1: \NeedsTeXFormat{LaTeX2e}
2: \documentclass[12pt,a4paper,fleqn]{article}
3:
4: \usepackage[DIV12]{typearea}
5: \usepackage{amsmath,amsfonts,amssymb,mathrsfs}
6: \usepackage{graphicx}
7: \usepackage[footnotesize]{caption}
8: \usepackage{cite}
9: \usepackage{subfigure}
10: \usepackage{color}
11:
12:
13:
14: \newcommand{\sfer}[1][]{{\widetilde f_{#1}}}
15: %%%%%%Squarks
16: \newcommand{\sqk}[1][]{{\widetilde q_{#1}}}
17: \newcommand{\squ}[1][]{{\widetilde u_{#1}}}
18: \newcommand{\sqd}[1][]{{\widetilde d_{#1}}}
19: \newcommand{\sqc}[1][]{{\widetilde c_{#1}}}
20: \newcommand{\sqs}[1][]{{\widetilde s_{#1}}}
21: \newcommand{\sqt}[1][]{{\widetilde t_{#1}}}
22: \newcommand{\sqb}[1][]{{\widetilde b_{#1}}}
23: %%%%%%Sleptons
24: \newcommand{\slep}[1][]{{\widetilde l_{#1}}}
25: \newcommand{\sle}[1][]{{\widetilde e_{#1}}}
26: \newcommand{\slmu}[1][]{{\widetilde\mu_{#1}}}
27: \newcommand{\sltau}[1][]{{\widetilde\tau_{#1}}}
28: \newcommand{\slneu}[1][]{{\widetilde\nu_{#1}}}
29:
30: \newcommand{\tbeta}{{\tan\beta}}
31:
32: %%%%%%%%%%%Higgs sector
33: %%%%%%Higgs
34: \newcommand{\Hu}[1][]{{H_u^{#1}}}
35: \newcommand{\Hd}[1][]{{H_d^{#1}}}
36: %%%%%%Higgsinos
37: \newcommand{\Ho}[1][]{{\widetilde H^{#1}}}
38: \newcommand{\Hou}[1][]{{\widetilde H_u^{#1}}}
39: \newcommand{\Hod}[1][]{{\widetilde H_d^{#1}}}
40:
41: %%%%%%%%%%%Gauginos
42: %%%%%%Gluinos
43: \newcommand{\go}{{\widetilde g}}
44: %%%%%%Winos
45: \newcommand{\Wo}{{\widetilde W}}
46: %\newcommand{\Wo+}{{\widetilde W^+}}
47: %\newcommand{\Wo-}{{\widetilde W^-}}
48: %%%%%%Zinos
49: \newcommand{\Zo}{{\widetilde Z}}
50: %\newcommand{\Zo0}{{\widetilde Z^0}}
51: %%%%%%Photino
52: \newcommand{\pho}{{\widetilde\gamma}}
53: %%%%%%%%%%%Neutralinos/Charginos
54: \newcommand{\No}[1][]{{\chi^0_{#1}}}
55: \newcommand{\Co}[1][]{{\chi^\pm_{#1}}}
56: \newcommand{\Cop}[1][]{{\chi^+_{#1}}}
57: \newcommand{\Com}[1][]{{\chi^-_{#1}}}
58:
59:
60: \newcommand{\gra}{\ensuremath{\widetilde{G}}}
61: \newcommand{\bin}{\ensuremath{\widetilde{B}}}
62: \newcommand{\slel}{\ensuremath{\widetilde{\ell}}}
63: \newcommand{\brac}{\ensuremath{\phantom{\frac{1}{1}}\!\!\!\!}}
64: \usepackage{feynmp}
65:
66:
67: \addtolength{\voffset}{12pt}
68: % Quick and dirty way to increase the upper margin. Ratio of upper and
69: % lower margin is now 2:3.
70:
71: \newcommand{\stau}{{\widetilde\tau}}
72: \newcommand{\snu}{{\widetilde\nu}}
73: \newcommand{\sel}{{\widetilde e}}
74: \newcommand{\slepton}{{\widetilde\ell}}
75: \newcommand{\gravitino}{\widetilde{G}}
76: \newcommand{\squark}{\ensuremath{\widetilde{Q}}}
77:
78: \newcommand{\Eqref}[1]{Eq.~\eqref{#1}}
79: \newcommand{\Figref}[1]{figure~\ref{#1}}
80: \newcommand{\Tabref}[1]{Tab.~\ref{#1}}
81:
82: \DeclareMathOperator{\im}{Im}
83: \DeclareMathOperator{\re}{Re}
84: \DeclareMathOperator{\tr}{Tr}
85: \DeclareMathOperator{\diag}{diag}
86:
87: \newcommand{\eVdist}{\kern-0.06em}
88: \newcommand{\Ev}{\text{e\eVdist V}} % solely as unit
89: \newcommand{\Kev}{\text{ke\eVdist V}}
90: \newcommand{\Mev}{\text{Me\eVdist V}}
91: \newcommand{\Gev}{\text{Ge\eVdist V}}
92: \newcommand{\Tev}{\text{Te\eVdist V}}
93: \newcommand{\ev}{\:\text{e\eVdist V}} % along with a number
94: \newcommand{\kev}{\:\text{ke\eVdist V}}
95: \newcommand{\mev}{\:\text{Me\eVdist V}}
96: \newcommand{\gev}{\:\text{Ge\eVdist V}}
97: \newcommand{\tev}{\:\text{Te\eVdist V}}
98: \newcommand{\s}{\:\text{s}}
99:
100: \hyphenation{FCNCs}
101: \hyphenation{gau-gi-no}
102: \hyphenation{im-port-ant}
103: \hyphenation{coup-lings}
104: \hyphenation{or-bi-fold}
105:
106:
107: \DeclareMathOperator{\ad}{ad}
108: \DeclareMathOperator{\quabla}{\boldsymbol{\square}}
109: \newcommand{\CenterObject}[1]{\ensuremath{\vcenter{\hbox{#1}}}}
110: \newcommand{\CenterEps}[2][1]{\ensuremath{\vcenter{\hbox{\includegraphics[scale=#1]{#2.eps}}}}}
111: \newcommand{\D}{\mathrm{d}}
112: \newcommand{\I}{\mathrm{i}}
113: \newcommand{\E}[1]{\ensuremath{\mathrm{E}_{#1}}} % e.g. \E{8}
114: \newcommand{\SO}[1]{\ensuremath{\mathrm{SO}(#1)}}
115: \newcommand{\SU}[1]{\ensuremath{\mathrm{SU}(#1)}}
116: \newcommand{\U}[1]{\ensuremath{\mathrm{U}(#1)}}
117: \newcommand{\Z}[1]{\ensuremath{\mathbbm{Z}_{#1}}} % Z_N ->\Z{N}
118:
119: \allowdisplaybreaks
120:
121: \unitlength=1mm
122: \allowdisplaybreaks[1]
123:
124: % ======================================================================
125:
126: \begin{document}
127:
128: \begin{titlepage}
129:
130: \begin{flushright}
131: TUM-HEP 695/08
132: \end{flushright}
133:
134: \vspace*{1.0cm}
135:
136: \begin{center}
137: {\Large\bf
138: A note on the primordial abundance of stau NLSPs
139: }
140:
141: \vspace{1cm}
142:
143: \textbf{
144: Michael Ratz\footnote[1]{Email: \texttt{mratz@ph.tum.de}},
145: Kai Schmidt-Hoberg\footnote[2]{Email: \texttt{kschmidt@ph.tum.de}}
146: and Martin Wolfgang Winkler\footnote[3]{Email: \texttt{mwinkler@ph.tum.de}}
147: }
148: \\[5mm]
149:
150:
151: Physik-Department T30, Technische Universit\"at M\"unchen, \\
152: James-Franck-Stra\ss e, 85748 Garching, Germany
153:
154: \end{center}
155:
156: \vspace{1cm}
157:
158: \begin{abstract}
159: \noindent
160: In scenarios with a gravitino LSP, there exist strong BBN constraints on the
161: abundance of a possible stau NLSP. We find that in settings with substantial
162: left-right mixing of the stau mass eigenstates these constraints can be evaded
163: even for very long-lived staus.
164: \end{abstract}
165:
166: \end{titlepage}
167:
168: \newpage
169:
170:
171:
172: \section{Introduction}
173:
174: Arguably, supersymmetry (SUSY) is one of the most plausible extensions of the
175: standard model (SM). Apart from its theoretical appeal, SUSY has the virtue of
176: providing a compelling dark matter candidate, the lightest supersymmetric
177: particle (LSP), which is stable if $R$ parity is conserved. A particular
178: interesting dark matter candidate is the gravitino, which evades direct
179: detection because all its interactions are suppressed by the Planck scale. The
180: hypothesis of a gravitino LSP may nevertheless be tested at the LHC if certain
181: conditions are met. First, the next-to-lightest
182: supersymmetric particle (NLSP) has to be charged, and second the gravitino mass
183: may not be too small. In this case, one could observe long-lived charged
184: particles in whose decays one could probe the properties of the LSP
185: \cite{Buchmuller:2004rq,Buchmuller:2004tm,Feng:2004mt,Brandenburg:2005he}.
186: The collider phenomenology of such
187: scenarios has been explored in various studies
188: \cite{Buchmuller:2004rq,Buchmuller:2004tm,Feng:2004mt,Hamaguchi:2004df,Feng:2004yi,Brandenburg:2005he,Martyn:2006as,%
189: Ellis:2006vu,Kitano:2008sa}.
190:
191: There are several theoretical reasons which make it appear desirable to have
192: gravitino masses $m_{3/2}$ not
193: much smaller than the masses of the SM superpartners.
194: For instance, simple explanations of the $\mu$ and $B\,\mu$ terms
195: seem to require not too small $m_{3/2}$ \cite{Kim:1983dt,Giudice:1988yz}.
196: Further, many simple mechanisms of baryogenesis, in particular leptogenesis
197: \cite{Fukugita:1986hr},
198: need rather high reheating temperatures $T_R$ \cite{Davidson:2002qv} which can be achieved
199: for gravitino masses of about $10\dots100\,$GeV
200: \cite{Bolz:2000fu,Pradler:2006hh} (i.e.\ at least the constraints from gravitino
201: overproduction can be satisfied).
202:
203:
204: However, there are severe constraints on such scenarios coming from cosmology.
205: The observed primordial abundances of light elements produced in big bang
206: nucleosynthesis (BBN) allow to place stringent constraints on the number density
207: of long-lived particles whose decays happen during or after BBN and induce
208: nuclear reactions that change the element abundances
209: \cite{Falomkin:1984eu,Khlopov:1984pf,Ellis:1984eq}. While a neutralino NLSP is
210: strongly disfavored for gravitino masses in the \gev\
211: range~\cite{Feng:2004mt,Cerdeno:2005eu}, scenarios with a sneutrino NLSP are
212: essentially unconstrained but very hard to test experimentally
213: \cite{Kanzaki:2006hm,Covi:2007xj}. This makes a charged slepton, specifically a
214: stau, particularly
215: appealing as an NLSP candidate. The stau NLSP abundance and lifetime can
216: satisfy the limits obtained from BBN by considering NLSP decays alone
217: [\citen{Asaka:2000zh},\citen{Fujii:2003nr,Feng:2004mt,Ellis:2003dn,Cerdeno:2005eu,Jedamzik:2005dh,%
218: Steffen:2006hw,Buchmuller:2006nx,Cyburt:2006uv,Pradler:2006hh}]. However, as
219: pointed out in \cite{Pospelov:2006sc}, charged NLSPs form bound states with
220: light nuclei, which leads to a drastic overproduction of $^6$Li. This process, known as
221: Catalyzed BBN (CBBN),
222: leads to strong constraints on the stau relic abundance, unless the NLSP lifetime
223: is shorter than a few thousand seconds.
224:
225:
226: Several ways to circumvent BBN constraints have been discussed in the
227: literature. For instance, entropy production between NLSP freeze-out and the
228: start of BBN can dilute the NLSP abundance sufficiently to satisfy all
229: constraints even for long lifetimes
230: \cite{Buchmuller:2006tt,Pradler:2006hh,Hamaguchi:2007mp}. However,
231: in order to arrive at such scenarios one usually relies on new sectors which are
232: typically very hard to access experimentally. Alternatively, the NLSP can be
233: sufficiently short lived if the gravitino is very light, $R$ parity is slightly broken
234: \cite{Takayama:2000uz,Buchmuller:2007ui} or the superpartner mass spectrum is
235: sufficiently heavy \cite{Kersten:2007ab}. However in these cases it is practically
236: impossible to test the nature of the LSP.
237:
238:
239: The purpose of this study is to point out that there are regions within the
240: parameter space of the minimal supersymmetric extension of the standard model
241: (MSSM) where the relic stau abundance is strongly suppressed such that the
242: bounds from CBBN can be evaded even for long stau lifetimes. As
243: we shall see, small thermal relic abundances of staus occur in parameter regions
244: with a substantial left-right mixing of the stau mass eigenstates, where the
245: annihilation into Higgs bosons is greatly enhanced.
246:
247: The paper is organized as follows. In the next section we will introduce the stau-Higgs coupling
248: and calculate the Higgs channel cross section.
249: Section~\ref{sec:muconstraints} is devoted to a discussion
250: of theoretical constraints on trilinear couplings between Higgs and $\stau$
251: fields.
252: In Section~\ref{primordial} we review the relevant BBN constraints and discuss
253: the stau relic abundance. Continuing with Section~\ref{sec:scenarios} we
254: introduce three scenarios within the MSSM in which a strong suppression of the
255: stau relic abundance can be achieved such that all cosmological constraints can
256: be evaded. Finally in Section~\ref{sec:Prospects} we briefly discuss the
257: implications of our scenario for the LHC.
258:
259:
260:
261: \section{Annihilation into Higgs bosons}
262:
263: In the early universe, superpartners are copiously produced; usually they are
264: assumed to be in thermal equilibrium. As the universe cools down, they will
265: cascade into staus, which we assume to be the NLSPs. Since staus are metastable,
266: until the BBN era their abundance will only decrease due to annihilation.
267: %
268: In most analyses performed so far, the lightest stau is assumed to be purely
269: right-handed. Then, for its freeze-out, only electroweak annihilation processes have to
270: be considered. The couplings governing the relevant reactions are either the
271: electric charge $e$ or the $\U1_Y$-coupling $g_Y=e / \cos{\theta_\mathrm{W}}$,
272: where $\theta_\mathrm{W}$ denotes the Weinberg angle. These couplings are rather
273: small, leading to a relatively large stau abundance after freeze-out
274: \cite{Asaka:2000zh},
275: \begin{equation}
276: Y_{\stau_\mathrm{R}}~\gtrsim~10^{-13} \quad \text{for} \quad {m_{\stau_1}} \gtrsim 100\gev\;,
277: \end{equation}
278: where the abundance $Y\equiv n/s$ is defined as the ratio of number and entropy densities.
279: Such a large relic stau abundance is allowed by CBBN only if the gravitino is
280: very light and the stau lifetime accordingly short. If the lighter stau has a
281: left-handed component, the electroweak annihilation cross section gets enhanced
282: due to its $\mathrm{SU}(2)_\mathrm{L}$ couplings, but as CBBN bounds are very
283: tight, the inclusion of further gauge interactions changes the situation only
284: marginally. On the other hand, we shall see that in the case of substantial
285: left-right mixing in the stau sector, the couplings between staus and Higgs
286: bosons can get significantly enhanced, thus greatly suppressing the stau relic
287: abundance.
288:
289: \subsection{Coupling of staus to Higgs bosons}
290:
291: To find the regions of parameter space where this annihilation reaction is
292: important, we now turn to the Lagrangean term which describes the couplings
293: between the light stau and the light Higgs. In our analytic discussion, we
294: make a couple of simplifying assumptions; later, in Section~\ref{sec:scenarios},
295: we will take into account all interactions and states.
296:
297: We shall assume that there is no generation mixing in the slepton sector which
298: is suggested by flavor constraints. Furthermore we take $\mu$ and $A^\tau$ to be
299: real parameters. Then the relevant terms read\footnote{There exist different
300: sign conventions for the A-parameter. Here we follow \cite{Drees:2004jm}.}
301: \begin{eqnarray}
302: \mathscr{L}_{\sltau[1] \sltau[1] h}&=&
303: \frac{g_2}{2M_W}
304: \left\lbrace \brac M_W^2\, \sin{(\alpha +\beta)}\,
305: \left[ \left( \tan^2{\theta_\mathrm{W}}-1\right) \,
306: \cos^2{\theta_{\sltau}} -2 \tan^2{\theta_\mathrm{W}}\,
307: \sin^2{\theta_{\sltau}} \right] \right. \nonumber\\
308: &&\left.\phantom{\frac{g_2}{2M_W}\lbrace }
309: {}+m_\tau\, \frac{\mu\, \cos\alpha- A^\tau\, \sin\alpha}{\cos\beta}
310: \sin{2\theta_{\sltau}}
311: + 2m_\tau^2\, \frac{\sin\alpha}{\cos\beta} \right\rbrace \; h
312: \,\sltau[1]^+\sltau[1]^- \;.
313: \end{eqnarray}
314: For simplicity we assume that the Higgs bosons except $h$ be relatively heavy
315: ($\gtrsim 300 \gev$), which is the case for all models we are considering later.
316: This allows us to work in the `decoupling limit' where the mixing parameter
317: $\alpha$ can be written as $\alpha \simeq \beta-\pi/2$ and therefore $\cos\alpha
318: \simeq \sin\beta$ and $\sin\alpha \simeq -\cos\beta$. The leading term of the
319: Lagrangean which couples the lightest stau to the Higgs is then given by
320: \begin{eqnarray}
321: \mathscr{L}_{\sltau[1] \sltau[1] h}&=&
322: \frac{g_2}{2M_W}\,
323: \sin{2\theta_{\sltau}}\; m_\tau\, \left\lbrace \mu\, \tan\beta+ A^\tau
324: \right\rbrace \; h \,\sltau[1]^+\sltau[1]^-
325: \nonumber \\
326: &=&
327: -\frac{g_2}{2M_W}\,
328: \sin{2\theta_{\sltau}} \; m_{\stau_{\mathrm{LR}}}^2\; h
329: \,\sltau[1]^+\sltau[1]^-\;.\label{eq:StauHiggsCoupling}
330: \end{eqnarray}
331: Here $m_{\stau_{\mathrm{LR}}}^2$ denotes the off-diagonal
332: element of the $2\times2$ stau mass matrix (cf.\
333: Appendix~\ref{app:StauMassMatrix}).
334:
335:
336: \subsection{Higgs channel cross section}
337:
338: With the stau-Higgs coupling \eqref{eq:StauHiggsCoupling}, we can now calculate
339: the cross section for the annihilation of the light staus into light Higgs
340: bosons,
341: \begin{equation}\label{eq:StauHiggsAnnihilation}
342: \stau_1^{+}+ \stau_1^-~\rightarrow ~h+h\;.
343: \end{equation}
344: The contributing Feynman diagrams are shown in
345: Figure~\ref{fig:StauHiggsAnnihilation}.
346: \begin{figure}[h]
347: \centerline{\subfigure[{}]{\CenterEps{Gravitino36}}\quad\quad\subfigure[{}]{\CenterEps{Gravitino37}}}
348: \caption{Stau annihilation into Higgs bosons.}
349: \label{fig:StauHiggsAnnihilation}
350: \end{figure}
351: In our approximation, we consider only the exchange of $\stau_1$, neglecting the
352: exchange of $\stau_2$.
353: In zeroth order of a velocity expansion, the thermally averaged annihilation cross section
354: $\langle\sigma_\mathrm{ann}\,v\rangle$ is equal to the cross section $\sigma_\mathrm{ann}$
355: times the relative velocity $v_\mathrm{rel}$ of the incoming staus. We obtain
356: \begin{equation}\label{eq:ThermalCrossSection}
357: \langle\sigma_\mathrm{ann}\,v\rangle~\simeq~\sigma_\mathrm{ann}\,v_\mathrm{rel}
358: ~=~
359: \frac{1}{16 \pi}\left( \frac{g_2}{2M_W}\,
360: \sin{2\theta_{\sltau}}\; m_{\stau_{\mathrm{LR}}}^2\right)^4
361: \frac{\sqrt{m_{\stau_1}^2-m_h^2}}{m_{\stau_1}^3(2 m_{\stau_1}^2-m_h^2)^2}
362: \;.
363: \end{equation}
364: Clearly, this annihilation cross section becomes important for sizable left-right mixing
365: and relatively small stau masses.
366:
367: \subsection{Comparison with electromagnetic cross section}
368:
369:
370: It is instructive to compare the cross section \eqref{eq:ThermalCrossSection}
371: with a typical electroweak cross section. For example the annihilation cross
372: section of staus into photons is given by \cite{Asaka:2000zh} $\langle
373: {\sigma_\mathrm{ann}}_{\;\sltau^+\,\sltau^-\rightarrow \gamma\,\gamma}\,v\rangle
374: \simeq 4 \pi \alpha^2/m_{\stau_1}^2$.
375: If $m_{\stau_1}$ and $m_h$ are not too close, one has
376: \begin{equation}\label{eq:RatioAnnihilationCrossSections}
377: \frac{\langle {\sigma_\mathrm{ann}}_{\;\stau^+\,\stau^-\rightarrow
378: h\,h}\,v\rangle}{%
379: \langle {\sigma_\mathrm{ann}}_{\;\stau^+\,\sltau^-\rightarrow \gamma\,\gamma}\,v\rangle}
380: ~ \sim~
381: \left( \frac{\tan\beta}{50}\right) ^4\,
382: \left( \frac{\mu_\mathrm{eff}}{2 m_{\stau_1}}\right) ^4\;,
383: \end{equation}
384: where $\mu_\mathrm{eff}=\mu\, \sin{2 \theta_{\sltau}}$.
385: Hence, for $\mu_\mathrm{eff}>2m_{\stau_1}$ the annihilation cross section is dominated by the
386: Higgs channel. Even for an order one ratio $\mu_\mathrm{eff}/2m_{\stau_1}$ one obtains a
387: dramatic reduction of the primordial stau abundance.
388: However, as we shall discuss next, there are constraints on the ratio
389: $\mu_\mathrm{eff}/2m_{\stau_1}$, implying that the reduction cannot be
390: arbitrarily strong.
391:
392:
393:
394: \section{Theoretical constraints on $\boldsymbol{\mu}$}
395:
396:
397: %\subsection{}
398: \label{sec:muconstraints}
399:
400: The enhancement of the stau annihilation cross section relies on a large
401: trilinear Higgs-stau coupling, which might lead to an unwanted (color and)
402: charge breaking (CCB) minimum of the potential. Such minima might be acceptable
403: if the `physical' vacuum is sufficiently long-lived. In what follows, we will
404: first analyze the tree-level potential and see that in the interesting regions
405: of parameter space there is indeed an unphysical minimum. We will discuss
406: tunneling to this vacuum. Then we will discuss quantum corrections to the
407: potential and see that they lift the unwanted minimum and possibly even make it
408: disappear. Finally, we will comment on constraints on $\mu$ arising from
409: unitarity.
410:
411:
412: \subsection{CCB constraints at tree level}
413:
414: The relevant field space is given by two real fields,
415: $\stau=\frac{1}{\sqrt{2}}\text{Re}(\stau_1)$ and $h$.
416: The corresponding scalar potential around the electroweak
417: vacuum can be written as
418: \begin{equation}
419: V~=~\frac{1}{2}m_{\stau_1}^2\,\stau^2+\frac{1}{2}m_h^2\,h^2
420: +b_h\,h^3+b_{\stau\stau h}\,h\,\stau^2
421: +\lambda_h\,h^4+\lambda_{\stau}\,\stau^4+\lambda_{\stau h}\,\stau^2\,h^2\;,
422: \end{equation}
423: where at tree level $m_h \simeq 90 \gev$, $b_h\simeq 17\gev$, $\lambda_h\simeq
424: \lambda_{\stau h}\simeq 0.018$, $\lambda_{\stau}\simeq 0.028$ and the trilinear
425: coupling $b_{h\stau\stau} = - \frac{y_\tau}{\sqrt{8}}\mu_\mathrm{eff}$. For very
426: negative $b_{h\stau\stau}$ values there is a second minimum of the tree-level
427: potential. We have searched for minimal paths connecting the electroweak vacuum
428: with the second, deeper minimum. It turns out that the relevant field direction
429: is always very close to $x=\frac{1}{\sqrt{2}}(\stau+h)$. The corresponding potential
430: along that direction can be written as
431: \begin{equation}
432: \label{pot}
433: V~=~
434: \frac{1}{4}(m_{\stau_1}^2+m_h^2)\,x^2 +
435: \left(\frac{b_h}{\sqrt{8}}- y_{\tau} \frac{\mu_{\text{eff}}}{8}\right)\,x^3
436: +\frac{1}{4} (\lambda_h + \lambda_{\stau} + \lambda_{h\stau})\, x^4 \;.
437: \end{equation}
438: In order to check whether the lifetime of the local minimum at $x=0$
439: exceeds the age of the universe we have to calculate the so called bounce action $S_\text{B}$
440: along the lines of \cite{Coleman:1977py}, which should satisfy $S_\text{B} \ge 400$ \cite{Kusenko:1996jn}.
441: Using the potential \eqref{pot} with the coefficients given by their tree-level values
442: we find an upper bound
443: on the coefficient of the trilinear term
444: $|\frac{b_h}{\sqrt{8}}-\frac{\mu_\mathrm{eff}}{16}| \lesssim 33.5 \,\mathrm{GeV}$,
445: which translates into
446: \begin{equation}
447: \mu_\mathrm{eff} ~\lesssim~630\,\mathrm{GeV}
448: \quad\text{(for $\tan\beta=50$ and $m_{\stau_1}=120\gev$)}\;.
449: \end{equation}
450: Note that for $\mu_\mathrm{eff} \lesssim 430\gev$ the tree-level potential
451: does not exhibit a second, deeper minimum at all.
452:
453:
454: \subsection{Quantum corrections to CCB constraints}
455: \label{sec:QuantumCorrections}
456:
457: It is well known that quantum corrections can change the tree-level picture drastically.
458: In order to analyze the situation properly, one has to take into
459: account radiative corrections to the potential.
460: This is a very complicated issue, and we refrain from performing a complete
461: analysis here.
462: In order to get a feeling for the impact of radiative corrections
463: we include the standard stop loop correction (cf.\ e.g.\
464: \cite[p.~245~f.]{Drees:2004jm}).
465: The resulting effective potential is significantly steeper in the Higgs
466: direction, such that the physical Higgs mass can
467: violate the tree-level bound $m_h\le m_Z$.
468: It also turns out that the cubic and
469: quartic coefficients in the Higgs potential get enhanced by $\sim 70\,\%$ and
470: $\sim 40\,\%$, respectively. This has important implications for the bounce action:
471: plugging the loop corrected values for these parameters into \eqref{pot} we find that
472: the metastable minimum has a sufficiently large lifetime for
473: \begin{equation}\label{eq:mubound2}
474: \mu_\mathrm{eff}~\lesssim~780\,\mathrm{GeV}
475: \quad\text{(for $\tan\beta=50$ and $m_{\stau_1}=120\gev$)}\;.
476: \end{equation}
477: The second, deeper minimum exists only if $\mu_\mathrm{eff} \gtrsim 500 \,\mathrm{GeV}$.
478:
479:
480: There are similar effects, in particular for the potential in $\stau$ direction.
481: This issue has been studied in~\cite{Ferreira:2001tk}, where it was found that
482: there are no charge breaking minima at all after quantum corrections are taken
483: into account.
484: Whether or not these statements also apply to parameter regions with large
485: $\tan\beta$ and $\mu_\mathrm{eff}$ will be studied elsewhere.
486:
487:
488:
489: \subsection{Unitarity bound}
490:
491: Further constraints on $\mu_\mathrm{eff}$ come from unitarity.
492: The unitary cross section for scalar particles can be calculated using partial
493: wave expansion. In the case of a non-elastic process it takes the form
494: \cite{Griest:1989wd}
495: \begin{equation}
496: \sigma_\mathrm{unit}~=~\frac{4 \pi(2J+1)}{|\vec{p}_\mathrm{in}|^2}\;,
497: \end{equation}
498: where $J$ is the angular momentum of the partial wave and
499: $\vec{p}_\mathrm{in}$ is the three-momentum of one incoming particle.
500: The dominant contribution to the stau annihilation into Higgs bosons comes from $s$-wave scattering.
501: Therefore the perturbative unitarity constraint relevant to our discussion is
502: \begin{equation}\label{eq:UnitarityConstraint}
503: {\sigma_\mathrm{ann}}_{\;\stau^+\,\stau^-\rightarrow
504: h\,h}
505: ~\leq ~\sigma_{\mathrm{unit},s}~=~\frac{4 \pi}{|\vec{p}_\mathrm{in}|^2}\;.
506: \end{equation}
507: If the bound is not respected, this signals that the perturbative calculation of the cross section
508: is no longer valid. In our case, the unitarity bound
509: translates into a constraint on the $\mu$ parameter. The bound is practically
510: independent of $m_h$, and reads
511: \begin{equation}\label{eq:UnitarityBound2}
512: \mu_\mathrm{eff}\,\frac{\tan\beta}{50}
513: ~\lesssim~4.1\,\mathrm{TeV}\times\left(\frac{m_{\widetilde{\tau}_1}}{150\gev}\right)\;.
514: \end{equation}
515: We find that the annihilation channel into Higgs pairs relative to the
516: annihilation into gauge bosons can well be larger by factors of
517: $\mathcal{O}(10^3)$ without violating the unitarity
518: bound~\eqref{eq:UnitarityBound2}.
519:
520: \section{Primordial staus}
521: \label{primordial}
522:
523: \subsection{BBN constraints}
524:
525: Various cosmological constraints on the stau yield $Y_\stau \equiv Y_{\stau^+}
526: +Y_{\stau^-}$ have been explored in the literature. In a scenario where the LSP
527: is very weakly coupled, the NLSP decays a considerable time after the start of
528: BBN. The decay products of such long-lived particles can alter the primordial
529: light element abundances. This leads to constraints on the released
530: electromagnetic and hadronic energy
531: \cite{Cyburt:2002uv,Kawasaki:2004qu,Jedamzik:2006xz}. The constraints on the
532: stau relic abundance from decays depend on the stau lifetime and mass as well
533: as on the electromagnetic and hadronic branching ratios. For reasonable values
534: of the stau mass, the hadronic branching fraction is typically $\lesssim
535: \mathcal{O}(10^{-3})$ which leads to rather weak constraints. Stronger
536: constraints come from electromagnetic energy injection, especially at late
537: times. Here the bounds can be as strong as $m_\stau Y_\stau \lesssim 10^{-13} \gev$
538: \cite{Kawasaki:2007xb}. However, typically the stau NLSP abundance and lifetime
539: can satisfy the limits obtained from BBN by considering NLSP decays alone
540: \cite{Fujii:2003nr,Feng:2004mt,Ellis:2003dn,Cerdeno:2005eu,Jedamzik:2005dh,%
541: Steffen:2006hw,Buchmuller:2006nx,Cyburt:2006uv,Pradler:2006hh}
542:
543: In addition to injecting energetic showers into the plasma through decays,
544: negatively charged particles can form bound states with light nuclei, which can
545: lead to a drastic overproduction of $^6$Li \cite{Pospelov:2006sc} and $^9$Be
546: \cite{Pospelov:2008ta}. This leads to strong constraints on the stau yield
547: $Y_{\stau^-}$ for lifetimes longer than a few thousand seconds. While
548: \cite{Jedamzik:2007qk} gives a rather conservative upper bound of $Y_{\stau^-}
549: \lesssim 10^{-14}$ derived from $^6$Li alone, \cite{Pospelov:2008ta}
550: takes into account $^6$Li as well as $^9$Be leading to $Y_{\stau^-} \lesssim
551: 10^{-15}$ which translates into $Y_{\stau} \lesssim 2 \cdot 10^{-15}$ in the
552: absence of a stau anti-stau asymmetry.
553:
554:
555:
556: \subsection{Stau relic abundance}
557:
558: The relic abundance of a (meta)stable particle can be calculated using the
559: Boltzmann equation. For the stau yield $Y_\stau$, (again in the absence of a
560: stau anti-stau asymmetry) it takes the form \cite{Gondolo:1990dk}
561: \begin{equation}\label{eq:Boltzmannyield}
562: \frac{\D Y_{\stau}}{\D x}
563: ~=~-\sqrt{\frac{2 g_*}{45}}\,\pi\, M_\mathrm{P}\,
564: \frac{m_{\stau_1}}{x^2}\,\left\langle\sigma_\mathrm{ann}\, v\right\rangle\,
565: \left( Y^2_{\sltau} -
566: Y^2_{\sltau,\mathrm{eq}}\right) \;,
567: \end{equation}
568: where $M_\mathrm{P}=2.43\cdot 10^{18} \gev$ is the reduced Planck mass,
569: $Y^2_{\sltau,\mathrm{eq}}$ is the abundance in thermal equilibrium and $g_*
570: \simeq 85$ represents the effective number of degrees of freedom at the stau
571: freeze out (cf.~\cite{Gondolo:1990dk}).
572:
573: It is well known that the relic abundance of a (meta)stable particle is
574: inversely proportional to its annihilation cross section times its mass (see
575: e.g.\ \cite{Drees:2004jm}). In the case where the stau freeze-out is dominated
576: by the Higgs channel, we can write the solution to~\eqref{eq:Boltzmannyield} as
577: \begin{equation}\label{StauYieldApproximation2}
578: Y_\stau ~=~10^{-15} \left( \frac{10^{-5} \gev^{-2}}{\langle\sigma v\rangle }\right)
579: \left( \frac{200 \gev}{m_{\stau_1}}\right) \;.
580: \end{equation}
581: If stau and Higgs are not mass degenerate ($m_{\stau_1} - m_h \gtrsim 5 \gev$),
582: the annihilation cross section~\eqref{eq:ThermalCrossSection} is practically
583: independent of $m_h$ and depends only on the stau mass $m_{\stau_1}$ and the
584: stau-Higgs coupling $\propto \mu\,\tan\beta\,
585: \sin{2\theta_\stau}=\mu_\mathrm{eff}\,\tan\beta$. Our result for
586: the yield can then be written as
587: \begin{equation}\label{eq:StauYieldApproximation}
588: Y_\stau =1.4\cdot10^{-15} \left( \frac{m_{\stau_1}}{150\gev}\right)^5 \left( \frac{1 \tev}{\mu}\right)^4
589: \left( \frac{50}{\tan\beta}\right)^4 \left( \frac{1}{\sin{2\theta_\stau}}\right)^4
590: \;.
591: \end{equation}
592: Here we neglected subleading effects of the order $10\,\%$ like e.g.~Sommerfeld
593: enhancement.\footnote{See e.g.~\cite{Berger:2008ti} for an explanation of the Sommerfeld
594: effect and an estimate of the errors in the general case of a charged relic.}
595: Combining our result~\eqref{eq:StauYieldApproximation} with the unitarity
596: bound~\eqref{eq:UnitarityBound2} we find that the lowest allowed abundance is
597: $Y^\text{min}\sim 4\cdot 10^{-18}$ for $m_{\stau_1}=120\gev$.
598: As explained in Section~\ref{sec:muconstraints}, the minimal, theoretically
599: viable abundance might well turn out to be larger than this value. From the conservative
600: bound \eqref{eq:mubound2}
601: we infer however that nevertheless $Y^\text{min} \lesssim 10^{-15}$.
602:
603:
604: \subsection{A comment on stau asymmetries}
605:
606: So far we have assumed that there is no asymmetry in the stau sector, that is,
607: there are as many $\widetilde{\tau}^+$ as $\widetilde{\tau}^-$ degrees of
608: freedom.
609: On the other hand, a large class of baryogenesis mechanisms rely on
610: $(B+L)$ violation by sphalerons \cite{Klinkhamer:1984di,Kuzmin:1985mm}, which
611: leads to an excess of baryons over anti-baryons if there are more anti-leptons
612: than leptons. In particular, leptogenesis \cite{Fukugita:1986hr} falls into this
613: class. From this point of view it appears natural to assume that there is also
614: an asymmetry in the slepton sector at the time of stau annihilation and freeze
615: out.
616: Now it is important to distinguish between slepton number conserving and slepton
617: number violating annihilation processes
618: (Figure~\ref{fig:SleptonNumberNonConservation}).
619:
620: \begin{figure}[h]
621: \centerline{\subfigure[]{\CenterEps{Gravitino34}}\quad\quad\subfigure[]{\CenterEps{Gravitino35}}}
622: \caption{Examples for slepton number conserving (a) and violating (b) annihilation processes.}
623: \label{fig:SleptonNumberNonConservation}
624: \end{figure}
625:
626: As we have seen, in settings with substantial left-right mixing in the stau
627: sector, the slepton number conserving processes dominate over the violating
628: ones. On the other hand, the slepton number violating processes are still
629: effective until the stau relic abundance has reached a value
630: $Y_{\stau}\sim10^{-12}\dots10^{-13}$ \cite{Asaka:2000zh}. It is then clear that
631: if the slepton asymmetry in the stau sector is of the order of the baryon
632: asymmetry, $\eta_b\sim 10^{-10}$, asymmetries will play no role. However, if
633: there are order one asymmetries in the (s)tau sector, this might have important
634: implications for the relic abundance of $\stau^-$: in a situation in which there
635: is a large excess of $\stau^+$ over $\stau^-$, the remaining $\stau^-$ are more
636: likely to find an annihilation partner, and hence their relic abundance can get
637: suppressed more strongly. For stau lifetimes where electromagnetic bounds are
638: not overly restrictive, $\tau_\stau \lesssim 10^7\,\mathrm{s}$, a relatively
639: large abundance of $\stau^+$, $Y_{\stau^+}\lesssim10^{-13}$, can still be
640: consistent with BBN because they do not form bound states with nuclei.
641:
642:
643:
644: \section{Scenarios with low $\boldsymbol{Y_{\stau}}$}
645: \label{sec:scenarios}
646:
647: In this section we present some `top-down motivated settings' in which the
648: previously discussed strong suppression of the stau relic abundance occurs. If a
649: scenario can have a large stau annihilation cross section is fully determined
650: by the stau spectrum. The necessary conditions are:
651: \begin{description}
652: \item[Condition 1:] Substantial left-right mixing of the stau mass eigenstates $\stau_1$ and $\stau_2$ driven by a
653: large off-diagonal stau mass matrix element $m_{\stau_{\mathrm{LR}}}^2\simeq
654: -m_\tau\,\mu\,\tan\beta$. This requires large $\mu$ and $\tan\beta$.
655: \item[Condition 2:] Moderate $m_{\stau_1}$, preferably $m_{\stau_1} \lesssim
656: 200\gev$.
657: \end{description}
658: We will check these conditions by looking at the entries of the stau mass
659: matrix $m_{\stau_{\mathrm{R}}}^2$, $m_{\stau_{\mathrm{L}}}^2$
660: and $m_{\stau_{\mathrm{LR}}}^2$ as well as on the mass eigenvalues
661: $m_{\stau_1}$ and $m_{\stau_2}$.
662:
663: It is clear that we could simply write down the desired stau mass matrices. The
664: purpose of this section however is to present soft mass patterns defined at the
665: unification scale $M_\mathrm{GUT}$ that lead to mass matrices with the above
666: properties. To this end, it is useful to recall the (one-loop) renormalization
667: group equations (RGEs) for the stau soft masses (see e.g.\ \cite{Drees:2004jm})
668: \begin{subequations}\label{eq:stauRGEs}
669: \begin{eqnarray}
670: 8\pi^2\,\frac{\D m_{\stau_{\mathrm{L\,\text{soft}}}}^2}{\D t}
671: & = &
672: y_\tau^2\,S_\tau-3g_2^2\,|M_2|^2-g_Y^2\,|M_1|^2-\frac{1}{2}g_Y^2\,S_Y\;,\\
673: 8\pi^2\,\frac{\D m_{\stau_{\mathrm{R\,\text{soft}}}}^2}{\D t}
674: & = &
675: 2\,y_\tau^2\,S_\tau-4\,g_Y^2\,|M_1|^2+g_Y^2\,S_Y\;,\label{eq:stauRightRGE}
676: \end{eqnarray}
677: \end{subequations}
678: where
679: \begin{subequations}
680: \begin{eqnarray}
681: S_\tau & = &
682: m_{H_1}^2+m_{\stau_{\mathrm{L\,\text{soft}}}}^2+m_{\stau_{\mathrm{R\,\text{soft}}}}^2+|A_\tau|^2
683: \;,\label{eq:STau}\\
684: S_Y & = & \frac{1}{2}\sum_i Y_i\,m_i^2
685: \end{eqnarray}
686: \end{subequations}
687: with $Y_i$ denoting the hypercharge of the scalar $i$. Note that
688: $m_{\stau_{\mathrm{L}}}^2 \simeq m_{\stau_{\mathrm{L\,\text{soft}}}}^2$ and
689: $m_{\stau_{\mathrm{R}}}^2 \simeq m_{\stau_{\mathrm{R\,\text{soft}}}}^2$ as other
690: contributions are tiny (cf.~\eqref{eq:StauMassMatrix}).
691:
692:
693: In order to obtain large mixing (condition 1) we have to demand that the right-hand sides (rhs)
694: of \eqref{eq:stauRGEs} be similar, assuming coincident stau masses at
695: the high scale. In addition, we need a large off-diagonal stau mass.
696: $|\mu|$ which determines the size of $m_{\stau_{\mathrm{LR}}}^2$ together with $\tan\beta$ is fixed at
697: the weak scale by the condition of correct electroweak symmetry
698: breaking which reads (at tree-level)~\cite{Drees:2004jm}
699: \begin{equation}\label{eq:MuEWSB}
700: |\mu|^2=\frac{m_{H_2}^2 \sin^2\beta-m_{H_1}^2 \cos^2\beta}{\cos{2\beta}}-\frac{M_Z^2}{2}
701: \stackrel{\text{large }\tan\beta}{\simeq} -m_{H_2}^2\;.
702: \end{equation}
703: An unsuppressed $m_{\stau_{\mathrm{LR}}}^2$ can typically be realized for
704: $\mu\sim 1 - 2\tev$. In principle it is not difficult to get $\mu$ in this
705: range, however one should mention here that a relatively large $\mu$ might be
706: associated with a significant amount of electroweak fine-tuning as can be seen
707: from \eqref{eq:MuEWSB}.In addition, very large values for $\mu$ might lead to charge
708: breaking vacua with unacceptably short lifetimes (cf.\ Section~\ref{sec:muconstraints}).
709:
710: The second condition is already partially fulfilled if the mixing is sizable.
711: To further reduce the stau masses, the rhs of \eqref{eq:stauRGEs}
712: should not be too negative.
713:
714:
715: In what follows we present three scenarios where the desired stau mass patterns
716: arise and low relic abundances through the Higgs channel can be achieved. We
717: use micrOMEGAs 2.0.7 \cite{Belanger:2001fz,Belanger:2006is} to calculate the
718: relic abundance of the stau NLSP numerically.
719: The superpartner spectrum is determined by SOFTSUSY 2.0.18
720: \cite{Allanach:2001kg} whereas the Higgs mass is calculated with the
721: specialized tool FeynHiggs 2.5.1
722: \cite{Heinemeyer:1998yj,Heinemeyer:1998np,Degrassi:2002fi,Frank:2006yh}. For
723: the top quark pole mass, we use the latest best-fit value of
724: $172.6\gev$~\cite{Group:2008nq}. Experimental constraints on the parameter space
725: arise primarily through mass limits. We employ the LEP Higgs bound $m_h\geq 114
726: \gev$ and $m_{\stau_1}\geq 100\gev$ \cite{staumass}.
727: Theoretical constraints, as discussed in
728: Section~\ref{sec:muconstraints}, are not shown explicitly.
729:
730:
731:
732:
733: \subsection{CMSSM with large $\boldsymbol{\tan\beta}$}
734:
735: Let us start with the constrained supersymmetric standard model (CMSSM), which
736: is defined through its free parameters $m_{1/2}$, $m_0$, $A_0$, $\tan\beta$ and
737: $\text{sign}\, \mu$. Although we will see that the annihilation can be more
738: efficient in other scenarios, we find it nevertheless worthwhile to point out
739: that also in this scheme a major suppression of the relic abundance is possible.
740: The important quantity here is $\tan\beta$ which we plot against the stau
741: spectrum in the left panel of Figure~\ref{fig:msugrarelic} for a typical stau
742: NLSP parameter point.
743:
744:
745: The plot shows that $m_{\stau_{\mathrm{R}}}$ and $m_{\stau_1}$ decrease strongly
746: with $\tan\beta$ through the tau Yukawa term in the RGE of
747: $m_{\stau_{\mathrm{R\,\text{soft}}}}^2$, because $y_\tau \propto \tan\beta$. Since
748: $\mu$ is practically independent of $\tan\beta$ the off-diagonal
749: mass\footnote{We use the definition
750: $m_{\stau_{\mathrm{LR}}}=\sqrt{|m_{\stau_{\mathrm{LR}}}^2|}$ in the following.}
751: $m_{\stau_{\mathrm{LR}}}$ grows like $\sqrt{\tan\beta}$ which leads to a further
752: reduction of $m_{\stau_1}$ through left right-mixing. However, in spite of a
753: strong off-diagonal stau mass, mixing cannot get close to maximal, because of a
754: large difference $m_{\stau_{\mathrm{L}}}^2-m_{\stau_{\mathrm{R}}}^2$.
755: %
756: We conclude that condition 2 can easily be satisfied while condition 1 only
757: partially. In summary, we obtain a significant enhancement of the
758: stau annihilation cross section through the Higgs channel at large $\tan\beta$
759: which is however limited by the mixing angle. To illustrate the effect,
760: Figure~\ref{fig:msugrarelic} shows the stau relic abundance in the CMSSM as a function of
761: $\tan\beta$.
762: \begin{figure}[t!]
763: \centering
764: \begin{minipage}[b]{7.2 cm}
765: \includegraphics[width=7.3cm,height=8cm]{cmssmmasses}
766: \end{minipage}\hspace*{1cm}
767: \begin{minipage}[b]{7.2 cm}
768: \includegraphics[width=7.3cm,height=8cm]{cmssmrelic}
769: \end{minipage}
770: \caption{ Stau relic abundance in the CMSSM. The left panel shows the dependence of the left- and right-handed
771: stau masses, the off-diagonal mass
772: and the mass eigenvalues on $\tan\beta$. In the right panel we plot the logarithm of the stau relic abundance
773: $\log_{10}(Y_\stau)$ in the ($m_{1/2}-\tan\beta$) --plane.
774: We find the minimal yield to be around $Y_{\sltau}^\text{min}\simeq 10^{-15}$.
775: The white region in the right panel is excluded due to direct searches.
776: }
777: \label{fig:msugrarelic}
778: \end{figure}
779:
780:
781:
782: \subsection{Non-universal Higgs masses (NUHM)}
783:
784: In the NUHM we can vary -- in addition to the parameters of the CMSSM -- the
785: down- and up-type soft Higgs masses squared at the GUT scale, $m_{H_1}^2$ and
786: $m_{H_2}^2$. We employ the GUT scale stability constraint $m_{H_{1,2}}^2+|\mu^2|
787: \geq 0$ to avoid dangerous vacua and electroweak symmetry breaking at the GUT
788: scale \cite{Ellis:2002iu}. It is instructive to investigate the additional
789: effects on the stau spectrum compared to the CMSSM which arise through the
790: variation of the soft Higgs masses: increasing $m_{H_1}^2$ leaves
791: $m_{\stau_{\mathrm{LR}}}$ unchanged, but it reduces $m_{\stau_{\mathrm{L}}}$ and
792: $m_{\stau_{\mathrm{R}}}$ dominantly through the Yukawa term in the soft mass
793: RGEs \eqref{eq:stauRGEs} which contains $m_{H_1}^2$. However more interesting
794: for us is the impact of $m_{H_2}^2$ which we show in the left panel of
795: Figure~\ref{fig:NuhmSpectrum}.
796: \begin{figure}[t!]
797: \centering
798: \begin{minipage}[b]{7.2 cm}
799: \includegraphics[width=7.3cm,height=8cm]{NUHMmasses}
800: \end{minipage}\hspace*{1cm}
801: \begin{minipage}[b]{7.2 cm}
802: \includegraphics[width=7.3cm,height=8.02cm]{NUHMrelic}
803: \end{minipage}
804: \caption{Stau relic abundance in the NUHM. The left panel shows the dependence of the left- and right-handed
805: stau masses, the off-diagonal mass
806: and the mass eigenvalues on the soft Higgs mass $m_{H_2}^2$.
807: In the right panel we plot the logarithm of the stau relic abundance
808: $\log_{10}(Y_\stau)$ for different $m_{H_1}^2$ and $m_{H_2}^2$.
809: Some of the low $\stau$ yield regions might be
810: excluded by theoretical constraints (cf.\
811: Section~\ref{sec:muconstraints}). Again, the white region in the right panel is excluded. }
812: \label{fig:NuhmSpectrum}
813: \end{figure}
814: We observe that a growth of $m_{H_2}^2$ reduces $m_{\stau_{\mathrm{R}}}$ and
815: enlarges $m_{\stau_{\mathrm{L}}}$. As $m_{H_2}^2$ does not appear in the Yukawa
816: term of the RGEs \eqref{eq:stauRGEs}, this effect arises through the $S_Y$-term.
817: For the off-diagonal stau mass $m_{\stau_{\mathrm{LR}}}$, it is
818: important to recall that $|\mu^2|\simeq -m_{H_2}^2$ at the weak scale,
819: cf.~\eqref{eq:MuEWSB}. Therefore, increasing $m_{H_2}^2$ leads to a
820: suppression of $m_{\stau_{\mathrm{LR}}}^2$.
821:
822: The low yield parameter region is at relatively large negative $m_{H_2}^2$, where the GUT
823: stability constraint can still be satisfied and again at large $\tan\beta$. Here both, the left-right mixing as well as the off-diagonal
824: stau mass $m_{\stau_{\mathrm{LR}}}$, can further be enhanced compared to the CMSSM case. Stau masses which are slightly
825: larger can be reduced by a positive $m_{H_1}^2$. We conclude that both conditions can be satisfied
826: in this region, leading to an extremely suppressed stau abundance.
827:
828:
829:
830:
831: \subsection{Scenarios with non-universal gaugino masses (NUGM)}
832:
833:
834:
835: \begin{figure}[t!]
836: \centering
837: \begin{minipage}[b]{7.3 cm}
838: \includegraphics[width=7.2cm,height=8cm]{NUGMmasses}
839: \end{minipage}\hspace*{1cm}
840: \begin{minipage}[b]{7.3 cm}
841: \includegraphics[width=7.2cm,height=8cm]{NUGMrelic}
842: \end{minipage}
843: \caption{Stau relic abundance in the NUGM. The left panel shows the dependence of the left- and right-handed
844: stau masses, the off-diagonal mass
845: and the mass eigenvalues on $m_{24}$. In the right panel we plot the logarithm of the stau relic abundance
846: $\log_{10}(Y_\stau)$ in the ($m_{1/2}-m_{24}$) --plane.
847: Some of the low $\stau$ yield regions might be
848: excluded by theoretical constraints (cf.\
849: Section~\ref{sec:muconstraints}).
850: Once more, the white region in the right panel is excluded.
851: }
852: \label{fig:Rossrelic}
853: \end{figure}
854: The possibility of having non-universal gaugino masses as high scale boundary
855: conditions even for unified theories has been realized long ago
856: \cite{Ellis:1985jn}. For concreteness we focus here on SU(5) GUTs and assume
857: that supersymmetry be broken by $F$-term vacuum expectation values of chiral
858: fields. These fields have to transform as the symmetric product of two adjoint
859: representations of the GUT group, but not necessarily as singlets. In the
860: following we concentrate on the two smallest possible representations for the
861: supersymmetry breaking fields, which are simply the singlet and the
862: $\boldsymbol{24}$plet.
863:
864: The high-scale mass patterns of the gauginos of SU(3)$_C$, SU(2)$_\mathrm{L}$
865: and U(1)$_Y$ turn out to be given as linear combinations of singlet ($m_{1/2}$)
866: and $\boldsymbol{24}$plet ($m_{24}$) contributions \cite{Huitu:1999ac},
867: \begin{eqnarray}
868: M_1&=&m_{1/2}-0.5\,m_{24} \;,\nonumber\\
869: M_2&=&m_{1/2}-1.5\,m_{24} \;,\nonumber\\
870: M_3&=&m_{1/2}+m_{24} \;.\label{eq:GUTGauginoMasses124}
871: \end{eqnarray}
872: The only new parameter compared to the CMSSM is the mass arising from the
873: $\boldsymbol{24}$plet, $m_{24}$. Its impact on the stau spectrum is illustrated
874: in the left panel of Figure~\ref{fig:Rossrelic}.
875:
876: $m_{\stau_{\mathrm{L}}}$ decreases quickly for growing $m_{24}$ as the smaller
877: $M_2$ reduces the SU(2)$_\mathrm{L}$ gauge contribution to
878: $m_{\stau_{\mathrm{L\,\text{soft}}}}^2$. $m_{\stau_{\mathrm{R}}}$ increases
879: slightly, although the U(1)$_Y$ gauge contribution to
880: $m_{\stau_{\mathrm{R\,\text{soft}}}}^2$ shrinks. This is because the reduction
881: of $m_{\stau_{\mathrm{R\,\text{soft}}}}^2$ through the Yukawa term in
882: \eqref{eq:stauRightRGE} is less effective for smaller
883: $m_{\stau_{\mathrm{L\,\text{soft}}}}^2$. The off-diagonal mass matrix entry
884: $m_{\stau_{\mathrm{LR}}}^2 \propto \mu$ grows slowly as $\mu$ gets a contribution
885: from the increasing gluino mass $M_3$.
886:
887:
888: It turns out that diagonal and off-diagonal stau soft masses can come very
889: close, leading to maximal left-right mixing in the stau
890: sector with a strong reduction of $m_{\sltau[1]}$ driven by a large
891: $m_{\stau_{\mathrm{LR}}}$. As this is the key ingredient for strongly enhanced
892: couplings of $\sltau[1]$ to the light Higgs, the stau relic abundance becomes very small, as can be seen in
893: the right panel of Figure~\ref{fig:Rossrelic}.
894:
895:
896:
897:
898: \subsection{Further remarks}
899:
900:
901:
902:
903:
904: Let us briefly summarize the main results of this section. We have seen that a
905: strong suppression of the stau relic abundance can be achieved such that all
906: cosmological constraints can be evaded. We also checked that the shown regions
907: of parameter space are not in conflict with precision measurements of $b
908: \rightarrow s\,\gamma$ and the Tevatron limit of $B_s \rightarrow \mu^+\mu^-$.
909: Furthermore, the parameter regions with low relic abundance are all within the
910: $2\,\sigma$ interval of the measured anomalous magnetic moment of the muon, the
911: only exception being the CMSSM for $m_{1/2}\gtrsim 1100 \gev$.
912: Comparing the low yield
913: regions with the CBBN bound $Y_\stau \lesssim 2\cdot 10^{-15}$ we see that
914: even in the CMSSM there remains some viable region in parameter space while for
915: the NUHM and NUGM the stau relic abundance can be even
916: smaller. Further analysis of the experimental signatures of such low stau yield
917: regions appears desirable.
918:
919:
920:
921: \section{Prospects for the LHC}
922: \label{sec:Prospects}
923:
924: A very appealing property of our scenario is that it will be tested at the LHC.
925: Given the stau spectrum, the stau-Higgs coupling is fully determined. It is then
926: easy to see the impact of the Higgs channel for stau annihilation in the early
927: universe.
928:
929: A common feature of the low yield regions is a rather small stau mass,
930: $m_{\stau_1} \lesssim 250\gev$, while the other SUSY particles may be quite
931: heavy. In general, the prospects for the LHC depend mainly on the mass scale of
932: the colored superpartners. If they are not too heavy, as can be the case e.g.~in
933: the NUHM, they will be produced in large numbers at the LHC due to their large
934: cross sections. In this case one has a good chance to measure the whole SUSY
935: spectrum in the cascade decays of gluinos and squarks and it should not be too
936: difficult to extract information about the stau-Higgs coupling.
937:
938: If, however, the mass scale of colored particles exceeds $2-2.5\tev$, their
939: production will become very rare or even impossible at the LHC
940: \cite{Baer:2003wx}. But even in this case, being rather light, staus could
941: still be pair produced, e.g.~in the Drell-Yan process $q \, \bar{q} \rightarrow
942: \stau_1^- \,\stau_1^+ $ through a virtual photon or $Z$ boson
943: \cite{Eichten:1984eu}. The number of produced stau pairs for different stau
944: masses are shown in Table~\ref{drellyan}. Further details on the stau spectrum might be extracted from the
945: Drell-Yan production of $\stau_1 \, \snu_\tau$, $\stau_1 \, \stau_2$ and
946: $\stau_2 \, \stau_2$, if kinematically accessible.
947:
948: %---------------------------Table--------------------------------------
949: \begin{table}[!ht!]
950: \centering
951: \renewcommand{\arraystretch}{1.7}
952: \begin{tabular}{||c||c|c|c|c||}
953: %%%%%
954: \hline
955: Mass & $150\gev$ & $200\gev$ & $300\gev$ & $400\gev$
956: \\ \hline
957: $\#_{\stau_\text{R}}$ & $2000$ & $800$ & $200$ & $60$
958: \\ \hline
959: $\#_{\stau_\text{L}}$ & $6000$ & $2000$ & $450$ & $150$
960: %%%%%%%%%%%%%
961: \\ \hline
962: \end{tabular}
963: \caption{Estimated number of produced stau pairs in Drell-Yan processes at the
964: LHC for integrated luminosity of $100 \, \text{fb}^{-1}$ extracted
965: from Figure~1 of \cite{Bozzi:2004qq}. The number of produced mass
966: eigenstates depends on the mixing angle and should lie in between the values given for left- and
967: right-handed staus.}
968: \label{drellyan}
969: \end{table}
970: %------------------------End Table-------------------------------------
971: Another interesting pair production mechanism in our scenario is gluon-gluon fusion.
972: Here two gluons generate a fermion loop (preferably a top-loop) to which a virtual Higgs boson is attached
973: which finally decays into a stau pair.
974: In an early study \cite{delAguila:1990yw} the cross section for this process was found to be three orders of magnitude
975: below the corresponding Drell-Yan cross section.
976: Note that in our scenario, due to the strong stau-Higgs coupling, this
977: suppression can at least partially be compensated such that gluon-gluon fusion could be comparable to
978: the Drell-Yan process.
979:
980:
981:
982: Altogether we see that even in scenarios in which some superpartners are beyond
983: the reach of the LHC, one may nevertheless establish the existence of
984: supergravity in nature along the lines of \cite{Buchmuller:2004rq,Buchmuller:2004tm,Feng:2004mt,Hamaguchi:2004df,Feng:2004yi,Brandenburg:2005he,Martyn:2006as,%
985: Ellis:2006vu,Kitano:2008sa}.
986:
987:
988: \section{Conclusions}
989:
990: We have analyzed stau NLSP scenarios. In contrast to previous studies, we have
991: not assumed that the stau mass eigenstates be purely right- or left-handed, but
992: have allowed for non-trivial left-right mixing. In the case of substantial
993: mixing, the annihilation into Higgs bosons can dominate over other channels,
994: such that the thermal relic stau abundance, i.e.\ the abundance obtained without
995: invoking late-time entropy production, can be strongly reduced. This makes it
996: possible to evade all BBN constraints. The emerging scenarios have the
997: advantage that they allow for rather large reheating temperatures, as required
998: for instance by leptogenesis, and the cold dark matter can be explained in terms of
999: `thermally' produced gravitinos. Most importantly, all ingredients of our low
1000: stau yield scenarios will be tested at the LHC.
1001:
1002:
1003:
1004:
1005: \section*{Acknowledgments}
1006: We would like to thank Alejandro Ibarra and David Straub for useful discussions.
1007: %
1008: We are indebted to the referee for important comments on an earlier version of
1009: this paper.
1010: This work has been supported by the SFB-Transregio 27 ``Neutrinos and Beyond''
1011: and by the DFG cluster of excellence ``Origin and Structure of the Universe''.
1012:
1013:
1014:
1015:
1016:
1017:
1018: \appendix
1019: \section{Stau Masses \& Mixings}
1020: \label{app:StauMassMatrix}
1021:
1022: Let us briefly introduce our conventions concerning the masses and mixings of the stau.
1023: We assume that there is no mixing between different slepton generations and take $\mu$ and $A^\tau$ to be real parameters,
1024: such the stau mass matrix can be written as
1025: \begin{align}
1026: \hspace{-3mm}
1027: \mathcal{M}_{\sltau}^2&=\begin{pmatrix}
1028: m_{\stau_{\mathrm{L\,\text{soft}}}}^2 + (\sin^2{\theta_\mathrm{W}}-\frac{1}{2}) M_Z^2 \cos{2\beta}+m_\tau^2 & -m_\tau (A^{\tau}
1029: + \mu \tan\beta) \nonumber\\
1030: -m_\tau (A^{\tau} + \mu \tan\beta) & m_{\stau_{\mathrm{R\,\text{soft}}}}^2 -\sin^2{\theta_\mathrm{W}} M_Z^2 \cos{2\beta} +m_\tau^2
1031: \end{pmatrix}\qquad \label{eq:StauMassMatrix}
1032: \\
1033: &\equiv\begin{pmatrix}
1034: m_{\stau_{\mathrm{L}}}^2 & m_{\stau_{\mathrm{LR}}}^2\\
1035: m_{\stau_{\mathrm{LR}}}^2 & m_{\stau_{\mathrm{R}}}^2
1036: \end{pmatrix}\;.
1037: \end{align}
1038: A non-zero off-diagonal element $m_{\stau_{\mathrm{LR}}}^2$ leads to a mixing of the chiral states $\stau_\text{L}$ and $\stau_\text{R}$.
1039: We can diagonalize the stau mass matrix by an orthogonal transformation
1040: \begin{equation}
1041: \mathcal{O^T}\, \mathcal{M}_{\sltau}^2\, \mathcal{O}
1042: ~=~\begin{pmatrix}m_{\sltau[1]}^2 & 0\\0& m_{\sltau[2]}^2\end{pmatrix}\;.
1043: \end{equation}
1044: The mass eigenvalues are
1045: \begin{equation}
1046: m_{\sltau[1,2]}^2~ =~\frac{1}{2} \left[ m_{\stau_{\mathrm{L}}}^2 + m_{\stau_{\mathrm{R}}}^2 \mp
1047: \sqrt{(m_{\stau_{\mathrm{L}}}^2 - m_{\stau_{\mathrm{R}}}^2)^2
1048: + 4 m_{\stau_{\mathrm{LR}}}^4}\:\right]
1049: \;.
1050: \end{equation}
1051: The orthogonal $2\times2$ matrix $\mathcal{O}$ is parameterized by the
1052: stau left-right mixing angle $\theta_{\sltau}$, which relates the mass
1053: eigenstates and the chiral states,
1054: \begin{equation}\label{eq:StauMixing}
1055: \left(\begin{array}{c}
1056: \sltau[1] \\ \sltau[2]
1057: \end{array}\right)
1058: ~=~
1059: \begin{pmatrix}
1060: \cos{\theta_{\sltau}} & \sin{\theta_{\sltau}} \\ -\sin{\theta_{\sltau}} & \cos{\theta_{\sltau}}
1061: \end{pmatrix}
1062: \left(\begin{array}{c}\sltau[\mathrm{L}] \\ \sltau[\mathrm{R}]\end{array}\right)
1063: \;.
1064: \end{equation}
1065: The mixing angle $\theta_{\sltau}$ is given by
1066: \label{eq:StauMixingAngle}
1067: \begin{eqnarray}
1068: \cos{\theta_{\sltau}}
1069: & = &
1070: \frac{-m_{\stau_{\mathrm{LR}}}^2}{\sqrt{(m_{\stau_{\mathrm{L}}}^2
1071: - m_{\sltau[1]}^2)^2+m_{\stau_{\mathrm{LR}}}^4}} \;.
1072: \end{eqnarray}
1073:
1074:
1075: \enlargethispage{0.8cm}
1076:
1077: \bibliography{GauginoMediation}
1078: \bibliographystyle{NewArXiv}
1079:
1080:
1081: \end{document}
1082: