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

2: \usepackage{epsf}

3: \usepackage{amsmath}

4: %\usepackage[vbox]{pdfsync}

5: \numberwithin{equation}{section}

6: 

7: %\def\ZZ{\rlx\leavevmode\ifmmode\mathchoice{\hbox{\cmss Z\kern-.4em Z}}

8:     %{\hbox{\cmss Z\kern-.4em Z}}{\lower.9pt\hbox{\cmsss Z\kern-.36em Z}}

9:     %{\lower1.2pt\hbox{\cmsss Z\kern-.36em Z}}\else{\cmss Z\kern-.4em Z}\fi}

10: 

11: \def\slash{\not{\! \! \rm}}

12: \def\id{\ 1 \! \! \! \! 1}

13: \def\Tr{{\rm Tr}}

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

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

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

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

18: \def\simlt{\mathrel{\lower2.5pt\vbox{\lineskip=0pt\baselineskip=0pt

19:              \hbox{$<$}\hbox{$\sim$}}}}

20: \def\simgt{\mathrel{\lower2.5pt\vbox{\lineskip=0pt\baselineskip=0pt

21:              \hbox{$>$}\hbox{$\sim$}}}}

22: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

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

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

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

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

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

28: \def\@eqnlabel{}

29: \def\@vacuum{}

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

31: \def\draft{\oddsidemargin -.5truein

32:          \def\@oddfoot{\sl preliminary draft \hfil

33:          \rm\thepage\hfil\sl\today\quad\militarytime}

34:          \let\@evenfoot\@oddfoot \overfullrule 3pt

35:          \let\label=\draftlabel

36:          \let\marginnote=\draftmarginnote

37: 

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

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

40: \def\preprint{\twocolumn\sloppy\flushbottom\parindent 1em

41:          \leftmargini 2em\leftmarginv .5em\leftmarginvi .5em

42:          \oddsidemargin -.5in    \evensidemargin -.5in

43:          \columnsep 15mm \footheight 0pt

44:          \textwidth 250mmin      \topmargin  -.4in

45:          \headheight 12pt \topskip .4in

46:          \textheight 175mm

47:          \footskip 0pt

48: 

49: \def\@oddhead{\thepage\hfil\addtocounter{page}{1}\thepage}

50:          \let\@evenhead\@oddhead \def\@oddfoot{} \def\@evenfoot{}  }

51: \def\titlepage{\@restonecolfalse\if@twocolumn\@restonecoltrue\onecolumn

52:       \else \newpage \fi \thispagestyle{empty}\c@page\z@

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

54: \def\endtitlepage{\if@restonecol\twocolumn \else  \fi

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

56:          \setcounter{footnote}{0}}  %\c@footnote\z@ }

57: \catcode`@=12

58: \relax

59: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

60: 

61: %\renewcommand{\theequation}{\thesection.\arabic{equation}}

62: 

63: \title{

64: \vspace*{-0.8cm}

65: \begin{flushright}

66: \normalsize{CERN--PH--TH/2004-217\\

67: \texttt{hep-th/0411032}}\\

68: \end{flushright}

69: \vspace{1cm}

70: %

71: \Large\textbf{Splitting Supersymmetry in String Theory}

72: %\vspace*{-0.5cm}

73: %\vspace*{.5cm}

74: \author{\large

75: {\bf I.~Antoniadis$^1$\footnote{On leave from CPHT 

76: (UMR CNRS 7644) Ecole Polytechnique, F-91128 Palaiseau},

77: S.~Dimopoulos$^{2}$}\\ \\

78: \emph{$^1$Department of Physics, CERN - Theory Division}\\

79: \emph{CH--1211 Geneva 23, Switzerland}\\

80: \emph{$^2$Physics Department, Stanford University}\\

81: \emph{Stanford, California 94309, USA}}}

82: \date{}

83: \begin{document}

84: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

85: %\draft

86: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

87: \maketitle

88: \thispagestyle{empty}

89: %\vspace*{.5cm}

90: 

91: \begin{abstract}

92: We point out that type I string theory in the presence of internal magnetic fields 

93: provides a concrete realization of split supersymmetry. To lowest order,

94: gauginos are massless while squarks and sleptons are superheavy. 

95: We build such realistic $U(3) \times U(2) \times U(1)$ models on stacks of 

96: magnetized D9-branes. Though not unified into a simple group, these theories 

97: preserve the successful supersymmetric relation of gauge couplings, 

98: as they start out with equal $SU(3)$ and $SU(2)$ couplings and the correct 

99: initial $\sin^2{\theta_W}$ at the compactification scale of 

100: $M_{\rm GUT}\simeq 2 \times 10^{16}$ GeV, and they have the 

101: minimal low-energy particle content of split supersymmetry. 

102: We also propose a mechanism in which the gauginos and higgsinos

103: are further protected by a discrete R-symmetry against gravitational corrections, 

104: as the gravitino gets an invariant Dirac mass by pairing with a member of a 

105: Kaluza-Klein tower of spin-3/2 particles.

106: In addition to the models proposed here, split supersymmetry offers novel 

107: strategies for realistic model-building. So, TeV-scale string models previously 

108: dismissed because of rapid proton decay, or incorrect $\sin^2{\theta_W}$, or 

109: because there were no unused dimensions into which to dilute the strength 

110: of gravity, can now be reconsidered as candidates for realistic split theories 

111: with string scale near $M_{\rm GUT}$, as long as the gauginos and 

112: higgsinos remain light.

113: 

114: 

115: \end{abstract}

116: %\vspace{5.cm}

117: \date

118: 

119: \newpage

120: 

121: \section{Introduction} \label{introduction}

122: 

123: Some recent developments challenge us to re-examine our preconceived notions of naturalness and our expectations for physics beyond the Standard Model at the LHC. First is the absence of any deviation from the Standard Model suggesting that, if there is new physics at a TeV, it appears to be fine-tuned at the per-cent level and does not comply with our notion of naturalness. Second, and most important, the cosmological constant problem (CCP) presents us with a fine-tuning much more severe than that of the gauge hierarchy problem (GHP). This raises the possibility that the mechanism which solves the CCP may also solve the GHP, and casts some doubts on all the mechanisms proposed so far to address the GHP (technicolor, low-energy supersymmetry (SUSY), low-scale strings, warping, little higgs), since none of them addresses the CCP problem. 

124: 

125: One concrete idea addressing the CCP is Weinberg's anthropic approach~\cite{weinberg} which postulated the existence of an enormous ``landscape'' of vacua, only a small fraction  of which have a vacuum energy small enough to  allow the formation of galaxies, which provide for a natural (and possibly necessary) habitat for observers such as ourselves. This approach has recently gained momentum because of the realization that string theory may have such a vast landscape of vacua ~\cite{bousso}. 

126: Such an environment may  drastically change what is a natural or likely theory. 

127: To see how this may happen, first recall that the standard measure of fine-tuning in SUSY theories is given by $ f_{standard} \sim {m_H^2 \over m_S^2}$, where $M_H \sim 100$ GeV is the Higgs mass and $m_S$ is the SUSY-breaking scale. 

128: Consider now a neighborhood in the landscape where the density of vacua increases with the scale of SUSY breaking proportional to $m_S^{2N}$ \cite{susskind}. Then, assigning equal a priori probability to each vacuum, the proper new measure of fine-tuning, which takes into account the ``entropy'' associated with the density of vacua, is $f_{new} \sim {m_H^2 \over m_S^2} \times  m_S^{2N}$. For $N > 1$, it thus favors large SUSY-breaking scale $m_S$. 

129: 

130: In such neighborhoods of the landscape low-scale SUSY is disfavored and, if we live in such a neighborhood, the simplest possibility is that we will discover the Standard Model (SM), rather than the supersymmetric Standard Model, at the LHC. This would

131: then account for why we have not seen any evidence for low-energy

132: supersymmetry,  at the expense of

133: giving up the two successes of the supersymmetric SM \cite{dg}: gauge

134: coupling unification~\cite{drw} and natural dark-matter (DM)

135: candidate~\cite{dg,gold}. A more interesting  possibility that

136: preserves these successes  is that  approximate chiral symmetries

137: protect the fermions of the supersymmetric SM  down to the TeV

138: scale ~\cite{savnim,noi,tutti,Arkani-Hamed:2004yi}. 

139: Actually, gauge coupling unification based on extrapolation 

140: of low-energy data to high energies is strictly speaking only 

141: an indirect indication of light gauginos and higgsinos, 

142: rather than of the full super-particle spectrum.

143: 

144: In these theories, the sparticle spectrum  is ``split" in two: (1) the scalars (squarks and

145: sleptons) that get a mass at the high-scale of supersymmetry

146: breaking $m_S$, which can be as large as the grand unification (GUT) scale, and (2)

147:  the fermions (gauginos and higgsinos) which are near the

148: electroweak scale and account for both gauge-coupling

149: unification and DM. The only light scalar in this theory is a

150: finely-tuned Higgs. Rather than the boring prediction that the

151: LHC will discover just the Higgs, these theories -- called Split

152: Supersymmetry -- predict gauginos and higgsinos at a TeV, maintain

153: the successes of the supersymmetric SM, and account for the

154: absence of any evidence for physics beyond the Standard Model, so far.

155: 

156: The main objective of this paper is to build models of split supersymmetry 

157: based on string theory.  It is clear that split supersymmetry offers novel, 

158: previously unavailable, strategies for realistic model-building, and some  

159: previously discarded classes of models can now be reconsidered.  

160: For example, classes of models of intersecting branes that were 

161: studied in the context of low-scale 

162: strings~\cite{Arkani-Hamed:1998rs,Antoniadis:1998ig}, and dismissed  

163: because of rapid proton decay or the value of the weak mixing angle 

164: will now be reconsidered, in section 6, and shown to contain good 

165: candidates for realistic theories. Some TeV-scale string models 

166: were also abandoned because of the absence of unused 

167: dimensions into which to dilute the strength of gravity~\cite{Uranga:2003pz}

168: should be reconsidered as candidates for realistic split 

169: theories with string scale near $2 \times10^{16}$ GeV, 

170: as long as the gauginos and higgsinos can remain light 

171: as a result of an approximate chiral symmetry. 

172: 

173: Another objective is  to build theories where the successful unification relation is preserved.  In split SUSY theories this is not a luxury but an essential ingredient, since unification is a fundamental phenomenological motivation for the ``split'' spectrum of these theories. This is a strong theoretical constraint, since it limits us to very economical fundamental theories with few relevant parameters in the gauge sector, small threshold corrections, minimal particle content, equal $SU(2)$ and $SU(3)$ couplings as well as the correct normalization of the weak mixing angle at the GUT scale ($\sin^2\theta_W=3/8$). 

174: 

175: Our paper is organized as follows.

176: In Section~\ref{framework} we discuss the theoretical framework, 

177: which is type I string theory with internally magnetized D9 branes. 

178: We show that the (tree-level) spectrum

179: of the resulting models is the one required by split supersymmetry.

180: In Section~\ref{unification}, we discuss the conditions that guarantee unification of

181: non-abelian gauge couplings and show that they can be naturally satisfied.

182: In Section~\ref{susybr}, we discuss the various mass scales and the

183: supersymmetry breaking in the gravity sector.

184: In Section~\ref{gauginos}, we propose mechanisms to keep gauginos

185: (and higgsinos) light in the presence of gravity.

186: In Section~\ref{models}, we study issues of model building

187: and present an explicit example of the SM embedding in the above framework,

188: with realistic particle spectrum, realizing the unification conditions and

189: predicting the correct weak mixing angle $\sin^2\theta_W=3/8$ at the GUT scale.

190: Finally, in Section~\ref{pheno}, we discuss some phenomenological consequences

191: and in particular constraints from gluino cosmology.

192: 

193: 

194: \section{The framework} \label{framework}

195: 

196: We start with type I string theory, or equivalently type IIB with 

197: orientifold planes and D-branes. Upon compactification 

198: in four dimensions on a Calabi-Yau manifold, 

199: one gets ${\cal N}=2$ supersymmetry in the bulk and 

200: ${\cal N}=1$ on the branes. Moreover, various fluxes can be turned on, to 

201: stabilize part or all of the closed string moduli. We then turn on 

202: internal magnetic fields~\cite{Bachas:1995ik, Angelantonj:2000hi}, 

203: which, in the T-dual picture, amounts to 

204: intersecting branes~\cite{Berkooz:1996km, bi}. 

205: For generic angles, or equivalently for 

206: arbitrary magnetic fields, supersymmetry is spontaneously broken and 

207: described by effective D-terms in the four-dimensional (4d) 

208: theory~\cite{Bachas:1995ik}. In the weak field limit, 

209: $|H|\alpha'<1$, the resulting mass shifts are given by:

210: \be

211: \delta M^2=(2k+1)|qH|+2qH\Sigma\qquad ;\qquad k=0,1,2,\dots\, ,

212: \label{deltam}

213: \ee

214: where $H$ is the magnetic field of an abelian 

215: gauge symmetry, corresponding to a  Cartan generator of the higher 

216: dimensional gauge group, on a non-contractible 2-cycle of the 

217: internal manifold. $\Sigma$ is the corresponding projection of the 

218: spin operator, $k$ is the Landau level and  $q=q_L+q_R$ is the charge 

219: of the state, given by the sum of the left and right charges of the 

220: endpoints of the associated open string. We recall that the exact 

221: string mass formula has the same form as (\ref{deltam}) with $qH$ 

222: replaced by:

223: \be

224: qH\longrightarrow\theta_L+\theta_R\qquad ;\qquad 

225: \theta_{L,R}=\arctan(q_{L,R}H\alpha')\, ,

226: \label{stringdeltam}

227: \ee

228: where $\alpha'$ is the string Regge slope.

229: Obviously, the field theory expression 

230: (\ref{deltam}) is reproduced in the weak field limit.

231: 

232: To illustrate 

233: the physics, consider an effective six-dimensional (6d) theory 

234: compactified on a magnetized ``2-cycle". From the mass formula 

235: (\ref{deltam}), it follows that all charged scalars become massive, 

236: since the internal spin $\Sigma$ either vanishes (for six-dimensional 

237: scalars), or has eigenvalues $\pm 1$ (for 6d vectors). 

238: Actually, one of the two spin-1 helicities becomes tachyonic, 

239: reflecting the Nielsen-Olesen instability. This tachyon can be 

240: avoided, either  when several magnetic fields are turned on in more than one 

241: internal 2-cycles~\cite{Bachas:1995ik}, 

242: or in more realistic models with ${\cal N}=1$ supersymmetry 

243: in four dimensions. In the former case, one provides a positive 

244: contribution to its mass-squared (see below), while in the latter, one uses an

245: orbifold-type projection which reduces the supersymmetry from its

246: maximal value of the toroidal compactification to ${\cal N}=1$. On the other 

247: hand, fermions have four-dimensional (4d) chiral zero modes, since they 

248: have internal helicities $\Sigma=\pm 1/2$ and only one of the two leads to

249: a massless mode for $k=0$.  Note that neutral 

250: states with respect to the magnetized $U(1)$ generator are 

251: not affected and form ${\cal N}=1$ supermultiplets. In particular, all gauge 

252: bosons of the unbroken gauge group are accompanied by massless 

253: gauginos.

254: 

255: In the general case of a magnetic field pointed in several 

256: directions of the six-dimensional internal manifold, $H$ and 

257: $H\Sigma$ are replaced by ${\rm Tr}JH$ and ${\rm Tr}H\Sigma$, where 

258: $J$ is the antisymmetric ``identity" matrix with elements $+1(-1)$ 

259: above (below) the diagonal and zero everywhere else. For instance, 

260: when the internal manifold is a product of three factorized tori 

261: $\prod_{I=1}^3 T^2_{(I)}$, one has $H=\sum_I H_I$ and 

262: $H\Sigma=\sum_IH_I\Sigma_I$, where $\Sigma_I$ is the projection of 

263: the internal helicity along the $I$-th plane. For a ten-dimensional (10d) spinor,

264: its eigenvalues are $\Sigma_I=\pm 1/2$, while for a 10d 

265: vector $\Sigma_I=\pm 1$ in one of the planes $I=I_0$

266: and zero in the other two $(I\ne I_0)$. Thus, charged higher dimensional

267: scalars become massive, fermions lead to chiral 4d zero modes if all 

268: $H_I\ne 0$, while the lightest scalars coming from 10d vectors have masses

269: \be

270: M_0^2=\left\{\begin{matrix}

271: |qH_1|+|qH_2|-|qH_3|\cr 

272: |qH_1|-|qH_2|+|qH_3|\cr

273: -|qH_1|+|qH_2|+|qH_3|\cr

274: \end{matrix}\right.

275: \label{scalars}

276: \ee

277: Note that all of them can be made positive definite if all $H_I\ne 0$.

278: Moreover, one can easily show that if a scalar mass vanishes, 

279: some supersymmetry remains unbroken~\cite{Angelantonj:2000hi}.

280: 

281: The Gauss law for the 

282: magnetic flux implies that the fields $H_i$ are quantized in terms of 

283: the area of the corresponding 2-cycles $A_i$:\footnote{The

284: index $i$ becomes identical to $I$ above, when the 6d internal manifold

285: is a product of three factorized tori. In the general case, $i$ denotes

286: all possible two-cycles, even non-factorizable.}

287: \be

288: H_i={m_i\over n_i 

289: A_i}\, ,

290: \label{Hquant}

291: \ee

292: where the integers $m_i,n_i$ correspond 

293: to the respective magnetic and electric charges; $m_i$ is the 

294: quantized flux and $n_i$ is the wrapping number of the higher 

295: dimensional brane around the corresponding internal 2-cycle. 

296: For a rectangular torus of radii $R_1$ and $R_2$ in the directions

297: $X_1$ and $X_2$, the area is $A=R_1R_2$.

298: Open string propagation in magnetic fields has a T-dual representation

299: in terms of D-branes at angles. For instance, starting with a D$p$ brane  

300: on a magnetized rectangular torus and applying a T-duality in the direction 

301: $X_2$, $R_2\to\alpha'/R_2$, leads to a D$(p-1)$ brane wrapped on a 

302: direction forming an angle $\theta$ relative to the $X_1$ 

303: axis, given by the dual of the magnetic field: 

304: \be

305: H\alpha'\to\tan\theta={mR_2\over nR_1}\, .

306: \label{dual}

307: \ee

308: Thus, the integers $m$ and $n$ in (\ref{Hquant}) become the wrapping 

309: numbers around the $X_2$ and $X_1$ directions, respectively.

310: 

311: We consider now 

312: several abelian magnetic fields $H_i^a$ of different Cartan generators 

313: $U(1)_a$, so that the gauge group is a product of unitary factors 

314: $\prod_a U(N_a)$ with $U(N_a)=SU(N_a)\times U(1)_a$. In an 

315: appropriate T-dual 

316: representation, it amounts to consider several stacks of 

317: D6-branes intersecting in the three internal tori at angles 

318: determined by the magnetic fields according to (\ref{dual}). An 

319: open string with one end on the $a$-th stack has charge $\pm 1$ under 

320: the $U(1)_a$, depending on its orientation, and is neutral with 

321: respect to all others. Using the results described above, the 

322: massless spectrum of the theory falls into three sectors~\cite{bi}:

323: \begin{enumerate}

324: \item Neutral open strings ending on 

325: the same stack, giving rise to ${\cal N}=1$ gauge supermultiplets of gauge 

326: bosons and gauginos.

327: \item Doubled charged open strings from a 

328: single stack, with charges $\pm 2$ under the corresponding $U(1)$, 

329: giving rise to massless fermions transforming in the antisymmetric or 

330: symmetric representation of the associated $SU(N)$ factor. Their 

331: bosonic superpartners become massive. For factorized toroidal 

332: compactifications $(T^2)^3$, the multiplicities of chiral fermions 

333: are given by:

334: \ba

335: {\rm Antisymmetric}&:&\quad {1\over 2}\left(\prod_I 

336: 2m_I^a\right)\left(\prod_J n_J^a+1\right)\nonumber\\

337: {\rm 

338: Symmetric}&:&\quad {1\over 2}\left(\prod_I 2m_I^a\right)\left(\prod_J 

339: n_J^a-1\right)

340: \label{dcmult}

341: \ea

342: where $a$ denotes the D-brane 

343: stack, $I$ is the label of the two-torus $T^2_{(I)}$, and $m_I^a, 

344: n_I^a$ are the integers entering in the expression of the magnetic 

345: field (\ref{Hquant}). For orbifolds or more general Calabi-Yau spaces, 

346: the above multiplicities may be further reduced by the corresponding 

347: supersymmetry projection down to ${\cal N}=1$.

348: 

349: In the degenerate case where 

350: a magnetic field vanishes, say, along one of the tori ($m_I^a=0$ for 

351: some $I$), there are no chiral fermions in $d=4$ dimensions, but the 

352: same formula with the products extending over the other two 

353: magnetized tori gives the multiplicities of chiral fermions in $d=6$. 

354: In this case, chirality in four dimensions may arise only when the 

355: last $T^2$ compactification is combined with some additional 

356: orbifold-type projection.

357: \item Open strings stretched between 

358: two different brane stacks, with charges $\pm 1$ under each of the 

359: corresponding $U(1)$s. They give rise to chiral fermions transforming 

360: in the bifundamental representation of the two associated unitary 

361: group factors. Their multiplicities, for toroidal compactifications, 

362: are given by:

363: \ba

364: (N_a,N_b)&:&\prod_I (m_I^a n_I^b+n_I^a 

365: m_I^b)

366: \nonumber\\

367: (N_a,{\overline N}_b)&:&\prod_I (m_I^a n_I^b-n_I^a 

368: m_I^b)\, .

369: \label{scmult}

370: \ea

371: As in the previous case, when a factor in the products of the above 

372: multiplicities vanishes, there are no 4d chiral 

373: fermions, but the same formula with the product extending over the 

374: other two magnetized tori gives the corresponding multiplicity of 

375: chiral fermions in $d=6$.

376: \end{enumerate}

377: 

378: As mentioned already above, 

379: all charged bosons are massive. Massless scalars can appear only when 

380: some supersymmetry remains unbroken. In case 2 of doubled charged 

381: strings from the same stack, the requirement of massless scalars is 

382: equivalent to unbroken supersymmetry on the corresponding brane stack. 

383: For toroidal compactifications, using the mass formula (\ref{scalars}), 

384: the condition for the $a$-th stack is

385: \be

386: \delta H_a\equiv

387: \epsilon_1H_1^a+\epsilon_2H_2^a+\epsilon_3H_3^a=0\, ,

388: \label{susya}

389: \ee

390: where $H_I^a$ is the magnetic field of $U(1)_a$ on the 

391: two-torus $T^2_{(I)}$, and $\epsilon_I$ are signs $\pm$ with one

392: at least different from the others

393: ($(\epsilon_1,\epsilon_2,\epsilon_3)=(+,+,-)$ or $(+,-,+)$ or $(-,+,+)$).

394: In case 3 of strings ending on two different sets of branes, massless

395: scalars arise when one has unbroken supersymmetry locally, at the 

396: intersection. The generalization of the above condition is:

397: \ba

398: \delta H_{ab}\equiv\delta H_a-\delta H_b=0\, .

399: \label{susyab}

400: \ea

401: In the T-dual representation, condition (\ref{susyab}) involves

402: the relative intersection angles $(\theta_I^a-\theta_I^b)$, defined

403: as in eq.~(\ref{dual}).

404: 

405: It is now clear that the above framework leads 

406: to models with a tree-level spectrum realizing the idea of split 

407: supersymmetry. Embedding the Standard Model (SM) in an appropriate 

408: configuration of D-brane stacks, one obtains tree-level

409: massless gauginos while 

410: all scalar superpartners of quarks and leptons typically get masses at the  scale 

411: of the magnetic fields, whose magnitude is set by the compactification scale 

412: of the corresponding internal space. 

413: 

414: On the other hand, the condition to obtain a (tree-level)

415: massless Higgs in the spectrum implies that 

416: supersymmetry remains unbroken in the Higgs sector, leading to a pair 

417: of massless higgsinos, as required by anomaly cancellation. Note that 

418: since the Higgs doublet has the same quantum numbers with leptons, it 

419: is likely that lepton doublets have the same open string origin as 

420: the Higgs scalar, and thus, left-handed sleptons are also massless at 

421: the tree-level.

422: 

423: \section{Gauge coupling unification} \label{unification}

424: 

425: On general grounds, there are two conditions to obtain unification of

426: Standard Model gauge interactions, consistently with extrapolation of

427: gauge couplings from low-energy data using the minimal supersymmetric

428: SM spectrum. (i) Equality of the $SU(3)$ color and weak $SU(2)$ 

429: non-abelian gauge couplings and (ii) the correct prediction for the weak 

430: mixing angle $\sin^2\theta_W=3/8$ at the grand unification (GUT) scale.

431: On the other hand, a generic D-brane model using several stacks, as

432: described in the framework of the previous section, does not satisfy

433: either of the two conditions. Indeed, this framework was developed

434: in connection to the idea of low-scale 

435: strings~\cite{Arkani-Hamed:1998rs,Antoniadis:1998ig}, where the concept of

436: unification is radically different from conventional GUTs.

437: In this section, we study precisely the general requirements for 

438: satisfying the first of the above two conditions, namely

439: natural unification of non-abelian gauge couplings.

440: The second condition is more involved and model dependent, 

441: since it is related with the particular hypercharge embedding and

442: will be discussed in section~\ref{models}.

443: 

444: The four-dimensional non-abelian gauge coupling $\alpha_{N_a}$ of the 

445: $a$-th brane stack is given by:

446: \ba

447: {1\over \alpha_{N_a}}={V^a\over 

448: g_s}\prod_I |n_I^a|\sqrt{1+(H_I^a\alpha')^2}\, ,

449: \label{ga}

450: \ea

451: where $g_s$ is the string coupling and $V^a$ the compactification volume 

452: in string units of the internal space of the $a$-th brane stack. 

453: The presence of the wrapping numbers $|n_I^a|$ can be understood from 

454: the fact that $|n_I^a|V^a_I$ is the effective area of the 2-torus $T^2_{(I)}$

455: wrapped $n_I^a$ times by the brane, and $V^a=\prod_IV^a_I$.

456: The additional factor in the square root follows from the non-linear

457: Dirac-Born-Infeld (DBI) action of the abelian gauge field, 

458: $\sqrt{\det (\delta_{ij}+F_{ij}\alpha')}$, which in the case of two dimensions

459: with $F_{ij}=\epsilon_{ij}H$, it is reduced to $\sqrt{1+(H\alpha')^2}$.

460: Obviously, the expression (\ref{ga}) holds at the compactification scale, 

461: since above it gauge couplings receive 

462: important corrections and become higher dimensional. 

463: Finally, the gauge couplings of the associated abelian factors, in our 

464: convention of $U(1)$ charges, are given by

465: \be 

466: \alpha_{_{U(1)_a}}={\alpha_{N_a}\over{2N_a}}\, . 

467: \label{gua}

468: \ee

469: Here, non-abelian generators are normalized according to 

470: ${\rm Tr}T^aT^b=\delta^{ab}/2$. 

471: 

472: From equation (\ref{ga}), it follows that 

473: unification of non-abelian gauge couplings holds if (i) $V^a$ and 

474: (ii) $\prod_I|n_I^a|$ are independent of $a$, while (iii) the 

475: magnetic fields are either $a$-independent as well, or they are much 

476: smaller than the string scale. 

477: \begin{itemize}

478: \item The first 

479: condition (i) is automatically satisfied for D9-branes, since then

480: $V^a=V$, the total volume of the six dimensional internal manifold. 

481: \item The second condition (ii) is satisfied for a large class 

482: of models with $|n_I^a|=1$, which is the point particle field theoretic value of 

483: Dirac quantization for magnetic fields (no multiple brane wrapping). 

484: Actually, this value follows also from eq.~(\ref{dcmult}), by 

485: requiring the absence of chiral fermions transforming in the 

486: symmetric representations of the non-abelian groups, {\em i.e.} no 

487: chiral $SU(3)$ color sextets and no weak $SU(2)$ 

488: triplets.\footnote{The vanishing of the 

489: multiplicity (\ref{dcmult}) is also realized when some $m_I^a=0$, 

490: which is a trivial solution since in this case the corresponding 

491: magnetic field vanishes.}

492: \item The third condition (iii) of weak 

493: magnetic fields is more quantitative. Allowing for $1\%$ error in the 

494: unification condition at high scale, one should have 

495: $H_I^a\alpha'\simlt 0.1$. From the quantization condition 

496: (\ref{Hquant}), this implies that the volume $V\simgt 10^3$ for three 

497: magnetized tori, which is rather high to keep the theory weakly 

498: coupled above the compactification scale. Indeed, eq.~(\ref{ga})

499: gives a string coupling $g_s$ of order ${\cal O}(10)$ for gauge couplings

500: $\alpha_{N_a}\simeq 1/25$ at the unification scale. On the other hand, 

501: for one or two magnetized tori one obtains $V\simgt 10-10^2$, 

502: which is compatible with a string weak 

503: coupling regime $(g_s\sim 0.1-1)$. Of course this discussion

504: should be taken with caution, because there is an uncertainty

505: in the relation of $g_s$ with the string loop expansion parameter.

506: Here, we were conservative and defined it as in a 4d gauge theory.

507: In a 10d theory however, there may be additional powers of $2\pi$

508: which would improve significantly perturbativity~\cite{ah}.

509: 

510: Actually, the condition of weak magnetic fields can 

511: be partly relaxed in some direction, by requiring the absence of 

512: chiral antiquark doublets in the spectrum. Indeed eq.~(\ref{scmult}),

513: for open strings stretched between the strong $SU(3)$ and 

514: weak $SU(2)$ interactions brane stacks, 

515: implies the vanishing of one of the factors in the product. 

516: This leads to the equality of the ratio $m_I^a/n_I^a$ for the 

517: two stacks and for some $I$, and thus, to the equality of the two 

518: corresponding magnetic fields via eq.~(\ref{Hquant}).\footnote{This 

519: argument is true only when the $U(1)$ accompanying the weak 

520: interactions brane stack participates in the hypercharge combination. 

521: Otherwise, quark anti-doublets are equivalent to quark doublets 

522: (see example in section~\ref{models}).} As a result, 

523: the condition of perturbativity is weakened and becomes 

524: possible even in the case of three factorized magnetized tori.

525: \end{itemize}

526: 

527: Note that in the T-dual representation of 

528: intersecting D6-branes, the unification conditions discussed above

529: appear less natural. 

530: In the expression (\ref{ga}) of gauge couplings, the numerator

531: ($V^a$ times the product) 

532: is replaced by the volume of the 3-cycle around which the 

533: D6 brane wraps. For instance, in the case of three factorized 

534: rectangular tori of radii $R^I_1$ and $R^I_2$ in string units, it is 

535: given by $\prod_I\sqrt{(n_I^aR_1^I)^2+(m_I^aR_2^I)^2}$. 

536: The same unification conditions then hold in this context, 

537: with the requirement of weak magnetic field replaced 

538: by the requirement of small angle, which is equivalent to

539: the inequality $R_2^I<<R_1^I$ (see eq.~(\ref{dual})).

540: 

541: The above analysis concerns mainly the QCD and 

542: $SU(2)_L$ gauge couplings $\alpha_3$ and $\alpha_2$. 

543: The case of hypercharge is more subtle since 

544: it can be in general a linear combination of several $U(1)$s coming 

545: from different brane stacks. 

546: In section~\ref{models}, for the purpose of illustration, 

547: we present an explicit example with the correct prediction of the

548: weak mixing angle. It is based on a minimal Standard 

549: Model embedding in three brane stacks with the hypercharge being a 

550: linear combination of two abelian factors. This provides an existence

551: proof that can be generalized in different constructions.

552: We notice for instance that in a class of supersymmetric

553: models with four brane stacks, the equality of the two non-abelian

554: couplings $\alpha_2=\alpha_3$ implies the value $3/8$ for

555: $\sin^2\theta_W$ at the unification scale~\cite{Blumenhagen:2003jy}.

556: 

557: \section{Mass scales and supersymmetry breaking}

558:  \label{susybr}

559: 

560: The supersymmetry breaking scale $m_{\rm S}$ 

561: on the brane stacks is given by

562: the lightest charged scalar masses (\ref{scalars}), or equivalently by 

563: $\delta H_a=\sum_I\epsilon_I H_I^a$ of eq.~(\ref{susya}), in the

564: weak field limit. In the case of strong magnetic fields, of order

565: of the string scale, $H_I^a$ should be replaced by the angles

566: $\theta_I^a$ according to eqs.~(\ref{stringdeltam}) and (\ref{dual}).

567: For magnetic fields in more than one internal planes, 

568: $m_{\rm S}$ can therefore be smaller 

569: than their magnitude, and consequently from the 

570: corresponding compactification scales (\ref{Hquant}).

571: Similarly, on brane intersections, the supersymmetry breaking 

572: scale is given by the differences $\delta H_{ab}$ of eq.~(\ref{susyab}),

573: and thus, can be again smaller than $H_I^a$ and the 

574: compactification scales.

575: 

576: Let us now discuss the various mass scales.

577: To preserve gauge coupling unification, the (non-gravitational part of the) theory must remain 4-dimensional up to the unification scale.  So  the compactification scale (actually the smallest, if there are several)

578: must be no smaller than the unification energy, 

579: $M_{\rm GUT}\simeq 10^{16}$ GeV, and we will take them to be of the same order. Above the compactification scale, gauge interactions

580: acquire a higher dimensional behavior.

581: So, to keep the theory weakly coupled, the string

582: scale $M_s\equiv{\alpha'}^{-1/2}$ should be close to the 

583: compactification scale and therefore to $M_{\rm GUT}$.

584: Moreover, as we discussed in the previous section, 

585: to ensure that corrections to the unification of gauge couplings

586: are within 1\%, the magnetic fields should be weak.

587: It follows that the string scale should be roughly a factor of 3 

588: higher than the compactification scale, 

589: \be

590: M_s\simgt 3\, M_{\rm GUT}\, .

591: \label{Ms}

592: \ee

593: 

594: On the other hand, as we pointed out above, $m_{\rm S}$

595: can be lower than $M_{\rm GUT}$. Although much lower

596: values require an apparent fine tuning of radii, such a 

597: tuning is technically natural since the supersymmetric 

598: point $m_{\rm S}=0$ is radiatively stable.

599: One can therefore treat $m_{\rm S}\sim|\delta H|^{1/2}$ 

600: as free parameter and drop for simplicity the brane stacks 

601: dependent index in $\delta H$.

602: 

603: All scalar masses are of order $|\delta H|^{1/2}$ except for

604: those coming from supersymmetric sectors, which are vanishing to 

605: lowest order, such as the Higgs and possibly the slepton doublets.

606: The latter are expected to acquire masses from one loop 

607: corrections, proportional to $|\delta H|^{1/2}$ but suppressed 

608: by a loop factor. 

609: Note that off diagonal elements of the $2\times 2$ Higgs 

610: mass matrix, usually denoted by $B\mu$, should also 

611: be generated at the same order as the diagonal elements, 

612: in the absence of a Peccei-Quinn (PQ) symmetry. For 

613: high $\delta H$, a fine tuning between $B\mu$ and the diagonal 

614: elements is then required to ensure a light higgs. 

615: 

616: It remains to discuss the corrections to gaugino and 

617: higgsino masses, $m_{1/2}$ and $\mu$, which are vanishing at the 

618: tree-level. In the absence of gravity, they are both protected by an 

619: R-symmetry. Actually, higgsino masses are protected in addition by a 

620: PQ symmetry which must be broken in order to generate a $B\mu$ 

621: mixing term in the Higgs mass matrix, as we argued above. 

622: Then, a $\mu$-term can be generated via $B\mu$, 

623: or directly using the PQ symmetry breaking, 

624: if R-symmetry is broken. Indeed,

625: R-symmetry is in general broken in the gravitational sector 

626: by the gravitino mass $m_{3/2}$ and thus, in the presence 

627: of gravity, $m_{1/2}$ and $\mu$ are not anymore protected.

628: 

629: Thus, the study of fermion masses requires some knowledge of

630: supersymmetry breaking in the gravity sector, which has been

631: ignored up to now. A related issue is the cancellation of the

632: cosmological constant between brane and bulk contributions,

633: in order to maintain the flat space background.

634: The brane contribution comes from the supersymmetry breaking

635: due to the magnetic field and scales as $(\delta H)^p$, in string units,

636: where the power $p$ depends on the number ${\cal N}$ of bulk 

637: supersymmetries broken by $\delta H$. For the maximal value

638: of ${\cal N}=4$ it was found that $p=3$~\cite{Bachas:1995ik}, 

639: while for ${\cal N}=1$ we expect that $p=1$, 

640: or more precisely $(\delta H)H$ from the

641: form of the DBI action (\ref{ga}). 

642: 

643: The vanishing of the vacuum energy implies that an additional 

644: source of supersymmetry breaking should probably be introduced

645: in the closed string sector (bulk).

646: The corresponding dominant bulk contribution to the cosmological

647: constant is proportional in general to $m_{3/2}^2\Lambda^2$, 

648: with $\Lambda$ the ultraviolet (UV) cutoff.

649: Combining the brane and bulk contributions, one obtains

650: \be

651: m_{3/2}\sim {|H|^{1/2}\over \Lambda}\times |\delta H|^{1/2}\, .

652: \label{mgrav}

653: \ee

654: Thus, for $\Lambda\simeq M_s$ the gravitino mass is

655: in general of the same order as the scalar masses,

656: while for $\Lambda\simeq M_P$, the 4d Planck mass, it is 

657: about three orders of magnitude lower.

658: 

659: In the following, we will consider for concreteness a source of

660: bulk supersymmetry breaking via Scherk-Schwarz (SS) ~\cite{SS} boundary

661: conditions along a ``gravitational" interval $S^1/Z_2$ 

662: of length $\pi R$~\cite{Antoniadis:1996hk}. This interval

663: can be identified either with the eleventh dimension of M-theory~\cite{Horava:1996ma}, or

664: with some internal orientifold direction of type I string theory, transverse

665: to all ``observable" brane stacks where the Standard Model is

666: localized~\cite{Antoniadis:1998ki}. The gravitino mass is then given by:

667: \be

668: m_{3/2}={\omega\over R}\, ,

669: \label{mgravitino}

670: \ee

671: where $\omega\in [0,1]$ is the parameter of the SS deformation.

672: It originates from the boundary conditions of the five dimensional 

673: fields which are periodic up to a phase 

674: of a symmetry transformation~\cite{SS}. 

675: The latter can be parametrized as $e^{2i\pi\omega/R}$ 

676: and corresponds to a discrete rotation in the internal compactified 

677: space, in order to give a mass to the gravitino. Therefore,

678: $\omega$ is quantized and equals $1/N$ for $Z_N$.

679: 

680: Using now the usual relations that express the 4d Planck mass 

681: $M_P$ and the gauge coupling at the unification scale $\alpha_G$ in

682: terms of the string parameters (scale, coupling and compactification

683: sizes), one finds upon eliminating the string 

684: coupling~\cite{Antoniadis:1996hk, Antoniadis:1998ig}:

685: \be

686: R^{-1}={2\over\alpha_G^2}{M_s^3\over M_P^2}V^{-1}\, ,

687: \label{RSS}

688: \ee

689: where $V$ is the internal compactification volume (in string units) 

690: of all SM branes. Substituting the values discussed above for 

691: $M_s$ and $V$, one finds $R^{-1}\simeq{\cal O}(10^{13}-10^{14})$ 

692: GeV. Following our previous discussion on the cancellation of

693: vacuum energy, when combining the SS bulk supersymmetry 

694: breaking with the brane magnetic fields, one expects that

695: $|\delta H|^{1/2}$ should also be of the same order $10^{13}$ GeV.

696: On the other hand, because of the uncertainty in the value of the

697: relevant UV cutoff in eq.~(\ref{mgrav}), the scalar masses could be

698: either significantly higher, of order of the unification scale,

699: or lower, of order of $10^{10}$ GeV for $\Lambda\simeq R^{-1}\simeq m_{3/2}$,

700: as was argued in the context of SS compactifications~\cite{Antoniadis:1990ew}.

701: Thus, in section~\ref{pheno}, we will study the phenomenology of the

702: whole range of scalar masses $m_S\simeq 10^{10}-10^{16}$ GeV.

703: 

704: 

705: \section{Light gaugino masses} \label{gauginos}

706: 

707: In the presence of R-symmetry, gluinos can only get a Dirac mass 

708: by pairing up with other color octet fermions, which spoils gauge 

709: coupling unification. So, gluinos must either be massless, 

710: which is phenomenologically strongly disfavored, or get an 

711: R-breaking Majorana mass. The latter requires a source for R-breaking, 

712: and would also permit, in combination with PQ breaking, 

713: the generation of higgsino masses.

714: 

715: One possible source of R-symmetry breaking is the Majorana 

716: mass for the gravitino. Such a mass is always present, 

717: as a result of canceling the tree-level cosmological constant, 

718: in theories where there is an energy regime in which 4d supergravity holds. 

719: A second possibility is that there is no such an energy regime, 

720: the gravitino gets an R-preserving Dirac mass by pairing up 

721: with another spin-3/2 fermion, and R-symmetry is broken 

722: spontaneously by a dynamical condensate. In this section, we will 

723: consider both possibilities, beginning with the first.

724: 

725: The first possibility has already been studied in some detail 

726: in the effective field theory~\cite{Antoniadis:1997ic,savnim,Arkani-Hamed:2004yi}. 

727: Once R-symmetry, as well as supersymmetry, is broken through 

728: the Majorana gravitino mass, the gauginos can get a mass 

729: in a number of ways. One is anomaly mediation~\cite{Randall:1998uk}, 

730: whose leading contribution

731: can be adequately suppressed  to allow for light gauginos in the presence of heavy gravitinos~\cite{Antoniadis:1997ic,Arkani-Hamed:2004yi}.

732: 

733: In string theory, gaugino masses mediated from closed string radiative corrections 

734: have been studied recently and shown to be generated at lowest order by 

735: string diagrams of ``one and a half" loop (``genus" 3/2)~\cite{Antoniadis:2004qn}. 

736: They contain for instance one handle and one boundary. 

737: It turns out that for generic compactifications and 

738: supersymmetry breaking mechanism, the resulting gaugino 

739: masses are proportional to the gravitino mass for small 

740: $m_{3/2}$ compared to the string scale: $m_{1/2}\sim\alpha^2 m_{3/2}$ 

741: with $\alpha$ the corresponding gauge coupling. 

742: Apart from the power of gauge coupling, this 

743: result is similar to the contribution of anomaly mediation~\cite{at}.

744: 

745: Suppression of the anomaly mediation contribution 

746: in radiative corrections may arise as follows.

747: A generic contribution to the gaugino mass

748: involves gravity and gauge loops and should contain a 

749: gravitino mass insertion that brings one power of $m_{3/2}$.

750: From an effective field theory analysis, one expects that the 

751: dominant contribution of each gravity loop is proportional to 

752: $\Lambda^2/M_P^2$, with $\Lambda$ the UV cutoff, 

753: since each gravitational vertex brings an inverse power of 

754: Planck mass $M_P$ and the loop is quadratically divergent. 

755: Moreover, gauge loops do not modify this power counting.

756: If the UV cutoff is set up by the Planck scale, $m_{1/2}$ would be

757: proportional to $m_{3/2}$.

758: However, in special models with supersymmetry breaking via 

759: Scherk-Schwarz compactifications, one expects a UV cutoff 

760: set by the compactification scale~\cite{Antoniadis:1990ew}, 

761: in which case the dominant contribution comes 

762: from one gravitational loop, leading to 

763: $m_{1/2}\propto m_{3/2}^3/M_P^2$~\cite{Antoniadis:1997ic,savnim}.

764: Thus, gauginos (and higgsinos) become light around the TeV 

765: scale, for instance when 

766: $m_{3/2}\sim|\delta H|^{1/2}\simeq 10^{13-14}$ GeV 

767: as we argued above.

768: 

769: On the other hand, in the string theory analysis,

770: it was found that for orbifold compactifications the 

771: corrections to $m_{1/2}$ are exponentially suppressed for small 

772: gravitino mass at the lowest non-trivial order 

773: of ``genus" 3/2~\cite{Antoniadis:2004qn}.

774: This is an indication that these models are indeed examples

775: of quantum gravitational suppressed anomaly mediation, 

776: which may be checked by going to the next non-trivial order.

777: 

778: We now present a different mechanism to protect gaugino 

779: masses in the presence of gravity, based on symmetry.

780: It involves theories in which there is no 4d supergravity energy regime, 

781: the gravitino gets an R-preserving Dirac mass by pairing up with 

782: another spin-3/2 fermion, and therefore does not feed a mass to the gluino. 

783: The breaking of R-symmetry, which is necessary to give masses to the 

784: gluinos and higgsinos, can subsequently occur via a dynamical condensate.

785: 

786: We begin from the observation that the lowest order

787: perturbative correction to $m_{1/2}$ is exactly vanishing for

788: the case of $Z_2$ SS deformation with $\omega=1/2$ (see 

789: eq.~(\ref{mgravitino}))~\cite{Gherghetta:2001sa, Antoniadis:2004qn}. 

790: Indeed, here we will argue that the 

791: usual SS compactification with a $Z_2$ shift in a 

792: direction transverse to the brane 

793: leaves unbroken a generalized R-type symmetry, which 

794: guarantees the vanishing of gaugino masses in the full theory. 

795: We will use the effective field theory description of the whole tower 

796: of Kaluza-Klein (KK) excitations for a generic SS compactification

797: on a circle of radius $R$~\cite{Bagger:2001ep}. 

798: A massless five-dimensional (5d)

799: spinor $\psi$, with SS boundary conditions twisted by the phase

800: $e^{2i\pi\omega/R}$ around the circle, gives rise in four 

801: dimensions to the following $2\times 2$ mass matrix for 

802: each pair of KK levels $|n>$ and $|-n>$~:

803: \ba

804: {1\over R}

805: \left(\psi^L_n\ 

806: \psi^R_n\right)

807: \left(\begin{matrix}

808: \omega&n\cr

809: n&\omega\cr

810: \end{matrix}\right)

811: \left(\begin{matrix}

812: \psi^L_n\cr

813: \psi^R_n\cr

814: \end{matrix}\right)\quad 

815: ;\quad n=0,1,2\dots\, ,

816: \label{KKmass}

817: \ea

818: where $\psi^{L,R}_n$ is 

819: the left (L) and right (R) component of the $n$-th KK excitation of 

820: the fermion. The diagonal element $\omega$ arises from

821: the SS deformation. The eigenvalues of the above mass 

822: matrix are $(\omega\pm n)/R$, reproducing the familiar shift of the 

823: KK number. 

824: 

825: Here, we consider a SS direction which is 

826: transverse to the brane stack, so that gauginos are not affected 

827: and supersymmetry remains unbroken 

828: on the branes in the presence of the SS deformation. 

829: The only source of supersymmetry breaking on the branes comes

830: from the magnetic fields on their world volume, which give 

831: masses to all charged scalars, as we described previously. 

832: Thus, in this case, the SS direction is not a circle but an interval 

833: $S^1/Z_2$, with $Z_2$ being the inversion of the extra coordinate.

834: Its action on the KK spectrum consists of sending $|n>\to|-n>$, 

835: while at the same time acts on 4d fermion chiralities:

836: left-hand components are invariant and right-handed change sign.

837: As a result, the $Z_2$ projection on the KK spectrum of a 5d spinor

838: keeps the left-handed symmetric and right-handed antisymmetric

839: combinations of states $(|n>_{_L}+|-n>_{_L})/\sqrt{2}$ 

840: and $(|n>_{_R}-|-n>_{_R})/\sqrt{2}$, having cosine and 

841: sine wave functions, respectively. Keeping the same notation

842: $\psi^L_n$ and $\psi^R_n$ for these $Z_2$ invariant combinations,

843: the fermion mass terms can be easily deduced from the

844: expression (\ref{KKmass}) and take the form:

845: \ba

846: {\omega\over R}\psi^L_0\psi^L_0+{1\over R}\sum_{n\ge 1}

847: \left(\psi^L_n\ \psi^R_n\right)

848: \left(\begin{matrix}

849: \omega&n\cr

850: n&\omega\cr

851: \end{matrix}\right)

852: \left(\begin{matrix}

853: \psi^L_n\cr

854: \psi^R_n\cr

855: \end{matrix}\right)\, .

856: \label{KKmassorb}

857: \ea 

858: 

859: For a generic SS deformation, corresponding for instance to a 

860: $Z_N$ shift with $\omega=1/N$ and $N>2$, a simple inspection of 

861: eq.~(\ref{KKmassorb}) shows that there is a tower of Majorana 

862: masses for the gravitino KK modes,

863: that break the R-symmetry of global supersymmetry.

864: Thus, gauginos are expected to acquire masses through

865: gravitational radiative corrections. Moreover, despite the

866: quantization of $\omega$, one can define three energy regimes

867: in the gravity sector. A low-energy 4d non-supersymmetric region

868: below the lightest KK gravitino mass $\omega/R$, a 5d supergravity

869: region at energies higher than the compactification scale $E>>1/R$,

870: and an intermediate regime at energies $\omega/R<E<1/R$,

871: where one can define a 4d ${\cal N}=1$ spontaneously broken 

872: supergravity. The latter can be obtained by integrating out all

873: heavier KK excitations with $n>1$ and describes the physics of

874: the gravitino ``zero mode" $n=0$.

875: 

876: This general picture breaks down in the $Z_2$ case $\omega=1/2$,

877: due to a new pairing that arises in the KK spectrum.

878: The ``zero mode" becomes degenerate with the lightest eigenstate 

879: of the $2\times 2$ mixing matrix for the first KK excitation $n=1$, 

880: with mass eigenvalue $1/2R$. This degeneracy continues 

881: similarly to all KK levels; the heaviest eigenstate at level $n$

882: with mass eigenvalue $(n+1/2)/R$, becomes degenerate with the 

883: lightest eigenstate at level $n+1$. 

884: Thus, all masses can be rewritten in a Dirac type form and one

885: can define a new unbroken R-symmetry that keeps gauginos 

886: on the transverse branes massless. Note also that in this

887: case, there is no intermediate energy regime

888: where one can define a 4d ${\cal N}=1$ supergravity, since

889: after the SS deformation, the 4d gravitino zero mode is degenerate

890: with another state coming from its $n=1$ KK excitation.

891: Including this extra spin-3/2 state in the effective theory, one

892: should also include its degenerate $n=1$ companion at the

893: symmetric phase, which however, after the SS deformation, 

894: becomes degenerate with the lightest eigenstate from the

895: next level $n=2$, and so on. In the effective supergravity,

896: one should therefore include the whole KK tower and the

897: intermediate energy regime is lost.  

898: 

899: The low-energy 4d non-supersymmetric region 

900: without any gravitino mode has obviously

901: a chiral symmetry associated to the massless gauginos. 

902: On the other hand, in order to describe the generalized 

903: R-symmetry in the presence of gravity, one has to go directly 

904: to the high energy 5d regime with the whole KK gravitino tower.

905: This phenomenon provides the first example of massive 

906: supergravity coupled to an exact supersymmetric gauge sector

907: (but non-supersymmetric chiral matter) 

908: that survives in the quantum theory.

909: As we mentioned earlier, the R-symmetry can be broken 

910: spontaneously by appropriate dynamics

911: within the effective field theory at much lower 

912: energies and generate gaugino and higgsino masses close to the 

913: electroweak scale, preserving the unification of gauge 

914: couplings. In this case, the corresponding breaking scale

915: is an extra parameter that requires separate dynamics.

916: 

917: \section{Model building} \label{models}

918: 

919: In this section, we present an explicit Standard Model embedding, 

920: in a minimal set of three brane stacks which has a realistic particle content,

921: satisfies the conditions of unification of strong and weak interactions

922: and predicts the correct weak angle $\sin^2\theta_W=3/8$ at the

923: unification scale. This model illustrates our general framework

924: and provides an explicit example where several problems can be

925: addressed and many general phenomenological consequences can

926: be discussed.

927: 

928: Model building with intersecting branes has been extensively 

929: studied in the recent literature, mainly in the context of low-scale string 

930: models~\cite{Arkani-Hamed:1998rs, Antoniadis:1998ig} 

931: (see for example ref.~\cite{Uranga:2003pz} and references therein). 

932: According to the general analysis of ref.~\cite{Antoniadis:2000en}, the 

933: SM  embedding requires usually four stacks of branes, 

934: the color $U(3)$, the weak $U(2)$, together with two abelian ones. The 

935: hypercharge is in general a linear combination of the four $U(1)$s, 

936: while the remaining three orthogonal combinations are usually broken 

937: by anomalies to their global counterparts corresponding to the baryon 

938: and lepton numbers and a Peccei-Quinn symmetry. Moreover, the 

939: value of the weak angle when $\alpha_3=\alpha_2$ is in general 

940: different from $3/8$, which in any case is not a desired value when 

941: the string scale is at low energies. Here, we will focus on a 

942: particular model that was dropped from the analysis of 

943: refs.~\cite{Antoniadis:2000en} because, although minimal 

944: and very economic, it was not appropriate for low string scale. 

945: Its two main defects were the value of the weak angle and the absence 

946: of baryon number as a symmetry to guarantee proton stability.

947: 

948: The model requires three stacks of branes giving rise to 

949: $U(3)\times U(2)\times U(1)$ gauge group. For completeness, 

950: below we will make a general study of SM embedding

951: in three brane stacks~\cite{ar}.

952: The quark and lepton doublets ($Q$ and $L$) correspond 

953: to open strings stretched between the weak and the color or $U(1)$ 

954: branes, respectively. On the other hand, the $u^c$ and $d^c$ antiquarks 

955: can come from strings that are

956: either stretched between the color and $U(1)$ branes, or that have 

957: both ends on the color branes and transform in the antisymmetric 

958: representation of $U(3)$ (which is an anti-triplet). There are 

959: therefore three possible models, depending on whether it is the $u^c$ 

960: (model A), or the $d^c$ (model B), or none of them (model C), the 

961: state coming from the antisymmetric representation of color branes. 

962: It follows that the antilepton $l^c$ comes in a similar way  from 

963: open strings with both ends either on the weak brane stack and 

964: transforming in the antisymmetric representation of $U(2)$ which is 

965: an $SU(2)$ singlet (in model A), or on the abelian brane 

966: and transforming in the ``symmetric" representation of $U(1)$

967: (in models B and C). The three

968: models are presented pictorially in Figure~\ref{fig_models}.

969: \begin{figure}

970: \centerline{

971: \epsfxsize=4.4cm

972: \epsfbox{ModelA.eps}~~~\epsfxsize=4.4cm

973: \epsfbox{ModelB.eps}~~~\epsfxsize=3.9cm

974: \epsfbox{ModelC.eps}}

975: \caption{\label{fig_models}{\it Pictorial representation of models} $A$,  $B$ and $C$.}

976: \end{figure}

977: 

978: Thus, the members of a family of quarks and leptons have the 

979: following quantum numbers:

980: \ba

981: &&{\rm Model\ A}\qquad\qquad\quad\quad 

982: {\rm Model\ B}

983: \qquad\qquad\quad\ {\rm Model\ 

984: C}\nonumber\\

985: \!\!\!\!\!\!\!\! Q && ({\bf 3},{\bf 

986: 2};1,1,0)_{1/6}\quad\quad\quad\ ({\bf 3},{\bf 2};1,\varepsilon_Q 

987: ,0)_{1/6}

988: \qquad\, ({\bf 3},{\bf 2};1,\varepsilon_Q 

989: ,0)_{1/6}\nonumber\\

990: \!\!\!\!\!\!\!\! u^c && (\bar{\bf 3},{\bf 

991: 1};2,0,0)_{-2/3}\quad\quad\ \ \, (\bar{\bf 3},{\bf 

992: 1};-1,0,1)_{-2/3}

993: \quad\ (\bar{\bf 3},{\bf 

994: 1};-1,0,1)_{-2/3}\nonumber\\

995: \!\!\!\!\!\!\!\! d^c && (\bar{\bf 

996: 3},{\bf 1};-1,0,\varepsilon_d )_{1/3}\qquad\, (\bar{\bf 3},{\bf 

997: 1};2,0,0)_{1/3}

998: \qquad\ \ \, (\bar{\bf 3},{\bf 

999: 1};-1,0,-1)_{1/3}

1000: \label{model}\\

1001: \!\!\!\!\!\!\!\! L && ({\bf 1},{\bf 

1002: 2};0,-1,\varepsilon_L)_{-1/2}\quad\ \, ({\bf 1},{\bf 

1003: 2};0,\varepsilon_L,1)_{-1/2}

1004: \quad\ \ \, ({\bf 1},{\bf 

1005: 2};0,\varepsilon_L,1)_{-1/2}\nonumber\\

1006: \!\!\!\!\!\!\!\! l^c && ({\bf 

1007: 1},{\bf 1};0,2,0)_1\qquad\quad\quad\, ({\bf 1},{\bf 

1008: 1};0,0,-2)_1

1009: \qquad\ \ ({\bf 1},{\bf 

1010: 1};0,0,-2)_1\nonumber\\

1011: \!\!\!\!\!\!\!\! \nu^c && ({\bf 1},{\bf 

1012: 1};0,0,2\varepsilon_\nu)_0\qquad\quad\ ({\bf 1},{\bf 

1013: 1};0,2\varepsilon_\nu,0)_0\qquad\ \, ({\bf 1},{\bf 

1014: 1};0,2\varepsilon_\nu,0)_0\nonumber

1015: \ea

1016: where the last three digits after the semi-column 

1017: in the brackets are the charges under the three 

1018: abelian factors $U(1)_3\times U(1)_2\times U(1)$, that we will call 

1019: $Q_3$, $Q_2$ and $Q_1$ in the following, while the subscripts denote 

1020: the corresponding hypercharges. The various sign ambiguities 

1021: $\varepsilon_i=\pm 1$ are due to the fact that the corresponding 

1022: abelian factor does not participate in the hypercharge combination 

1023: (see below). 

1024: In the last line, we also give the quantum numbers of a possible 

1025: right-handed neutrino in each of the three models. These are in fact 

1026: all possible ways of embedding the SM spectrum in three sets of 

1027: branes.

1028: 

1029: The value of the weak angle can be easily computed from the 

1030: hypercharge combination: 

1031: \be

1032: Y=\sum_ic_iQ_i\quad\Rightarrow\quad

1033: \sin^2\theta_W={1\over 

1034: 1+4c_2^2+2c_1^2\alpha_2/\alpha_1+6c_3^2\alpha_2/\alpha_3}\, ,

1035: \label{Ysin}

1036: \ee

1037: where $\alpha_i$ are the non-abelian couplings and the 

1038: numerical coefficients are due to our normalization of $U(1)$ 

1039: charges according to eq.~(\ref{gua}). In our models, the hypercharge 

1040: combination is:

1041: \ba

1042: \label{hyper}

1043: {\rm Model\ A}\quad\ &:&\quad Y=-{1\over 3}Q_3+{1\over 2}Q_2\\

1044: {\rm Model\ B, C}&:&\quad 

1045: Y=\ \ \, {1\over 6}Q_3-{1\over 2}Q_1\nonumber

1046: \ea

1047: leading to 

1048: the following expressions for the weak angle:

1049: \ba

1050: {\rm Model\ A}\quad\ 

1051: &:&\quad \sin^2\theta_W={1\over 2+2\alpha_2/3\alpha_3}

1052: ={3\over 8}\, {\bigg|}_{\alpha_{_2}=\alpha_{_3}}\\

1053: {\rm Model\ B, C}&:&\quad 

1054: \sin^2\theta_W={1\over 1+\alpha_2/2\alpha_1+\alpha_2/6\alpha_3}=

1055: {6\over 

1056: 7+3\alpha_2/\alpha_1}\, 

1057: {\bigg|}_{\alpha_{_2}=\alpha_{_3}}\nonumber

1058: \label{sintheta}

1059: \ea

1060: 

1061: In the second part of the above equalities, we used the unification relation 

1062: $\alpha_2=\alpha_3$, that can be naturally imposed as described in 

1063: section~\ref{unification}. Indeed, it follows by requiring the absence of 

1064: chiral fermions that transform in the symmetric representation of $SU(3)$ 

1065: and $SU(2)$ and the magnetic fields to be roughly an order of magnitude 

1066: smaller than the string scale. The last condition can be partly relaxed 

1067: in model A from the requirement of absence of chiral quark 

1068: ``anti"-doublets in the spectrum. Notice that such states have wrong 

1069: hypercharge, since their $Q_2$ charge is opposite from quark

1070: doublets in eq.~(\ref{hyper}).

1071: This cannot be used in models B 

1072: and C because $Q_2$ does not participate in the hypercharge 

1073: combination, and thus, doublets and anti-doublets are 

1074: indistinguishable. In any case, in these models the unification of 

1075: the two non-abelian couplings is not sufficient to predict the weak 

1076: angle and further conditions are needed for the $U(1)$ coupling 

1077: $\alpha_1$. Such an analysis goes beyond the scope of this paper, 

1078: which is to describe the general framework and present 

1079: a simple example. Indeed, model A admits natural gauge 

1080: coupling unification of strong and weak interactions, realizing the 

1081: conditions we described in section~\ref{unification}, 

1082: and predicts the correct value for 

1083: $\sin^2\theta_W=3/8$ at the unification scale $M_{\rm GUT}$.

1084: 

1085: The spectrum (\ref{model}) can be easily implemented 

1086: with a Higgs sector, since the Higgs field $H$ has the same 

1087: quantum numbers as the lepton doublet or its complex conjugate:

1088: \ba

1089: && {\rm Model\ 

1090: A}\qquad\qquad\quad\quad {\rm Model\ B, C}

1091: \nonumber\\

1092: \!\!\!\!\!\!\!\! H\ && ({\bf 1},{\bf 

1093: 2};0,-1,\varepsilon_H)_{-1/2}\quad\ ({\bf 1},{\bf 2};0,\varepsilon_H,1)_{-1/2}

1094: \\

1095: \!\!\!\!\!\!\!\! H' && ({\bf 

1096: 1},{\bf 2};0,1,\varepsilon_{H'})_{1/2}\qquad\ \, 

1097: ({\bf 1},{\bf 

1098: 2};0,\varepsilon_{H'},-1)_{1/2}

1099: \nonumber

1100: \label{higgs}

1101: \ea

1102: Actually, as explained in the general framework of 

1103: section~\ref{framework}, the Higgs sector should be locally 

1104: supersymmetric, so that the Higgs scalars are massless

1105: at the tree level,

1106: and thus $H$ and $H'$ correspond to two Higgs chiral supermultiplets.

1107: 

1108: Besides the hypercharge combination, there are two additional 

1109: $U(1)$s. It is easy to check that one of the two can be identified 

1110: with $B-L$.  For instance, in model A choosing the signs 

1111: $\varepsilon_d=\varepsilon_L=-\varepsilon_\nu=

1112: -\varepsilon_H=\varepsilon_{H'}$, it is given by:

1113: \be

1114: B-L=-{1\over 6}Q_3+{1\over 2}Q_2-{\varepsilon_d\over 2}Q_1\, .

1115: \label{BL}

1116: \ee

1117: The other $U(1)$ corresponds to a Peccei-Quinn (PQ) type symmetry. 

1118: $B-L$ can be broken by a vacuum expectation value (VEV) of 

1119: a SM singlet field of the type of $\nu^c$, at a high scale. 

1120: In any case, this model has no baryon number conservation 

1121: and thus proton is unstable by dimension six 

1122: effective operators suppressed by the string scale.

1123: 

1124: The second $U(1)$ combination of PQ type is anomalous. 

1125: The corresponding gauge field should become 

1126: massive via the Green-Schwarz mechanism, by absorbing an axion from 

1127: the Ramond-Ramond (RR) closed string sector~\cite{Sagnotti:1992qw}. 

1128: Usually, its global counterpart survives and remains unbroken 

1129: in perturbation theory at the orbifold point~\cite{Poppitz:1998dj}. 

1130: To avoid the presence of an electroweak axion, one 

1131: should either move away from this point, or 

1132: find some appropriate extension of the model which allows to break

1133: PQ by a scalar VEV at a high scale.

1134: On the other hand, in the presence of magnetic fields, it was noticed

1135: that the RR axions involved in the anomaly cancellation

1136: come from the untwisted orbifold sector~\cite{Angelantonj:2000hi}.

1137: In this case, the global symmetry will be in general broken at the scale

1138: of the anomalous $U(1)$ mass, as in heterotic string models. 

1139: As a result, the axion becomes invisible 

1140: and no PQ symmetry should survive at low energies.

1141: 

1142: 

1143: \section{Constraints from Gluino Cosmology} \label{pheno}

1144: 

1145: The most distinctive signature of split SUSY, decisively

1146: differentiating it from the usual supersymmetric SM, 

1147: is the long-lived gluino, which is the smoking gun of this framework.

1148:  Because the scale of supersymmetry breaking is high,

1149: the squarks are heavy and the lifetime for the gluino to decay

1150: into a quark, antiquark and LSP -- which is mediated by virtual

1151: squark exchange -- is:

1152: \begin{equation}

1153: \tau = 3 \times 10^{-2} {\rm sec} \Big(\frac{m_S}{10^9\,

1154: {\rm GeV}}\Big)^4 \Big( \frac{1\, {\rm TeV}}{m_{\tilde{g}}}\Big)^5,

1155: \end{equation}

1156: where $m_S$ is the squark mass and $m_{\tilde{g}}$ the gluino mass.

1157: We have included a QCD enhancement factor of $\sim 10$ in the

1158: rate, and another factor $\sim 10$ for the number of decay

1159: channels. The longevity of the gluino can lead to a host of interesting 

1160: signatures at the LHC such as displaced vertices, intermittent tracks, 

1161: late decaying gluinos captured near the detector etc., 

1162: which have been discussed in refs.~\cite{savnim,noi,tutti,Arkani-Hamed:2004yi}. 

1163: These signatures depend on the lifetime, which in turn depends 

1164: sensitively, through the above equation, on the gluino mass and 

1165: the squark mass $m_S \sim \sqrt{\delta H}$. These quantities are 

1166: constrained by cosmological considerations, to which we turn next.

1167: 

1168: The most natural value for the squark mass, one that does not require 

1169: tuning the ratio $\delta H/H$ to be small, is 

1170: $m_S \sim \sqrt{\delta H} \sim M_{GUT} \sim 10^{16}$ GeV. 

1171: For this value of $m_S$, and for  $m_{\tilde{g}} \sim 1$ TeV,

1172: the gluino lifetime is of order $3 \times 10^{26}$ sec, much longer 

1173: than the age of the universe. Such cosmologically stable gluinos are 

1174: expected to assemble into color singlet ``R-hadrons" by combining 

1175: with gluons, quarks and antiquarks during the QCD phase transition. 

1176: Subsequently, during the primordial big bang nucleosynthesis, the 

1177: R-hadrons are expected to assemble, often together with ordinary 

1178: nucleons, into nuclei, which will eventually form atoms. 

1179: These atoms will be chemically similar to a familiar atom, 

1180: but will instead have a heavy $\sim$ TeV mass nucleus. 

1181: Searches for such anomalous heavy isotopes are very restrictive. 

1182: The limits on heavy hydrogen isotopes in the mass range up to 

1183: $\sim$ 1 TeV is one such atom in $10^{27}$ nucleons, and for 

1184: isotopes in the mass range from 1 TeV to 10 TeV is $10^{22}$ 

1185: per nucleon~\cite{Smith:1982qu}. The upper limits for heavy (up to 10 TeV) 

1186: isotopes of Helium, Carbon and Oxygen nuclei are 

1187: as small as one atom in $\sim 10^{17}$ nucleons. 

1188: 

1189: These suggest similar limits to the abundances of gluinos relative to 

1190: ordinary matter, since most gluinos and R-hadrons are expected to 

1191: end-up in nuclei. Although one cannot prove that this is inescapable, 

1192: it is hard to imagine that none of the many possible ways in which 

1193: R-hadrons and ordinary hadrons  can combine with nucleons into 

1194: some low-Z nuclei is realized. More precisely, there is a multitude 

1195: of ways in multi-quark states can combine with a gluino into a 

1196: color-singlet state of charge 0,1, 2, 6 or 8, and it is unlikely that none 

1197: of these bound states form. So the abundance of gluinos relative to 

1198: ordinary matter should probably be as small as $10^{-27}$ per nucleon, 

1199: or $10^{-37}$ per photon, to account for the absence of heavy 

1200: hydrogen isotopes of mass up to 1 TeV, or less than $10^{-32}$ per 

1201: photon, if the gluino weighs up to 10 TeV. The absence of heavy, 

1202: up to 10 TeV, isotopes of Helium, Carbon and Oxygen gives an 

1203: upper limit of about  not bigger than  $10^{-17}$ per nucleon, 

1204: or about $10^{-27}$ per photon. We now estimate the 

1205: cosmological abundance of gluinos relative to photons 

1206: before they have a chance to decay.

1207: 

1208: Before decaying, the gluinos can only reduce their number 

1209: density by annihilating with each other. They can do so as either 

1210: bare gluinos, before the QCD phase transition, or as gluinos 

1211: clothed into R-hadrons, after the QCD phase transition. 

1212: The cross section for bare gluinos is perturbative and 

1213: scales as  $\sim m_{\tilde{g}}^{-2}$. The cross section for 

1214: two R-hadrons to annihilate in the early universe is more 

1215: subtle, and is still an open question; in part because even if the 

1216: two R-hadrons combine into an ``R-molecule" bound state,

1217: this can be dissociated by collisions with the medium before 

1218: the gluinos in the molecule have a chance to annihilate each other. 

1219: Making, nevertheless, the plausible hypothesis that the cross 

1220: section for two R-hadron annihilation scales as the square of the QCD size,

1221: of order $\sigma \sim 30$ mb,  results in a

1222: gluino abundance which we estimate by equating expansion and

1223: reaction rates, 

1224: $n\, \sigma\, v \sim T^2/M_{Pl}$

1225: with $T \sim \Lambda_{QCD}$. This translates to

1226: \begin{equation}

1227: \frac {n_{\tilde{g}}}{n_{\gamma}} = 10^{-18} \left(\frac

1228: {m_{\tilde{g}}}{ 1\, \mbox{TeV}}\right)^{1/2}\, .

1229: \label{eq:abundance1}

1230: \end{equation}

1231: This is larger than the the maximal allowed abundance of $10^{-27}$ 

1232: relative to photons from the absence of anomalous heavy isotopes, 

1233: and much larger than the upper limit of $10^{-37}$ ($10^{-32}$) 

1234: per photon from the absence of anomalous isotopes of heavy 

1235: hydrogen of mass up to 1 TeV ($10$ TeV). Though the estimate 

1236: of the R-hadron annihilation cross section is uncertain, 

1237: the discrepancy is so large that we can exclude the possibility 

1238: of stable gluinos with some confidence. 

1239: 

1240: Some remaining loopholes are: 

1241: 1) The reheat temperature of the universe after inflation is 

1242: so low that no gluinos are made. 

1243: 2) Gluinos do not form any heavy nuclei, which, as mentioned before, 

1244: we find implausible. 

1245: 3) The simplest loophole: the gluino is in fact not cosmologically stable, 

1246: and lives much less than the age of the universe. This can be 

1247: accomplished most simply by a combination of a heavier gluino 

1248: and lighter $m_S$. For example, if $m_S \sim \sqrt {\delta H} \leq 10^{15}$ GeV 

1249: and the gluino mass is more than about $10$ TeV, 

1250: the lifetime drops to $\sim 10^{16}$ 

1251: years or less, which is acceptable because most gluinos will have 

1252: decayed by now and will not be around to form heavy isotopes. 

1253: Note that this requires some fine-tuning to make the ratio 

1254: $\delta H/H$ small. This tuning though is radiatively stable

1255: as was already pointed out in section \ref{susybr},

1256: since it is protected by supersymmetry, which is broken by 

1257: $\delta H$ but not $H$. Note however that for gluino masses 

1258: much heavier than $\sim 10$ TeV, the successful gauge 

1259: coupling unification will be distorted. In addition, such 

1260: heavy gluinos will not be accessible to the LHC. 

1261: A phenomenologically more appealing case is that of a 

1262: TeV-mass gluino and $m_S \sim \sqrt {\delta H} \leq 3 \times 10^{13}$ GeV. 

1263: This requires more of a tuning, which as before is radiatively stable, 

1264: and maintains both unification and accessibility of gluinos at the LHC.

1265: 

1266: Gluinos, such as those just discussed, can also be cosmologically 

1267: dangerous if their lifetime is shorter than the age of the universe 

1268: but longer than a second, and their abundance is not adequately small. 

1269: This is because their decay products can

1270: distort the photon background or destroy nuclei synthesized during

1271: primordial nucleosynthesis, which began when the universe was one

1272: second old. A gluino that decays in less than a second is

1273: harmless, as its decay products thermalize

1274: and the heat bath erases any trace of its existence.

1275: Gluinos that live longer than a second can be safe, as long as

1276: their abundance is small. This is easily satisfied as long as the 

1277: R-hadrons annihilate with a QCD-size cross section of order of 30 mb. 

1278: The relevant quantity then, more important that the plain abundance, is:

1279: \begin{equation}

1280: m_{\tilde{g}}\, \frac {n_{\tilde{g}}}{n_{\gamma}} = 10^{-15}

1281: \left(\frac {m_{\tilde{g}}}{ 1\,\mbox{TeV}}\right)^{3/2} \,

1282: \mbox{GeV}\, .

1283: \label{eq:abundance2}

1284: \end{equation}

1285: It measures the destructive power of the decaying gluino gas, 

1286: as it depends on both the mass and the concentration of gluinos.

1287: The abundance of gluinos with lifetime up to $10^{13}$ sec must be

1288: small to  avoid  spectral distortions of the CMBR~\cite{Hu:1993gc}.

1289: This constraint is mild, and equation (\ref{eq:abundance2}) easily

1290: satisfies it. The abundance of gluinos with lifetime in the range

1291: from $10^{-1}$ sec to $10^{12}$ sec must also be small to avoid

1292: the destruction of the light nuclei synthesized during the

1293: BBN~\cite{Dimopoulos:1988ue,Kawasaki:1994af}. Although this

1294: constraint is strong, especially for lifetimes between $10^{4}$

1295: sec to $10^{7}$ sec, equation (\ref{eq:abundance2}) satisfies it.

1296: Other constraints from possible distortions of the diffuse photon

1297: background  are also easily satisfied.

1298: 

1299: In summary, as long as its lifetime is much shorter than the age of the universe,

1300: and the R-hadrons annihilate with QCD-size cross sections $\sim 30$ mb, 

1301: the gluino is cosmologically safe, and does not distort either photon 

1302: backgrounds or nuclear abundances. For a TeV mass gluino this entails 

1303: a $\leq 10^{-3}$ fine-tuning that makes $\delta H$ smaller than $H$, 

1304: and is protected by supersymmetry. Heavier gluinos require less tuning, 

1305: at the expense  of distorting the successful unification 

1306: and losing the gluinos at the LHC.

1307: 

1308: What if the R-hadrons  do not annihilate with QCD-size cross sections, 

1309: and the only mechanism for the disappearance of gluinos before they 

1310: decay is standard perturbative annihilation? Then to avoid distorting the 

1311: photon spectrum or the nuclear abundances via the gluino decay products, 

1312: its lifetime must be less than a second, which implies a squark mass 

1313: $m_S \sim \sqrt {\delta H} \simlt 3 \times 10^9$ GeV, for a gluino mass 

1314: of a TeV. Again, such a small $\delta H$ will require a tuning which is 

1315: stable and protected by supersymmetry.

1316: 

1317: 

1318: \section*{Acknowledgements}

1319: 

1320: We would like to thank Fabio Zwirner for useful discussions on the 

1321: Scherk-Schwarz effective field theory. We also acknowledge valuable discussions

1322: with Nima Arkani-Hamed, Gian Giudice and Marc Tuckmantel.

1323: This work was supported in 

1324: part by the European Commission under the RTN contracts 

1325: HPRN-CT-2000-00148 and MRTN-CT-2004-503369. 

1326: SD is supported by the NSF grant 0244728.

1327: 

1328: 

1329: \setcounter{equation}{0}

1330: 

1331: 

1332: \begin{thebibliography}{}

1333: 

1334: \bibitem{weinberg}

1335: S.~Weinberg,

1336: %``Anthropic Bound On The Cosmological Constant,''

1337: Phys.\ Rev.\ Lett.\  {\bf 59}, 2607 (1987).

1338: %%CITATION = GR-QC 9406010;%%

1339: For earlier related work see T.~Banks,

1340: %``T C P, Quantum Gravity, The Cosmological Constant And All That..,''

1341: Nucl.\ Phys.\ B {\bf 249}, 332 (1985) and A.~D.~Linde, in ``300

1342: Years of Gravitation" (Editors: S.Hawking and W. Israel, Cambridge

1343: University Press, 1987), 604. This constraint was sharpened in

1344: A.~Vilenkin,

1345: %``Predictions From Quantum Cosmology,''

1346: Phys.\ Rev.\ Lett.\  {\bf 74}, 846 (1995) [arXiv:gr-qc/9406010]. A

1347: nice review of these ideas can be found in  C.~J.~Hogan,

1348: %``Why the universe is just so,''

1349: Rev.\ Mod.\ Phys.\  {\bf 72}, 1149 (2000) [arXiv:astro-ph/9909295]

1350: and M.~J.~Rees,

1351: %``Numerical coincidences and 'tuning' in cosmology,''

1352: arXiv:astro-ph/0401424.

1353: %%CITATION = ASTRO-PH 0401424;%%

1354: 

1355: \bibitem{bousso}

1356: R. Bousso and J. Polchinski, 

1357: %``Quantization of Four-Form Fluxes and

1358: %Dynamical Neutralization of the Cosmological Constant,'' 

1359: JHEP

1360: {\bf 0006} (2000) 006, hep-th/0004134;

1361: A. Maloney, E. Silverstein and A. Strominger,

1362: % ``de Sitter Space in

1363: %Noncritical String Theory,'' 

1364: hep-th/0205316;

1365: S. Kachru, R. Kallosh, A. Linde and S. Trivedi,

1366: % ``de Sitter Vacua

1367: %in String Theory,'' 

1368: Phys. Rev. {\bf D68} (2003) 046005, hep-th/0301240;

1369: L. Susskind, 

1370: %``The Anthropic Landscape of String Theory,''

1371: hep-th/0302219;

1372: M. Douglas, 

1373: %``The Statistics of String/M Theory Vacua,'' JHEP {\bf 0305}

1374: %(2003) 046, 

1375: hep-th/0303194;

1376: S. Giddings, S. Kachru and J. Polchinski, 

1377: %``Hierarchies from Fluxes

1378: %in String Compactifications,'' 

1379: Phys. Rev. {\bf D66} (2002) 106006,

1380: hep-th/0105097;

1381: S. Ashok and M. Douglas, 

1382: %``Counting Flux Vacua,''

1383:  JHEP {\bf 0401}

1384: (2004) 060, hep-th/0307049;

1385: F. Denef and M. Douglas, 

1386: %``Distributions of Flux Vacua,''

1387:  hep-th/0404116;

1388: A. Giryavets, S. Kachru and P. Tripathy, hep-th/0404243;

1389: J. Conlon and F. Quevedo, 

1390: %``On the explicit construction and statistics

1391: %of Calabi-Yau flux vacua,'' 

1392: hep-th/0409215;

1393: O. DeWolfe, A. Giryavets, S. Kachru and W. Taylor, to appear.

1394: Early arguments along these lines can be found in 

1395: %\cite{Lerche:1986cx}

1396: %\bibitem{Lerche:1986cx}

1397: W.~Lerche, D.~Lust and A.~N.~Schellekens,

1398: %``Chiral Four-Dimensional Heterotic Strings From Selfdual Lattices,''

1399: Nucl.\ Phys.\ B {\bf 287} (1987) 477.

1400: %%CITATION = NUPHA,B287,477;%%

1401: 

1402: \bibitem{susskind}

1403: L. Susskind, 

1404: hep-th/0405189;

1405: M. Douglas, hep-th/0405279;

1406: M. Dine, E. Gorbatov and S. Thomas, hep-th/0407043;

1407: E. Silverstein, hep-th/0407202;

1408: A. Linde and R. Kallosh, hep-th/0411011.

1409: 

1410: \bibitem{dg}

1411: %\cite{Dimopoulos:1981zb}

1412: %\bibitem{Dimopoulos:1981zb}

1413: S.~Dimopoulos and H.~Georgi,

1414: %``Softly Broken Supersymmetry And SU(5),''

1415: Nucl.\ Phys.\ B {\bf 193} (1981) 150.

1416: %%CITATION = NUPHA,B193,150;%% 

1417: 

1418: \bibitem{drw}

1419: %\bibitem{stuart}

1420: S.~Dimopoulos, S.~Raby and F.~Wilczek,

1421: %``Supersymmetry And The Scale Of Unification,''

1422: Phys.\ Rev.\ D {\bf 24} (1981) 1681.

1423: %%CITATION = PHRVA,D24,1681;%%

1424: 

1425: \bibitem{gold}

1426: H.~Goldberg,

1427: %``Constraint On The Photino Mass From Cosmology,''

1428: Phys.\ Rev.\ Lett.\  {\bf 50} (1983) 1419.

1429: %%CITATION = PRLTA,50,1419;%%

1430: 

1431: \bibitem{savnim}

1432: N.~Arkani-Hamed and S.~Dimopoulos,

1433: %``Supersymmetric unification without low energy supersymmetry and signatures

1434: %for fine-tuning at the LHC,''

1435: arXiv:hep-th/0405159.

1436: %%CITATION = HEP-TH 0405159;%%

1437: 

1438: \bibitem{noi}

1439: G.~F.~Giudice and A.~Romanino,

1440: %``Split supersymmetry,''

1441: arXiv:hep-ph/0406088.

1442: %%CITATION = HEP-PH 0406088;%

1443: 

1444: \bibitem{tutti}

1445: A.~Arvanitaki, C.~Davis, P.~W.~Graham and J.~G.~Wacker,

1446: %``One loop predictions of the finely tuned SSM,''

1447: arXiv:hep-ph/0406034;

1448: %%CITATION = HEP-PH 0406034;%%

1449: A.~Pierce,

1450: %``Dark matter in the finely tuned minimal supersymmetric standard model,''

1451: arXiv:hep-ph/0406144;

1452: %%CITATION = HEP-PH 0406144;%%

1453: S.~h.~Zhu,

1454: %``Chargino pair production at linear collider and split supersymmetry,''

1455: arXiv:hep-ph/0407072;

1456: %%CITATION = HEP-PH 0407072;%%

1457: B.~Mukhopadhyaya and S.~SenGupta,

1458: %``Sparticle spectrum and phenomenology in split supersymmetry: Some

1459: %possibilities,''

1460: arXiv:hep-th/0407225;

1461: %%CITATION = HEP-TH 0407225;%%

1462: W.~Kilian, T.~Plehn, P.~Richardson and E.~Schmidt,

1463: %``Split supersymmetry at colliders,''

1464: arXiv:hep-ph/0408088;

1465: %%CITATION = HEP-PH 0408088;%%

1466: R.~Mahbubani,

1467: %``Bounds on the Higgs mass in variations of Split Supersymmetry,''

1468: arXiv:hep-ph/0408096;

1469: %%CITATION = HEP-PH 0408096;%%

1470: M.~Binger,

1471: %``The Higgs boson mass at 2 loops in the finely tuned split supersymmetric

1472: %standard model,''

1473: arXiv:hep-ph/0408240;

1474: %%CITATION = HEP-PH 0408240;%%

1475: J.~L.~Hewett, B.~Lillie, M.~Masip and T.~G.~Rizzo,

1476: %``Signatures of long-lived gluinos in split supersymmetry,''

1477: arXiv:hep-ph/0408248;

1478: %%CITATION = HEP-PH 0408248;%%

1479: L.~Anchordoqui, H.~Goldberg and C.~Nunez,

1480: %``Probing split supersymmetry with cosmic rays,''

1481: arXiv:hep-ph/0408284.

1482: %%CITATION = HEP-PH 0408284;%%

1483: 

1484: %\cite{Arkani-Hamed:2004yi}

1485: \bibitem{Arkani-Hamed:2004yi}

1486: N.~Arkani-Hamed, S.~Dimopoulos, G.~F.~Giudice and A.~Romanino,

1487: %``Aspects of split supersymmetry,''

1488: arXiv:hep-ph/0409232.

1489: %%CITATION = HEP-PH 0409232;%% 

1490: 

1491: %\cite{Arkani-Hamed:1998rs}

1492: \bibitem{Arkani-Hamed:1998rs}

1493: N.~Arkani-Hamed, S.~Dimopoulos and G.~R.~Dvali,

1494: %``The hierarchy problem and new dimensions at a millimeter,''

1495: Phys.\ Lett.\ B {\bf 429} (1998) 263

1496: [arXiv:hep-ph/9803315].

1497: %%CITATION = HEP-PH 9803315;%%

1498: 

1499: %\cite{Antoniadis:1998ig}

1500: \bibitem{Antoniadis:1998ig}

1501: I.~Antoniadis, N.~Arkani-Hamed, S.~Dimopoulos and G.~R.~Dvali,

1502: %``New dimensions at a millimeter to a Fermi and superstrings at a TeV,''

1503: Phys.\ Lett.\ B {\bf 436} (1998) 257

1504: [arXiv:hep-ph/9804398].

1505: %%CITATION = HEP-PH 9804398;%%

1506: 

1507: %\cite{Uranga:2003pz}

1508: \bibitem{Uranga:2003pz}

1509: For recent reviews, see e.g.

1510: A.~M.~Uranga,

1511: %``Chiral four-dimensional string compactifications with intersecting

1512: %D-branes,''

1513: Class.\ Quant.\ Grav.\  {\bf 20} (2003) S373

1514: [arXiv:hep-th/0301032];

1515: D.~Cremades, L.~E.~Ibanez and F.~Marchesano,

1516: %``More about the standard model at intersecting branes,''

1517: arXiv:hep-ph/0212048;

1518: %\cite{Lust:2004ks}

1519: D.~Lust,

1520: %``Intersecting brane worlds: A path to the standard model?,''

1521: Class.\ Quant.\ Grav.\  {\bf 21} (2004) S1399

1522: [arXiv:hep-th/0401156];

1523: and references therein.

1524: %%CITATION = HEP-TH 0401156;%%

1525: 

1526: \bibitem{Bachas:1995ik}

1527: C.~Bachas,

1528: %``A Way to break supersymmetry,''

1529: arXiv:hep-th/9503030.

1530: %%CITATION = HEP-TH 9503030;%%

1531: 

1532: %\cite{Angelantonj:2000hi}

1533: \bibitem{Angelantonj:2000hi}

1534: C.~Angelantonj, I.~Antoniadis, E.~Dudas and A.~Sagnotti,

1535: %``Type-I strings on magnetised orbifolds and brane transmutation,''

1536: Phys.\ Lett.\ B {\bf 489} (2000) 223

1537: [arXiv:hep-th/0007090].

1538: %%CITATION = HEP-TH 0007090;%%

1539: 

1540: \bibitem{Berkooz:1996km}

1541: M.~Berkooz, M.~R.~Douglas and R.~G.~Leigh,

1542: %``Branes intersecting at angles,''

1543: Nucl.\ Phys.\ B {\bf 480} (1996) 265

1544: [arXiv:hep-th/9606139];

1545: %%CITATION = HEP-TH 9606139;%%

1546: 

1547: \bibitem{bi}

1548: R.~Blumenhagen, L.~Goerlich, B.~Kors and D.~Lust,

1549: %``Noncommutative compactifications of type I strings on tori with

1550: %magnetic background flux,''

1551: JHEP {\bf 0010} (2000) 006

1552: [arXiv:hep-th/0007024];

1553: %%CITATION = HEP-TH 0007024;%%

1554: %\cite{Aldazabal:2000dg}

1555: %\bibitem{Aldazabal:2000dg}

1556: G.~Aldazabal, S.~Franco, L.~E.~Ibanez, R.~Rabadan and A.~M.~Uranga,

1557: %``D = 4 chiral string compactifications from intersecting branes,''

1558: J.\ Math.\ Phys.\  {\bf 42} (2001) 3103

1559: [arXiv:hep-th/0011073].

1560: %%CITATION = HEP-TH 0011073;%%

1561: 

1562: \bibitem{ah}

1563: A.~Hebecker, private communication.

1564: 

1565: %\cite{Blumenhagen:2003jy}

1566: \bibitem{Blumenhagen:2003jy}

1567: R.~Blumenhagen, D.~Lust and S.~Stieberger,

1568: %``Gauge unification in supersymmetric intersecting brane worlds,''

1569: JHEP {\bf 0307} (2003) 036

1570: [arXiv:hep-th/0305146].

1571: %%CITATION = HEP-TH 0305146;%%

1572: 

1573: \bibitem{SS}

1574: J.~Scherk and J.~H.~Schwarz,

1575: %``Spontaneous Breaking Of Supersymmetry Through Dimensional Reduction,''

1576: Phys.\ Lett.\ B {\bf 82} (1979) 60.

1577: 

1578: %\cite{Antoniadis:1996hk}

1579: \bibitem{Antoniadis:1996hk}

1580: I.~Antoniadis and M.~Quiros,

1581: %``Large radii and string unification,''

1582: Phys.\ Lett.\ B {\bf 392} (1997) 61

1583: [arXiv:hep-th/9609209];

1584: %%CITATION = HEP-TH 9609209;%%

1585: %\cite{Antoniadis:1997xk}

1586: %\bibitem{Antoniadis:1997xk}

1587: %I.~Antoniadis and M.~Quiros,

1588: %``On the M-theory description of gaugino condensation,''

1589: Phys.\ Lett.\ B {\bf 416} (1998) 327

1590: [arXiv:hep-th/9707208];

1591: %%CITATION = HEP-TH 9707208;%%

1592: %\cite{Dudas:1997jn}

1593: %\bibitem{Dudas:1997jn}

1594: E.~Dudas and C.~Grojean,

1595: %``Four-dimensional M-theory and supersymmetry breaking,''

1596: Nucl.\ Phys.\ B {\bf 507} (1997) 553

1597: [arXiv:hep-th/9704177].

1598: %%CITATION = HEP-TH 9704177;%%

1599: 

1600: %\cite{Horava:1996ma}

1601: \bibitem{Horava:1996ma}

1602: P.~Horava and E.~Witten,

1603: %``Eleven-Dimensional Supergravity on a Manifold with Boundary,''

1604: Nucl.\ Phys.\ B {\bf 475}, 94 (1996)

1605: [arXiv:hep-th/9603142];

1606: %%CITATION = HEP-TH 9603142;%%

1607: %\cite{Horava:1996vs}

1608: %\bibitem{Horava:1996vs}

1609: P.~Horava,

1610: %``Gluino condensation in strongly coupled heterotic string theory,''

1611: Phys.\ Rev.\ D {\bf 54} (1996) 7561

1612: [arXiv:hep-th/9608019].

1613: %%CITATION = HEP-TH 9608019;%%

1614: 

1615: %\cite{Antoniadis:1998ki}

1616: \bibitem{Antoniadis:1998ki}

1617: I.~Antoniadis, E.~Dudas and A.~Sagnotti,

1618: %``Supersymmetry breaking, open strings and M-theory,''

1619: Nucl.\ Phys.\ B {\bf 544} (1999) 469

1620: [arXiv:hep-th/9807011].

1621: %%CITATION = HEP-TH 9807011;%%

1622: 

1623: %\cite{Antoniadis:1990ew}

1624: \bibitem{Antoniadis:1990ew}

1625: I.~Antoniadis,

1626: %``A Possible New Dimension At A Few Tev,''

1627: Phys.\ Lett.\ B {\bf 246} (1990) 377;

1628: %%CITATION = PHLTA,B246,377;%%

1629: %\cite{Antoniadis:1997zg}

1630: %\bibitem{Antoniadis:1997zg}

1631: I.~Antoniadis, S.~Dimopoulos and G.~R.~Dvali,

1632: %``Millimeter range forces in superstring theories with weak-scale

1633: %compactification,''

1634: Nucl.\ Phys.\ B {\bf 516} (1998) 70

1635: [arXiv:hep-ph/9710204];

1636: %%CITATION = HEP-PH 9710204;%%

1637: %\cite{Antoniadis:1998sd}

1638: %\bibitem{Antoniadis:1998sd}

1639: I.~Antoniadis, S.~Dimopoulos, A.~Pomarol and M.~Quiros,

1640: %``Soft masses in theories with supersymmetry breaking by

1641: %TeV-compactification,''

1642: Nucl.\ Phys.\ B {\bf 544} (1999) 503

1643: [arXiv:hep-ph/9810410].

1644: %%CITATION = HEP-PH 9810410;%%

1645: 

1646: %\cite{Antoniadis:1997ic}

1647: \bibitem{Antoniadis:1997ic}

1648: I.~Antoniadis and M.~Quiros,

1649: %``Supersymmetry breaking in M-theory and gaugino condensation,''

1650: Nucl.\ Phys.\ B {\bf 505} (1997) 109

1651: [arXiv:hep-th/9705037];

1652: %%CITATION = HEP-TH 9705037;%%

1653: %\bibitem{Luty:2002hj}

1654: M.~A.~Luty and N.~Okada,

1655: %``Almost no-scale supergravity,''

1656: JHEP {\bf 0304} (2003) 050

1657: [arXiv:hep-th/0209178];

1658: %%CITATION = HEP-TH 0209178;%%

1659: %\cite{Rattazzi:2003rj}

1660: %\bibitem{Rattazzi:2003rj}

1661: R.~Rattazzi, C.~A.~Scrucca and A.~Strumia,

1662: %``Brane to brane gravity mediation of supersymmetry breaking,''

1663: Nucl.\ Phys.\ B {\bf 674} (2003) 171

1664: [arXiv:hep-th/0305184].

1665: %%CITATION = HEP-TH 0305184;%%

1666: 

1667: %\cite{Randall:1998uk}

1668: \bibitem{Randall:1998uk}

1669: L.~Randall and R.~Sundrum,

1670: %``Out of this world supersymmetry breaking,''

1671: Nucl.\ Phys.\ B {\bf 557} (1999) 79

1672: [arXiv:hep-th/9810155];

1673: %%CITATION = HEP-TH 9810155;%%

1674: G.~F.~Giudice, M.~A.~Luty, H.~Murayama and R.~Rattazzi,

1675: %``Gaugino mass without singlets,''

1676: JHEP {\bf 9812} (1998) 027

1677: [arXiv:hep-ph/9810442].

1678: %%CITATION = HEP-PH 9810442;%%

1679: 

1680: %\cite{Antoniadis:2004qn}

1681: \bibitem{Antoniadis:2004qn}

1682: I.~Antoniadis and T.~R.~Taylor,

1683: %``Topological masses from broken supersymmetry,''

1684: Nucl.\ Phys.\ B {\bf 695} (2004) 103

1685: [arXiv:hep-th/0403293].

1686: %%CITATION = HEP-TH 0403293;%%

1687: 

1688: \bibitem{at}

1689: I.~Antoniadis and T.~R.~Taylor, in preparation.

1690: 

1691: %\cite{Gherghetta:2001sa}

1692: \bibitem{Gherghetta:2001sa}

1693: T.~Gherghetta and A.~Riotto,

1694: %``Gravity-mediated supersymmetry breaking in the brane-world,''

1695: Nucl.\ Phys.\ B {\bf 623} (2002) 97

1696: [arXiv:hep-th/0110022].

1697: %%CITATION = HEP-TH 0110022;%%

1698: 

1699: %\cite{Bagger:2001ep}

1700: \bibitem{Bagger:2001ep}

1701: J.~Bagger, F.~Feruglio and F.~Zwirner,

1702: %``Brane induced supersymmetry breaking,''

1703: JHEP {\bf 0202} (2002) 010

1704: [arXiv:hep-th/0108010];

1705: %%CITATION = HEP-TH 0108010;%%

1706: %\cite{Biggio:2002rb}

1707: %\bibitem{Biggio:2002rb}

1708: C.~Biggio, F.~Feruglio, A.~Wulzer and F.~Zwirner,

1709: %``Equivalent effective Lagrangians for Scherk-Schwarz compactifications,''

1710: JHEP {\bf 0211} (2002) 013

1711: [arXiv:hep-th/0209046].

1712: %%CITATION = HEP-TH 0209046;%%

1713: 

1714: %\cite{Antoniadis:2000en}

1715: \bibitem{Antoniadis:2000en}

1716: I.~Antoniadis, E.~Kiritsis and T.~N.~Tomaras,

1717: %``A D-brane alternative to unification,''

1718: Phys.\ Lett.\ B {\bf 486} (2000) 186

1719: [arXiv:hep-ph/0004214];

1720: %%CITATION = HEP-PH 0004214;%%

1721: %\cite{Antoniadis:2002qm}

1722: %\bibitem{Antoniadis:2002qm}

1723: I.~Antoniadis, E.~Kiritsis, J.~Rizos and T.~N.~Tomaras,

1724: %``D-branes and the standard model,''

1725: Nucl.\ Phys.\ B {\bf 660} (2003) 81

1726: [arXiv:hep-th/0210263].

1727: %%CITATION = HEP-TH 0210263;%%

1728: 

1729: \bibitem{ar}

1730: I.~Antoniadis and J.~Rizos, 2003 unpublished work.

1731: 

1732: %\cite{Sagnotti:1992qw}

1733: \bibitem{Sagnotti:1992qw}

1734: A.~Sagnotti,

1735: %``A Note on the Green-Schwarz mechanism in open string theories,''

1736: Phys.\ Lett.\ B {\bf 294} (1992) 196

1737: [arXiv:hep-th/9210127];

1738: %%CITATION = HEP-TH 9210127;%%

1739: %\cite{Ibanez:1998qp}

1740: %\bibitem{Ibanez:1998qp}

1741: L.~E.~Ibanez, R.~Rabadan and A.~M.~Uranga,

1742: %``Anomalous U(1)'s in type I and type IIB D = 4, N = 1 string vacua,''

1743: Nucl.\ Phys.\ B {\bf 542} (1999) 112

1744: [arXiv:hep-th/9808139].

1745: %%CITATION = HEP-TH 9808139;%%

1746: 

1747: %\cite{Poppitz:1998dj}

1748: \bibitem{Poppitz:1998dj}

1749: E.~Poppitz,

1750: %``On the one loop Fayet-Iliopoulos term in chiral four dimensional type I

1751: %orbifolds,''

1752: Nucl.\ Phys.\ B {\bf 542} (1999) 31

1753: [arXiv:hep-th/9810010].

1754: %%CITATION = HEP-TH 9810010;%%

1755:  

1756:  \bibitem{Smith:1982qu}

1757: P.~F.~Smith, J.~R.~J.~Bennett, G.~J.~Homer, J.~D.~Lewin, H.~E.~Walford and W.~A.~Smith,

1758: %``A Search For Anomalous Hydrogen In Enriched D-2 O, Using A Time-Of-Flight

1759: %Spectrometer,''

1760: Nucl.\ Phys.\ B {\bf 206} (1982) 333.

1761: %%CITATION = NUPHA,B206,333;%%

1762: 

1763: %\cite{Hu:1993gc}

1764: \bibitem{Hu:1993gc}

1765: W.~Hu and J.~Silk,

1766: %``Thermalization constraints and spectral distortions for massive unstable

1767: %relic particles,''

1768: Phys.\ Rev.\ Lett.\  {\bf 70} (1993) 2661.

1769: %%CITATION = PRLTA,70,2661;%

1770: 

1771: %\cite{Dimopoulos:1988ue}

1772: \bibitem{Dimopoulos:1988ue}

1773: S.~Dimopoulos, R.~Esmailzadeh, L.~J.~Hall and G.~D.~Starkman,

1774: %``Limits On Late Decaying Particles From Nucleosynthesis,''

1775: Nucl.\ Phys.\ B {\bf 311} (1989) 699.

1776: %%CITATION = NUPHA,B311,699;%%

1777: 

1778: %\cite{Kawasaki:1994af}

1779: \bibitem{Kawasaki:1994af}

1780: M.~Kawasaki and T.~Moroi,

1781: %``Gravitino production in the inflationary universe and the effects on big

1782: %bang nucleosynthesis,''

1783: Prog.\ Theor.\ Phys.\  {\bf 93} (1995) 879

1784: [arXiv:hep-ph/9403364].

1785: %%CITATION = HEP-PH 9403364;%% 

1786: 

1787: \end{thebibliography}

1788: 

1789: \end{document}

1790: 

1791: