma}$ of branes with definite

7140: signature in a given group.

7141: They are defined to satisfy

7142: \beq

7143:   m_i = m_i^{+}+2m_i^{0}+m_i^{-},

7144: \eeq

7145: where $m_i$ is the total multiplicity of branes in the $i$-th group.

7146: For later convenience we give a table of these numbers below.

7147: 

7148: \begin{table}[tbh]

7149: {\scriptsize\begin{center}

7150: \begin{tabular}{|c||c|c|c|c|c|c|c|c|}

7151: \hline

7152: $\#$

7153:  &$\!\!P_{0;+++++}\!\!$&$\!\!P_{0;+-+++}\!\!$

7154:  &$\!\!P_{0;++-++}\!\!$&$\!\!P_{0;+--++}\!\!$

7155:  &$\!\!P_{0;++--+}\!\!$&$\!\!P_{0;+---+}\!\!$

7156:  &$\!\!P_{0;++---}\!\!$&$\!\!P_{0;+----}\!\!$\\

7157: \hline

7158: 1,2,7,8&(1,0,0)&(0,1,0)&(0,1,0)&(0,0,1)&(0,0,1)&(0,1,0)&(0,1,0)&(1,0,0)\\

7159:  3,4 &(1,0,0)&(0,1,0)&(0,1,0)&(1,0,0)&(0,0,1)&(0,1,0)&(0,1,0)&(0,0,1)\\

7160:  5,6 &(2,0,0)&(0,2,0)&(0,2,0)&(0,0,2)&(0,0,2)&(0,2,0)&(0,2,0)&(2,0,0)\\

7161:  9,14&(3,0,1)&(0,0,4)&(2,0,2)&(1,0,3)&(3,0,1)&(2,0,2)&(0,0,4)&(1,0,3)\\

7162: 10,15&(0,6,0)&(0,0,6)&(4,0,2)&(0,6,0)&(0,6,0)&(2,0,4)&(6,0,0)&(0,6,0)\\

7163: 11   &(5,0,11)&(15,0,1)&(7,0,9)&(11,0,5)&(7,0,9)&(9,0,7)&(9,0,7)&(9,0,7)\\

7164: 12   &(6,20,12)&(20,10,8)&(12,10,16)&(10,20,8)

7165:      &(12,20,6)&(16,10,12)&(8,10,20)&(8,20,10)\\

7166: 13   &(0,12,0)&(0,0,12)&(8,0,4)&(0,12,0)&(0,12,0)&(4,0,8)&(12,0,0)&(0,12,0)\\

7167: \hline

7168: \end{tabular}

7169: \vskip5mm

7170: \begin{tabular}{|c||c|c|c|c|c|c|c|c|}

7171: \hline $\#$

7172:  &$\!\!P_{1;+++++}\!\!$&$\!\!P_{1;+-+++}\!\!$

7173:  &$\!\!P_{1;++-++}\!\!$&$\!\!P_{1;+--++}\!\!$

7174:  &$\!\!P_{1;++--+}\!\!$&$\!\!P_{1;+---+}\!\!$

7175:  &$\!\!P_{1;++---}\!\!$&$\!\!P_{1;+----}\!\!$\\

7176: \hline

7177: 1,2,3,4&(2,0,0)&(0,2,0)&(0,2,0)&(1,0,1)&(0,0,2)&(0,2,0)&(0,2,0)&(1,0,1)\\

7178: 5,7    &(8,0,8)&(8,0,8)&(8,0,8)&(8,0,8)&(8,0,8)&(8,0,8)&(8,0,8)&(8,0,8)\\

7179: 6,8    &(0,24,4)&(12,4,12)&(12,4,12)&(2,24,2)

7180:        &(4,24,0)&(12,4,12)&(12,4,12)&(2,24,2)\\

7181: \hline

7182: \end{tabular}

7183: \end{center}}

7184: \caption{Multiplicities $(m_i^+,2m_i^0,m_i^-)$ of branes

7185:          with different gauge groups.}

7186: \end{table}

7187: 

7188: 

7189: \subsubsection{Counting the Solutions}

7190: 

7191: Let us now count the number of supersymmetric vacua in various

7192: orientifolds of two parameter model.

7193: As compared to the case with quintic, there arise a new

7194: complication due to the presence of $Sp$ or $U$-type branes.

7195: We explain the detail of the counting in the cases $P^B_{0,+++++}$

7196: and $P^B_{0,+-+++}$, and state the results for other cases briefly.

7197: 

7198: \subsubsection*{Parity $P^B_{0;+++++}$}

7199: 

7200: Using (\ref{chBk/2}) and the expression in the table {\ref{Otwopar1}},

7201: this O-plane is shown to have the RR-charge

7202: \beq

7203: \bigl[O_P\bigr] ~=~ ( 2,  6, 16, 26, 30, 26, 16,  6)^t

7204:           ~\simeq~ -(28, 20, 14,  0,  0,  0, 14, 20)^t.

7205: \eeq

7206: Under the choice $\tilomega^{\frac12}=-1$, it preserves a

7207: spacetime supersymmetry corresponding to $M_O=0$.

7208: We denote by $n_i$, $i=1,2,\ldots,15$ the number of times

7209: a given charge appears in a tadpole canceling configuration.

7210: Computing the charges, projecting onto $6$ independent ones

7211: and equating brane and O-plane charges leads to the following

7212: three linearly independent equations on the $n_i$'s.

7213: \beq

7214: \left(

7215: \begin{array}{ccccccccccccccc}

7216: 1& 1& 2& 2& 1& 1& 3& 3& 0& 1& 2& 2& 1& 4& 3\\

7217: 0& 0& 1& 1& 1& 1& 2& 2& 1& 0& 2& 1& 1& 3& 2\\

7218: 0& 1& 0& 2& 0& 1& 1& 2& 0& \frac12& 1& 1& \frac12& 2& \frac32

7219: \end{array}

7220: \right)

7221: \left(

7222: \begin{array}{c}

7223: n_1\\

7224: n_2\\

7225: \vdots

7226: \\

7227: n_{15}

7228: \end{array}

7229: \right)

7230: =\left(

7231: \begin{array}{c}

7232: 28\\

7233: 20\\

7234: 14

7235: \end{array}

7236: \right)

7237: \label{btcc2p1}

7238: \eeq

7239: Note that again, there is a finite number of solutions

7240: to these equations, and we can count the total number of

7241: tadpole canceling D-brane configurations.

7242: 

7243: The counting goes in the following way.

7244: Suppose that the numbers $\{n_i\}$~($i=1,\cdots,15$ or $8$) give

7245: a solution to (\ref{btcc2p1}).

7246: For each solution, we have to distribute the

7247: charge over the various branes with fixed charge,

7248: taking into account their signature $\sigma$.

7249: The first 8 charges are carried only by $O$-type branes,

7250: but the other 7 are carried by branes with various signatures.

7251: We note that even though the charges $i>8$ are invariant

7252: under the parity, the branes themselves need not be

7253: because of the action on the $\psi$-label.

7254: Consider all the possible decompositions of each $n_i$~($i=9,\cdots,15$)

7255: into

7256: \beq

7257:   n_i = n_i^++n_i^0+n_i^-

7258: \eeq

7259: with the condition that $n_i^0$ and $n_i^-$ be even.

7260: The combinatorial factor associated to a solution $\{n_i\}$

7261: is given by the sum over all the possible decompositions

7262: \beq

7263: \#(n_i)=

7264:   (n_5+1)(n_6+1)

7265:   \prod_{i=9}^{15}\sum_{n_i=\Sigma_\sigma n_i^\sigma}

7266: \left(\!\!

7267: \begin{array}{c} 

7268: \frac{n_i^-}2 + m^-_i-1 \\[0cm] \frac{n_i^-}2

7269: \end{array}\!\!\right)

7270: \left(\!\!

7271: \begin{array}{c} 

7272: \frac{n_i^0}2 + m_i^0-1 \\ \frac{n_i^0}2

7273: \end{array}\!\!\right)

7274: \left(\!\!

7275: \begin{array}{c} 

7276: n_i^++m_i^+-1 \\ n_i^+

7277: \end{array}\!\!\right).

7278: \label{comb1}

7279: \eeq

7280: The number of vacua is therefore the sum of this over

7281: all the solutions $\{n_i\}$.

7282: The total number turns out to be $13213511375147\approx 10^{13}$.

7283: 

7284: \subsubsection*{Parity $P^B_{0;+-+++}$}

7285: 

7286: Under the choice $\tilomega^{\frac12}=-i$, this O-plane preserves

7287: a spacetime supersymmetry corresponding to $M_O=4$. Its RR-charge is

7288: \beq

7289: \bigl[O_P\bigr]~=~  (0,-4,-6,-4, 0, 4, 6, 4)^t

7290:           ~\simeq~ -(6, 8,12, 8, 6, 0, 0, 0)^t.

7291: \eeq

7292: One can easily see that the first $8$ branes on the list are

7293: mapped onto each other under this parity, and give rise to

7294: a unitary gauge group.

7295: We have to require $n_i=n_{i+1}$ for $i=1,3,5,7$.

7296: Equating crosscap and brane charge then leads to the

7297: two independent conditions

7298: \beq

7299: \left(

7300: \begin{array}{ccccccccccc}

7301: 2& 4& 2& 6&  0& 1& 2& 2& 1& 4& 3 \\

7302: 0& 2& 2& 4&  1& 0& 2& 1& 1& 3& 2

7303: \end{array}

7304: \right)

7305: \left(

7306: \begin{array}{c}

7307: n_1\\

7308: n_3\\

7309: \vdots

7310: \\

7311: n_{15}

7312: \end{array}

7313: \right)

7314: =\left(

7315: \begin{array}{c}

7316: 12\\

7317: 8

7318: \end{array}

7319: \right).

7320: \label{btcc2p2}

7321: \eeq

7322: The number of vacua is given by the sum of the combinatoric factors

7323: \beq

7324: \# (n_i) = \bigl(n_5+1\bigr)\;

7325: \prod_{i=9}^{15} 

7326: \sum

7327: \left(\!\!

7328: \begin{array}{c} 

7329: \frac{n_i^-}2 + m^-_i-1 \\[0cm] \frac{n_i^-}2

7330: \end{array}\!\!\right)

7331: \left(\!\!

7332: \begin{array}{c} 

7333: \frac{n_i^0}2 + m_i^0-1 \\ \frac{n_i^0}2

7334: \end{array}\!\!\right)

7335: \left(\!\!

7336: \begin{array}{c} 

7337: n_i^++m_i^+-1 \\ n_i^+

7338: \end{array}\!\!\right),

7339: \label{comb2}

7340: \eeq

7341: over all the solutions of (\ref{btcc2p2}).

7342: The total number is $47803952$.

7343: 

7344: \subsubsection*{Other parities (with even quantum symmetry dressings)}

7345: One can analyze the cases with other parities in a similar way.

7346: Let us denote by $|\epsilon|$ the number of minus signs in $\epsilon_i$.

7347: Choosing $\tilomega^{\frac12}=-i^{|\epsilon|}$ for the parity

7348: $P^B_{0;\epsilon_1\epsilon_2\epsilon_3\epsilon_4\epsilon_5}$,

7349: the O-plane preserves a spacetime supersymmetry corresponding

7350: to $M_O=4|\epsilon|$.

7351: The O-planes have the RR-charges

7352: \beqa

7353: \bigl[O_{P^B_{0;+++++}}\bigr]

7354:  &\simeq& -(28,20,14, 0, 0, 0,14,20)^t, \nn\\

7355: \bigl[O_{P^B_{0;+-+++}}\bigr] ~=~

7356: \bigl[O_{P^B_{0;++-++}}\bigr]

7357:  &\simeq& -(6, 8,12, 8, 6, 0, 0, 0)^t, \nn\\

7358: \bigl[O_{P^B_{0;+--++}}\bigr] ~=~

7359: \bigl[O_{P^B_{0;++--+}}\bigr]

7360:  &\simeq& -( 0, 0, 2, 4, 4, 4, 2, 0)^t, \nn\\

7361: \bigl[O_{P^B_{0;+---+}}\bigr] ~=~

7362: \bigl[O_{P^B_{0;++---}}\bigr]

7363:  &\simeq& -( 2, 0, 0, 0, 2, 0, 4, 0)^t, \nn\\

7364: \bigl[O_{P^B_{0;+----}}\bigr]

7365:  &\simeq& -(-4, 4,-2, 0, 0, 0,-2, 4)^t.

7366: \label{ochvec1}

7367: \eeqa

7368: So the tadpole cancellation conditions for other O-planes are

7369: given by the replacements $(28,20,14)\to (x,y,z)$ in (\ref{btcc2p1}),

7370: where $x,y$ and $z$ are the $(2|\epsilon|+1)$-st, $(2|\epsilon|+2)$-nd

7371: and $(2|\epsilon|+3)$-rd components of the above vectors.

7372: When $|\epsilon|$ is odd, the branes in the groups $i=1,\cdots,8$

7373: are all U-type and the parity maps the charges $1\leftrightarrow2$,

7374: $3\leftrightarrow4$ and so on.

7375: So we have to put $n_i=n_{i+1}$ for $i={1,3,5,7}$ in these cases,

7376: and there remain only two linearly independent equations

7377: for $n_1,n_3,\cdots,n_{15}$.

7378: 

7379: The total number of vacua is given by the sum of combinatoric

7380: factors $\#(n_i)$, defined in a similar way as (\ref{comb1})

7381: or (\ref{comb2}), over all the solutions $\{n_i\}$ of tadpole

7382: cancellation condition.

7383: The result is summarized below.

7384: \begin{center}

7385: {\footnotesize

7386: \begin{tabular}{|c|c|c|c|c|c|c|c|}

7387: \hline

7388: $\!\!B_{0;+++++}\!\!$&

7389: $\!\!B_{0;+-+++}\!\!$&

7390: $\!\!B_{0;++-++}\!\!$&

7391: $\!\!B_{0;+--++}\!\!$&

7392: $\!\!B_{0;++--+}\!\!$&

7393: $\!\!B_{0;+---+}\!\!$&

7394: $\!\!B_{0;++---}\!\!$&

7395: $\!\!B_{0;+----}\!\!$ \\

7396: \hline

7397: 13213511375147 &

7398: 47803952 &

7399: 434841441 &

7400: 1051 &

7401: 2162 &

7402: 35 &

7403: 148 &

7404: 0

7405: \\ \hline

7406: \end{tabular}}

7407: \end{center}

7408: 

7409: \subsubsection*{Other parities (with odd quantum symmetry dressings)}

7410: 

7411: For parities $P^B_{1;\epsilon_1\epsilon_2\epsilon_3\epsilon_4\epsilon_5}$

7412: with odd quantum symmetry dressings,

7413: we choose $\tilomega^{\frac12}=-i^{|\epsilon|}$ and consider the

7414: systems of an O-plane and D-branes with $M_O=1+4|\epsilon|$.

7415: In solving the tadpole cancellation condition, note first that

7416: for the parities $P^B_{+-***}$ the O-planes have no RR-charge so that

7417: the O-planes by themselves give the unique consistent superstring

7418: backgrounds.

7419: For other parities, the O-planes have the RR-charges

7420: \beqa

7421: \bigl[O_{P^B_{1;+++++}}\bigr]

7422:  &\simeq& -(28,28,20, 8, 0, 0, 8,20)^t, \nn\\

7423: \bigl[O_{P^B_{1;++-++}}\bigr]

7424:  &\simeq& -( 4, 8,12,12, 8, 4, 0, 0)^t, \nn\\

7425: \bigl[O_{P^B_{1;++--+}}\bigr]

7426:  &\simeq& -( 0, 0, 0, 4, 4, 4, 4, 0)^t, \nn\\

7427: \bigl[O_{P^B_{1;++---}}\bigr]

7428:  &\simeq& -( 0, 4, 0, 0, 4, 0, 4, 4)^t.

7429: \label{ochvec2}

7430: \eeqa

7431: 

7432: Tadpole cancellation condition is then given by three linearly

7433: independent equations on 8 numbers.

7434: For the parity $P^B_{1;+++++}$ it becomes

7435: \beq

7436: \left(\begin{array}{cccccccc}

7437: 1& 1& 2& 2& 1& 1     & 3& 2 \\

7438: 0& 1& 1& 2& 1& \frac12& 2& \frac32 \\

7439: 0& 1& 0& 1& 0& \frac12& 1& \frac12

7440: \end{array}\right)

7441: \left(\begin{array}{c} n_1\\n_2\\\vdots\\n_8\end{array}\right)

7442: ~=~ 

7443: \left(\begin{array}{c} 28\\20\\8\end{array}\right),

7444: \eeq

7445: and for other parities we only have to replace the numbers $(28,20,8)$

7446: on the right hand side with

7447: $(2|\epsilon|+2,2|\epsilon|+3,2|\epsilon|+4)$-th components of the

7448: vectors (\ref{ochvec2}).

7449: When $|\epsilon|$ is odd, the branes in the group 1 and 2, 3 and 4

7450: are mapped to each other.

7451: So we have to put $n_1=n_2, n_3=n_4$ for such cases,

7452: and the number of independent equations on $n_{1,3,5,6,7,8}$

7453: turns out to be reduced by one.

7454: 

7455: The total number of vacua is then calculated as a sum of a certain

7456: combinatoric factor over all the solutions $n_{i}$ of the above

7457: equations.

7458: The results are summarized below.

7459: \begin{center}

7460: {\footnotesize

7461: \begin{tabular}{|c|c|c|c|c|}

7462: \hline

7463: $\!\!B_{1;+++++}\!\!$&

7464: $\!\!B_{1;++-++}\!\!$&

7465: $\!\!B_{1;++--+}\!\!$&

7466: $\!\!B_{1;++---}\!\!$&

7467: $\!\!B_{1;+-***}\!\!$ \\

7468: \hline

7469: 28956442028638 & 1093287843 & 654 & 0 & 1

7470: \\ \hline

7471: \end{tabular}}

7472: \end{center}

7473: 

7474: \subsubsection{Distribution of gauge group rank}

7475: 

7476: Let us see the distribution of the rank of gauge group over

7477: the supersymmetric vacua.

7478: For the brane configuration with gauge group

7479: $G=\prod O(N_{j^+})\times\prod U(N_{j^0})\times\prod Sp(N_{j^-})$,

7480: the rank is counted as

7481: $$

7482:  \sum_{j^+}\left[\frac{N_{j^+}}{2}\right]

7483: +\sum_{j^0}N_{j^0}

7484: +\sum_{j^-}N_{j^-}.

7485: $$

7486: The results are summarized in the table below.

7487: 

7488: As the table shows, in the type I cases the rank of the gauge group

7489: is peaked around $9$, differently from the quintic case.

7490: Also, the number of vacua with low ranks are more strongly suppressed

7491: because of the presence of $U$ or $Sp$ type branes.

7492: For other orientifolds the maximum allowed rank is reduced, and

7493: the distributions are peaked at lower values of rank. 

7494: 

7495: 

7496: {\scriptsize

7497: \begin{center}

7498: \begin{tabular}{|c||r|r|r|r|r|r|r|}

7499: \hline

7500: \lw{rank}

7501:  &$P_{0;+++++}$

7502:  &$P_{0;+-+++}$

7503:  &$P_{0;++-++}$

7504:  &$P_{0;+--++}$

7505:  &$P_{0;++--+}$

7506:  &$P_{0;+---+}$

7507:  &$P_{0;++---}$ \\

7508: &13213511375147

7509: &47803952

7510: &434841441

7511: &1051

7512: &2162

7513: &35

7514: &148

7515: \\ \hline\hline

7516:  0

7517: &646540

7518: &508725

7519: &4926687

7520: &166

7521: &480

7522: &0

7523: &15

7524: \\

7525:  1

7526: &44771470

7527: &2554170

7528: &29783246

7529: &330

7530: &721

7531: &7

7532: &105

7533: \\

7534:  2

7535: &1031791551

7536: &7173709

7537: &76613078

7538: &397

7539: &719

7540: &28

7541: &28

7542: \\

7543:  3

7544: &11643923756

7545: &11188898

7546: &113881856

7547: &88

7548: &172

7549: &&\\

7550:  4

7551: &75080785790

7552: &11195422

7553: &102828964

7554: &70

7555: &70

7556: &&\\

7557:  5

7558: &302754231919

7559: &8532104

7560: &66661000

7561: &&&&\\

7562:  6

7563: &816375589073

7564: &4126724

7565: &28541380

7566: &&&&\\

7567:  7

7568: &1555478380691

7569: &1860600

7570: &9347940

7571: &&&&\\

7572:  8

7573: &2202010164391

7574: &501900

7575: &1980090

7576: &&&&\\

7577:  9

7578: &2424675084374

7579: &129360

7580: &244860

7581: &&&&\\

7582: 10

7583: &2159636846181

7584: &32340

7585: &32340

7586: &&&&\\

7587: 11

7588: &1607633137394

7589: &&&&&&\\

7590: 12

7591: &1023393658328

7592: &&&&&&\\

7593: 13

7594: &567624907070

7595: &&&&&&\\

7596: 14

7597: &277143210040

7598: &&&&&&\\

7599: 15

7600: &120183191993

7601: &&&&&&\\

7602: 16

7603: &46373508969

7604: &&&&&&\\

7605: 17

7606: &15919273033

7607: &&&&&&\\

7608: 18

7609: &4851273490

7610: &&&&&&\\

7611: 19

7612: &1288731061

7613: &&&&&&\\

7614: 20

7615: &300818948

7616: &&&&&&\\

7617: 21

7618: &56875115

7619: &&&&&&\\

7620: 22

7621: &9505650

7622: &&&&&&\\

7623: 23

7624: &961928

7625: &&&&&&\\

7626: 24

7627: &106392

7628: &&&&&&\\

7629: \hline

7630: \end{tabular}

7631: \end{center}

7632: }

7633: 

7634: {\scriptsize

7635: \begin{center}

7636: \begin{tabular}{|c||r|r|r|}

7637: \hline

7638: \lw{rank}

7639:  &$P_{1;+++++}$

7640:  &$P_{1;++-++}$

7641:  &$P_{1;++--+}$ \\

7642: &28956442028638

7643: &1093287843

7644: &654

7645: \\ \hline\hline

7646:  0

7647: &131418

7648: &10073409

7649: &70

7650: \\

7651:  1

7652: &12060448

7653: &98476432

7654: &448

7655: \\

7656:  2

7657: &355100152

7658: &281607952

7659: &136

7660: \\

7661:  3

7662: &5191991568

7663: &398177360

7664: &\\

7665:  4

7666: &44410530386

7667: &248315690

7668: &\\

7669:  5

7670: &245511738472

7671: &52357760

7672: &\\

7673:  6

7674: &928376315288

7675: &4279240

7676: &\\

7677:  7

7678: &2485035106608

7679: &&\\

7680:  8

7681: &4801648669394

7682: &&\\

7683:  9

7684: &6693313716784

7685: &&\\

7686: 10

7687: &6689330341632

7688: &&\\

7689: 11

7690: &4555609978656

7691: &&\\

7692: 12

7693: &2009612464368

7694: &&\\

7695: 13

7696: &457636777344

7697: &&\\

7698: 14

7699: &40397106120

7700: &&\\

7701: \hline

7702: \end{tabular}

7703: \end{center}

7704: }

7705: 

7706: \subsection{Particle Spectrum in Some Supersymmetric Models}

7707: \label{spectrum}

7708: 

7709: A closer look at the annulus and the M\"{o}bius strip

7710: amplitudes gives us a more detailed information on the spectrum

7711: of open strings.

7712: Let us now turn to count the number of matter fields between

7713: the same branes.

7714: There are many branes and each has quite a few scalar fields, so

7715: the best way to count them is again to use computers.

7716: As in the analysis of A-branes, the massless chiral fields in the

7717: spacetime theory are in one to one correspondence with the

7718: chiral primary states in the internal theory.

7719: Let us summarize the necessary materials.

7720: \begin{itemize}

7721: \item

7722: The open string NS state $\otimes_i(l_i,n_i,s_i)$ between the branes

7723: $B^B_{{\bf L},M}$ and $B^B_{{\bf L}',M'}$ obey

7724: $\sum_in_iw_i = M'-M~({\rm mod}~ 2H)$, as well

7725: as the usual selection rule $l_i+n_i\in 2\Z$ and the $SU(2)$ fusion

7726: rule constraint.

7727: The states between short-orbit branes with the same $\II$

7728: and additional labels $\psi, \psi'=\pm1$ are subject to a projection

7729: $\prod_{i\in \bfS}(-1)^{\frac12(l_i+n_i)}=\psi\psi'$.

7730: \item

7731: The open string states on parity-invariant D-branes have definite

7732: eigenvalues of parity. The M\"{o}bius strip amplitude between

7733: $B_{{\bf L},M}$ and its image under $P^B_{M_\omega;\epsilon_i}$

7734: contains NS states $\otimes_i(l_i,2\nu_i,s_i)$ satisfying

7735: $\sum_i\nu_iw_i=M_\omega-M$ (mod $H$), as well as

7736: the $SU(2)$ fusion constraint on $l_i$.

7737: The contribution of massless states to M\"{o}bius strip amplitude

7738: is given by the sum of chiral primary states satisfying these condition,

7739: with the phase

7740: $$

7741:  i^{-1+|\epsilon|+\{\#{\rm of}(s_i=2)\}}\prod_i\epsilon_i^{L_i+\nu_i}.

7742: $$

7743: \end{itemize}

7744: 

7745: 

7746: 

7747: 

7748: We present here the relevant amplitudes from which the result follows.

7749: The annulus amplitudes between two long-orbit B-branes

7750: $B^B_{{\bf L},M}$ and $B^B_{{\bf L}',M'}$ has the following 

7751: NS part:

7752: \beqa

7753: \lefteqn{\langle B^B_{{\bf L},M}|q^H|B^B_{{\bf L}',M'}\rangle|_{\NS}}

7754:  \nn\\ &=&

7755: \frac12\sum_{n_i}

7756: \delta^{(2H)}_{\sum_iw_in_i+M-M'}

7757: \sum_{l_i}\prod_{i=1}^rN_{L_iL'_i}^{~l_i}

7758: \times \left\{

7759:  \chi^{(\rm st)NS+}\prod_{i=1}^r\chi^{\rm NS+}_{l_i,n_i}

7760: -\chi^{(\rm st)NS-}\prod_{i=1}^r\chi^{\rm NS-}_{l_i,n_i}\right\}

7761: \eeqa

7762: Here $\chi^{(\rm st)NS\pm}$ and

7763: $\chi^{\rm NS\pm}_{l,n}=\chi_{l,n,0}\pm\chi_{l,n,2}$ are the same

7764: as those used in the discussion of A-branes.

7765: The delta symbol represents that the sum over $n_i$ is

7766: taken over the $\tilGamma$-orbit but is shifted by $M$ and $M'$.

7767: For two short-orbit branes $\widehat{B}^B_{{\bf L},M}$ and

7768: $\widehat{B}^B_{{\bf L}',M'}$ the above amplitude has to be divided

7769: by $2^{[|\bfS|/2]}2^{[|\bfS'|/2]}$, where $\bfS$ and $\bfS'$ are

7770: the sets of $i$'s such that $L_i$ or $L'_i$ coincide with $\frac{k_i}{2}$.

7771: When $\bfS=\bfS'$ and $|\bfS|=2$ or $4$, the twisted parts of boundary

7772: states yield

7773: \beqa

7774: &&

7775: \pm\frac{1}{2^{1+|\bfS|}}\sum_{n_i}

7776: \delta^{(2H)}_{\sum_iw_in_i+M-M'}

7777: \sum_{l_i}\prod_{i=1}^rN_{L_iL'_i}^{~l_i}\times

7778: \prod_{i\in\bfS}(-1)^{\frac12(l_i+n_i)}

7779: \nn\\ && \hskip20mm

7780: \times \left\{

7781:  \chi^{(\rm st)NS+}\prod_{i=1}^r\chi^{\rm NS+}_{l_i,n_i}

7782: -\chi^{(\rm st)NS-}\prod_{i=1}^r\chi^{\rm NS-}_{l_i,n_i}\right\},

7783: \eeqa

7784: so the states with $\prod_{i\in\bfS}(-1)^{\frac12(l_i+n_i)}=1\,(-1)$

7785: propagates between the branes with the same (opposite) signs.

7786: All these maintain the integrality of the open string spectrum, and

7787: ensure that every single B-brane supports a $U(1)$ gauge symmetry

7788: in the absence of orientifolds.

7789: 

7790: For parity-invariant branes, we have to find the

7791: action of parity on the matter fields on their worldvolume.

7792: The NS part of M\"{o}bius strip amplitude between a long-orbit B-brane

7793: $B^B_{{\bf L},M}$ and its image under the parity $P^B_{\omega,{\bf m}}$

7794: reads (recall $\omega = \e^{\frac{2\pi i M_\omega}{H}}$ and

7795:        $\tilomega_i= \e^{-\frac{2\pi i m_i}{k_i+2}}=\pm1$)

7796: \beqa

7797: \lefteqn{\langle B^B_{{\bf L},M}|q^H|C^B_{\omega,{\bf m}}\rangle|_{\NS}}

7798:  \nn\\ &=&

7799:   {\rm Re}\left\{

7800:    i\e^{\frac{\pi i(r-d)}{4}}\tilomega^{{\bf L}-\frac12}

7801:    \sum_{\tilnu_i\in\Z_{k_i+2}}

7802:    \delta^{(1)}_{\sum_i\frac{\tilnu_i}{k_i+2},\frac{M_\omega-M}{H}}

7803:    \sum_{l_i}\tilomega^{\tilnu}\hat\chi^{(\rm st)NS+}

7804:    \prod_{i=1}^rN_{L_iL_i}^{~l_i}\hat

7805:    \chi^{\rm NS+}_{l_i,2\tilnu_i}

7806:   \right\}

7807: \eeqa

7808: where the characters $\hat\chi^{(\rm st)NS\pm}$, $\hat\chi^{\rm NS\pm}_{l,n}$

7809: are the same as those appeared in the discussion of A-branes.

7810: The expressions are the same for short-orbit branes except for obvious

7811: change of normalizations.

7812: The states contributing to the above M\"{o}bius strip amplitude

7813: with $+(-)$ signs are the eigenstates of the parity

7814: with eigenvalues $P^B_{\omega;{\bf m}}=+1(-1)$.

7815: We also see that, in the $q_l$-expansion of this amplitude,

7816: the terms of order $q_l^0$ corresponding to gauge fields

7817: appear with the sign $-\epsilon^{\omega;{\bf m}}_{{\bf L},M}$.

7818: 

7819: \subsubsection{Quintic}

7820: 

7821: %Ilka

7822: For the quintic, there exists only a single parity of interest. 

7823: We are considering  supersymmetry preserving branes,

7824: which in the case of the quintic are invariant under the parity.

7825: The following

7826: table lists the number of massless scalars on these branes

7827: and their transformation properties

7828: under parity: $(n_1,n_2)$ denotes the number of (symmetric, antisymmetric)

7829: massless scalars.

7830: \begin{center}

7831: {\scriptsize

7832: \begin{tabular}{|c||c|}

7833: \hline

7834: ($L_i$)&$(n_1,n_2)$\\

7835: \hline

7836: (00000)&$(0,0)$\\

7837: (10000)&$(4,0)$\\

7838: (11000)&$(8,3)$\\

7839: (11100)&$(15,9)$\\

7840: (11110)&$(28,22)$\\

7841: (11111)&$(51,50)$\\

7842: \hline

7843: \end{tabular}}

7844: \end{center}

7845: It is now straightforward

7846: to find the matter content of supersymmetric tadpole canceling

7847: configurations. Two such solutions have been given in \cite{Blumenhagen}, 

7848: the standard solution with $4$ branes of type $(11111)$

7849: and the one with $12$ branes of type $(00000)$ and $20$ of type $(10000)$,

7850: which is the configuration with the highest possible rank in this example.

7851: Just for the purposes of illustration, we give a 

7852: third configuration, which is chosen completely randomly. We consider a

7853: setup consisting of $4$ branes of type $(11000)$ and $8$ of $(00111)$. 

7854: The matter content

7855: under the gauge group $O(4) \times O(8)$ is

7856: \beq

7857: 8({\bf 10},{\bf 1} ) \oplus 3({\bf 6},\bf{1}) 

7858: \oplus 15 ({\bf 1},{\bf 36}) \oplus 9({\bf 1},{\bf 28}) \oplus 101 ({\bf 4}, {\bf 8}). 

7859: \eeq

7860: The part of the spectrum involving only one type of brane can be directly

7861: read off from the above table. For those strings that connect one type

7862: of brane to another, note that there is always a linear combination of

7863: any open string operator and its parity image that survives the projection.

7864: As a consequence,this part of the spectrum can be determined using the results

7865: on the open string spectrum without orientifolds.

7866: 

7867: 

7868: 

7869: %end Ilka

7870: 

7871: 

7872: \subsubsection{Two parameter model}

7873: 

7874: 

7875: The tables \ref{cp-2pb1} and \ref{cp-2pb2} list the  gauge groups

7876: $G$ and the number of matters on the D-branes for various choices

7877: of orientifolds of the two parameter model.

7878: For $G=O$ or $Sp$, the two numbers in $G_{(n_1,n_2)}$ mean there

7879: are massless scalars in $n_1$ symmetric and $n_2$ antisymmetric

7880: tensor representations of $G$.

7881: For $G=U$ the three numbers in $U_{(n_1,n_2,n_3)}$ mean

7882: there are $n_1$ adjoint, $n_2$ symmetric tensor and $n_3$ antisymmetric

7883: tensor representations.

7884: 

7885: %\begin{table}

7886: \begin{center}

7887: {\scriptsize

7888: \begin{tabular}{|c||l|l|l|l|l|l|}

7889: \hline

7890: ($L_i$)&$P_{0;+++++}$&$P_{0;++-++}$&$P_{0;++++-}$

7891:        &$P_{0;++--+}$&$P_{0;+++--}$&$P_{0;++---}$\\

7892: \hline

7893: (00000)&$O_{(0,0)}$&$U_{(0,2,1)}$&$U_{(0,2,1)}$&$Sp_{(0,0)}$&$Sp_{(0,0)}$&$U_{(0,3,0)}$\\

7894: (00100)&$O_{(2,1)}$&$O_{(1,2)}$&$Sp_{(2,1)}$&$O_{(2,1)}$&$Sp_{(3,0)}$&$Sp_{(0,3)}$\\

7895: (00110)&$U_{(2,1,0)}$&$O_{(1,1)}$&$Sp_{(2,0)}$&$U_{(2,0,1)}$&$U_{(2,0,1)}$&$O_{(2,0)}$\\

7896: (00111)&$Sp_{(3,0)}$&$O_{(1,2)}$&$O_{(1,2)}$&$O_{(2,1)}$&$O_{(2,1)}$&$Sp_{(0,3)}$\\

7897: \hline

7898: (11000)&$O_{(5,0)}$&$U_{(5,4,6)}$&$U_{(5,4,6)}$&$Sp_{(2,3)}$&$Sp_{(2,3)}$&$U_{(5,4,6)}$\\

7899: (11100)&$O_{(9,6)}$&$O_{(11,4)}$&$Sp_{(6,9)}$&$O_{(7,8)}$&$Sp_{(6,9)}$&$Sp_{(8,7)}$\\

7900: (11110)&$U_{(9,3,3)}$&$O_{(6,3)}$&$Sp_{(3,6)}$&$U_{(9,5,1)}$&$U_{(9,3,3)}$&$O_{(4,5)}$\\

7901: (11111)&$Sp_{(6,9)}$&$O_{(9,6)}$&$O_{(9,6)}$&$O_{(9,6)}$&$O_{(9,6)}$&$Sp_{(10,5)}$\\

7902: \hline

7903: (20000)&$O_{(4,0)}$&$U_{(4,3,4)}$&$U_{(4,3,4)}$&$Sp_{(2,2)}$&$Sp_{(2,2)}$&$U_{(4,4,3)}$\\

7904: (20100)&$O_{(7,4)}$&$O_{(8,3)}$&$Sp_{(5,6)}$&$O_{(5,6)}$&$Sp_{(6,5)}$&$Sp_{(5,6)}$\\

7905: (20110)&$U_{(6,3,2)}$&$O_{(4,2)}$&$Sp_{(3,3)}$&$U_{(6,4,1)}$&$U_{(6,2,3)}$&$O_{(3,3)}$\\

7906: (20111)&$Sp_{(6,5)}$&$O_{(6,5)}$&$O_{(6,5)}$&$O_{(7,4)}$&$O_{(7,4)}$&$Sp_{(7,4)}$\\

7907: \hline

7908: (22000)&$O_{(13,3)}$&$U_{(16,7,12)}$&$U_{(16,7,12)}$&$Sp_{(7,9)}$&$Sp_{(7,9)}$&$U_{(16,6,13)}$\\

7909: (22100)&$O_{(20,15)}$&$O_{(25,10)}$&$Sp_{(14,21)}$&$O_{(16,19)}$&$Sp_{(13,22)}$&$Sp_{(20,15)}$\\

7910: (22110)&$U_{(18,8,9)}$&$O_{(12,6)}$&$Sp_{(5,13)}$&$U_{(18,13,4)}$&$U_{(18,9,8)}$&$O_{(7,11)}$\\

7911: (22111)&$Sp_{(13,22)}$&$O_{(21,14)}$&$O_{(21,14)}$&$O_{(20,15)}$&$O_{(20,15)}$&$Sp_{(24,11)}$\\

7912: \hline

7913: (31000)&$O_{(9,6)}$&$Sp_{(6,9)}$&$Sp_{(6,9)}$&$Sp_{(6,9)}$&$Sp_{(6,9)}$&$O_{(5,10)}$\\

7914: (31100)&$U_{(9,3,3)}$&$O_{(6,3)}$&$Sp_{(3,6)}$&$U_{(9,3,3)}$&$U_{(9,1,5)}$&$Sp_{(5,4)}$\\

7915: (31110)&$Sp_{(6,9)}$&$O_{(9,6)}$&$Sp_{(4,11)}$&$O_{(9,6)}$&$Sp_{(8,7)}$&$O_{(7,8)}$\\

7916: (31111)&$Sp_{(0,5)}$&$U_{(5,6,4)}$&$U_{(5,6,4)}$&$O_{(3,2)}$&$O_{(3,2)}$&$U_{(5,6,4)}$\\

7917: \hline

7918: (33000)&$U_{(9,3,3)}$&$Sp_{(3,6)}$&$Sp_{(3,6)}$&$U_{(9,1,5)}$&$U_{(9,1,5)}$&$O_{(0,9)}$\\

7919: (33100)&$Sp_{(6,9)}$&$O_{(9,6)}$&$Sp_{(4,11)}$&$Sp_{(8,7)}$&$O_{(1,14)}$&$Sp_{(10,5)}$\\

7920: (33110)&$Sp_{(0,5)}$&$U_{(5,6,4)}$&$U_{(5,0,10)}$&$O_{(3,2)}$&$Sp_{(4,1)}$&$U_{(5,4,6)}$\\

7921: (33111)&$Sp_{(0,15)}$&$Sp_{(10,5)}$&$Sp_{(10,5)}$&$O_{(7,8)}$&$O_{(7,8)}$&$O_{(9,6)}$\\

7922: \hline

7923: \end{tabular}}

7924: \end{center}

7925: %\end{table}

7926: 

7927: \begin{table}[htb]

7928: \begin{center}

7929: {\scriptsize

7930: \begin{tabular}{|c||l|l|l|l|l|l|}

7931: \hline

7932: ($L_i$)&$P_{1;+++++}$&$P_{1;++-++}$&$P_{1;++++-}$

7933:        &$P_{1;++--+}$&$P_{1;+++--}$&$P_{1;++---}$\\

7934: \hline

7935: (10000)&$O_{(1,0)}$&$U_{(1,3,3)}$&$U_{(1,3,3)}$&$Sp_{(0,1)}$&$Sp_{(0,1)}$&$U_{(1,3,3)}$\\

7936: (10100)&$O_{(4,3)}$&$O_{(4,3)}$&$Sp_{(3,4)}$&$O_{(4,3)}$&$Sp_{(3,4)}$&$Sp_{(3,4)}$\\

7937: (10110)&$U_{(5,1,1)}$&$O_{(3,2)}$&$Sp_{(2,3)}$&$U_{(5,1,1)}$&$U_{(5,1,1)}$&$O_{(3,2)}$\\

7938: (10111)&$Sp_{(3,4)}$&$O_{(4,3)}$&$O_{(4,3)}$&$O_{(4,3)}$&$O_{(4,3)}$&$Sp_{(3,4)}$\\

7939: \hline

7940: (21000)&$O_{(9,0)}$&$U_{(9,5,9)}$&$U_{(9,5,9)}$&$Sp_{(4,5)}$&$Sp_{(4,5)}$&$U_{(9,5,9)}$\\

7941: (21100)&$O_{(14,9)}$&$O_{(18,5)}$&$Sp_{(9,14)}$&$O_{(10,13)}$&$Sp_{(9,14)}$&$Sp_{(13,10)}$\\

7942: (21110)&$U_{(13,5,5)}$&$O_{(9,4)}$&$Sp_{(4,9)}$&$U_{(13,9,1)}$&$U_{(13,5,5)}$&$O_{(5,8)}$\\

7943: (21111)&$Sp_{(9,14)}$&$O_{(14,9)}$&$O_{(14,9)}$&$O_{(14,9)}$&$O_{(14,9)}$&$Sp_{(17,6)}$\\

7944: \hline

7945: (30000)&$O_{(4,3)}$&$Sp_{(3,4)}$&$Sp_{(3,4)}$&$Sp_{(3,4)}$&$Sp_{(3,4)}$&$O_{(4,3)}$\\

7946: (30100)&$U_{(5,1,1)}$&$O_{(3,2)}$&$Sp_{(2,3)}$&$U_{(5,1,1)}$&$U_{(5,1,1)}$&$Sp_{(2,3)}$\\

7947: (30110)&$Sp_{(3,4)}$&$O_{(4,3)}$&$Sp_{(3,4)}$&$O_{(4,3)}$&$Sp_{(3,4)}$&$O_{(4,3)}$\\

7948: (30111)&$Sp_{(0,1)}$&$U_{(1,3,3)}$&$U_{(1,3,3)}$&$O_{(1,0)}$&$O_{(1,0)}$&$U_{(1,3,3)}$\\

7949: \hline

7950: (32000)&$O_{(14,9)}$&$Sp_{(9,14)}$&$Sp_{(9,14)}$&$Sp_{(9,14)}$&$Sp_{(9,14)}$&$O_{(6,17)}$\\

7951: (32100)&$U_{(13,5,5)}$&$O_{(9,4)}$&$Sp_{(4,9)}$&$U_{(13,5,5)}$&$U_{(13,1,9)}$&$Sp_{(8,5)}$\\

7952: (32110)&$Sp_{(9,14)}$&$O_{(14,9)}$&$Sp_{(5,18)}$&$O_{(14,9)}$&$Sp_{(13,10)}$&$O_{(10,13)}$\\

7953: (32111)&$Sp_{(0,9)}$&$U_{(9,9,5)}$&$U_{(9,9,5)}$&$O_{(5,4)}$&$O_{(5,4)}$&$U_{(9,9,5)}$\\

7954: \hline

7955: \end{tabular}}

7956: \end{center}

7957: \caption{Gauge group and number of massless scalar fields}

7958: \label{cp-2pb1}

7959: \end{table}

7960: 

7961: %\begin{table}

7962: \begin{center}

7963: {\scriptsize

7964: \begin{tabular}{|c||l|l|l|l|l|l|}

7965: \hline

7966: ($L_i$)&$P_{0;+-+++}$&$P_{0;+--++}$&$P_{0;+-++-}$

7967:        &$P_{0;+---+}$&$P_{0;+-+--}$&$P_{0;+----}$\\

7968: \hline

7969: (00000)&$U_{(0,0,3)}$&$Sp_{(0,0)}$&$Sp_{(0,0)}$&$U_{(0,1,2)}$&$U_{(0,1,2)}$&$O_{(0,0)}$\\

7970: (00100)&$Sp_{(0,3)}$&$O_{(0,3)}$&$Sp_{(1,2)}$&$Sp_{(2,1)}$&$O_{(1,2)}$&$Sp_{(1,2)}$\\

7971: (00110)&$Sp_{(0,2)}$&$U_{(2,0,1)}$&$U_{(2,0,1)}$&$O_{(0,2)}$&$Sp_{(1,1)}$&$U_{(2,1,0)}$\\

7972: (00111)&$Sp_{(0,3)}$&$Sp_{(1,2)}$&$Sp_{(1,2)}$&$O_{(1,2)}$&$O_{(1,2)}$&$O_{(0,3)}$\\

7973: \hline

7974: (11000)&$U_{(5,10,0)}$&$O_{(1,4)}$&$O_{(1,4)}$&$U_{(5,6,4)}$&$U_{(5,6,4)}$&$Sp_{(2,3)}$\\

7975: (11100)&$O_{(11,4)}$&$Sp_{(14,1)}$&$O_{(7,8)}$&$O_{(5,10)}$&$Sp_{(8,7)}$&$O_{(9,6)}$\\

7976: (11110)&$O_{(6,3)}$&$U_{(9,5,1)}$&$U_{(9,3,3)}$&$Sp_{(9,0)}$&$O_{(4,5)}$&$U_{(9,1,5)}$\\

7977: (11111)&$O_{(9,6)}$&$O_{(9,6)}$&$O_{(9,6)}$&$Sp_{(10,5)}$&$Sp_{(10,5)}$&$Sp_{(14,1)}$\\

7978: \hline

7979: (20000)&$U_{(4,1,6)}$&$Sp_{(2,2)}$&$Sp_{(2,2)}$&$U_{(4,2,5)}$&$U_{(4,2,5)}$&$O_{(0,4)}$\\

7980: (20100)&$Sp_{(3,8)}$&$O_{(3,8)}$&$Sp_{(4,7)}$&$Sp_{(7,4)}$&$O_{(2,9)}$&$Sp_{(6,5)}$\\

7981: (20110)&$Sp_{(1,5)}$&$U_{(6,2,3)}$&$U_{(6,0,5)}$&$O_{(1,5)}$&$Sp_{(4,2)}$&$U_{(6,3,2)}$\\

7982: (20111)&$Sp_{(1,10)}$&$Sp_{(6,5)}$&$Sp_{(6,5)}$&$O_{(4,7)}$&$O_{(4,7)}$&$O_{(3,8)}$\\

7983: \hline

7984: (22000)&$U_{(16,9,10)}$&$Sp_{(7,9)}$&$Sp_{(7,9)}$&$U_{(16,8,11)}$&$U_{(16,8,11)}$&$O_{(5,11)}$\\

7985: (22100)&$Sp_{(16,19)}$&$O_{(18,17)}$&$Sp_{(15,20)}$&$Sp_{(18,17)}$&$O_{(13,22)}$&$Sp_{(19,16)}$\\

7986: (22110)&$Sp_{(7,11)}$&$U_{(18,9,8)}$&$U_{(18,5,12)}$&$O_{(9,9)}$&$Sp_{(10,8)}$&$U_{(18,8,9)}$\\

7987: (22111)&$Sp_{(12,23)}$&$Sp_{(19,16)}$&$Sp_{(19,16)}$&$O_{(17,18)}$&$O_{(17,18)}$&$O_{(18,17)}$\\

7988: \hline

7989: (31000)&$O_{(11,4)}$&$O_{(7,8)}$&$O_{(7,8)}$&$Sp_{(8,7)}$&$Sp_{(8,7)}$&$Sp_{(8,7)}$\\

7990: (31100)&$O_{(6,3)}$&$U_{(9,5,1)}$&$U_{(9,3,3)}$&$O_{(4,5)}$&$Sp_{(5,4)}$&$U_{(9,3,3)}$\\

7991: (31110)&$O_{(9,6)}$&$O_{(9,6)}$&$Sp_{(8,7)}$&$Sp_{(10,5)}$&$O_{(7,8)}$&$O_{(7,8)}$\\

7992: (31111)&$U_{(5,6,4)}$&$O_{(3,2)}$&$O_{(3,2)}$&$U_{(5,6,4)}$&$U_{(5,6,4)}$&$Sp_{(4,1)}$\\

7993: \hline

7994: (33000)&$O_{(6,3)}$&$U_{(9,3,3)}$&$U_{(9,3,3)}$&$Sp_{(5,4)}$&$Sp_{(5,4)}$&$U_{(9,5,1)}$\\

7995: (33100)&$O_{(9,6)}$&$O_{(9,6)}$&$Sp_{(8,7)}$&$O_{(7,8)}$&$Sp_{(10,5)}$&$Sp_{(6,9)}$\\

7996: (33110)&$U_{(5,6,4)}$&$O_{(3,2)}$&$Sp_{(4,1)}$&$U_{(5,6,4)}$&$U_{(5,4,6)}$&$O_{(3,2)}$\\

7997: (33111)&$Sp_{(10,5)}$&$O_{(7,8)}$&$O_{(7,8)}$&$O_{(9,6)}$&$O_{(9,6)}$&$Sp_{(8,7)}$\\

7998: \hline

7999: \end{tabular} }

8000: \end{center}

8001: %\end{table}

8002: \begin{table}[htb]

8003: \begin{center}

8004: {\scriptsize

8005: \begin{tabular}{|c||l|l|l|l|l|l|}

8006: \hline

8007: ($L_i$)&$P_{1;+-+++}$&$P_{1;+--++}$&$P_{1;+-++-}$

8008:        &$P_{1;+---+}$&$P_{1;+-+--}$&$P_{1;+----}$\\

8009: \hline

8010: (10000)&$U_{(1,0,6)}$&$Sp_{(1,0)}$&$Sp_{(1,0)}$&$U_{(1,2,4)}$&$U_{(1,2,4)}$&$O_{(0,1)}$\\

8011: (10100)&$Sp_{(1,6)}$&$O_{(0,7)}$&$Sp_{(3,4)}$&$Sp_{(5,2)}$&$O_{(2,5)}$&$Sp_{(3,4)}$\\

8012: (10110)&$Sp_{(1,4)}$&$U_{(5,0,2)}$&$U_{(5,0,2)}$&$O_{(0,5)}$&$Sp_{(3,2)}$&$U_{(5,2,0)}$\\

8013: (10111)&$Sp_{(1,6)}$&$Sp_{(3,4)}$&$Sp_{(3,4)}$&$O_{(2,5)}$&$O_{(2,5)}$&$O_{(0,7)}$\\

8014: \hline

8015: (21000)&$U_{(9,10,4)}$&$O_{(4,5)}$&$O_{(4,5)}$&$U_{(9,8,6)}$&$U_{(9,8,6)}$&$Sp_{(5,4)}$\\

8016: (21100)&$O_{(14,9)}$&$Sp_{(15,8)}$&$O_{(12,11)}$&$O_{(10,13)}$&$Sp_{(13,10)}$&$O_{(12,11)}$\\

8017: (21110)&$O_{(8,5)}$&$U_{(13,6,4)}$&$U_{(13,6,4)}$&$Sp_{(9,4)}$&$O_{(6,7)}$&$U_{(13,4,6)}$\\

8018: (21111)&$O_{(14,9)}$&$O_{(12,11)}$&$O_{(12,11)}$&$Sp_{(13,10)}$&$Sp_{(13,10)}$&$Sp_{(15,8)}$\\

8019: \hline

8020: (30000)&$Sp_{(1,6)}$&$Sp_{(3,4)}$&$Sp_{(3,4)}$&$O_{(2,5)}$&$O_{(2,5)}$&$O_{(0,7)}$\\

8021: (30100)&$Sp_{(1,4)}$&$U_{(5,0,2)}$&$U_{(5,0,2)}$&$Sp_{(3,2)}$&$O_{(0,5)}$&$U_{(5,2,0)}$\\

8022: (30110)&$Sp_{(1,6)}$&$Sp_{(3,4)}$&$O_{(0,7)}$&$O_{(2,5)}$&$Sp_{(5,2)}$&$Sp_{(3,4)}$\\

8023: (30111)&$U_{(1,0,6)}$&$Sp_{(1,0)}$&$Sp_{(1,0)}$&$U_{(1,2,4)}$&$U_{(1,2,4)}$&$O_{(0,1)}$\\

8024: \hline

8025: (32000)&$Sp_{(9,14)}$&$Sp_{(11,12)}$&$Sp_{(11,12)}$&$O_{(10,13)}$&$O_{(10,13)}$&$O_{(8,15)}$\\

8026: (32100)&$Sp_{(5,8)}$&$U_{(13,4,6)}$&$U_{(13,4,6)}$&$Sp_{(7,6)}$&$O_{(4,9)}$&$U_{(13,6,4)}$\\

8027: (32110)&$Sp_{(9,14)}$&$Sp_{(11,12)}$&$O_{(8,15)}$&$O_{(10,13)}$&$Sp_{(13,10)}$&$Sp_{(11,12)}$\\

8028: (32111)&$U_{(9,4,10)}$&$Sp_{(5,4)}$&$Sp_{(5,4)}$&$U_{(9,6,8)}$&$U_{(9,6,8)}$&$O_{(4,5)}$\\

8029: \hline

8030: \end{tabular}}

8031: \end{center}

8032: \caption{Gauge group and number of massless scalar fields (continued)}

8033: \label{cp-2pb2}

8034: \end{table}

8035: 

8036: Note that the symmetric and antisymmetric tensors of $U(n)$ are

8037: complex representations, and are supported on strings between a

8038: $U$-type brane and its parity image.

8039: The analysis of the spectrum also shows that all the $U$-type branes

8040: in the tables support equal number of (anti)symmetric tensors and

8041: their conjugates, namely, all of them support non-chiral matters.

8042: We will extend this observation later and show the non-chirality

8043: of the spectrum for general supersymmetric brane configurations

8044: in any type IIB orientifolds of Gepner model.

8045: 

8046: \subsubsection{Spectrum in Sample Examples}

8047: 

8048: It is straightforward to determine the matter contents

8049: in sample tadpole canceling configurations.

8050: Let us consider a few examples as an exercise.

8051: 

8052: As the first example, let us take the parity $P^B_{0;+++++}$

8053: and take six branes from the group $\#14$ and two from $\#12$ to

8054: cancel the tadpole.

8055: There are still many ways to do so.

8056: For example, there are the following two inequivalent configurations

8057: supporting $O(6)\times O(2)$ gauge group:

8058: \beqa

8059:  6(22100)+2(20100) &:&

8060: 20({\bf 21,1})\oplus 15({\bf 15,1})\oplus

8061:  7({\bf  1,3})\oplus  4({\bf  1,1})\oplus 12({\bf  6,2}), \nn\\

8062:  6(22100)+2(20010) &:&

8063: 20({\bf 21,1})\oplus 15({\bf 15,1})\oplus

8064:  7({\bf  1,3})\oplus  4({\bf  1,1})\oplus  6({\bf  6,2}).

8065: \eeqa

8066: Here ${\bf 2}$ and ${\bf 3}$ of $O(2)$ mean the reducible

8067: representations ${\bf [1]\oplus[-1]}$ and

8068: ${\bf [2]\oplus[0]\oplus[-2]}$ of $U(1)$.

8069: The configurations supporting $O(6)\times U(1)$ are

8070: \beqa

8071:  6(22100)+(11110)^++(11110)^- &:&

8072: 20({\bf 21})\oplus 15({\bf 15})

8073: \oplus 9({\bf 1})\oplus 3({\bf 1})^{\pm\pm}\oplus 8({\bf 6})^{\pm}

8074: \nn\\

8075:  6(22100)+(11011)^++(11011)^- &:&

8076: 20({\bf 21})\oplus 15({\bf 15})

8077: \oplus 9({\bf 1})\oplus 3({\bf 1})^{\pm\pm}\oplus 2({\bf 6})^{\pm}

8078: \nn\\

8079:  6(22100)+(31100)^++(31100)^- &:&

8080: 20({\bf 21})\oplus 15({\bf 15})

8081: \oplus 9({\bf 1})\oplus 3({\bf 1})^{\pm\pm}\oplus 12({\bf 6})^{\pm}

8082: \nn\\

8083:  6(22100)+(31010)^++(31010)^- &:&

8084: 20({\bf 21})\oplus 15({\bf 15})

8085: \oplus 9({\bf 1})\oplus 3({\bf 1})^{\pm\pm}\oplus 6({\bf 6})^{\pm}

8086: \nn\\

8087:  6(22100)+(33000)^++(33000)^- &:&

8088: 20({\bf 21})\oplus 15({\bf 15})

8089: \oplus 9({\bf 1})\oplus 3({\bf 1})^{\pm\pm}\oplus 10({\bf 6})^{\pm}

8090: \eeqa

8091: Here the $\pm$ signs represent $U(1)$ charge.

8092: $18({\bf 1})$ are neutral scalars corresponding to open strings

8093: with both ends on the same short-orbit brane, while

8094: $({\bf 1})^{\pm\pm}$ correspond to strings stretching between a short-orbit

8095: brane and its parity image.

8096: There are six configurations supporting $O(6)\times Sp(1)$:

8097: \beqa

8098:  6(22100)+2(20111) &:&

8099: 20({\bf 21,1})\oplus 15({\bf 15,1})

8100: \oplus 6({\bf 1,3})\oplus 5({\bf 1,1})\oplus 6({\bf 6,2})

8101: \nn\\

8102:  6(22100)+2(31111) &:&

8103: 20({\bf 21,1})\oplus 15({\bf 15,1})

8104: \oplus 5({\bf 1,1})\oplus 6({\bf 6,2})

8105: \nn\\

8106:  6(22100)+2(33110)^\pm &:&

8107: 20({\bf 21,1})\oplus 15({\bf 15,1})

8108: \oplus 5({\bf 1,1})\oplus 10({\bf 6,2})

8109: \nn\\

8110:  6(22100)+2(33011)^\pm &:&

8111: 20({\bf 21,1})\oplus 15({\bf 15,1})

8112: \oplus 5({\bf 1,1})\oplus 4({\bf 6,2})

8113: \eeqa

8114: There are indeed a lot more tadpole-canceling configurations

8115: with various choices of orientifold and D-branes, and the spectrum

8116: of massless states can be obtained in the same way.

8117: 

8118: 

8119: 

8120: \subsection{Chirality --- Vanishing Theorem}

8121: 

8122: 

8123: As was explained before, chirality of the theory is measured

8124: by the Witten index.

8125: Given a tadpole-free set of an O-plane and D-branes, the theory

8126: is chiral if there is a pair of D-branes with nonzero open string Witten

8127: index, or any D-brane and its parity image with nonzero twisted Witten index.

8128: 

8129: The index between two long-orbit B-branes in Gepner model can be easily

8130: computed as the diagonal elements\footnote

8131:   {All the diagonal elements take the same value because $g$ is

8132:    a $H$-dimensional shift matrix.}

8133: of the following polynomial of the $H$-dimensional shift matrix $g$,

8134: \beq

8135:   Q_{{\bf L},M}(g)Q_{{\bf L}',M'}(g^{-1})\prod_{i=1}^r(1-g^{w_i}),

8136: \label{bindex}

8137: \eeq

8138: where $Q_{{\bf L},M}(g)$ is the polynomial defined in (\ref{branec}).

8139: The parity twisted Witten index is given by replacing one of the two

8140: polynomials with the one representing the O-plane charge, which are

8141: given in (\ref{ochvec1}) and (\ref{ochvec2}).

8142: 

8143: Using the index formula (\ref{bindex}), one can show that any

8144: tadpole-free configurations of an O-plane and long-orbit D-branes

8145: are non-chiral.

8146: To do this, notice first that the polynomial $Q_{{\bf L},M}(g)$

8147: is symmetric under $g^i\to g^{M-i}$.

8148: Similarly, the polynomials representing the O-plane charges

8149: are symmetric under $g^i\to g^{M_O-i}$, where

8150: $M_O=M_\omega+4|\epsilon|$ characterize the spacetime supersymmetry

8151: preserved by the O-planes.

8152: On the other hand, under the assumption $\sum_iw_i=H$ the last factor

8153: in (\ref{bindex}) is transformed to $(-1)^r$ times itself under

8154: $g\to g^{-1}$.

8155: Using all these one finds that, in standard four-dimensional models

8156: with $r=5$, the polynomial (\ref{bindex}) has no $g^0$ term

8157: for any susy-preserving pairs of long-orbit D-branes and the O-plane.

8158: Thus the index vanishes for all such pairs.

8159: 

8160: There is still a possibility of having chiral models with B-type

8161: orientifolds of Gepner models.

8162: The point is that, in some Gepner models with even $H$, there are

8163: RR-charges carried by some short-orbit B-branes and none of

8164: long-orbit B-branes.

8165: In general, the number of RR-charges in type IIB orientifolds

8166: is $2h_{1,1}+2$, and the number of RR-charges carried by

8167: B-branes is fewer than this:

8168: \beqa

8169: 2h_{1,1}+2 &\ge&

8170: \#\mbox{(charges carried by all the short- and long-orbit B-branes)}

8171: \nn\\ &\ge&

8172: \#\mbox{(charges carried by $L_i=0$ B-branes)}.

8173: \eeqa

8174: The first inequality shows that there can be RR-charges carried

8175: by none of rational B-type boundary states constructed in this paper.

8176: It is expected that such RR-charges are associated with non-toric

8177: blowups, as there are non-polynomial deformations in the IIA case.

8178: In the $k=(66222)$ model both of the above equalities hold,

8179: so there is no chiral brane configurations.

8180: 

8181: As an example where neither of the two equalities hold, let us

8182: consider the $k=(22444)$ model which is known to have $h_{1,1}=6$.

8183: Let us first work out the $14=2h_{1,1}+2$ RR ground states.

8184: First, take the RR ground states

8185: $|0\rangle_\nu$ ($\nu=1,\cdots,k+1$) in the level $k$ minimal model

8186: \beq

8187:  |0\rangle_\nu = |\nu-1,\nu,1\rangle \times |\nu-1,\nu,1\rangle,

8188: \eeq

8189: and construct the ground states of the form

8190: $|0\rangle_{(\nu_i)}=\prod_i |0\rangle_{\nu_i}$,

8191: with $\nu_i= \nu$ (mod $k_i+2$) for all $i$.

8192: There are only eight such states:

8193: \beqa

8194: &

8195: |0\rangle_{(11111)},~~

8196: |0\rangle_{(22222)},~~

8197: |0\rangle_{(33333)},~~

8198: |0\rangle_{(33111)},\nn\\ &

8199: |0\rangle_{(11555)},~~

8200: |0\rangle_{(11333)},~~

8201: |0\rangle_{(22444)},~~

8202: |0\rangle_{(33555)}.

8203: \eeqa

8204: Other states are obtained by looking for mixed products

8205: of $|0\rangle_\nu$ and $|l\rangle_\RR$, where

8206: \beq

8207:  |l\rangle_\RR = |l,l+1,1\rangle \times |l,-l-1,-1\rangle.

8208: \eeq

8209: One finds six additional states of the form

8210: $|l_1,l_2\rangle_\RR\times|0\rangle_{(\nu_3\nu_4\nu_5)}$:

8211: \beqa

8212: &

8213: |2,0\rangle_\RR|0\rangle_{(222)},~~

8214: |1,1\rangle_\RR|0\rangle_{(222)},~~

8215: |0,2\rangle_\RR|0\rangle_{(222)},\nn\\

8216: &

8217: |2,0\rangle_\RR|0\rangle_{(444)},~~

8218: |1,1\rangle_\RR|0\rangle_{(444)},~~

8219: |0,2\rangle_\RR|0\rangle_{(444)}.

8220: \eeqa

8221: (Note that the state $|1,1\rangle_\RR|0\rangle_{(222)}$ is different

8222:  from $|0\rangle_{(22222)}$, although they are labeled by the same

8223:  quantum numbers. Recall that in Gepner model certain closed string

8224:  states appear more than once in the toroidal partition function,

8225:  and we should distinguish them as they are sitting in

8226:  different twisted sectors.)

8227: The $L_i=0$ B-branes can only couple to the first eight states,

8228: and the short-orbit branes with $L_1=L_2=1$ couple also to the

8229: the two states with $l_1=l_2=1$ in the second group.

8230: The remaining four RR ground states have no overlaps

8231: with any B-branes.

8232: 

8233: Unfortunately, one can also show that the index vanishes

8234: for pairs of short-orbit branes with these extra RR charges,

8235: using a similar index formula as before.

8236: Let us take two short-orbit B-branes with the same $\bfS$

8237: of even order.

8238: As was given in (\ref{fixed1}) and (\ref{fixed2}), the

8239: boundary states are sums of two terms orthogonal to each other.

8240: So the index is also a sum of two terms,

8241: one of which is $2^{-|\bfS|}$ times the expression for long-orbit

8242: branes (\ref{bindex}) and the other represents the new contribution

8243: \beq

8244:  2^{-|\bfS|}

8245:  \widetilde{Q}_{{\bf L},M}(g)\widetilde{Q}_{{\bf L}',M'}(g^{-1})

8246:  \prod_{i\;/\!\!\!\!\in\bfS}(1-g^{w_i})

8247:  \prod_{i\in\bfS}(1+g^{w_i}+\cdots+g^{w_i(k_i+1)}),

8248: \eeq

8249: where

8250: \beq

8251:  \widetilde{Q}_{{\bf L},M}(g)= g^{M/2}\prod_{i\;/\!\!\!\!\in\bfS}

8252:  \left(\sum_{j_i=0}^{L_i}g^{w_i(\frac{L_i}{2}-j_i)}\right).

8253: \eeq

8254: Using the symmetry or antisymmetry of each factor under the

8255: inversion $g\to g^{-1}$ one finds that no supersymmetry-preserving

8256: pair of short-orbit B-branes can have non-zero index.

8257: 

8258: Thus, we find

8259: 

8260: \noindent

8261: {\bf Theorem:}\\

8262: {\it The index of any pair

8263: of branes in a tadpole canceling and supersymmetric rational brane

8264: configuration vanishes in Type IIB orientifolds of Gepner models.

8265: In particular, there is no chiral and supersymmetric theory

8266: in this class of solutions.}

8267: 

8268: 

8269: \noindent

8270: {\bf Remarks.}\\{\small

8271: {\bf (i)}~This theorem applies only to the Gepner model obtained as

8272: the orbifold of the product of minimal models by a

8273: {\it single} cyclic group $\Z_H$, and may not hold for

8274: orbifolds with more than one cyclic group factors.

8275: For example, Type IIA models we considered in

8276: Section~\ref{sec:TCCA} is nothing but Type IIB models on

8277: orbifolds with four cyclic group factors \cite{GreenePlesser},

8278: and we indeed found

8279: chiral supersymmetric models there.

8280: Actually there is an existence proof of chiral model

8281: if the orbifold group has two cyclic factors (next to minimal):

8282: In Appendix~\ref{orbifold}, we analyze

8283: the condition for a Type IIB orientifold of

8284: the model $M_3^5/\Z_5\times\Z_5$ corresponding to

8285: the $\Z_5$-orbifold of the quintic.

8286: There we find some chiral solutions.\\

8287: {\bf (ii)}~The theorem applies

8288: to more general models with $r\geq 5$, as long as

8289: the orbifold group is $\Z_H$.

8290: The essential point we have used is that $r$ is odd.

8291: In our supersymmetric formulation, we indeed need

8292: $r$ to be odd as discussed in Appendix~\ref{app:Gepner}

8293: (if $r=6$ or $8$ in the formulation

8294: as in \cite{Gepnerclassification} we need to add $k=0$ factor(s)

8295: to make $r$ odd).}

8296: 

8297: 

8298: 

8299: 

8300: 

8301: 

8302: 

8303: 

8304: 

8305: 

8306: 

8307: 

8308: 

8309: \section{Continuation to Geometry}\label{sec:continuation}

8310: 

8311: 

8312: In this section, we compare the results at the Gepner point

8313: and what is expected at the large volume regions.

8314: Namely we compare two different domains of the K\"ahler moduli space.

8315: The story is very much different between Type IIA and

8316: Type IIB cases since the role of the K\"ahler moduli

8317: are different.

8318: 

8319: 

8320: In Type IIA orientifolds, K\"ahler class and B-field form complex moduli

8321: fields. The large volume region, if consistent with orientifold,

8322: is always smoothly connected to the Gepner point and the comparison

8323: makes sense.

8324: The K\"ahler moduli can enter into the tree level

8325: superpotential.

8326: The comparison of the two regions may be useful to find

8327: out the set of low energy fields and the global determination of the

8328: tree level superpotential.

8329: 

8330: 

8331: 

8332: 

8333: In Type IIB orientifolds, K\"ahler moduli are real and

8334: are complexified by RR potentials.

8335: In some cases the large volume regions are separated from the Gepner point,

8336: but in some other cases they are smoothly connected.

8337: It is only in the latter case where the comparison makes sense.

8338: The K\"ahler moduli do not enter into the

8339: tree level superpotential, though they may enter into FI parameters

8340: as well as non-perturbative superpotential.

8341: 

8342: 

8343: 

8344: The main focus of this section will  be on the Type IIB cases.

8345: One technical advantage in these cases is that

8346: the large volume interpretation of the branes at the Gepner

8347: model has been worked out in detail.

8348: Thus, Sections~\ref{subsec:LV} through \ref{subsec:twoparaLV}

8349: are about Type IIB orientifolds.

8350: However, in the last subsection, we make some remarks

8351: on the Type IIA cases.

8352: 

8353: 

8354: 

8355: 

8356: 

8357: 

8358: 

8359: 

8360: 

8361: \subsection{Consistency Condition at Large Volume}

8362: \label{subsec:LV}

8363: 

8364: \newcommand{\sfX}{{\bf X}}

8365: 

8366: 

8367: Let us first

8368: present the tadpole cancellation condition in the large volume regime.

8369: We consider a spacetime manifold $\sfX$ with an involution $\tau$, and a

8370: D-brane supporting a complex vector bundle $E$ with an antilinear map

8371: that descends to $\tau$. 

8372: The tadpole cancellation condition for the $\tau$-orientifold

8373: of this system is 

8374: \beq

8375: {\rm ch}(E)\e^{-B}\sqrt{\widehat{\rm A}(\sfX)}

8376: =2^{2\dim_c\sfX^{\tau}-\dim_c \sfX}

8377: \epsilon [\sfX^{\tau}] \sqrt{{L({1\over 4}T\sfX^{\tau})\over

8378: L({1\over 4}N\sfX^{\tau})}}.

8379: \label{tccLV}

8380: \eeq

8381: This is found by comparing the formulae for the RR-overlaps

8382: with the boundary state

8383: $\widetilde{\Pi}^E_i$ and 

8384: the crosscap state $\widetilde{\Pi}^{\tau\Omega}_i$

8385: computed in the non-linear sigma models (see e.g. page 27 of \cite{BH2}).

8386: Some remarks are in order:\\

8387: $\bullet$~ $B$ is the B-field. In this section, we normalize it

8388: so that $B$ is trivial for closed strings if and only if

8389: $B\in H^2(\sfX,\Z)$.\\

8390: $\bullet$~ $\sfX^{\tau}$ is the O-plane, the fixed point set of

8391: $\tau$. $\sfX^{\tau}$ may consist of

8392: several connected components. In such a case the right hand side

8393:  is regarded as the sum over

8394: components.\\

8395: $\bullet$~ In the power of $2$, $\dim_c\sfX^{\tau}$ and $\dim_c\sfX$

8396: include the $\R^4$ directions (counted as 2 complex dimensions) as well as

8397: the internal dimensions.

8398: So, the power is $32$ for O9-plane, $8$ for O7,

8399: $2$ for O5, and $1/2$ for O3.\\

8400: $\bullet$~ $[\sfX^{\tau}]$ is the Poincar\'e dual of

8401: (the component of) the O-plane.

8402: ``$\epsilon$'' stands for a sign which is determined by the orientation

8403: of O-plane.\\

8404: $\bullet$~

8405: Useful identities to be remembered (on a Calabi-Yau three-fold $M$) are

8406: \beqa

8407: &&

8408: \widehat{\rm A}(M)={\rm td}(M)=1+{c_2(M)\over 12},

8409: \nn\\

8410: &&

8411: L({1\over 4}V)=1+{p_1(V)\over 48}

8412: =1-{c_2(V\otimes \C)\over 48},\quad \mbox{for a real vector bundle $V$}

8413: \nn

8414: \eeqa

8415: 

8416: Let us apply this to Type I string theory compactified

8417: on a Calabi-Yau 3-fold $M$ --- Type IIB orientifold

8418: of $\sfX=M\times \R^4$ associated with $\tau={\rm id}_{\sfX}$.

8419: In this case, $\sfX^{\tau}=\sfX$ and $[\sfX^{\tau}]=1$.

8420: Applying the useful formula we find

8421:  $\sqrt{{\rm td}(M)}=1+{1\over 24}c_2(M)$

8422: and $\sqrt{L({1\over 4}TM)}=1-{1\over 48}c_2(M)$.

8423: The condition is therefore

8424: $$

8425: {\rm ch}(E)\e^{-B}=32+2{\rm ch}_2(M),

8426: $$

8427: which is the rank and the anomaly cancellation condition in the

8428: standard form.

8429: 

8430: 

8431: We will examine whether the condition (\ref{tccLV})

8432: is satisfied

8433: for the D-brane configuration at the Gepner model

8434: continued to the large volume, whenever the continuation is

8435: possible. We work in two examples  --- the quintic case $(5,5,5,5,5)$

8436: and the two parameter model $(8,8,4,4,4)$.

8437: 

8438: 

8439: 

8440: \subsection{Quintic}\label{subsec:quinticLV}

8441: 

8442: Let us first discuss the model $(k_i+2)=(5,5,5,5,5)$

8443: that continues to the sigma model on the quintic hypersurface 

8444: $M$ of $\CP^4$.

8445: As we have seen in Section~\ref{subsub:quintic},

8446: the moduli space of the orientifold model is real,

8447: $\e^t\in \R$: The Gepner point $\e^t=0$ is separated from the $B=0$

8448: large volume ($\e^t\ll -1$)

8449: by the conifold point $\e^t=-5^5$, but is connected

8450: to the large volume region with $B={H\over 2}$ ($\e^t\gg 1$),

8451: where $H=c_1({\mathcal O}(1))|_M$ is the integral

8452: generator of $H^2(M,\Z)$.

8453: Thus, we expect the match of the condition only with

8454: the large volume with $B={H\over 2}$.

8455: 

8456: 

8457: Let us first write down the tadpole cancellation condition at the

8458: large volume. The Chern character of $M$ can be read from

8459: the exact sequence $0\to T_M\to T_{\CP^4}\to N_{M/\CP^4}\to 0$

8460: as ${\rm ch}(T_M)={\rm ch}(T_{\CP^4})|_M

8461: -{\rm ch}(N_{M/\CP^4})$. We know that $N_{M/\CP^4}={\mathcal O}(5)|_M$

8462: since $M$ is quintic, and

8463: also that

8464: ${\rm ch}(T_{\CP^4})={\rm ch}({\mathcal O}(1)^5)-{\rm ch}({\mathcal O})$

8465: from the tautological sequence.

8466: Thus, ${\rm ch}(T_M)=5\e^H-1-\e^{5H}

8467: =3-10H^2-20H^3$, and in particular

8468: ${\rm ch}_2(M)=-10 H^2$.

8469: Thus, the tadpole cancellation condition in the large volume region

8470: is

8471: \beq

8472: {\rm ch}(E)\e^{-B}=32-20 H^2.

8473: \eeq

8474: 

8475: Now we would like to compare this with the condition we obtained in

8476: Section~\ref{sec:TCCB}.

8477: In order to make the comparison, we need to know the relation of

8478: the basis of the D-brane charges at the Gepner model

8479: and the basis at the large volume region.

8480: This has been studied in \cite{BDLR}, and the result is

8481: \beq

8482: B_{{\bf L},M}=B_{(00000),2m+2n}\longleftrightarrow V_{m}

8483: \label{iden}

8484: \eeq

8485: for some $n\in \Z_5$ where

8486: \beqa

8487: V_0={\mathcal O},&& {\rm ch}(V_0)=1,

8488: \nn\\

8489: V_1=\overline{T^*_{\CP^4}(1)},&&

8490: {\rm ch}(V_1)=-4+H+{1\over 2}H^2+{1\over 6}H^3,

8491: \nn\\

8492: V_2=\wedge^2T^*_{\CP^4}(2),&&

8493: {\rm ch}(V_2)=6-3H-{1\over 2}H^2+{1\over 2}H^3,

8494: \nn\\

8495: V_3=\overline{\wedge^3T^*_{\CP^4}(3)},&&

8496: {\rm ch}(V_3)=-4+3H-{1\over 2}H^2-{1\over 2}H^3,

8497: \nn\\

8498: V_4=\wedge^4T^*_{\CP^4}(4),&&

8499: {\rm ch}(V_4)=1-H+{1\over 2}H^2-{1\over 6}H^3.

8500: \nn

8501: \eeqa

8502: We found in Section~\ref{sec:TCCB}

8503: that the O-plane has the D-brane charge

8504: $4[B_{{\bf 1},5}]=4(2[B_{{\bf 0},0}]

8505: +5[B_{{\bf 0},2}]

8506: +10[B_{{\bf 0},4}]

8507: +10[B_{{\bf 0},6}]

8508: +5[B_{{\bf 0},8}])$

8509: We try all the 5 possible identifications (\ref{iden}) to compute

8510: the rank of the tadpole canceling brane:

8511: \beqa

8512: V_{m}\leftrightarrow M=2m&\Longrightarrow&

8513: {\rm rank}=28,

8514: \nn\\

8515: V_{m}\leftrightarrow M=2m+2&\Longrightarrow&

8516: {\rm rank}=28,

8517: \nn\\

8518: V_{m}\leftrightarrow M=2m+4&\Longrightarrow&

8519: {\rm rank}=-32,

8520: \nn\\

8521: V_{m}\leftrightarrow M=2m+6&\Longrightarrow&

8522: {\rm rank}=-12,

8523: \nn\\

8524: V_{m}\leftrightarrow M=2m+8&\Longrightarrow&

8525: {\rm rank}=-12,

8526: \nn

8527: \eeqa

8528: Thus, we find that the identification

8529: $V_{m}\leftrightarrow M=2m+4$ may work.

8530: Indeed under this identification the full charge of the tadpole canceling

8531: D-brane is

8532: $$

8533: {\rm ch}(E)=4\left(-8+4H+4H^2-{7\over 3}H^3\right),

8534: $$

8535: and for the choice $B=-{H\over 2}$, we find

8536: \beq

8537: {\rm ch}(E)\e^{-B}=-32+20 H^2,

8538: \eeq

8539: which is nothing but the large volume condition.

8540: 

8541: 

8542: 

8543: 

8544: \subsection{The Two Parameter Model}\label{subsec:twoparaLV}

8545: 

8546: 

8547: Let us now discuss the two parameter model that includes the Gepner model

8548: with $(k_i+2)=(8,8,4,4,4)$. As we have seen in Section~\ref{subsub:two},

8549: the Gepner point and the large volume regions are separated in

8550: the K\"ahler moduli space of the orientifold models by the parities

8551: $P^B_{0,\epsilon_1...\epsilon_5}$.

8552: Thus in this case, we do not expect that the tadpole cancellation condition

8553: at the Gepner point matches with that in the large volume.

8554: On the other hand, for the orientifolds by $P_{1;\epsilon_1...\epsilon_5}^B$,

8555: the Gepner point is connected to the large volume regions

8556: in the moduli space.

8557: In fact, there are two separate large volume regions --- one with

8558: $B={L\over 2}$ and another with $B={H\over 2}+{L\over 2}$.

8559: Thus, the condition at the Gepner point must match with the conditions

8560: at {\it both} of the large volume region.

8561: We will check this in what follows.

8562: 

8563: 

8564: 

8565: 

8566: \subsubsection{Topology of the manifold and O-planes}

8567: 

8568: We first describe the topology of the Calabi-Yau manifold $M$ itself.

8569: Let $X$ be the toric manifold associated to the $U(1)^2$ gauge theory

8570: with six matter fields of the following charge

8571: $$

8572: \begin{array}{ccccccc}

8573: &X_1&X_2&X_3&X_4&X_5&X_6\\

8574: U(1)_1&0&0&1&1&1&1\\

8575: U(1)_2&1&1&0&0&0&-2

8576: \end{array}

8577: $$

8578: Our Calabi-Yau manifold $M$ is a hypersurface of $X$ given by

8579: $X_6^4(X_1^8+X_2^8)+X_3^4+X_4^4+X_5^4=0$.

8580: The cohomology ring of $X$ is generated by the divisor class

8581: $H=(X_3=0)=(X_4=0)=(X_5=0)$ and $L=(X_1=0)=(X_2=0)$ that obey the relations

8582: \beqa

8583: &&L^2=H^3(H-2L)=0,\nn\\

8584: &&\int_XH^3L=1.\nn

8585: \eeqa

8586: Holomorphic tangent bundle of $X$ fits into an exact sequence

8587: $$

8588: 0\to {\mathcal O}\oplus {\cal O}\to 

8589: {\mathcal L}_1\oplus

8590: {\mathcal L}_2\oplus

8591: {\mathcal L}_3\oplus

8592: {\mathcal L}_4\oplus

8593: {\mathcal L}_5\oplus

8594: {\mathcal L}_6

8595: \to T_X\to 0

8596: $$

8597: where ${\mathcal L}_i$ is the line bundle with section $X_i$.

8598: Chern lass of $X$ is therefore given by

8599: $$

8600: c(X)=(1+L)^2(1+H)^3(1+H-2L).

8601: $$

8602: The hypersurface $M$ yields the divisor class $[M]=4H$ and the normal

8603: bundle has $c_1(N_{M/X})=4H|_M$. We shall hereafter denote

8604: $H|_M,L|_M$ simply by $H,L$. They obey

8605: $$

8606: \int_MH^2L=4,\quad

8607: \int_MH^3=8.

8608: $$

8609: Chern class of $M$ is given by $c(M)=c(X)|_Mc(N_{M/X})^{-1}$

8610: namely,

8611: \beqa

8612: &&c_1(M)=0\\

8613: &&c_2(M)=2HL+6H^2\\

8614: &&c_3(M)=-21 H^3

8615: \eeqa

8616: 

8617: 

8618: Now, we write down the tadpole cancellation condition 

8619: (\ref{tccLV}) for the various involutions we discussed in

8620: Section~\ref{subsub:two}.

8621: 

8622: 

8623: \subsubsection*{\underline{$(+++++)$}}

8624: 

8625: When $\tau:M\to M$ is identity (the case for Type I string theory),

8626: the consistency condition for the

8627: bundle $E$ is

8628: \beq

8629: {\rm ch}(E)\e^{-B}

8630: =32-4HL-12 H^2.

8631: \eeq

8632: 

8633: \subsubsection*{\underline{$(++-++)$ etc}}

8634: 

8635: The fixed point set of $\tau:

8636: (X_1,...,X_6)\mapsto (X_1,X_2,-X_3,X_4,X_5,X_6)$ is the divisor

8637: $X_3=0$. For this we have

8638: \beqa

8639: [M^{\tau}]&=&H,

8640: \nn\\

8641: N_{M^{\tau}/M}&=&{\mathcal L}_3|_{M^{\tau}},\quad

8642: c(N_{M^{\tau}/M})=1+H

8643: \nn\\

8644: c(T_{M^{\tau}})&=&c(M)|_{M^{\tau}}c(N_{M^{\tau}/M})^{-1}

8645: =1-H+7H^2+2HL

8646: \nn\\

8647: p_1(TM^{\tau})

8648: &=&-c_2(T_{M^{\tau}}\oplus \overline{T}_{M^{\tau}})

8649: =-13 H^2-4HL,

8650: \nn\\

8651: p_1(NM^{\tau})

8652: &=&-c_2(N_{M^{\tau}}\oplus \overline{N}_{M^{\tau}})

8653: =H^2

8654: \nn\\

8655: {\rm td}(M)|_{M^{\tau}}

8656: &=&\left.1+{1\over 6}HL+{1\over 2}H^2\right|_{M^{\tau}}

8657: =1+{7\over 12}H^2

8658: \nn\\

8659: L({1\over 4}TM^{\tau})&=&

8660: 1-{1\over 48}(13H^2+4HL)=1-{15\over 48}H^2,

8661: \nn\\

8662: L({1\over 4}NM^{\tau})&=&

8663: 1+{1\over 48}H^2,

8664: \nn\\

8665: {L({1\over 4}TM^{\tau})\over

8666: L({1\over 4}NM^{\tau})}

8667: {\rm td}(M)^{-1}&=&1-{11\over 12}H^2.

8668: \nn

8669: \eeqa

8670: Thus, consistency condition for this orientifold is

8671: \beq

8672: {\rm ch}(E)\e^{-B}

8673: =\pm \left(8 H -{11\over 3} H^3\right).

8674: \eeq

8675: 

8676: \subsubsection*{\underline{$(++--+)$ etc}}

8677: 

8678: For

8679: $\tau:(X_1,...,X_6)\mapsto (X_1,X_2,-X_3,-X_4,X_5,X_6)$,

8680: the fixed point sets are the curve

8681: $C=\{X_3=X_4=0\}$ and four lines

8682: $\ell_a=\{X_5=X_6=0,X_3=\e^{{\pi i\over 4}+{\pi i a\over 2}}X_4\}$

8683: ($a=1,2,3,4$).

8684: Their Poincar\'e duals are

8685: \beqa

8686: &&[C]=[X_3=0]\cup [X_4=0]=H^2,

8687: \nn\\

8688: &&\sum_{a=1}^4[\ell_a]=[X_5=0]\cup[X_6=0]=H(H-2L).

8689: \nn

8690: \eeqa

8691: Thus, if the four O-planes at $\ell_a$ are of the same type, the

8692: consistency condition is

8693: \beqa

8694: {\rm ch}(E)\e^{-B}

8695: &=&2\{\pm H^2\pm H(H-2L)\}

8696: \nn\\

8697: &=&\left\{

8698: \begin{array}{l}

8699: \pm 4(H^2-HL)\\

8700: \mbox{or}\\

8701: \mp 4HL

8702: \end{array}

8703: \right.

8704: \eeqa

8705: The first line of the RHS 

8706: is when $C$ and $\ell_a$ contributes to the O-plane charge 

8707: with $\epsilon[C]=\pm [C]$ and $\sum_{a=1}^4\epsilon[\ell_a]

8708: =\pm \sum_{a=1}^4[\ell_a]$, while the second line is when

8709: they contributes with

8710: $\epsilon[C]=\mp [C]$ and $\sum_{a=1}^4\epsilon[\ell_a]

8711: =\pm \sum_{a=1}^4[\ell_a]$.

8712: 

8713: 

8714: 

8715: \subsubsection*{\underline{$(++---)$}}

8716: 

8717: For $\tau:(X_1,...,X_6)\mapsto 

8718: (X_1,X_2,-X_3,-X_4,-X_5,X_6)$,

8719: the fixed point sets are the divisor

8720: $D=\{X_6=0\}$

8721: and eight points

8722: $p_a=\{X_3=X_4=X_5=0, X_1=\e^{\pi i({1\over 8}+{a\over 4})}X_2\}$

8723: ($a=1,2,...,8$).

8724: Their Poincar\'e duals are

8725: \beqa

8726: &&[D]=[X_6=0]=H-2L,\nn\\

8727: &&\sum_{a=1}^8[p_a]=[X_3=X_4=X_5=0]=H^3.\nn

8728: \eeqa

8729: We have

8730: \beqa

8731: &&\int_DH^2=\int_M(H-2L)H^2=0 \,\,(\mbox{thus $H^2=0$ on $D$}),\nn\\

8732: &&\int_DHL=\int_M(H-2L)HL=4.\nn

8733: \eeqa

8734: and

8735: \beqa

8736: c(N_D)&=&1+H-2L,\nn\\

8737: c(T_D)&=&c(M)|_Dc(N_D)^{-1}

8738: =1+2L-H-2HL,\nn\\

8739: p_1(TD)&=&-c_2(T_D\oplus\overline{T}_D)=0,\nn\\

8740: p_1(ND)&=&-c_2(N_D\oplus\overline{N}_D)=-4HL,

8741: \nn\\

8742: L({1\over 4}TD)&=&0,\nn\\

8743: L({1\over 4}ND)&=&1-{1\over 12}HL,\nn\\

8744: {\rm td}(M)|_D&=&1+{1\over 6}HL,\nn\\

8745: {L({1\over 4}TD)\over L({1\over 4}ND)}

8746: {\rm td}(M)|_D^{-1}&=&1-{1\over 12}HL.

8747: \nn

8748: \eeqa

8749: If all the eight O3-planes are of the same type,

8750: the consistency condition is

8751: \beqa

8752: {\rm ch}(E)\e^{-B}

8753: &=&\pm 8(H-2L)(1-{1\over 24}HL)\pm {1\over 2}H^3

8754: \nn\\

8755: &=&\pm \left\{ 8H-16L-{1\over 6}H^3 \right\}

8756: \pm {1\over 2}H^3\nn\\

8757: &=&\pm\left\{

8758: \begin{array}{l}

8759: 8H-16L+{1\over 3}H^3

8760: \\

8761: \mbox{or}\\

8762: 8H-16L-{2\over 3}H^3

8763: \end{array}

8764: \right.

8765: \eeqa

8766: The first line of RHS is when $D$ and the eight points $p_a$

8767: contributes to the O-plane charge with

8768: $\epsilon[D]=\pm [D]$ and

8769: $\sum_{a=1}^8\epsilon[p_a]=\pm \sum_{a=1}^8[p_a]$,

8770: while the second line is when they contribute with

8771: $\epsilon[D]=\pm [D]$ and

8772: $\sum_{a=1}^8\epsilon[p_a]=\mp \sum_{a=1}^8[p_a]$.

8773: 

8774: \subsubsection*{\underline{$(+-***)$}}

8775: 

8776: In all the cases with $\epsilon_1=-\epsilon_2$,

8777: we have seen that the fixed point set

8778: consists of a pair of homologous components.

8779: Thus, one possible consistency condition is

8780: \beq

8781: {\rm ch}(E)\e^{-B}=0.

8782: \eeq

8783: There are of course other possibilities as well.

8784: 

8785: 

8786: \subsubsection{Gepner model to the large volume with

8787: $B={1\over 2}H+{1\over 2}L$}

8788: 

8789: Let us now see whether the set of Cardy

8790: branes obeying the tadpole cancellation condition, when transported in

8791: the orientifold moduli space, obey the condition at large volume.

8792: 

8793: 

8794: For the parities $P_{1;+-***}^B$, we have seen that the O-plane has no

8795: charge and therefore the tadpole canceling set of branes must have zero

8796: total RR-charge. This is indeed one of the possibilities as we have just seen

8797: --- the case where the two (set of) O-planes have the opposite

8798: RR-charge. In particular, this is realized by the supersymmetric

8799: O-plane configurations, where one of them is of $SO$-type and the

8800: other is of $Sp$-type.

8801: 

8802: 

8803: 

8804: For the parities $P_{1;++***}^B$, the O-plane has non-zero RR-charge

8805: and the check is non-trivial.

8806: For the comparison,

8807: we need to know the relation of the RR-charge of the Cardy branes

8808: at the Gepner model

8809: and the charge associated with the vector bundles at the large volume.

8810: One relation is found in \cite{DiaDou}

8811: \beqa

8812: &&{\rm ch}(V_1)=1-H+L+2\ell+{2\over 3}v

8813: \nn\\

8814: &&{\rm ch}(V_2)=-1+H-2L+4h-2\ell-{8\over 3}v

8815: \nn\\

8816: &&{\rm ch}(V_3)=-3+2H-L-{4\over 3}v

8817: \nn\\

8818: &&{\rm ch}(V_4)=3-2H+4L-8h+{4\over 3}v

8819: \nn\\

8820: &&{\rm ch}(V_5)=3-H-L-2\ell+{2\over 3}v

8821: \nn\\

8822: &&{\rm ch}(V_6)=-3+H-2L+4h+2\ell+{4\over 3}v

8823: \nn\\

8824: &&{\rm ch}(V_7)=-1+L

8825: \nn\\

8826: &&{\rm ch}(V_8)=1

8827: \nn\\

8828: \label{DiDo}

8829: \eeqa

8830: where

8831: $$

8832: \ell:={H^2-2HL\over 4},\quad

8833: h:={HL\over 4},\quad

8834: v:={H^3\over 8}={H^2L\over 4}.

8835: $$

8836: Up to cyclic permutation, $V_1,...,V_8$ are identified as a certain

8837: analytic continuation of the Cardy branes with

8838: $L=(00000)$ and $M=0,2,4,6,8,10,12,14$.

8839: We would first like to see which cyclic permutation is the relevant one.

8840: To find it, we compute the rank (D9-brane charge) of the tadpole canceling

8841: D-brane for the case $(+++++)$.

8842: We need it to be 32.

8843: A tadpole canceling D-brane has charge

8844: $(-20,-8,-12,12,8,20)$ with respect to the first six of the

8845: $L=(00000)$ Cardy branes.

8846: We find

8847: \beqa

8848: V_m\leftrightarrow M=2m&\Longrightarrow& {\rm rank}=32,

8849: \nn\\

8850: V_m\leftrightarrow M=2m-2&\Longrightarrow& {\rm rank}=24,

8851: \nn\\

8852: V_m\leftrightarrow M=2m-4&\Longrightarrow& {\rm rank}=0,

8853: \nn\\

8854: V_m\leftrightarrow M=2m-6&\Longrightarrow& {\rm rank}=-24,

8855: \nn\\

8856: V_m\leftrightarrow M=2m-8&\Longrightarrow& {\rm rank}=-32,

8857: \nn\\

8858: V_m\leftrightarrow M=2m+6&\Longrightarrow& {\rm rank}=-24,

8859: \nn\\

8860: V_m\leftrightarrow M=2m+4&\Longrightarrow& {\rm rank}=0,

8861: \nn\\

8862: V_m\leftrightarrow M=2m+2&\Longrightarrow& {\rm rank}=24,

8863: \nn

8864: \eeqa

8865: Thus, the identification $V_m\leftrightarrow M=2m$ is the correct one.

8866: ($V_m\leftrightarrow M=2m+8$ may also have a chance,

8867: but it is simply the sign flip of $V_m\leftrightarrow M=2m$.)

8868: Under this identification,

8869: the tadpole cancellation condition

8870: at the Gepner point continues to the condition at the large volume

8871: with $B=-{1\over 2}H+{1\over 2}L$, as we now see.

8872: 

8873: 

8874: 

8875: \subsubsection*{\underline{$(+++++)$}}

8876: 

8877: 

8878: The charge of a tadpole canceling D-brane $E$ is

8879: \beqa

8880: {\rm ch}(E)&=&

8881: -20 {\rm ch}(V_8)

8882: -8{\rm ch}(V_1)

8883: -12{\rm ch}(V_2)

8884: +12{\rm ch}(V_3)

8885: +8{\rm ch}(V_4)

8886: +20{\rm ch}(V_5)

8887: \nn\\

8888: &=&32-16H+16L-8H^2-12HL+{13\over 3}H^3.

8889: \nn

8890: \eeqa

8891: If we choose $B=-{1\over 2}H+{1\over 2}L$, we find

8892: \beq

8893: {\rm ch}(E)\e^{-B}=32-12H^2-4HL.

8894: \eeq

8895: This is nothing but the tadpole cancellation condition in the large volume

8896: regime.

8897: 

8898: 

8899: 

8900: \subsubsection*{\underline{$(++-++)$ etc}}

8901: 

8902: 

8903: \beqa

8904: {\rm ch}(E)&=&

8905: -4 {\rm ch}(V_8)

8906: -8{\rm ch}(V_1)

8907: -12{\rm ch}(V_2)

8908: -12{\rm ch}(V_3)

8909: -8{\rm ch}(V_4)

8910: -4{\rm ch}(V_5)

8911: \nn\\

8912: &=&-8H+4(H^2-HL)+{11\over 3}H^3

8913: \nn

8914: \eeqa

8915: and, for $B=-{1\over 2}H+{1\over 2}L$,

8916: \beq

8917: {\rm ch}(E)\e^{-B}

8918: =-8H+{11\over 3}H^3.

8919: \eeq

8920: This agrees with the large volume condition with $\epsilon[M^{\tau}]

8921: =-[M^{\tau}]$.

8922: 

8923: 

8924: \subsubsection*{\underline{$(++--+)$ etc}}

8925: 

8926: 

8927: \beqa

8928: {\rm ch}(E)&=&

8929: 4 {\rm ch}(V_8)

8930: +0{\rm ch}(V_1)

8931: +4{\rm ch}(V_2)

8932: -4{\rm ch}(V_3)

8933: +0{\rm ch}(V_4)

8934: -4{\rm ch}(V_5)

8935: \nn\\

8936: &=&4HL-H^3

8937: \nn

8938: \eeqa

8939: and, for $B=-{1\over 2}H+{1\over 2}L$,

8940: \beq

8941: {\rm ch}(E)\e^{-B}

8942: =4HL.

8943: \eeq

8944: This agrees with the condition in the large volume regime where 

8945: $C$ and $\ell_a$ contribute to the O-plane charge with

8946: $\epsilon[C]=[C]$ and 

8947: $\epsilon[\ell_a]=-[\ell_a]$.

8948: 

8949: 

8950: \subsubsection*{\underline{$(++---)$}}

8951: 

8952: 

8953: 

8954: \beqa

8955: {\rm ch}(E)&=&

8956: 4 {\rm ch}(V_8)

8957: +0{\rm ch}(V_1)

8958: +4{\rm ch}(V_2)

8959: +4{\rm ch}(V_3)

8960: +0{\rm ch}(V_4)

8961: +4{\rm ch}(V_5)

8962: \nn\\

8963: &=&8H-16L-4(H^2-3HL)-{5\over 3}H^3

8964: \nn

8965: \eeqa

8966: and, for $B=-{1\over 2}H+{1\over 2}L$,

8967: \beq

8968: {\rm ch}(E)\e^{-B}

8969: =8H-16L-{2\over 3}H^3.

8970: \eeq

8971: This agrees with the condition in the large volume regime where 

8972: $D$ and $p_a$ contribute to the O-plane charge with

8973: $\epsilon[D]=[D]$ and

8974: $\epsilon[p_a]=-[p_a]$.

8975: 

8976: 

8977: 

8978: \subsubsection{Gepner Model to the large volume with

8979: $B={1\over 2}L$}

8980: 

8981: For the parities $P^B_{1;\epsilon_1...\epsilon_5}$

8982: we are considering, the orientifold

8983: moduli space contains another large volume region

8984: --- the region with $B={L\over 2}$ (mod $\Z H+\Z L$).

8985: Since this region is not separated from

8986: the Gepner point in the moduli space,

8987: the tadpole cancellation condition

8988: at the Gepner point should match with the one at this large volume.

8989: Let us confirm this.

8990: 

8991: 

8992: The main task is to find

8993: the transformation rule of the D-brane charge

8994: --- from the Gepner point to the large volume.

8995: Let $(\phi,\psi)$ be the coordinate

8996: of the cover of the moduli space (before orientifold) that

8997: are used in \cite{Candelas}. These are the natural parameters of the

8998: superpotential of the mirror LG model (\ref{dualW2para}),

8999: $\widetilde{W}=\widetilde{W}_G-8\psi \widetilde{X}_1\cdots\widetilde{X}_5

9000: -2\phi\widetilde{X}_1^4\widetilde{X}_2^4$,

9001: and is related to the linear sigma model parameters as

9002: \beqa

9003: &&\e^{t_1}=-2^{11}\psi^4\phi^{-1},\nn\\

9004: &&\e^{t_2}=4\phi^2.\nn

9005: \eeqa

9006: The singular loci are described as

9007: $$

9008: C_1=\Bigl\{\,\phi^2=1\,\Bigr\},\quad

9009: C_{\rm con}=\Bigl\{\,(\phi+8\psi^4)^2=1\,\Bigr\}.

9010: $$

9011: The $\omega=\e^{2\pi i/8}$ orientifolds impose constraints

9012: $\e^{2\pi i /8}\psi=\overline{\psi}$ and

9013: $-\phi=\overline{\phi}$, or

9014: \beq

9015: \psi\in \e^{-{\pi i\over 8}}\R\quad

9016: \phi\in i\R.

9017: \eeq

9018: Let us consider

9019: a path in this moduli space, ${\mathcal P}_0$:

9020: $\psi=\e^{-\pi i/8}t^{3\over 8}$,

9021: $\phi=\e^{-\pi i/2}\sqrt{t}$, $0\leq t<+\infty$.

9022: (In the $(t_1,t_2)$ coordinates, it is

9023: $\e^{t_1}=-2^{11}t$, $\e^{t_2}=-4t$.) 

9024: It goes from the Gepner point to the large volume region with

9025: $B={H\over 2}+{L\over 2}$.

9026: The identification (\ref{DiDo}) for

9027: $\Pi^{\rm Cardy}=(B_{{\bf 0},0},B_{{\bf 0},2},B_{{\bf 0},4},

9028: B_{{\bf 0},6},B_{{\bf 0},8},B_{{\bf 0},10})^T$

9029: and $\Pi^{\rm LV}=(1,H,L,H^2,HL,H^3)^T$: 

9030: $$

9031: \Pi^{\rm Cardy}=M_{{\mathcal P}_0}\Pi^{\rm LV};\qquad

9032: M_{{\mathcal P}_0}=\left(\begin{array}{cccccc}

9033: 1&0&0&0&0&0\\

9034: 1&-1&1&{1\over 2}&-1&{2\over 3}\\

9035: -1&1&-2&-{1\over 2}&2&-{8\over 3}\\

9036: -3&2&-1&0&0&-{4\over 3}\\

9037: 3&-2&4&0&-2&{4\over 3}\\

9038: 3&-1&-1&-{1\over 2}&1&{2\over 3}

9039: \end{array}\right)

9040: $$

9041: can be regarded as the

9042: transformation of charges for this choice of path.

9043: We would like to find the transformation with respect to

9044: the other path,

9045: ${\mathcal P}_1$:

9046: $\psi=\e^{-\pi i/8}t^{3\over 8}$,

9047: $\phi=-\e^{-\pi i/2}\sqrt{t}$

9048: ($\e^{t_1}=2^{11}t$, $\e^{t_2}=-4t$), 

9049: that goes to the large volume with

9050: $B={L\over 2}$. In order to find it,

9051: let us find a homotopy of paths from

9052: the Gepner point to large volume,

9053: that deforms

9054: ${\mathcal P}_0$ to ${\mathcal P}_1$.

9055: The following does the job:

9056: \beq

9057: {\mathcal P}_s

9058: :\,\left\{\begin{array}{l}

9059: \psi=\e^{-{\pi i\over 8}}t^{3\over 8}\\

9060: \phi=\e^{-{\pi i\over 2}}\e^{\pi i s}\sqrt{t},

9061: \end{array}\right.\quad 0\leq s\leq 1.

9062: \eeq

9063: \begin{figure}[htb]

9064: \centerline{\includegraphics{homotopy.eps}}

9065: \caption{The homotopy ${\mathcal P}_s$. The shaded region is the

9066: orientifold moduli space.}

9067: \label{homotopyF}

9068: \end{figure}

9069: It intersects only with $C_1$ of the singular locus at

9070: $t=1$ and $s={1\over 2}$. 

9071: See Figure~\ref{homotopyF}.

9072: Thus, we find that ${\mathcal P}_1$ is homotopic to

9073: $-{\mathcal P}_{C_1}+{\mathcal P}_0+{\mathcal P}_{\infty}$

9074: where ${\mathcal P}_{C_1}$ is the contour

9075: that goes once around the singular locus $C_1$ and

9076: ${\mathcal P}_{\infty}$ is a contour that stays in the large volume limit.

9077: See another figure, Fig.~\ref{paths}.

9078: 

9079: \begin{figure}[htb]

9080: \centerline{\includegraphics{paths.eps}}

9081: \caption{The paths}

9082: \label{paths}

9083: \end{figure}

9084: 

9085: In \cite{Candelas}, the monodromy of the RR-charge for the contour

9086: ${\mathcal P}_{C_1}$ is computed

9087: in a basis $\Pi^{\rm G}$ as

9088: $$

9089: \Pi^{\rm G}\to B\Pi^{\rm G};\qquad

9090: B=\left(\begin{array}{cccccc}

9091: 1&0&0&0&0&0\\

9092: 0&1&-1&1&-1&1\\

9093: 0&0&2&-1&1&-1\\

9094: 3&-3&4&-3&3&-3\\

9095: -3&3&-4&4&-2&3\\

9096: -3&3&-3&3&-2&3

9097: \end{array}\right)

9098: $$

9099: We also know that the intersection matrices with respect to the two bases are

9100: related as $I^{\rm Cardy}=(1-A)I^{\rm G}(1-A)^T$

9101: where

9102: $$

9103: A=\left(\begin{array}{cccccc}

9104: 0&1&0&0&0&0\\

9105: 0&0&1&0&0&0\\

9106: 0&0&0&1&0&0\\

9107: 0&0&0&0&1&0\\

9108: 0&0&0&0&0&1\\

9109: -1&0&-1&0&-1&0

9110: \end{array}\right)

9111: $$

9112: This implies that the two bases are related by

9113: $$

9114: \Pi^{\rm Cardy}=U\Pi^{\rm G};\quad

9115: U=\pm (1-A)A^n,

9116: $$

9117: for some $n$.

9118: In the Cardy basis, the monodromy along the contour $-{\mathcal P}_{C_1}$

9119: is given by $\Pi^{\rm Cardy}\to UB^{-1}U^{-1}\Pi^{\rm Cardy}$.

9120: Thus, the transformation of the charge basis along the path

9121: ${\mathcal P}_1$ is given by

9122: \beq

9123: \Pi^{\rm Cardy}=M_{{\mathcal P}_1}\Pi^{\rm LV};\quad

9124: M_{{\mathcal P}_1}=UB^{-1}U^{-1}M_{{\mathcal P}_0}.

9125: \eeq

9126: It turns out that $U=(1-A)A^6$ is the right choice so that the Cardy branes 

9127: canceling the tadpole at the Gepner point obey

9128: the condition at the large volume with $B=-H+{3\over 2}L$:\\

9129: \underline{$(+++++)$}

9130: \beq

9131: {\rm ch}(E)\e^{-B}

9132: =32-12 H^2-4HL.

9133: \eeq

9134: This agrees with the large volume condition.\\

9135: \underline{$(++-++)$ etc}

9136: \beq

9137: {\rm ch}(E)\e^{-B}

9138: =-8H+{11\over 3}H^3

9139: \eeq

9140: This agrees with the large volume condition with

9141: $\epsilon[M^{\tau}]=-[M^{\tau}]$.

9142: \\

9143: \underline{$(++--+)$ etc}

9144: \beq

9145: {\rm ch}(E)\e^{-B}

9146: =4H^2-4HL

9147: \eeq

9148: This agrees with the large volume condition with

9149: $\epsilon[C]=[C]$ and

9150: $\epsilon[\ell_a]=[\ell_a]$.

9151: \\

9152: \underline{$(++---)$}

9153: \beq

9154: {\rm ch}(E)\e^{-B}

9155: =-8H+16L-{1\over 3}H^3.

9156: \eeq

9157: The is agrees with the large volume condition

9158: with $\epsilon[D]=-[D]$ and

9159: $\epsilon[p_a]=-[p_a]$.

9160: 

9161: 

9162: 

9163: 

9164: \subsubsection{Type of the O-planes}

9165: 

9166: We have determined the orientation $\epsilon[\sfX^{\tau}]$

9167: of the O-plane in the two large volume regions, for

9168: the $P_{1;\epsilon_1...\epsilon_5}$-orientifolds.

9169: We would now like to know the type of each component of the O-plane.

9170: We recall that there are roughly two types of O-planes

9171: --- O$^-$ and O$^+$:

9172: $N$ D$p$-brane on top of an O$^-p$-plane support $O(N)$ gauge

9173: group while $N$ D$p$-brane on top of an O$^+p$-plane support

9174: $Sp(N/2)$.\footnote{There is actually a finer classification

9175: labeled by the discrete RR-flux \cite{WHWSH}, which we do not

9176: discuss in the present paper.}

9177: O$^-$-plane has a negative tension and O$^+$-plane has a positive tension.

9178: Thus, in a supersymmetric configuration where the NSNS tadple

9179: is also cancelled, it is impossible to have O-planes of type O$^+$ only.

9180: In particular, if the O-plane consists of a single component, that must be

9181: an O$^-$-plane.

9182: Thus, the O9-plane for the $P_{*;+++++}$-orientifold

9183: and the O7-plane for the $P_{*;++-++}$-orientifold must be both O$^-$,

9184: as long as there are supersymmetric brane configurations at large volume.

9185: For the $P_{1;+-***}$-orientifolds,

9186: the O-plane has two (sets of) components

9187: which are homologous to each other.

9188: At the Gepner point,

9189: we found that a configuration without a brane is supersymmetric

9190: and tadpole canceling. Thus the two (sets of) O-plane components

9191: are of the opposite types, as we have mentioned.

9192: In what follows, we discuss the remaining cases, $P_{1;++--+}$

9193: and $P_{1;++---}$.

9194: 

9195: 

9196: 

9197: Let us consider Type II orientifold on

9198: a ten-dimensional manifold $\sfX$ with respect to an involution

9199: $\tau$ of $\sfX$.

9200: Let

9201: $$

9202: \sfX^{\tau}=\bigcup_iW_i

9203: $$

9204: be the decomposition of the O-plane into connected components.

9205: Let $o_i=\pm 1$ be the type of the O-plane at $W_i$ ---

9206: it is $\pm 1$ if $W_i$ is an O$^{\pm}$-plane.

9207: As before we denote the orientation of the O-plane

9208: by $\epsilon[W_i]$.

9209: Then, the D-brane wrapped on $W_i$ preserves the same spacetime

9210: supersymmetry as this O-plane at $W_i$ if its orientation

9211: is $o_i\epsilon[W_i]$.

9212: This also tells us the phase of the supersymmetry in the large volume limit.

9213: To be specific, we consider the Type IIB orientifolds on a Calabi-Yau

9214: three-fold. The overlap of the boundary state for a brane

9215: wrapped on $W_i$ and the RR-ground state

9216: $|0\rangle_{\RR}$ of the lowest R-charge

9217: is given by

9218: \begin{eqnarray}

9219: {}_{\RR}\langle 0|B_{W_i}\rangle

9220: &=&

9221: \int_{[W_i]}\e^{-i\omega}+\cdots

9222: \nonumber\\

9223: &=&

9224: (-i)^{\dim W_i}\int_{[W_i]}{\omega^{\dim W_i}\over

9225: \dim W_i!}+\cdots

9226: \nonumber

9227: \end{eqnarray}

9228: where $+\cdots$ are small in the limit $\omega\gg 1$.

9229: Note that if $W_i$ is a complex submanifold and $[W_i]$ is the

9230: standard orientation, $\int_{[W_i]}\omega^{\dim W_i}$ is positive.

9231: Thus, in the large volume limit,

9232: the phase of the supersymmetry preserved by the O-plane of type $o_i$

9233: and orientation $\epsilon[W_i]=\epsilon_i\cdot [W_i]$ is given by

9234: $$

9235: \e^{i\theta_i}=o_i\epsilon_i(-i)^{\dim W_i}.

9236: $$

9237: In particular, in a supersymmetric orientifold, all the components

9238: $W_i$ must have the same $o_i\epsilon_i(-i)^{\dim W_i}$.

9239: 

9240: 

9241: This can be used to find the type of O-plane components

9242: (up to an overall sign)

9243: for the $P_{1;*****}$-orientifolds.

9244: We recall the O-plane orientations for the four relevant cases:

9245: $$

9246: \begin{array}{l|l|l}

9247: &B={H\over 2}+{L\over 2}&B={L\over 2}\\[0.1cm]

9248: \hline

9249: P_{1;++--+}&

9250: \begin{array}{l}

9251: \epsilon[C]=-[C]\\

9252: \epsilon[\ell_a]=[\ell_a]

9253: \end{array}

9254: &

9255: \begin{array}{l}

9256: \epsilon[C]=-[C]\\

9257: \epsilon[\ell_a]=-[\ell_a]

9258: \end{array}

9259: \\[0.2cm]\hline

9260: P_{1;++---}&

9261: \begin{array}{l}

9262: \epsilon[D]=-[D]\\

9263: \epsilon[p_a]=[p_a]

9264: \end{array}

9265: &

9266: \begin{array}{l}

9267: \epsilon[D]=[D]\\

9268: \epsilon[p_a]=[p_a]

9269: \end{array}

9270: \end{array}

9271: $$

9272: Thus, for a common supersymmetry to be conserved, we need that

9273: the types must be related as follows

9274: $$

9275: \begin{array}{l|l|l}

9276: &B={H\over 2}+{L\over 2}&B={L\over 2}\\[0.1cm]\hline

9277: P_{1;++--+}&

9278: o_C=-o_{\ell_a}

9279: &

9280: o_C=o_{\ell_a}

9281: \\[0.2cm]\hline

9282: P_{1;++---}&

9283: o_D=o_{p_a}

9284: &

9285: o_D=-o_{p_a}

9286: \end{array}

9287: $$

9288: To fix the overall sign, we need to look at the tension of the O-plane.

9289: Since we need some branes to cancel the tadpole in all cases,

9290: we need the total tension to be negative.

9291: Let us first discuss the $P_{1;++--+}$ orientifolds.

9292: For a K\"ahler form $\omega=r_1H+r_2L$, we find

9293: \begin{eqnarray}

9294: &&

9295: \int_{[C]}\omega+\sum_{a=1}^4\int_{[\ell_a]}\omega

9296: =8r_1+8r_2

9297: \nonumber\\

9298: &&\int_{[C]}\omega-\sum_{a=1}^4\int_{[\ell_a]}\omega

9299: =8r_1

9300: \nonumber

9301: \end{eqnarray}

9302: Thus, the type of the O-plane at $C$ determines the sign of the total

9303: tension, and it must always be O$^-$, $o_C=-1$.

9304: Next, let us consider the $P_{1;++---}$ orientifolds.

9305: In this case, the O-plane at $D$ has clearly

9306: larger tension than the ones at $p_a$ in the large volume limit.

9307: Thus, O-plane at $D$ must always be O$^-$, $o_D=-1$.

9308: To summarize, we find that the type of the O-plane

9309: components are given by 

9310: $$

9311: \begin{array}{|l|l|l|}

9312: \hline

9313: &B={H\over 2}+{L\over 2}&B={L\over 2}\\[0.1cm]\hline

9314: P_{1;++--+}&

9315: \begin{array}{l}

9316: C:O^-\\

9317: \ell_a:O^+

9318: \end{array}

9319: &

9320: \begin{array}{l}

9321: C:O^-\\

9322: \ell_a:O^-

9323: \end{array}

9324: \\[0.2cm]\hline

9325: P_{1;++---}&

9326: \begin{array}{l}

9327: D:O^-\\

9328: p_a:O^-

9329: \end{array}

9330: &

9331: \begin{array}{l}

9332: D:O^-\\

9333: p_a:O^+

9334: \end{array}

9335: \\\hline

9336: \end{array}

9337: $$

9338: 

9339: 

9340: We found an interesting phenomenon.

9341: {\it As we move from one large volume region to another,

9342: through the non-geometric region of the constrained K\"ahler moduli space,

9343: the type of O-plane changes}:

9344: For the $P_{1;++--+}$-orientifold, the O5-plane at the rational curves

9345: $\ell_a$ change from O$5^+$

9346: to O$5^-$ while the O5-plane at the genus 9 curve

9347: $C$ stays as O$5^-$.

9348: For the $P_{1;++---}$-orientifold,

9349: the O3-planes at $p_a$ change from O$3^-$ to O$3^+$

9350: while the O7-plane at the divisor $D$ remains to be O$7^-$.

9351: In this discussion, we have assumed that the sign of the total

9352: O-plane tension remains the same as the Gepner point.

9353: This can be justified by showing that the overlap

9354: $\Pi^{\tau\Omega}_0$ does not vanish on a path from the Gepner point

9355: to the large volume. (However, even if this assumption turns out

9356: to be wrong, the change in the type of O-plane we have just discussed

9357: remains true.)

9358: 

9359: 

9360: This provide a challenge in finding

9361: the classification scheme of D-brane charges

9362: using K-theory that is valid uniformly in the moduli space.

9363: For flat tori, it is known that

9364: the D-brane charges in Type II orientifolds

9365: are classified by using KR group \cite{WK,HT,BGH}

9366: (see also \cite{Gukov,Stefanski}):

9367: For $T^{9-p}/\Z_2$-orientifold

9368: with all O$p^-$-planes (resp. all O$p^+$-planes),

9369: the D-brane charges is classified by

9370: KR$^{-(9-p)}$-group (resp. KR$^{-(5-p)}$-group).

9371: One may guess that a similar rule applies when the space is

9372: curved.

9373: But the above phenomenon tells us that we need something very different

9374: to describe the D-brane charges uniformly on the moduli space.

9375: 

9376: 

9377: 

9378: 

9379: 

9380: 

9381: \subsection{Comments on Type IIA Orientifolds}

9382: 

9383: 

9384: Let us consider

9385: Type IIA orientifold on a large volume Calabi-Yau manifold

9386: $M$ with respect to an antiholomorphic involution $\tau$ of $M$.

9387: The O-plane is the fixed point set $M^{\tau}$ which is

9388: a special Lagrangian submanifold, and 

9389: the RR-flux generated by this is determined by the homology class

9390: $[M^{\tau}]\in H_3(M;\Z)$.

9391: The tadpole cancellation condition $n_i$ D6-branes wrapped on special

9392: Lagrangian submanifolds $L_i$ are given by\footnote{We

9393: assume that $M$ is simply connected, in which case

9394: $K^1(M)=H_3(M,\Z)$.}

9395: \beq

9396: \sum_{i=1}^Nn_i[L_i]=4[M^{\tau}],

9397: \eeq

9398: where $[L_i]$ is the homology class of the submanifold

9399: $L_i$.

9400: One obvious solution to this condition is the configuration of

9401: four D6-branes wrapped on $M^{\tau}$,

9402: but other solutions may exist as well.

9403: There can be a spacetime superpotential

9404: depending on the $b_1(L_i)$ open string fields as well as

9405: the K\"ahler moduli. There could even be open string fields

9406: which are heavy at large volume limit but become light

9407: in some interior regions of the K\"ahler moduli.

9408: We would like to compare this with some results obtained at the Gepner

9409: point.

9410: 

9411: 

9412: 

9413: 

9414: 

9415: Let us first consider the odd $H$ cases.

9416: In each of such cases, we found a supersymmetric and tadpole canceling

9417: configuration consisting of four copies of one brane

9418: $B_{\bf {k-1\over2},{k-1\over 2}}$.

9419: By comparison with the large volume condition,

9420: it is natural to identify the brane $B_{\bf {k-1\over2},{k-1\over 2}}$

9421: as the brane wrapped on the fixed point set $M^{\tau}$.

9422: For example, in the case of the quintic,

9423: $B_{\bf 1,1}$ is identified as the D6-brane wrapped on

9424: the real quintic which has a topology of $\RP^3$.

9425: In any of the odd $H$ cases,

9426: the open string spectrum at the Gepner point includes

9427: massless chiral multiplets charged under

9428: $O(4)$ --- one or more symmetric tensors and in some cases

9429: antisymmetric tensors as well (see Section~\ref{subsubsec:oddH}).

9430: Let us consider the quintic case, where

9431: there is a single massless matter in the symmetric

9432: representation.

9433: Can it be consistent with the large volume result?

9434: At large volume, there is no open string moduli since

9435: the D-brane is wrapped on a simply connected submanifold $\RP^3$.

9436: However, as noted in \cite{BDLR}, there are choices in specifying

9437: the supersymmetric configuration --- the choice of the flat gauge

9438: connection on the brane.

9439: In the present case where the gauge group is $O(4)$, this is given by the

9440: $\pi_1(\RP^3)=\Z_2$ holonomy, namely, an element of

9441: $O(4)$ which squares to $1$. This is up to gauge transformation, and thus the

9442: vacuum manifold is

9443: \beq

9444: {\it V}_{\rm LV}=\Bigl\{\,g\in O(4)\,\Bigl|\,

9445: g^2={\bf 1}_4\,\Bigl\}/{\rm ad} O(4),

9446: \label{LVVac}

9447: \eeq

9448: where $/{\rm ad} O(4)$ is quotient by the action

9449: $g\mapsto hgh^{-1}$, $h\in O(4)$.

9450: Is there a field theory model consistent with this and

9451: the massless spectrum at the Gepner point?

9452: This problem is encountered in \cite{BDLR} in the context of a single

9453: brane in a theory without orientifold in which

9454: there are two vacua at large volume ${\it V}^{\rm single}_{\rm LV}=\{\pm 1\}$:

9455: The model is

9456: given by the superpotential $W=\phi \psi+\psi^3$ where $\phi$ is

9457: a closed string field representing the K\"ahler

9458: class and $\psi$ is the open string field.

9459: If we identify $\phi=0$ as the Gepner point, $\psi$ is massless at the Gepner

9460: point and at large volume there are two vacua

9461: $\psi=\pm \sqrt{-\phi/3}$, which is consistent with

9462: ${\it V}^{\rm single}_{\rm LV}=\{\pm 1\}$.

9463: A natural extension to the current situation is 

9464: the theory with superpotential

9465: \beq

9466: W=\phi\Tr\psi +\Tr\psi^3,

9467: \label{Wphipsi}

9468: \eeq

9469: where $\phi$ is the closed string field

9470: representing the K\"ahler class and

9471: and $\psi$ is the symmetric tensor for $O(4)$,

9472: $\psi_{ij}=\psi_{ji}$, $i,j=1,2,3,4$.

9473: The Gepner point is identified as $\phi=0$ where $\psi$ is massless.

9474: Away from that point, say $\phi=-3$, the vacuum equations for

9475: $\psi$ are

9476: $$

9477: [\psi,\psi^{\dag}]=0,\quad

9478: \psi^2={\bf 1}_4,

9479: $$

9480: where the first equation is the D-term equation

9481: (with $\psi^T=\psi$ taken into account) and the second equation

9482: is the F-term equation, $\partial_{\psi}W=0$.

9483: The vacuum manifold is obtained by moding out the solution space

9484: by the adjoint $O(4)$ action.

9485: By the D-term equation, $\psi$ can be diagonalized by $U(4)$ matrix

9486: and it then follows from $\psi^T=\psi$ that $\psi$ is a real matrix.

9487: Thus, $\psi$ is a four-by-four real matrix with

9488: the constraint $\psi^T\psi=\psi^2={\bf 1}_4$.

9489: Namely, the vacuum manifold

9490: agrees with one at the large volume (\ref{LVVac}).

9491: In this discussion, we have treated the K\"ahler modulus

9492: $\phi$ as a parameter. Of course, 

9493: in the full string theory,  we must treat $\phi$ as a dynamical field and

9494: include $\partial_{\phi}W=0$ into the

9495: vacuum equations. Then, we obtain an extra constraint

9496: $\Tr \psi=0$ which means that $\psi$ has the same number of

9497: $+1$ or $-1$ eigenvalues.

9498: It would be interesting to find a similar field theory model for the odd $H$

9499: cases other than the quintic.

9500: 

9501: 

9502: 

9503: 

9504: Let us next consider an even $H$ case, the two parameter model

9505: with $k_i=(6,6,2,2,2)$.

9506: There is a freedom to dress by quantum symmetry, but

9507: we only consider those without dressing for which

9508: the large volume region is included in the moduli space.

9509: There are six such cases $P^A_{+;00000}$,

9510: $P^A_{+;00001}$, $P^A_{+;00011}$, $P^A_{+;00111}$

9511: $P^A_{+;01000}$. $P^A_{+;01001}$.

9512: At the Gepner point,

9513: only one of them, $P^A_{+;01000}$, admits a tadpole canceling

9514: supersymmetric solution

9515: with exactly four copies of an elementary brane.

9516: This brane, $B_{(30111),(30111)}$, is identified in the large volume limit

9517: as the O-plane which has topology of $S^3$.

9518: Since this is simply connected, there is a unique supersymmetric

9519: configuration at the large volume.

9520: At the Gepner point, we found no massless matter field.

9521: A theory consisting of all these is the one

9522: with only $O(4)$ super-Yang-Mills without matter and

9523: exactly flat K\"ahler moduli space.

9524: In all of the five other cases, we found that

9525: there is no consistent supersymmetric configuration

9526: with only four branes at the Gepner point

9527: (Section~\ref{subsec:general}),

9528: while ``four D6-branes wrapped

9529: on the O-plane'' is always a solution at large volume.

9530: Note that the O-plane has $b_1\geq 1$ in these cases,

9531: and there are massless open string fields that correspond

9532: to moving pairs of D6-branes away from the O-plane,

9533: breaking $O(4)$ to $U(2)$ or further to $U(1)^2$.

9534: One can expect a non-trivial superpotential depending on such open string

9535: fields as well as K\"ahler moduli, and it is conceivable that the

9536: supersymmetric vacua with unbroken $O(4)$ misses the Gepner point.

9537: It is an interesting problem to verify it by explicit computation of

9538: superpotential. Another interesting problem is

9539: to analyze the interaction of

9540: the supersymmetric solutions we found at the Gepner point and try to

9541: connect to the large volume limit.

9542: 

9543: 

9544: 

9545: 

9546: 

9547: 

9548: 

9549: 

9550: 

9551: 

9552: 

9553: 

9554: 

9555: 

9556: 

9557: \section*{Acknowledgement}

9558: 

9559: We would like to thank M. Cvetic, C. Doran,

9560: M. Douglas, J. Giedt, J. Gomis, M. Haack, M. Kapranov,

9561: G. Mikhalkin, E. Poppitz,

9562: R. Rabadan, A. Uranga for discussions. K.H. thanks Banff 

9563: International Research Centre, Stanford University, and KITP 

9564: Santa Barbara for hospitality during various stages of this work.

9565: K.H. and K.H. were supported in part by Natural Sciences and Engineering

9566: Research Council of Canada. K.H. is also supported by the Alfred P. Sloan

9567: Foundation. The research is also supported in part by the National Science

9568: Foundation under Grant No. PHY99-07949 and the PPARC grant

9569: PPA/G/O/2000/00451.

9570: 

9571: 

9572: 

9573: \appendix{More General Gepner Models}\label{app:Gepner}

9574: 

9575: 

9576: A Gepner model is defined as the orbifold of the product of

9577: minimal models $\prod_{i=1}^rM_{k_i}$ 

9578: by the group $\Gamma\simeq \Z_H$ ($H:={\rm l.c.m}\{k_i+2\}$)

9579: generated by $\gamma=(\g,...,\g)$ with

9580: $\g=\e^{-2\pi i J_0}(-1)^{\widehat{F}}$.

9581: It can be used to define a fine compactification to

9582: $3+1$ dimensions under the central charge condition

9583: \beq

9584: c=\sum_{i=1}^r{3k_i\over k_i+2}=9,

9585: \label{ccondition}

9586: \eeq

9587: and the condition on the number of factors

9588: \beq

9589: r:\,\,\,{\rm odd}.

9590: \label{oddc}

9591: \eeq

9592: The second condition is needed in order for

9593: the RR-charge of the lowest R-charge (corresponding to the holomorphic

9594: volume-form of the corresponding Calabi-Yau) to survive the orbifold

9595: projection.

9596: 

9597: 

9598: 

9599: Gepner models coming from the

9600: linear sigma models of the type described in Section~\ref{subsec:CYLG}

9601: always have $r=5$.

9602: But there are other models as well.

9603: The equation (\ref{ccondition}) has

9604: solutions with various number of factors, starting with $r=4$.

9605: $r=4$ solutions can be made into $r=5$ by adding a single $k=0$ factor.

9606: But there are solutions with $r>5$:

9607: According to \cite{Gepnerclassification},

9608: there are twenty-one models with minimal $r\geq 6$

9609:  --- fourteen with $r_{\rm min}=6$,

9610: four with $r_{\rm min}=7$, two with

9611: $r_{\rm min}=8$ and one with $r_{\rm min}=9$.

9612: (A solution with even $r$ {\it must be}

9613:  added an odd number of $k=0$ factors

9614: so that the condition (\ref{oddc}) is obeyed.)

9615: To be complete we consider these more general cases in this Appendix.

9616: 

9617: 

9618: 

9619: One important identity is

9620: \beq

9621: \mu:=\sum_{i=1}^r\left(1-{1\over k_i+2}\right)

9622: ={r+3\over 2}

9623: \label{mudef}

9624: \eeq

9625: where we have used the central charge condition (\ref{ccondition}).

9626: If we use the second condition (\ref{oddc}),

9627: we find that $\mu$ is an integer. Namely $\sum_i{1\over k_i+2}$

9628: is an integer. We have implicitly assumed this

9629: in the construction of the crosscap states: Look at the

9630: the sign factor $(-1)^{\sum_i{\nu\over k_i+2}}$

9631: in the RR part of the crosscap state (\ref{CAAA}).

9632: This does not make sense unless $\sum_i{1\over k_i+2}$

9633: is an integer.

9634: 

9635: 

9636: 

9637: In the main text of the paper starting Section~\ref{subsec:total}

9638: we have assumed that $r=5$. Here we would like to present some formula

9639: that is valid for all cases (with odd $r$).

9640: The relation between

9641: $\e^{\pi i J_0}|\Scr{C}_{P}\rangle$ and $|\Scr{C}_{(-1)^FP}\rangle$

9642: changes (for both A and B types) by the sign factor $(-1)^{\mu}$.

9643: Thus, the formulae for the total crosscap states are modified as

9644: \beqa

9645: (\ref{CARtotal})\longrightarrow&&

9646: |C_{\omega;{\bf m}}\rangle_{\RR}

9647: =|\Scr{C}_{P^A_{\omega;{\bf m}}}\rangle\otimes

9648: |\Scr{C}_+^{\rm \,st}\rangle_{\RR}

9649: -(-1)^{\mu}\omega

9650: |\Scr{C}_{(-1)^FP^A_{\omega;{\bf m}}}\rangle\otimes

9651: |\Scr{C}_-^{\rm \,st}\rangle_{\RR},

9652: \nn\\

9653: (\ref{CBRtotal})\longrightarrow&&

9654: |C^B_{\omega;{\bf m}}\rangle_{\RR}

9655: =|\Scr{C}_{P^B_{\omega;{\bf m}}}\rangle\otimes

9656: |\Scr{C}_+^{\rm \,st}\rangle_{\RR}

9657: -(-1)^{\mu}\tilomega^{-1}

9658: |\Scr{C}_{(-1)^FP^B_{\omega;{\bf m}}}\rangle\otimes

9659: |\Scr{C}_-^{\rm \,st}\rangle_{\RR},

9660: \nn

9661: \eeqa

9662: If $\mu$ is even (note that $\mu=4$ (even) if $r=5$),

9663: there is no difference in the discussion after 

9664: Section~\ref{subsec:total}.

9665: But there is some change if $\mu$ is odd.

9666: The largest effect is in the action of parities on the branes.

9667: The transformation rules (\ref{PAonD})-(\ref{PAonDshort})

9668:  and (\ref{PBonBb})-(\ref{PBonBshort})

9669: changes by sign (orientation).

9670: As a consequence, this affects the set of parity invariant D-branes.

9671: The analysis of the structure of

9672: Chan-Paton factor goes through as in the discussion of

9673: $r=5$, with an obvious modification of the result.

9674: 

9675: Addition of two $k=0$ factors shifts even $\mu$ to odd $\mu$ and vice versa.

9676: What we have seen is that

9677: this has a non-trivial effect on the physics involving branes in the

9678: orientifold model.

9679: In fact, without orientifold, addition of

9680: even number of $k=0$ factors makes no difference since

9681: the orbifold action is trivial on such pair of $k=0$ factors.

9682: This is also true for the case involving D-branes (before orientifold):

9683: for A-branes, the orbifold group flips the orientation of

9684: the brane in a $W=X^2$ factor

9685: but a pair of such flips cancel against each other.

9686: For B-branes the same can be said (this is known as

9687: the Kn\"orrer periodicity \cite{Knorrer}).

9688: However, with an orientifold, this step 2 periodicity

9689: is doubled to step 4.

9690: This reminds us of the Bott periodicity:

9691: complex K-theory has periodicity 2 but

9692: Real K-theory has periodicity 8.

9693: 

9694: 

9695: 

9696: 

9697: \appendix{Some Detail}\label{app:detail1}

9698: 

9699: We explain the projection factors (\ref{AshProj}) and (\ref{AshPA})

9700: that are used to read the parity action on short-orbit branes.

9701: 

9702: We first compute the $\Z_2$ projection factor for the open

9703: string stretched from a short-orbit brane

9704: $\widehat{B}^{(\varepsilon)}_{\bf \bar{L},\bar{M}}$

9705: to another $\widehat{B}^{(\varepsilon')}_{\bf L,M}$.

9706: To this end, let us consider the loop-channel expansion of

9707: the relevant overlaps in the minimal model with even $k$,

9708: \beqa

9709:  && \langle\Scr{B}_{\frac{k}{2},\bar{M},\bar{S}}|

9710:   q_t^{H}|\Scr{B}_{\frac{k}{2},M,S}\rangle_{\NSNS\atop \RR}

9711:  = \sum_{l\atop s=0,1}N_{\frac{k}{2}\frac{k}{2}}^l

9712:      (\pm1)^s\chi_{l,\bar{M}-M,\bar{S}-S+2s}(\tau_l),

9713:  \nonumber \\

9714:  && \langle\Scr{B}_{\frac{k}{2},\bar{M},\bar{S}}|

9715:   q_t^{H}|\Scr{B}_{\frac{k}{2},M,S}\rangle_{(\mp1)^Fa^{k+2}}

9716:  = \sum_{l\atop s=0,1}N_{\frac{k}{2}\frac{k}{2}}^l

9717:      (\pm1)^s(-1)^{\frac{1}{2}(l+\bar{M}-M-\bar{S}^2+S^2)}

9718: \chi_{l,\bar{M}-M,\bar{S}-S+2s}(\tau_l)

9719:  \nonumber\\

9720: \eeqa

9721: This is enough to see 

9722: the open string states labeled by

9723: $\otimes_{i=1}^r(l_i,n_i,s_i)$ are subject to the projection

9724: \beq

9725:   \frac12\left(1+\gamma^{H/2}\right) ~=~

9726:   \frac12\left(1+\varepsilon\varepsilon'

9727:  \prod_{w_i\,{\rm odd}}(-1)^{\frac{1}{2}(l_i+n_i-s)}\right),

9728: %\label{AshProj}

9729: \eeq

9730: where $s=0$ for NS states and $1$ for R ones.

9731: This is nothing but (\ref{AshProj}).

9732: 

9733: 

9734: Let us next find the projection factor that appears in the

9735: parity twisted partition function.

9736: To do this, let us consider the minimal model with even $k$,

9737: and take the M\"{o}bius strip amplitude

9738: \beqa

9739: \lefteqn{

9740:   _{\NSNS}\langle\Scr{B}_{\frac{k}{2},M,S}|q_t^H

9741:   |\Scr{C}^A_{(\pm1)^Fg^m\widetilde{P}_A}\rangle

9742: } \nonumber \\ &=&

9743:    \sum_l\delta^{(2)}_l(-1)^{m+\frac{l}{2}}

9744:    \left\{e^{\mp\frac{\pi i}{4}}\hat{\chi}_{l,2M-2m,2S}

9745:          -e^{\pm\frac{\pi i}{4}}\hat{\chi}_{l,2M-2m,2S+2}\right\}(\tau)

9746: \label{mobminev}

9747: \eeqa

9748: where

9749: \beq

9750:  \hat{\chi}_{l,n,s}(\tau) =

9751:  e^{-\pi i(\frac{l(l+2)-n^2}{4k+8}+\frac{s^2}{8}-\frac{c}{24})}

9752:  \chi_{l,n,s}(\tau+1/2).

9753: \eeq

9754: Let us see how it changes under the shift

9755: $m\to m+\frac{k+2}{2}$.

9756: This tell us that the open string state $(l,n,s)$ on the M\"{o}bius

9757: strip (\ref{mobminev}) has the following eigenvalue of $g^{\frac{k+2}{2}}$:

9758: \beq

9759:  g^{\frac{k+2}{2}} = \pm i(-1)^{\frac{l+n}{2}+S}.

9760: \label{gkact}

9761: \eeq

9762: One can also perform a similar analysis on the RR sector states.

9763: Let us then take a short orbit A-brane

9764: $\widehat{B}^{\pm,A}_{{\bf L,M}}$ and an A-parity

9765: $P^A_{\omega,{\bf m}}$.

9766: Using (\ref{gkact}) for each minimal model, we can now easily

9767: find the eigenvalue of $\gamma^{H/2}$ for the open string state

9768: $\otimes_{i=1}^r(l_i,n_i,s_i)$ on the M\"{o}bius strip.

9769: The result is independent of the choice of branes and

9770: reads

9771: \beq

9772:   \gamma^{H/2} = \omega^{\frac{H}{2}}(-1)^{\frac{\sigma}{2}}

9773:                  \prod_{w_i~{\rm odd}}(-1)^{\frac{1}{2}(l_i+n_i-s)},

9774: \eeq

9775: where $s=0$ for NS states and $1$ for R ones as before, and

9776: $\sigma$ is the number of $i$'s such that $w_i$ is odd.

9777: This leads to the projection factor (\ref{AshPA}).

9778: 

9779: \appendix{Integral bases of three-cycles in the quintic}

9780: \label{integral}

9781: 

9782: In the main text, we have seen that the most convenient

9783: way to write down and solve the tadpole conditions is to

9784: find an integral basis of the charge lattice, so that the

9785: coefficients in the equations (\ref{tadcan}) are manifestly

9786: integer. In this appendix, we describe how such a basis

9787: can be found for A-type branes on the quintic. We will also

9788: describe an alternative way of solving the tadpole

9789: conditions using $g_i$ polynomials.

9790: 

9791: \subsubsection*{Integral basis}

9792: 

9793: In a single minimal model with $k=3$, the charges of

9794: A-type branes span a 4-dimensional lattice which is

9795: generated by the $5$ branes $\Scr{B}^A_{0,M,1}$ with

9796: $M=1,3,5,7,9$. As is easiest to see in the Landau-Ginzburg

9797: picture, these $5$ charges satisfy one linear relation:

9798: Their sum is zero. We can fit the set $\Lambda$ of 

9799: five charges modulo this one relation into an

9800: exact sequence

9801: \beq

9802: 0\exact \{R\}\exact \A_5\exact \Lambda

9803: \exact 0

9804: \label{exactsingle}

9805: \eeq 

9806: where $\A_5$ stands for the set $\{n=0,1,2,3,4\}$

9807: representing the charges $Q_n$ of the brane

9808: $\Scr{B}^A_{0,2n+1,1}$, and $R$ is the relation

9809: $$

9810: R \mapsto \sum_{n} Q_n

9811: $$

9812: There is an obvious $\Z_5$ action on $\A_5$ and on

9813: (\ref{exactsingle}) which cyclically

9814: permutes the five elements, and leaves $R$ invariant.

9815: 

9816: If we now take the tensor product of $5$ such minimal models,

9817: the charge lattice has dimension $4^5=1024$. It is generated

9818: by the tensor products $Q({\bf n}) = \prod_i Q_{n_i}$ modulo

9819: the relations

9820: $$

9821: R_1(i;{\bf n}) = \sum_{n_i} Q(n_1,\ldots,n_5)

9822: $$

9823: where ${\bf n}=(n_1,\ldots,\widehat{n_i},\ldots n_5)$. Thus

9824: we have $5^5$ charges with $5^5$ relations, but these 

9825: relations are not all independent. Namely, we have the

9826: relations between relations

9827: $$

9828: R_2(i,j;{\bf n}) = \sum_{n_i} R_1(j;({\bf n},n_i))

9829: - \sum_{n_j} R_1(i;({\bf n},n_j))

9830: $$

9831: where now ${\bf n}=(n_1,\ldots,\widehat{n_i},

9832: \ldots,\widehat{n_j},\ldots, n_5)$.

9833: Continuing this way, we obtain the long-exact sequence for

9834: the set of charges $\Lambda^{\rm ten}$ of the tensor product

9835: \beq

9836: 0

9837: \exact \{R_5\}

9838: \exact 5 \A_5

9839: \exact 10 (\A_5)^2

9840: \exact 10 (\A_5)^3

9841: \exact 5 (\A_5)^4

9842: \exact (\A_5)^5

9843: \exact \Lambda^{\rm ten}

9844: \exact 0

9845: \label{exacttensor}

9846: \eeq

9847: from which we see that the dimension of the charge lattice

9848: is indeed $4^5$. The advantage of this representation is that

9849: it is now trivial to take the $\Z_5$ orbifold. All

9850: relations $R_s$ with $s<5$ are related to one another under the

9851: diagonal $\Z_5$ action, while $R_5$ is invariant. Thus,

9852: the untwisted charges can be represented by the sequence

9853: \beq

9854: 0\exact \{R_5\}

9855: \exact 5 (\A_5)^0

9856: \exact 10 \A_5

9857: \exact 10 (\A_5)^2

9858: \exact 5 (\A_5)^3

9859: \exact (\A_5)^4

9860: \exact \Lambda^{\rm Gep}

9861: \exact 0

9862: \label{exactgepner}

9863: \eeq

9864: from which we read off the dimensions of the charge lattice

9865: of the Gepner model to be $204$, as expected. To obtain a basis 

9866: of $\Lambda^{\rm Gep}$, we take a section of (\ref{exactgepner}). 

9867: In view of solving the tadpole conditions, it is most useful 

9868: to do this in such a way that respects the action of the parity. 

9869: The parity acts in a single minimal model on $Q_n$ as $0\mapsto 4$,

9870: $1\mapsto 3$, and $2\mapsto 2$, and similarly on all the

9871: relations. We will show how this language simplifies finding

9872: the explicit form of (\ref{tadcan}) in appendix

9873: \ref{orbifold}.

9874: 

9875: 

9876: 

9877: 

9878: \subsubsection*{Solving the tadpole conditions with

9879: $g_i$ polynomials}

9880: 

9881: The tadpole cancellation condition can also be written in a

9882: simple form by using the $g_i$-polynomials.

9883: Here we again restrict our attention to the RR-charges and

9884: tadpoles sitting in the untwisted sector.

9885: Let us introduce the $(k_i+2)$-dimensional shift matrices

9886: $g_i$ satisfying $g_i^{k_i+2}=1$,

9887: and associate the following polynomial to each D-brane

9888: \beq

9889:  Q_{\bf L,M}(g_i)=\prod_{i=1}^5

9890:  \left(\sum_{n_i=0}^{L_i}g_i^{n_i+(M_i-L_i)/2}\right),

9891: \label{QLM}

9892: \eeq

9893: representing its RR-charge.

9894: Let us also associate similar $g_i$-polynomials to the orientifolds

9895: by first expressing their RR-charges in terms of D-branes and

9896: then using the above formula.

9897: These polynomials of $g_i$ are useful in computing the (twisted)

9898: Witten indices between D-branes and O-planes.

9899: The index between the branes $B$ and $B'$ is given by

9900: the diagonal element(more precisely, the product of diagonal elements)

9901: of the matrix

9902: \begin{equation}

9903:   I = Q_B(g_i)Q_{B'}(g_i^{-1})\prod_{i=1}^5(1-g_i^{-1})

9904:   \sum_{\nu=1}^H(g_1g_2g_3g_4g_5)^\nu

9905: \end{equation}

9906: Since the last factor in the right hand side is the projection

9907: onto the states on which $g_5^{-1}=g_1g_2g_3g_4$, we can eliminate

9908: $g_5$ and obtain

9909: \begin{eqnarray}

9910:   I &=& Q_B(g_i)Q_{B'}(g_i^{-1})

9911:  (1-g_1^{-1})(1-g_2^{-1})(1-g_3^{-1})(1-g_4^{-1})(1-g_1g_2g_3g_4)

9912:  \nn\\ && \times

9913:  \sum_{\nu=1}^{w_5}(g_1g_2g_3g_4)^{\nu(k_5+2)},

9914: \end{eqnarray}

9915: where the index is read off as the diagonal elements.

9916: This agrees with the formula for quintic given in \cite{BDLR}.

9917: The RR-charges of any configurations of branes and the O-plane

9918: are therefore expressed as polynomials of $g_i$,

9919: \beq

9920:   (\mbox{RR-charge})=

9921:   \sum_{m_i=0}^{k_i+1}

9922:   N_{m_1m_2m_3m_4m_5}g_1^{m_1}g_2^{m_2}g_3^{m_3}g_4^{m_4}g_5^{m_5}.

9923: \label{gpol}

9924: \eeq

9925: The configuration is free of tadpoles when the sum of $g_i$-polynomials

9926: of the constituent D-branes and the O-plane vanishes up to

9927: \beq

9928:  1+g_i+\cdots+g_i^{k_i+1}=0,~~~

9929:  g_1g_2g_3g_4g_5=1.

9930: \label{defgi}

9931: \eeq

9932: It is cumbersome to have these equivalence relations in

9933: analyzing the polynomial.

9934: Therefore, it is more convenient to use the relations

9935: (\ref{defgi}) to bring the polynomial into the following gauge

9936: \beq

9937:  \sum_{m_a=0}^{k_a+1}N_{m_1\cdots m_a\cdots m_5}=0,~~~

9938:  N_{m_1m_2m_3m_4m_5}=N_{m_1+1,m_2+1,m_3+1,m_4+1,m_5+1}

9939: \label{gauge}

9940: \eeq

9941: and see whether each coefficient is vanishing or not.

9942: This is certainly possible because each of the polynomials

9943: $Q_{\bf L,M}(g_i)$ can be brought to this gauge (\ref{gauge})

9944: in the following way.

9945: \beqa

9946:  Q_{\bf L,M}(g_i) &=&

9947:  \frac1H\sum_{\nu=0}^{H-1}

9948:  \sum_{n_i=0}^{k_i+1}\prod_{i=1}^5

9949:  (f_{L_i,n_i}g_i^{n_i+\nu+(M_i-L_i)/2)}),\nn\\

9950:  f_{L_i,n_i} &=& \left\{

9951:  \begin{array}{ll}

9952:   1-{\textstyle\frac{L_i+1}{k_i+2}}&(0\le n_i\le L_i)\\

9953:    -{\textstyle\frac{L_i+1}{k_i+2}}&({\rm otherwise})

9954:  \end{array}\right.

9955: \eeqa

9956: 

9957: As was noted before, these $g_i$-polynomials can only express the

9958: RR-charges corresponding to polynomial deformations of hypersurfaces

9959: defining the target space.

9960: In the two parameter model with $(k_i+2)=(8,8,4,4,4)$ there are

9961: six missing RR charges sitting in the twisted sector.

9962: One can develop a similar argument using polynomials

9963: for those RR-charges, too.

9964: 

9965: We can again see that the number of independent components of 

9966: $N_{m_1m_2m_3m_4m_5}$ agrees with the number of RR ground states in 

9967: the untwisted sector, which take the form (\ref{ground}),

9968: \beq

9969:   |l_i\rangle = \prod_{i=1}^5|l_i,l_i+1,1\rangle\times|l_i,-l_i-1,-1\rangle~~~

9970:   (1\le l_i+1\le k_i+1,~\sum_{i}{\textstyle\frac{l_i+1}{k_i+2}}\in\Z)

9971: \eeq

9972: To see this, let us take the Fourier transform of $N_{m_1m_2m_3m_4m_5}$:

9973: \beq

9974:   \widetilde{N}_{n_1n_2n_3n_4n_5}

9975:  = \sum_{m_i\in\Z_{k_i+2}}

9976:    \exp(\sum_i\frac{2\pi im_in_i}{k_i+2})N_{m_1m_2m_3m_4m_5}

9977: \eeq

9978: Then $\widetilde{N}_{n_1n_2n_3n_4n_5}$ are nonzero only when

9979: $n_i$ are all nonzero mod $k_i+2$ and $\sum_i\frac{n_i}{k_i+2}\in\Z$,

9980: which is the same condition as the RR ground states satisfy

9981: under the identification $n_i\leftrightarrow l_i+1$ (mod $k_i+2$).

9982: So $\widetilde{N}_{n_1n_2n_3n_4n_5}$ has as many independent

9983: components as there are untwisted RR vacua.

9984: 

9985: Of all the linear equations, there are some equations

9986: among the tadpole cancellation conditions in which

9987: all the D-branes appear with positive definite coefficients.

9988: These essentially come from the overlaps with the RR ground states

9989: $|\nu\rangle$, namely, those with $l_i=\nu$ mod $k_i+2$ for all $i$.

9990: The reason for the positivity is that the overlaps of D-branes

9991: or O-planes with any of these states have the same phases

9992: if they preserve the same spacetime supersymmetry.

9993: These equations are particularly important,

9994: because they ensure that there are only finite number of

9995: tadpole canceling configurations.

9996: These equations also contain the condition that the sum of D-brane

9997: tensions must cancel the O-plane tension.

9998: One can obtain these special equations from

9999: $g_i$-polynomials by setting $g_i=g^{w_i}$, where $g$ is

10000: a $H$-dimensional shift matrix.

10001: 

10002: 

10003: 

10004: \appendix{Tadpole conditions for $\Z_5$ orbifold of quintic}

10005: \label{orbifold}

10006: 

10007: In this appendix, we discuss the tadpole conditions for Type

10008: IIB orientifolds of the orbifold of the quintic by the $\Z_5$

10009: symmetry called $[1,4,0,0,00]$ in the notation of \cite{GreenePlesser}.

10010: By mirror symmetry, this is equivalent to Type IIA orientifold

10011: of $(\Z_5)^2$ orbifold of the quintic. In B-type language, the

10012: model has $2h^{1,1}+2=12$ RR charges to cancel (the fact

10013: that $h^{2,1}=49$ will not be important). The analogs of RS

10014: branes in such orbifolds have been discussed for instance in 

10015: \cite{BD}, and are also straightforward to obtain in the 

10016: Landau-Ginzburg picture. It is easy to see that the branes 

10017: are labeled as $\Scr{B}_{{\bf L},{\bf M},S}$ with ${\bf L}$ 

10018: as before and ${\bf M}=(M_1,M_2,M_3,M_4,M_5)$ modulo a $(\Z_5)^3$ 

10019: identification.

10020: 

10021: We can present an integral basis of the charge lattice similarly

10022: to (\ref{exactgepner}) via

10023: \beq

10024: 0\exact {R_3}

10025: \exact 3 (\A_5)^0

10026: \exact 3 \A_5

10027: \exact (\A_5)^2

10028: \exact \Lambda^{\rm orbi}

10029: \exact 0

10030: \label{exactorbi}

10031: \eeq

10032: such that $\Lambda^{\rm orbi}$ indeed has dimension $25-15+3-1=12$.

10033: We now take a section through (\ref{exactorbi}) and specify an

10034: integral basis as the charges of the branes with ${\bf L}={\bf 0}$

10035: and ${\bf M}=2 {\bf n}+1$ with

10036: $$

10037: \begin{array}{l}

10038: {\bf n}\in\Bigl\{

10039: [0,0,2,2,2],

10040: [0,4,2,2,2], 

10041: [1,0,2,2,2], 

10042: [1,1,2,2,2], 

10043: [1,3,2,2,2], 

10044: [1,4,2,2,2],\\

10045: \qquad

10046: [4,4,2,2,2], 

10047: [4,0,2,2,2], 

10048: [3,4,2,2,2], 

10049: [3,3,2,2,2], 

10050: [3,1,2,2,2], 

10051: [3,0,2,2,2]\Bigr\}

10052: \end{array}

10053: $$

10054: where $[\cdots]$ denotes $(\Z_5)^3$ orbits and where the second 

10055: line is obviously the parity image of the first.

10056: 

10057: As before, branes preserving the same supersymmetry as the O-plane have 

10058: arbitrary ${\bf L}$ and $\sum M_i=0\bmod 5$ and appropriate $S$ label 

10059: depending on the parity of $\sum L_i$. Obviously, the charge of

10060: such branes does not depend on the permutations of $L_3,L_4,L_5$, 

10061: so we have the following representatives of ${\bf L}$ labels

10062: \beq

10063: \begin{array}{ll}

10064: {\bf L}= 

10065: &(0,0,0,0,0), \\

10066: &(1,0,0,0,0),(0,1,0,0,0),(0,0,1,0,0),\\

10067: &(1,1,0,0,0),(1,0,1,0,0),(0,1,1,0,0),(0,0,1,1,0),\\

10068: &(1,1,1,0,0),(1,0,1,1,0),(0,1,1,1,0),(0,0,1,1,1),\\

10069: &(1,1,1,1,0),(1,0,1,1,1),(0,1,1,1,1)\\

10070: &(1,1,1,1,1)

10071: \end{array}

10072: \eeq

10073: where we have ordered the branes according to increasing tension.

10074: The possible ${\bf M}$ labels are ($\bmod 5$ and modulo $(\Z_5)^3$):

10075: \beq

10076: {\bf M}=

10077: [0,0,0,0,0],

10078: [1,4,0,0,0], 

10079: [4,1,0,0,0],

10080: [2,3,0,0,0],

10081: [3,2,0,0,0] 

10082: \eeq

10083: The first of these is obviously invariant under the parity,

10084: while the others are each others image. Thus, we have a total

10085: of $16\times 3=48$ different charges to consider in the

10086: tadpole cancellation. We will denote by $n_1,\ldots

10087: n_{48}$ the number of times a given charge appears.

10088: 

10089: By utilizing the well-known expressions for the charges of

10090: branes in the minimal model (see (\ref{QLM}) or (\ref{branec})),

10091: we can compute the RR charges of all branes on the list in

10092: the basis described above. This leads to the following

10093: $6$ equations on the $48$ $n_i$'s

10094: \beq

10095: (n_1,\ldots,n_{48})

10096: \left(

10097: \hbox{\scriptsize

10098: $\begin{array}{cccccc}

10099: 2&2&0&0&0&0\\

10100: %

10101: 2&2&1&0&0&0\\

10102: %

10103: 1&1&1&-1&0&0\\

10104: %

10105: 2&0&0&1&0&-1\\

10106: %

10107: 2&0&0&1&-1&1\\

10108: %

10109: 1&0&1&0&1&0\\

10110: %

10111: 2&0&0&1&0&1\\

10112: %

10113: 2&0&1&1&1&-1\\

10114: %

10115: 1&0&0&0&-1&0\\

10116: %

10117: 2&0&-1&2&0&0\\

10118: %

10119: 2&0&1&0&0&0\\

10120: %

10121: 1&0&1&0&0&0\\

10122: %

10123: 4&2&2&0&0&0\\

10124: %

10125: 4&2&1&0&0&0\\

10126: %

10127: 2&1&0&1&0&0\\

10128: %

10129: 4&2&0&1&-1&-1\\

10130: %

10131: 4&2&2&0&1&0\\

10132: %

10133: 2&1&1&0&0&1\\

10134: %

10135: 4&2&1&1&1&1\\

10136: %

10137: 4&2&1&0&-1&0\\

10138: %

10139: 2&1&1&0&0&-1\\

10140: %

10141: 4&2&-1&2&0&0\\

10142: %

10143: 4&2&2&0&0&0\\

10144: %

10145: 2&1&2&-1&0&0\\

10146: %

10147: 6&2&3&0&0&0\\

10148: %

10149: 6&2&1&2&0&0\\

10150: %

10151: 3&1&0&1&0&0\\

10152: %

10153: 6&2&0&2&-1&-2\\

10154: %

10155: 6&2&2&1&0&1\\

10156: %

10157: 3&1&2&0&1&1\\

10158: %

10159: 6&2&1&2&1&2\\

10160: %

10161: 6&2&2&1&0&-1\\

10162: %

10163: 3&1&1&0&-1&-1\\

10164: %

10165: 6&2&-2&4&0&0\\

10166: %

10167: 6&2&3&0&0&0\\

10168: %

10169: 3&1&3&-1&0&0\\

10170: %

10171: 10&4&5&0&0&0\\

10172: %

10173: 10&4&2&2&0&0\\

10174: %

10175: 5&2&0&2&0&0\\

10176: %

10177: 10&4&0&3&-2&-3\\

10178: %

10179: 10&4&4&1&1&1\\

10180: %

10181: 5&2&3&0&1&2\\

10182: %

10183: 10&4&2&3&2&3\\

10184: %

10185: 10&4&3&1&-1&-1\\

10186: %

10187: 5&2&2&0&-1&-2\\

10188: %

10189: 16&6&8&0&0&0\\

10190: %

10191: 16&6&3&4&0&0\\

10192: %

10193: 8&3&0&3&0&0

10194: \end{array}$}

10195: \right)

10196: =(32, 12, 0, 12, 0, 0)

10197: \label{orbicond}

10198: \eeq

10199: Inspection reveals that the first two of these equations are nothing

10200: but the equations (\ref{condition}) that we have solved in the context

10201: of B-type orientifold of quintic. To see this, one has to take into 

10202: account that for branes not invariant under parity, $n_i$ denotes the 

10203: number of times the brane and its image under parity appear, and the 

10204: fact that the mass depends only on the number of $1$'s in the ${\bf L}$ 

10205: label. Thus, to find solutions of (\ref{orbicond}), we can take some

10206: solution of (\ref{condition}) and scan through all ways of distributing

10207: this mass among the branes on the list of $48$ with the same tension.

10208: 

10209: Here are a few examples of tadpole canceling brane configurations for the

10210: $\Z_5$ orbifold of the quintic obtained in this way.

10211: \\

10212: \underline{Example 1}

10213: $$

10214: (B_{(00111), (20111)} + \mbox{\footnotesize image})

10215: +(\overline{B_{(11110), (39116)}} + \mbox{\footnotesize image}

10216: )

10217: +2 B_{(11111), (11111)}

10218: $$

10219: \underline{Example 2}

10220: \beqa

10221: &4 (\overline{B_{(00000), (20666)}} +\mbox{\footnotesize image})

10222: +2 (B_{(00100), (20166)} +\mbox{\footnotesize image}) 

10223: +B_{(00100), (66166)} 

10224: \nn\\ 

10225: &+(B_{(01110), (25116)}+\mbox{\footnotesize image}) + 

10226: (B_{(10000), (34666)}+\mbox{\footnotesize image}) + 

10227: 3(B_{(10000), (70666)}+\mbox{\footnotesize image})

10228: \nn\\

10229: & + 

10230: \overline{B_{(11110), (11116)}}

10231: \nn

10232: \eeqa

10233: \underline{Example 3}

10234: \beqa

10235: &

10236: 3 (\overline{B_{(00000), (20666)}}+\mbox{\footnotesize image})

10237:  + 

10238: \overline{B_{(00000), (66666)}} + 

10239: (\overline{B_{(00000), (84666)}}+\mbox{\footnotesize image})

10240: \nn\\

10241: & + 

10242: 5 (B_{(00100), (20166)}+\mbox{\footnotesize image}) + 

10243: B_{(00100), (66166)} + 

10244: 2B_{(01000), (61666)} + 

10245: 2 B_{(10000), (16666)} \nn\\

10246: & + B_{(11111), (11111)}

10247: \nn

10248: \eeqa

10249: 

10250: \subsubsection*{Particle spectrum}

10251: 

10252: The spectrum of massless matters for these brane configurations

10253: is analyzed in a similar way as in the case of ordinary quintic.

10254: Here we only present the results.

10255: {\footnotesize

10256: \begin{center}

10257: \begin{tabular}{|c|cc||cc|cc|c|}

10258: \hline

10259: ${}^{\rm gauge}_{\rm group}$ &

10260: $\sharp$&${\bf L},~{\bf M},~S$ &

10261: 1&2&3&4&5 \\\hline\hline

10262: \lw{$U(1)$}

10263: &1&(00111),(20111),0 & 12 &  0 &  0 &  3 &  0 \\

10264: &2&(00111),(02111),0 &  0 & 12 &  3 &  0 &  0 \\ \hline

10265: \lw{$U(1)$}

10266: &3&(11110),(39116),2 &  0 &  3 & 12 & 7+3& 10 \\

10267: &4&(11110),(93116),2 &  3 &  0 & 7+3& 12 & 10 \\ \hline

10268: $O(2)$

10269: &5&(11111),(11111),0 &  0 &  0 & 10 & 10 & 13+12 \\

10270: \hline

10271: \end{tabular}

10272: \vskip2mm

10273: Example 1

10274: \end{center}

10275: }

10276: 

10277: The first example consists of five kinds of branes,

10278: $B_1+B_2+B_3+B_4+2B_5$, where $B_5$ is parity invariant

10279: and $B_2,B_4$ are parity images of $B_1,B_3$.

10280: The spectrum of chiral matters is summarized in the table above.

10281: 

10282: The gauge group is $U(1)^2\times O(2)$, and the labels

10283: $({\bf L,M},S)$ of five D-branes are presented in the

10284: second column.

10285: The $5\times5$ numbers give the multiplicities of chiral primary

10286: states on $i\!-\!j$ string ($i,j=1,\cdots,5$).

10287: $3\!-\!4, 4\!-\!3$ and $5\!-\!5$ strings are parity invariant,

10288: and they belong to symmetric or antisymmetric tensor representations

10289: of gauge group according to their parity eigenvalues.

10290: The numbers $7+3$ or $13+12$ represent the multiplicities of

10291: symmetric and antisymmetric representations.

10292: 

10293: The table contains nine blocks.

10294: Upper off-diagonal blocks are related with lower

10295: off-diagonal ones by parity, namely,

10296: the multiplicity of matters on $i\!-\!j$ string is the same

10297: as that of $P(j)\!-\!P(i)$ string.

10298: As was explained in section \ref{sec:FI}, the spectrum is chiral

10299: if there is a block with the following property

10300: \begin{itemize}

10301: \item An off-diagonal block corresponding to one

10302:       unitary and one non-unitary groups, with numbers

10303: $$

10304:   \left(\begin{array}{c} a\\b\end{array}\right)~~{\rm or}~~

10305:   (a~b),~~~

10306:   a\ne b

10307: $$

10308: \item An off-diagonal block corresponding to two different

10309:       unitary groups, with numbers

10310: $$

10311:   \left(\begin{array}{cc} a&b\\c&d\end{array}\right), ~~~

10312:   a\ne d ~~{\rm or}~~ b\ne c.

10313: $$

10314: \item A diagonal block for a unitary group, with numbers

10315: $$

10316:   \left(\begin{array}{cc} a&b_s+b_a\\c_s+c_a&a\end{array}\right), ~~~

10317:   b_s\ne c_s~~{\rm or}~~b_a\ne c_a

10318: $$

10319: \end{itemize}

10320: The table shows that the first example is non-chiral.

10321: 

10322: The other two examples are chiral, as can be seen

10323: from the tables below.

10324: Note that there are no antisymmetric tensor representations of $U(1)$

10325: or $O(1)$.

10326: \newpage

10327: {\footnotesize

10328: \begin{center}

10329: \begin{tabular}{|c|cc||cc|cc|c|cc|cc|cc|c|}

10330: \hline

10331: ${}^{\rm gauge}_{\rm group}$ &

10332: $\sharp$&${\bf L},~{\bf M},~S$ &

10333: 1&2&3&4&5&6&7&8&9&10&11&12 \\\hline\hline

10334: \lw{$U(4)$}

10335: &1&(00000),(20666),2 &  0&0&2&0&1&0&1&0&0&0&1&0 \\

10336: &2&(00000),(02666),2 &  0&0&0&2&1&0&0&1&3&0&0&0 \\ \hline

10337: \lw{$U(2)$}

10338: &3&(00100),(20166),0 &  2&0&2&0&1&0&3&1&2&0&0&0 \\

10339: &4&(00100),(02166),0 &  0&2&0&2&1&0&0&0&0&2&0&0 \\ \hline

10340: $O(1)$

10341: &5&(00100),(66166),0 &  1&1&1&1&2&2&1&2&0&1&0&3 \\ \hline

10342: \lw{$U(1)$}

10343: &6&(01110),(25116),0 &  0&1&0&3&1&5&1&0&1&2&1&5 \\

10344: &7&(01110),(07116),0 &  0&0&0&0&2&1&5&1&0&1&2&2 \\ \hline

10345: \lw{$U(1)$}

10346: &8&(10000),(34666),0 &  3&0&0&2&0&0&1&0&0&3&0&2 \\

10347: &9&(10000),(98666),0 &  1&0&0&1&2&1&0&3&0&1&0&1 \\ \hline

10348: \lw{$U(3)$}

10349: &10&(10000),(70666),0 &  0&1&0&0&0&2&1&0&0&0&1&0 \\

10350: &11&(10000),(52666),0 &  0&0&2&0&1&1&2&1&3&0&0&2 \\ \hline

10351: $O(1)$

10352: &12&(11110),(11116),2 &  0&0&0&0&3&2&5&1&2&2&0&6 \\

10353: \hline

10354: \end{tabular}

10355: \vskip0mm

10356: Example 2

10357: \vskip5mm

10358: \begin{tabular}{|c|cc||cc|c|cc|cc|c|c|c|c|}

10359: \hline

10360: ${}^{\rm gauge}_{\rm group}$ &

10361: $\sharp$&${\bf L},~{\bf M},~S$ &

10362: 1&2&3&4&5&6&7&8&9&10&11 \\\hline\hline

10363: \lw{$U(3)$}

10364: &1&(00000),(20666),2 &0&0&0&0&0&2&0&1&0&3&0 \\

10365: &2&(00000),(02666),2 &0&0&0&0&0&0&2&1&3&0&0 \\ \hline

10366: $O(1)$

10367: &3&(00000),(66666),2 &0&0&0&0&0&1&1&2&0&0&1 \\ \hline

10368: \lw{$U(1)$}

10369: &4&(00000),(84666),2 &0&0&0&0&0&1&0&0&0&1&0 \\

10370: &5&(00000),(48666),2 &0&0&0&0&0&0&1&0&1&0&0 \\ \hline

10371: \lw{$U(5)$}

10372: &6&(00100),(20166),0 &2&0&1&1&0&2&0&1&1&0&0 \\

10373: &7&(00100),(02166),0 &0&2&1&0&1&0&2&1&0&1&0 \\ \hline

10374: $O(1)$

10375: &8&(00100),(66166),0 &1&1&2&0&0&1&1&2&0&0&3 \\ \hline

10376: $O(2)$

10377: &9&(01000),(61666),0 &3&0&0&1&0&0&1&0&0&3&1 \\ \hline

10378: $O(2)$

10379: &10&(10000),(16666),0 &0&3&0&0&1&1&0&0&3&0&1 \\ \hline

10380: $O(1)$

10381: &11&(11111),(11111),0 &0&0&1&0&0&0&0&3&1&1&13 \\

10382: \hline

10383: \end{tabular}

10384: \vskip1mm

10385: Example 3

10386: \end{center}

10387: }

10388: 

10389: 

10390: \newpage

10391: 

10392: 

10393: \begin{thebibliography}{99}

10394: 

10395: \small

10396: \parskip=0pt plus 2pt

10397: 

10398: 

10399: \bibitem{DougString}

10400: M.~Douglas, ``Mastering N=1'', talk at Strings 2002, Cambridge,

10401: July 17, 2002;\\

10402: http://www.damtp.cam.ac.uk/strings02/avt/douglas/

10403: 

10404: \bibitem{GKP}

10405: S.~B.~Giddings, S.~Kachru and J.~Polchinski,

10406: ``Hierarchies from fluxes in string compactifications,''

10407: Phys.\ Rev.\ D {\bf 66} (2002) 106006

10408: [hep-th/0105097].

10409: 

10410: 

10411: \bibitem{KKLT}

10412: S.~Kachru, R.~Kallosh, A.~Linde and S.~P.~Trivedi,

10413: ``De Sitter vacua in string theory,''

10414: Phys.\ Rev.\ D {\bf 68} (2003) 046005

10415: [hep-th/0301240].

10416: 

10417: \bibitem{Gepner}

10418: D.~Gepner,

10419: ``Space-Time supersymmetry in compactified string

10420: theory and superconformal models,''

10421: Nucl.\ Phys.\ B {\bf 296} (1988) 757;

10422: ``Exactly solvable string

10423: compactifications on manifolds of $SU(N)$ holonomy,''

10424: Phys.\ Lett.\ B {\bf 199} (1987) 380.

10425: 

10426: \bibitem{Cardy}

10427: J.~L.~Cardy,

10428: ``Boundary Conditions, Fusion Rules And The Verlinde Formula,''

10429: Nucl.\ Phys.\ B {\bf 324}, 581 (1989).

10430: %%CITATION = NUPHA,B324,581;%%

10431: 

10432: \bibitem{RS}

10433: A.~Recknagel and V.~Schomerus,

10434: ``D-branes in Gepner models,''

10435: Nucl.\ Phys.\ B {\bf 531} (1998) 185

10436: [hep-th/9712186].

10437: 

10438: \bibitem{BDLR}

10439: I.~Brunner, M.~R.~Douglas, A.~E.~Lawrence and C.~Romelsberger,

10440: ``D-branes on the quintic,''

10441: JHEP {\bf 0008} (2000) 015

10442: [hep-th/9906200].

10443: 

10444: 

10445: 

10446: \bibitem{PSS}

10447: G.~Pradisi, A.~Sagnotti and Y.~S.~Stanev,

10448: ``Planar duality in SU(2) WZW models'',

10449: Phys.\ Lett.\ B {\bf 354} (1995) 279

10450: [hep-th/9503207];

10451: ``The Open descendants of nondiagonal SU(2) WZW models,''

10452: Phys.\ Lett.\ B {\bf 356},  (1995) 230

10453: [hep-th/9506014];

10454: ``Completeness Conditions for Boundary Operators in

10455: 2D Conformal Field Theory,''

10456: Phys.\ Lett.\ B {\bf 381},  (1996) 97

10457: [hep-th/9603097].

10458: 

10459: 

10460: \bibitem{bantay}

10461: P.~Bantay,

10462: ``The Frobenius-Schur indicator in conformal field theory'',

10463: Phys.\ Lett.\ B {\bf 394},  (1997) 89

10464: [hep-th/9610192].

10465: 

10466: \bibitem{HSS}

10467: L.~R.~Huiszoon, A.~N.~Schellekens and N.~Sousa,

10468: ``Klein bottles and simple currents,''

10469: Phys.\ Lett.\ B {\bf 470} (1999) 95

10470: [hep-th/9909114].

10471: 

10472: 

10473: \bibitem{FSHSS}

10474: J.~Fuchs, L.~R.~Huiszoon, A.~N.~Schellekens, C.~Schweigert and J.~Walcher,

10475: ``Boundaries, crosscaps and simple currents,''

10476: Phys.\ Lett.\ B {\bf 495} (2000) 427

10477: [hep-th/0007174].

10478: 

10479: \bibitem{SaAn}

10480: C.~Angelantonj and A.~Sagnotti,

10481: ``Open strings,''

10482: Phys.\ Rept.\  {\bf 371}, 1 (2002)

10483: [hep-th/0204089].

10484: 

10485: 

10486: \bibitem{FS1}

10487: J.~Fuchs and C.~Schweigert,

10488: ``Symmetry breaking boundaries. I: General theory,''

10489: Nucl.\ Phys.\ B {\bf 558}, 419 (1999)

10490: [hep-th/9902132].

10491: %%CITATION = HEP-TH 9902132;%%

10492: 

10493: 

10494: \bibitem{BH1}

10495: I.~Brunner and K.~Hori,

10496: ``Notes on orientifolds of rational conformal field theories,''

10497: [hep-th/0208141].

10498: 

10499: \bibitem{BlWi}

10500: R.~Blumenhagen and A.~Wisskirchen,

10501: ``Spectra of 4D, N = 1 type I string vacua on non-toroidal CY threefolds'',

10502: Phys.\ Lett.\ B {\bf 438} (1998) 52

10503: [hep-th/9806131].

10504: 

10505: \bibitem{ABPSS}

10506: C.~Angelantonj, M.~Bianchi, G.~Pradisi, A.~Sagnotti and Y.~S.~Stanev,

10507: ``Comments on Gepner models and type I vacua in string theory'',

10508: Phys.\ Lett.\ B {\bf 387} (1996) 743

10509: [hep-th/9607229].

10510: 

10511: \bibitem{BH2}

10512: I.~Brunner and K.~Hori,

10513: ``Orientifolds and mirror symmetry,''

10514: hep-th/0303135.

10515: 

10516: \bibitem{GovMaj}

10517: S.~Govindarajan and J.~Majumder,

10518: ``Orientifolds of type IIA strings on Calabi-Yau manifolds,''

10519: hep-th/0305108;

10520: ``Crosscaps in Gepner models and type IIA orientifolds,''

10521: hep-th/0306257.

10522: 

10523: \bibitem{HuiSch}

10524: L.~R.~Huiszoon and K.~Schalm,

10525: ``BPS orientifold planes from crosscap states in Calabi-Yau

10526: compactifications,''

10527: JHEP {\bf 0311} (2003) 019

10528: [hep-th/0306091].

10529: 

10530: 

10531: 

10532: 

10533: \bibitem{FS}

10534: J.~Fuchs and C.~Schweigert,

10535: ``Branes: From free fields to general backgrounds,''

10536: Nucl.\ Phys.\ B {\bf 530}, 99 (1998)

10537: [hep-th/9712257].

10538: %%CITATION = HEP-TH 9712257;%%

10539: 

10540: \bibitem{BS}

10541: I.~Brunner and V.~Schomerus,

10542: ``D-branes at singular curves of Calabi-Yau compactifications,''

10543: JHEP {\bf 0004}, 020 (2000)

10544: [hep-th/0001132].

10545: %%CITATION = HEP-TH 0001132;%%

10546: 

10547: \bibitem{FSW}

10548: J.~Fuchs, C.~Schweigert and J.~Walcher,

10549: ``Projections in string theory and boundary states for Gepner models,''

10550: Nucl.\ Phys.\ B {\bf 588}, 110 (2000)

10551: [hep-th/0003298].

10552: %%CITATION = HEP-TH 0003298;%%

10553: 

10554: \bibitem{FKLLSW}

10555: J.~Fuchs, P.~Kaste, W.~Lerche, C.~A.~Lutken, C.~Schweigert and J.~Walcher,

10556: ``Boundary fixed points, enhanced gauge symmetry and singular bundles on K3,''

10557: Nucl.\ Phys.\ B {\bf 598}, 57 (2001)

10558: [hep-th/0007145].

10559: %%CITATION = HEP-TH 0007145;%%

10560: 

10561: 

10562: \bibitem{GreenePlesser}

10563: B.~R.~Greene and M.~R.~Plesser,

10564: ``Duality In Calabi-Yau Moduli Space,''

10565: Nucl.\ Phys.\ B {\bf 338} (1990) 15.

10566: 

10567: 

10568: \bibitem{MMS}

10569: J.~M.~Maldacena, G.~W.~Moore and N.~Seiberg,

10570: ``Geometrical interpretation of D-branes in gauged WZW models,''

10571: JHEP {\bf 0107} (2001) 046

10572: [hep-th/0105038].

10573: 

10574: \bibitem{DiaDou}

10575: D.~E.~Diaconescu and M.~R.~Douglas,

10576: ``D-branes on stringy Calabi-Yau manifolds,''

10577: hep-th/0006224.

10578: 

10579: \bibitem{DiaRo}

10580: D.~E.~Diaconescu and C.~Romelsberger,

10581: ``D-branes and bundles on elliptic fibrations,''

10582: Nucl.\ Phys.\ B {\bf 574}, 245 (2000)

10583: [hep-th/9910172].

10584: %%CITATION = HEP-TH 9910172;%%

10585: 

10586: \bibitem{KLLW}

10587: P.~Kaste, W.~Lerche, C.~A.~Lutken and J.~Walcher,

10588: ``D-branes on K3-fibrations,''

10589: Nucl.\ Phys.\ B {\bf 582}, 203 (2000)

10590: [hep-th/9912147].

10591: %%CITATION = HEP-TH 9912147;%%

10592: 

10593: \bibitem{Scheidegger}

10594: E.~Scheidegger,

10595: ``D-branes on some one- and two-parameter Calabi-Yau hypersurfaces,''

10596: JHEP {\bf 0004}, 003 (2000)

10597: [hep-th/9912188].

10598: %%CITATION = HEP-TH 9912188;%%

10599: 

10600: 

10601: \bibitem{Mayr}

10602: P.~Mayr,

10603:  ``Phases of supersymmetric D-branes on Kaehler manifolds and the McKay

10604: %correspondence,''

10605: JHEP {\bf 0101}, 018 (2001)

10606: [hep-th/0010223].

10607: %%CITATION = HEP-TH 0010223;%%

10608: 

10609: 

10610: 

10611: \bibitem{Candelas}

10612: P.~Candelas, X.~De La Ossa, A.~Font, S.~Katz and D.~R.~Morrison,

10613: ``Mirror symmetry for two parameter models. I,''

10614: Nucl.\ Phys.\ B {\bf 416} (1994) 481

10615: [hep-th/9308083].

10616: 

10617: 

10618: \bibitem{Kennaway}

10619: K.~D.~Kennaway,

10620: ``A geometrical construction of

10621: Recknagel-Schomerus boundary states in  linear sigma models'',

10622: Nucl.\ Phys.\ B {\bf 647} (2002) 471

10623: [hep-th/0203266].

10624: 

10625: 

10626: \bibitem{HoriString}

10627: K. Hori, ``D-branes, Orientifolds of Calabi-Yau, And N=1 Supersymmetry'',

10628: talk at Strings 2003, Kyoto, July 10, 2003;\\

10629: http://www2.yukawa.kyoto-u.ac.jp/$\sim$str2003/speakerspro.html\#hori

10630: 

10631: 

10632: \bibitem{Aldazabal}

10633: G.~Aldazabal, E.~C.~Andres, M.~Leston and C.~Nunez,

10634: ``Type IIB orientifolds on Gepner points,''

10635: JHEP {\bf 0309} (2003) 067

10636: [hep-th/0307183].

10637: 

10638: \bibitem{Blumenhagen}

10639: R.~Blumenhagen,

10640: ``Supersymmetric orientifolds of Gepner models,''

10641: hep-th/0310244.

10642: 

10643: \bibitem{phases}

10644: E.~Witten,

10645: ``Phases of N = 2 theories in two dimensions,''

10646: Nucl.\ Phys.\ B {\bf 403} (1993) 159

10647: [hep-th/9301042].

10648: 

10649: 

10650: 

10651: \bibitem{Gepnerclassification}

10652: J.~Fuchs, A.~Klemm, C.~Scheich and M.~G.~Schmidt,

10653: ``Spectra And Symmetries Of Gepner Models Compared To Calabi-Yau 

10654: Compactifications,''

10655: Annals Phys.\  {\bf 204}, 1 (1990);

10656: %%CITATION = APNYA,204,1;%%

10657: M.~Lynker and R.~Schimmrigk,

10658: ``A-D-E Quantum Calabi-Yau Manifolds,''

10659: Nucl.\ Phys.\ B {\bf 339}, 121 (1990);

10660: %%CITATION = NUPHA,B339,121;%%

10661: The most complete list of Gepner models and Calabi-Yau

10662: manifolds is available online at 

10663: http://hep.itp.tuwien.ac.at/$\sim$kreuzer/CY/

10664: 

10665: \bibitem{MorrisonPlesser}

10666: D.~R.~Morrison and M.~Ronen Plesser,

10667: ``Summing the instantons: Quantum cohomology and mirror symmetry in toric

10668: varieties,''

10669: Nucl.\ Phys.\ B {\bf 440} (1995) 279

10670: [hep-th/9412236].

10671: 

10672: 

10673: 

10674: \bibitem{LVW}

10675: W.~Lerche, C.~Vafa and N.~P.~Warner,

10676: ``Chiral Rings In N=2 Superconformal Theories,''

10677: Nucl.\ Phys.\ B {\bf 324} (1989) 427.

10678: 

10679: 

10680: \bibitem{Strominger}

10681: A.~Strominger,

10682: ``Massless black holes and conifolds in string theory,''

10683: Nucl.\ Phys.\ B {\bf 451} (1995) 96

10684: [hep-th/9504090].

10685: 

10686: 

10687: \bibitem{VafaLGorb}

10688: C.~Vafa,

10689: ``String Vacua And Orbifoldized L-G Models,''

10690: Mod.\ Phys.\ Lett.\ A {\bf 4} (1989) 1169.

10691: 

10692: 

10693: 

10694: \bibitem{HV}

10695: K.~Hori and C.~Vafa,

10696: ``Mirror symmetry,''

10697: hep-th/0002222.

10698: 

10699: 

10700: \bibitem{ASI} 

10701: A.~Sagnotti, ``Open strings and their symmetry groups'',

10702: ROM2F-87-25, {\it Talk presented at the Carg\`ese summer

10703: institute on non-perturbative methods in fields theory,

10704: Carg\`ese, France, July 1987}, [hep-th/0208020].

10705: 

10706: \bibitem{BSII}

10707: M.~Bianchi and A.~Sagnotti,

10708: ``On the systematics of open string theories,''

10709: Phys.\ Lett.\ B {\bf 247}, (1990) 517.

10710: 

10711: \bibitem{ASII}

10712: M.~Bianchi and A.~Sagnotti,

10713: ``Twist symmetry and open string Wilson lines'',

10714: Nucl.\ Phys.\ B {\bf 361}, 519 (1991).

10715: 

10716: \bibitem{HoI}

10717: P.~Horava,

10718: ``Strings on world sheet orbifolds'',

10719: Nucl.\ Phys.\ B {\bf 327}, 461 (1989).

10720: 

10721: \bibitem{HoII}

10722: P.~Horava,

10723: ``Background duality of open string models'',

10724: Phys.\ Lett.\ B {\bf 231}, (1989) 251.

10725: 

10726: 

10727: 

10728: \bibitem{RRW}

10729: R.~Roiban, C.~Romelsberger and J.~Walcher,

10730: ``Discrete torsion in singular G(2)-manifolds and real LG,''

10731: Adv.\ Theor.\ Math.\ Phys.\ 6 (2003) 207-278

10732: [hep-th/0203272].

10733: 

10734: 

10735: \bibitem{AGM}

10736: P.~S.~Aspinwall, B.~R.~Greene and D.~R.~Morrison,

10737: ``Calabi-Yau moduli space, mirror manifolds and spacetime topology  change in

10738: string theory,''

10739: Nucl.\ Phys.\ B {\bf 416} (1994) 414

10740: [hep-th/9309097].

10741: 

10742: 

10743: \bibitem{PolCai}

10744: J.~Polchinski and Y.~Cai,

10745: ``Consistency Of Open Superstring Theories,''

10746: Nucl.\ Phys.\ B {\bf 296} (1988) 91.

10747: %%CITATION = NUPHA,B296,91;%%

10748: 

10749: \bibitem{Callan}

10750: C.~G.~Callan, C.~Lovelace, C.~R.~Nappi and S.~A.~Yost,

10751: ``Loop Corrections To Superstring Equations Of Motion,''

10752: Nucl.\ Phys.\ B {\bf 308} (1988) 221.

10753: 

10754: 

10755: 

10756: \bibitem{discretetorsion}

10757: M.~R.~Douglas,

10758: ``D-branes and discrete torsion,''

10759: hep-th/9807235.

10760: %%CITATION = HEP-TH 9807235;%%

10761: 

10762: 

10763: \bibitem{HIV}

10764: K.~Hori, A.~Iqbal and C.~Vafa,

10765: ``D-branes and mirror symmetry,''

10766: hep-th/0005247.

10767: 

10768: \bibitem{HKKPTVVZ}

10769: K.~Hori, S.~Katz, A.~Klemm, R.~Pandharipande,

10770: R.~Thomas,

10771: C.~Vafa, R.~Vakil and E.~Zaslow,

10772: ``Mirror Symmetry'',

10773: (AMS/CMI, 2003), Chapter 39.

10774: 

10775: \bibitem{GimPol}

10776: E.~G.~Gimon and J.~Polchinski,

10777: ``Consistency Conditions for Orientifolds and D-Manifolds,''

10778: Phys.\ Rev.\ D {\bf 54} (1996) 1667

10779: [hep-th/9601038].

10780: 

10781: 

10782: \bibitem{recknagel}

10783: A.~Recknagel,

10784: ``Permutation branes,''

10785: JHEP {\bf 0304}, 041 (2003)

10786: [hep-th/0208119].

10787: %%CITATION = HEP-TH 0208119;%%

10788: 

10789: 

10790: 

10791: 

10792: 

10793: 

10794: 

10795: 

10796: 

10797: \bibitem{GS}

10798: M.~B.~Green and J.~H.~Schwarz,

10799: ``Anomaly Cancellation In Supersymmetric D=10 Gauge Theory And Superstring

10800: Theory,''

10801: Phys.\ Lett.\ B {\bf 149} (1984) 117.

10802: 

10803: 

10804: \bibitem{GSW}

10805: M.~B.~Green, J.~H.~Schwarz and E.~Witten,

10806: ``Superstring Theory. Vol. 2: Loop Amplitudes, Anomalies And Phenomenology,''

10807: (Cambridge University Press, 1987).

10808: 

10809: \bibitem{DSW}

10810: M.~Dine, N.~Seiberg and E.~Witten,

10811: ``Fayet-Iliopoulos Terms In String Theory,''

10812: Nucl.\ Phys.\ B {\bf 289} (1987) 589.

10813: 

10814: 

10815: 

10816: \bibitem{DM}

10817: M.~R.~Douglas and G.~W.~Moore,

10818: ``D-branes, Quivers, and ALE Instantons,''

10819: hep-th/9603167.

10820: 

10821: 

10822: 

10823: \bibitem{Alda1}

10824: G.~Aldazabal, S.~Franco, L.~E.~Ibanez, R.~Rabadan and A.~M.~Uranga,

10825: ``D = 4 chiral string compactifications from intersecting branes,''

10826: J.\ Math.\ Phys.\  {\bf 42} (2001) 3103

10827: [hep-th/0011073].

10828: 

10829: 

10830: \bibitem{CSU}

10831: M.~Cvetic, G.~Shiu and A.~M.~Uranga,

10832: ``Chiral four-dimensional N = 1 supersymmetric type IIA orientifolds from

10833: intersecting D6-branes,''

10834: Nucl.\ Phys.\ B {\bf 615} (2001) 3

10835: [hep-th/0107166].

10836: 

10837: 

10838: \bibitem{IRU1}

10839: L.~E.~Ibanez, R.~Rabadan and A.~M.~Uranga,

10840: ``Anomalous U(1)'s in type I and type IIB D = 4, N = 1 string vacua,''

10841: Nucl.\ Phys.\ B {\bf 542} (1999) 112

10842: [hep-th/9808139].

10843: 

10844: 

10845: 

10846: \bibitem{KM}

10847: S.~Kachru and J.~McGreevy,

10848: ``Supersymmetric three-cycles and (super)symmetry breaking,''

10849: Phys.\ Rev.\ D {\bf 61} (2000) 026001

10850: [hep-th/9908135].

10851: 

10852: 

10853: \bibitem{DFRI}

10854: M.~R.~Douglas, B.~Fiol and C.~R\"omelsberger,

10855: ``Stability and BPS branes,''

10856: [hep-th/0002037].

10857: 

10858: \bibitem{DFRII}

10859: M.~R.~Douglas, B.~Fiol and C.~Romelsberger,

10860: ``The spectrum of BPS branes on a noncompact Calabi-Yau,''

10861: [hep-th/0003263].

10862: 

10863: \bibitem{Dougcat}

10864: M.~R.~Douglas,

10865: ``D-branes, categories and N = 1 supersymmetry,''

10866: J.\ Math.\ Phys.\  {\bf 42}, 2818 (2001)

10867: [hep-th/0011017].

10868: 

10869: 

10870: 

10871: 

10872: \bibitem{poppitz}

10873: E.~Poppitz,

10874: ``On the one loop Fayet-Iliopoulos term in chiral four dimensional type I

10875: orbifolds,''

10876: Nucl.\ Phys.\ B {\bf 542} (1999) 31

10877: [hep-th/9810010].

10878: 

10879: \bibitem{Cvetic}

10880: M.~Cvetic, L.~L.~Everett, P.~Langacker and J.~Wang,

10881: ``Blowing-up the four-dimensional Z(3) orientifold,''

10882: JHEP {\bf 9904} (1999) 020

10883: [hep-th/9903051].

10884: 

10885: \bibitem{Lalak}

10886: Z.~Lalak, S.~Lavignac and H.~P.~Nilles,

10887: ``String dualities in the presence of anomalous U(1) symmetries,''

10888: Nucl.\ Phys.\ B {\bf 559} (1999) 48

10889: [hep-th/9903160].

10890: 

10891: 

10892: \bibitem{IRU2}

10893: L.~E.~Ibanez, R.~Rabadan and A.~M.~Uranga,

10894: ``Sigma-model anomalies in compact D = 4, N = 1 type IIB orientifolds and

10895: Fayet-Iliopoulos terms,''

10896: Nucl.\ Phys.\ B {\bf 576} (2000) 285

10897: [hep-th/9905098].

10898: 

10899: \bibitem{IbaQue}

10900: L.~E.~Ibanez and F.~Quevedo,

10901: ``Anomalous U(1)'s and proton stability in brane models,''

10902: JHEP {\bf 9910} (1999) 001

10903: [hep-ph/9908305].

10904: 

10905: 

10906: \bibitem{WHWSH}

10907: E.~Witten,

10908: ``New *gauge* theories in six dimensions,''

10909: JHEP {\bf 9801} (1998) 001

10910: [Adv.\ Theor.\ Math.\ Phys.\  {\bf 2} (1998) 61]

10911: [hep-th/9710065];

10912: 

10913: K.~Hori,

10914: ``Consistency condition for fivebrane in M-theory on R**5/Z(2) orbifold,''

10915: Nucl.\ Phys.\ B {\bf 539} (1999) 35

10916: [hep-th/9805141];

10917: 

10918: E.~G.~Gimon,

10919: ``On the M-theory interpretation of orientifold planes,''

10920: hep-th/9806226;

10921: 

10922: E.~Witten,

10923: ``Baryons and branes in anti de Sitter space,''

10924: JHEP {\bf 9807} (1998) 006

10925: [hep-th/9805112];

10926: 

10927: S.~Sethi,

10928: ``A relation between N = 8 gauge theories in three dimensions,''

10929: JHEP {\bf 9811} (1998) 003

10930: [hep-th/9809162];

10931: 

10932: A.~Hanany and B.~Kol,

10933: ``On orientifolds, discrete torsion, branes and M theory,''

10934: JHEP {\bf 0006} (2000) 013

10935: [hep-th/0003025];

10936: 

10937: 

10938: A.~Hanany, B.~Kol and A.~Rajaraman,

10939: ``Orientifold points in M theory,''

10940: JHEP {\bf 9910} (1999) 027

10941: [hep-th/9909028].

10942: 

10943: 

10944: 

10945: 

10946: 

10947: \bibitem{WK}

10948: E.~Witten,

10949: ``D-branes and K-theory,''

10950: JHEP {\bf 9812} (1998) 019

10951: [hep-th/9810188].

10952: 

10953: 

10954: \bibitem{HT}

10955: K.~Hori,

10956: ``D-branes, T-duality, and index theory,''

10957: Adv.\ Theor.\ Math.\ Phys.\  {\bf 3} (1999) 281

10958: [hep-th/9902102].

10959: %%CITATION = HEP-TH 9902102;%%

10960: 

10961: 

10962: \bibitem{BGH}

10963: O.~Bergman, E.~G.~Gimon and P.~Horava,

10964: ``Brane transfer operations and T-duality of non-BPS states,''

10965: JHEP {\bf 9904} (1999) 010

10966: [hep-th/9902160].

10967: 

10968: 

10969: \bibitem{Gukov}

10970: S.~Gukov,

10971: ``K-theory, reality, and orientifolds,''

10972: Commun.\ Math.\ Phys.\  {\bf 210} (2000) 621

10973: [hep-th/9901042].

10974: 

10975: \bibitem{Stefanski}

10976: B.~J.~Stefanski,

10977: ``D-branes, orientifolds and K-theory,''

10978: hep-th/0206174;

10979: 

10980: V.~Braun and B.~J.~Stefanski,

10981: ``Orientifolds and K-theory,''

10982: hep-th/0206158.

10983: 

10984: \bibitem{Knorrer}

10985: Horst~Kn\"orrer,

10986: ``Cohen-Macaulay modules on hypersurface singularities I''

10987: Invent.\ Math.\ {\bf 88} (1987) 153-164.

10988: 

10989: \bibitem{BD}

10990: I.~Brunner and J.~Distler,

10991: ``Torsion D-branes in nongeometrical phases,''

10992: Adv.\ Theor.\ Math.\ Phys.\  {\bf 5}, 265 (2002)

10993: [hep-th/0102018].

10994: %%CITATION = HEP-TH 0102018;%%

10995: 

10996: 

10997: \bibitem{statistics}

10998: M.~R.~Douglas,

10999: ``The statistics of string / M theory vacua,''

11000: JHEP {\bf 0305}, 046 (2003)

11001: [hep-th/0303194];

11002: 

11003: S.~Ashok and M.~R.~Douglas,

11004: ``Counting flux vacua,''

11005: hep-th/0307049.

11006: 

11007: \end{thebibliography}

11008: 

11009: 

11010: \end{document}

11011: