1: % On Orientifolds, Discrete Torsion, Branes and M Theory

2: % Amihay Hanany, Barak Kol

3: 

4: \documentclass[letterpaper]{JHEP}

5: % hep-th requires letterpaper, while JHEP uses DIN A4.

6: \usepackage{epsfig}

7: %\usepackage[active]{srcltx}

8: 

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

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

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

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

13: 

14: \newcommand{\IZ}{{\bf Z}} \newcommand{\IR}{{\bf R}}

15: \newcommand{\IC}{\bf C} \newcommand{\IT}{\bf T}

16: \newcommand{\IQ}{\bf Q} \newcommand{\IP}{\bf P}

17: \newcommand{\IS}{\bf S} \newcommand{\IRP}{\bf RP}

18: \newcommand{\qm}{{\bf ?}}

19: \newcommand{\IZZ}{{\bf Z}_2}

20: \newcommand{\OMtp}{$OM2^+~$}

21: \newcommand{\OMtm}{$OM2^-~$}

22: \newcommand{\SLtz}{$SL(2,{\bf Z})~$}

23: \newcommand{\Ltb}{$L_{IIB}~$}

24: \newcommand{\Ttm}{${\bf T}^2_M~$}

25: \newcommand{\pq}{$(p,q)~$}

26: \newcommand{\wtilde}{\widetilde}

27: \newcommand{\Othm}{$O3^-~$}

28: \newcommand{\Othp}{$O3^+~$}

29: \newcommand{\Othmt}{$\widetilde{O3^-}~$}

30: \newcommand{\Othpt}{$\widetilde{O3^+}~$}

31: 

32: \newcommand{\wwwspires}{http://www.slac.stanford.edu/spires/find/hep/www}

33: \def\func#1{\mathop{\rm #1}\nolimits}

34: 

35: 

36: \preprint{MIT-CTP-2946\\ TAUP-2609-2000\\ NSF-ITP-00-07\\ {\tt

37: hep-th/0003025}}

38: 

39: \title{On Orientifolds, Discrete Torsion, Branes and  M Theory}

40: 

41: \author{Amihay Hanany

42: \\

43: Center for Theoretical Physics,

44: \\ Massachusetts Institute of Technology\\ Cambridge MA 02139\\

45: \email{hanany@mit.edu}

46: }

47: \author{Barak Kol

48: \\

49: School of Physics and Astronomy\\ Tel Aviv University\\ Ramat Aviv 69978,

50: Israel\\

51: \email{barak1k@post.tau.ac.il}

52: }

53: 

54: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

55: \abstract{ We find some lifts to M theory of orientifold and

56: orbifold planes including the O1, O3 and O5 planes of Type

57: IIB and their transformations under \SLtz. The possible

58: discrete torsion variants (or K theory classes) are

59: explored, and are interpreted as arising from brane

60: intersections with planes. We find new variants of the O0

61: and of an orbifold line (OF1) and determine their tensions

62: in some cases. A systematic review of orientifolds, M

63: orientifolds, and known M lifts, with some new

64: clarifications is included together with a discussion of

65: the role of T duality.}

66: \keywords{M-theory, p-branes}

67: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

68: 

69: \begin{document}

70: 

71: \section{Introduction}

72: 

73: Orientifold planes are objects in string theory which are

74: defined perturbatively by gauging a discrete symmetry

75: which involves reversing the sign of coordinates

76: transverse to the plane while changing the orientation of

77: the string \cite{Sagnotti:1987tw,Horava:1989vt,GP}. Before

78: we delve into the details of their construction and their

79: dynamics let us mention some motivation and interest in

80: their study.

81: 

82: Orientifold planes are of interest in many aspects of

83: string theory. They turn out to be useful in studying

84: disconnected components in moduli space of various string

85: compactifications. They are crucial in the brane

86: construction of gauge theories with $Sp/SO$ gauge

87: groups.\footnote{See \cite{Clifford} for a recent

88: interesting application.} In addition they give a simple

89: realization of matter in symmetric and antisymmetric

90: second rank representations in $SU(N)$ gauge theories.

91: 

92: As will be discussed in detail in the following sections,

93: orientifold planes turn out to be characterized by

94: discrete fluxes. Some of these fluxes originate from NS

95: fields (See \cite{Zurab} for a recent discussion) and some

96: other fluxes come from RR fields. While the first (NS) are

97: visible from the usual perturbative description of

98: orientifolds, the latter (RR) cannot be treated in a

99: perturbative formulation of the Type II string since, as

100: of now, there is no formalism which includes RR

101: backgrounds in string theory. For this reason we need to

102: rely on other methods in order to study such objects, and

103: once such a formalism is available, these questions should

104: be revisited. The best analysis which is available at the

105: moment will include essentially two tools

106: - various dualities, and discrete fluxes represented by

107: intersections of branes with the orientifold plane.

108: 

109: Here we study two subjects related to orientifolds - their

110: lift to M theory (See a recent related discussion in

111: \cite{Angel}), and the different variants they appear in.

112: We begin with a review of the relevant features of

113: orientifolds in section \ref{review}. We describe the 4

114: types of $Op$ planes which are known to exist for $p \le

115: 5$. We explain how these 4 types are classified by a pair

116: of $\IZZ$ parameters, one perturbative coming from the

117: $B_{NS}$ form and the other from one of the RR forms.

118: These parameters are ``discrete torsion'' parameters which

119: arise whenever the field strength of a p-form potential

120: has a corresponding non trivial discrete cohomology in

121: space time \cite{Vafa:1986wx,Vafa:1995rv} (see also recent

122: related work

123: \cite{Dijkgraaf:1999za,Gomis:2000ej,Klein:2000tf,Angelantonj:1999xf}).

124: Equivalently we show that these variants arise as brane

125: configurations in the spirit of \cite{HananyWitten} (see

126: recent related work \cite{Bergman:1999na}) from

127: intersections of an O plane with a brane which divides it

128: into two parts. We search for all the possible discrete

129: torsion variants, and we find that for $p \le 1$ there are

130: some additional ones. Later in that section, in

131: \ref{duality}, we explain how T duality relates a wrapped

132: orientifold with a pair of lower dimensional orientifolds

133: and the mapping between the various discrete torsion

134: variants.

135: 

136: In section \ref{revM} we systematically review the $\IZZ$

137: orbifolds of M theory. There are only a few of these and

138: upon compactification they give rise to the various $\IZZ$

139: orientifolds and orbifolds of string theory. We review the

140: M lifts of the O0, O2, O4

141: \cite{DasguptaMukhi,HananyKolRajaraman,

142: Sethi:1998zk,BerkoozKapustin,Witten_OM, Hori,Gimon:1998be}

143: in terms of 11d O planes - the OM1, OM2 and OM5. In the

144: case of the $O2$ we give a new M theory explanation for

145: the difference between the two variants $O2^+$ and

146: $\wtilde{O2}^+$.

147: 

148: In section \ref{Mlift} we describe the M lift of the O1,

149: O3 and O5 in terms of the same OM planes. In all cases we

150: pay special attention to the \SLtz symmetry

151: \cite{SchwarzSL2Z} (see also related work \cite{Kol5d}).

152: For the O3 we show how the M lift realizes the \SLtz

153: geometrically, as usual. For the O1 and O5 the S

154: transformation of \SLtz  transforms the orientifold planes

155: into orbifolds, which originate in the same OM planes.

156: 

157: In section \ref{orb} we discuss in detail the

158: interrelations between $Op$ planes and orbifolds for

159: $p=1,5$. We explore the possible discrete torsion variants

160: and discuss new ones for the orbifold and orientifold

161: lines and for the orientifold point. We compare the

162: discrete torsion classification with a recent claim on a K

163: theory classification \cite{MooreWitten} (see

164: \cite{Minasian:1997mm,Witten:1998cd,Horava:1999jy,

165: Bergman:1999ta,Gukov:1999yn,Hori:1999me}

166: for earlier and related work) and find groups of the same

167: order but with a different group structure. We are able to

168: find the tension of some of the variants, and it remains

169: to do so for the other cases.

170: 

171: Finally we discuss some miscellaneous applications in

172: section \ref{misc}. We discuss the dyon spectrum for 4d

173: ${\cal N}=4$ with $SO/Sp$ gauge groups, where the $Sp$

174: theory has two versions. Then the discussion on the

175: monopole spectrum of such a theory is generalized to a

176: field theory in an arbitrary dimension. Finally, we

177: discuss the problems associated with a \pq web of

178: orientifolds.

179: 

180: Let us mention some open questions

181: \begin{itemize}

182: \item We show how an intersection of an orientifold, or an OM plane, with a

183: brane causes a tension jump, and so a {\it fractional} charge is

184: deposited on the brane. This jump should be understood and

185: probably required by the worldvolume theory on the brane. In

186: particular consistency with Dirac quantization is required.

187: \item We find the tension of some of the discrete torsion variants of the

188:  orbifold lines, and it would be interesting to know all of them.

189: \end{itemize}

190: 

191: Preliminary results of this

192: \href{http://online.itp.ucsb.edu/online/susy99/hanany/}{work} were

193: presented in Santa Barbara and can be found in the following link:

194: \href{http://online.itp.ucsb.edu/online/susy99/hanany/}

195: {http://online.itp.ucsb.edu/online/susy99/hanany/}.

196: 

197: \vspace{2cm}

198: \section{\protect\bigskip Introduction to Orientifolds}

199: \label{review}

200: 

201: An orientifold plane in $p+1$ dimensions is defined as

202: Type II string theory on $R^{p,1}\times

203: R^{9-p}/(I\cdot\Omega\cdot J)$ where $I$ is the inversion

204: of all coordinates in the transverse space $R^{9-p}$,

205: $\Omega$ is the orientation reversal on the world sheet of

206: the fundamental string and $J$ is the identity operator

207: for $p=0,1  ~{\rm mod}~4$ and $(-1 )^{F_L}$, the left

208: moving spacetime fermion number operator, for $p=2,3 ~{\rm

209: mod}~4$ \cite{DabholkarPark}. We will denote an

210: orientifold p-plane as an ``$Op$ plane''.

211: 

212: By definition the orientifold acts on $B_{NS}$ (with

213: components parallel to the $Op$ plane) with a sign

214: reversal. The action on the other NS fields is trivial, so

215: let us specify the action on all of the RR forms. The $O9$

216: which projects Type IIB to Type I keeps invariant the

217: left-right symmetric part of the spectrum, which in the RR

218: sector is the 2-form $B_{RR}$. By T dualizing we arrive at

219: the following rule for the action of an $Op$ plane on the

220: RR forms $C_{p'}$

221: \bea C_{p'} \to C_{p'} ~~p'=p+1~{\rm mod}~4  \nonumber \\

222:  C_{p'} \to - C_{p'} ~~p'=p+3~{\rm mod}~4 .

223:  \eea

224: Note that the above sign comes in addition to a component

225: dependent sign which comes from the tensor transformation

226: rules. Forms which get reversed by the action of the group

227: are termed ``twisted'', and this notation should not be

228: confused with forms from a twisted sector.

229: 

230: The transverse space to an $Op$ plane is an $\IRP^{8-p}$.

231: This space has some (discrete) torsion cohomologies which

232: are summarized in appendix \ref{cohomology}. Whenever a

233: space has discrete torsion, forms of an appropriate rank

234: are topologically classified by it. Namely

235: \be

236: [G_{p+1}] \in H^{p+1}(X,\IZ) \ee where $H^{p+1}(X,Z)$ is

237: the integral cohomology of the space $X$ and $G_{p+1}$ is

238: the RR field strength $p+1$-form. Twisted forms require

239: the use of twisted cohomologies, $\tilde{H}^{p+1}$. In the

240: case of orientifolds we will see that discrete torsion

241: allows for the orientifold to appear in several variants.

242: 

243: For a general orientifolds ($p \le 5$) there are at least

244: four types of orientifold planes which are distinguished

245: by two $\IZZ$ charges. One $\IZZ$ charge comes from the NS

246: sector and the other comes from the RR sector. For this

247: reason, the former charge is seen in perturbative string

248: theory while the latter is not. The perturbative

249: distinction was first discussed in modern language in

250: \cite{GP}. Later, the other $\IZZ$ charge was discovered

251: in \cite{Uranga,WittenAdSBaryons}\footnote{In

252: \cite{Witten6d} there is an earlier discussion of the four

253: variants of 5 planes in an M theory language which is

254: directly related to discrete torsions.} for the case $p=3$

255: essentially by using field theory intuition which is based

256: on Montonen Olive Duality or Type IIB self duality. The

257: case $p=4$ was studied in detail by

258: \cite{Hori,Gimon:1998be}.

259: 

260: For an $Op$ plane, the two $\IZZ$ charges are given by

261: $b$, which is the class $[dB_{NS}]=[H_{NS}] \in

262: \tilde{H}^3(\IRP^{8-p})$ and by $c$, which is the class of an

263:  RR form $[dC_{5-p}]=[G_{6-p}] \in \IZZ$

264: (the relevant cohomology, either twisted or untwisted, is

265: found to be $\IZZ$). In the next subsection we will give

266: these a physical interpretation via brane-orientifold

267: intersections. The above classes can be defined for $p \le

268: 5$, while for $6

269: \le p \le 8$ one can still define $b$ through the

270: perturbative action on open strings.  In general the

271: existence of a discrete torsion implies the existence of

272: variants, though variants could possibly exist without it,

273: as in the case of the $b$ variant. The absence of a $c$

274: charge for $p=6$ suggests that, contrary to naive

275: expectations, one can not have an $\wtilde{O6}^-$ plane,

276: i.e. half a D6 brane stuck on an $O6^-$ plane. That turns

277: out to be the case as implied by the results of

278: \cite{Dijk}.

279: 

280: One can systematically check for any $p$ which additional

281: forms can get discrete torsions. For low $p$ some extra

282: variants exist (see section \ref{orb}): for $p

283: \le 1$ one can define an additional charge $c'$

284: by $[dC_{1-p}]=[G_{2-p}]

285: \in \IZZ$, and for $p=1$ one has $*H_{NS} \in

286: \tilde{H}^7(\IRP^7)=\IZZ$. In addition there are two other discrete

287: torsion whose variants we ignore - $H^0(\IRP^n)=\IZ$ leads

288: to a integral discrete variant for $p=2 ~{\rm mod}~ 4$

289: which we interpret as describing the possible massive

290: theories of type IIA (this interpretation is based on the

291: brane intersection picture, where this torsion is seen to

292: be due to an intersection with a D8 plane). The other type

293: which we ignore is $H^{8-p}=\IZ$ for odd $p$ and

294: $\tilde{H}^{8-p}=\IZ$ for even $p$, which simply means

295: that one can add to an $Op$ plane any integer number of

296: $Dp$ branes.

297: 

298: The orientifold planes will be denoted according to their

299: $\IZZ$ charges. A trivial $b$ charge will be denoted by a

300: $^-$ superscript and a non-trivial $b$ will be denoted by

301: $^+$. A non-trivial $c$ charge will be denoted by adding a

302: $\widetilde{}

303: \,$.

304:  The charges of these planes are $-2^{p-5}$

305: for the $Op^-$ plane, $+2^{p-5}$ for both the $Op^+$ and the

306: $\widetilde{Op^+}$ and ${\frac{1}{2}}-2^{p-5}$ for the

307: $\widetilde{Op^-}$ plane. The tensions of these objects are

308: measured in units of the Dp brane tension and are identical to

309: their charge. Note a change of notation for the various

310: orientifold planes from the papers \cite{HZ,HananyKolRajaraman}.

311: The orientifold planes are denoted there $Op^-, Op^+,

312: \widehat{Op},$ and $\widetilde{Op}$, respectively. The new

313: notation is based on the two $\IZZ$ charges and makes it more

314: simple to work with. $n$ physical $Dp$ branes stacked upon an $Op$

315: plane leads to a gauge group $SO(2n), Sp(n), SO(2n+1)$ and $Sp(n)$

316: for $Op^-, Op^+, \widetilde{Op^-}, \widetilde{Op^+}$,

317: respectively. The two $Sp(n)$ theories differ, in some cases, by a

318: theta angle and in their monopole spectrum. This will be discussed

319: in detail in sections \ref{spectrum} and \ref{BPS}.

320: 

321: \vspace{2cm}

322: \subsection{Brane Realization of Discrete Torsion}

323: 

324: \EPSFIGURE[h]{basic}{A half NS brane and an Orientifold plane. The

325: solid line represents a half NS brane and the dashed line

326: represents an $Op$ plane. The type of the $Op$ plane

327: changes from $Op^-$ to $Op^+$ as it crosses the half NS

328: brane. \label{basic}}

329: 

330: In this section we will use a basic fact about orientifold planes

331: and NS five branes and will develop a set of relations between

332: intersecting branes and orientifold planes. The basic observation,

333: made in \cite{EJS}, can be described in figure \ref{basic}.

334: Consider a NS five brane which spans the 012345 coordinates and

335: intersecting an $Op$ plane which spans the $0,\ldots, p-1$, and 6

336: coordinates, $p\leq6$. It splits the orientifold plane into two

337: different parts, one to its left and one to its right and the

338: orientifold plane changes its type as in figure \ref{basic}.

339: 

340: The NS brane in this configuration has the special property that

341: it is reflected onto itself by the orientifold. In a compact

342: configuration such a brane would have half of the charge of a

343: brane which does not intersect the orientifold and so we call it a

344: ${\frac{1}{2}}$ NS brane. Note that a 1/2 brane which extends

345: outside of the orientifold has the same charge density as a unit

346: brane, only it covers half of the volume due to the projection. On

347: the other hand, a half brane inside the orientifold has half the

348: charge density. A 1/2 brane cannot move in directions transverse

349: to the orientifold plane, because only integral branes can move

350: away off the orientifold. In the following brane configurations

351: one should bear in mind that all branes which intersect with an

352: orientifold plane are 1/2 branes.

353: 

354: \EPSFIGURE[ht]{smooth}{A physical NS brane (and its image)

355:  near an orientifold plane. In the left

356: figure (a), the NS brane and its image are away from the

357: orientifold plane. As they move and meet along the $Op$

358: plane they can split into a pair of half NS branes and

359: form a new type of orientifold plane (b). The $Op$ types

360: are denoted by $+$ and $-$ signs.

361: \label{smooth}}

362: 

363: Let us emphasize this point in more detail. Consider a

364: configuration in which a physical NS brane is located far

365: away from the orientifold, as in figure \ref{smooth}a. Let

366: this brane move slowly towards the orientifold plane. It

367: moves together with its image under the orientifold

368: projection. As the two images meet on the orientifold

369: plane, they can split along it as in figure \ref{smooth}b.

370: At this point the rule in figure \ref{basic} can be used

371: and a new type of orientifold plane is generated in

372: between the two half NS branes.

373: 

374: We can look at it from a different point of view starting

375: from figure \ref{smooth}b and moving to figure

376: \ref{smooth}a. As observed in \cite{EJS}, the type of the

377: orientifold plane changes as one crosses the

378: ${\frac{1}{2}}$ NS brane. This is determined by the $b$

379: charge of the orientifold which changes as one crosses the

380: ${\frac{1}{2}}$ NS brane. On the other hand, if we add an

381: additional ${\frac{1}{2}}$ NS brane, the type of the

382: orientifold changes back to its original value, as in

383: figure \ref{smooth}b. In this case we can have a dynamical

384: process in which the two half NS branes combine together

385: along the 6 coordinate and leave the orientifold along the

386: 789 directions as in figure \ref{smooth}a. This is the

387: inverse process to the one described in the last

388: paragraph.

389: 

390: One can relate such configurations to the value of the two form

391: flux $b$ by the following reasoning. The combined configuration of

392: a $Dp$ brane and half an NS brane is located at a point in 789

393: directions. There is a corresponding $\IRP^{2}$ which surrounds

394: the configuration. The field which couples magnetically to the NS

395: brane is the 2-form NS field. Consequently, the integral of the

396: two form over the $\IRP^{2}$ measures the number of 1/2 NS branes,

397: mod 2, which are located within the $\IRP^{2}$ \footnote{A

398: discussion in a similar spirit can be found in \cite{Tsabar}.}.

399: \begin{equation}

400: \exp\left(i\int_{\IRP^{2}}B_{NS}\right)=(-)^{\#\frac{1}{2} \rm{NS

401: ~branes}}.

402: \end{equation}

403: This holonomy (``Wilson loop'') $\in H^p(X,U(1))$ of a

404: general p-form potential has the same discrete torsion as

405: the field strength class $\in H^{p+1}(X,Z)$ which was

406: mentioned before. This is seen via the exact sequences $0

407:  \to \IZ \to \IR \to U(1) \to 0$ and

408: \bea

409: \dots \to& H^{p}(X,\IR) \to& H^{p}(X,U(1)) \to \nonumber \\

410: \to H^{p+1}(X,\IZ) \to& H^{p+1}(X,\IR) \to&  \dots

411: \eea

412: where the cohomologies with real coefficients do not have

413: torsion parts.

414: 

415: We will assume that the configuration of figure

416: \ref{basic} exists and will apply some S and T dualities

417: to see what other results can be obtained from this basic

418: configuration. First let us apply S duality to this

419: configuration in the case $p=3$ (see also \cite{Mukhi3}).

420: For this we need to know what are the S transformations of

421: the various orientifold planes. This is easily done by

422: looking at their $\IZZ$ charges which are the NS 3-form

423: and the RR 3-form field strengths. Let us recall that

424: $O3^-$ has charge $(b,c)=(0,0)$ and the $\widetilde{O3^+}$

425: has charge (1,1). These orientifolds are self-dual under S

426: duality. $O3^+$ has (1,0) and transforms under S duality

427: to (0,1) which is the $\widetilde{O3^-}$.

428: 

429: Equipped with this data let us make an S duality on the

430: configuration of figure \ref{basic}. One gets that an

431: $O3^-$ transforms to an $\widetilde{O3^-}$ while crossing

432: a half D5 brane. By crossing another half NS brane the $b$

433: charge of the orientifold jumps and one gets

434: $\widetilde{O3^+}$. This is expected if one assumes that

435: the configuration of half D5 and half NS brane are self

436: dual under S duality and that the order of the D5 and NS

437: brane is not important. The summary is that when crossing

438: a half NS brane the $b$ charge changes while when crossing

439: a half D5 brane the $c$ charge of the orientifold changes.

440: $b$ measures the number of half NS branes enclosed by an

441: $\IRP^2$ surrounding the intersection of O3 plane and NS

442: brane while $c$ measures the number of half D5 branes

443: enclosed by an $\IRP^2$ surrounding the intersection of O3

444: plane and D5 brane.

445: 

446: Next we can perform T duality on these results. Consider

447: dualizing this system in a supersymmetric fashion to a

448: system of a NS brane along 012345 and a D6 brane along

449: 0123789. The orientifold plane dualizes to an O4 plane

450: along 01236. The two $\IZZ$ charges are measured by $b$

451: and by $c$ which is a Wilson line of the RR one form of

452: Type IIA. T duality then implies that when an O4 plane

453: crosses a half NS brane its $b$ charge changes while when

454: crossing a half D6 brane its $c$ charge changes (A Wilson

455: line $c$ can be associated to our configuration by

456: repeating the arguments for the O3 plane). The system is

457: located at a point in the 45 directions. The object which

458: couples magnetically to the D6 brane is the RR 1-form of

459: Type IIA. Correspondingly $c$ measures the number of half

460: D6 branes, mod 2, trapped inside the $\IRP^1$.

461: 

462: Let us summarize the situation after further applying T

463: duality to other directions. An $Op$ plane along

464: 012,\ldots,$p-1$ and 6, a half NS brane along 012345 and a

465: half D(p+2) brane along 012,\ldots,$p-1$, 789. {\it When

466: crossing a half NS brane the $Op$ plane changes its $b$

467: charge and when crossing a half D(p+2) brane the $Op$

468: plane changes its $c$ charge} which is measured by a

469: `Wilson loop integral' of the RR $(5-p)$ form potential.

470: This configuration exists for any $p$ which is 5 or less.

471: $b$ measures the number of half NS branes in an $\IRP^2$

472: enclosing the configuration while $c$ measures the number

473: of half $D(p+2)$ branes inside an $\IRP^{5-p}$ which

474: surrounds the configuration.

475: 

476: \subsection{T Duality}

477: \label{duality}

478: 

479: Let us apply T-duality in a direction along the orientifold plane.

480: An $O(p+1)$ plane wrapping a circle of radius $L$ turns after T

481: duality to a pair of $Op$ planes on a circle of radius $L'=1/L$.

482: Since we have (at least) 4 possible types for each $Op$ plane we

483: expect 16 possible types for the wrapped $O(p+1)$ plane.

484: 

485: Although this may sound surprising at first, we will

486: demonstrate the 16 possibilities by analyzing the possible

487: discrete torsions. The transverse space is $\IRP^{7-p}$

488: with cohomologies given in appendix \ref{cohomology}, and

489: we should find the 9 dimensional fields that can have

490: discrete torsions. Before wrapping the $O(p+1)$ plane the

491: fields with $\IZZ$ discrete torsions are $B_{NS}$ and

492: $C_{5-(p+1)=4-p}$. After compactifying on $L$ there are

493: two more such fields

494: - the reduction of the metric on $L$ (an untwisted 1

495: form), and the reduction of $C_{6-p}$ on $L$. To find the

496: relation with the T dual picture it is useful to list the

497: various discrete torsions in both pictures

498: 

499: \be

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

501: \hline

502: $O(p+1)$ & $B_{NS}$ & $C_{4-p}$ & $g_{\mu \nu}/L$ & $C_{6-p}/L$ \\

503: \hline

504: $O(p)$ & $B_{NS}$ & $C_{5-p}/L'$ & $B_{NS}/L'$ & $C_{5-p}$ \\

505: \hline

506: \end{tabular}

507: \label{Tduality}

508:  \ee

509:  where the notation $C_q/L$ means ``the form

510: $C_q$ reduced on the circle $L$''. Let us restate this

511: mapping in terms of brane intersections

512:  \be

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

514: \hline

515: $O(p+1)$ & NS5 & $D(p+3)$ & KK monopole & $\hat{D}(p+1)$ \\

516: \hline

517: $O(p)$ & NS5 & $\hat{D}(p+2)$ & $\widehat{NS}5$ & $D(p+2)$ \\

518: \hline

519: \end{tabular}

520: \label{TdualityBranes}

521:  \ee

522:  where the hat above a brane means that

523: it does not wrap the circle.

524: 

525: Let us discuss some examples. An NS5 intersecting the

526: $O(p+1)$ turns into an NS5 which intersects both of the

527: $Op$'s while an intersection with a $D(p+3)$ turns into a

528: $D(p+2)$ which intersects only one out of the pair. If we

529: want to get an RR discrete torsion on both $Op$'s then we

530: should intersect the $O(p+1)$ with a $D(p+1)$, one which

531: does not wrap the circle (and still has 4 mixed directions

532: relative to the O plane). If we take an $Op^-,~Op^+$ pair,

533: which has zero charge, then it transforms into an $O(p+1)$

534: intersecting a KK monopole on the circle $L'$ and must

535: have zero charge as well.

536: 

537: \section{Review of M-lifts of Orientifolds}

538: \label{revM} In this section we review known lifts of orientifold

539: planes to M theory. Such a lift requires lifting the

540: $\IZZ$ action to M theory. The objects we describe are

541: lifts of both orientifolds and orbifolds and accordingly

542: are denoted as $OMp$ planes. Since the worldsheet

543: formulation is lost in the lift, orientation reversal is

544: meaningless. Nevertheless we shall sometimes continue to

545: call them ``orientifolds''.

546: 

547: Lacking a fundamental definition of M theory, we are satisfied by

548: specifying the $\IZZ$ action on the 11d supergravity fields. An

549: $OMp$ plane includes a transverse spatial reflection, so we look

550: at M theory on ${\bf R}^{p,1} \times {\bf R}^{10-p}/ {\bf Z}_2$

551: where the first factor is the worldvolume of the $OMp$ plane, the

552: ${\bf Z}_2$ in the second part is the reflection, and often it

553: will be more convenient to replace the last factor with ${\bf

554: T}^{10-p}/{\bf Z}_2$. The reflection determines the action on the

555: metric ($g \to g$). The action on the 3-form $C$ is determined by

556: requiring invariance of the topological term in the action $\int C

557: \wedge G \wedge G$, where $G=dC$, to be

558: \be

559: C \to (-)^p C.

560: \ee

561: 

562: Supersymmetry provides another constraint. When acting on

563: fermions, the inversion of $10-p$ coordinates squares to

564: the identity for $10-p=0,1 ~{\rm mod} ~4$ and to $(-)^F$

565: for $10-p=2,3 ~{\rm mod} ~4$ (this is a consequence of

566: $(\Gamma_1 \Gamma_2 ... \Gamma_n)^2=(-)^{n(n-1)/2}$). Thus

567: in order to have a supersymmetric orbifold and a $\IZZ$

568: action rather than a $\IZ_4$ action we require

569: \be

570: p=1,2 ~{\rm mod} ~4 ~.

571: \label{susyp}

572: \ee

573: 

574: We see that the $\IZZ$ objects intrinsic to M theory, are

575: the OM1, OM2, OM5, OM6 and OM9. It is no surprise that the

576: orientifolds intrinsic to M theory include a 2 plane and a

577: 5 plane which we denote by OM2 and OM5. We shall review

578: the definition of these orientifolds and the way they were

579: used to find the M lift of the O4 and the O2. Then we look

580: for intersections of branes and orientifolds in M theory,

581: where it turns out that there is only one such

582: configuration: a $1/2$ M5 stuck on an OM2. After that we

583: review the lift of another Type IIA orientifold, the O0,

584: through an M theory object which we may call OM1.

585: 

586: We summarize here the results. The OM5 and OM2 carry charges given by

587: \be

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

589: \hline

590: $OM2^-$ & $OM2^+$ & OM5 \\

591: \hline

592: -1/16   & +3/16   & -1/2 \\

593: \hline

594: \end{tabular}

595: \ee where the $OM2^-, ~OM2^+$ are two discrete torsion variants of

596: the $OM2$ plane (which will be discussed later in

597: subsection \ref{OM5OM2}), and the charge is in units of

598: bulk OM2 and OM5 branes. The other planes support a

599: ``twisted sector'' matter rather than charge (they are

600: neutral in 11d)

601: \bea \nonumber

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

603: \hline

604: $OM1$ & $OM6$ & OM9 \\

605: \hline

606: a chiral fermion & a 7d $SU(2)$ vector multiplet & a 10d $E(8)$ vector multiplet

607: \\

608: \hline

609: \end{tabular}

610: \eea

611: 

612: The OM6 and OM9 will not be discussed any further in this paper.

613: The OM6 is M theory on $\IT^4/\IZZ$ and the OM9 is a Ho\v{r}ava -

614: Witten plane \cite{HoravaWitten}.

615: 

616: \subsection{The OM5 plane}

617: The ``orientifold'' M on ${\bf T}^5/ {\bf Z}_2$ was studied in

618: \cite{DasguptaMukhi, Witten_OM}. The ``untwisted'' sector has a 6d

619: gravitational anomaly that can be canceled by 16 tensor

620: multiplets. Moreover, local anomaly cancellation would seem to

621: require adding a twisted sector of 1/2 a tensor multiplet at each

622: of the 32 fixed points. This problem is avoided by using a

623: different method of canceling the anomaly - assigning to each one

624: of the fixed points a charge

625: \be

626: Q(OM5)= -1/2, \ee in units of physical M5 branes. Note that while

627: we can usually put a half brane on top of an orientifold plane,

628: one cannot put half an M5 on an OM5 due to  M theory flux

629: quantization \cite{WittenFluxQuant}.

630: 

631: \subsubsection{M-lift of the O4}

632: \label{OM5/O4}

633: 

634: The OM5 can account for the O4 planes of Type IIA

635: \cite{Hori,Gimon:1998be} (see also \cite{Ahn:ads7}). When

636: lifting the $\IZZ$ action from Type IIA to M theory it is

637: required to specify the action on the circle $R_{11}$.

638: Topologically there are three possible actions: the

639: identity, a reversal and a shift through half a circle.

640: The identity is interpreted as an OM5 wrapped on $R_{11}$.

641: Reversal is not allowed since it would accumulate to an

642: inversion in 6 directions which would break supersymmetry

643: (\ref{susyp}). A shift is the same as a non-trivial

644: discrete torsion for the Type IIA 1-form $A$. It is

645: exactly the discrete torsion which is present in the

646: $\widetilde{O4}^{\pm}$ planes. So ``untilded'' ($c=0$)

647: orientifolds M-lift to $\IR ^5/

648: \IZZ$ while the ``tilded'' ones ($c=1$) M-lift to

649: $(\IR^5 \times \IS ^1)/ \IZZ$ with the $\IZZ$ acting on

650: the circle by a shift.

651: 

652: Let us look more at the shift orbifold $(\IR^5 \times

653: \IS^1)/\IZZ$. It has two kinds of minimal 1-cycles, one of

654: them wraps $\IS^1$ (``the circle'') and the other is a

655: straight line between points identified by the $\IZZ$

656: action (``antipodal line''). For points away from the

657: origin of $\IR^5$ the circle is smaller, but as we

658: approach the origin the antipodal line becomes smaller,

659: and at the origin itself it is half the size of the

660: circle.

661: 

662: We summarize the different O4 planes and their M theory origin

663: \begin{enumerate}

664: \item $O4^-$. Charge $Q=-1/2$. This is simply an OM5 wrapping $R_{11}$.

665: \item $\widetilde{O4^-}$. Charge $Q=0$. This is the smooth shift orbifold,

666: and as such indeed does not carry charge. (Recall that a bound

667: state of an OM5 and a half M5 is not allowed).

668: \item $O4^+$. $Q=+1/2$. This is an OM5 with a full stuck M5. (It is stuck

669: by imposing as Wilson loop an element of $O(2)$ which is not

670: connected to the identity \cite{Gimon:1998be}).

671: \item $\widetilde{O4^+}$. Charge $Q=+1/2$. This is the shift orbifold with

672: a stuck M5 at the origin on the circle of half radius.

673: 

674: The two types of $\widetilde{O4}$ planes were conjectured to be

675: related to the two elements of $\pi_4(Sp(n))= \IZZ$.

676: \end{enumerate}

677: 

678: \subsection{The OM2 plane}

679: The OM2 orientifold was studied in \cite{Sethi:1998zk}

680: (see also \cite{Ahn:ads4}). The charge can be found by

681: considering the interaction $-\int C \wedge I_8(R)$ for M

682: theory on $X=\IT^8 / \IZZ$. The effective charge is

683: $-\int_X I_8(R)=-\chi/24$. Although this space is

684: singular, we can define its ``resolved cohomologies'' by

685: adding to the invariant (untwisted) cohomologies an extra

686: 256 cohomologies (in $H^{2,2}$) from the RR twisted

687: sector. This totals the Euler characteristic to 384 and so

688: the total charge to

689: -16. Dividing by 256, the number of fixed points, we find

690: $Q=-1/16$ in units of M2 charge.

691: 

692: This orientifold allows a variant due to a discrete flux. The

693: transverse space is ${\bf RP} ^7$ and the only field strength form in

694: M theory is the 4-form $G$, so we are interested in the cohomology

695:  $H^4({\bf RP} ^7,\IZ)=\IZZ$ (Appendix \ref{cohomology}). The discrete torsion

696:   adds a charge of a $+1/4= -{1 \over 2}\int_{{\bf RP}

697:   ^7} {C \over 2\pi} \wedge {G \over 2\pi}$ in units of M2.

698:  We denote the OM2 with trivial discrete torsion by $OM2^-$ and the

699:  one with non-trivial torsion by $OM2^+$. Their charges are summarized by

700: \bea

701: Q(OM2^-)=-1/16 \nonumber \\

702:  Q(OM2^+)=Q(OM2^-) +1/4=+3/16.

703: \eea

704: 

705: \subsubsection{The brane - orientifold intersection in M theory}

706: \label{OM5OM2}

707: 

708: The \OMtp can be realized by branes.

709:  To do that we put a half M5 brane on an

710: \OMtm plane. This is an analogue (the only one) of the Type II

711: configurations in section (\ref{review}).

712: 

713: Note that like the case of the OM5, one cannot attach a half M2

714: brane on top of an OM2 due to M theory flux quantization

715:  \cite{WittenFluxQuant}.

716: 

717: Once we compactify M theory we get (at least) two more

718: possibilities for brane intersections. One possible configuration

719: comes from lifting OF1 and D6 with charge jump of 1/16. It lifts

720: to a wound OM2 intersecting a KK6 - a Kaluza Klein monopole

721: (section \ref{o1}). Another configuration is the lift of an OF1

722: intersecting a D2, that is a wound OM2 intersecting with a

723: transverse M2 with a tension jump of $+1$ (sections \ref{o1-M},

724: \ref{o1}). An inspection of this example shows that the jump of

725: one unit of F1 charge is actually represented by a physical wound

726: M2 brane which is stretched, with its mirror, like a T shaped

727: brane (this is the M-lift of a double F1 ending on a D2

728: \cite{ASY}).

729: 

730: \subsubsection{M-lift of the O2}

731: \label{OM2}

732: 

733: Once one identifies the M theory objects, the \OMtm and

734: the \OMtp,  one can go on and find the M-lift of the O2

735: planes in Type IIA \cite{Sethi:1998zk,BerkoozKapustin}.

736: Actually, originally it must have been easier to go in the

737: opposite direction and determine the charges of the OM

738: planes from the O2 planes. Like the case of M-lifting the

739: O4, we need to lift the $\IZZ$ action to $R_{11}$. This

740: time it must be a reversal - it cannot be the identity

741: because of the susy constraint on $p$ (\ref{susyp}). So in

742: M theory there are actually two fixed planes located at

743: the two fixed points on the circle.

744: 

745: The different O2 planes and their M theory origin are described by

746: \begin{enumerate}

747: \item $O2^-$. Charge $Q=-1/8$. This is a pair of \OMtm

748:   planes: $-1/8=2 \times -1/16$.

749: \item $\widetilde{O2^-}$. Charge $Q=-1/8 ~+1/2=+3/8$. This is a pair

750:   of \OMtp planes: $-3/8=2 \times -3/16$.

751: \item $O2^+, ~\widetilde{O2^+}$. Both have $Q=+1/8$. This is a

752:   composite pair of an \OMtm with an \OMtp, $1/8=-1/16 ~+3/16$. The two possible

753:   O2's correspond to the possible ordering of the

754:   OM2's. This can be seen by intersecting an $O2^+$ with a D4

755:   brane. After lifting to M theory and using the intersection rule

756:   explained in section (\ref{review}), one finds that the $\widetilde{O2^+}$ has the

757:   reversed order of OM2's.

758: \end{enumerate}

759: 

760: \subsection{The OM1 line}

761: \label{OM1}

762: 

763: Let us consider the M-lift of the orientifold point, the O0, of

764: Type IIA. We take the action on $R_{11}$ to be the identity (a

765: reversal is not allowed by equation \ref{susyp}),

766:  so we consider

767: M theory on $R_{11} \times \IS ^1 \times \IR ^9/\IZZ$. A

768: computation of 2d Gravitational anomalies for M theory on

769: $\IT^9/\IZZ$ suggests that there is a chiral fermion on

770: every fixed line as an ``M theory twisted sector''

771: \cite{DasguptaMukhi,rey}. An independent evidence for the

772: existence of a chiral fermion on the fixed line comes from

773: the computation of the Witten index of $Sp(N), ~SO(N)$

774: matrix quantum mechanics

775: \cite{KacSmilga,HananyKolRajaraman}. We may call this line

776: an OM1 orientifold.

777: 

778: It is interesting to get the ``twisted sector'' which is

779: described above truly from a twisted sector of string

780: theory (see also \cite{DasguptaMukhi,DasguptaMukhi2}). In

781: order to get a 2d model one needs to compactify on an 8

782: manifold, and Type IIB on $\IT^8/

783: \IZZ$ has the right action on the fields (this is the

784: orbifold which we will call $OP1_B$ in section \ref{o1}).

785: Type IIA divided by $I_8 \cdot (-)^{F_L}$ would also do,

786: but we will stay with the more geometric example. The

787: $OP1_B$ has a twisted sector from the 4 form wrapping the

788: 256 resolved $H^{2,2}$ cohomologies (\cite{orbifolds}

789: describes the computation of twisted sectors in general).

790: These scalars are chiral as they inherit their self

791: duality from the 4 form. Thus we get one chiral scalar for

792: each of the 256 fixed lines. Comparing with the 512 chiral

793: fermions of M theory on $\IT^9 / \IZZ$ we see that they

794: could match by bosonization provided the periodicity of

795: the scalars is at the free fermionic value, as it should.

796: This actually gives a nice realization of 2 dimensional

797: bosonization as implied by a lift to M theory.

798: 

799: As we consider here the OM1 on the $R_{11}$ circle there

800: are two possible boundary conditions for the fermion. The

801: Neveu-Schwarz boundary conditions correspond to an O0 with

802: trivial $RR$ discrete torsion ($c_{RR}=0$), and Ramond

803: corresponds to $c_{RR}=1$. This is verified by a

804: computation of the Casimir energy in the two cases, which

805: matches the O0 mass

806: \be

807: M(O0)= \pm 1/32

808: \ee

809: where the units are of momentum quanta along the circle.

810: 

811: Let us  summarize the different O0 planes and their M theory

812: origin

813: \begin{enumerate}

814: \item $O0^-$. Charge $Q=-1/32$. This is an OM1 with NS boundary

815:   conditions, and with integral momentum.

816: \item $\widetilde{O0^-}$. Charge $Q=-1/32 ~+1/2$. This OM1 has NS

817:   boundary conditions, but carries half-integer momentum.

818: \item $O0^+, ~\widetilde{O0^+}$. Both have $Q=+1/32$. These OM1's are

819:   in the Ramond sector, and there are two of them due to the zero

820:   mode which generates a degenerate ground state.

821: \end{enumerate}

822: 

823: \section{M lifts of Type IIB Orientifolds}

824: \label{Mlift}

825: In this section we describe (new) M-lifts of various

826: orientifold planes in Type IIB, while paying special

827: attention to the transformation properties under \SLtz.

828: 

829: We start by M-lifting the O3 plane, where we get a nice geometric/

830: microscopic realization of the \SLtz symmetry. Then we discuss the

831: O1 and O5 planes. Their S duals, which we call OF1 and ON5, are

832: constructed, and will be further discussed in section \ref{orb}.

833: The ON5 was already discussed in some works (see \cite{HZ} for a

834: recent review on this plane), and the OF1 was discussed in

835: \cite{Sen:1996na}.

836: 

837: The method is to recall the M theory origin of a Dp brane in terms

838: of M branes and then to find the analogous construction of an Op

839: plane in terms of the OM planes which were reviewed in the

840: previous section. The basic correspondence is between M theory on

841: $\IT^2_M$ and IIB on a circle of radius $L_{IIB}$, so we should

842: always distinguish two cases according to whether the Dp brane

843: wraps \Ltb or not. At weak coupling the \Ttm has a short side and

844: a long side, such that their ratio is the string coupling (when

845: the RR axion vanishes).

846: 

847: \subsection{O3}

848: Let us recall the M-lift of the D3 brane. A D3 which does

849: not wrap \Ltb is an M5 wrapping the torus, while a D3

850: which wraps \Ltb is the M2.

851: 

852: The four kinds of O3 planes have the following charges (in

853: D3 units)

854: \bea

855:  Q(O3^-) = -1/4, \nonumber \\

856:  Q(\widetilde{O3^-}) ~=Q(O3^+) ~= Q(\widetilde{O3^+}) = ~+1/4.

857: \eea

858: 

859: We start by lifting an O3 which wraps \Ltb, for otherwise

860: the circle

861: \Ltb is inverted as well and we get two O3 planes at the two fixed

862: points. Since the M-lift of a D3 brane which wraps

863: \Ltb is an M2 which does not wrap \Ttm, we should take an

864: OM2 which does not wrap \Ttm. Because of the compactness

865: of

866: \Ttm we are considering actually {\it four OM2 planes}.

867: The simplest possibility is to take four

868: \OMtm planes. One checks that the total charge $4 \times

869: (-1/16)= ~-1/4$ fits the $O3^-$ plane as expected.

870: 

871: \EPSFIGURE[ht] {o3fig,width=10cm}{The four types of $O3$ orientifolds

872: lifted to M theory. The $\pm$ signs stand for $OM2^\pm$

873: planes inside the M theory torus. \label{o3fig}}

874: 

875: 

876: To get the other O3 planes

877: \footnote{A.H. would like to thank Jacques Distler for discussions

878: on related points.}

879:  we use the brane intersection picture

880: of section \ref{review}. For example, to get the M-lift of

881: the $O3^+$ we should intersect the $O3^-$ with an NS5

882: brane. The lift of the NS5 is an M5 wrapping the long

883: side. Using the basic intersection in M theory (section

884: \ref{OM5OM2}) we see that we end up with two \OMtp planes

885: along the long side, and two other \OMtm planes, as in

886: figure \ref{o3fig}. The charges fit since $2 \times

887: (-1/16) + 2 \times (+3/16)=+1/4$. A similar argument works

888: for the $\widetilde{O3^-}$ and the $\widetilde{O3^+}$

889: planes by replacing the NS5 branes with a D5 brane or a

890: (1,1) brane respectively. Figure (\ref{o3fig}) summarizes

891: the various configurations.

892: 

893: Recall that the \SLtz properties of O3 planes can be described in

894: a diagram such as figure \ref{o3fig} \cite{Uranga,

895: WittenAdSBaryons}, where \SLtz acts on the torus in the diagram

896: according the its natural action on $(\IZZ)^2$. This action is

897: clearly visible from our M-lift into \Ttm.

898: 

899: There is an alternative way of finding the M lift of an O3

900: which uses T duality (section \ref{duality}). Under T

901: duality an O3 that wraps \Ltb turns into a pair of O2

902: planes, each one of which can be lifted to a pair of OM2

903: planes, as in section \ref{OM2}, giving 4 OM2 planes as

904: above. In this way one can recover the different lifts for

905: the different variants.

906: 

907: One may wonder about other choices for the signs of the

908: four OM2 planes. The ones with an odd number of signs

909: cannot be constructed by intersecting the $O3^-$ with a

910: Type IIB 5-brane as the 5-brane must intersect exactly two

911: OM2 planes (because the M5 is oriented). Nevertheless,

912: such configurations can be constructed making use of the

913: large but compact circle \Ltb, by intersecting the wound

914: O3 (wound on \Ltb) with a KK monopole (on \Ltb) as in

915: (\ref{TdualityBranes}). One gets a configuration with

916: three $OM2^-$ and one $OM2^+$ and total charge zero.

917: Intersecting now with an NS5 would give the other

918: possibility - one $OM2^-$ and three $OM2^+$. The case of 4

919: \OMtp planes is probably equivalent to four \OMtm planes

920: with an additional M2 brane.

921: 

922: Now we turn to an O3 which does not wrap \Ltb. It is

923: actually a pair of O3 planes, and we may T- Dualize them

924: into an O4 plane as in section (\ref{duality}). The M-lift

925: of the latter was described already in section

926: \ref{OM5/O4} in terms of OM5 planes.

927: 

928: \subsection{O5 - ON5}

929: Let us discuss O5 planes which wrap \Ltb, both because we

930: are less interested in a pair of O5's which we would have

931: had if the compact direction were inverted, and since we

932: are interested in configurations which lift to M5 branes

933: rather than KK monopoles.

934: \footnote{An alternative M lift

935: of the O5, the $ON5_B$ and their variants was given in

936: \cite{Witten6d}, using OM6 planes and more elaborate

937: quotients rather than OM5 planes.}

938:  For 5 branes, A \pq 5 brane which wraps \Ltb lifts to an

939: M5 wrapping a \pq cycle of

940: \Ttm. By analogy, we attempt to lift the O5 to an OM5

941: wrapping the short side of \Ttm. Such an OM5 plane is

942: actually a pair of OM5's because of the transverse compact

943: coordinate on \Ttm. We check that the charges match: $2

944: \times Q(OM5)= 2 \times (-1/2)= -1=Q(O5)$.

945: 

946: So far we have discussed the $O5^-$. We would like to

947: construct other discrete torsion variants of the O5, ones

948: which are independent of the compactification on

949: \Ltb, namely, those which are not related to forms that were

950: reduced on \Ltb. This can be done by turning on discrete

951: torsions for M theory on the torus.

952: 

953: Performing \SLtz we can get a family of \pq O5 planes,

954: such that the charge of a \pq $O5^-$ plane is $-1$ in

955: units of a \pq 5-brane. In particular we can consider a

956: $(0,1)$ O5 plane which we call $ON5_B$ because it is

957: charged under the same field that couples to the NS5 brane

958: of Type IIB (the charge is -1 in units of the NS5 charge).

959: The system of $ON5_B^-$ together with an NS5 brane can be

960: identified to be the $IIB/I_4 (-)^{F_L}$ orbifold and can

961: be called $ON5_B^0$ (section \ref{o5}; see \cite{HZ} for a

962: more detailed discussion). A set of N NS5 branes in the

963: vicinity of an $ON5_B$ results in 6d worldvolume gauge

964: theory, the same as a set of D5 branes near an O5 plane,

965: with the gauge group being one of $SO(2N),SO(2N+1),

966: ~Sp(N)$ according to the type of the $ON5_B$.

967: 

968: Since NS5 branes exist both in Type IIA and in Type IIB, one might

969: expect the ON5 to exist in Type IIA as well. Indeed, the $IIA/I_4

970: (-)^{F_L}$ orbifold (section \ref{o5}), which we call an

971: $ON5_A^0$, is a system composed of an $ON5_A$ with an NS5 brane.

972: The $ON5_A^-$ has the property that when N NS5 branes coincide

973: with it, the worldvolume theory is a (2,0) CFT with a global

974: symmetry group $SO(2N)$, and this is the only possible variant.

975: 

976: \subsection{O1 - OF1}

977: \label{o1-M} The case of O1 is quite similar to the O5. Consider

978: an O1 which does not wrap \Ltb (so there is actually a pair of

979: O1's). As a D1 which does not wrap \Ltb M-lifts to a membrane

980: which wrap the long side of the torus, we should try an OM2 plane

981: wrapping the long side (so again there are actually two of them

982: because of the transverse short side). Let us check the charges:

983: $2 \times (-1/16)$ for the pair of O1's, indeed equals $2 \times

984: (-1/16)$ for the pair of OM2's.

985: 

986: For O1 planes which wrap \Ltb, we recall that  a D1

987: wrapping \Ltb is described by a unit of momentum along the

988: short side of \Ttm. So we try to wrap an OM1 along the

989: short side of \Ttm as its mass scales like units of

990: momentum (actually it is a pair of OM1's due to the

991: transverse long side). The charges work out for an $O1^-$

992: being made of a pair of $OM1^-$: $-1/16= 2 \times

993: (-1/32)$.

994: 

995: One can get other variants of the O1 by lifting brane intersection

996: to M theory. A new configuration happens for a pair of $O1^-$'s

997: which do not wrap \Ltb and are intersected by a D3 which does.

998: After the intersection we get a pair of $\wtilde{O3^-}$, and so

999: the tension jump is $2 \times +1/2= ~+1$. By lifting to M theory

1000: we learn that the intersection of a wound OM2 with a transverse M2

1001: gives a $+1$ tension jump.

1002: 

1003: Performing \SLtz we can get a family of \pq O1 planes,

1004: such that the charge of a \pq $O1^-$ plane is the same as

1005: for a \pq string. In particular we can consider a $(1,0)$

1006: O1 plane which we call $OF1_B$ because it is charged under

1007: the same field that couples to the fundamental string. We

1008: will see that the $OF1_B$ can be identified with the

1009: $IIB/I_8(-)^{F_L}$ orbifold (section \ref{o1}). Since

1010: fundamental strings exist both in IIA and in IIB, one

1011: might expect the OF1 to exist in Type IIA as well. Indeed,

1012: this is the $IIA/(-)^{F_L}$ orbifold (section \ref{o1}),

1013: and we call it an $OF1_A$. The perturbative orbifold

1014: variants may be referred to as $OF1_B^0, ~OF1_A^0$.

1015: 

1016: \section{Orbifolds, Orientifolds and New Variants}

1017: \label{orb}

1018: 

1019: Here we discuss the relations between orientifolds and

1020: orbifolds and their variants. We start with lines, then 5

1021: planes and then the O0. Throughout this section, when we

1022: identify a perturbative orbifold with some plane which is

1023: a dual of an orientifold, it should be borne in mind that

1024: the identification holds only for one variant of the

1025: plane, possibly with some extra matter, and all other

1026: variants are produced by changing non-perturbative

1027: discrete torsions.

1028: 

1029: \subsection{Orbifold lines}

1030: \label{o1}

1031: Orbifold lines together with the O1 form a family

1032: connected by dualities. Table \ref{o1map} is our roadmap

1033: for these connections. We will first explore this map and

1034: then present some results on the tensions of discrete

1035: torsion variants and a relation with K theory.

1036: 

1037: \TABLE{

1038: \begin{tabular}{lccc}

1039:           & \fbox{$IIB/I_8 ~\Omega ~~~~O1$} &   &\vspace{1mm}\\

1040:           &  S $~~~\updownarrow$                    & & \\ \vspace{1mm}

1041: \parbox{2cm}{(8,8) in 2d; OF (OM2)\\+variants}  & \fbox{$IIB/I_8 ~(-)^{F_L} ~~~~OF1_B$}&

1042: $\longleftrightarrow ~~T \perp$ & \fbox{$IIA/I_8 ~~~~OF1_A$}

1043: \vspace{1mm}\\

1044:           & T $\parallel$ $~~~\updownarrow$   & & \vspace{1mm}\\

1045: (16,0) in 2d;\\ OP (OM1)  & \fbox{$IIA/I_8 ~(-)^{F_L}

1046: ~~~~OP1_A$}& & \fbox{$IIB/I_8 ~~~~OP1_B$}\vspace{1mm}\\

1047: \end{tabular}

1048: \caption{Orbifold and orientifold lines. In this table the charges

1049: and fluxes of the various O planes are not specified and may

1050: change, depending on the particular case.} \label{o1map} }

1051: 

1052: Our starting point is the O1, that is, $IIB/I_8 \Omega$.

1053: It carries D1 charge and the forms $B_{NS},C_0,C_4$ are

1054: odd (twisted) under it. It has discrete torsion variants

1055: due to $H_{NS},*H_{NS},G_5,G_1$ or in terms of brane

1056: intersections due to the NS5, F1, D3 and a 7 brane. The 7

1057: brane must allow a D1 charged object to end on it, and so

1058: it should be a (0,1) 7 brane rather than a (1,0) D7.

1059: 

1060: We denote the S dual of the O1 by $OF1_B$ as S duality

1061: replaces a D1 charge with an F1 charge. S duality replaces

1062: $\Omega$ with $(-)^{F_L}$ and so this orbifold is $IIB/I_8

1063: (-)^{F_L}$\footnote{To be more precise, one should note

1064: that the $O1^-$, which is the O1 plane with no discrete

1065: fluxes carries charge $-{1\over16}$, which must be

1066: cancelled by adding non-perturbative discrete torsion and/

1067: or extra matter, since the orbifold Type $IIB$ on

1068: $\IT^8/\IZZ$ has 0 charge.}. The odd (twisted) forms under

1069: the projection are $C_2,C_0,C_4,C_6$ and we get 16 (!)

1070: discrete torsion variants from all four, or in terms of

1071: branes from intersections with D1,D3,D5,D7. Since an F

1072: string can end on any D brane, it is natural that each

1073: intersection with a D brane is allowed and gives a new

1074: variant.

1075: 

1076: Operating on the $OF1_B$ with a T duality in a direction

1077: transverse to the fixed line gives an orbifold which we

1078: denote by $OF1_A$, just as this operation acting on the F

1079: string of IIB would give the F string of Type IIA. Such a

1080: T duality is accompanied by an additional $(-)^{F_L}$, so

1081: operating on $IIB/I_8 (-)^{F_L} \equiv OF1_B$ we get the

1082: $IIA/I_8$ orbifold. This orbifold has 8 variants due to

1083: intersections with D branes $D2,D4,D6$ or their respective

1084: forms $G_2,G_4,G_6$. A D8 intersection is different

1085: because the associated $G_0$ form has $H^0(\IRP^7)=\IZ$

1086: cohomology rather than $\IZZ$ and is interpreted as a

1087: change in the Type IIA cosmological constant. The D0 is

1088: not in the list since it has no cohomology $H^8(\IRP^7)=0$

1089: (but it may produce variants nevertheless).

1090: 

1091: Other orbifolds can be constructed now by compactifying an

1092: OF1 line on a circle and performing parallel T duality.

1093: This time one does not add an extra $(-)^{F_L}$. We get

1094: $IIB/I_8$ and $IIA/I_8 (-)^{F_L}$. These orbifolds do not

1095: have discrete torsion variants (when uncompactified). By T

1096: duality they carry a momentum charge, so we denote them by

1097: $OP1_A, OP1_B$.

1098: 

1099: The M lift of the O-lines can be found by looking at their

1100: charges. After recalling the M lift of the F string we

1101: conclude that the OF1 planes must be wrapping modes of the

1102: OM2. The OP1 planes, on the other hand, are an unwrapped

1103: OM1. Since the OM1 carries a chiral fermion (section

1104: \ref{OM1}) after being compactified its Casimir energy

1105: will give the required momentum charge.

1106: 

1107: Let us now find the \href{\hash ref-of1Tensions}{tensions} of

1108: some of the discrete torsion variants of the OF1 lines

1109: (table \ref{of1Tensions}). The tension of a bare OF1 is

1110: $-1/16$ (in F string units) by S duality with the O1 . It

1111: is consistent with the M description as an OM2 wrapping

1112: the 11'th dimension, since the OM2 has tension $-1/16$ (in

1113: M2 units).

1114: 

1115: \TABLE{

1116: \begin{tabular}{|c|c|c|c|c|c|c|}

1117: \hline

1118: D1 & D2 & D3 & D4 & D5 & D6 & D7 \\

1119: \hline

1120: +2? & +1 & +1/2 & +1/4 & +1/8 &  +1/16? & +1/32? \\

1121: \hline

1122: \end{tabular}

1123: \caption{Tension jumps of the OF1 after various intersections.

1124: Question marks denote charges for which more consistency

1125: checks are needed.} \label{of1Tensions} }

1126: \begin{itemize}

1127: \item To compute the tension of an $OF1_B$ after intersecting a D5, consider

1128: performing S duality to an O1 which upon intersecting an

1129: NS5 changes from $O1^-$ of tension $-1/16$ to an $O1^+$ of

1130: tension $+1/16$ (so the jump is $+1/8$).

1131: \item The intersection of an $OF1_A$ with a D4 can be M lifted to the basic

1132: intersection of an OM2 with an M5 (section \ref{OM5OM2}),

1133: and so the tension jump is $+1/4$.

1134: \item The tension jump of an $OF1_B$ intersecting with a D3 is found again by S duality to

1135: be $+1/2$.

1136: \item The tension jump of an $OF1_A$ intersecting a D2 is

1137: $+1$. This is a consequence of the M theory configuration

1138: found in section \ref{o1-M}, where a wound OM2 intersects

1139: a transverse M2. Since the tension jump is integral there

1140: is nothing to prevent a whole F1 to separate from this

1141: variant.

1142: \item We seem to get a rule that after intersecting a Dp brane the jump is

1143: $2^{2-p}$, which is consistent with T duality. For the D6

1144: this rule has an independent check. A D6 intersection

1145: corresponds after an M lift to a shift in $R_{11}$. So

1146: this $OF1_A$ variant lifts to M theory on the smooth

1147: manifold $(\IT^8 \times \IS^1)/\IZZ$ where the $\IZZ$ acts

1148: by inversion on the first factor and by a shift on the

1149: second. As a smooth manifold it carries zero tension, in

1150: agreement with a $-1/16$ jump.

1151: \item

1152: The intersection of an $OF1_A$ with a D0 does not have a

1153: discrete cohomology as $H^8(\IRP^7)=0$ so it is not clear

1154: whether it gives a new variant. If there is a new variant

1155: corresponding to intersection of the OF1 with D0 the jump

1156: in its charge is 4. Conservation of the fundamental string

1157: charge then implies that 4 physical fundamental strings

1158: must enter the D0 together with the $OF1^-$ plane. This is

1159: indeed consistent with the ``fork'' configuration of

1160: \cite{Berlin} which gives some support for its existence.

1161: \end{itemize}

1162: 

1163: 

1164: One can discuss OF1 orbifolds with a discrete torsion from

1165: several forms turned on at the same time. It would be

1166: interesting to find their tensions.

1167: 

1168: The above variants have an interesting relation with

1169: \underline{K theory}. It is simpler to consider $OF1_A=

1170: ~IIA/I_8$. Discrete torsion variants are classified by the

1171: reduced cohomology $H^*(\IRP^7)=\IZZ \oplus \IZZ \oplus

1172: \IZZ$. However, it was recently claimed that the correct

1173: classifying group is the reduced K group

1174: \cite{MooreWitten} $K(\IRP^7)=\IZ_8$ (``reduced'' simply

1175: means in both cases that we do not write down a trivial

1176: $\IZ$ factor). The K group is actually a ring which

1177: differs from the standard $\IZ_8$ and is defined by the

1178: following relations on its generator $x$

1179: \bea

1180: 8x=0 \nonumber \\

1181:  x^2 =-2x

1182: \eea

1183: Note that both the cohomology and the K ring have the same order (8) whereas

1184: their structures differ. It would be interesting to elucidate the role of the

1185: algebraic structure.

1186: 

1187: \subsection{Orbifold 5 planes}

1188: \label{o5}

1189: We can take a similar tour of orbifold 5 planes, with

1190: table \ref{O5s} as our road map.

1191: 

1192: \TABLE{

1193: \begin{tabular}{lccc}

1194:           & \fbox{$IIB/I_4 ~\Omega ~~~~O5$} &   &\vspace{1mm}\\

1195:           &  S $~~~\updownarrow$                    & & \\ \vspace{1mm}

1196: 

1197: \parbox{2cm}{(1,1) in 6d\\+variants}  & \fbox{$IIB/I_4 ~(-)^{F_L} ~~~~ON5_B$}&

1198:  $\longleftrightarrow ~~T \perp$ & \fbox{$IIA/I_4 ~(OM6)$} \vspace{1mm}\\

1199: 

1200:           & T $\parallel$ $~~~\updownarrow$   & & \vspace{1mm}\\

1201: 

1202: (2,0) in 6d  & \fbox{$IIA/I_4 ~(-)^{F_L} ~~~~ON5_A ~(OM5)$}&  &

1203: \fbox{$IIB/I_4$}\vspace{1mm}\\

1204: \end{tabular}

1205: \caption{Orbifold and orientifold 5 planes}

1206: \label{O5s}

1207: }

1208: 

1209: We start with the O5 orientifold. It has charge -1 in

1210: units of D5, and it has a pair of discrete torsions due to

1211: the forms $G_1,H_{NS}$, or in terms of branes due to

1212: intersection with $D7, NS5$.

1213: 

1214: S duality creates the orbifold $IIB/I_4 (-)^{F_L}$, which

1215: we denote by $ON5_B$ since it carries NS5 charge. In order

1216: to cancel the charge one needs to add to the $ON_5$ plane

1217: an NS5 and this is actually the perturbative orbifold. The

1218: matter living on the NS5 of Type IIB is, of course, in

1219: (1,1) 6d multiplets.  It has variants from the RR forms

1220: $G_1,G_3$, allowing all the $SO,Sp$ gauge groups.

1221: 

1222: Performing T duality parallel to an $ON5_B$ gives Type

1223: IIA/$I_4 (-)^{F_L}$, which we denote by $ON5_A$, because

1224: it has NS5 charge as well. It carries matter in (2,0) 6d

1225: multiplets. If N NS5 branes are added to the orbifold we

1226: get a (2,0) theory with $SO(2N)$ group.

1227: 

1228: We can also perform T duality transverse to the orbifold

1229: planes. We get $IIA/I_4$ and $IIB/I_4$. The first orbifold

1230: has a variant due to $G_2$. These orbifolds are known to

1231: carry non perturbative matter

1232: - a (1,1) theory in the first case, and (2,0) in the

1233: second.

1234: 

1235: Let us consider the M lifts of Type IIA orbifold 5-planes.

1236: We can imagine two M theory orbifolds that would give us 5

1237: planes in Type IIA, either an OM6 wound on $R_{11}$ or an

1238: unwound OM5. By comparing the action on the fields we find

1239: that the $ON5_A=IIA/I_4 ~(-)^{F_L}$ lifts to the unwound

1240: OM5 while $IIA/I_4$ lifts to a wound OM6. The case of Type

1241: IIB can be discussed as well, but it has more detail

1242: because one needs to specify whether the 5 plane wraps

1243: \Ltb or not.

1244: 

1245: \subsection{New variants of the O0 plane}

1246: For the low dimensional orientifolds, the O0 and the O1, a

1247: discrete torsion analysis predicts the existence of more

1248: than two $\IZZ$ parameters. For the O1 we saw that there

1249: are four $\IZZ$ parameters. In addition to the usual

1250: $(b,c)=H_{NS},G_5$ there are also a pair of $\IZZ$'s from

1251: $G_1$ and $*H_{NS}$.

1252: 

1253: Similarly, for the O0 we can analyze the possible discrete

1254: torsions. In addition to the expected pair of $\IZZ$ parameters

1255: $(b,c)=H_{NS},G_6$, there is a third $\IZZ$ from $G_2$. After an M

1256: lift, this additional discrete torsion is nothing but a shift in

1257: $R_{11}$, the possibility which was not considered section

1258: \ref{OM1}. It would be interesting to find the mass of the 4

1259: variants with $[G_2] \ne 0$.

1260: 

1261: 

1262: \section{Miscellaneous applications}

1263: \label{misc}

1264: \subsection{The spectrum of 4d ${\cal N}=4$ with $SO, ~Sp~$ gauge group}

1265: \label{spectrum}

1266: We know that every orientifold 3-plane gives rise

1267: to a 4 dimensional gauge theory on D3 probes parallel to it. Each

1268: O3 plane gives rise to a theory with a different gauge group $G$,

1269: according to

1270: \begin{enumerate}

1271: \item For $O3^- ~G=SO(2N)$.

1272: \item For $\wtilde{O3^-} ~G=SO(2N+1)$.

1273: \item For both $O3^+, ~\wtilde{O3^+}$ the gauge group

1274: is $G=Sp(N)$.

1275: \end{enumerate}

1276: 

1277: Since both \Othp and \Othpt have the same gauge group one may ask

1278: how do the two theories differ. It is clear that the theory with

1279: \Othpt is an \SLtz transform of the one with \Othp by the element

1280: \be

1281: T=\left[\begin{array}{cc}

1282:  1 & 1 \\

1283:  0 & 1

1284:  \end{array}\right].

1285: \ee

1286: 

1287:    We would like to show how this difference manifests itself in

1288: one of the basic measurables of the theory, the spectrum of 1/2

1289: BPS states.

1290: 

1291: By Definition we know that the W bosons lie in the root

1292: lattice of the gauge group $G$. Knowing the lattice of the

1293: W bosons and the

1294: \SLtz transformation which acts naturally both on the

1295: charges of the states and on the discrete charges of the

1296: orientifolds allows us to characterize the 1/2 BPS

1297: spectrum as follows

1298: \be

1299: \begin{tabular}{|r|c|c|c|c|}

1300: \hline

1301:                 & \Othm & \Othmt & \Othp & \Othpt \\

1302:         $(b,c)$= & (0,0) & (0,1)  & (1,0) & (1,1) \\

1303: \hline \hline

1304: \pq=(1,0) mod 2 & D     & B      & C     & C      \\

1305: \pq=(0,1) mod 2 & D     & C      & B     & C      \\

1306: \pq=(1,1) mod 2 & D     & C      & C     & B      \\ \hline

1307: \end{tabular}

1308: \ee Here $D$ denotes that the states lie in a $D=SO(2N)$ lattice,

1309: $B$ is a $B=SO(2N+1)$ lattice and $C$ is a $C=Sp(N)$ lattice. We

1310: get that the charge lattice is of type B iff $(p,q)=(c,b) ~\rm{mod

1311: ~2}$.

1312: 

1313: Note that the difference between \Othp and \Othpt is manifest in

1314: their spectrum of monopoles and dyons - whereas the monopoles of

1315: the \Othp theory lie in the $B$ lattice (the dual lattice), the

1316: monopoles of the \Othpt theory lie in a $C$ lattice just like the

1317: W bosons.

1318: 

1319: \subsection{A comment on allowed BPS stated and $\IZZ$ charges}

1320: \label{BPS}

1321: 

1322: Consider a configuration with an $Op$ plane and a physical Dp

1323: brane away from it. Our aim is to ask what are the allowed BPS

1324: configurations which can stretch in between the Dp brane and its

1325: image. For the simplest case, with $Op^-$ the gauge group is

1326: $SO(2)$. It is known that a fundamental string stretched between

1327: the Dp brane and its image does not lead to a BPS state but

1328: rather, as Sen shows in detail in \cite{Sen}, to a non-BPS state.

1329: This happens because the BPS ground state is projected out, so the

1330: lowest state is the next massive level which is not  BPS.

1331: 

1332: We would like to extend this discussion to monopole

1333: solutions and, when possible, to dyonic states. The first

1334: example is for $p=3$. Using the results of the previous

1335: subsection we find that the allowed BPS states are

1336: described by the following table

1337: \be

1338: \begin{tabular}{|c|l|l|l|l|}

1339: \hline (b,c) & (0,0) & (1,0) & (0,1) & (1,1) \\ \hline \hline

1340: Fundamental String & no & yes & no & yes \\ D1 brane & no & no &

1341: yes & yes \\ \hline

1342: \end{tabular}

1343: \ee

1344: 

1345: It is easy to see from this table that a brane is allowed

1346: as a BPS state whenever the $\IZZ$ of the two form which

1347: couples to it electrically is non trivial.

1348: 

1349: This leads to the following generalization for any $p$. Consider a

1350: Dp brane and its image under a reflection by an $Op$ plane. Then

1351: BPS states arise when either a fundamental string or a Dp-2 brane

1352: is stretched between the brane and its image, according to the

1353: following table:

1354: \be

1355: \begin{tabular}{|c|c|c|c|c|}

1356: \hline (b,c) & (0,0) & (1,0) & (0,1) & (1,1) \\ \hline \hline

1357: Fundamental String & no & yes & no & yes \\ Dp-2 brane & no & no &

1358: yes & yes \\ \hline

1359: \end{tabular}

1360: \ee

1361: 

1362: Unlike the previous case, the simple rule that a brane is

1363: allowed to be stretched in between the heavier brane and

1364: its image whenever the corresponding $\IZZ$ flux of the

1365: form which couples to it electrically is non-trivial does

1366: not apply here. It is not clear how the picture

1367: generalizes.

1368: 

1369: Similar statements hold for D3 branes stretching between

1370: NS5 branes in the presence of the different types of

1371: $ON5_B$ planes.

1372: 

1373: \subsection{Orientifold webs}

1374: Since we identified \pq O1 lines and \pq O5 planes one may wonder

1375: whether \pq webs of orientifolds are possible. We would like to show

1376: that those may be possible in some special cases, but in general

1377: they do not make sense.

1378: 

1379: Let us consider the basic junction at $\tau=i$, where

1380: $\tau$ is the complex scalar of Type IIB. We have a

1381: horizontal $(0,1)$ O1 meeting a vertical $(1,0)$ OF1 and a

1382: $(1,1)$ O line which leaves the junction at 45 degrees.

1383: Each orientifold plane requires a $\IZZ$ projection. If we

1384: ignore the orientation reversal and consider only the

1385: spatial inversion we see that these 3 reflections generate

1386: a group of order 8, isomorphic to $D_4$, the dihedral

1387: group of 4 elements. It is not clear how do the different

1388: orientation reversals combine.

1389: 

1390: However, if there are two O lines at an irrational relative angle

1391: $\alpha$ (measured in radians/$2\pi$) then a composition of both

1392: reflections gives a rotation by $2\alpha$. Since we assumed

1393: $\alpha$ irrational, then this element generates an infinite group

1394: of identifications on the plane, generating a non-discrete image

1395: set from a single point. Thus we cannot hope to have an ordinary

1396: orbifold/orientifold, and this is the case for generic $\tau$.

1397: 

1398: \vspace{1cm}

1399: \noindent {\large {\bf Acknowledgements}}

1400: \vspace{.5cm}

1401: 

1402: We thank O. Bergman, E. Gimon, S. Gukov, E. Shustin, C. Vafa, E.

1403: Witten and A. Zaffaroni for enjoyable discussions.

1404: 

1405: A.H. would like to thank the Institute for Theoretical

1406: Physics at Santa Barbara and Tel-Aviv University for their

1407: kind hospitality while various stages of this work were

1408: completed. B.K. would like to thank J. Sonnenschein, S.

1409: Yankielowicz and the rest of the group at Tel Aviv.

1410: 

1411: A. H. is partially supported by the National Science

1412: Foundation under grant no. PHY94-07194, by the DOE under

1413: grant no. DE-FC02-94ER40818, by an A. P. Sloan Foundation

1414: Fellowship and by a DOE OJI award. The research of BK was

1415: supported in part by the US-Israeli Binational Science

1416: Foundation, the German--Israeli Foundation for Scientific

1417: Research (GIF), and the Israel Science Foundation.

1418: 

1419: \appendix

1420: \section{Appendix - Cohomology of $\IRP^n$}

1421: \label{cohomology}

1422: We distinguish between two kinds of cohomologies. The twisted

1423: ones, denoted by $\tilde{H}$ classify ``twisted'' forms. These are

1424: forms which reverse sign under the projection (such forms are not

1425: related to a twisted sector). The ordinary ``untwisted''

1426: cohomologies are denoted just by $H$ and are appropriate to

1427: classify forms which do not change sign under the projection.

1428: 

1429: The integral cohomologies of $\IRP^n$ are

1430: \bea

1431: H^q =&& \left\{

1432: \begin{array}{ll}

1433: \IZZ & \mbox{ $q$ even but } q \neq 0\\

1434:  \IZ & q=0, \mbox{ and for odd n } ~q=n \\

1435:  0   & \mbox{otherwise}

1436: \end{array} \right. \nonumber \\

1437: \tilde{H}^q =&& \left\{

1438: \begin{array}{ll}

1439: \IZZ & \mbox{$q$ odd}\\

1440:  \IZ & \mbox{for even n } ~q=n \\

1441:   0 & \mbox{otherwise}

1442: \end{array} \right.

1443: \eea

1444: 

1445: 

1446: These results can be easily deduced from the chain complexes

1447: \be

1448: \begin{array}{c}

1449: \leftarrow

1450: C^2 =\IZ

1451: \stackrel{\times 2}{\longleftarrow}

1452: C^1 =\IZ

1453: \stackrel{0}{\longleftarrow}

1454: C^0 =\IZ \leftarrow 0 \\

1455: \stackrel{\times 2}{\longleftarrow}

1456: C^{2m+1} =\IZ

1457: \stackrel{0}{\longleftarrow}

1458: C^{2m} =\IZ

1459: \stackrel{\times 2}{\longleftarrow}

1460: \dots \\

1461: 0 \leftarrow C^n =\IZ

1462: \leftarrow

1463: \dots

1464: \end{array}

1465: \ee

1466: and

1467: \be

1468: \begin{array}{c}

1469: \leftarrow

1470: \tilde{C}^2 =\IZ

1471: \stackrel{0}{\longleftarrow}

1472: \tilde{C}^1 =\IZ

1473: \stackrel{\times 2}{\longleftarrow}

1474: \tilde{C}^0 =\IZ \leftarrow 0 \\

1475: \stackrel{0}{\longleftarrow}

1476: \tilde{C}^{2m+1} =\IZ

1477: \stackrel{\times 2}{\longleftarrow}

1478: \tilde{C}^{2m} =\IZ

1479: \stackrel{0}{\longleftarrow}

1480: \dots \\

1481: 0 \leftarrow \tilde{C}^n =\IZ

1482: \leftarrow

1483: \dots

1484: \end{array}

1485: \ee

1486: where $C^q$ are the q-cochains and $\tilde{C}^q$ are the twisted q-cochains.

1487: 

1488: For completeness we list the integral homology groups as

1489: well

1490: \bea

1491: H_q =&& \left\{

1492: \begin{array}{ll}

1493: \IZZ & q \mbox{ odd but }  q \neq n\\

1494:  \IZ & q=0, \mbox{ and for odd n } ~q=n \\

1495:  0   & \mbox{otherwise}

1496: \end{array} \right. \nonumber \\

1497: \tilde{H}_q =&& \left\{

1498: \begin{array}{ll}

1499: \IZZ & \mbox{$q$ even } q \neq n\\

1500:  \IZ & \mbox{for even n } ~q=n \\

1501:   0 & \mbox{otherwise}

1502: \end{array} \right.

1503: \eea

1504: 

1505: For odd $n$ $\IRP^n$ is orientable and Poincare duality

1506: holds

1507: \be

1508: H_i=H^{n-i}, ~~~~~~ \tilde{H}_i=\tilde{H}^{n-i}

1509: \ee

1510: For even $n$, on the other hand

1511: \be

1512: H_i=\tilde{H}^{n-i}, ~~~~~  \tilde{H}_i=H^{n-i}

1513: \ee

1514: 

1515: \bibliographystyle{JHEP}

1516: 

1517: \begin{thebibliography}{99}

1518: 

1519: %\cite{Sagnotti:1987tw}

1520: \bibitem{Sagnotti:1987tw}

1521: A.~Sagnotti, ``Open Strings And Their Symmetry Groups,'' ROM2F-87/25 {\it Talk

1522: presented at the Cargese Summer Institute on Non-Perturbative Methods in Field

1523: Theory, Cargese, Italy, Jul 16-30, 1987}.

1524: 

1525: %\cite{Horava:1989vt}

1526: \bibitem{Horava:1989vt}

1527: P.~Ho\v{r}ava, ``Strings On World Sheet Orbifolds,'' Nucl.\ Phys.\

1528: {\bf B327}, 461 (1989).

1529: %%CITATION = NUPHA,B327,461;%%

1530: \href{http://www.slac.stanford.edu/spires/find/hep/www?j=NUPHA,B327,461}{SPIRES}

1531: 

1532: %\cite{Gimon:1996rq}

1533: \bibitem{GP}

1534: E.~G.~Gimon and J.~Polchinski, ``Consistency Conditions for Orientifolds and

1535: D-Manifolds,'' Phys.\ Rev.\  {\bf D54}, 1667 (1996) [hep-th/9601038].

1536: %%CITATION = HEP-TH 9601038;%%

1537: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9601038}{SPIRES}

1538: 

1539: \bibitem{Clifford}

1540: L.~Jarv, C.~V.~Johnson, ``Orientifolds, M-Theory, and the

1541: ABCD's of the Enhancon,'' hep-th/0002244.

1542: \href{http://xxx.lanl.gov/abs/hep-th/0002244}{hep-th/0002244.}

1543: 

1544: %\cite{Kakushadze:2000hm}

1545: \bibitem{Zurab}

1546: Z.~Kakushadze, ``Geometry of orientifolds with NS-NS B-flux,''

1547: hep-th/0001212.

1548: %%CITATION = HEP-TH 0001212;%%

1549: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/0001212}{SPIRES}

1550: 

1551: %\cite{Uranga:1999ib}

1552: \bibitem{Angel}

1553: A.~M.~Uranga, ``Comments on non-supersymmetric orientifolds at

1554: strong coupling,'' hep-th/9912145.

1555: %%CITATION = HEP-TH 9912145;%%

1556: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9912145}{SPIRES}

1557: 

1558: %\cite{Vafa:1986wx}

1559: \bibitem{Vafa:1986wx}

1560: C.~Vafa, ``Modular Invariance And Discrete Torsion On Orbifolds,'' Nucl.\ Phys.\

1561: {\bf B273}, 592 (1986).

1562: %%CITATION = NUPHA,B273,592;%%

1563: \href{http://www.slac.stanford.edu/spires/find/hep/www?j=NUPHA,B273,592}{SPIRES}

1564: 

1565: %\cite{Vafa:1995rv}

1566: \bibitem{Vafa:1995rv}

1567: C.~Vafa and E.~Witten, ``On orbifolds with discrete torsion,'' J.\ Geom.\ Phys.\

1568: {\bf 15}, 189 (1995) [hep-th/9409188].

1569: %%CITATION = HEP-TH 9409188;%%

1570: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9409188}{SPIRES}

1571: 

1572: %%%%%%% recent related papers

1573: %\cite{Dijkgraaf:1999za}

1574: \bibitem{Dijkgraaf:1999za}

1575: R.~Dijkgraaf, ``Discrete torsion and symmetric products,'' hep-th/9912101.

1576: %%CITATION = HEP-TH 9912101;%%

1577: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9912101}{SPIRES}

1578: 

1579: %\cite{Gomis:2000ej}

1580: \bibitem{Gomis:2000ej}

1581: J.~Gomis, ``D-branes on orbifolds with discrete torsion

1582: and topological  obstruction,'' hep-th/0001200.

1583: %%CITATION = HEP-TH 0001200;%%

1584: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/0001200}{SPIRES}

1585: 

1586: %\cite{Klein:2000tf}

1587: \bibitem{Klein:2000tf}

1588: M.~Klein and R.~Rabadan, ``Orientifolds with discrete

1589: torsion,'' hep-th/0002103.

1590: %%CITATION = HEP-TH 0002103;%%

1591: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/0002103}{SPIRES}

1592: 

1593: %\cite{Angelantonj:1999xf}

1594: \bibitem{Angelantonj:1999xf}

1595: C.~Angelantonj and R.~Blumenhagen, ``Discrete deformations in Type

1596: I vacua,'' hep-th/9911190.

1597: %%CITATION = HEP-TH 9911190;%%

1598: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9911190}{SPIRES}

1599: %%%%%%%%end recent related

1600: 

1601: %\cite{Hanany:1997ie}

1602: \bibitem{HananyWitten}

1603: A.~Hanany and E.~Witten, ``Type IIB superstrings, BPS monopoles, and

1604: three-dimensional gauge  dynamics,'' Nucl.\ Phys.\  {\bf B492}, 152 (1997)

1605: [hep-th/9611230].

1606: %%CITATION = HEP-TH 9611230;%%

1607: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9611230}{SPIRES}

1608: 

1609: %\cite{Bergman:1999na}

1610: \bibitem{Bergman:1999na}

1611: O.~Bergman, A.~Hanany, A.~Karch and B.~Kol, ``Branes and supersymmetry breaking

1612: in 3D gauge theories,'' JHEP {\bf 9910}, 036 (1999) [hep-th/9908075].

1613: %%CITATION = HEP-TH 9908075;%%

1614: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9908075}{SPIRES}

1615: %\cite{Armoni:1999gc}

1616: %\bibitem{Armoni:1999gc}

1617: A.~Armoni and B.~Kol, ``Non-supersymmetric large N gauge theories from type 0

1618: brane  configurations,'' JHEP {\bf 9907}, 011 (1999) [hep-th/9906081].

1619: %%CITATION = HEP-TH 9906081;%%

1620: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9906081}{SPIRES}

1621: 

1622: %\cite{Dasgupta:1996zm}

1623: \bibitem{DasguptaMukhi}

1624: K.~Dasgupta and S.~Mukhi, ``Orbifolds of M-theory,'' Nucl.\ Phys.\

1625: {\bf B465}, 399 (1996) [hep-th/9512196].

1626: %%CITATION = HEP-TH 9512196;%%

1627: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9512196}{SPIRES}

1628: 

1629: %\cite{Hanany:1999jy}

1630: \bibitem{HananyKolRajaraman}

1631: A.~Hanany, B.~Kol and A.~Rajaraman, ``Orientifold points in M theory,'' JHEP

1632: {\bf 9910}, 027 (1999) [hep-th/9909028].

1633: %%CITATION = HEP-TH 9909028;%%

1634: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9909028}{SPIRES}

1635: 

1636: %\cite{Sethi:1998zk}

1637: \bibitem{Sethi:1998zk}

1638: S.~Sethi,

1639: ``A relation between N = 8 gauge theories in three dimensions,''

1640: JHEP {\bf 11}, 003 (1998)

1641: hep-th/9809162.

1642: %%CITATION = JHEPA,9811,003;%%

1643: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9809162}{SPIRES}

1644: 

1645: %\cite{Berkooz:1998sn}

1646: \bibitem{BerkoozKapustin}

1647: M.~Berkooz and A.~Kapustin,

1648: ``New IR dualities in supersymmetric gauge theory in three dimensions,''

1649: JHEP {\bf 02}, 009 (1999)

1650: hep-th/9810257.

1651: %%CITATION = JHEPA,9902,009;%%

1652: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9810257}{SPIRES}

1653: 

1654: %\cite{Witten_OM}

1655: \bibitem{Witten_OM}

1656: E.~Witten, ``Five-branes and M-theory on an orbifold,'' Nucl.\

1657: Phys.\  {\bf B463}, 383 (1996) [hep-th/9512219].

1658: %%CITATION = HEP-TH 9512219;%%

1659: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9512219}{SPIRES}

1660: 

1661: %\cite{Hori:1998iv}

1662: \bibitem{Hori}

1663: K.~Hori,

1664: ``Consistency condition for fivebrane in M-theory on R**5/Z(2) orbifold,''

1665: Nucl.\ Phys.\ {\bf B539}, 35 (1999)

1666: hep-th/9805141.

1667: %%CITATION = NUPHA,B539,35;%%

1668: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9805141}{SPIRES}

1669: 

1670: %\cite{Gimon}

1671: \bibitem{Gimon:1998be}

1672: E.G.~Gimon,

1673: ``On the M-theory interpretation of orientifold planes,''

1674: hep-th/9806226.

1675: %%CITATION = HEP-TH 9806226;%%

1676: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9806226}{SPIRES}

1677: 

1678: %\cite{Schwarz:1995dk}

1679: \bibitem{SchwarzSL2Z}

1680: J.~H.~Schwarz, ``An SL(2,Z) multiplet of Type IIB superstrings,''

1681: Phys.\ Lett.\ {\bf B360}, 13 (1995) [hep-th/9508143].

1682: %%CITATION = HEP-TH 9508143;%%

1683: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9508143}{SPIRES}

1684: %\cite{Aspinwall:1996fw}

1685: %\bibitem{Aspinwall:1996fw}

1686: P.~S.~Aspinwall, ``Some relationships between dualities in string theory,''

1687: Nucl.\ Phys.\ Proc.\ Suppl.\  {\bf 46}, 30 (1996) [hep-th/9508154].

1688: %%CITATION = HEP-TH 9508154;%%

1689: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9508154}{SPIRES}

1690: 

1691: %\cite{Kol:1999vi}

1692: \bibitem{Kol5d}

1693: B.~Kol, ``On 6d {``gauge''} theories with irrational theta angle,'' JHEP {\bf

1694: 9911}, 017 (1999) [hep-th/9711017].

1695: %%CITATION = HEP-TH 9711017;%%

1696: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9711017}{SPIRES}

1697: %\cite{Kol:1999fv}

1698: %\bibitem{Kol:1999fv}

1699: B.~Kol,``5d field theories and M theory,'' JHEP {\bf 9911}, 026 (1999)

1700: [hep-th/9705031].

1701: %%CITATION = HEP-TH 9705031;%%

1702: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9705031}{SPIRES}

1703: 

1704: %\cite{Moore:1999gb}

1705: \bibitem{MooreWitten}

1706: G.~Moore and E.~Witten, ``Self-duality, Ramond-Ramond fields, and K-theory,''

1707: hep-th/9912279.

1708: %%CITATION = HEP-TH 9912279;%%

1709: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9912279}{SPIRES}

1710: 

1711: %%%%%%K theory -

1712: %\cite{Minasian:1997mm}

1713: \bibitem{Minasian:1997mm}

1714: R.~Minasian and G.~Moore, ``K-theory and Ramond-Ramond charge,'' JHEP {\bf

1715: 9711}, 002 (1997) [hep-th/9710230].

1716: %%CITATION = HEP-TH 9710230;%%

1717: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9710230}{SPIRES}

1718: 

1719: %\cite{Witten:1998cd}

1720: \bibitem{Witten:1998cd}

1721: E.~Witten, ``D-branes and K-theory,'' JHEP {\bf 9812}, 019 (1998)

1722: [hep-th/9810188].

1723: %%CITATION = HEP-TH 9810188;%%

1724: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9810188}{SPIRES}

1725: 

1726: %\cite{Horava:1999jy}

1727: \bibitem{Horava:1999jy}

1728: P.~Ho\v{r}ava, ``Type IIA D-branes, K-theory, and matrix

1729: theory,'' Adv.\ Theor.\ Math.\ Phys.\  {\bf 2}, 1373

1730: (1999) [hep-th/9812135].

1731: %%CITATION = HEP-TH 9812135;%%

1732: \href{\wwwspires?eprint=HEP-TH/9812135}{SPIRES}

1733: 

1734: %\cite{Bergman:1999ta}

1735: \bibitem{Bergman:1999ta}

1736: O.~Bergman, E.~G.~Gimon and P.~Ho\v{r}ava, ``Brane transfer

1737: operations and T-duality of non-BPS states,'' JHEP {\bf 9904}, 010

1738: (1999) [hep-th/9902160].

1739: %%CITATION = HEP-TH 9902160;%%

1740: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9902160}{SPIRES}

1741: 

1742: %\cite{Gukov:1999yn}

1743: \bibitem{Gukov:1999yn}

1744: S.~Gukov, ``K-theory, reality, and orientifolds,'' hep-th/9901042.

1745: %%CITATION = HEP-TH 9901042;%%

1746: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9901042}{SPIRES}

1747: 

1748: %\cite{Hori:1999me}

1749: \bibitem{Hori:1999me}

1750: K.~Hori, ``D-branes, T-duality, and index theory,''

1751: hep-th/9902102.

1752: %%CITATION = HEP-TH 9902102;%%

1753: \href{\wwwspires?eprint=HEP-TH/9902102}{SPIRES}

1754: %%%%%end K theory

1755: 

1756: %\cite{Dabholkar:1996pc}

1757: \bibitem{DabholkarPark}

1758: A.~Dabholkar and J.~Park, ``Strings on Orientifolds,''

1759: Nucl.\ Phys.\  {\bf B477}, 701 (1996) [hep-th/9604178].

1760: %%CITATION = HEP-TH 9604178;%%

1761: \href{\wwwspires?eprint=HEP-TH/9604178}{SPIRES}

1762: 

1763: %\cite{Uranga:1998uj}

1764: \bibitem{Uranga}

1765: A.~M.~Uranga, ``Towards mass deformed N = 4 SO(n) and Sp(k)

1766: theories from brane  configurations,'' Nucl.\ Phys.\  {\bf B526},

1767: 241 (1998) [hep-th/9803054].

1768: %%CITATION = HEP-TH 9803054;%%

1769: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9803054}{SPIRES}

1770: 

1771: \bibitem{WittenAdSBaryons}

1772: E.~Witten, ``Baryons and branes in anti de Sitter space,'' JHEP {\bf 9807}, 006

1773: (1998) [hep-th/9805112].

1774: %%CITATION = HEP-TH 9805112;%%

1775: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9805112}{SPIRES}

1776: 

1777: %\cite{Witten:1998kz}

1778: \bibitem{Witten6d}

1779: E.~Witten, ``New *gauge* theories in six dimensions,''

1780: JHEP {\bf 9801}, 001 (1998) [hep-th/9710065].

1781: %%CITATION = HEP-TH 9710065;%%

1782: \href{\wwwspires?eprint=HEP-TH/9710065}{SPIRES}

1783: 

1784: \bibitem{Dijk}

1785: J.~de Boer, R.~Dijkgraaf, A.~Keurentjes, J.~Morgan,

1786: D.~Morrison, S.~Sethi as reported in the talk of R.

1787: Dijkgraaf at "String Theory at The Millennium" conference

1788: \href{http://quark.theory.caltech.edu/people/rahmfeld/Dijkgraaf/fs1.html}

1789: {http://quark.theory.caltech.edu/people/rahmfeld/Dijkgraaf/fs1.html},

1790: p.19-26.

1791: 

1792: %\cite{Hanany:1999sj}

1793: \bibitem{HZ}

1794: A.~Hanany and A.~Zaffaroni, ``Issues on orientifolds: On the brane

1795: construction of gauge theories  with SO(2n) global symmetry,''

1796: JHEP {\bf 9907}, 009 (1999) [hep-th/9903242].

1797: %%CITATION = HEP-TH 9903242;%%

1798: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9903242}{SPIRES}

1799: 

1800: %\cite{Evans:1997hk}

1801: \bibitem{EJS}

1802: N.~Evans, C.~V.~Johnson and A.~D.~Shapere, ``Orientifolds, branes, and duality

1803: of 4D gauge theories,'' Nucl.\ Phys.\  {\bf B505}, 251 (1997) [hep-th/9703210].

1804: %%CITATION = HEP-TH 9703210;%%

1805: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9703210}{SPIRES}

1806: 

1807: %\cite{Elitzur:1998ju}

1808: \bibitem{Tsabar}

1809: S.~Elitzur, A.~Giveon, D.~Kutasov and D.~Tsabar, ``Branes,

1810: orientifolds and chiral gauge theories,'' Nucl.\ Phys.\  {\bf

1811: B524}, 251 (1998) [hep-th/9801020].

1812: %%CITATION = HEP-TH 9801020;%%

1813: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9801020}{SPIRES}

1814: 

1815: %\cite{Mukhi:1999hs}

1816: \bibitem{Mukhi3}

1817: S.~Mukhi and N.~V.~Suryanarayana, ``Gravitational

1818: couplings, orientifolds and M-planes,'' JHEP {\bf 9909},

1819: 017 (1999) [hep-th/9907215].

1820: %%CITATION = HEP-TH 9907215;%%

1821: \href{\wwwspires?eprint=HEP-TH/9907215}{SPIRES}

1822: 

1823: %\cite{Berlin}

1824: \bibitem{Berlin}

1825: I.~Brunner, A.~Hanany, A.~Karch and D.~Lust, ``Brane

1826: dynamics and chiral non-chiral transitions,'' Nucl.\

1827: Phys.\  {\bf B528}, 197 (1998) [hep-th/9801017].

1828: %%CITATION = HEP-TH 9801017;%%

1829: %\href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9801017}{SPIRES}

1830: 

1831: %\cite{Horava:1996ma}

1832: \bibitem{HoravaWitten}

1833: P.~Ho\v{r}ava and E.~Witten, ``Eleven-Dimensional Supergravity on

1834: a Manifold with Boundary,'' Nucl.\ Phys.\  {\bf B475}, 94 (1996)

1835: [hep-th/9603142].

1836: %%CITATION = HEP-TH 9603142;%%

1837: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9603142}{SPIRES}

1838: 

1839: %\cite{Witten:1996md}

1840: \bibitem{WittenFluxQuant}

1841: E.~Witten,

1842: ``On flux quantization in M-theory and the effective action,''

1843: J.\ Geom.\ Phys.\ {\bf 22}, 1 (1997)

1844: hep-th/9609122.

1845: %%CITATION = JGPHE,22,1;%%

1846: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9609122}{SPIRES}

1847: 

1848: %\cite{Ahn:ads7}

1849: \bibitem{Ahn:ads7}

1850: C.~Ahn, H.~Kim and H.~S.~Yang, ``SO(2N)(0,2) SCFT and M

1851: theory on AdS(7) x RP(4),'' Phys.\ Rev.\  {\bf D59},

1852: 106002 (1999) [hep-th/9808182].

1853: %%CITATION = HEP-TH 9808182;%%

1854: \href{\wwwspires?eprint=HEP-TH/9808182}{SPIRES}

1855: 

1856: %\cite{Ahn:ads4}

1857: \bibitem{Ahn:ads4}

1858: C.~Ahn, H.~Kim, B.~Lee and H.~S.~Yang, ``N = 8 SCFT and M

1859: theory on AdS(4) x RP(7),'' Phys.\ Rev.\  {\bf D61},

1860: 066002 (2000) [hep-th/9811010].

1861: %%CITATION = HEP-TH 9811010;%%

1862: \href{\wwwspires?eprint=HEP-TH/9811010}{SPIRES}

1863: 

1864: %\cite{Aharony:1996xr}

1865: \bibitem{ASY}

1866: O.~Aharony, J.~Sonnenschein and S.~Yankielowicz,

1867: ``Interactions of strings and D-branes from M theory,''

1868: Nucl.\ Phys.\  {\bf B474}, 309 (1996) [hep-th/9603009].

1869: %%CITATION = HEP-TH 9603009;%%

1870: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9603009}{SPIRES}

1871: 

1872: \bibitem{rey}

1873: N.~Kim and S.~Rey,

1874: ``M(atrix) theory on T(9)/Z(2) orbifold and twisted zero-branes,''

1875: hep-th/9710245.

1876: %%CITATION = HEP-TH 9710245;%%

1877: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9710245}{SPIRES}

1878: 

1879: %\cite{Kac:1999av}

1880: \bibitem{KacSmilga}

1881: V.~G.~Kac and A.~V.~Smilga, ``Normalized vacuum states in N = 4

1882: supersymmetric Yang-Mills quantum  mechanics with any gauge

1883: group,'' hep-th/9908096.

1884: %%CITATION = HEP-TH 9908096;%%

1885: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9908096}{SPIRES}

1886: 

1887: %\cite{Dasgupta:1997yh}

1888: \bibitem{DasguptaMukhi2}

1889: K.~Dasgupta and S.~Mukhi, ``A note on low-dimensional

1890: string compactifications,'' Phys.\ Lett.\  {\bf B398}, 285

1891: (1997) [hep-th/9612188].

1892: %%CITATION = HEP-TH 9612188;%%

1893: \href{\wwwspires?eprint=HEP-TH/9612188}{SPIRES}

1894: 

1895: %\cite{Dixon:1985jw}

1896: \bibitem{orbifolds}

1897: L.~Dixon, J.~A.~Harvey, C.~Vafa and E.~Witten, ``Strings

1898: On Orbifolds,'' Nucl.\ Phys.\  {\bf B261}, 678 (1985).

1899: %%CITATION = NUPHA,B261,678;%%

1900: %\href{http://www.slac.stanford.edu/spires/find/hep/www?j=NUPHA%2cB261%2c678}{SPIRES}

1901: %\cite{Dixon:1986jc}

1902: %\bibitem{Dixon:1986jc}

1903: L.~Dixon, J.~A.~Harvey, C.~Vafa and E.~Witten, ``Strings

1904: On Orbifolds. 2,'' Nucl.\ Phys.\  {\bf B274}, 285 (1986).

1905: %%CITATION = NUPHA,B274,285;%%

1906: %\href{http://www.slac.stanford.edu/spires/find/hep/www?j=NUPHA%2cB274%2c285}{SPIRES}

1907: 

1908: %\cite{Sen:1996na}

1909: \bibitem{Sen:1996na}

1910: A.~Sen, ``Duality and Orbifolds,'' Nucl.\ Phys.\  {\bf B474}, 361 (1996)

1911: [hep-th/9604070].

1912: %%CITATION = HEP-TH 9604070;%%

1913: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9604070}{SPIRES}

1914: 

1915: %\cite{Sen:1998ii}

1916: \bibitem{Sen}

1917: A.~Sen, ``Stable non-BPS bound states of BPS D-branes,'' JHEP {\bf

1918: 9808}, 010 (1998) [hep-th/9805019].

1919: %%CITATION = HEP-TH 9805019;%%

1920: \href{http://www.slac.stanford.edu/spires/find/hep/www?eprint=HEP-TH/9805019}{SPIRES}

1921: 

1922: \end{thebibliography}

1923: 

1924: \end{document}

1925: