1: \documentclass[12pt]{article}

2: %\documentstyle[psfig,12pt]{article}

3: 

4: \textheight=8.57truein

5: \textwidth=6.1truein

6: \overfullrule=0pt

7: \parskip=2pt

8: \parindent=12pt

9: \headheight=0in

10: \headsep=0in

11: \topmargin=0.3in

12: \oddsidemargin=0in

13: 

14: %--------+---------+---------+---------+---------+---------+---------+

15: \newsavebox{\ns}

16: \newsavebox{\dbrane}

17: \newsavebox{\dbshort}

18: 

19: \renewcommand{\arraystretch}{1.2}

20: 

21: \usepackage{epsfig}

22: \usepackage{amssymb}

23: 

24: %\newcommand{\sect}[1]{\setcounter{equation}{0}\section{#1}}

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

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

27: %       {\setcounter{section}{0}

28: %       \gdef\thesection{\Alph{section}}}\section}

29: 

30: \def\be{\begin{eqnarray}}

31: \def\ee{\end{eqnarray}}

32: 

33: 

34: \newcommand{\nn}{\nonumber}

35: \newcommand\para{\paragraph{}}

36: \newcommand{\ft}[2]{{\textstyle\frac{#1}{#2}}}

37: \newcommand{\eqn}[1]{(\ref{#1})}

38: \newcommand\balpha{\mbox{\boldmath $\alpha$}}

39: \newcommand\bbeta{\mbox{\boldmath $\beta$}}

40: \newcommand\bgamma{\mbox{\boldmath $\gamma$}}

41: \newcommand\bomega{\mbox{\boldmath $\omega$}}

42: \newcommand\blambda{\mbox{\boldmath $\lambda$}}

43: \newcommand\bmu{\mbox{\boldmath $\mu$}}

44: \newcommand\bphi{\mbox{\boldmath $\phi$}}

45: \newcommand\bzeta{\mbox{\boldmath $\zeta$}}

46: \newcommand\bsigma{\mbox{\boldmath $\sigma$}}

47: \newcommand\bepsilon{\mbox{\boldmath $\epsilon$}}

48: \newcommand\btau{\mbox{\boldmath $\tau$}}

49: 

50: \def\Dslash{\,\,{\raise.15ex\hbox{/}\mkern-12mu D}}

51: \def\Dbarslash{\,\,{\raise.15ex\hbox{/}\mkern-12mu {\bar D}}}

52: \def\delslash{\,\,{\raise.15ex\hbox{/}\mkern-9mu \partial}}

53: \def\delbarslash{\,\,{\raise.15ex\hbox{/}\mkern-9mu {\bar\partial}}}

54: \def\pslash{\,\,{\raise.15ex\hbox{/}\mkern-9mu p}}

55: \def\calDslash{\,\,{\raise.15ex\hbox{/}\mkern-12mu {\cal D}}}

56: \newcommand\Bprime{B${}^\prime$}

57: \newcommand{\sign}{{\rm sign}}

58: 

59: 

60: \def\pp{\mathbf{\Phi}}

61: \def\be{\begin{eqnarray}}

62: \def\ee{\end{eqnarray}}

63: \def\tr{{\rm Tr}}

64: \def\dd{{\rm d}}

65: \def\D{{\cal D}}

66: \def\vac{|\,0\rangle}

67: \def\nn{\nonumber}

68: % bold face in formulas

69: \def\mbf#1{\mathchoice{\hbox{\boldmath $\displaystyle #1$}}

70:         {\hbox{\boldmath $\textstyle #1$}}

71:         {\hbox{\boldmath $\scriptstyle #1$}}

72:         {\hbox{\boldmath $\scriptscriptstyle #1$}}}

73: 

74: \def\dblone{\hbox{$1\hskip -1.2pt\vrule depth 0pt height 1.6ex width 0.7pt

75:                   \vrule depth 0pt height 0.3pt width 0.12em$}}

76: 

77: 

78: \def\bR{\mathbb{R}}

79: \def\bC{\mathbb{C}}

80: \def\bH{\mathbb{H}}

81: \def\bO{\mathbb{O}}

82: \def\bF{\mathbb{F}}

83: \def\bP{\mathbb{P}}

84: \def\bK{\mathbb{K}}

85: 

86: 

87: 

88: \begin{document}

89: \pagestyle{plain}

90: \setcounter{page}{1}

91: \newcounter{bean}

92: \baselineskip16pt

93: 

94: 

95: \begin{titlepage}

96: 

97: \begin{center}

98: \today

99: \hfill {.}\\

100: %\hfill MIT-CTP-3390 \\

101: 

102: 

103: \vskip 1.5 cm {\large \bf The Berry Phase of D0-Branes}

104: \vskip 1cm

105: {Chris Pedder, Julian Sonner and David Tong}\\

106: \vskip 1cm {\sl Department of Applied Mathematics and Theoretical

107: Physics, \\ University of Cambridge, UK}

108: 

109: \end{center}

110: 

111: 

112: \vskip 0.5 cm

113: \begin{abstract}

114: We study $SU(2)$ Yang-Mills quantum mechanics with $N=2,4,8$ and

115: $16$ supercharges. This describes the non-relativistic dynamics of

116: a pair of D0-branes moving in $d=3,4,6$ and $10$ spacetime

117: dimensions respectively. We show that as the D0-branes orbit,

118: states undergo a Berry holonomy described by the four Hopf maps. For the $N=2$

119: theory, the associated Hopf map is the ${\bf Z}_2$ M\"obius bundle

120: and its effect is to turn the D0-branes into anyons with exchange

121: statistics $e^{i\pi/2}$. For the $N=4,8$ and $16$ theories, the

122: Hopf maps give rise to Berry connections that are familiar to

123: physicists: the $U(1)$ Dirac monopole; the $SU(2)$ Yang monopole;

124: and the $SO(8)$ octonionic monopole.

125: 

126: 

127: 

128: \end{abstract}

129: 

130: 

131: \end{titlepage}

132: 

133: \subsection*{1. Introduction}

134: 

135: Many years ago, Kugo and Townsend \cite{kugo} pointed out a

136: relationship between  supersymmetric field theories with $N=2,4,8$

137: and $16$ supercharges and the four normed division algebras

138: $\bK\cong \bR$, $\bC$, $\bH$ and $\bO$. The key observation is

139: algebraic. Theories with $N=2,4,8$ and $16$ supercharges naturally

140: live in $d=3,4,6$ and $10$ spacetime dimensions respectively. The

141: Lorentz Lie algebra in these dimensions is isomorphic to the

142: algebra of $2\times 2$ Hermitian matrices with vanishing trace and

143: elements in $\bK$,

144: %

145: \be sl(2;\bK) \cong so(d-1,1) \label{the}\ee

146: %

147: This allows us to express a spinor in $d$ dimensions as a

148: 2-component $\bK$-vector, generalizing the well-known result for

149: $d=4$. This construction was elaborated upon in \cite{jon}.

150: 

151: \para

152: For theories with $N=2,4$ and $8$ supercharges, the relationship

153: to the division algebra $\bK$ also manifests itself in more

154: physical and dynamical matters. This includes familiar features of

155: supersymmetric theories, such as the holomorphy of the

156: superpotential and the hyperK\"ahler/quaternionic structure of

157: Calabi-Yau moduli spaces. However, so far the tantalizing idea

158: that an octonionic structure underlies theories with 16

159: supercharges has not led to major insight about quantum dynamics.

160: 

161: \para

162: The purpose of this short note is to point out that the

163: isomorphism \eqn{the} has a simple, physical consequence in

164: the framework of supersymmetric quantum mechanics. We focus on

165: $SU(2)$ gauged quantum mechanics. The theory with $N$ supercharges

166: can be thought of as the dimensional reduction of the minimal

167: super Yang-Mills theory in $d=3,4,6$ and 10 dimensions, and

168: describes the non-relativistic dynamics of two D0-branes moving in

169: $d-1=N/2+1$ spatial dimensions. We show that the states of a pair

170: of orbiting D0-branes undergo a holonomy described by the Hopf map

171: associated to the division algebra $\bR$, $\bC$, $\bH$ and $\bO$.

172: 

173: \para

174: The four Hopf maps take $S^{N-1}\rightarrow S^{N/2}$:

175: %

176: \be \begin{array}{c} S^{1} \\ \ \,\Bigg{\downarrow} {}^{S^0\cong Z_2} \\

177: S^1

178: \end{array}\ \ \ \ \begin{array}{c} S^{3} \\ \ \,\Bigg{\downarrow} {}^{S^1} \\

179: S^2

180: \end{array}\ \ \ \ \begin{array}{c} S^{7} \\ \ \,\Bigg{\downarrow} {}^{S^3} \\

181: S^4

182: \end{array}\ \ \ \ \begin{array}{c} S^{15} \\ \ \,\Bigg{\downarrow} {}^{S^7} \\

183: S^8

184: \end{array}\nn\ee

185: %

186: %

187: %In each case, the total manifold $S^{N-1}$ can be thought of as

188: %arising from the pair $(q_1,q_2)\in \bK^2$, subject to the

189: %constraint $|q_1|^2 + |q_2|^2=1$. Meanwhile, the base manifold is

190: %the projective space, $S^{N/2}\cong \bK\bP^1$. We will describe

191: %the Hopf map in more detail in Section 3.

192: %

193: The case of $N=2$ supercharges is special. The map ${\bf

194: Z}_2\hookrightarrow S^1\rightarrow S^1$, sometimes known as the

195: zeroth Hopf map, describes the M\"obius bundle. We will show in

196: Section 3.1 that the ground state of the associated quantum

197: mechanics undergoes a discrete ${\bf Z}_2$ holonomy, changing by a

198: sign as D0-branes orbit. Since a single orbit is equivalent to two

199: exchanges, this implies that the D0-branes in $N=2$ Yang-Mills

200: quantum mechanics are anyons: they obey exchange statistics

201: $e^{\pm i \pi/2}=\pm i$. As we review in Section 3, the remaining

202: three Hopf maps induce connections over the base space $S^{d-2}$

203: which subsequently arise as Berry connections in the quantum

204: mechanics. Each is familiar to physicists: they are the $U(1)$

205: Dirac monopole, the $SU(2)$ Yang monopole \cite{yang}, and the

206: $SO(8)$ octonionic monopole \cite{grossman,tchrak}.

207: 

208: \para

209: This quartet of Hopf bundles are the fab four of geometrical

210: phases. The discrete ${\bf Z}_2$ holonomy was one of the earliest

211: known Berry phases \cite{chemist}, while the Dirac monopole

212: appeared as an example in Berry's original paper \cite{berry}.

213: Both the Yang monopole and the octonionic monopole have also

214: arisen in various contexts. The former was first introduced as a

215: non-Abelian Berry phase in \cite{avron,feher}, and has subsequently

216: appeared in several condensed matter systems

217: \cite{bigzhang,middlezhang,smallzhang}, with various properties

218: explored in \cite{levay}. The $SO(8)$ octonionic monopole appeared

219: previously in association with the eight-dimensional quantum Hall

220: effect \cite{bern}. In this paper we will see how all these

221: connections arise as (non-Abelian) Berry phases in Yang-Mills

222: quantum mechanics with $N$ supercharges.

223: 

224: \para

225: The present paper can be thought of as a follow-up to our recent

226: work \cite{us,second} exploring the Berry phase that emerges in

227: various supersymmetric quantum mechanics. In \cite{us}, we showed

228: that certain $N=4$  quantum systems naturally give rise to the

229: Dirac monopole and associated constructs such as the smooth 't

230: Hooft-Polyakov monopole. In \cite{second}, we studied a quantum

231: theory with $N=8$ supercharges and showed that a deformation of

232: the Yang-monopole arises as the spin connection of a dual

233: gravitational background. The Yang monopole and octonionic monopole 

234: also appeared in a related context in \cite{usp}.

235: 

236: 

237: 

238: 

239: 

240: 

241: 

242: \subsection*{2. D0-Brane Dynamics}

243: 

244: In this paper we study the $SU(2)$ super-Yang-Mills quantum

245: mechanics with $N=2,4,8$ and $16$ supercharges. The Lagrangians

246: are the dimensional reduction of minimal super Yang-Mills in

247: $d=3,4,8$ and $10$ dimensions respectively, and each takes the

248: form

249: %

250: \be \label{lag} L = \frac{1}{2g^2}\,\tr \left( \left( \D_0 X_i

251: \right)^2 + \sum_{i<j}\left[ X_i , X_j \right]^2 +i \bar\psi\D_0

252: \psi + \bar\psi \Gamma^i \left[ X_i , \psi \right]

253: \right)\,, \ee %

254: where $X_i$ and $\psi_\alpha$ are both $su(2)$ valued fields. The

255: inverse coupling constant $1/g^2$ is the mass of the D0-brane.

256: Lagrangians with different amounts of supersymmetry are

257: distinguished by the number of scalar fields $X_i$, and the

258: structure of the Grassmann parameters $\psi$. The scalar index

259: runs over $i=1,\ldots, d-1$ and the Lagrangian describes the

260: non-relativistic relative motion of a pair of D0-branes moving in

261: $d-1=2,3,5$ and $9$ spatial dimensions\footnote{More precisely,

262: the theory with $N=16$ supercharges  describes the dynamics of a

263: pair of D0-branes in flat ten-dimensional spacetime. Lagrangians

264: with less supersymmetry describe the dynamics of a pair of

265: fractional D0-branes, trapped to lie at a suitable singularity in

266: a K3, a CY 3-fold, or a $G_2$ holonomy manifold.}. The matrices

267: $\Gamma^i$ satisfy the $d-1$ dimensional Euclidean Clifford

268: algebra: $\{\Gamma^i,\Gamma^j\}=2\delta^{ij}$. The spinors are

269: real or complex, depending on the possible representations of the

270: Clifford algebra:

271: %

272: \begin{itemize}

273: \item $N=2$: For $d-1=2$, we may choose the real Pauli matrices

274: $\Gamma^1=\sigma^1$, $\Gamma^2=\sigma^3$. The spinor

275: $\psi_\alpha$, with $\alpha=1,2$, is also real.

276: %

277: \item $N=4$: For $d-1=3$, there are only complex representations

278: of the Clifford algebra. We take the Pauli matrices

279: $\Gamma^i=\sigma^i$. The spinor $\psi_\alpha$, with $\alpha=1,2$,

280: is complex.

281: %

282: \item $N=8$: The Clifford algebra in $d-1=5$ dimensions is again

283: complex. The complex spinor $\psi_\alpha$ has $\alpha=1,\ldots,4$.

284: %

285: \item $N=16$: The $d-1=9$ Clifford algebra admits a real

286: representation. The real spinor $\psi_\alpha$ has

287: $\alpha=1,\ldots,16$.

288: \end{itemize}

289: %

290: 

291: 

292: \subsubsection*{2.1 The Born-Oppenheimer Approximation}

293: 

294: We work in the Born-Oppenheimer approximation, considering

295: well-localized wavepackets describing D0-branes with separation

296: $\hat{X}^i$. To this end, we expand around the classical

297: background

298: %

299: \be \langle X^i\rangle = \ft12 \hat{X}^i\sigma^3 \label{bo}\ee

300: %

301: and study the effects of high-frequency modes of open strings

302: stretched between the D0-branes. Our results are applicable in the

303: weak coupling limit $\hat{X}^3\gg g^2$. This typically excludes

304: the ground state of the full system, for which the wavefunction is

305: localized close to the origin $\hat{X} \sim g^{2/3}$.

306: 

307: \para

308: The background \eqn{bo} breaks the gauge symmetry to the Cartan

309: subalgebra: $SU(2)\rightarrow U(1)$. The remnant ${\bf Z}_2$ Weyl

310: group acts as $\hat{X}^i\rightarrow -\hat{X}^i$. The configuration

311: space of the D0-branes is thus ${\bf R}^d/{\bf Z}_2$,  reflecting

312: the indistinguishability of the particles.  We expand,

313: %

314: \be \label{bmasses} X^i = \langle X^i\rangle + \ft12 x^i_m

315: \sigma^m \ee

316: %

317: The bosonic commutator term in \eqn{lag} provides oscillation

318: frequencies for the fluctuations $x^i_m$. The fluctuations

319: $x^i_3$, lying in the unbroken $U(1)$, are zero modes and describe

320: the relative motion of the D0-brane pair in $\bR^{d-1}$. In

321: contrast, the $m=1,2$ components, lying in the broken part of the

322: gauge group, pick up non-zero frequencies and describe the

323: excitation of open strings stretched between the D0-branes. We

324: form the complex combination

325: %

326: \be z^i=x^i_1+ix^i_2\ ,\ \ \ \ i=1,\ldots, d-1\ee

327: %

328: which has charge $+1$ under the unbroken $U(1)$ gauge symmetry.

329: Introducing the complex conjugate momentum $\pi_i$, the leading

330: order free Hamiltonian for these off-diagonal bosonic modes is

331: given by,

332: %

333: \be H_B = \frac{1}{2g^2}\left[ {\cal P}_{ij}\left( \pi_i

334: \bar\pi_j+ \hat X^2 z_i\bar z_j \right) + \ldots

335: %{\cal O}\left(\frac{g}{|\hat{X}|^{1/3}}\right)

336: \right] \ee

337: %

338: where $\ldots$ are interaction parts of the Hamiltonian which are

339: perturbations of order $g/\hat{X}^{1/3}$. The Hamiltonian includes

340: the projection operator

341: %

342: \be {\cal P}_{ij} = \delta_{ij}-\frac{\hat X^i \hat X^j}{\hat

343: X^2}\label{bproj}\ee

344: %

345: and therefore describes only $d-2$ complex harmonic oscillators.

346: These oscillators can be thought of as the transverse excitations

347: of a string stretched between the D0-branes. The remaining degree

348: of freedom falls victim to the broken gauge symmetry on the

349: D0-branes; it is analogous to the scalar that is eaten by the

350: Higgs mechanism in higher dimensions.

351: 

352: \para

353: We may also expand the fermions around the background \eqn{bo}.

354: Once again, the fermions $(\psi_\alpha)_3$, lying in the Cartan

355: subalgebra, provide zero modes. Upon quantization, these fill out

356: a $2^{N/2}$ dimensional multiplet of states whose degeneracy is

357: split only by interaction terms. In contrast, the off-diagonal

358: components $(\psi_\alpha)_1$ and $(\psi_\alpha)_2$ have non-zero

359: frequencies. Here it is convenient to differentiate between the

360: $N=2,16$ cases, which have real fermions and the $N=4,8$ cases,

361: which have complex fermions. In the former case, we introduce the

362: complexified Grassmann parameter,

363: %

364: \be \label{eq:complexspinor} \Psi=\psi_1 + i \psi_2\,. \ee

365: %

366: in terms of which the fermionic Hamiltonian is written as

367: %

368: \be H_F^{N=2,16} = \frac{1}{2g^2}\left[\Psi^\dagger \left( \hat X

369: \cdot \Gamma \right)\Psi + \ldots\right]\label{hf216}\ee

370: %

371: The $N=4,8$ cases have complex spinors from the outset. We may now

372: form two linearly independent complex combinations

373: %

374: \be \Psi =  \frac{1}{\sqrt{2}} \left(  \psi_1 + i \psi_2 \right)\

375: \ \ ,\ \ \  \tilde{\Psi} =  \frac{1}{\sqrt{2}} \left( \psi_1 - i

376: \psi_2 \right) \ee

377: %

378: with respective charges $+1$ and $-1$ under the unbroken $U(1)$

379: gauge group. The fermionic part of the Hamiltonian is now given by

380: %

381: \be H_F^{N=4,8}=\frac{1}{2g^2}\left[\Psi^\dagger \left(\hat X

382: \cdot \Gamma \right)\Psi - \tilde{\Psi}^\dagger \left(  \hat X

383: \cdot \Gamma \right)\tilde{\Psi}+\ldots\right] \label{hf48}\ee

384: %

385: To summarize, the free part of the Hamiltonian $H=H_B+H_F$ for the

386: massive oscillators contains $N/2$ complex scalars and $N$ complex

387: Grassmann parameters. The interaction Hamiltonian is of order

388: $g/\hat{X}^{1/3}$.

389: 

390: 

391: \subsubsection*{2.2 Quantization and the Hilbert Space}

392: 

393: In the Born-Oppenheimer approximation, we treat the massive

394: oscillator states $z_i$, $\Psi_\alpha$ and $\tilde{\Psi}_\alpha$

395: quantum mechanically in the classical background $\hat{X}$. The

396: zero frequency modes $\hat{X}^i$ and $(\psi_\alpha)_3$ are

397: quantized subsequently. The approximation holds as long as

398: $\hat{X}\gg g^{1/3}$ ensuring that we keep a separation of scales,

399: meaning that the wavefunction for $\hat{X}$ should not have

400: significant support near the origin where the two D0-branes

401: approach. For example, this will be the case for excited states of

402: orbiting D0-branes that carry large angular momentum. In the

403: following section we will study the holonomy of the excited states

404: of the massive oscillators as the two D0-branes orbit. We first

405: briefly describe the Hilbert space of these states.

406: 

407: \para

408: Working in the regime $\hat{X}\gg g^{1/3}$, we may restrict

409: attention to the free theory. The $d-2$ massive complex scalars

410: have ground state energy  $E_B=(d-2)\hat{X}/2g^2$. One may

411: construct the Hilbert space by acting with (suitably projected)

412: creation operators ${\cal P}_{ij}\,a^\dagger_j$ and ${\cal

413: P}_{ij}\,\bar{a}^\dagger_j$, where

414: %

415: \be a^\dagger_i=\frac{(\hat{X}z_i

416: -i\bar{\pi}_i)}{\sqrt{2\hat{X}}}\ \ \ ,\ \ \ \bar{a}_i^\dagger =

417: \frac{(\hat{X} \bar{z}_i-i\pi_i)}{\sqrt{2\hat{X}}} \ee

418: %

419: create quanta of charge $+1$ and $-1$ respectively under the

420: unbroken $U(1)$ gauge symmetry. Canonical quantization for

421: fermions gives the brackets $\{ \Psi_\alpha,\Psi_{\beta}^\dagger

422: \} = \delta_{\alpha\beta}$ and

423: $\{\tilde{\Psi}_\alpha,\tilde{\Psi}_{\beta}^\dagger \} =

424: \delta_{\alpha\beta}$. We build the fermionic Hilbert space ${\cal

425: H}_F$ by picking a reference state $\vac$, satisfying

426: %

427: \be \Psi_\alpha\vac=\tilde{\Psi}_\alpha\vac=0\label{ref}\ee

428: %

429: We may then act  upon $\vac$ with $\Psi^\dagger$ in the $N=2,16$

430: theories, and $\Psi^\dagger$ and $\tilde{\Psi}^\dagger$ in the

431: $N=4,8$ theories, to construct a Hilbert space of dimension

432: $\dim({\cal H}_F)=2^N$.

433: 

434: \para

435: The free fermionic Hamiltonians \eqn{hf216} and \eqn{hf48} have a

436: unique ground state $|\Omega\rangle$ with energy

437: $E_F=-N\hat{X}/4g^2$. This ensures that the ground state energy of

438: the full theory is $E_0=E_B+E_F=0$. The fermionic ground state

439: always lies in the sector with half of the fermions excited. To

440: describe it, we first introduce the fermionic projection operators

441: %

442: \be P_\pm = \frac{1}{2}\left( 1 \pm \frac{\hat X\cdot

443: \Gamma}{|\hat X|}\right)\label{fproj}\ee

444: %

445: Then, for the $N=2,16$ theories, the (un-normalized) ground state

446: is given schematically by,

447: %

448: \be\label{eq:ground} |\Omega\rangle = \left(P_-

449: \Psi^\dagger\right)^{N/2} \vac \label{vac1}\ee

450: %

451: while, for the $N=4,8$ theories, the (un-normalized) ground state

452: is given by

453: %

454: \be |\Omega\rangle = \left(P_- \Psi^\dagger\right)^{N/4} \left(P_+

455: \tilde{\Psi}^\dagger\right)^{N/4}   \vac \label{vac2}\ee

456: %

457: Note that $|\Omega\rangle$ is the ground state for the high

458: frequency modes in the Born-Oppenheimer approximation. The

459: question of whether a normalizable ground state of the full theory

460: exists is more subtle, and irrelevant for our considerations. (It

461: does for $N=16$, but is expected not to for $N=2,4,8$).

462: 

463: \para

464: The physical Hilbert space of the theory is subject to Gauss' law

465: for the unbroken $U(1)\subset SU(2)$ gauge symmetry which ensures

466: that all physical states are gauge neutral. The vacuum

467: $|\Omega\rangle$ is assigned charge zero under the $U(1)$ and so

468: survives Gauss' purge. Other states in the fermionic Hilbert space

469: arise by acting on $|\Omega\rangle$ with combinations of the

470: charge $+1$ states $P_-\Psi$ and $P_-\tilde{\Psi}^\dagger$ and the

471: charge $-1$ states $P_+\Psi^\dagger$ and $P_+\tilde{\Psi}$. In

472: each case, one may construct a state in the physical Hilbert space

473: by dressing the fermionic state with appropriate powers of the

474: charged bosonic operator ${\cal P}_{ij}a_j^\dagger$ and ${\cal

475: P}_{ij}\bar{a}_j^\dagger$.

476: 

477: 

478: 

479: \subsection*{3. Berry Phase and Hopf Maps}

480: 

481: In the previous section we constructed the Hilbert space over each

482: background separation $\hat{X}^i$. In this section we are

483: interested in how these Hilbert spaces evolve as the D0-branes

484: orbit, and $\hat{X}^i$ traces a closed path $\Gamma$ in

485: configuration space. The dynamical phase of a set of degenerate

486: states $|\phi_a\rangle$ is accompanied by a (possibly non-Abelian)

487: Berry holonomy, described by

488: %

489: \be |\phi_a\rangle \longrightarrow P \exp\left(-i \oint_\Gamma

490: {A}_{ab}\cdot \hat{X} \right)|\phi_b\rangle \ \ \ {\rm with}\ \ \

491: (A_i)_{ab}=i\langle\phi_b|\frac{\partial}{\partial\hat{X}^i}|\phi_a\rangle\ee

492: %

493: We will see that for certain states in quantum mechanics with $N$ supercharges, the

494: Berry connection $A$ is given by the associated Hopf map. We start

495: by reviewing the the Hopf maps and connections in more detail.

496: 

497: \para

498: The Hopf map from $S^{N-1}\rightarrow S^{N/2}$ is defined in the

499: following manner. One starts with a commuting spinor

500: $\chi_\alpha$, of the type described in Section 2: real for

501: $N=2,16$ and complex for $N=4,8$. Imposing the normalization

502: condition $\chi^\dagger \chi =1$ ensures that $\chi$ defines a

503: point on $S^{N-1}$. The map to the base manifold $S^{N/2}$ is then

504: given by the bi-linear form

505: %

506: %

507: \be n^i = \chi^\dagger \Gamma^i \chi \ee

508: %

509: where $\Gamma^i$ obey the $SO(d-1)$ Clifford algebra. One may

510: check that $n^in^i=1$ and hence $n^i$ defines a point on

511: $S^{N/2}$.

512: 

513: 

514: \para

515: It is illustrative to stress the connection to the division

516: algebras by reformulating the Hopf maps over the algebra $\bK$

517: using the isomorphism \eqn{the}. (For more details see, for example,

518: \cite{baez}). The total manifold $S^{N-1}$ can be thought of as

519: arising from the pair $(q_1,q_2)\in \bK^2$, subject to the

520: constraint $|q_1|^2 + |q_2|^2=1$. From this pair we can define the

521: vector $n^i$, $i=1\ldots,d-1$,

522: %

523: \be n^i = q^\dagger \Gamma^i q\ee

524: %

525: where $\Gamma^i$ still satisfy the $SO(d-1)$ Clifford algebra,

526: hence justifying their name, but can now be thought of as basis

527: elements of $\Gamma^i \in sl(2;\bK)$, defined by

528: %

529: \be \Gamma^i=\left(\begin{array}{cc} 0 & e_i \\ e_i^\star & 0

530: \end{array}\right)\ \  \ i=1,\ldots , N/2\ \ \ ,\ \ \

531: \Gamma^{d-1}=\left(\begin{array}{cc} 1 & 0 \\ 0 & -1

532: \end{array}\right)\label{gam}\ee

533: %

534: where $e_i$ are the generators of the division algebra $\bK$. The

535: fact that these matrices obey the Clifford algebra reflects  the

536: isomorphism \eqn{the}. One can check once again that $n^in^i=1$,

537: ensuring that this defines a map to the base manifold

538: $S^{N/2}\cong \bK\bP^1$.

539: 

540: \para

541: The zeroth Hopf map associated to $\bR$ is the M\"obius bundle

542: ${\bf Z}_2\hookrightarrow S^1\rightarrow

543: \bR\bP^1$. We will shortly see how this

544: arises in the theory with $N=2$ supercharges: the holonomy of the

545: ground state is a minus sign whose role is to render the D0-branes

546: anyonic. The remaining Hopf maps, associated to $\bC$, $\bH$ and

547: $\bO$, each define a connection $A$ over the base space $S^{N/2}$,

548: which describes how the base manifold is twisted inside the total

549: space. These connections are given in terms of the projection

550: operators $P_\pm$ defined in \eqn{fproj}, where we identify

551: $\hat{X}^i\equiv \hat{X}n^i$. Let $\lambda_a$ be the non-vanishing

552: orthonormal eigenvectors of $P_-$: i.e. $P_-\lambda_a=\lambda_a$.

553: The Hopf connection is defined by,

554: %

555: \be A_{ab} = i\lambda_b^\dagger\, d\lambda_a\label{hopfcon}\ee

556: %

557: For $N=4,8$ and $16$, the maximum value of the index $a$ is

558: $1,2,8$ and $A_{ab}$ is therefore a $U(1)$, $U(2)$ and $SO(8)$

559: connection respectively. In the following, we will describe these

560: connections in more detail and see how they arise as the Berry

561: phase of certain states in the quantum mechanics.

562: 

563: 

564: 

565: \subsubsection*{3.1 $N=2$ and the Zeroth Hopf Map}

566: 

567: For the $N=2$ case, the ground state wavefunction undergoes

568: a discrete holonomy as the D0-branes orbit.  While the existence

569: of this phase follows on general grounds from the degenerate

570: nature of $H_F$ at the origin, it is instructive to review

571: explicitly how it occurs.

572: 

573: \para

574: The fermionic Hilbert space ${\cal H}_F$ consists of four states

575: $\vac$, $\Psi^\dagger_\alpha\vac$ and

576: $\Psi^\dagger_1\Psi^\dagger_2\vac$. The ground state may be

577: expanded as a linear combination of the middle sector,

578: %

579: \be |\Omega\rangle=\lambda_\alpha \Psi^\dagger_\alpha\vac\ee

580: %

581: with $\lambda_\alpha$ the negative eigenvector of $\hat{X}\cdot

582: \Gamma$.

583: %They have energy

584: %%

585: %\be H_F\vac=H_F\Psi^\dagger_1\Psi^\dagger_2\vac=0\ \ \ {\rm and}\

586: %\ \ H_F\Psi_\alpha^\dagger\vac =

587: %\frac{1}{2g^2}(\hat{X}\cdot\Gamma)_{\alpha\beta}\Psi^\dagger_\beta\vac\ee

588: %%

589: We choose to work with the real basis of gamma matrices

590: $\Gamma^1=\sigma^1$ and $\Gamma^2=\sigma^3$ and introduce polar

591: coordinates

592: %

593: \be X^1=X\sin\theta\ \ \ ,\ \ \ X^2=X\cos\theta\ee

594: %

595: Then the groundstate eigenvector is

596: %

597: \be \vec{\lambda} =

598: \frac{1}{\sqrt{2+2\cos\theta}}\left(\begin{array}{c}-\sin\theta \\

599: 1+ \cos\theta \end{array}\right) \label{turn}\ee

600: %

601: where we have resolved the square-root sign ambiguity in favour of

602: the positive. The evolution of the eigenvector as $\theta$ is

603: adiabatically varied is shown in the figure. We see that as

604: $\theta$ varies from 0 to $2\pi$, $\vec{\lambda}$ returns to

605: $-\vec{\lambda}$. This is the manifestation of the zeroth Hopf

606: map: $S^1\rightarrow S^1$.

607: 

608: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

609: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

610: \newcommand{\onefigurenocap}[1]{\begin{figure}[h]

611:          \begin{center}\leavevmode\epsfbox{#1.eps}\end{center}

612:          \end{figure}}

613: \newcommand{\onefigure}[2]{\begin{figure}[htbp]

614:          \begin{center}\leavevmode\epsfbox{#1.eps}\end{center}

615:          \caption{\small #2\label{#1}}

616:          \end{figure}}

617: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

618: \begin{figure}[htbp]

619: \begin{center}

620: \epsfxsize=2in\leavevmode\epsfbox{z2berry.eps}

621: \end{center}

622: {\small Figure 1: The ${\bf Z}_2$ holonomy as the eigenvector

623: encircles the degeneracy at the origin.}

624: %\label{mon}

625: \end{figure}

626: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

627: 

628: \para

629: This ${\bf Z}_2$ Berry phase is inherited by the ground state of

630: the D0-branes. After a single orbit, the wavefunction returns to

631: $|\,\Omega\rangle \mapsto -|\,\Omega\rangle$. Yet such an orbit is

632: equivalent to two exchanges of the D0-branes. Since the branes are

633: indistinguishable particles, upon a single exchange,

634: the D0-brane wavefunction picks up the phase $e^{\pm i\pi/2}=\pm

635: i$, where the $\pm$ sign depends on whether the exchange proceeds

636: clockwise or anti-clockwise. We learn that the D0-branes are

637: anyons \cite{anyon1,anyon2}.

638: 

639: \para

640: It is worthwhile re-deriving the fact that the D0-branes are

641: anyons by studying a single exchange, rather than a complete

642: orbit.  Let us start with the D0-branes separated by

643: $\hat{X}^i=(0,\hat{X})$, for which the ground state is given by

644: $|\Omega\rangle_+ = \Psi_2^\dagger\vac$. After an anti-clockwise

645: rotation to $\hat{X}^i=(0,-\hat{X})$, we see from \eqn{turn} that

646: the ground state adiabatically evolves into $|\Omega\rangle_- =

647: -\Psi_1^\dagger|0\rangle$. The ${\bf Z}_2$ Weyl symmetry ensures

648: that the Hilbert space constructed over $\hat{X}^i$ is physically

649: identified with the Hilbert space over $-\hat{X}^i$. Thus in order

650: to understand the phase picked up by the ground state, we need to

651: understand the map between the Hilbert spaces constructed over

652: $\pm\hat{X}^i$ induced by the Weyl group.

653: 

654: \para

655: The action of the ${\bf Z}_2$ Weyl group is given by,

656: %

657: \be &{\bf Z}_2:& \hat{X}^i \mapsto -\hat{X}^i\ \ \ ;\ \ \

658: z^i\mapsto -z^{\dagger i}\ \ \ ;\ \ \ \Psi\mapsto -\Psi^\dagger

659: \ee

660: %

661: From the latter of these actions, we see that the reference state

662: $\vac_+$ defined by \eqn{ref} at $\hat{X}^i$ is mapped to

663: $\Psi_1^\dagger \Psi_2^\dagger \vac_-$ at $-\hat{X}^i$, up to a

664: phase $\omega$,

665: %

666: \be &{\bf Z}_2:& \vac_+ \mapsto \omega \Psi_1^\dagger

667: \Psi_2^\dagger \vac_-\ee

668: %

669: The question is: what is $\omega$? We can answer this by following the

670: fate of the state $\Psi_1^\dagger\vac_+$ under two actions of the

671: Weyl group,

672: %

673: \be \Psi_1^\dagger\vac_+\mapsto -\omega\Psi_2^\dagger\vac_-\mapsto

674: -\omega^2 \Psi_1^\dagger\vac_+\ee

675: %

676: Insisting that the Weyl group action squares to the unit operator,

677: we learn that $\omega=\pm i$. We now use this knowledge to map the

678: state $|\Omega\rangle_-$ above back to an element of the Hilbert

679: space over $\hat{X}^i=(0,\hat{X})$. We see,

680: %

681: \be |\Omega\rangle_+ \stackrel{\tiny{\rm

682: adiabatic}}{\longrightarrow} |\Omega\rangle_- \stackrel{\tiny{\rm

683: {\bf Z}_2}}{\longrightarrow}\ \mp i|\Omega\rangle_+\ee

684: %

685: The upshot is that the ground state of the two D0-branes indeed

686: picks up the phase factor $\pm i$ upon exchange.

687: 

688: \para

689: Note that we have still to quantize the $U(1)\subset SU(2)$ Cartan

690: subalgebra. In particular, this includes the massless fermions

691: $(\psi_3)_\alpha$, which double the Hilbert space. The resulting

692: states have conjugate exchange statistics. If one state picks up a

693: phase $+i$ upon an anti-clockwise rotation, the other state picks

694: up a phase $-i$.

695: 

696: \subsubsection*{3.2 $N=4$ and the First Hopf Map}

697: 

698: We now turn to D0-branes moving in $d-1=3$ spatial dimensions.

699: As explained in Section 2, the Hilbert space is constructed by

700: acting with bosonic and fermionic creation operators on the

701: reference state $\vac$. Each of these creation operators is

702: accompanied by a projection operator: ${\cal P}$ for bosons,

703: defined in \eqn{bproj}, and $P_\pm$ for fermions, defined in

704: \eqn{fproj}. These two types of projection operators give rise to

705: two types of Berry holonomy:

706: 

707: \para

708: The projection operators ${\cal P}$ restrict the bosonic

709: excitations to lie tangent to the sphere $S^2$ at fixed

710: $|\hat{X}|$. As we saw above, this can be traced to the

711: implementation of the $SU(2)$ gauge symmetry and, from the string

712: theory perspective, is the familiar statement that stretched

713: strings have only transverse excitations. The net result is that as the

714: D0-branes orbit, excited states that include bosonic excitations

715: undergo a Berry holonomy as tangent vectors on the sphere $S^2$.

716: This is the same kind of

717: non-Abelian holonomy that was described in the original

718: paper of Wilczek and Zee \cite{wz}.

719: 

720: \para

721: Fermionic excitations are accompanied by the projection operators

722: $P_\pm$ \eqn{fproj}. To be concrete, let's focus on the operator

723: $P_-$. We introduce the normalized eigenvector

724: $P_-\lambda=\lambda$. Acting with the creation operator

725: $P_-\Psi^\dagger$ means creating the normalized state

726: $|\lambda\rangle =\lambda_\alpha\Psi_\alpha^\dagger\vac$. As

727: $\hat{X}^i$ varies adiabatically, this state undergoes a holonomy

728: described by the Berry connection,

729: %

730: \be A_i = i \langle \lambda | \frac{\partial}{\partial

731: \hat{X}^i}|\lambda\rangle\ee

732: %

733: which is the connection of the Hopf map \eqn{hopfcon}. An explicit

734: form of the connection requires a choice of gauge which, in this

735: context, means a chosen phase for the eigenvector $\lambda$. We

736: choose $\lambda_1\in \bR$, a choice which is valid everywhere

737: except along the positive $\hat{X}^3$ axis. In this gauge, the

738: Berry connection takes the familiar form of the Dirac monopole,

739: %

740: \be A_i =

741: \frac{-\hat{X}^j}{2\hat{X}(\hat{X}-\hat{X}^3)}\,\epsilon_{ij}

742: \qquad i = 1,2 \ \ \ \ ,\ \ \ \ A_3 =0 \label{dirac}\ee

743: %

744: The Hopf-Berry connection has first Chern class $-1$ over the

745: sphere $S^2$, meaning that the integral of the field strength

746: $F=dA$ yields,

747: %

748: \be c_1=\frac{1}{2\pi}\int_{S^2} F = -1\ee

749: %

750: Acting with the operator $P_+\Psi^\dagger$ results is a state

751: whose Berry connection has Chern class $+1$. The ground state in

752: the $N=4$ theory is given by

753: %

754: \be |\Omega\rangle = N^\prime(P_-\

755: \Psi^\dagger)(P_+\tilde{\Psi}^\dagger)\vac\equiv (\lambda_\alpha

756: \Psi_\alpha^\dagger)(\tilde{\lambda}_\beta\tilde{\Psi}_\beta^\dagger)\vac\ee

757: %

758: where $N^\prime$ is a normalization factor, while $\lambda$

759: ($\tilde{\lambda}$) is the normalized non-zero eigenvector of

760: $P_\mp$. The presence of the two, opposite, projection operators

761: ensures that the ground state does not pick up a Berry phase as

762: the D0-branes orbit.

763: 

764: \para

765: Excited states do pick up a Berry phase, given by the sum of the

766: phases associated to the relevant projection operators. For

767: example, a degenerate pair of states obeying Gauss' law

768: is given

769: by

770: %

771: \be |\phi_-\rangle = (\lambda_\alpha\Psi_\alpha^\dagger)

772: (\lambda_\alpha\tilde{\Psi}_\alpha^\dagger)\vac \ \ \ ,\ \ \

773: |\phi_+\rangle = (\tilde{\lambda}_\alpha\Psi_\alpha^\dagger)

774: (\tilde{\lambda}_\alpha\tilde{\Psi}_\alpha^\dagger)\vac \ee

775: %

776: These states have energy $E=E_B+E_F=\hat{X}/g^2$ and describe two

777: strings attached to the D0-branes. (They are part of a triplet of

778: excited spin 1 states).

779: %

780: %transform with charge with charge $\pm 2$ under the unbroken

781: %$U(1)_R$ R-symmetry of the problem, reflecting the fact that they

782: %come from a spin 1 state in three dimensions. (The third state in

783: %the triplet, $|\phi_0\rangle =

784: %(P_+\Psi^\dagger)(P_-\tilde{\Psi}^\dagger)\vac$, has had it's

785: %energy split by the separation of the D0-branes).

786: %

787: From the discussion above, we see that these states pick up a

788: Berry phase arising from a magnetic monopole of charge $q=\pm2$.

789: Physically, this means that the D0-branes orbit as charged

790: particles as if in the presence of a magnetic monopole fixed at

791: their centre. In a semi-classical analysis, the orbits are no

792: longer restricted to lie on a plane, but rather lie on a cone with

793: opening angle $\cos\theta = -q/(2J+q)$, where $J$ is angular

794: momentum of the spinning D0-branes. The energy $E$ these rotating

795: states scales as $E^3\sim g^2 J(J+q)$.

796: 

797: 

798: 

799: \subsubsection*{3.3 $N=8$ and the Quaternionic Hopf Map}

800: 

801: The discussion for the $N=8$ theory is very similar to the $N=4$

802: theory above. The unique ground state of the system \eqn{vac2}

803: does not undergo a Berry phase.

804: %

805: %\be |\Omega\rangle \sim

806: %(P_-\Psi^\dagger)^2(P_+\tilde{\Psi}^\dagger)^2

807: %%(\lambda_{1\alpha}\Psi_\alpha^\dagger)(\lambda_{2\beta}\Psi_\beta^\dagger)

808: %%(\tilde{\lambda}_{1\gamma}\Psi_\gamma^\dagger)

809: %%(\tilde{\lambda}_{2\delta}\Psi_\delta^\dagger)

810: %\vac\ee

811: %%

812: %This state does not undergo a Berry phase. Further, excited states

813: %do undergo a Berry holonomy.

814: %

815: Excited states do. The typical state undergoes a non-Abelian

816: holonomy arising from the sum of bosonic and fermionic Berry

817: connections. Once again, bosonic excitations give rise to a

818: holonomy in which states transform as tangent vectors on $S^5$.

819: More interesting for the present discussion are the fermionic

820: excitations. The projection operator $P_-$ now has a pair of

821: orthonormal eigenvectors: $P_-\lambda_a=\lambda_a$, $a=1,2$. This

822: means that the resulting Berry holonomy of the states

823: $|\lambda_a\rangle= \lambda_{a\alpha}\Psi^\dagger_\alpha\vac$ is

824: described by a $U(2)$ connection,

825: %

826: \be (A_i)_{ab} = i\langle \lambda_a |\frac{\partial}{\partial

827: \hat{X}^i}|\lambda_b\rangle \ee

828: %

829: Explicit computation shows that this is actually an $SU(2)$

830: connection, known as the Yang-monopole \cite{yang}. A suitable

831: choice of gamma matrices and gauge, can be found in \cite{second}.

832: The resulting connection is,

833: %

834: \be (A_i)_{ab} =

835: \frac{-\hat{X}^j}{2\hat{X}(\hat{X}-\hat{X}^5)}\,\eta^m_{ij}\,\sigma^{m}_{ab}

836: \qquad i = 1,2,3,4 \ \ \ \ ,\ \ \ \ A_5 =0 \label{yang}\ee

837: %

838: where $\eta_{\mu\nu}^m$ are the self-dual $4\times 4$ 't Hooft

839: matrices and $\sigma_{ab}$ are the Pauli matrices. The Yang

840: monopole is perhaps more familiar when viewed as a connection

841: restricted to $S^4$, where it is simply the $SO(5)$ invariant

842: instanton satisfying $F=-{}^\star F$ with second Chern class,

843: %

844: \be c_2=\frac{1}{8\pi^2}\int_{S^4}\,{\rm tr}(F\wedge F)=-1 \ee

845: %

846: where the non-Abelian field strength is defined by

847: $F_{ij}=\partial_iA_j-\partial_jA_i-i[A_i,A_j]$. States

848: constructed from the projection operators $P_+$ undergo a Berry

849: holonomy associated with a Yang-monopole of Chern class $c_2=+1$.

850: 

851: %\para

852: %Again, we can focus on the fermionic Berry connection by looking

853: %at the two pairs of states,

854: %%

855: %\be |\phi\rangle = N(P_+\Psi^\dagger)(P_-\Psi)|\Omega\rangle\ \ \

856: %,\ \ \ |\tilde{\phi}\rangle = \tilde{N}

857: %(P_-\tilde{\Psi}^\dagger)(P_+\tilde{\Psi})|\Omega\rangle\ee

858: %%

859: %Each of these states removes one fermion from the ground state

860: %with lower energy, and replaces it with an excited state. We have

861: %neglected indices, but these describe two pairs of states, all

862: %with energy $E=2\hat{X}$. The pair of states $|\phi\rangle$

863: %undergo holonomy from the Yang monopole; the states

864: %$|\tilde{\phi}\rangle$ undergo holonomy from the Yin monopole.

865: %Yang-monopoles. {\bf Question: Is this right? This seems to be the

866: %direct sum of connections that we had in the previous paper,

867: %rather than the direct product that we might have expected}.

868: 

869: 

870: 

871: 

872: \subsubsection*{3.4 $N=16$ and the Last Hopf Map}

873: 

874: Details for the theory with $N=16$ supercharges are again similar

875: to those above: the vacuum \eqn{vac1} does not pick up a Berry

876: connection, while bosonic excitations transform as tangent vectors

877: on $S^8$. The fermionic excitations are associated to projection

878: operators $P_\pm$, which are now $16\times 16$ real matrices. We

879: once again define the $8$ orthonormal, real eigenvectors

880: $P_-\lambda_a=\lambda_a$, with $a=1,\ldots,8$. States in the

881: quantum mechanics involving fermions undergo a holonomy arising

882: from the $SO(8)$ Berry connection,

883: %

884: \be (A_i)_{ab} = i\langle \lambda_b |\frac{\partial}{\partial

885: \hat{X}^i}|\lambda_a\rangle\ee

886: %

887: with $|\lambda_a\rangle =

888: \lambda_{a\alpha}\Psi_\alpha^\dagger\vac$. We refer to the

889: resulting connection as the $SO(8)$ octonionic monopole. It was

890: constructed in \cite{grossman,tchrak}. A simple expression for the

891: connection can be found in \cite{bern}

892: %

893: \be (A_i)_{ab} =

894: \frac{-\hat{X}^j}{2\hat{X}(\hat{X}-\hat{X}^9)}\,\Sigma_{ij} \qquad

895: i = 1,\ldots,8 \ \ \ \ ,\ \ \ \ A_9 =0 \label{so8}\ee

896: %

897: where $\Sigma_{ij}$ are the 28 generators of $SO(8)$ Lie algebra,

898: defined in terms of the $\Gamma$ matrices. An explicit form can be

899: given if we choose a suitable representation for the generators of

900: the octonions $e_i$, $i=1\ldots 8$ in \eqn{gam}, in terms of

901: $8\times 8$ real matrices. We choose the unit matrix to correspond

902: to $e_8=1$. Then $\Sigma_{ij} = -\ft12 [e_i,e_j]$, for

903: $i=1,\ldots,7$ and $\Sigma_{i8}=e_i$ \cite{bern}. The non-Abelian

904: field strength restricted to $S^8$ satisfies the generalized

905: self-duality condition  $F\wedge F = -{}^\star F\wedge F$, with

906: %

907: \be \frac{1}{4!(2\pi)^4}\int_{S^8}{\rm tr}(F\wedge F\wedge

908: F\wedge F) = -1\ee

909: %

910: %We can now see how this connection acts on the states of the

911: %theory. Once again, the ground state $|\Omega\rangle =

912: %(P_-\Psi^\dagger)^8\vac$ has no Berry phase. The simplest states

913: %which do pick up a Berry connection are the $64$ states with

914: %energy $E=2\hat{X}$, with purely fermionic excitations

915: %%

916: %\be |\phi\rangle = N(P_+\Psi^\dagger)(P_-\Psi)|\Omega\rangle\ee

917: %%

918: %These excited states, corresponding to two strings stretched

919: %between the D0-branes, undergo a non-Abelian holonomy given by the

920: %tensor product of the octonionic instanton and anti-instanton.

921: %

922: %

923: As in previous cases, states associated to $P_+$ undergo a Berry

924: holonomy with $c_4=1$.

925: 

926: \para

927: It is intriguing to see the octonionic monopole appearing in the

928: Matrix theory for D0-branes in this fashion, although its physical

929: significance in M-theory remains obscure. Nonetheless, it is

930: tempting to speculate. Firstly, recall that there is a precedent

931: for the appearance of Berry's phase in Matrix theory: the membrane

932: feels the magnetic field of the five-brane through the appearance

933: of a Berry connection \cite{dberk}. In this case, the abstract

934: Berry connection is recast as a physical magnetic field. It would

935: be very interesting if a similar M-theory interpretation could be

936: given in the present case. We note in passing that a physical

937: octonionic monopole is conjectured to act as the end-point for an

938: open heterotic string \cite{pol}.

939: 

940: \subsection*{Acknowledgments}

941: 

942: We would like to thank Freddy Cachazo, Joe Polchinski, Savdeep

943: Sethi and Paul Townsend for useful discussions. J.S. gratefully

944: acknowledges the University of Chicago, Harvard and the Perimeter

945: Institute for hospitality and travel support while this paper was

946: being written up. D.T. thanks the Isaac Newton Institute for their

947: kind hospitality while this work was undertaken. C.P. is supported

948: by an EPSRC studentship. J.S. is supported by a research

949: fellowship from Trinity College, Cambridge. D.T. is supported by

950: the Royal Society.

951: 

952: 

953: 

954: \begin{thebibliography}{99}

955: 

956: \small

957: \parskip=0pt plus 2pt

958: 

959: %\cite{Kugo:1982bn}

960: \bibitem{kugo}

961:   T.~Kugo and P.~K.~Townsend,

962:   {\it Supersymmetry And The Division Algebras,}

963:   Nucl.\ Phys.\  B {\bf 221}, 357 (1983).

964:   %%CITATION = NUPHA,B221,357;%%

965: 

966: \bibitem{jon}   J.~M.~Evans, ``{\it Supersymmetric Yang-Mills Theories 

967: and Division Algebras}'',   Nucl.\ Phys.\  B {\bf 298}, 92 (1988).

968:   %%CITATION = NUPHA,B298,92;%%

969: 

970: \bibitem{yang} C.~N.~Yang,

971:   ``{\it Generalization Of Dirac's Monopole To SU(2) Gauge Fields},''

972:   J.\ Math.\ Phys.\  {\bf 19}, 320 (1978).

973:   %%CITATION = JMAPA,19,320;%%

974: %\cite{Demler:1998pm}

975: 

976: 

977: \bibitem{grossman}

978:   B.~Grossman, T.~W.~Kephart and J.~D.~Stasheff,

979:   ``{\it Solutions To Yang-Mills Field Equations In Eight-Dimensions And The Last

980:   Hopf Map},''

981:   Commun.\ Math.\ Phys.\  {\bf 96}, 431 (1984)

982:   [Erratum-ibid.\  {\bf 100}, 311 (1985)].

983:   %%CITATION = CMPHA,96,431;%%

984: 

985: \bibitem{tchrak}  D.~H.~Tchrakian,

986:   ``{\it Spherically Symmetric Gauge Field Configurations With Finite Action In 4

987:   P-Dimensions (P = Integer)},''

988:   Phys.\ Lett.\  B {\bf 150}, 360 (1985).

989:   %%CITATION = PHLTA,B150,360;%%

990: 

991: \bibitem{chemist} G. Herzberg and H. C. Longuet-Higgins,

992: ``{\it Intersection of potential energy surfaces in polyatomic

993: molecules}", Discuss. Faraday Soc., 1963, 35, 77

994: 

995: \bibitem{berry} M.~V.~Berry,

996:   ``{\it Quantal phase factors accompanying adiabatic changes},''

997:   Proc.\ Roy.\ Soc.\ Lond.\  A {\bf 392}, 45 (1984).

998:   %%CITATION = PRSLA,A392,45;%%

999: 

1000: \bibitem{avron}   J.~E.~Avron, L.~Sadun, J.~Segert and B.~Simon,

1001:   ``{\it Topological Invariants in Fermi Systems with Time-Reversal

1002:   Invariance}", Phys.\ Rev.\ Lett.\  {\bf 61}, 1329 (1988); ``{\it

1003:   Chern Numbers, Quaternions and Berry's Phases in Fermi

1004:   Systems}", Commun. Math. Phys. {\bf 124} (1989) 595.

1005:   %%CITATION = PRLTA,61,1329;%%

1006: 

1007: \bibitem{feher}   M.~G.~Benedict, L.~G.~Feher and Z.~Horvath,

1008:   ``{\it Monopoles and Instantons from Berry's Phase},''

1009:   J.\ Math.\ Phys.\  {\bf 30}, 1727 (1989).

1010:   %%CITATION = JMAPA,30,1727;%%

1011: 

1012: \bibitem{bigzhang}

1013:   E.~Demler and S-C.~Zhang,

1014: ``{\it Non-Abelian holonomy of BCS and SDW quasiparticles,}"

1015:   Annals Phys.\  {\bf 271} (1999) 83

1016:   [arXiv:cond-mat/9805404].

1017:   %%CITATION = APNYA,271,83;%%

1018: 

1019: \bibitem{middlezhang}  S. Murakami, N. Nagaosa, S-C. Zhang,

1020: ``{\it SU(2) Non-Abelian Holonomy and Dissipationless Spin Current

1021: in Semiconductors}", Phys. Rev. {\bf B69}, 235206 (2004),

1022: [arXiv:cond-mat/0310005v3]

1023: 

1024: \bibitem{smallzhang} C-H. Chern, H-D. Chen, C. Wu, J-P. Hu, S-C.

1025: Zhang, ``{\it Non-abelian Berry's phase and Chern numbers in

1026: higher spin pairing condensates}", Phys. Rev. {\bf B69}, 214512

1027: (2004) [arXiv:cond-mat/0310089v2].

1028: 

1029: 

1030: %\cite{Levay:1990}

1031: \bibitem{levay}

1032:   P. L\'evay,

1033:   {\it Geometrical description of SU(2) Berry phases,}

1034:   Phys.\ Rev.\ A \ {\bf 41} {1990} 2837;

1035:   ``{\it Quaternionic Gauge Fields And The Geometric Phase},''

1036:   J.\ Math.\ Phys.\  {\bf 32}, 2347 (1991);

1037:   ``{\it NonAbelian

1038:   Born-Oppenheimer electric gauge force and the natural metric on

1039:   Hilbert subspaces},''

1040:   Phys.\ Rev.\  A {\bf 45}, 1339 (1992);

1041:   %%CITATION = PHRVA,A45,1339;%%

1042:   %%CITATION = JMAPA,32,2347;%%

1043: 

1044: \bibitem{bern}

1045:  B.~A.~Bernevig, J.~p.~Hu, N.~Toumbas and S.~C.~Zhang,

1046:   ``{\it The Eight Dimensional Quantum Hall Effect and the Octonions},''

1047:   Phys.\ Rev.\ Lett.\  {\bf 91}, 236803 (2003)

1048:   [arXiv:cond-mat/0306045].

1049:   %%CITATION = PRLTA,91,236803;%%

1050: 

1051: \bibitem{us} C.~Pedder, J.~Sonner and D.~Tong,

1052:   ``{\it The Geometric Phase in Supersymmetric Quantum Mechanics},''

1053:   arXiv:0709.0731 [hep-th].

1054:   %%CITATION = ARXIV:0709.0731;%%

1055: 

1056: %\cite{Pedder:2007wp}

1057: \bibitem{second}

1058:   C.~Pedder, J.~Sonner and D.~Tong,

1059:   ``{\it The Geometric Phase and Gravitational Precession of D-Branes},''

1060:   arXiv:0709.2136 [hep-th].

1061:   %%CITATION = ARXIV:0709.2136;%%

1062: 

1063: 

1064: 

1065: \bibitem{usp}   B.~Chen, H.~Itoyama and H.~Kihara,

1066:   ``{\it Nonabelian Berry phase, Yang-Mills instanton and USp(2k) matrix model},''

1067:   Mod.\ Phys.\ Lett.\  A {\bf 14}, 869 (1999)

1068:   [arXiv:hep-th/9810237];

1069:    ``{\it Nonabelian monopoles from matrices:

1070:   Seeds of the spacetime structure},''

1071:   Nucl.\ Phys.\  B {\bf 577}, 23 (2000)

1072:   [arXiv:hep-th/9909075].

1073:   %%CITATION = NUPHA,B577,23;%%

1074: 

1075: 

1076: 

1077: 

1078: \bibitem{baez}

1079:   J.~C.~Baez,

1080:   ``{\it The Octonions},''

1081:   arXiv:math/0105155.

1082:   %%CITATION = MATH/0105155;%

1083: 

1084: \bibitem{anyon1}  J.~M.~Leinaas and J.~Myrheim,

1085:   ``{\it On the theory of identical particles},''

1086:   Nuovo Cim.\  B {\bf 37} (1977) 1.

1087:   %%CITATION = NUCIA,37B,1;%%

1088: 

1089: 

1090: \bibitem{anyon2}  F.~Wilczek,

1091:   ``{\it Quantum Mechanics Of Fractional Spin Particles},''

1092:   Phys.\ Rev.\ Lett.\  {\bf 49}, 957 (1982).

1093:   %%CITATION = PRLTA,49,957;%%

1094: 

1095: 

1096: \bibitem{wz}  F.~Wilczek and A.~Zee,

1097:   ``{\it Appearance Of Gauge Structure In Simple Dynamical Systems},''

1098:   Phys.\ Rev.\ Lett.\  {\bf 52}, 2111 (1984).

1099:   %%CITATION = PRLTA,52,2111;%%

1100: 

1101: 

1102: \bibitem{dberk}  M.~Berkooz and M.~R.~Douglas,

1103:   ``{\it Five-branes in M(atrix) theory},''

1104:   Phys.\ Lett.\  B {\bf 395}, 196 (1997)

1105:   [arXiv:hep-th/9610236].

1106:   %%CITATION = PHLTA,B395,196;%%

1107: 

1108: \bibitem{pol}  J.~Polchinski,

1109:   ``{\it Open heterotic strings},''

1110:   JHEP {\bf 0609}, 082 (2006)

1111:   [arXiv:hep-th/0510033].

1112:   %%CITATION = JHEPA,0609,082;%%

1113: 

1114: 

1115: 

1116: \end{thebibliography}

1117: 

1118: 

1119: 

1120: 

1121: 

1122: 

1123: 

1124: 

1125: 

1126: \end{document}

1127: 

1128: 

1129: 

1130: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1131: %\newcommand{\onefigure}[2]{\begin{figure}[htbp]

1132: %

1133: %         \caption{\small #2\label{#1}(#1)}

1134: %         \end{figure}}

1135: %\newcommand{\onefigurenocap}[1]{\begin{figure}[h]

1136: %         \begin{center}\leavevmode\epsfbox{#1.eps}\end{center}

1137: %         \end{figure}}

1138: %\renewcommand{\onefigure}[2]{\begin{figure}[htbp]

1139: %         \begin{center}\leavevmode\epsfbox{#1.eps}\end{center}

1140: %         \caption{\small #2\label{#1}}

1141: %         \end{figure}}

1142: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1143: 

1144: %\begin{figure}[htb]

1145: %\begin{center}

1146: %\epsfxsize=4in\leavevmode\epsfbox{mir.eps}

1147: %\end{center}

1148: %\caption{Mirror symmetry at various scales.}

1149: %\label{mir}

1150: %\end{figure}

1151: