1: \documentclass[a4paper,12pt]{article}

2: \usepackage{epsfig}

3: \usepackage{amssymb}

4: \usepackage{amsfonts}

5: %\usepackage[english]{babel}

6: \usepackage{graphics}

7: \usepackage{color,amsmath,bm}

8: %\usepackage[dvips]{graphicx}

9: 

10: \newskip\humongous \humongous=0pt plus 1000pt minus 1000pt

11: \def\caja{\mathsurround=0pt}

12: \def\eqalign#1{\,\vcenter{\openup1\jot \caja

13:  \ialign{\strut \hfil$\displaystyle{##}$&$

14:  \displaystyle{{}##}$\hfil\crcr#1\crcr}}\,}

15: \newif\ifdtup

16: \def\panorama{\global\dtuptrue \openup1\jot \caja

17:  \everycr{\noalign{\ifdtup \global\dtupfalse

18:  \vskip-\lineskiplimit \vskip\normallineskiplimit

19:  \else \penalty\interdisplaylinepenalty \fi}}}

20: \def\eqalignno#1{\panorama \tabskip=\humongous

21:  \halign to\displaywidth{\hfil$\displaystyle{##}$

22:  \tabskip=0pt&$\displaystyle{{}##}$\hfil

23:  \tabskip=\humongous&\llap{$##$}\tabskip=0pt

24:  \crcr#1\crcr}}

25: 

26: 

27: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

28: 

29: \jot = 1.5ex

30: \def\baselinestretch{1.2}

31: \parskip 5pt plus 1pt

32: 

33: \catcode`\@=11

34: 

35: %%%%%%%%%%%%%% This causes equations to be numbered by section

36: 

37: \@addtoreset{equation}{section}

38: %\def\theequation{\arabic{section}.\arabic{equation}}

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

40: 

41: \def\@normalsize{\@setsize\normalsize{15pt}\xiipt\@xiipt

42: \abovedisplayskip 14pt plus3pt minus3pt%

43: \belowdisplayskip \abovedisplayskip

44: \abovedisplayshortskip \z@ plus3pt%

45: \belowdisplayshortskip 7pt plus3.5pt minus0pt}

46: 

47: \def\small{\@setsize\small{13.6pt}\xipt\@xipt

48: \abovedisplayskip 13pt plus3pt minus3pt%

49: \belowdisplayskip \abovedisplayskip

50: \abovedisplayshortskip \z@ plus3pt%

51: \belowdisplayshortskip 7pt plus3.5pt minus0pt

52: \def\@listi{\parsep 4.5pt plus 2pt minus 1pt

53:      \itemsep \parsep

54:      \topsep 9pt plus 3pt minus 3pt}}

55: 

56: \relax

57: 

58: 

59: \catcode`@=12

60: 

61: %%%%%%%%%%%%%%%%%%%%%%%%% set page size

62: \evensidemargin 0.0in \oddsidemargin 0.0in \textwidth 6.0in

63: \textheight 8.5in \hoffset .7 cm \voffset -1 cm \headsep .75in

64: 

65: %%%%%%%%%%%%%%%%%%%%%%%%%% reset section commands

66: 

67: 

68: \catcode`\@=11

69: 

70: \def\section{\@startsection{section}{1}{\z@}{3.5ex plus 1ex minus

71:    .2ex}{2.3ex plus .2ex}{\large\bf}}

72: 

73: \def\thesection{\arabic{section}}

74: \def\thesubsection{\arabic{section}.\arabic{subsection}}

75: \def\thesubsubsection{\arabic{section}.\arabic{subsection}.\arabic{subsubsection}}

76: \def\appendix{\setcounter{section}{0}

77: \def\thesection{\Alph{section}}}

78: % \def\thesubsection{\Alph{section}.\arabic{subsection}}

79: % \def\theequation{\Alph{section}.\arabic{equation}}}

80: 

81: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

82: %%%%%%%%%%%%%%%%%% end of mypaper.sty %%%%%%%%%%%%%%%%%%%%%%%%%

83: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

84: 

85: 

86: 

87: 

88: 

89: \begin{document}

90: 

91: 

92: 

93: %%%%%%Define some new commands and  macros

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

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

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

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

98: \newcommand{\beas}{\begin{eqnarray*}}

99: \newcommand{\eeas}{\end{eqnarray*}}

100: \newcommand{\defi}{\stackrel{\rm def}{=}}

101: \newcommand{\non}{\nonumber}

102: \newcommand{\bquo}{\begin{quote}}

103: \newcommand{\enqu}{\end{quote}}

104: \newcommand{\mat}{\mathbf}

105: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% definitions

106: \def\d{\mathrm{d}}

107: \def\de{\partial}

108: \def\Tr{ \hbox{\rm Tr}}

109: \def\const{\hbox {\rm const.}}

110: \def\o{\over}

111: \def\im{\hbox{\rm Im}}

112: \def\re{\hbox{\rm Re}}

113: \def\bra{\langle}\def\ket{\rangle}

114: \def\Arg{\hbox {\rm Arg}}

115: \def\Re{\hbox {\rm Re}}

116: \def\Im{\hbox {\rm Im}}

117: \def\diag{\hbox{\rm diag}}

118: \def\longvert{{\rule[-2mm]{0.1mm}{7mm}}\,}

119: \def\Z{\mathbb Z}

120: \def\N{{\cal N}}

121: \def\tq{{\widetilde q}}

122: \def\W{{\cal W}}

123: \def\tQ{{\widetilde Q}}

124: \def\dag{{}^{\dagger}}

125: \def\p{{}^{\,\prime}}

126: \def\a{\alpha}

127: \def\Tr{ \hbox{\rm Tr}}

128: \def\tM{{\widetilde M}}

129: \def\tm{{\widetilde m}}

130: \def\T{{\cal T}}

131: \def\t{T}

132: \def\J{{\cal J}}

133: \def\V{{\sf V}}

134: \def\lcm{\mathrm{lcm}}

135: 

136: \begin{titlepage}

137: \begin{flushright}

138: hep-th/0507273\\

139: \end{flushright}

140: 

141: \bigskip

142: 

143: 

144: 

145: \begin{center}

146: {\Large

147: 

148: {\bf Domain Walls and Flux Tubes }

149:  }

150: \end{center}

151: 

152: 

153: \renewcommand{\thefootnote}{\fnsymbol{footnote}}

154: \bigskip

155: \begin{center}

156: {\large   Stefano Bolognesi }

157:  \vskip 0.20cm

158: \end{center}

159: 

160: 

161: \begin{center}

162: {\it      \footnotesize

163: Scuola Normale Superiore - Pisa, Piazza dei Cavalieri 7, Pisa, Italy \\

164: \vskip 0.10cm

165: and\\

166: \vskip 0.10cm

167: Istituto Nazionale di Fisica Nucleare -- Sezione di Pisa, \\

168: Via Buonarroti 2, Ed. C, 56127 Pisa,  Italy   \\  } \vskip 0.15cm

169: s.bolognesi@sns.it\\

170: \end {center}

171: 

172: 

173: \setcounter{footnote}{0}

174: 

175: \bigskip

176: \bigskip

177: 

178: 

179: 

180: 

181: \noindent

182: \begin{center} {\bf Abstract} \end{center}

183: 

184: We present a new vortex solution made of a domain wall

185: compactified into a cylinder and stabilized by the magnetic flux

186: within. When the thickness of the wall is much less than the

187: radius of the vortex some precise results can be obtained, such as

188: the tension spectrum and profile functions. This vortex can

189: naturally end on the wall that has created it, making the simplest

190: junction between a wall and a vortex. We then classify every kind

191: of junction between a flux tube and domain wall. The criteria for

192: classification are as follows: a flux can or can not end on the

193: wall, and when it ends, the flux  must go somewhere. Various

194: examples are discussed, including abelian and nonabelian theories,

195: as well as supersymmetric and non-supersymmetric theories.

196: 

197: 

198: 

199: 

200: 

201: 

202: 

203: \vfill

204: 

205: 

206: 

207: \begin{flushleft}

208: July, 2005

209: \end{flushleft}

210: 

211: 

212: 

213: \end{titlepage}

214: 

215: \bigskip

216: 

217: \hfill{}

218: 

219: 

220: \tableofcontents

221: 

222: 

223: 

224: 

225: \section{Introduction}

226: 

227: 

228: We present a new kind of vortex which is formed by a wall. It

229: occurs whenever a theory has two degenerate vacua, one in the

230: Coulomb phase, and one in the Higgs phase. The ordinary vortex in

231: the Higgs vacuum \cite{A,NO} can be thought of as the domain wall

232: interpolating between the two vacua compactified into a cylinder.

233: The rolled wall is stabilized by the magnetic field

234: inside.\footnote{This is reminiscent of the relation between

235: $Q$-kinks and $Q$-lumps discussed in \cite{Abraham:1992qv}.}

236: 

237: 

238: %%%%%%%%%%%%%%%%%%%%%%%%The-wall-limit%%%%%%%%%%%%%%%%%%%%%%%

239: We then study the extent to which the thickness of the wall

240: $\Delta_W$ is much less than the radius of the vortex $R_V$. This

241: limit is reached for a large number of quanta of the magnetic

242: flux. Within these limits the spectrum of vortices is simple to

243: compute by means of classical arguments.  They are particular kind

244: of type I superconducting vortices with a tension that scales like

245: $n^{2/3}$. In addition, the profile functions of the vortex are

246: exactly computable within this limit.

247: 

248: 

249: 

250: %%%%%%%%%%%%%%%%%%%%%%%Coulomb-junction%%%%%%%%%%%%%%%%%%%%

251: This kind of flux tube naturally ends on the wall that has created

252: it. When a wall vortex ends on its wall, the magnetic flux it

253: carries is spread radially through the Coulomb vacuum on the other

254: side of the wall. The point on the wall at which the vortex ends

255: is seen from the Coulomb vacuum as a monopole of double charge.

256: The junction between the wall and the wall vortex can be thought

257: of as the final stage of a process in which a monopole in the

258: Coulomb phase is pushed against the wall.

259: 

260: 

261: %%%%%%%%%%%%%%%%%%%%non-abelian%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

262: This kind of wall vortex arises  in the nonabelian gauge theories

263: every time a domain wall interpolates between two vacua with

264: different confinement indices.

265: 

266: 

267: 

268: %%%%%%%%%%%%%%%%%%%%%%Classification-of-junctions%%%%%%%%%%%%%%%%%%%%%

269: We then classify the junctions between domain walls and flux

270: tubes. The  previously discussed junction is only one particular

271: variety, and perhaps the most natural.  The two basic criteria for

272: distinguish junctions are:

273: \begin{itemize}

274: \item The tube can or can not end on the wall. \item  Where the

275: flux goes.

276: \end{itemize}

277: 

278: 

279: 

280: %%%%%%%%%%%%%%%%%%%%%%%%%%D-brane-like%%%%%%%%%%%%%%%%%%%%

281: When the vortex ends on the wall and spread its flux across the

282: $2$-dimensional surface of the wall is where we find a D-brane

283: junction. This example has been widely studied in recent years

284: \cite{Gauntlett:2000de,Shifman:2002jm,nahiggs}. The driving force

285: behind these studies was, in fact, to find some field theoretical

286: analogues of the D-brane physics in string theory. This variety of

287: wall, in order to completely resemble a D-brane, must also

288: supports massless gauge fields in the effective low-energy action.

289: 

290: 

291: 

292: %%%%%%%%%%%%%%%%%%%%%%%%%%%Bound-state%%%%%%%%%%%%%%%%%%%%%%%%%%

293: Another type of junction is found when the string ends on the wall

294: and the flux is captured by a particle confined within the wall.

295: In this case, there is no localization of the gauge field inside

296: the wall. The first example has been found in \cite{Witten:1997ep}

297: where it was shown that the $\mathbb{Z}_N$ strings of pure $SU(N)$

298: $\N=1$ super Yang-Mills can end on the domain wall that

299: interpolates between two chirally adjacent vacua.

300: 

301: 

302: 

303: %%%%%%%%%%%%%%%%%%%%%%%%%%%%Crossing-junction%%%%%%%%%%%%%%%%%%%

304: The last type is a cross-junction, which is characterized by the

305: tube, rather than ending on the wall, crossing it instead and

306: becoming another flux tube in the opposite vacuum.  We can find

307: examples of it in pure $SU(N)$ $\N=1$ SYM when the domain wall

308: interpolates between two vacua whose phase shift has a common

309: divisor with $N$.

310: 

311: 

312: 

313: %%%%%%%%%%%%%%%%%%%%%%%%%All-junctions%%%%%%%%%%%%%%%%%%%%%%%

314: Finally we put our classification to the test in a particular

315: theory. In $\N=2$ pure SYM broken to $\N=1$ by a generic

316: superpotential for the adjoint field, there are a lot of vacua

317: with different confinement index \cite{Cachazo:2002zk}. We provide

318: examples in which all three types  of junctions are presented

319: simultaneously.

320: 

321: %%%%%%%%%%%%%%%%%%%%%%%%%Paper-organization%%%%%%%%%%%%%%%%%%%%%%%

322: The paper is divided into two parts.  Section \ref{made} is

323: devoted to the study of the  wall vortex. In  \ref{abelian} we

324: consider the abelian gauge theory and its supersymmetric

325: extension. In \ref{naexistence} we study the wall vortex in

326: nonabelian gauge theories. In  Section \ref{classification} we

327: classify the domain wall/flux tube junctions. In \ref{local} we

328: consider a flux tube which ends on the wall, and in

329: \ref{crossubsection}, a flux tube which crosses the wall. Finally,

330: in  \ref{super} we give an example in which every junction studied

331: in the paper is presented simultaneously. In Appendix

332: \ref{yappendix} we provide a mechanism to obtain a flux tube/flux

333: tube junction, and we use it to build a Y junction for baryons in

334: $SU(3)$.

335: 

336: 

337: 

338: 

339: \section{The Vortex Formed by a Wall \label{made}}

340: 

341: We consider a flux tube that can be thought as made of wall. The

342: simplest example in which it can arise is a $U(1)$ gauge theory

343: that has two degenerate vacua: one in the Coulomb phase and the

344: other in the Higgs phase. Consider the domain wall of tension

345: $T_W$ that interpolates between these two vacua. We can build a

346: flux tube rolling the wall in a cylinder of radius $R$, keeping

347: the Coulomb phase  inside the tube and turning on a magnetic flux

348: inside (see Figure \ref{springroll}). \begin{figure}[h!t]

349: \begin{center}

350: \leavevmode \epsfxsize 11 cm \epsffile{springroll.eps}

351: \end{center}

352: \caption{\footnotesize The wall vortex. A wall of thickness

353: $\Delta_W$ is compactified on a circle of radius $R_V$ and

354: stabilized by the magnetic field inside.} \label{springroll}

355: \end{figure} The tension of the wall

356: gives a contribution $T_W 2\pi R$ to the energy density. The

357: magnetic flux is $\Phi_B=B \pi R^2$ where $B$ is the magnetic

358: field. Varying the radius $R$ the flux $\Phi_B$ must remain

359: constant, so the contribution of the flux to the energy density is

360: $\Phi_B / 2\pi R^2$.  The magnetic flux is quantized in integer

361: values: $\Phi_B=2\pi n/e$, where $e$ is the coupling constant. The

362: tension of the tube is the sum of two pieces, one that comes from

363: the flux and the other that comes from the wall: \beq \label{qual}

364: T(R)=\frac{{\Phi_B}^2}{2 \pi R^2}+T_W 2 \pi R \ .\eeq  The stable

365: configuration is the one that minimizes the tension: \beq

366: \label{eqscal} R_V = \sqrt[3]{2} \; \sqrt[3]{\frac{n^{2}}{ e^{2}

367: {T_W}}} \ , \qquad T_V=3 \sqrt[3]{2} \pi \; \left(\frac{n

368: T_W}{e}\right)^{2/3}  \ .\eeq This result can be trusted only when

369: the thickness of the wall $\Delta_W$ is much less than the radius

370: of the vortex.  In general the profile of the vortex will be a

371: mixture of the magnetic field and the wall. In any case this kind

372: of flux tube can be though as made of the same stuff of the wall.

373: 

374: 

375: 

376: 

377: 

378: 

379: 

380: \subsection{Abelian Higgs-Coulomb model \label{abelian}}

381: 

382: Now we are going to consider an explicit example. The simplest

383: that we can imagine is a $U(1)$ gauge theory coupled to a charged

384: scalar $q$

385: \begin{equation}

386: \label{BasicVortex} {\cal

387: L}=-\frac{1}{4}F_{\mu\nu}F^{\mu\nu}-\frac12

388: |(\de_{\mu}-ieA_{\mu})q|^2-V(|q|)\ ,

389: \end{equation}

390: and a suitable potential (see Figure \ref{potential}) that has two

391: degenerate  minima, one in the Coulomb phase $q=0$ and the other

392: in the Higgs phase $|q|=q_0$. \begin{figure}[h!t]

393: \begin{center}

394: \leavevmode \epsfxsize 9 cm \epsffile{potential.eps}

395: \end{center}

396: \caption{\footnotesize A potential with two degenerate vacua.

397: $q=0$ is in the Coulomb phase while $|q|=q_0$ is in the Higgs

398: phase.} \label{potential}

399: \end{figure}

400: This theory admits a kink that interpolates between the two vacua.

401: The profile of the kink, $q(z)$, is constant in $x,y$ and has

402: boundary conditions $q(-\infty)=0$ and $q(+\infty)=q_0$. The

403: profile $q(z)$ has constant phase since the potential is a real

404: function that depends only on $|q|$. In the Higgs phase there is a

405: vortex obtained by choosing an element $n$ of the homotopy group

406: $[n]\in \pi_1(\mathbf{S}^1)$: \bea \label{vortex}  q &=& e^{i n

407: \theta} q(r) \ , \\  A_{\theta} &=& \frac{n}{e r} A(r) \ ,

408: \nonumber \eea where we have used cylindrical coordinates

409: $r,\theta,z$ instead of $x,y,z$. Clearly the wall and the tube are

410: continuously related. Consider a kink in the radial direction at a

411: radius much greater than its thickness. We can roll it around the

412: axe $z$ giving it a gauge rotation $e^{in\theta}$. This

413: automatically turns on a pure gauge field at large distance that,

414: for continuity, must have a flux inside. This configuration in

415: general is not a minimum of the energy density, so it will start

416: to loose energy until it reach the stable vortex.

417: 

418: 

419: 

420: 

421: \subsubsection{The wall limit \label{wallim}}

422: 

423: 

424: The differential equations for the profile functions of the vortex

425: (\ref{vortex}) are \bea \label{eqdiff} && \frac{\d^2 q}{\d

426: r^2}+\frac{1}{r}\frac{\d q}{\d r}-n^2

427: \frac{(1-A)^2}{r^2}q-\frac{\delta V}{\delta q}=0 \ , \\ &&

428: \frac{\d^2 A}{\d r^2}-\frac{1}{r}\frac{\d A}{\d r}+e^2(1-A)q^2=0 \

429: , \nonumber \eea where $n$ is the winding number. We are looking

430: for some limit of the parameters so that the vortex really looks

431: like a rolled wall. In this limit the profile functions should be:

432: \bea  \label{theta} q(r) &=& q_0 \theta_H(r-R_V)\ ,  \\

433: A(r)&=&r^2/R_V^2  \qquad  0 \leq r \leq R_V \ , \nonumber \\

434: A(r)&=&1\  \qquad  r \geq R_V \ , \nonumber \eea  where $\theta_H$

435: is the step function. First of all we manipulate a bit the

436: differential equations (\ref{eqdiff}) to simplify them.  The

437: potential can be written as a dimensionless function \beq V(q)=v_0

438: \V \left(\frac{q}{q_0}\right)\ , \eeq where $v_0$ is the scale of

439: the potential and $q_0$ the vev. In the following we will not use

440: the shape of the dimensionless potential $\V$. The only important

441: thing is that $\V(0)=\V(1)=0$ and its height is of order $1$ (the

442: simplest example is $\V(\chi)=\chi^2(\chi^2-1)^2$). We also scale

443: the scalar field $q=q_0 \chi$. After these scalings the equations

444: (\ref{eqdiff}) for the profiles becomes: \bea \label{eqscal1} &&

445: \frac{\d^2 \chi}{\d r^2}+\frac{1}{r}\frac{\d \chi}{\d r}-n^2

446: \frac{(1-A)^2}{r^2} \chi- \alpha \frac{\delta \V}{\delta \chi}=0 \

447: , \\  \label{eqscal2} && \frac{\d^2 A}{\d r^2}-\frac{1}{r}\frac{\d

448: A}{\d r}+ \beta  (1-A)\chi^2=0 \ . \eea So there are three

449: parameters that enter the game: \beq n\ , \qquad

450: \alpha=\frac{v_0}{{q_0}^2}\ , \qquad \beta=e^2 {q_0}^2\ .\eeq The

451: domain wall  tension and  thickness are respectively:\footnote{For

452: this qualitative result it is sufficient to take the wall

453: lagrangian ${\cal L}=-\de q \de q-V(q)$ and then write the tension

454: as a function of the thickness $ T(\Delta) \sim

455: \frac{q_0}{\Delta}+v_0 \Delta$. Minimizing $T(\Delta)$ with

456: respect to $\Delta$ we find (\ref{qualitative}).}

457:  \beq \label{qualitative} T_W \sim  q_0 \sqrt{v_0}\ , \qquad \Delta_W

458: \sim  \frac{q_0}{\sqrt{v_0}} \ .\eeq

459: 

460: 

461: 

462: Now comes the first non trivial hint. If the wall limit exists,

463: then formula (\ref{eqscal}) can be trusted in this limit. But the

464: radius of the vortex $R_V$ comes out from equations (\ref{eqdiff})

465: and so must depend only on the three relevant parameters

466: $n,\alpha,\beta$. In general a function of $n,e,v_0,q_0$ cannot be

467: expressed as a function of $n,\alpha,\beta$, but for

468: (\ref{eqscal}) this is possible: \beq R_V \sim \frac{n^{2/3}}

469: {\alpha^{1/6} \beta^{1/3}} \ . \eeq If we wouldn't have found such

470: expression, we would have concluded that the wall limit doesn't

471: exist. This result encourages us to go on.

472: 

473: 

474: 

475: 

476: Now we discuss the wall limit. In this paragraph we look for a

477: limit in which the radius of the vortex remain constant and the

478: solution approaches the wall vortex (\ref{theta}) (our goal is

479: described in Figure \ref{limite}). \begin{figure}[h!t]

480: \begin{center}

481: \leavevmode \epsfxsize 14 cm \epsffile{limite.eps}

482: \end{center}

483: \caption{\footnotesize Increasing the winding number $n$ and

484: keeping fixed the radius of the vortex $R_V$, the profile

485: functions approach the wall vortex.} \label{limite}

486: \end{figure}

487: The $\chi$ profile must become a step function: it is zero inside

488: $R_V$, it goes from zero to one in a distance $\Delta_W$, and then

489: remains constant at one.  Thus $\chi''$ in equation

490: (\ref{eqscal1}) develops a $\delta'(r-R_V)$ singularity or,

491: thinking in terms of $\Delta_W$, $\chi'' \sim 1/{\Delta_W}^2$.

492: Thus to counterbalance this divergence in (\ref{eqscal1}) we must

493: have $\alpha$ that goes to infinity like $1/{{\Delta_W}^2}$.  The

494: encouraging fact is that $\alpha \frac{\delta \V}{\delta \chi}$

495: really resembles a $\delta'(r-R_V)$ in this limit (note that this

496: is true only if the two vacua are exactly degenerate).  Now

497: consider the second equation (\ref{eqscal2}) where $A''(R_V)$ has

498: a $(2 / R_V) \delta(r-R_v) $ singularity, or thinking in terms of

499: $\Delta_W$, $A''(R_V) \sim 1/(R_V \Delta_W)$. Since $(1-A)\chi$ is

500: of order $\Delta_W / R_v$ around $R_v$, we must also send $\beta$

501: to infinity like $1/{\Delta_W}^2$.  Now we can make the conjecture

502: of the wall limit.

503: \newline {\bf Mathematical Conjecture:}  { \it Consider the

504: succession of parameters $\alpha_n=n^{4/3} \alpha_1$ and $

505: \beta_n=n^{4/3} \beta_1$ and call the solution of (\ref{eqscal1})

506: and (\ref{eqscal2}) with the vortex boundary conditions,

507: $\chi_{n,\alpha_n,\beta_n}(r)$ and $A_{n,\alpha_n,\beta_n}(r)$. In

508: the limit $n \to \infty$ \bea && \lim_{n \to \infty}

509: \chi_{n,\alpha_n,\beta_n}(r) \to \theta_H(r-R_v)\ , \\ && \lim_{n

510: \to \infty} A_{n,\alpha_n,\beta_n}(r) \to \left\{

511: \begin{array}{cc} r^2/{R_V}^2 & 0 \leq r \leq R_V \ , \\

512: 1 & r>R_V  \ . \\ \end{array}\right. \nonumber \eea } \newline

513: This limit has been chosen so that the radius of the vortex remain

514: constant $R_V \sim n^{2/3} {\alpha_n}^{-1/3}

515: {\beta_n}^{-1/6}={\alpha_1}^{-1/3} {\beta_1}^{-1/6}$ and also the

516: ratio $\alpha_n / \beta_n$ remains constant. The ratio $\alpha /

517: \beta$ disappears in the wall limit and is related only to the

518: shape of the limiting functions $\chi_{n,\alpha_n,\beta_n}(r)$ and

519: $A_{n,\alpha_n,\beta_n}(r)$. Probably a stronger version of the

520: conjecture is  true:  the ratio $\alpha_n /\beta_n$ is kept

521: limited from above and from below during the limit, so that it

522: doesn't go neither to infinity nor to zero.

523: 

524: 

525: 

526: 

527: The limit discussed above has been called mathematical since it is

528: easy to express in a mathematical language. In this paragraph we

529: discuss a more physical situation in which we keep fixed the

530: parameters of the theory and we only change the winding number

531: $n$. When $n$ is increased the radius of the vortex becomes large

532: while the thickness of the wall remains constant since it doesn't

533: depend on $n$. We are going to use (\ref{eqscal}) in an self

534: consistent way. Suppose that $n$ is enough large so that

535: (\ref{eqscal}) is true, then we use the expression of $R_V$ to

536: obtain the condition for $n$ so that $R_V \gg \Delta_W$.

537: \newline {\bf Physical Conjecture:} {\it We keep $\alpha$ and $\beta$ fixed

538: and increase $n$. The condition $R_V \gg \Delta_W$ is $n \gg

539: (\beta/\alpha)^{1/4}$. When this is satisfied the vortex resembles

540: a wall vortex and its tension is given by \beq \label{spectrum}

541: T_V=3  \sqrt[3]{2} \pi \; \left(\frac{n {T_W}}{e}\right)^{2/3} \ .

542: \eeq}

543: 

544: 

545: 

546: 

547: 

548: \subsubsection{Pushing a monopole against the wall}

549: 

550: There is another intuitively way in which we can see the wall

551: continuously transformed into the vortex. Consider a monopole in

552: the Coulomb phase. The lines of its magnetic flux will be tangent

553: to the wall surface. Now we move the monopole in the wall

554: direction as in Figure \ref{push}. \begin{figure}[h!t]

555: \begin{center}

556: \leavevmode \epsfxsize 11 cm \epsffile{push.eps}

557: \end{center}

558: \caption{\footnotesize A monopole in the Coulomb phase is pushed

559: against that domain wall. At the end of the process we are left

560: with a monopole in the Higgs phase confined on a vortex that ends

561: on the wall.} \label{push}

562: \end{figure}

563: The wall is repelled and change its shape. Far from the $x,y$

564: coordinate of the monopole the shape of the wall is logarithmic $z

565: \propto \log{(x^2+y^2)}$. If we continue to push the monopole in

566: the negative $z$ the configuration will resemble more and more a

567: monopole in Higgs phase, attached to a flux tube that ends on the

568: wall. This physical picture well explains a wall that is

569: continuously transformed into a vortex. It's also clear that this

570: kind of vortices can naturally end on the wall that has created

571: them since they are made of the same stuff.

572: 

573: 

574: 

575: A more detailed and quantitative analysis of this junction will

576: appear in \cite{ioeken}.

577: 

578: 

579: 

580: 

581: 

582: 

583: 

584: 

585: 

586: \subsubsection{$\N=2$ super QED broken to $\N=1$ \label{susy}}

587: 

588: Now we consider $\N =2$ SQED broken to $\N=1$ by a superpotential

589: for the adjoint field.  The $U(1)$ gauge multiplied is composed by

590: the superfields $W_{\alpha}$ and $\Phi$, while the matter

591: superfields are $Q$ of charge $+1$ and $\tQ $ of charge $-1$. The

592: Lagrangian is the following: \bea \label{classicalSQED}

593: {\cal L}&=&\int d^2\theta \, \frac{1}{4e^2}W^{\a}W_{\a}+h.c.\\

594: &&+\int d^2\theta d^2\bar{\theta} \, (\frac{1}{e^2}\Phi\dag\Phi+Q\dag e^{V}Q+{\widetilde Q}\dag e^{-V}{\widetilde Q})\nonumber\\

595: &&+\int d^2\theta \, \sqrt{2} ({\widetilde Q}\Phi Q-m{\widetilde

596: Q}Q+W(\Phi))+h.c.\ .\nonumber \eea The potential for the scalar

597: fields is \beq \label{pot}

598: V=2|(\phi-m)q|^2+2|(\phi-m)\tq|^2+2e^2|\tq

599: q+W\p(\phi)|^2+\frac{e^2}{2}(|q|^2-|\tq|^2)^2\ . \eeq This

600: potential has two kind of minima: one is the Higgs vacuum \beq

601: \phi=m \ , \qquad |q|=|\tq| \ ,  \qquad \tq q = - W\p(m)\ , \eeq

602: and the others are a set of Coulomb vacua, each for every

603: stationary point of the superpotential \beq \phi=a_i\ ,\qquad

604: q=\tq=0 \ . \eeq

605: 

606: 

607: 

608: 

609: In the Higgs vacuum the gauge group $U(1)$ is completely broken by

610: the quark condensate, so the theory admits  vortex solutions that

611: belongs to the homotopy group $\pi_1(U(1))=\mathbb{Z}$. We don't

612: loose any information choosing \beq \tq = -q\dag

613: \frac{W\p(\phi)}{|W\p(\phi)|} \ . \eeq We are left with the

614: following theory: \beq {\cal

615: L}=-\frac{1}{4e^2}F_{\mu\nu}F^{\mu\nu}-\frac{1}{e^2}

616: \de_{\mu}\phi^{\dagger}\de^{\mu}\phi-(D_{\mu}q)^{\dagger}(D^{\mu}q)-V(\phi,|q|)\

617: , \eeq where the potential is \beq

618: V(\phi,|q|)=2|(\phi-m)q|^2+\frac{e^2}{2} (|q|^2-2|W\p(\phi)|)^2  \

619: .\eeq The vortex solution in terms of the profile functions is:

620: \bea   q &=& e^{i n \theta} q(r) \ , \\

621: \phi &=& \phi(r) \ , \nonumber \\  A_{\theta} &=& \frac{n}{e r}

622: A(r) \ . \nonumber \eea The boundary conditions for $q(r)$ and

623: $A(r)$ are the usual ones. The boundary conditions for $\phi(r)$

624: are $\phi(\infty)=m$ and $\phi(0)=a_j$ where the root $a_j$ will

625: be the one that makes the vortex lighter.  When all the roots

626: $a_i$ are reals, it is clear that $a_j$ is the one that minimize

627: $|m-a_i|$. This vortex can be considered as made off the wall

628: connecting the Higgs vacuum and the $j$-Coulomb vacuum. It is thus

629: straightforward to generalize the physical conjecture made in

630: \ref{wallim}. For $n$ sufficiently large so that $R_V \gg

631: \Delta_W$, the spectrum of these vortices is (\ref{spectrum}).

632: 

633: 

634: 

635: 

636: {\it Almost-BPS solution}

637: 

638: 

639: The theory under consideration has been widely studied since it

640: arise as an effective description of more complicated nonabelian

641: theories \cite{SW,DS,vainyung,Hanany:1997hr}. This SQED is

642: sometimes magnetic dual with respect to the original degrees of

643: freedom and thus our vortices describes confinement of quarks.

644: Usually these vortices have been studied in an almost BPS regime,

645: that is by neglecting the second derivative of the superpotential.

646: In this limit we can make a stronger ansatz for the vortex

647: solution:

648:  \beq \label{BPSansatz}

649: \phi=m \ . \eeq  Thus the Lagrangian becomes the usual one in the

650: BPS limit \cite{Bogomolny:1975de,Edelstein:1993bb}: \beq

651: {\cal

652: L}=-\frac{1}{4e^2}F_{\mu\nu}F^{\mu\nu}-(D_{\mu}q)^{\dagger}(D^{\mu}q)-\frac{e^2}{2}(|q|^2-2|W\p|)^2

653: \ . \eeq The spectrum of vortices is the well known BPS

654: proportionality between tension and charge: \beq

655: \label{BPSsemiclassical} T=4\pi| n W\p(m) |\ . \eeq

656: 

657: 

658: 

659: 

660: Clearly (\ref{BPSansatz}) is only an approximation. It is enough

661: to look at the equation of motion for $\phi$ and see that it is

662: not satisfied inside the vortex: \beq \frac{\Box

663: \phi}{e^2}=2(\phi-m)|q|^2+e^2 {W\p}\p(\phi)\dag

664: (|q|^2-2|W\p(\phi)|)\frac{W\p(\phi)}{|W\p(\phi)|}\ . \eeq An

665: important point is to find the condition under which

666: (\ref{BPSsemiclassical}) is a good approximation to the spectrum.

667: In \cite{Bolognesi:2004yh} we have done this for the winding

668: number $n=1$, and the condition has been found to be: \beq

669: \label{goodone} \frac{e^2{{W\p}\p}^2}{{W\p}} \ll 1\ . \eeq It is

670: clear also from \cite{Bolognesi:2004yh} that as the winding $n$

671: becomes bigger, so do the non-BPS corrections. For consistency

672: with the wall vortex limit, we should have a spectrum like Figure

673: \ref{winding}.

674: \begin{figure}[h!t]

675: \begin{center}

676: \leavevmode \epsfxsize 10 cm \epsffile{winding.eps}

677: \end{center}

678: \caption{\footnotesize Spectrum of vortices in $\N=2$ SQED broken

679: by a superpotential. If $\frac{e^2{{W\p}\p}^2}{W\p} \ll 1$ an

680: almost BPS region is present for small winding numbers. Increasing

681: $n$ non-BPS corrections become larger and not negligible. When the

682: winding $n$ is sufficiently big we are in the wall vortex regime.}

683: \label{winding}

684: \end{figure}

685: For sufficiently small $n$ the tension is given by the BPS formula

686: (\ref{BPSsemiclassical}) and is proportional to $n$. For

687: sufficiently big $n$ the tension is given by (\ref{spectrum}) and

688: is proportional to $n^{2/3}$. There must not be superposition

689: between the two regions.

690: 

691: 

692: 

693: {\it $\N=2$ SQED broken by adjoint mass}

694: 

695: 

696:  Now we make the consistency check the the two regions do not

697: overlapped. For simplicity we limit ourselves to the quadratic

698: superpotential $ W(\Phi)=\frac{1}{4} \mu \Phi^2 $. The energy

699: density for the wall is

700:  \beq

701: T=\int \d x  ( \frac{1}{e^2}\de_x \phi\dag \de_x \phi+ \de_x q\dag

702: \de_x q+ 2|(\phi-m)q|^2+ \frac{e^2}{2}(|q|^2 -\mu \phi)^2 ) \ .

703: \eeq We rewrite it in terms of the dimensionless variables $

704: \phi=m\varphi$ and $q=\sqrt{\mu m}\chi$:

705:  \beq \label{tt}

706: T=\int \d x  ( \frac{m^2}{e^2}\de_x \varphi\dag \de_x \varphi+

707: \de_x \chi\dag \de_x \chi+ 2m^3\mu|(\varphi-1)\chi|^2+ \frac{e^2

708: m^2 \mu^2}{2}(|\chi|^2 -\varphi)^2 ) \ . \eeq The condition for

709: the existence of an almost BPS region is

710:  \beq \label{alm} \frac{e^2 \mu}{m} \ll 1 .\eeq

711:  The $n=1$ vortex would also belong to the wall vortex region is the condition $R_V \gg \Delta_W$ would be satisfied.

712:  In terms of our variables this condition would be

713:  \beq  \label{wal} \frac{1}{e^{2/3}{T_W}^{1/3}} \gg \Delta_W .\eeq

714:  We are going to prove that, if (\ref{alm}) is satisfied, then (\ref{wal}) cannot be

715:  satisfied.

716: 

717: 

718: 

719: Under the assumption that (\ref{alm}) is satisfied, the tension

720: (\ref{tt}) can be approximated by \beq \label{var} T \sim

721: \frac{m^2}{e^2\Delta}  +m^3\mu \Delta \ , \eeq where $\Delta$ is

722: the length scale of variation of the fields. Minimizing

723: (\ref{var}) with respect to $\Delta$  we obtain: \beq \Delta_W

724: \sim \frac{1}{e\sqrt{\mu m}}\ ,\qquad T_W \sim

725: \frac{m^{5/2}\mu^{1/2}}{e} \ .\eeq Thus the condition for the wall

726: vortex is \beq \frac{1}{e^{2/3} {T_W}^{1/3} \Delta_W} \sim

727: \left(\frac{e^2 \mu}{m}\right)^{1/3} \gg 1 \ ,\eeq and it is

728: clearly not satisfied if (\ref{alm}) is true.

729: 

730: 

731: 

732: 

733: 

734: \subsubsection{Quantum corrections}

735: 

736: The non-supersymmetric theory (\ref{BasicVortex}) with the

737: potential of Figure \ref{potential} has a problem: it is not

738: stable under quantum corrections since there is no spontaneously

739: broken discrete symmetry that can relate an Higgs vacuum with a

740: Coulomb vacuum. Only with a fine tuning of the parameters we can

741: have this degeneracy.

742: 

743: Degeneracy of vacua, not related by any spontaneously broken

744: discrete symmetry, is a common feature of supersymmetric gauge

745: theories. In fact in the supersymmetric theory described in

746: \ref{susy} the degeneracy is protected by non renormalization

747: theorems. We can thus ask what happens to the relations

748: (\ref{eqscal}) when we consider quantum corrections. Note that to

749: obtain (\ref{eqscal}) we have used only general principles and so

750: they should be valid also in a full quantum theory, even if the

751: various parameters that enter in the game $T_V, R_V, T_W,

752: \Delta_W$ are subject to quantum corrections. Special attention

753: should be given to the coupling constant $e(\mu)$ since in the

754: full quantum theory depends on the energy scale $\mu$. But since

755: we want to study vortices with large $n$ and so large radius

756: $R_V$, we should keep the low-energy coupling constant

757: $e(m_{lightest})$ where $m_{lightest}$ is the mass of the lightest

758: charged particle.

759: %Having understood in which sense (\ref{eqscal})

760: %is true in the quantum theory, it is straightforward to repeat the

761: %physical conjecture given in \ref{wallim}.

762: 

763: 

764: 

765: 

766: \subsubsection{Zero modes}

767: 

768: The low-energy dynamics of a $p$-soliton is described by an

769: effective $p+1$ dimensional field theory on the world volume of

770: the soliton. By low-energy we mean length greater that the

771: thickness of the soliton. The target space of the effective action

772: is the space of zero modes. this space can be divided in

773: displacement zero modes, the ones that describe fluctuation of the

774: brane in the space, and internal zero modes. Up to now it is the

775: standard technique for describe soliton dynamics. In the case of

776: the wall vortex it is not hard to imagine that there should be a

777: relation between the effective $2+1$ theory of the domain wall and

778: the effective $1+1$ theory of the vortex. In fact the vortex

779: dynamics is described by the $2+1$ theory of the wall with a

780: spatial direction compactified on a circle of radius $R_V$. For

781: wave lengths $\lambda$ so that $\Delta_W \ll \lambda \ll R_V$,  is

782: a vibrating membrane and for wave lengths $\lambda \gg R_V$ it is

783: a vibrating string. In Figure \ref{zeromodes} we have the

784: displacement zero modes in the two different regimes. It is also

785: clear that the wall vortex share the same internal zero modes

786: target space with the wall that makes it.

787: \begin{figure}[h!t]

788: \begin{center}

789: \leavevmode \epsfxsize 10 cm \epsffile{zeromodes.eps}

790: \end{center}

791: \caption{\footnotesize Fluctuation of the displacement zero modes.

792: For wave length $\lambda \gg R_V$ we have a vibrating string. For

793: wave length $\Delta_W \ll \lambda \ll R_V$ we have a vibrating

794: wall compactified on a circle.} \label{zeromodes}

795: \end{figure}

796: 

797: 

798: 

799: \subsection{Nonabelian theories and the existence at strong

800: coupling \label{naexistence}}

801: 

802: It is a natural question if these wall vortices can arise in

803: nonabelian theory and at strong coupling. The answer is yes and,

804: as we will see in a moment, it is enough to consider confining

805: vacua with different confinement index.

806: 

807: Take a generic $SU(N)$ gauge theory.  The confinement index $t$ is

808: defined as the minimal integer so that $Q^t$ in unconfined. By

809: $Q^t$ we mean the representation of $t$ antisymmetrized quarks $Q$

810: in the fundamental of the gauge group. When  $t=N$ the theory is

811: completely confined, like ordinary QCD. When $t=1$ the theory is

812: completely unconfined. $t$ must be a divisor of $N$. The

813: confinement is caused by strings with topological number

814: $\mathbb{Z}_t$. A $k$-string, with $k \in \mathbb{Z}_t$, confines

815: the $Q^k$ representation. The $t$-string is trivial and in fact

816: $Q^t$ is unconfined.

817: 

818:  Now consider a case in which this theory has

819: one vacuum $A$ with confinement index $t_A$ and another vacuum $B$

820: with confinement index $t_B > t_A$. Suppose that exists a domain

821: wall interpolating between these two vacua. There will be at least

822: one representation $Q^k$ that is confined in one vacuum and not

823: confined in the other. For example $Q^{t_A}$ is not confined in

824: $A$ but confined in $B$.   We can make a continuous

825: transformation, such as the one previously described in Figure

826: \ref{push}. We take a $Q^{t_A}$ particle in vacuum $A$ and we push

827: it against the wall. We will end with a $Q^{t_A}$ in vacuum $B$

828: connected to a $t_A$-string that ends on the wall. This continuous

829: interpolation between the wall and the string shows that the

830: $t_A$-string is indeed a wall vortex.

831: 

832: 

833: 

834: In general every $k$-string in vacuum $B$ is a wall vortex with

835: respect to the $A$/$B$ wall if $k$ is a multiple of $t_A$ but not

836: of $t_B$. The same can be said on the other side of the wall.

837: Every $k$-string in vacuum $A$ is a wall vortex with respect to

838: the $B$/$A$ wall if $k$ is a multiple of $t_B$ but not of $t_A$.

839: To be more precise we should think of the strings as living in an

840: extended group $\mathbb{Z}_{\lcm{(t_A,\,t_B)}}$ that contains both

841: $\mathbb{Z}_{t_A}$ and $\mathbb{Z}_{t_B}$. Strings in vacuum $A$

842: that belong to the subgroup $\mathbb{Z}_{\lcm{(t_A,\,t_B)}/t_B}$

843: are wall vortices with respect to the $A$/$B$ wall. Strings in

844: vacuum $B$ that belong to the subgroup

845: $\mathbb{Z}_{\lcm{(t_A,\,t_B)}/t_A}$ are wall vortices with

846: respect to the $B$/$A$ wall. We will came back to this point in

847:  \ref{super}.

848: 

849: 

850: 

851: In general we can prove that a domain wall interpolating between

852: vacua with different confinement index exists. Consider the set of

853: vacua of the theory and group them in sets with the same

854: confinement index like in Figure \ref{wall}.

855: \begin{figure}[h!t]

856: \begin{center}

857: \leavevmode \epsfxsize 9 cm \epsffile{wall.eps}

858: \end{center}

859: \caption{\footnotesize In general it not true that given any

860: couple of vacua $A$ and $B$ a domain wall that interpolates

861: between them exist. But if it doesn't exist then it decays in a

862: path of stable domain walls that connect $A$ and $B$ passing

863: through other vacua. If the $t_A \neq t_B$, surely there will be a

864: stable domain wall at some point of the path that connects vacua

865: with different confinement index. } \label{wall}

866: \end{figure}

867: It is not true that for any couple of vacua $A$ and $B$ the domain

868: wall interpolating between them exist. But, for sure, it will

869: exist at least a path connecting the two vacua so that in every

870: segment of the path the domain wall exist and it's stable. In this

871: path there will be some domain wall interpolating between vacua

872: with different confinement index.

873: 

874: 

875: The study of the wall limit for nonabelian strings, such as the

876: one done in \ref{wallim}, will appear in \cite{miononabeliani}.

877: 

878: 

879: 

880: 

881: \section{Classification of the Domain Wall/Flux Tube Junctions \label{classification}}

882: 

883: The wall vortex studied in  Section \ref{made} is one particular

884: case of wall/tube junction. The purpose of this Section is to

885: classify the various kind of junctions between walls and tubes.

886: The junctions can be distinguished by simple properties:

887: \begin{center}

888: \begin{itemize}

889: \item If the tube can end or not on the wall. \item Where the flux

890: goes into.

891: \end{itemize}

892: \end{center}

893: In this spirit, the Coulomb junction of Section \ref{made}  can be

894: schematically represented as in Figure \ref{Coulomb}. The tube

895: ends on the wall and the flux goes in the opposite half space.

896: \begin{figure}[h!t]

897: \begin{center}

898: \leavevmode \epsfxsize 10 cm \epsffile{Coulomb.eps}

899: \end{center}

900: \caption{\footnotesize The Coulomb junction. A vortex ends on the

901: wall and the flux is spread into the other half space. The point

902: where it ends is seen in the Coulomb vacuum as a monopole of

903: double charge.} \label{Coulomb}

904: \end{figure}

905:  To analyze

906: other examples we make the first distinction. In \ref{local} we

907: consider vortices that can end on the wall. In \ref{crossing} we

908: consider vortices that cross the wall and continue on the other

909: side.

910: 

911: 

912: 

913: 

914: 

915: \subsection{The Flux Tube Ends on the Wall \label{local}}

916: 

917: 

918: In string theory, a string can have Dirichlet boundary conditions

919: if it ends on a dynamical object called D-brane. By analogy, in

920: some QFT, similar phenomenon can happen when a flux tube ends on

921: the wall.  We will briefly describe two examples in which a flux

922: tube can terminate on a wall. In the first case there is a

923: localization of the flux on the wall. Thus the tube spread out its

924: flux into the wall.  In the second example there is a bound state

925: confined on the wall that captures the flux. The bound state is an

926: object composed by two pieces that belong to the two different

927: vacua.

928: 

929: 

930: 

931: \subsubsection{\label{nahiggs} Localization of the flux}

932: 

933: 

934: This case is probably the more studied, particularly in the last

935: few years. We have a domain wall interpolating between two vacua

936: both in the Higgs phase. The magnetic strings on both side of the

937: wall can end on it and spread the flux inside it (see Figure

938: \ref{Local}).

939: \begin{figure}[h!t]

940: \begin{center}

941: \leavevmode \epsfxsize 10 cm \epsffile{Local.eps}

942: \end{center}

943: \caption{\footnotesize The D-brane junction. A vortex ends an a

944: wall and spread its flux into it. The effective theory on the wall

945: must contain massless gauge field. The point where the tube ends

946: is an electric charge with respect to the $2+1$ effective gauge

947: theory.} \label{Local}

948: \end{figure}

949: Inside the wall there must be a localized gauge theory. Note that

950: if we are able to prove that the $2+1$ dimensional theory contains

951: the massless gauge field, the confining string must end on the

952: wall (the other possible configurations would be not favorite

953: energetically).

954: 

955: 

956: 

957: 

958: Recently there has been research in the the so called non-abelian

959: Higgs model \cite{Shifman:2002jm,nahiggs}. It is a $U(N_c)$ $\N=2$

960: gauge theory with $N_f$ hypermultliplets of mass $m_i$. The theory

961: is then broken to $\N=1$ adding a Fayet-Iliopoulos term for the

962: $U(1)$ subgroup. When $N_f \geq N_c$ the theory has

963: $\binom{N_f}{N_c}$ vacua since every color must be locked to some

964: flavor. This theory is a nice lab because it possesses all the

965: three kinds of solitons: monopoles, vortices and domain walls,

966: that can appear in various junctions that preserves supersymmetry.

967: Consider in particular a fundamental wall that separates two vacua

968: $\dots,i,\widehat{i+1},\dots$ and $\dots,\widehat{i},i+1,\dots$

969: (with the hat we indicate the flavors that are not locked). In the

970: first vacua a $U(1)$ is locked to the flavor $i$ while the flavor

971: $i+1$ is unlocked, the opposite in the other vacua. In both vacua

972: the $U(1)$ is Higgsed, in the first by the condensation of the

973: flavor $Q_i$ and in the second by the flavor $Q_{i+1}$. They admit

974: a magnetic vortex that breaks $1/2$ of supersymmetries and

975: saturates the BPS bound. The central charge for the vortex is

976: given essentially by the Fayet-Iliopoulos term. It has been shown

977: that the walls are also $1/2$-BPS and in particular the one under

978: consideration (from $i, \widehat{i+1}$ to $\widehat{i}, i+1$), has

979: a Coulomb phase inside with respect to this $U(1)$. The junction

980: vortex-wall has been shown to be $1/4$-BPS. For generalizations to

981: nonabelian strings and nonabelian walls see \cite{nonabelian}.

982: 

983: 

984: 

985: 

986: 

987: \subsubsection{\label{sy} Bound state localized on the wall}

988: 

989: 

990: Another example of strings anding on walls is provided by $\N=1$

991: $SU(N)$ super Yang-Mills. This theory has $N$ vacua obtained by

992: the spontaneous breaking of the anomaly free residual

993: $R$-symmetry. We label these vacua with the index $0 \leq r < N$

994: (see Figure \ref{SYM}).

995: \begin{figure}[h!t]

996: \begin{center}

997: \leavevmode \epsfxsize 13 cm \epsffile{SYM.eps}

998: \end{center}

999: \caption{\footnotesize The $\N=1$ pure Super Yang Mills has $N$

1000: different vacua labelled with $r$. The confinement in each vacuum

1001: is explained by the condensation of a particle that carries

1002: magnetic charge $1$ and electric charge $Q^r$. In every vacuum

1003: strings belong to the topological group $\mathbb{Z}_N$. A

1004: $k$-string connect a $Q^k$ with a $\overline{Q}^k$ and creates a

1005: meson.} \label{SYM}

1006: \end{figure}

1007: Every vacuum has confinement index $t=N$ and so there are

1008: $\mathbb{Z}_{N}$ strings responsible for the confinement of

1009: sources in any non trivial representation of the center of the

1010: gauge group. The string $k \in \mathbb{Z}_{N}$ confines the

1011: representation $Q^k$.   The confinement in the $r$ vacua can be

1012: understood as caused by the condensation a monopole bounded

1013: together with a $Q^r$. We indicate this object with $(Q^r,1)$

1014: where the $1$ refers to the magnetic charge.

1015: 

1016: 

1017: It has been shown in \cite{Witten:1997ep} that the domain

1018: wall\footnote{ Domain walls in $\N=1$ super Yang-Mills have been

1019: studied in \cite{Dvali:1996xe,Shifman,Dvali}. The effective action

1020: on these walls has been found in \cite{Acharya:2001dz}.}

1021: interpolating between two adjacent vacua $r$ and $r+1$ is a

1022: D-brane for the $k$-strings. This phenomenon happens also for

1023: every wall interpolating between a vacuum $r_1$ and a vacuum $r_2$

1024: so that $r_2-r_1$ is prime with $N$. The easiest way to understand

1025: it is the Rey's argument reported in \cite{Witten:1997ep}. The

1026: confinement in the $r$-vacuum is due to the condensation of a

1027: $(Q^r,1)$ particle. To prove that the $k$-string can end on the

1028: wall we need a bound state confined on the wall that carries the

1029: same charge of $Q^{-k}$. As showed in Figure \ref{boundSYM} this

1030: object can be obtained putting together $k$ $(Q^r,1)$ and $-k$

1031: $(Q^{r+1},1)$ on the opposite sides of the

1032: wall.\footnote{Actually, to have a rigorous prove, we should show

1033: the composite object has a binding energy.}

1034: \begin{figure}[h!t]

1035: \begin{center}

1036: \leavevmode \epsfxsize 8 cm \epsffile{boundSYM.eps}

1037: \end{center}

1038: \caption{\footnotesize The End junction. The vortex ends on a

1039: boundstate confined on the wall. } \label{boundSYM}

1040: \end{figure}

1041: 

1042: 

1043: 

1044: 

1045: \subsection{The Flux Tube Crosses the Wall \label{crossubsection}}

1046: 

1047: 

1048: When a  flux tube ends on a wall, the flux must go somewhere.

1049: There are only two possibility: it goes on the other half space

1050: (see Section \ref{made}), or it is confined inside the wall (see

1051:  \ref{local}). Now we consider the case in which the

1052: vortex cannot end on the wall and so  must  continue into another

1053: vortex in the other vacua (see Figure \ref{crossing}).

1054: \begin{figure}[h!t]

1055: \begin{center}

1056: \leavevmode \epsfxsize 9 cm \epsffile{crossing.eps}

1057: \end{center}

1058: \caption{\footnotesize The Cross junction. The flux tube crosses

1059: the wall and becomes a vortex in the other vacuum.}

1060: \label{crossing}

1061: \end{figure}

1062: They could also bring different fluxes since there could be a

1063: particle localized on the wall, in the junction between the two

1064: vortices, that carries the difference of the fluxes.

1065: 

1066: 

1067: 

1068: The first example is the trivial one. Take the usual $U(1)$ gauge

1069: theory coupled to a charged scalar in the Higgs phase. Than add a

1070: real scalar field responsible for the creation of a domain wall.

1071: We don't put interactions between the sector of the theory

1072: responsible for the vortex and the sector responsible for the

1073: wall. The domain wall is transparent to the vortex that can cross

1074: it without any modification. This trivial junction also arise in

1075: the non-abelian Higgs model considered in \ref{nahiggs}.  For

1076: example, we saw that the wall interpolating between the  vacua

1077: $\dots,i,\widehat{i+1},\dots$ and $\dots,\widehat{i},i+1,\dots$,

1078: forms a D.brane junction with respect to the vortex that carry the

1079: flux of the $U(1)$ that changes locking. The same wall makes a

1080: trivial junction with respect to the other $U(1)$'s that are left

1081: unchanged.

1082: 

1083: 

1084: 

1085: \subsubsection{A vortex with a wall around it}

1086: 

1087: 

1088: Now we move to a non trivial case. Consider the $U(1)$ gauge

1089: theory (\ref{BasicVortex}), where the potential has two different

1090:  minima at $q_A < q_B$ (see Figure \ref{wallaroundpot}).

1091: \begin{figure}[h!t]

1092: \begin{center}

1093: \leavevmode \epsfxsize 9 cm \epsffile{wallaroundpot.eps}

1094: \end{center}

1095: \caption{\footnotesize A potential with two degenerate Higgs

1096: vacua.} \label{wallaroundpot}

1097: \end{figure}

1098: There is a wall that interpolates between the two vacua, that has

1099: a profile $q=q(z)$ with boundary conditions $q(-\infty)=q_A$ and

1100: $q(+\infty)=B$. The wall has not have a Coulomb phase inside it,

1101: since $q(z)$ is always different from zero.  Both the vacua $A$

1102: and $B$, admit magnetic flux tubes and the tension $T_A$ will be

1103: smaller than the tension $T_B$.  The Coulomb junction is excluded

1104: since $A$ and $B$ are both in the Higgs phase. The D-brane

1105: junction is also excluded since there is not a Coulomb phase

1106: inside the wall. The only possibility is that the vortex in $A$

1107: crosses the wall and becomes the vortex in $B$ (see Figure

1108: \ref{wallaroundfig}).

1109: \begin{figure}[h!t]

1110: \begin{center}

1111: \leavevmode \epsfxsize 9 cm \epsffile{wallaroundfig.eps}

1112: \end{center}

1113: \caption{\footnotesize A vortex crosses the wall and becomes a

1114: heavier vortex that can be though as the previous one with a wall

1115: wrapped around.} \label{wallaroundfig}

1116: \end{figure}

1117: Since they carry the same flux there non need of any particle

1118: localized on the wall. There is a nice interpretation of the fact

1119: that $T_A$ is smaller than $T_B$. We can imagine that the vortex

1120: in $B$ is composed by a vortex in $A$ and the wall rolled around

1121: it. Thus the vortex in $B$ is more heavy because it has a wall

1122: around.

1123: 

1124: 

1125: 

1126: 

1127: 

1128: \subsubsection{Crossing in nonabelian theories \label{crossna}}

1129: 

1130: The last example comes from $\N=1$ $SU(N)$ super Yang-Mills.  We

1131: choose $r_A$, for the vacuum $A$, and $r_B$, for the vacuum $B$,

1132: so that $r_B-r_A$ is a divisor of $N$. We use again the Rey's

1133: argument to determine which string ends on the wall and which one

1134: crosses the wall.  Is possible to build a bound state on which a

1135: $k$-string could end, only if $k$ is a multiple of

1136: $\gcd{(r_B-r_A,N)}$. Thus we have a subgroup

1137: $\mathbb{Z}_{\gcd{(r_B-r_A,N)}} \subset \mathbb{Z}_N$ of strings

1138: that can end on the wall with an End junction. Strings in the

1139: quotient group $\mathbb{Z}_{N/\gcd{(r_B-r_A,N)}}$ instead, cross

1140: the wall and goes into another string (Cross junction). Like in

1141: Figure \ref{crossSYM} there can be a bound state multiple of

1142: $\gcd{(r_B-r_A,N)}$ that change the flux of the string but cannot

1143: absorb completely its flux.

1144: \begin{figure}[h!t]

1145: \begin{center}

1146: \leavevmode \epsfxsize 8 cm \epsffile{crossSYM.eps}

1147: \end{center}

1148: \caption{\footnotesize The vortex crosses the wall. A bound state

1149: confined on the wall can change the flux of the vortex. This

1150: happens whenever $k_A-k_B$ is not a multiple of

1151: $\gcd{(r_A-r_B,N)}$.} \label{crossSYM}

1152: \end{figure}

1153: 

1154: We analyze the simplest example. Take $N$ even and a wall between

1155: $r_A$ and $r_B=r_A+2$. Even strings can end on the wall. Odd

1156: strings must cross the wall and they becomes odd strings on the

1157: other side. For example a $1$-string can end on a bound state that

1158: carries charge $Q^{-2}$ and then emerges on the other side like a

1159: $-1$-string.

1160: 

1161: 

1162: \subsection{A complete example \label{super}}

1163: 

1164: 

1165: An interesting theory is $U(N)$ $\N=2$ super Yang-Mills broken to

1166: $\N=1$ by a generic superpotential. This theory is enough rich to

1167: present all the junctions just described.

1168: 

1169: 

1170: 

1171: The simplest example is a cubic superpotential so that

1172: $W\p=(\Phi-a_1)(\Phi-a_2)$ has two roots. For simplification we

1173: will consider the regime $\Lambda \gg a_1,a_2,a_1-a_2$. In this

1174: case  the theory must distribute the eigenvalues of $\Phi$ at a

1175: scale where the gauge group is still weakly coupled, and at this

1176: scale, $U(N)$ is broken down to $U(N_1) \times U(N_2)$. At the

1177: scale $\Lambda$ the gauge group confines and we have in global

1178: $N_1 \cdot N_2$ vacua labelled by two integers $r_{N_1}$ and

1179: $r_{N_2}$. A complete characterization of a vacua is thus given

1180: specifying three integers $N_1$, $r_1$  and $r_2$. The confinement

1181: index $t$ is the greatest common divisor between $N_1$, $N$, and

1182: $r_1-r_2$ \cite{Cachazo:2002zk, Homotopy}. A domain wall is given

1183: choosing a couple of vacua, call the first vacuum $A$ and the

1184: second vacuum $B$.

1185: 

1186: 

1187: \subsubsection{All the junctions simultaneously}

1188: 

1189: The simplest case in which all the three kind of junctions appear

1190: simultaneously is the one schematically represented if Figure

1191: \ref{tutti}.

1192: \begin{figure}[h!t]

1193: \begin{center}

1194: \leavevmode \epsfxsize 9 cm \epsffile{tutti.eps}

1195: \end{center}

1196: \caption{\footnotesize $U(16) \to U(8) \times U(8)$.}

1197: \label{tutti}

1198: \end{figure}

1199: It is a $U(16)$ theory that breaks down to $U(8) \times U(8)$. For

1200: vacuum $A$ we chose $r_{A 1}=r_{A 2}=0$ and the confinement index

1201: is $t_A=8$. For vacuum $A$ we chose $r_{B 1}=2$, $r_{B 2}=6$ and

1202: the confinement index is $t_B=4$. Vacuum $A$ has $\Z_8$ strings

1203: while vacuum $B$ has $\Z_4$ strings. The state $Q^4$ is confined

1204: in $A$ and non confined in $B$, thus we have a $\Z_2$ subgroup of

1205: strings in $A$ that make a Coulomb junction with the wall. The

1206: quotient $\Z_8 / \Z_2 =\Z_4$ is now equal to the strings in vacuum

1207: $B$. The subgroup $\Z_2$ that confines even powers of $Q$ can end

1208: on the wall and thus they form an End junction. The quotient $\Z_4

1209: / \Z_2 =\Z_2$ are the strings that confine odd powers of $Q$ and

1210: they form a Cross junction.

1211: 

1212: 

1213: In general all the three kind of junctions appear when we have

1214: vacua with different confinement indices. What we are going to say

1215: is presented schematically in Figure \ref{groups}.

1216: \begin{figure}[h!t]

1217: \begin{center}

1218: \leavevmode \epsfxsize 6 cm \epsffile{groups.eps}

1219: \end{center}

1220: \caption{\footnotesize  The web of groups that describe wall

1221: vortex junctions between vacua of a $SU(N)$ gauge theory with

1222: confinement index $t_A$ and $t_B$. The smallest group that

1223: contains both the $A$ and the $B$ strings is

1224: $\mathbb{Z}_{\lcm(t_A,t_B)}$. The subgroup

1225: $\mathbb{Z}_{\lcm(t_A,t_B)/t_B}$ identifies strings in vacuum $A$

1226: that end on the wall and forms a Coulomb junction. The opposite is

1227: true for strings in vacuum $B$. $\mathbb{Z}_{cross}$ is the group

1228: of stable strings in the domain wall background, that is the group

1229: of strings that cross the wall. Strings in vacuum $A$ that belongs

1230: to the subgroup $\mathbb{Z}_{t_A/ cross} \in \mathbb{Z}_{t_A}$ end

1231: on the wall and form an end junction. The broup of crossing

1232: strings $\mathbb{Z}_{cross}$ is the quotient

1233: $\mathbb{Z}_{t_A}/\mathbb{Z}_{t_A/ cross}$.  The opposite is true

1234: for strings in vacuum $B$. } \label{groups}

1235: \end{figure}

1236: The confining strings $\mathbb{Z}_{t_A}$ of vacuum $A$ and

1237: $\mathbb{Z}_{t_B}$ of vacuum $B$, must be though embedded in the

1238: smallest group that contains both of them

1239: $\mathbb{Z}_{\lcm{(t_A,\,t_B)}}$. Strings that are multiple of

1240: $t_B$ but not of $t_A$ are wall vortices in vacuum $B$ and they

1241: belongs to the group $\mathbb{Z}_{\lcm{(t_A,\,t_B)}/t_B}$. In the

1242: same way  strings that are multiple of $t_A$ but not of $t_B$ are

1243: wall vortices in vacuum $A$ and they belongs to the group

1244: $\mathbb{Z}_{\lcm{(t_A,\,t_B)}/t_A}$. For the remaining strings we

1245: must decide if they form an End junction or a Cross junction. The

1246: Cross-junctions has a nice interpretation. They can be thought as

1247: the stable strings in the domain wall background. So in the same

1248: way $\mathbb{Z}_{t_A}$ is the group of string in vacuum $A$ and

1249: $\mathbb{Z}_{t_B}$ is the group of strings in vacuum $B$, so

1250: $\mathbb{Z}_{cross}$ is the group of stable strings in the domain

1251: wall background where $cross$ must be a divisor of both $t_A$ and

1252: $t_B$. The remaining strings $\mathbb{Z}_{t_A/cross}$ are End

1253: junction in vacuum $A$ and $\mathbb{Z}_{t_B/cross}$ are End

1254: junctions in vacuum $B$. In Figure \ref{esempi} we give two

1255: examples of the groups relation of the scheme in Figure

1256: \ref{groups}.

1257: \begin{figure}[h!t]

1258: \begin{center}

1259: \leavevmode \epsfxsize 14 cm \epsffile{esempi.eps}

1260: \end{center}

1261: \caption{\footnotesize  Two examples of web of groups in Figure

1262: \ref{groups}. The first refers to $U(N)$ super Yang-Mills where

1263: vacuum $A$ has index $r_A$ and vacuum $B$ has index $r_B$. The

1264: second example is $U(N) \to U(N_1) \times U(N_2)$ with indices

1265: $r_{A 1}$, $r_{A 2}$ for vacuum $A$ and $r_{B 1}$, $r_{B 2}$ for

1266: vacuum $B$. } \label{esempi}

1267: \end{figure}

1268: 

1269: 

1270: 

1271: 

1272: 

1273: \subsubsection{Decay of junctions}

1274: 

1275: 

1276: In the theory at hand, strings can appear in different regimes

1277: distinguished by the magnitude of the superpotential. $\sqrt{W\p}$

1278: is the energy scale where $\N=2$ is broken and $W\p$ is the

1279: tension scale of the strings. Thus we have three cases.

1280: $\sqrt{W\p} \ll \Lambda \ll a$ is the strong coupling regime,

1281: where the superpotential is a small perturbation that can be added

1282: to the low energy dynamics of $\N=2$. In this case flux tubes are

1283: nothing but ANO vortices of the low-energy $U(1)$'s. In the

1284: intermediate regime $\Lambda \ll \sqrt{W\p} \ll a$, the dynamics

1285: is approximatively that of various independent $U(N_i)$, note that

1286: the number of strings is the same as before but now they are

1287: governed by the group $\oplus_i \mathbb{Z}_{N_i}$. The transition

1288: between the strong coupling regime and the pure $\N=1$ regime has

1289: been studied in \cite{Hanany:1997hr} and there is nothing new to

1290: say about it. The last regime is $\Lambda \ll a \ll \sqrt{W\p}$,

1291: where some strings that where metastable before now can decay.

1292: Their dynamics is governed by the group

1293: $\mathbb{Z}_{\gcd{(N_i,b_i)}}$.

1294: 

1295: 

1296: 

1297: We consider in detail the case $N=4$. We choose both the vacua $A$

1298: and $B$ so that classically $U(4) \to U(2) \times U(2)$. In vacuum

1299: $A$ both the $U(2)$ factors are in the monopole vacuum $r_{A

1300: 1}=r_{A 2}=0$. In vacuum $B$ the first $U(2)$ is in the monopole

1301: vacuum $r_{B 1}=0$, and the second in the dyon vacuum $r_{B 2}=1$

1302: (see Figure \ref{decayth}).

1303: \begin{figure}[h!t]

1304: \begin{center}

1305: \leavevmode \epsfxsize 9 cm \epsffile{decayth.eps}

1306: \end{center}

1307: \caption{\footnotesize $U(4) \to U(2) \times U(2)$.}

1308: \label{decayth}

1309: \end{figure}

1310: \begin{description}

1311:     \item[$\mathbf{\Lambda \ll \sqrt{W\p} \ll a}$.]    This is

1312: the most easiest region to analyze. Both vacua have two confining

1313: strings that are the non trivial elements of the group

1314: $\mathbb{Z}_2 \oplus \mathbb{Z}_2$. The first group refers to the

1315: first $U(2)$ and, since $A$ and $B$ are both in the monopole

1316: vacua, these strings cannot end on the wall. So we are forced to

1317: conclude that they form a Cross junction. The second

1318: $\mathbb{Z}_2$ refers to the second $U(2)$. Since vacuum $A$ is in

1319: the monopole vacuum and $B$ in the dyon vacuum, the strings can

1320: end on the wall and so they forms an End junction.

1321: \item[$\mathbf{\Lambda \ll a \ll \sqrt{W\p}}$.]   In this region

1322: of parameters the two theories are governed by the confinement

1323: index. Vacuum $A$ has $\mathbb{Z}_2$ strings and so only one non

1324: trivial element. The interpretation is that the two elements

1325: founded in the previous regime belongs to the same topological

1326: element and can decay one into the other. Vacuum $B$ has $t_B=1$

1327: so is completely unconfined. The two strings founded previously

1328: are only metastable. The conclusion is obvious, we are in the

1329: presence of a Coulomb junction. Thus we have seen an example of

1330: the decay schematically presented in Figure \ref{decay}.

1331: \end{description}

1332: \begin{figure}[h!t]

1333: \begin{center}

1334: \leavevmode \epsfxsize 10 cm \epsffile{decay.eps}

1335: \end{center}

1336: \caption{\footnotesize Decay of junctions.} \label{decay}

1337: \end{figure}

1338: 

1339: 

1340: \appendix

1341: 

1342: \section{The Flux Tube/Flux Tube Junction} \label{yappendix}

1343: 

1344: 

1345: 

1346: In this Appendix we describe, using the things studied in this

1347: paper, a way to obtain a  junction between a flux tube and another

1348: flux tube. This junction can be used to build a baryon vertex in

1349: some particular confining gauge theory. In the string model of

1350: hadrons, mesons are identified with a string with a quark and an

1351: antiquark attached at the endpoints. For the baryons things are

1352: not so natural and various configurations are possible. One of the

1353: hypothesis is the so called   ``Y'' junction: the baryon is

1354: composed by three strings with one end in common and three quarks

1355: at the other ends \cite{Artru:1974zn}. In general the construction

1356: of the baryon vertex is not at all trivial (see for example

1357: \cite{Hanany:1997hr} for  MQCD and \cite{Witten:1998xy} for the

1358: AdS/CFT correspondence).      In what follows we describe another

1359: possible mechanism for the formation of the Y junction using the

1360: flux tube/flux tube junction.

1361: 

1362: 

1363: 

1364:  Suppose to have a theory with two gauge

1365: groups $U(1)_1$ and $U(1)_2$, and two discrete vacua $A$ and $B$.

1366: For the first $U(1)_1$ both vacua are in the Higgs phase and, on

1367: the domain wall interpolating between the two, there is a

1368: localization of the gauge field. Thus we have a wall/vortex

1369: junction like a D-brane. For the second group $U(1)_2$ the vacuum

1370: $A$ is in the Higgs phase, while vacuum $B$ is in the Coulomb

1371: phase and we can have a vortex made of wall like the one studied

1372: in Section \ref{made}. The vortex in vacuum $A$ that carries the

1373: magnetic flux of $U(1)_1$, can end on the domain wall and, as a

1374: consequence, can also end on the magnetic vortex that carries flux

1375: under $U(1)_2$. In fact the last one is nothing but the wall

1376: rolled in a cylinder.  In this way we we have obtained a ``Y''

1377: junction.

1378: 

1379: 

1380: 

1381: 

1382: Now we are going to show take a particular confining $SU(3)$ gauge

1383: theory that admits the  Y junction previously constructed as its

1384: baryon vertex.

1385: 

1386: 

1387: The theory that we consider is $\N=2$ $SU(3)$ SQCD with two

1388: hypermultiplets in the foundamental representation. the hypermultiplets have mass $m_A$ and $m_B$. In a

1389: generic point of the moduli space ($2$ complex dimensions) the

1390: theory is in a $U(1)_1 \times U(1)_2$ Coulomb phase. At some

1391: critical lines ($1$ complex dimension), there are massles

1392: hypermultiplets for one of the $U(1)$'s. At some crtical points ($0$

1393: complex dimensions), both the U(1)'s have charged massless

1394: hypermultiplets. We take vacuum $A$ to be the critical point where

1395: $U(1)_1$ is locked to the flavor $m_A$, and some massless

1396: monopole is charged under $U(1)_2$. We choose vacuum $B$

1397: to be on the critical line where $U(1)_1$ is locked to $m_2$, while

1398: $U(1)_2$ has no massless charged particles.  We know that is

1399: possible to choose some superpotential $W(\Phi)$ so that vacua $A$

1400: and $B$ are among the ones that survive the perturbation. After

1401: the perturbation, charged  massless particles condense and creates flux

1402: tubes. If these particles have magnetic charge,  the flux tubes carry electric flux and are responsible for confinement.

1403: 

1404: 

1405: It is possible to send the mass $m_1$  into the strong coupling region ($\sim \Lambda$),

1406:  so that  the charge of the locked massless flavor becomes magnetic, and vacuum A is in the confining phase.

1407: 

1408: 

1409: With respect to $U(1)_1$ both vacua are both in the Higgs phase

1410: but the breaking of the gauge group is due to two different

1411: charged particles. Thanks to the same mechanism of

1412: \cite{Shifman:2002jm}, the $U(1)_1$ gauge field is localized on the wall that interpolates between $A$ and $B$.

1413:  With respect to $U(1)_2$, vacuum $A$ is in

1414: the Higgs phase while vacuum $B$ in the Coulomb phase. The two flux tubes in vacuum $A$ can thus form a Y junction like the one previously described.

1415: So we can

1416: have a baryon vertex like the one of Figure \ref{yjunction}.

1417: \begin{figure}[h!t]

1418: \begin{center}

1419: \leavevmode \epsfxsize 10 cm \epsffile{yjunction.eps}

1420: \end{center}

1421: \caption{\footnotesize A Y junction for the baryon vertex. The

1422: three quarks are choosen so that the foundamental of $SU(3)$ is

1423: $(Q_1,Q_2,Q_3)$. $Q_1$ and $Q_2$ are connected by a wall vortex

1424: that carries the flux of the $t_3$ generator. The quark $Q_3$ is

1425: connected to another tube that carries the $t_8$ charge makes a

1426: junction like \ref{nahiggs} with the wall vortex. The charge of

1427: $Q_3$ is divided in two equal parts when the $t_8$ tube join the

1428: $t_3$ wall vortex, and this is consistent with the fact that $t_8

1429: \propto \mathrm{diag}(1,1,-2)$. } \label{yjunction}

1430: \end{figure}

1431: 

1432: 

1433: %\section{Conclusions}

1434: 

1435: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%Riassunto-dei-risultati%%%%%%%%%%%%%%%%%%

1436: 

1437: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%wall vortex non e' mai BPS%%%%%%%%%%%%%%%%%%%%%

1438: 

1439: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%non conosco il wall limit per na theoires%%%%%%%%%%%%%%%%%%%%%%

1440: 

1441: %%%%%%%%%%%%%%%%%%%%%%%%%%%la classificazione non pretende di essere completa%%%%%%%%%%%%%%%%%%%

1442: 

1443: %\appendix

1444: 

1445: 

1446: 

1447: 

1448: \section* {Acknowledgement}

1449:  I want to thank the following people for useful comments and

1450:  discussions: J.~Evslin, K.~Konishi, H.~B.~Nielsen, A.~Ritz and A.~Smilga.

1451: 

1452: 

1453: 

1454: \begin{thebibliography}{}

1455: 

1456: 

1457: \bibitem{A}

1458:   A.~A.~Abrikosov,

1459:   %``On The Magnetic Properties Of Superconductors Of The Second Group,''

1460:   Sov.\ Phys.\ JETP {\bf 5} (1957) 1174

1461:   [Zh.\ Eksp.\ Teor.\ Fiz.\  {\bf 32} (1957) 1442].

1462:   %%CITATION = SPHJA,5,1174;%%

1463: 

1464: \bibitem{NO}

1465: 

1466:   H.~B.~Nielsen and P.~Olesen,

1467:    %``Vortex-Line Models For Dual Strings,''

1468:   %

1469:   Nucl.\ Phys.\ B {\bf 61}, 45 (1973).

1470:   %%CITATION = NUPHA,B61,45;%%

1471: 

1472: 

1473: %\cite{Abraham:1992qv}

1474: \bibitem{Abraham:1992qv}

1475: 

1476: 

1477:   E.~R.~C.~Abraham and P.~K.~Townsend,

1478:   %``Q kinks,''

1479:   Phys.\ Lett.\ B {\bf 291} (1992) 85.

1480:   %%CITATION = PHLTA,B291,85;%%

1481: 

1482: 

1483: 

1484:   E.~R.~C.~Abraham and P.~K.~Townsend,

1485:   %``More on Q kinks: A (1+1)-dimensional analog of dyons,''

1486:   Phys.\ Lett.\ B {\bf 295} (1992) 225.

1487:   %%CITATION = PHLTA,B295,225;%%

1488: 

1489: 

1490: 

1491: \bibitem{Gauntlett:2000de}

1492:   J.~P.~Gauntlett, R.~Portugues, D.~Tong and P.~K.~Townsend,

1493:   %``D-brane solitons in supersymmetric sigma-models,''

1494:   Phys.\ Rev.\ D {\bf 63} (2001) 085002

1495:   [arXiv:hep-th/0008221].

1496:   %CITATION = HEP-TH 0008221;%%

1497: 

1498: 

1499: \bibitem{Shifman:2002jm}

1500:   M.~Shifman and A.~Yung,

1501:    %``Domain walls and flux tubes in N = 2 SQCD: D-brane prototypes,''

1502:   %

1503:   Phys.\ Rev.\ D {\bf 67}, 125007 (2003)

1504:   [arXiv:hep-th/0212293].

1505:   %%CITATION = HEP-TH 0212293;%%

1506: 

1507: 

1508: \bibitem{nahiggs}

1509:  D.~Tong,

1510:   %``Monopoles in the Higgs phase,''

1511:   Phys.\ Rev.\ D {\bf 69} (2004) 065003

1512:   [arXiv:hep-th/0307302].

1513:   %%CITATION = HEP-TH 0307302;%%

1514: 

1515:  A.~Hanany and D.~Tong,

1516:   %``Vortices, instantons and branes,''

1517:   JHEP {\bf 0307} (2003) 037

1518:   [arXiv:hep-th/0306150].

1519:   %%CITATION = HEP-TH 0306150;%%

1520: 

1521:  A.~Hanany and D.~Tong,

1522:   %``Vortex strings and four-dimensional gauge dynamics,''

1523:   JHEP {\bf 0404} (2004) 066

1524:   [arXiv:hep-th/0403158].

1525:   %%CITATION = HEP-TH 0403158;%%

1526: 

1527: 

1528:   Y.~Isozumi, M.~Nitta, K.~Ohashi and N.~Sakai,

1529:   %``All exact solutions of a 1/4 Bogomol'nyi-Prasad-Sommerfield equation,''

1530:   Phys.\ Rev.\ D {\bf 71} (2005) 065018

1531:   [arXiv:hep-th/0405129].

1532:   %%CITATION = HEP-TH 0405129;%%

1533: 

1534: 

1535:   N.~Sakai and D.~Tong,

1536:   %``Monopoles, vortices, domain walls and D-branes: The rules of interaction,''

1537:   JHEP {\bf 0503} (2005) 019

1538:   [arXiv:hep-th/0501207].

1539:   %%CITATION = HEP-TH 0501207;%%

1540: 

1541: 

1542: 

1543:   R.~Auzzi, M.~Shifman and A.~Yung,

1544:   %``Studying boojums in N = 2 theory with walls and vortices,''

1545:   arXiv:hep-th/0504148.

1546:   %%CITATION = HEP-TH 0504148;%%

1547: 

1548:  M.~Eto, Y.~Isozumi, M.~Nitta and K.~Ohashi,

1549:   %``1/2, 1/4 and 1/8 BPS Equations in SUSY Yang-Mills-Higgs Systems -- Field

1550:   %Theoretical Brane Configurations --,''

1551:   arXiv:hep-th/0506257.

1552:   %%CITATION = HEP-TH 0506257;%%

1553: 

1554: 

1555: K.~Lee and H.~U.~Yee,

1556:   %``New BPS Objects in N=2 Supersymmetric Gauge Theories,''

1557:   arXiv:hep-th/0506256.

1558:   %%CITATION = HEP-TH 0506256;%%

1559: 

1560:  A.~Hanany and D.~Tong,

1561:   %``On Monopoles and Domain Walls,''

1562:   arXiv:hep-th/0507140.

1563:   %%CITATION = HEP-TH 0507140;%%

1564: 

1565: 

1566: 

1567: 

1568: 

1569: 

1570: 

1571: 

1572: 

1573: 

1574: %\cite{Witten:1997ep}

1575: \bibitem{Witten:1997ep}

1576:   E.~Witten,

1577:   %``Branes and the dynamics of {QCD},''

1578:   Nucl.\ Phys.\ B {\bf 507} (1997) 658

1579:   [arXiv:hep-th/9706109].

1580:   %%CITATION = HEP-TH 9706109;%%

1581: 

1582: 

1583: 

1584: 

1585: %\cite{Cachazo:2002zk}

1586: \bibitem{Cachazo:2002zk}

1587:   F.~Cachazo, N.~Seiberg and E.~Witten,

1588:   %``Phases of N = 1 supersymmetric gauge theories and matrices,''

1589:   JHEP {\bf 0302} (2003) 042

1590:   [arXiv:hep-th/0301006].

1591:   %%CITATION = HEP-TH 0301006;%%

1592: 

1593: \bibitem{ioeken}

1594: S.~Bolognesi and K.~Konishi, {\it in preparation}.

1595: 

1596: 

1597: 

1598: 

1599: 

1600: \bibitem{SW}

1601:   N.~Seiberg and E.~Witten,

1602:   %``Electric - magnetic duality, monopole condensation, and confinement in N=2

1603:   %supersymmetric Yang-Mills theory,''

1604:   Nucl.\ Phys.\ B {\bf 426} (1994) 19

1605:   [Erratum-ibid.\ B {\bf 430} (1994) 485]

1606:   [arXiv:hep-th/9407087].

1607:   %%CITATION = HEP-TH 9407087;%%

1608: 

1609: 

1610: \bibitem{DS}

1611:   M.~R.~Douglas and S.~H.~Shenker,

1612:    %``Dynamics of SU(N) supersymmetric gauge theory,''

1613:   %

1614:   Nucl.\ Phys.\ B {\bf 447}, 271 (1995)

1615:   [arXiv:hep-th/9503163].

1616:   %%CITATION = HEP-TH 9503163;%%

1617: 

1618: 

1619: \bibitem{vainyung}

1620:   A.~I.~Vainshtein and A.~Yung,

1621:   %``Type I superconductivity upon monopole condensation in Seiberg-Witten

1622:   %theory,''

1623:   Nucl.\ Phys.\ B {\bf 614} (2001) 3

1624:   [arXiv:hep-th/0012250].

1625:   %%CITATION = HEP-TH 0012250;%%

1626: 

1627: 

1628: 

1629: 

1630: 

1631: %\cite{Hanany:1997hr}

1632: \bibitem{Hanany:1997hr}

1633:   A.~Hanany, M.~J.~Strassler and A.~Zaffaroni,

1634:   %``Confinement and strings in M{QCD},''

1635:   Nucl.\ Phys.\ B {\bf 513} (1998) 87

1636:   [arXiv:hep-th/9707244].

1637:   %%CITATION = HEP-TH 9707244;%%

1638: 

1639: %\cite{Bogomolny:1975de}

1640: \bibitem{Bogomolny:1975de}

1641:   E.~B.~Bogomolny,

1642:   %``Stability Of Classical Solutions,''

1643:   Sov.\ J.\ Nucl.\ Phys.\  {\bf 24} (1976) 449

1644:   [Yad.\ Fiz.\  {\bf 24} (1976) 861].

1645:   %%CITATION = SJNCA,24,449;%%

1646: 

1647: 

1648: \bibitem{Edelstein:1993bb}

1649:   J.~D.~Edelstein, C.~Nunez and F.~Schaposnik,

1650:   %``Supersymmetry and Bogomolny equations in the Abelian Higgs model,''

1651:   Phys.\ Lett.\ B {\bf 329} (1994) 39

1652:   [arXiv:hep-th/9311055].

1653:   %%CITATION = HEP-TH 9311055;%%

1654: 

1655: 

1656: 

1657: 

1658:   %\cite{Bolognesi:2004yh}

1659:    \bibitem{Bolognesi:2004yh}

1660:     S.~Bolognesi,

1661:   %``The holomorphic tension of vortices,''

1662:   JHEP {\bf 0501} (2005) 044

1663:   [arXiv:hep-th/0411075].

1664:   %%CITATION = HEP-TH 0411075;%%

1665: 

1666: 

1667: R.~Auzzi, S.~Bolognesi and J.~Evslin,

1668:   %``Monopoles can be confined by 0, 1 or 2 vortices,''

1669:   JHEP {\bf 0502}, 046 (2005)

1670:   [arXiv:hep-th/0411074].

1671:   %%CITATION = HEP-TH 0411074;%%

1672: 

1673: 

1674: 

1675: \bibitem{miononabeliani}

1676: 

1677: S.~Bolognesi, {\it in preparation}.

1678: 

1679: 

1680: \bibitem{nonabelian}

1681: 

1682: 

1683:   M.~Shifman and A.~Yung,

1684:   %``Localization of non-Abelian gauge fields on domain walls at weak  coupling

1685:   %(D-brane prototypes II),''

1686:   Phys.\ Rev.\ D {\bf 70} (2004) 025013

1687:   [arXiv:hep-th/0312257].

1688:   %%CITATION = HEP-TH 0312257;%%

1689: 

1690:  Y.~Isozumi, M.~Nitta, K.~Ohashi and N.~Sakai,

1691:   %``Construction of non-Abelian walls and their complete moduli space,''

1692:   Phys.\ Rev.\ Lett.\  {\bf 93} (2004) 161601

1693:   [arXiv:hep-th/0404198].

1694:   %%CITATION = HEP-TH 0404198;%%

1695: 

1696: 

1697:  Y.~Isozumi, M.~Nitta, K.~Ohashi and N.~Sakai,

1698:   %``Non-Abelian walls in supersymmetric gauge theories,''

1699:   Phys.\ Rev.\ D {\bf 70} (2004) 125014

1700:   [arXiv:hep-th/0405194].

1701:   %%CITATION = HEP-TH 0405194;%%

1702: 

1703: 

1704:  M.~Eto, Y.~Isozumi, M.~Nitta, K.~Ohashi, K.~Ohta and N.~Sakai,

1705:   %``D-brane construction for non-Abelian walls,''

1706:   Phys.\ Rev.\ D {\bf 71} (2005) 125006

1707:   [arXiv:hep-th/0412024].

1708:   %%CITATION = HEP-TH 0412024;%%

1709: 

1710: 

1711: 

1712: \bibitem{Dvali:1996xe}

1713:   G.~R.~Dvali and M.~A.~Shifman,

1714:    %``Domain walls in strongly coupled theories,''

1715:   %

1716:   Phys.\ Lett.\ B {\bf 396}, 64 (1997)

1717:   [Erratum-ibid.\ B {\bf 407}, 452 (1997)]

1718:   [arXiv:hep-th/9612128].

1719:   %%CITATION = HEP-TH 9612128;%%

1720: 

1721: 

1722: 

1723: \bibitem{Shifman}

1724: 

1725:  A.~Kovner and M.~A.~Shifman,

1726:   %``Chirally symmetric phase of supersymmetric gluodynamics,''

1727:   Phys.\ Rev.\ D {\bf 56} (1997) 2396

1728:   [arXiv:hep-th/9702174].

1729:   %%CITATION = HEP-TH 9702174;%%

1730: 

1731: 

1732:  A.~Kovner, M.~A.~Shifman and A.~Smilga,

1733:   %``Domain walls in supersymmetric Yang-Mills theories,''

1734:   Phys.\ Rev.\ D {\bf 56} (1997) 7978

1735:   [arXiv:hep-th/9706089].

1736:   %%CITATION = HEP-TH 9706089;%%

1737: 

1738: 

1739: I.~I.~Kogan, A.~Kovner and M.~A.~Shifman,

1740:   %``More on supersymmetric domain walls, N counting and glued potentials,''

1741:   Phys.\ Rev.\ D {\bf 57} (1998) 5195

1742:   [arXiv:hep-th/9712046].

1743:   %%CITATION = HEP-TH 9712046;%%

1744: 

1745: 

1746: 

1747: \bibitem{Dvali}

1748: 

1749:  G.~R.~Dvali, G.~Gabadadze and Z.~Kakushadze,

1750:   %``BPS domain walls in large N supersymmetric {QCD},''

1751:   Nucl.\ Phys.\ B {\bf 562} (1999) 158

1752:   [arXiv:hep-th/9901032].

1753:   %%CITATION = HEP-TH 9901032;%%

1754: 

1755:   G.~R.~Dvali and Z.~Kakushadze,

1756:   %``Large N domain walls as D-branes for N = 1 {QCD} string,''

1757:   Nucl.\ Phys.\ B {\bf 537} (1999) 297

1758:   [arXiv:hep-th/9807140].

1759:   %%CITATION = HEP-TH 9807140;%%

1760: 

1761: 

1762: 

1763: 

1764: 

1765: \bibitem{Acharya:2001dz}

1766:   B.~S.~Acharya and C.~Vafa,

1767:   %``On domain walls of N = 1 supersymmetric Yang-Mills in four dimensions,''

1768:   arXiv:hep-th/0103011.

1769:   %%CITATION = HEP-TH 0103011;%%

1770: 

1771: 

1772: 

1773: 

1774: 

1775: 

1776: 

1777: 

1778: 

1779: 

1780: 

1781: 

1782: 

1783: 

1784: 

1785: 

1786: \bibitem{Homotopy}

1787: S.~Bolognesi and J.~Evslin,

1788:   %``Stable vs Unstable Vortices in SQCD,''

1789:   arXiv:hep-th/0506174.

1790:   %%CITATION = HEP-TH 0506174;%%

1791: 

1792: 

1793: 

1794: 

1795: 

1796: 

1797: 

1798: 

1799: 

1800: \bibitem{Artru:1974zn}

1801:   X.~Artru,

1802:   %``String Model With Baryons: Topology; Classical Motion,''

1803:   Nucl.\ Phys.\ B {\bf 85} (1975) 442.

1804:   %%CITATION = NUPHA,B85,442;%%

1805: 

1806: 

1807: 

1808: 

1809: 

1810: 

1811: 

1812:  \bibitem{Witten:1998xy}

1813:    E.~Witten,

1814:    %``Baryons and branes in anti de Sitter space,''

1815:    JHEP {\bf 9807} (1998) 006

1816:    [arXiv:hep-th/9805112].

1817:    %%CITATION = HEP-TH 9805112;%%

1818: 

1819: 

1820:  \bibitem{Rajaraman}

1821: R.\ Rajaraman, {\it Solitons and Instantons}, (North Holland,

1822: Amsterdam, 1982).

1823: 

1824: 

1825: 

1826: \end{thebibliography}

1827: 

1828: \end{document}

1829: