1: %\documentstyle[12pt,epsfig]{article}

2: \documentclass[12pt,dvips]{article}

3: \usepackage{graphicx}

4: %\usepackage{epsfig}

5: %\headheight 0.0 in

6: % \headsep 0in

7: %\textwidth 6.25 in

8: % \textheight 9 in

9: %\oddsidemargin 0.125 in

10: %\evensidemargin 0.125 in

11: %\pagestyle{plain}

12: \pagenumbering{arabic}

13: \usepackage{amsmath}

14: \usepackage[psamsfonts]{amssymb}

15: \usepackage{amsthm}

16: \usepackage{euscript}

17: 

18: \usepackage{latexsym}

19: %\pagestyle{headings}

20: %\setlength{\parindent}{0em}

21: 

22: \setlength{\topmargin}{0in}

23: \setlength{\headheight}{0in}

24: \setlength{\headsep}{0in}

25: \setlength{\topskip}{.6in}

26: \setlength{\textheight}{8.8in}

27: \setlength{\oddsidemargin}{0in}

28: \setlength{\evensidemargin}{0in}

29: \setlength{\textwidth}{6.5in}

30: %\setlength{\mathindent}{.3in}

31: \renewcommand{\theequation}{\arabic{section}.\arabic{equation}}

32: \renewcommand{\(}{\begin{equation}}

33: \renewcommand{\)}{end{equation} \vspace{-.05in}\linebreak}

34: \newcommand{\numeq}{\renewcommand{\theequation}{\arabic{section}.\arabic{equation}}}

35: \newcommand{\nonumeq}{\renewcommand{\theequation}{}}

36: \newcounter{saveeqn}

37: \newcounter{savealpheqn}

38: \newcounter{savesection}

39: \newcommand{\alpheqn}{\setcounter{saveeqn}{\value{equation}}%

40:  \stepcounter{saveeqn}\setcounter{equation}{0}%

41:  \renewcommand{\theequation}{\mbox{\arabic{section}.\arabic{saveeqn}\alph{equation}}}

42:  \renewcommand{\)}{\end{equation}}}

43: \def\part#1{\frac{\partial}{\partial{#1}}}%

44: \def\group#1{\refstepcounter{equation}\setcounter{saveeqn}{\value{equation}}%

45:  \label{#1}\setcounter{equation}{0}%

46:  \renewcommand{\theequation}{\mbox{\arabic{section}.\arabic{saveeqn}\alph{equation}}}

47:  \renewcommand{\)}{\end{equation}}}

48: \newcommand{\reseteqn}{\setcounter{equation}{\value{saveeqn}}%

49:  \renewcommand{\theequation}{\arabic{section}.\arabic{equation}}%

50:  \renewcommand{\)}{\end{equation}}}

51: \newcounter{alphcount}

52: %\def\getletter#1{\renewcommand{\theequation}{\alph{equation}}%

53: %                 \setcounter{alphcount}{\value{equation}}

54: %                 \refstepcounter{equation}%

55: %                 %\begin{eqnarray}%

56: %                 \label{#1}%

57: %                 \setcounter{equation}{\value{alphcount}}%

58: %                 %\nonumber\end{eqnarray}\vspace{-.666in}%

59: %                 \renewcommand{\theequation}{\mbox{\arabic{section}.\arabic{saveeqn}\alph{equation}}}}

60: %%\newcommand{\writeletters}{}

61: %\def\writeletter#1{\renewcommand{\theequation}{\alph{#1}}%

62: %                 \begin{eqnarray}%

63: %                 \label{#1}%

64: %                 \nonumber\end{eqnarray}\vspace{-.666in}}

65: %\def\getlette2r#1{\newcounter{#1}%

66: %                 \setcounter{#1}{\value{equation}}%

67: %                 \providecommand{\writeletters}{\writeletters\writeletter{#1}}}

68: %\def\agetletter#1{\renewcommand{\theequation}{\alph{equation}}%

69: %                 \begin{eqnarray}%

70: %                 \label{#1}%

71: %                 \nonumber\end{eqnarray}\vspace{-.666in}%

72: %                 \renewcommand{\theequation}{\mbox{\Alph{subsection}.\arabic{saveeqn}\alph{equation}}}}

73: \newcommand{\aalpheqn}{\setcounter{saveeqn}{\value{equation}}%

74:  \stepcounter{saveeqn}\setcounter{equation}{0}%

75:  \renewcommand{\theequation}{\mbox{\Alph{subsection}.\arabic{saveeqn}\alph{equation}}}

76:   \renewcommand{\)}{\end{equation}}}

77: \newcommand{\areseteqn}{\setcounter{equation}{\value{saveeqn}}%

78:  \renewcommand{\theequation}{\Alph{subsection}.\arabic{equation}}%

79:  \renewcommand{\)}{\end{equation}}}

80: \newcommand{\unalph}{\setcounter{savealpheqn}{\value{equation}}%

81:  \renewcommand{\theequation}{\mbox{\arabic{section}.\arabic{saveeqn}}}}

82: \newcommand{\realph}{\setcounter{equation}{\value{savealpheqn}}}

83: \renewcommand{\=}{\hspace{-.03in}=\hspace{-.02in}}

84: %\newcommand{\+}{\hspace{-.03in}+\hspace{-.02in}}

85: \renewcommand{\thefootnote}{\alph{footnote}}

86: \renewcommand{\(}{\begin{equation}}

87: \renewcommand{\)}{\end{equation}}

88: \newcommand{\ba}{\begin{eqnarray}}

89: \newcommand{\ea}{\end{eqnarray}}

90: \renewcommand{\l}{\lambda}

91: \renewcommand{\a}{\alpha}

92: \renewcommand{\b}{\beta}

93: \renewcommand{\r}{\rho}

94: \newcommand{\bp}{\mathop{\vtop{\ialign{##\crcr

95:   $\hfil\displaystyle{}\hfil$\crcr\noalign{\kern-13pt\nointerlineskip}

96:   \BIG{(}\hskip0pt\crcr\noalign{\kern3pt}}}}}

97: \newcommand{\cbp}{\mathop{\vtop{\ialign{##\crcr

98:   $\hfil\displaystyle{}\hfil$\crcr\noalign{\kern-13pt\nointerlineskip}

99:   \BIG{)}\hskip0pt\crcr\noalign{\kern3pt}}}}}

100: \newcommand{\pa}{\mathop{\vtop{\ialign{##\crcr

101:   $\hfil\displaystyle{\oplus}\hfil$\crcr\noalign{\kern+1pt\nointerlineskip}

102:   \hspace{.08in}$^{\alpha=0}$\hskip6pt\crcr\noalign{\kern3pt}}}}}

103: \renewcommand{\sp}{,\hspace{.3in}}

104: %\newcommand{\newsection}{\setcounter{equation}{0}\section}

105: \newcommand{\p}{^\prime}

106: \newcommand{\pp}{^{\prime\prime}}

107: \newcommand{\w}{\omega}

108: %\newcommand{\mod}{{\textup{\scriptsize{ mod }}}}

109: \newcommand{\rank}{{\textup{\scriptsize{rank}}}}

110: \newcommand{\rrank}{{\textup{rank}}}

111: \newcommand{\rmod}{{\textup{ mod }}}

112: \newcommand{\appendixa}

113:  {\renewcommand{\theequation}{\Alph{subsection}.\arabic{equation}}%

114:   \renewcommand{\thesubsection}%

115:                {Appendix \Alph{subsection}.\setcounter{equation}{0}}%

116:   \renewcommand{\alpheqn}{\aalpheqn}%

117:   \renewcommand{\reseteqn}{\areseteqn}

118:   \newcounter{savesec}}

119: \newcommand{\appendices}{\appendix\appendixa}

120: \newcommand{\R}{\ensuremath{\mathbb R}}

121: \newcommand{\Rn}{\ensuremath{{\mathbb R}^n}}

122: \newcommand{\C}{\ensuremath{\mathbb C}}

123: \newcommand{\Cn}{\ensuremath{{\mathbb C}^n}}

124: \newcommand{\PR}{\ensuremath{\mathbb P}}

125: \newcommand{\PRn}{\ensuremath{{\mathbb P}^n}}

126: \newcommand{\Zn}{\ensuremath{\mathbb Z}^n}

127: \newcommand{\Q}{\ensuremath{\mathbb Q}}

128: \newcommand{\T}{\ensuremath{\mathbb T}}

129: \newcommand{\Tn}{\ensuremath{{\mathbb T}^n}}

130: \newcommand{\Z}{\ensuremath{\mathbb Z}}

131: \newcommand{\ci}{\ensuremath{{C^\infty}}}

132: \def\H{\ensuremath{\ES{H}}}

133: \def\i{\ensuremath{\dot\imath}}

134: \def\S{\ensuremath{\ES{S}}}

135: \def\L{\ensuremath{{\cal L}}}

136: \def\D{\ensuremath{{\cal D}}}

137: \def\O{\ensuremath{{\cal O}}}

138: \newcommand{\del}{\ensuremath{\partial}}

139: \newcommand{\Del}{\ensuremath{\nabla}}

140: \newcommand{\eps}{\ensuremath{\varepsilon}}

141: \newcommand{\half}{\ensuremath{\frac{1}{2}}}

142: \newcommand{\quart}{\ensuremath{\frac{1}{4}}}

143: \newcommand{\N}{\ensuremath{\mathcal N}}

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

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

146: \newcommand{\sub}{\subset}

147: \newcommand{\osub}{\overset{\circ}{\subset}}

148: \newcommand{\tensor}{\otimes}

149: \newcommand{\xtensor}{\boxtimes}

150: \newcommand{\contract}{\lrcorner}

151: \newcommand{\iso}{\cong}

152: \newcommand{\heq}{\simeq}

153: \newcommand{\oo}{\ensuremath{\infty}}

154: \newcommand{\iseq}{_{i=1}^{\infty}}

155: \newcommand{\onto}{\twoheadrightarrow}

156: \newcommand{\monoto}{\rightarrowtail}

157: \newcommand{\into}{\hookrightarrow}

158: 

159: \newcommand{\pderiv}[1]{\ensuremath{\frac{\del}{\del #1}}}

160: \newcommand{\pder}[2]{\ensuremath{\frac{\del #1}{\del #2}}}

161: \newcommand{\abs}[1]{\ensuremath{\lvert#1\rvert}}

162: \newcommand{\norm}[1]{\ensuremath{\lVert#1\rVert}}

163: \newcommand{\ip}[1]{\ensuremath{\langle#1\rangle}}

164: \newcommand{\op}[1]{\ensuremath{\operatorname{#1}}}

165: \newcommand{\limop}[1]{\operatornamewithlimits{#1}}

166: \newcommand{\bb}[1]{\ensuremath{\mathbb{#1}}}

167: \newcommand{\ES}[1]{\ensuremath{\EuScript{#1}}}

168: \newcommand{\EF}[1]{\ensuremath{\mathfrak{#1}}}

169: \newcommand{\expi}[1]{e^{-2\pi\i #1}}

170: 

171: %Specific newcommands

172: \newcommand{\cliff}{\ensuremath{\op{Cliff}(V)}}

173: \newcommand{\cliffc}{\ensuremath{\op{Cliff}_{\C}(V)}}

174: \newcommand{\End}{\ensuremath{\op{End}}}

175: 

176: %\theoremstyle{plain} \newtheorem{thm}{Theorem}

177: %\theoremstyle{plain} \newtheorem{lemma}[thm]{Lemma}

178: %\newtheorem{corollary}[thm]{Corollary}

179: %\newtheorem{proposition}[thm]{Proposition}

180: %\newtheorem{defn}[thm]{Definition}

181: 

182: \newcommand{\remark}{\noindent {\bf Remark.\ }}

183: \newcommand{\convention}{\noindent {\bf Convention.\ }}

184: \newcommand{\theproof}{\noindent {\bf Proof.\ }}

185: \newcommand{\sketchproof}{\noindent {\bf Sketch of Proof.\ }}

186: \newcommand{\done}{$\hfill \hfill \blacksquare$ \bigskip}

187: 

188: \numberwithin{equation}{section}

189: %\numberwithin{thm}{section}

190: 

191: %\def\app#1#2{\renewcommand{\thesubsection}{\Alph{subsection}}%

192: %        \refstepcounter{subsection}%

193: %        \setcounter{subsection}{\value{savesec}}%

194: %        \stepcounter{savesec}\label{#1}%

195: %        \renewcommand{\thesubsection}%

196: %               {Appendix \Alph{subsection}.\setcounter{equation}{0}}%

197: %        \subsection{#2}}

198: %\def\big#1{\mbox{\Large $#1$}}

199: %\def\BIG#1{\mbox{\Huge $#1$}}

200: %\def\gb            {\mbox{$\mathfrak g$}}

201: %\def\hb            {\mbox{$\mathfrak h$}}

202: %\def\ssz           {\mbox{\tiny $\mathbb  Z$}}

203: %\def\sz            {\mbox{\scriptsize $\mathbb  Z$}}

204: %\def\R             {\textup{R}\hsp{-.104}\textup{I}\hsp{.03}}

205: %\def\H             {\textup{H}\hsp{-.104}\textup{I}\hsp{.03}}

206: %\def\zl            {\mbox{$\mathbb  Z_\l$}}

207: %\def\subsecz       {\mbox{\large $\mathbb  Z$}}

208: %\def\mz            {\mbox{\Large $\mathbb  Z$}}

209: %\def\mzl           {\mbox{\Large $\mathbb  Z_\l$}}

210: %\def\omzl          {\mbox{\Large /$\mathbb  Z_\l$}}

211: %\def\lz            {\mbox{\Huge $\mathbb  Z$}}

212: %\def\d             {\mbox{$\mathbb D$}}

213: %\def\dl            {\mbox{$\mathbb D_\l$}}

214: %\def\subsecd       {\mbox{\large $\mathbb  D$}}

215: %\def\md            {\mbox{\Large $\mathbb  D$}}

216: 

217: 

218: \def\journal{\topmargin .5in    \oddsidemargin .5in

219:         \headheight 0pt \headsep 0pt

220:         \textwidth 5.625in % 1.2 preprint size  %6.5in

221:         \textheight 8.25in % 1.2 preprint size 9in

222:         \marginparwidth 1.5in

223:         \parindent 2em

224:         \parskip .5ex plus .1ex         \jot = 1.5ex}

225: %

226: %       The default is set to be journal!

227: %\journal

228: \def\baselinestretch{1.2}

229: %\mathchardef\endbar="375

230: %\font\fivesans=cmss10 at 4.61pt

231: %\font\sevensans=cmss10 at 6.81pt

232: %\font\tensans=cmss10 at 12pt %added ``at 12pt''

233: %\newfam\sansfam

234: %\textfont\sansfam=\tensans\scriptfont\sansfam=\sevensans\scriptscriptfont

235: %\sansfam=\fivesans

236: %\def\sans{\fam\sansfam\tensans}

237: %\def\Z{{\mathchoice

238: %{\hbox{$\sans\textstyle Z\kern-0.455em Z$}} %was .4

239: %{\hbox{$\sans\textstyle Z\kern-0.455em Z$}} %was .4

240: %{\hbox{$\sans\scriptstyle Z\kern-0.355em Z$}} %was .3

241: %{\hbox{$\sans\scriptscriptstyle Z\kern-0.255em Z$}}}} %was .2

242: 

243: 

244: %\font\tensans=cmss10 at 14pt 

245: %\newfam\sansfamb

246: %\textfont\sansfamb=\tensans\scriptfont\sansfamb=\sevensans\scriptscriptfont

247: %\sansfamb=\fivesans

248: %\def\sansb{\fam\sansfamb\tensans}

249: %\font\tensans=cmss10 at 17pt

250: %\newfam\sansfamc

251: %\textfont\sansfamc=\tensans\scriptfont\sansfamc=\sevensans\scriptscriptfont

252: %\sansfamc=\fivesans

253: %\def\sansc{\fam\sansfamc\tensans}

254: %\def\contr#1#2{\mathop{\vtop{\ialign{##\crcr

255: %  $\hfil\displaystyle{#2}\hfil$\crcr\noalign{\kern3pt\nointerlineskip}

256: %  \hspace{.09in}\rule[0in]{.01in}{.1in}\rule[0in]{#1in}{.01in}\rule[0in]{.01in}{.1in}\hskip6pt\crcr\noalign{\kern3pt}}}}}

257: %\def\contrb#1#2#3{\mathop{\vtop{\ialign{##\crcr

258: %  $\hfil\displaystyle{#3}\hfil$\crcr\noalign{\kern3pt\nointerlineskip}

259: %  \hspace{#1in}\rule[0in]{.01in}{.1in}\rule[0in]{#2in}{.01in}\rule[0in]{.01in}{.1in}\hskip6pt\crcr\noalign{\kern3pt}}}}}

260: %\def\namegroup#1{\begin{eqnarray}\label{#1}\nonumber\end{eqnarray}\vspace{-.5in}}

261: \def\hsp#1{\hspace{#1in}}

262: \def\rf#1{\ref{ref#1}}

263: \def\comment#1{\hsp{.3}\textup{#1}}

264: 

265: \def\lsim{\mathrel{\mathpalette\vereq\langle}}

266: \def\gsim{\mathrel{\mathpalette\vereq\rangle}}

267: \catcode`\@=11

268: \def\vereq#1#2{\lower3pt\vbox{\baselineskip1.5pt \lineskip1.5pt

269: \ialign{$\m@th#1\hfill##\hfil$\crcr#2\crcr\sim\crcr}}}

270: \catcode`\@=12

271: 

272: \renewcommand{\textfraction}{0.15}

273: \renewcommand{\topfraction}{0.85}

274: \renewcommand{\bottomfraction}{0.65}

275: \renewcommand{\floatpagefraction}{0.60}

276: \makeatletter

277: \newcommand\figcaption{\def\@captype{figure}\caption}

278: \newcommand\tabcaption{\def\@captype{table}\caption}

279: \makeatother

280: 

281: \begin{document}

282: \begin{titlepage}

283: \begin{flushright}

284: UCB-PTH-01/24 \\

285: hep-th/0107072

286: \end{flushright}

287: 

288: \vspace{2em}

289: \def\thefootnote{\fnsymbol{footnote}}

290: 

291: \begin{center}

292: {\Huge Dial M for Flavor Symmetry Breaking}

293: \end{center}

294: \vspace{1em}

295: 

296: \begin{center}

297: Jarah Evslin\footnote{E-Mail: jarah@uclink4.berkeley.edu}, Hitoshi Murayama\footnote{E-Mail: murayama@lbl.gov}, Uday Varadarajan\footnote{E-Mail: udayv@socrates.berkeley.edu} and John E. Wang\footnote{E-Mail:

298:   hllywd2@physics.berkeley.edu}

299: \end{center}

300: 

301: \begin{center}

302: \vspace{1em}

303: {\em Department of Physics,

304:      University of California\\

305:      Berkeley, California 94720}\\

306: and\\

307: {\em Theoretical Physics Group\\

308:      Ernest Orlando Lawrence Berkeley National Laboratory\\

309:      University of California,

310:      Berkeley, California 94720}        

311: 

312: \end{center}

313: 

314: \vspace{3em}

315: \begin{abstract}

316: 

317: \noindent

318: We introduce a realization of Seiberg duality in MQCD that does not

319: involve moving branes. Using related arguments in M Theory and IIA,

320: we investigate the flavor symmetry breaking via flavored magnetic

321: monopole condensation of hep-th/0005076. We verify our results by

322: considering visualizations of M5 branes corresponding to $\N=2$ vacua

323: which survive soft breaking by an adjoint mass term.

324: 

325: \end{abstract}

326: 

327: \vfill

328: August 8, 2001

329: 

330: \end{titlepage}

331: \setcounter{footnote}{0} 

332: \renewcommand{\thefootnote}{\arabic{footnote}}

333: 

334: 

335: \pagebreak

336: \renewcommand{\thepage}{\arabic{page}}

337: %\tableofcontents

338: \pagebreak 

339: 

340: \section{Introduction}

341: 

342: Four years ago we learned that superQCD is contained in the low energy

343: description of certain configurations of M5 branes in multi-centered

344: Taub-NUT space \cite{Witten}.  As a result, brane configurations in M

345: theory are able to provide a concrete realization of otherwise

346: mysterious quantum phenomena, such as confinement via magnetic

347: monopole condensation \cite{SeibergWitten1,Witten2,Rabinovici} and

348: Douglas-Shenker strings \cite{DS,HananyZaffaroni}.  This approach is

349: known as MQCD and is reviewed in \cite{Giveon}. 

350: 

351: In this paper, we introduce a novel MQCD description of Seiberg

352: duality in $\N=2$ superQCD softly broken to $\N=1$ by adding a

353: mass term for the adjoint chiral multiplets. We then proceed to

354: construct the M theory realization of the U($N_f$) $\rightarrow$

355: U($r$)$\times$U($N_f-r$) flavor symmetry breaking mechanism in $\N=1$

356: SU($N_c$) gauge theories studied in Refs.~\cite{Murayama1, Murayama,

357:   Murayamatalk}.

358: 

359: In section~\ref{revsec}, we will review the semi-classical description

360: of supersymmetric gauge theories as effective field theories of

361: parallel D4 branes suspended between NS5 branes in type IIA string

362: theory.  To understand the quantum gauge theories we lift this

363: description to M theory.  We then review the field theory results of

364: Refs.~\cite{Murayama1, Murayama, Murayamatalk} in section 3.  Then, in

365: section 4 we describe Seiberg duality as a choice of two D brane

366: configurations in IIA which lift to the same M5 brane configuration as

367: seen by M2 brane probes at different energy scales.  We finally turn

368: to the MQCD description of flavor symmetry breaking in section 5. A

369: detailed example of the M5 brane configurations corresponding to

370: r-vacua in SU(3) with 4 flavors is presented in Appendix A.

371: 

372: \section{A Review of MQCD} \label{revsec}

373: 

374: We begin with a review of the standard embedding of $3+1$ dimensional

375: $SU(N_c)$ SQCD in IIA string theory and its M theory lift.

376: 

377: \subsection{Classical $\mathcal{N}=2$ SQCD and IIA}

378: 

379: \begin{figure}[ht]

380:   \centering \includegraphics[width=6in]{basic.eps}

381: \caption{Type IIA string theory realization of $\N=2$ SU(4) SQCD with

382:   3 flavors. The effective worldvolume of the gauge theory (which lives on the

383:   D4 color branes) is spanned by the $x^0$, $x^1$, $x^2$, and $x^3$

384:   directions.}

385:   \label{basic}

386: \end{figure}

387: 

388: Consider type IIA string theory on $\R^{9,1}$ with coordinates $x^0,

389: \ldots, x^9$ and complex coordinates

390: \begin{equation}

391: v=x^4+i x^5\sp w=x^8+ix^9. 

392: \end{equation}

393: A stack of $N_c$ parallel D4 ``color'' branes extend along directions

394: $x^0$, $x^1$, $x^2$, $x^3$, and $x^6$.  The low energy theory on these

395: branes is SU($N_c$) 4+1 dimensional super Yang-Mills with 16

396: supercharges.  We can Kaluza-Klein reduce this to the desired 3+1

397: dimensional SYM with 8 supercharges by adding two parallel NS5 branes

398: extended along the $x^0$, $x^1$, $x^2$, $x^3$, $x^4$, and $x^5$

399: directions and placed at positions $0$ and $L_6$ along the $x^6$

400: direction with the color branes suspended between them.

401: 

402: %\begin{figure}[ht]

403: %\begin{center}

404: %\epsfxsize=6in\leavevmode\epsfbox{3color.eps}

405: %\caption{IIA realization of supersymmetric SU(3) gauge theory.  The three D4 b%ranes are stacked on top of each other} \label{3color}

406: %\end{center}

407: %\end{figure}

408: The effective gauge coupling, $g$, of the 3+1 dimensional theory is given by 

409: \begin{equation}

410: \frac{1}{g^2}=\frac{L_6}{g_s l_s}

411: \end{equation}

412: where $g_s$, $l_s$, and $L_6$ are the string coupling constant, the

413: string length and the distance between the two NS5 branes.  To

414: decouple the degrees of freedom of the bulk from the

415: color branes we take the limits

416: \begin{equation}

417: g_s \ \rightarrow \ 0 \sp \frac{L_6}{l_s} \ \rightarrow \ 0 \sp g=\textup{constant}.

418: \end{equation}

419: The light perturbative degrees of freedom of the gauge theory are

420: strings which stretch between the color branes, yielding an

421: $\mathcal{N}=2$ vector multiplet transforming in the adjoint

422: representation of SU($N_c$).  The distances between color branes

423: correspond to the vacuum expectation values of the adjoint scalars in

424: the vector multiplet and so parameterize the Coulomb branch. 

425: 

426: %Magnetic monopoles are hypermultiplets transforming in the

427: %fundamental representation of the gauge theory and whose mass which

428: %is proportional to the area enclosed by two D4 branes and the two NS5

429: %branes 

430: %\begin{equation}

431: %m_m \ \sim \phi \times L_6 \ \sim \ \frac{ \phi}{g^2} \ .

432: %\end{equation}

433: %For small values of the gauge coupling, the magnetic monopoles are

434: %heavy relative to the W bosons and classically are not important in

435: %the infrared. 

436: 

437: %\begin{figure}[ht]

438: %\begin{center}

439: %\epsfxsize=6in\leavevmode\epsfbox{color.eps}

440: %\caption{IIA realization of spontaneously broken SU(3) gauge theory.} \label{color}

441: %\end{center}

442: %\end{figure}

443: 

444: Quark hypermultiplets transforming in the fundamental representation

445: of SU($N_c$) may be included by attaching $N_f$ D4 ``flavor'' branes

446: stretching between one of the NS5 branes and a D6 flavor brane.  To

447: preserve $\N=2$ supersymmetry no two D4 branes may connect the same

448: NS5 and D6 brane.  This is known as the s-rule {\cite{HananyWitten}}

449: and is U-dual {\cite{Bachas1, Bachas2}} to Pauli's exclusion

450: principle. The quarks and squarks are strings which stretch from one

451: D4 flavor brane to one D4 color brane and so they transform in the

452: fundamental representations of both the SU($N_c$) gauge group and the

453: global flavor symmetry group. Semiclassical magnetic monopole and dyon

454: {\cite{Witten, HY, Russian}} states are realized by D2 branes with the

455: topology of a disk bounded by a D4-NS5-D4-NS5 cycle. The brane

456: configuration is summarized in Table~\ref{bral}.

457: 

458: \begin{figure}[h]

459: \centering

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

461: Brane & $x^0$ & $x^1$ & $x^2$ & $x^3$ & $x^4$ & $x^5$ & $x^6$ & $x^7$

462: & $x^8$ & $x^9$ \\ \hline

463: D4           & X & X & X & X &     &     & X &    &     &     \\

464: NS5          & X & X & X & X &  X  &  X  &   &    &     &     \\

465: %NS5$\p$     & X & X & X & X &     &     &   &    &  X  &  X  \\

466: NS5$_\theta$ & X & X & X & X & (X) & (X) &   &    & (X) & (X) \\

467: D6           & X & X & X & X &     &     &   & X  &  X  &  X  \\

468: D2           & X &   &   &   & (X) & (X) & X &    &     &     \\

469: D0           & (X) &   &   &   &     &     & (X) &    &     &     \\

470: \multicolumn{10}{c}{}\\

471: \end{tabular}

472: \tabcaption{Alignments of branes in IIA. Parentheses indicate that the

473: brane may be aligned at an angle between the given directions. For the

474: D0 branes they are used to indicate the presence of both

475: D-instantons and dynamical D0 particles.} \label{bral}

476: \end{figure}

477: 

478: There is an unbroken global U($N_f$) symmetry when the flavor branes

479: are placed at the same $v$ coordinate, although this symmetry is

480: broken to U(1$)^{N_f}$ when they are placed at distinct positions

481: $v=m_i$, $i=1, \ldots, N_f$.  The $m_i$ are the bare quark masses.

482: Generally, a quark with flavor $i$ and color $a=1, \ldots,

483: N_c$ has mass

484: \begin{equation}

485: m_i^a=|m_i - \phi^a|

486: \end{equation}

487: which is the shortest distance between color brane $a$ and flavor

488: brane $i$.  Alternately the quark mass can be read from the

489: superpotential terms:

490: \begin{equation}

491: W \supset \int d^2 \theta \ \{ {\tilde{Q}}_i \Phi Q^i + m_i

492: {\tilde{Q}}_i Q^i \} 

493: \end{equation}

494: where $Q_i, {\tilde{Q}}_i$ are the $\N=1$ chiral multiplets of the

495: quark hypermultiplet and $\Phi$ is the $\N=1$ chiral multiplet of the

496: $\N=2$ vector multiplet.

497: 

498: \begin{figure}[htb]

499: \centering

500: \includegraphics[width=6in]{quark.eps} 

501: \hsp{-1}\caption{Type IIA brane configurations corresponding to the a)

502:   Coulomb branch b) a mixed Coulomb and Higgs branch.}

503: \label{quark}

504: \end{figure}

505: 

506: %\begin{figure}[ht]

507: %\begin{center}

508: %\epsfxsize=6in\leavevmode\epsfbox{3flavor.eps}

509: %\caption{A system of 3 stacked flavor branes has U(3) flavor symmetry} \label{%flavor}

510: %\end{center}

511: %\end{figure}

512: When two flavor branes $i$ and $j$ and a color brane $a$ are at the

513: same $v$ coordinate, $\phi^a=m_i=m_j$, then it is possible to enter

514: the Higgs branch of the gauge theory.  This is done by connecting

515: color brane $a$ to flavor brane $i$ and then breaking flavor brane $j$

516: on D6 brane $i$.  At this point we are allowed to move the portion of

517: D4 brane $j$ which is between D6 branes $i$ and $j$, corresponding to

518: generating vacuum expectation values for the squarks in the

519: hypermultiplet.  These vacuum expectation values are parameterized by

520: the position of the D4 brane in the $x^7$, $x^8$, and $x^9$ directions as

521: well as the Wilson line of the gauge field $A_6$.

522: 

523: %In this discussion we have not mentioned the degrees of freedom which

524: %would correspond to strings stretched between the flavor branes.  Due

525: %to the boundary conditions at the NS5 brane and the D6 brane, the

526: %flavor branes are fixed.  However, these degrees of freedom will

527: %become important when we go to $\mathcal{N}=1$ and will turn out to be

528: %the mesons in the theory.  In that case, the D4 branes will not be

529: %fixed because the NS5 brane will be rotated, leading to a different

530: %boundary condition which allows the D4 brane to move.

531: 

532: The U(1$)_R$ $\times$ SU(2$)_R$ R-symmetry of the classical $\N=2$

533: theory is manifested as a rotational symmetry of the brane cartoon.

534: The U(1$)_R$ symmetry corresponds to rotations of the $v$-plane, while

535: the SU(2$)_R$ is the universal cover of the SO(3) acting on $x^7$,

536: $x^8$, and $x^9$ by rotations.

537: 

538: \subsection{Quantum Gauge Theory and M Theory}

539: 

540: By lifting the above brane configuration to M theory {\cite{Witten}},

541: we can consider non-perturbative (in $g_s$) quantum corrections and

542: thereby gain considerable insight into the origin of quantum phenomena

543: in SQCD. In particular, we consider the above IIA brane configuration

544: in the limit of large $g_s$ and $L_6$ with $\frac{1}{g^2}=\frac{L_6}{R}$

545: fixed, where the classical description of M theory is valid.

546: 

547: In M theory, the D4 brane is an M5 brane which wraps the M theory

548: circle ($x^{10}\sim x^{10}+2\pi$) once while an NS5 brane is an M5

549: brane which does not wrap the $x^{10}$ direction. The collection of D4

550: and NS5 branes can therefore be described as a single M5 brane

551: $\R^{3,1} \times \Sigma$ which fills the $x^0$, $x^1$, $x^2$, $x^3$

552: space and is a Riemann surface $\Sigma$ in the $x^4$, $x^5$, $x^6$,

553: and $x^{10}$ directions.

554: 

555: If we introduce the new holomorphic coordinate,

556: \begin{equation}

557: t=\exp{\left( \frac{-x^6}{R}-i x^{10} \right)}

558: \end{equation}

559: where $R=g_s l_s$ is the radius of the M theory circle, we can construct

560: $\Sigma$ explicitly as the vanishing locus of a polynomial $F(v,t)$,

561: \begin{equation}

562: F(v,t)=t^2+t\prod_{a=1}^{N_c} (v-\phi^a)+\Lambda^{2N_c-N_f}\prod_{i=1}^{N_f}(v-m_i)=0.

563: \end{equation}

564: where $\Lambda$ is the dynamically generated QCD scale of the theory.

565: Note that $\Sigma$ is precisely the Seiberg-Witten curve

566: \cite{SeibergWitten1} of the gauge theory. For example, the lift of

567: the brane configuration shown in Fig.~\ref{m5} is the single M5 brane

568: pictured in Fig.~\ref{m5lift}.

569: 

570: %This can be understood by noting that in limit, 

571: %the relation to M

572: %theory is $R=g_s l_s$, $l_P = g^{1/3}_s l_s$. Membrane tension

573: %$1/l^3_P$, wrapped gives string tension $2\pi R/l^3_P = 2\pi /

574: %l^2_s$.

575: %\begin{equation}

576: %g_s \ \rightarrow \ 0 \sp \frac{L_6}{l_s} \ \rightarrow \ 0 \sp

577: %g=\textup{constant}.

578: %\end{equation}

579: 

580: \begin{figure}[htb]

581: \centering

582: \includegraphics[width=4in]{su3nf32a.eps}

583: \caption{IIA realization of the Coulomb branch of $\N=2$ SU(3) SQCD

584:   with 3 flavors} \label{m5}

585: \includegraphics[width=4in]{su3nf3g.eps}

586: \caption{The M theory lift of the above IIA configuration to a single

587:   M5 brane. The directions $x^4$, $x^5$, and $x^6$ are shown explicitly

588:   while the $x^{10}$ coordinate is parameterized by darkness.}

589:   \label{m5lift}

590: \end{figure}

591: 

592: We can immediately see the running of the gauge coupling from the form

593: of the curve. Since the couplings are functions of the mass scales of

594: the lightest charged states in the theory, and as these scales just

595: correspond to distances in $v$ {\cite{Witten}}, we find that

596: \begin{equation} \label{grun}

597: \frac{1}{g(v)^2}=\frac{L_6(v)}{g_s l_s} \ \sim \ \log |v| \ .

598: \end{equation}

599: Also, the U(1$)_R$ anomaly is particularly easy to see in the M theory

600: picture. Recall that D4 branes wrap the M theory direction and end on

601: the NS5 branes, which do not wrap the M theory direction.  This means

602: that the end of a D4 brane on an NS5 brane is a vortex in the

603: embedding coordinate of the NS5 brane in the M theory direction (see

604: Figs.~\ref{m5lift} and \ref{Uday}). In particular, at large $v$ if one

605: follows a circle along the NS5 brane whose interior contains all of

606: the D4 branes, this circle will wrap the M theory direction as many

607: times as there are colors minus flavors attached to this NS5 brane.

608: Thus the naive U(1$)_R$ rotational R-symmetry of the brane must be

609: combined with a simultaneous rotation of the M theory circle. Such a

610: redefinition is not possible with both NS5 branes as the $x^{10}$

611: redefinitions would have to be in opposite directions for the two

612: branes and so the U(1$)_R$ is anomalous. There is a residual

613: $\Z_{2N_c-N_f}$ symmetry in the U(1) redefined to include a rotation

614: of the M theory circle in opposite directions on both NS5 branes. For

615: these special angles the two redefinitions only disagree by a multiple

616: of $2\pi$ in the M theory direction.

617: 

618: The reader may verify these claims visually by considering the

619: asymptotic $x^{10}$ dependence of the regions corresponding to the NS5

620: branes (large positive and negative values of $x^6$) on the M5 branes

621: pictured in Figs.~\ref{m5lift} and \ref{Uday}.

622: 

623: Another benefit of the M theory lift is that all the matter in the

624: gauge theory is realized in M theory by open M2 branes ending on the

625: M5 brane. In reducing to IIA, the M2 branes which wrap the M theory

626: circle carry fundamental string charge, while unwrapped M2 branes are

627: D2 branes. The M2 brane configurations corresponding to BPS states

628: were analyzed critically in \cite{HY,Russian}. They found that BPS

629: states were M2 branes with the minimal possible areas in their

630: homology classes.

631: 

632: We can also gain intuition for baryons and mesons in SQCD using M2

633: branes. For example, baryons are M2 branes {\cite{Rabinovici}} with

634: $k+1$ boundaries as seen in Figure~\ref{baryon}.  One boundary wraps

635: $k$ color branes and the other $k$ boundaries each wrap a flavor

636: brane.  In the IIA limit, the baryon is a collection of $k$ quarks

637: attached by a k-string {\cite{Witten2,HananyZaffaroni}}.  Mesons are

638: tubular M2 branes which start by wrapping one flavor brane, connect to

639: a color brane, and then extend in the space directions ($x^1$, $x^2$,

640: and $x^3$) before wrapping another flavor brane.

641: 

642: \begin{figure}[htb]

643: \centering

644: \includegraphics[width=6in]{baryontest.eps}

645: \caption{a) a baryon which wraps 2 color and 2 flavor branes in SU(2)

646:   SQCD with 2 flavors\hsp{.5} {}b) a meson which wraps 2 flavor branes

647:   and wraps and unwraps one color brane} \label{baryon}

648: \end{figure}

649: 

650: We certainly have gained a great deal in this lift, but we don't quite

651: get it for free. Since this lift involves going to large string

652: coupling, one cannot be sure that the low energy effective theory on

653: the M5 brane is the gauge theory we started with. In fact, it is

654: actually a six dimensional theory. The M theory limit is exactly the

655: opposite limit needed to obtain the 4D gauge theory. It requires

656: taking the radius of the M theory circle $R=g_s l_s$ as well as $L_6$

657: large leaving $\frac{1}{g^2}=\frac{L_6}{R}$ fixed.  However, since

658: unbroken supersymmetries protect certain holomorphic quantities (like

659: the masses of BPS states and superpotentials) from perturbative $g_s$

660: quantum corrections, one can still use classical computations

661: involving the M5 brane to determine them exactly.  The breakdown of

662: the M theory picture for the computation of non-holomorphic quantities

663: (like the masses of non-BPS states) can then be understood as due to

664: the presence of KK modes (dynamical D0 branes and $A_6$ fluctuations)

665: which become light and strongly coupled in this limit \cite{Witten2}.

666: 

667: For example, one can solve the $\N=2$ gauge theories constructed above

668: at generic points in their moduli spaces as their effective actions

669: are governed by holomorphic quantities ($\N=2$ prepotentials ${\cal

670:   F}$) which can be computed exactly using $\Sigma$ {\cite{Witten}}.

671: Upon breaking to $\N=1$, the K\"{a}hler potential is no longer

672: protected from such corrections. However, as the superpotential of the

673: $\N=1$ theory is still protected, we will see that we can still learn

674: what we need about $\N=1$ SQCD from M theory.

675: 

676: \subsection{Soft Breaking to $\mathcal{N}=1$ SQCD} {\label{N1section}}

677: 

678: Rotating one of the NS5 branes

679: \begin{equation} 

680: \left( \begin{array}{c} v\p \\  w\p \end{array} \right) = \left(

681:   \begin{array}{cc} \cos \theta & \sin\theta \\  -\sin \theta & \cos

682:     \theta \end{array} \right) \left( \begin{array}{c} v \\  w

683:   \end{array} \right) \ 

684: \end{equation} 

685: results in a brane configuration which preserves only four

686: supercharges {\cite{Berkooz}} and softly breaks ${\mathcal{N}=2}$

687: supersymmetry to ${\mathcal{N}=1}$.  We will refer to the rotated NS5

688: brane as the NS$5_\theta$ brane. The NS$5_{\pi/2}$ brane is commonly

689: referred to as the NS$5\p$ brane in the literature.

690: 

691: \begin{figure}[htb]

692: \centering

693: \includegraphics[width=6in]{3tflavor.eps}

694: \caption{a) $r$ color branes and flavor brane connects.  b) After

695:   rotating the five brane, color branes which are not connected to

696:   flavor branes move to the origin of $v$.  This is the realization of

697:   the $r$ vacua in \cite{Plesser}.} \label{3tflavor}

698: \end{figure}

699: 

700: The symmetry breaking classical superpotential generated by this

701: process can be understood as follows. After the rotation, color branes

702: at generic positions in $v$ no longer minimize their lengths. In fact,

703: to reach equilibrium, all color branes much either slide to $v\sim

704: O(\Lambda)$ or attach to flavor branes.  Since translations along $v$

705: (which correspond to adjoint scalar vevs) now cause the color branes

706: to stretch, the $\N=1$ chiral multiplets containing these adjoint

707: scalars acquire a mass $\mu$ via the superpotential term

708: \begin{equation}

709: W\supset\mu \ \textup{Tr} \Phi^2\sp 

710: \mu \ \sim \ \tan \theta \ .

711: \end{equation}

712: If $r$ color branes connect to flavor branes, classically the only

713: surviving vacuum is

714: \begin{equation}

715: \phi^i=m_i \ \mbox{  for  } \ i=1, \ldots, r\hsp{.05} \hsp{.666}\phi^a=0

716:  \ \mbox{  for  } \ a=r+1, \ldots N_c.

717: \end{equation}

718: However, when quantum corrections are considered the surviving vacua

719: are those where $N_c-r$ or $N_c-r-1$ monopoles and dyons become

720: massless and condense with vevs of order $O(\mu\Lambda)$. In the M

721: theory picture, this occurs when the bounding cycles of the

722: corresponding M2 branes degenerate, as in Fig.~{\ref{Uday}}.

723: 

724: \begin{figure}[htb]

725: \centering

726: \includegraphics[width=3in]{su3nf0g.eps}

727: \caption{An M5 brane configuration corresponding to $\N=2$ SU(3) SYM

728:   near a vacuum which survives upon softly breaking to $\N=1$. The two

729:   small holes are the degenerating cycles. The M2 branes corresponding

730:   to nearly massless monopoles are disks bounded by these cycles. The

731:   M theory direction is parameterized by darkness.}

732: \label{Uday}

733: \end{figure}

734: 

735: More precisely, the vacua which survive the $\mu \mbox{Tr} \Phi^{2}$

736: perturbation are those in which the cycles that have degenerated

737: result in an M5 brane configuration $\Sigma \times \R^4$ such that

738: $\Sigma$ is of genus zero, as argued in Refs.~{\cite{SeibergWitten1,

739:     Witten2, HoriOz}}. 

740: 

741: To preserve SUSY the flavor branes must continue to extend along the

742: $x^0$, $x^1$, $x^2$, $x^3$, and $x^6$ directions, in particular they

743: cannot rotate into the $w$ directions.  Therefore when the

744: NS$5_\theta$ brane rotates, the flavor branes ending on it must

745: translate in $w$, sliding along the corresponding D6 brane.  This

746: translation corresponds to meson vevs \cite{Giveon,HoriOz}.

747: 

748: \section{Summary of Field Theory Results} \label{sumfield}

749: 

750: The dynamics of the $\N=2$ supersymmetric SU($N_{c}$) gauge theories

751: constructed above and the dynamical breaking of flavor symmetry have

752: been studied in detail in Refs.~{\cite{Murayama1, Murayama,

753: Murayamatalk}}.  Here we briefly summarize the results.

754: 

755: The theory with $N_{f}$ massless quark hypermultiplets has U($N_{f}$) flavor

756: symmetry, SU(2$)_{R}$ symmetry, and a non-anomalous discrete

757: $\Z_{2N_{c}-N_{f}}$ subgroup of U(1$)_{R}$.  We are interested in the

758: $\N=1$ perturbation of the theory by the adjoint mass term $\mu

759:  \mbox{ Tr }\Phi^{2}$.  The moduli space contracts to the set of points

760: that give maximally degenerate (genus zero) Seiberg--Witten curves.

761: 

762: In the semi-classical regime with a large adjoint VEV, there are

763: `t~Hooft--Polyakov magnetic monopoles.  Zero modes of the quarks

764: around a monopole generate flavor quantum numbers for the magnetic

765: monopoles.  It was shown that they come in completely anti-symmetric

766: rank-$r$ tensor representations with $_{N_{f}}C_{r}$ multiplicities.

767: 

768: The strongly coupled regime was studied with a variety of techniques.

769: When $N_{f} < N_{c}$, there are vacua parameterized by an integer $r =

770: 0, 1, \cdots, [N_{f}/2]$ ($[x]$ is the Gauss' symbol) where the flavor

771: symmetry is broken dynamically as U($N_{f}$) $\rightarrow$ U($r$)

772: $\times$ U($N_{f}-r$). If $r<N_f/2$, the physics around the vacuum can

773: be described by an IR free effective Lagrangian \cite{Plesser} (a

774: ``magnetic dual'' to the asymptotically free semiclassical SU($N_c$)

775: description) with SU($r$) $\times$ U(1$)^{N_{c}-r-1}$ gauge group.

776: $N_{f}$ ``magnetic quark'' hypermultiplets transform as the

777: fundamental representation of SU($r$) while there are ``magnetic

778: monopoles'' for each of the ``magnetic'' U(1) factors.  When perturbed

779: by the adjoint mass term, all gauge groups are Higgsed by the

780: condensates of magnetic objects, corresponding to the confinement of

781: the electric theory.  It was argued in Ref.~\cite{Murayama1, Murayama}

782: that the semi-classical monopoles in the rank-$r$ anti-symmetric

783: tensor representation smoothly match to the baryonic composites of

784: magnetic quarks of the low-energy SU($r$) theory based on

785: circumstantial evidence.  Upon mass perturbations, one can count the

786: number of vacua:

787: \begin{equation}

788:     \N_{1} = (2N_{c}-N_{f})2^{N_{f}-1},

789: \end{equation}

790: originating from $r$-vacua ($r \leq [N_{f}/2]$) with $(2N_{c}-N_{f})$

791: copies due to the $\Z_{2N_{c}-N_{f}}$ symmetry.  Therefore {\it the

792:   flavor symmetry breaking and confinement have a common origin in

793:   these theories}\/: condensation of magnetic objects with non-trivial

794: flavor quantum numbers.  Strictly speaking, however, the existence of

795: monopoles in the anti-symmetric tensor representations was

796: demonstrated only in the semi-classical regime and its extrapolation

797: to the strongly coupled regime and the matching to the baryonic

798: composite was a conjecture.  When $r=N_{f}/2$ (possible obviously only

799: when $N_{f}$ is even), the low-energy magnetic gauge group is

800: superconformal with an infinitely strong coupling $\tau = -1$.  Due to

801: some reason, the same low-energy effective action seems to describe

802: the dynamics of flavor symmetry breaking even though there is no

803: weakly coupled description of the theory.

804: 

805: When $N_{f} > N_{c}$, there is a new vacuum without flavor symmetry

806: breaking.  It is at the same point on the moduli space as the

807: $r=N_{f}-N_{c}$ vacuum, while the finite quark mass perturbation shows

808: that there are additional

809: \begin{equation}

810:     \N_{2} = \sum_{r=0}^{N_{f}-N_{c}-1} (N_{f}-N_{c}-r)

811:     {}_{N_{f}}C_{r}

812: \end{equation}

813: vacua with unbroken flavor symmetry. 

814: %Thus, the total number of vacua

815: %calculated via semiclassical arguments for this case is

816: %\begin{equation} \label{scvac}

817: %\N_{sc}= \N_1 + \N_2 = \sum_{r=0}^{min(N_c+1,N_f)}(N_c-r)(^{N_f}_{\ r}).

818: %\end{equation}

819: 

820: \section{Seiberg Duality from M Theory} \label{MSeiberg}

821: 

822: As we noted in the last section, the effective theories at the

823: $r$-vacua which survive SUSY breaking are actually IR free theories

824: for $r < N_f/2$. In particular, at the baryonic root we find a weakly

825: coupled theory whose low energy physics is well described by an IR

826: free SU($\tilde{N}_c$) gauge theory, where $\tilde{N}_c = N_f - N_c$.

827: This is very reminiscent of Seiberg duality in $\N=1$ theories

828: \cite{Seiberg1}. In \cite{Plesser}, this observation was used to

829: ``derive'' $\N=1$ Seiberg duality by mass perturbing the $\N=2$

830: theory\footnote{Unfortunately, the result was not quite Seiberg

831:   duality as there was an extra non-renormalizable coupling in the

832:   effective action associated with integrating out the massive adjoint

833:   scalar which becomes relevant in the low energy limit.}.  We will

834: consider this scenario using M theory.

835: 

836: Seiberg duality was first realized in IIA string theory in

837: Ref.~{\cite{Elitzur}} via the exchange of two NS5 branes, as

838: illustrated in Fig.~\ref{sduality}.  To avoid a singular configuration

839: the authors first displaced one of the NS5 branes in the 7 direction,

840: which corresponds to turning on a Fayet-Illipolous term in the field

841: theory description.  This process corresponds to Seiberg duality with

842: one caveat, the full U($N_f$) $\times$ U($N_f$) flavor symmetry is not

843: realized in the above IIA configuration nor in its M theory lift.

844: 

845: \begin{figure}[htb]

846: \centering

847: \includegraphics[width=6in]{sduality.eps}

848: \caption{In IIA Seiberg duality is realized by brane exchange.}

849: \label{sduality}

850: \end{figure}

851: 

852: An alternate proposal for realizing Seiberg duality via string/M

853: theory was made later that year by Schmaltz and Sundrum \cite{SS}.

854: Using the M theory lift of the above brane setup, they found that

855: Seiberg duality could be understood by taking the $\Lambda\rightarrow

856: 0$ and $\Lambda\rightarrow\infty$ limits of the resulting single M5

857: brane configuration. In particular, they found that the two limits

858: were related by an exchange of branes and therefore correspond to an

859: electric theory and its magnetic dual. However, their arguments

860: depended on turning on finite bare quark masses.

861: 

862: The following year this scenario was clarified and extended to the

863: massless case by Hori {\cite{Kentaro}}.  He observed that in M-theory,

864: the M5 branes can be crossed with no singularity even without the

865: Fayet-Illiopolous parameter.  In fact, at the root of the baryonic

866: branch, when $N_f \geq N_c+1$ the M5 brane consists of two connected

867: components \cite{HoriOz} and the duality corresponds to simply

868: translating one past the other along $x^6$.  The $x^6$ coordinate of a

869: connected component does not affect the field theory description and

870: so Hori argues that this process clearly preserves its universality

871: class.

872: 

873: Most recently, in Ref.~\cite{Dasgupta} Seiberg duality was understood

874: within the context of geometric engineering as a birational flop in a

875: T dual description consisting of D5 branes which are dual to the D4

876: branes in IIA and degenerations in the T-dualized circle which are

877: dual to the NS5 branes.

878: 

879: We propose a new description of Seiberg duality which has no moving

880: branes and is valid at any fixed, finite value of $\Lambda$.  Consider

881: the case $N_f > N_c$ at the root of the baryonic branch in an

882: $\cal{N}=$2 theory with no bare quark masses.  Then according to

883: Ref.~{\cite{HoriOz}} the connected components of the M5 brane are

884: described by

885: \begin{equation}

886: t=v^{N_c}\hsp{.5}\textup{and}\hsp{.5} t=v^{\tilde{N}_c}, \hsp{.3}

887: \tilde{N}_c=N_f-N_c

888: \end{equation}

889: as seen in Figs.~\ref{duality} and \ref{mduality}.  These two M5

890: branes intersect at $N_c-\tilde{N}_c$ points.

891: 

892: \begin{figure}[htb]

893: \centering

894: \includegraphics[width=6in]{duality1.eps}

895: \caption{At the root of the baryonic branch, the low energy physics

896:   (to the right of the intersection) is the SU($\tilde{N}_c$) $\times$

897:   U(1$)^{N_c-\tilde{N}_c}$ magnetic theory.} \label{duality}

898: \end{figure}

899: 

900: \begin{figure}[htb]

901: \centering

902: \includegraphics[width=4in]{su8nf13.eps}

903: \caption{Baryonic root of SU(8) with 13 flavors} \label{mduality}

904: \end{figure}

905: 

906: The crucial realization is that the reduction of this configuration to

907: IIA and in particular the $x^6$ location of the NS5 branes is not

908: uniquely defined \cite{Witten}.  Rather, we claim, that the effective

909: location of the NS5 branes depends on the energy scale probed, $E$.

910: We consider charged hypermultiplet matter corresponding to M2 branes of disk

911: topology stretched between the two branches of the M5 brane. It is

912: clear from Fig.~\ref{mduality} that the area of such an M2 brane (and

913: therefore its energy) is proportional to the distance in $v$ between

914: the two branches of M5 brane at its position in $x^6$. This restricts

915: the regions on the M5 brane that a quark of a given energy can

916: probe. Thus, for such a probe, one may effectively place an NS5 brane

917: at the intersection of the corresponding M5 brane and $v \sim E$. A

918: high energy $E_1 \gg \Lambda$ (semiclassical) M2 brane probe is

919: restricted to probe large features at small $x^6$. Thus, to such a

920: probe, the configuration consists of an NS5 brane with an NS$5_\theta$

921: brane on its left, $N_c$ color branes connecting them and $N_f$

922: semi-infinite flavor branes which extend to the right. That is, the

923: reduction to type IIA is Fig.~{\ref{sduality}}a, which can roughly be

924: obtained by drawing an NS5 brane wherever the line $E_1$ crosses an

925: $M5$ brane.  This corresponds to the SU($N_c$) asymptotically free

926: electric theory.  However, low energy quarks can only exist at

927: sufficiently large $x^6$.  Thus, a low energy $E_2 \ll \Lambda$ M2

928: brane probe corresponding to a charged quark is only sensitive to the

929: configuration of the M5 brane at large $x^6$, the right side of

930: Fig.~\ref{duality}. Thus, if we consider the portion of the M5 brane

931: configuration accessible to such a probe, we would find that the

932: corresponding reduction to IIA at $E_2$ is Fig.~{\ref{sduality}}d.

933: 

934: Now, note that the two M5 branes cross at the QCD scale

935: $\Lambda$, and at energy scales below this, that is, further right in

936: the figure, the M5 brane is to the left of the M5$_\theta$ brane.

937: Thus a probe at energies below the QCD scale will see the two M5

938: branes interchanged, which is the usual description of the magnetic

939: theory.  This theory is IR free because the branes separate as $v$

940: increases, and has a Landau pole where the branes cross.

941: 

942: \section{Flavor Symmetry Breaking} \label{flavbreak}

943: 

944: \subsection{Flavored Magnetic Monopoles}

945: 

946: Recall that a flavorless magnetic monopole (or dyon) in a IIA

947: realization of the electric picture is a D2 brane with disk topology

948: bounded by a circle which extends along one color brane, down an NS5

949: brane, back along another color brane and then finally back along the

950: other NS5 brane, as in Fig.~\ref{basic}.  More generally magnetic

951: monopoles may be charged under the U($N_f$) flavor symmetry.  

952: In this case the monopole may include fundamental strings extending from

953: the D2 brane to a flavor brane.  Another version of the s-rule states

954: that at most one such fundamental string can extend from a given

955: monopole to a given flavor brane, which identifies these strings as

956: excitations of the fermionic zeromodes present in flavored monopoles.

957: 

958: The M theory lift of this monopole configuration is a single M2 brane

959: with the topology of a disk. Thus, its boundary is a circle which

960: wraps the M theory direction a total of zero times, as seen in

961: Fig.~{\ref{magmon}}.

962: 

963: The microscopic mechanism behind the flavor symmetry breaking of

964: Refs.~{\cite{Murayama1, Murayama, Murayamatalk}} is the condensation

965: of magnetic monopoles in an antisymmetric tensor flavor representation

966: drawn in Fig.~\ref{magmon}a.  The transformation properties under

967: U($N_f$) can be read off from the brane cartoon.  If the monopole does

968: not wrap any flavor branes it transforms as a flavor singlet.  A

969: monopole that wraps one of the flavor branes transforms in the

970: fundamental representation, while wrappings of more than one flavor

971: brane transform in the antisymmetric representation of the flavor

972: group. To see why this representation is antisymmetric, notice that in

973: IIA, if the D6 branes are moved between the NS5 branes, a monopole

974: consists of a D2 brane connected by strings to D6 branes.  The s-rule

975: provides an exclusion principle, restricting the number of strings

976: connecting a monopole to a D6 brane to 0 or 1. As a consistency check

977: on this picture, notice that there are $2^{N_f}$ configurations of

978: wrappings which agrees with the known number of states in the

979: representation.

980: 

981: In order to understand how semiclassical magnetic monopoles in the UV

982: theory are related to the IR degrees of freedom, we can consider an M2

983: brane corresponding to a high energy monopole configuration and follow

984: its decay into the IR. Deforming the UV theory away from the baryonic

985: root to clearly visualize its charges, such a monopole is pictured in

986: Fig.~\ref{magmon}a. We deform back to the baryonic root and then allow

987: it to decay, requiring that its wrappings (i.e. charges) are

988: preserved. Its energy will become sufficiently low that it is best

989: described using the dual magnetic description, that is, with the NS5

990: branes switched. Now, if we deform the IR theory to a generic point in

991: its Coulomb branch and keep track of the wrappings, this configuration

992: corresponds to magnetic baryons as seen in Fig.~\ref{magmon}b.  The

993: magnetic theory is IR free, and so these baryons decay into magnetic

994: quarks whose condensation provides the order parameter of the flavor

995: symmetry breaking.

996: 

997: \begin{figure}[htb]

998: \centering

999: \includegraphics[width=6in]{magmons.eps}

1000: \caption{a) A magnetic monopole in an antisymmetric tensor flavor

1001:   representation. \hsp{.5} b) In the dual magnetic picture it becomes

1002:   a baryon.} \label{magmon}

1003: \end{figure}

1004: 

1005: It is also easy to see that the correlation between the electric

1006: charge and chirality of dyons discussed for SU(2) gauge theories in

1007: Ref.{\cite{SeibergWitten2}} follows from the fact that the monopole is

1008: topologically a disk.  All monopoles with even chirality come from

1009: monopoles whose boundary wraps an even number of flavor branes and

1010: therefore the M theory direction an even number of times along the M5

1011: brane as well.  Each flavor brane wrapping introduces a

1012: hypermultiplet into the antisymmetric representation and hence the

1013: monopole has chirality $(-1)^H=1$.  Each color brane wrapped yields a

1014: unit (using the conventions of {\cite{SeibergWitten2}}) of electric

1015: charge and so the monopole acquires an even electric charge.  This

1016: argument works similarly for odd chirality and odd electric charge.

1017: 

1018: \subsection{Symmetry Breaking Pattern}

1019: 

1020: As we reviewed in section \ref{sumfield}, field theory

1021: calculations in a variety of limits showed that flavor symmetry is

1022: generically broken

1023: \begin{equation}

1024: U(N_f)\hsp{.1}\rightarrow\hsp{.1}U(r)\times U(N_f-r)

1025: \end{equation}

1026: in softly broken $\N=1$ asymptotically free ($N_f<2N_c$) SQCD in the

1027: limit that bare quark masses vanish.  We will draw string and M theory

1028: realizations of these limits and use them to reproduce the qualitative

1029: results of several field theory calculations.  In particular, in the

1030: semiclassical limit, we will relate $r$ to the number of color branes

1031: attached to flavor branes.

1032: 

1033: \begin{figure}[htb]

1034: \centering

1035: \includegraphics[width=6in]{semivevs.eps}

1036: \caption{a) Semiclassically when SUSY is broken by rotating the

1037:   NS$5_\theta$ brane, $r$ color branes connect to flavor branes while

1038:   the rest slide to $v=0$.\hsp{.4}b)The dual magnetic description of

1039:   the nonbaryonic branches can be understood similarly to the

1040:   semiclassical case.} \label{semivevs}

1041: \end{figure}

1042: 

1043: \subsection{Semiclassical Analysis}

1044: 

1045: Following Refs.~{\cite{Murayama1, Murayama, Murayamatalk}} we begin by

1046: considering bare quark masses much larger than the QCD scale so that a

1047: semiclassical (IIA) analysis is valid.  This will allow us to count

1048: the total number of vacua for comparison with later calculations.

1049: Recall from section~\ref{N1section} that after rotating the

1050: NS$5_\theta$ brane all color branes must either slide to $v=0$ or

1051: connect to a flavor brane as in Fig.~\ref{semivevs}a.  The number of

1052: color branes connecting to flavor branes, $r$, clearly can neither

1053: exceed the number of color branes nor the number of flavor branes. We

1054: illustrate the simple case of the vacua arising this way with the bare

1055: quark masses all equal $m=m_i \gg \Lambda$ in Appendix A

1056: for the case of SU(3) with $N_f=4$.

1057: 

1058: %\begin{figure}[htb]

1059: %\centering

1060: %\includegraphics[width=3in]{su3nf4r2.eps}

1061: %\caption{The semiclassical $r=2$ configuration in SU(3) with 4

1062: %  flavors.}  \label{semic}

1063: %\end{figure}

1064: 

1065: To count the number of vacua with generic quark masses, notice that

1066: $r$ flavor branes can attach to color branes in $(^{N_f}_{\ r})$ ways

1067: (recall that there is no combinatoric factor from choosing which color

1068: branes to attach as these choices are gauge equivalent), leaving

1069: $N_c-r$ color branes which form a line\footnote{Actually they form an

1070:   ellipse whose semi-minor axis scales with $\theta$.  This ellipse

1071:   degenerates to a line when $\theta=0$ and a circle with an

1072:   $A_{N_c-r-1}$ singularity in its center at $\theta=\pi/2$.}

1073: {\cite{HananyZaffaroni}} centered at $v=0$, as shown in

1074: Fig.~\ref{Uday}.  This line can be rotated by integer multiples of

1075: $\pi/(N_c-r)$ without affecting the $x^{10}$ coordinates

1076: asymptotically far away, corresponding to the anomaly-free R-symmetry

1077: subgroup.  The $N_c-r$ inequivalent orientations of the line result in

1078: $N_c-r$ different vacua, in agreement with the Witten index of this

1079: theory.  Thus the total number of semiclassical vacua (assuming all

1080: bare quark masses are distinct and nonvanishing) is

1081: \begin{equation}

1082: \N_{sc}=\sum_{r=0}^{min(N_c+1,N_f)}(N_c-r)(^{N_f}_{\ r})

1083: \end{equation}

1084: in agreement with computations from field theory considerations in

1085: Refs.~{\cite{Murayama1, Murayama, Murayamatalk}}.

1086: 

1087: \subsection{Nonbaryonic Branches}

1088: 

1089: Semiclassically flavor symmetry is broken by meson vevs equal to $\mu

1090: m_i \sim \mu \ \textup{tan}\theta$ which is the distance in the $w$

1091: plane shown in Fig.~\ref{semivevs}a.  These vevs vanish when the bare

1092: quark masses vanish, apparently restoring the explicitly broken flavor

1093: symmetry.  However we will see that, if we include quantum effects, in

1094: some vacua the flavor symmetry remains broken even in this limit.  For

1095: simplicity let all of the bare quark masses be equal $m=m_i \ll

1096: \Lambda$.  

1097: 

1098: The $r=N_f/2$ theory is superconformal in the IR and generally

1099: difficult to understand. An example of M5 brane associated with such a

1100: vacuum is drawn in Fig.~\ref{scir}.  The two branches, corresponding

1101: to different NS5 branes upon reduction to IIA, intersect at exactly

1102: two points.  Cross-sections to the left and right of these two

1103: singularities are seen in the first and second lines of

1104: Fig.~\ref{scircs}.  They are topologically distinct and correspond to

1105: distinct reductions to IIA, yielding different effective field

1106: theories.

1107: 

1108: \begin{figure}[htb]

1109: \centering

1110: \includegraphics[width=4in]{su3nf4r2m=0.eps}

1111: \caption{This M5 brane configuration corresponds to the $r=2$

1112:   nonbaryonic root of SU(3) with 4 flavors.  In the IR (the right side

1113:   of the picture) a reduction to IIA produces two parallel, almost

1114:   coincident NS5 branes indicating that the theory is strongly

1115:   coupled.  The fact that the distance between the branes converges

1116:   indicates that the IR theory is superconformal.}  \label{scir}

1117: \end{figure}

1118:  

1119: \begin{figure}[htb]

1120: \begin{minipage}[c]{.25\textwidth}

1121: \centering

1122:   \includegraphics[width=1.6in]{su3nf4r2z=-1.45.eps}

1123: \end{minipage}%

1124: \begin{minipage}[c]{.25\textwidth}

1125: \centering

1126:   \includegraphics[width=1.6in]{su3nf4r2z=-1.23.eps}

1127: \end{minipage}%

1128: \begin{minipage}[c]{.25\textwidth}

1129: \centering

1130:   \includegraphics[width=1.6in]{su3nf4r2z=-1.13.eps}

1131: \end{minipage}%

1132: \begin{minipage}[c]{.25\textwidth}

1133: \centering

1134:   \includegraphics[width=1.6in]{su3nf4r2z=-1.12.eps}

1135: \end{minipage}

1136: \begin{minipage}[c]{.25\textwidth}

1137: \centering

1138:   \includegraphics[width=1.2in]{su3nf4r2z=-1.00.eps}

1139: \end{minipage}%

1140: \begin{minipage}[c]{.25\textwidth}

1141: \centering

1142:   \includegraphics[width=1.5in]{su3nf4r2z=-0.50.eps}

1143: \end{minipage}%

1144: \begin{minipage}[c]{.25\textwidth}

1145: \centering

1146:   \includegraphics[width=1.5in]{su3nf4r2z=0.00.eps}

1147: \end{minipage}%

1148: \begin{minipage}[c]{.25\textwidth}

1149: \centering

1150:   \includegraphics[width=1.2in]{su3nf4r2z=3.00.eps}

1151: \end{minipage}

1152: \caption{Above are M5 brane cross-sections at constant $x^6$

1153:   corresponding roughly to constant energy scales of the $r=2$

1154:   nonbaryonic root of the SU(3) theory with 4 flavors and no bare

1155:   quark masses.  Reading from left to right and top to bottom the

1156:   energy scale of the corresponding M2 brane probes decreases.  The M

1157:   theory direction is parameterized by darkness.}  \label{scircs}

1158: \end{figure}

1159: 

1160: We will be interested in the IR free case of $0<r<N_f/2$,

1161: corresponding to nonbaryonic branches.  The fundamental degrees of

1162: freedom must be magnetic because in the UV this theory has a Landau

1163: pole (M5 branes cross) separating it from the semiclassical region

1164: with electric degrees of freedom.  Thus flavor symmetry breaking can

1165: only be caused by magnetic quark vevs, which like meson vevs (meson

1166: vevs are quadratic in quark vevs) correspond to distances in the $w$

1167: plane.  Semiclassically this distance was $\mu m_i$ and so vanished

1168: when the bare quark masses were taken to zero.  Quantum mechanically

1169: these flavor branes have width $O(\Lambda)$, critically changing the

1170: distances between them.  As a result the $r$ nonvanishing magnetic

1171: quark vevs

1172: \begin{equation}

1173: q=\sqrt{\mu(m_i-\Lambda/r)}

1174: \end{equation}

1175: do not vanish when the bare masses are eliminated, as seen in

1176: Fig.~\ref{semivevs}b.  These quark vevs break the flavor symmetry

1177: \begin{equation}

1178: U(N_f)\ \rightarrow\ U(r)\times U(N_f-r).

1179: \end{equation}

1180: These vevs also break the global $\Z_{2N_c-N_f}$ symmetry and

1181: so there must be $2N_c-N_f$ copies of this configuration.  When $r=0$,

1182: $q=\sqrt{\mu m_i}$ and flavor symmetry is unbroken.  Again there are

1183: $(^{N_f}_{\ r})$ ways to choose $r$ quarks, leading to a total of

1184: \begin{equation}

1185: \N_1=(N_c-\tilde{N}_c)2^{N_f-1}

1186: \end{equation}

1187: vacua of this type.  Notice that when $N_f<N_c$, $\N_1$ agrees with

1188: $\N_{sc}$ and therefore by supersymmetry this is a complete

1189: classification of the vacua.

1190: 

1191: \subsection{Baryonic Branch}

1192: The above analysis is incomplete when $N_f\ge N_c$ because every color

1193: brane can be broken by a D6 brane and the two halves displaced from

1194: each other along the $w$ plane by a distance (baryon vev) exactly

1195: canceling the displacement measured by the dual quark vev.  Clearly

1196: this requires $N_f\ge N_c$ because $N_f$ is the number of D6 branes

1197: while $N_c$ is the number of color branes, and each color brane

1198: requires a D6 brane along which to break, as illustrated in

1199: Fig.~\ref{brootfig}.

1200: 

1201: \begin{figure}[htb]

1202: \centering

1203: \includegraphics[width=6in]{broot.eps}

1204: \caption{On the baryonic branch, the baryon vev $\mu\Lambda$ is

1205:   realized as a relative slide of the two halves of the configuration

1206:   along a collection of D6 branes.} \label{brootfig}

1207: \end{figure}

1208: 

1209: More concretely, at the root of the baryonic branch the unattached

1210: $N_c-\tilde{N}_c$ color branes form a circle.  This means that

1211: $N_c-\tilde{N}_c$ magnetic monopoles (or dyons), which are between

1212: adjacent color branes, become massless and can acquire vevs.  The

1213: baryonic branch has one more massless monopole compared to the

1214: nonbaryonic branch, which is enough to completely Higgs the

1215: U(1$)^{N_c-\tilde{N}_c}$, and therefore there are enough vevs to

1216: control the center of mass motion of this set of color branes.  The

1217: center of mass mode of the entire system is infinitely massive and so

1218: a shift in the center of mass of the $N_c-\tilde{N}_c$ branes leads to

1219: an opposite center of mass shift of the Seiberg dual SU($\tilde{N}_c$)

1220: gauge theory along the $w$ plane, the same shift parametrized by all

1221: quark vevs.  One result is that the magnetic monopoles are charged

1222: under the U(1$)_B$ baryon number.  The crucial implication is that a

1223: slide along the $w$ plane in the U(1$)^{N_c-\tilde{N}_c}$ system can

1224: undo the $w$-shift in the SU($\tilde{N}_c$) configuration which led to

1225: the flavor symmetry breaking quark vev.  Thus flavor symmetry is

1226: unbroken on the baryonic branch.

1227: 

1228: These geometrical quantities can be related to the corresponding field

1229: theory calculation ~\cite{Murayama1, Murayama, Murayamatalk} by

1230: considering the following superpotential terms {\cite{Plesser}} in the

1231: magnetic description of the baryonic root

1232: \begin{equation}

1233: W\supset

1234: \frac{1}{\tilde{N}_c}\textup{Tr}(q\tilde{q})(\sum_{k=1}^{N_c-\tilde{N}_c}\psi_k)-\sum_{k=1}^{N_c-\tilde{N}_c}\psi_k

1235: e_k\tilde{e}_k+\mu\Lambda\sum_{k=1}^{N_c-\tilde{N}_c}x_k\psi_k.

1236: \end{equation}

1237: Here $q$ and $\tilde{q}$ are magnetic quarks, $e_k$ and $\tilde{e}_k$

1238: are flavorless magnetic monopoles, $x_k$ are constants and $\psi_k$

1239: are the dual photons of the abelian part of the

1240: SU($\tilde{N}_c$) $\times$ U(1$)^{N_c-\tilde{N}_c}$ dual gauge group.

1241: From the superpotential we see that each magnetic monopole is charged

1242: under a U(1) while the dual quarks are charged under all of the

1243: U(1)'s.\footnote{This is in contrast to the nonbaryonic root, where

1244:   there is one less massless magnetic monopole and so after a basis

1245:   change the dual quarks and magnetic monopoles are charged under

1246:   disjoint U(1)'s.}  As a result the D term equation for each U(1) is

1247: \begin{equation}

1248: \frac{1}{\tilde{N}_c}\textup{Tr}(q\tilde{q})-e_k\tilde{e}_k+\mu\Lambda

1249: x_k = 0. \label{dequation}

1250: \end{equation}

1251: This means that, because there are as many $e_k$'s as $x_k$'s,

1252: Tr$(q\tilde{q})$ can vanish if each $e_k\tilde{e}_k$ is chosen

1253: correctly and, in fact, this solution is consistent with the rest of

1254: the D and F term constraints.  Thus we see that the D term equation

1255: for the difference of two vevs is interpreted in M theory as the

1256: following statement.  If the connected components of the M5 brane

1257: slide apart in the $w$ plane along the $D6$ branes bisecting the

1258: $\tilde{N}_c-r$ color branes then, because the components are rigid,

1259: they also separate along the $w$ plane at the semi-infinite flavor

1260: branes.  In other words by trading a monopole vev for a magnetic quark

1261: vev, we find vacua which preserve flavor symmetry.  These two

1262: distances, whose differences are preserved, are marked with

1263: double-headed arrows in Fig.~\ref{brootfig}.

1264: 

1265: To count vacua on this branch, consider the dual SU($\tilde{N}_c$)

1266: theory whose gauge symmetry is broken by the adjoint scalar vevs of

1267: the $r$ color branes attached to flavor branes.  As in the

1268: semiclassical case, the remaining $\tilde{N}_c-r$ color branes form a

1269: line which can have $\tilde{N}_c-r$ orientations preserving the M

1270: theory coordinate asymptotically far away, yielding a multiplicity of

1271: $\tilde{N}_c-r$ times the combinatoric factor of $(^{N_f}_{\ r})$ from

1272: the choice of which flavor branes to connect.  In all, this provides

1273: \begin{equation}

1274: \N_2=\sum_{r=0}^{\tilde{N}_c}(\tilde{N}_c-r)(^{N_f}_{\

1275:   r})=\N_{sc}-\N_1

1276: \end{equation}

1277: states, and thus completes the classification of vacua.

1278: 

1279: \section{Conclusion}

1280: 

1281: We have constructed a new realization of Seiberg duality that relies

1282: on an energy scale dependent reduction of M theory to IIA.  We have

1283: found the M2 branes that correspond to flavored magnetic monopoles and

1284: argued that they correspond to magnetic baryons in the dual magnetic

1285: theory, which in turn decay to magnetic quarks.  And finally we have

1286: interpreted baryon vevs as the relative sliding of two halves of an M

1287: theory configuration along a Taub-NUT singularity.

1288: 

1289: As an application of the above constructions, we have reproduced the

1290: field theory results of \cite{Murayama}.  In particular we have

1291: correctly reproduced the flavor symmetry breaking patterns, the order

1292: parameters of the symmetry breaking and the counting of states in

1293: various regimes. We can interpret these countings in terms of discrete

1294: rotations of a line of M5 brane in the $v$ plane. 

1295: 

1296: The case $r=N_f/2$ is difficult to analyze using traditional field

1297: theory techniques, as it is superconformal and strongly coupled in the

1298: IR.  However it is possible that by deforming the corresponding curve,

1299: an analysis similar to the Seiberg duality of monopoles above may be

1300: possible in this M theory setting.

1301: 

1302: A natural next step is to add orientifold planes and try to repeat

1303: this analysis for SO($N$) and SP($N$) gauge theories.

1304: 

1305: 

1306: \noindent {\bf Acknowledgements}

1307: 

1308: 

1309: We would like to thank Dan Brace, Gordon Chalmers, Sergei Cherkis, Jan

1310: de Boer, Scott Thomas, Angel Uranga, Ed Witten and Bruno Zumino for

1311: enlightening conversations.  The work of HM and UV was supported in

1312: part by the Director, Office of Science, Office of High Energy and

1313: Nuclear Physics, Division of High Energy Physics of the U.S.

1314: Department of Energy under Contract DE-AC03-76SF00098 and in part by

1315: the National Science Foundation under grant PHY-95-14797.

1316: 

1317: \appendixa

1318: 

1319: \subsection{An Example, SU(3) with $N_f =4$} \label{Example}

1320: 

1321: We will consider in some detail the M5 branes associated with the

1322: $r$-vacua which survive breaking to $\N=1$ for the case of SU(3) with

1323: $N_f=4$ and equal quark masses, $m_i=m$. This case is rather

1324: special in that the coordinates of these vacua in the moduli

1325: space can be analytically determined as a function of mass. Thus, we

1326: can explicitly draw out the corresponding M5 brane configurations and

1327: test our claims for the brane interpretations of the $r$ vacua in

1328: various regimes. 

1329: 

1330: In the case of equal and large bare quark masses, we can clearly see

1331: that the M5 brane configurations which survive breaking to $\N=1$ in

1332: Fig.~\ref{semicm5} do indeed correspond to $r$ flavor branes and color

1333: branes connected to each other.

1334: \begin{figure}[htbp]

1335: %\begin{center}

1336: \centering

1337: \begin{minipage}[c]{0.50\textwidth}

1338: {(a) The baryonic root}

1339: \centering

1340: \includegraphics[width=3in]{su3nf4brm=20.eps}

1341: \end{minipage}%

1342: \begin{minipage}[c]{0.50\textwidth}

1343: {(b) A non-baryonic r=0 root}

1344: \centering

1345: \includegraphics[width=3in]{su3nf4r0bm=20.eps}

1346: \end{minipage}

1347: \begin{minipage}[c]{0.50\textwidth}

1348: {(c) A non-baryonic r=1 root}

1349: \centering

1350: \includegraphics[width=3in]{su3nf4r1bm=20.eps}

1351: \end{minipage}%

1352: \begin{minipage}[c]{0.50\textwidth}

1353: {(d) The non-baryonic r=2 root}

1354: \centering

1355: \includegraphics[width=3in]{su3nf4r2m=20.eps}

1356: \end{minipage}

1357: %\end{center}

1358: \caption{The M5 brane configurations corresponding to $r$-vacua in SU(3)

1359:   with 4 flavors in the semiclassical limit, i.e. with $m_i = m \gg

1360:   \Lambda $. The flavor branes and color branes which have connected

1361:   are represented by semi-infite tubes extending from the left branch

1362:   of the M5 brane. The connections between the two NS5 branes are the

1363:   remaining color branes, pairs of which have condensed massless

1364:   monopoles between them.}  \label{semicm5}

1365: \end{figure}

1366: In the massless case, the relevant curves are more difficult to

1367: interpret, though we include them for completeness in Fig.~\ref{m0m5}.

1368: Note that in the limit that $m \rightarrow 0$, the $r=1$ non-baryonic

1369: root and the baryonic root converge to the same point in moduli space,

1370: the $m=0$ baryonic root.  Just as in the superconformal case, it is

1371: easier to interpret these brane configuration if we consider their

1372: cross-sections in Figs.~\ref{m0m5csr0} and \ref{m0m5csbr}, though we

1373: will leave the interpretation of these cross-sections for future work.

1374: \begin{figure}[htbp]

1375: %\begin{center}

1376: \centering

1377: \begin{minipage}[c]{0.333\textwidth}

1378: {(a) The baryonic root}

1379: \centering

1380: \includegraphics[width=2.2in]{su3nf4brm=0.eps}

1381: \end{minipage}%

1382: \begin{minipage}[c]{0.333\textwidth}

1383: {(b) An r=0 vacuum}

1384: \centering

1385: \includegraphics[width=2.2in]{su3nf4r0bm=0.eps}

1386: \end{minipage}%

1387: \begin{minipage}[c]{0.333\textwidth}

1388: {(c) The r=2 vacuum}

1389: \centering

1390: \includegraphics[width=2.2in]{su3nf4r2m=0.eps}

1391: \end{minipage}

1392: %\end{center}

1393: \caption{M5 brane configurations at the r vacua in SU(3) with 4 flavors

1394:   and $\forall m_i =0$.}

1395:   \label{m0m5}

1396: \end{figure}

1397: 

1398: \begin{figure}[htbp]

1399: \begin{center}

1400: \begin{minipage}[c]{.25\textwidth}

1401: \centering

1402:   \includegraphics[width=1.65in]{su3nf4r0z=-1.00.eps}

1403: \end{minipage}%

1404: \begin{minipage}[c]{.25\textwidth}

1405: \centering

1406:   \includegraphics[width=1.45in]{su3nf4r0z=-0.50.eps}

1407: \end{minipage}%

1408: \begin{minipage}[c]{.25\textwidth}

1409: \centering

1410:   \includegraphics[width=1.35in]{su3nf4r0z=0.00.eps}

1411: \end{minipage}

1412: \begin{minipage}[c]{.25\textwidth}

1413: \centering

1414: \includegraphics[origin=c,width=1.25in]{su3nf4r0z=1.00.eps}

1415: \end{minipage}%

1416: \begin{minipage}[c]{.25\textwidth}

1417: \centering

1418:   \includegraphics[origin=c,width=0.75in]{su3nf4r0z=3.00.eps}

1419: \end{minipage}%

1420: \begin{minipage}[c]{.25\textwidth}

1421: \centering

1422:   \includegraphics[origin=c,width=1.0in]{su3nf4r0z=7.00.eps}

1423: \end{minipage}

1424: \end{center}

1425: \caption{Cross sections at constant $x^6$ of the M5 brane of Fig.~\ref{m0m5}b (the r=0 root).} \label{m0m5csr0}

1426: \end{figure}

1427: 

1428: \begin{figure}[htbp]

1429: \begin{center}

1430: \begin{minipage}[c]{.25\textwidth}

1431: \centering

1432:   \includegraphics[width=1.5in]{su3nf4brz=-0.60.eps}

1433: \end{minipage}%

1434: \begin{minipage}[c]{.25\textwidth}

1435: \centering

1436:   \includegraphics[width=1.5in]{su3nf4brz=-0.05.eps}

1437: \end{minipage}%

1438: \begin{minipage}[c]{.25\textwidth}

1439: \centering

1440:   \includegraphics[width=1.5in]{su3nf4brz=0.50.eps}

1441: \end{minipage}%

1442: \begin{minipage}[c]{.25\textwidth}

1443: \centering

1444:   \includegraphics[width=1.5in]{su3nf4brz=2.00.eps}

1445: \end{minipage}%

1446: \end{center}

1447: \caption{Cross sections at constant $x^6$ of the M5 brane of

1448:   Fig.~\ref{m0m5}a (the baryonic root).} \label{m0m5csbr}

1449: \end{figure}

1450: 

1451: \bibliographystyle{ieeetr} 

1452: \bibliography{m5}

1453: \end{document}

1454: 

1455: 

1456: 

1457: %%% Local Variables: 

1458: %%% mode: latex

1459: %%% TeX-master: t

1460: %%% TeX-master: t

1461: %%% End: 

1462: