1: \documentclass[a4paper,11pt,twoside]{article}
2:
3: \usepackage{geometry}
4: \geometry{width=0.75\paperwidth,height=0.75\paperheight} % default is 0.7, 0.7
5: \geometry{marginratio={1:1,2:3}} % default is {2:3,2:3} for twoside, {1:1,2:3} for oneside
6: % \usepackage[notref,notcite]{showkeys}
7: \usepackage{amsmath}
8: \usepackage{amssymb}
9: \usepackage{graphicx}
10: \usepackage{epstopdf}
11: \usepackage{color}
12: % \usepackage[italian]{babel}
13: % \usepackage{mycommands}
14:
15: %\setlength{\hoffset}{-1in}
16: %\setlength{\voffset}{-1in}
17: %\setlength{\oddsidemargin}{30mm}
18: %\setlength{\textwidth}{150mm}
19: %\setlength{\textheight}{220mm}
20: %\setlength{\footnotesep}{0.9\baselineskip}
21: %\setlength{\footskip}{1.5\baselineskip}
22: %\setlength{\topmargin}{28mm}
23: %\setlength{\headheight}{0mm}
24: %\setlength{\headsep}{0mm}
25:
26: \newcommand{\capdef}{}
27: \newcommand{\mycaption}[2][\capdef]{\renewcommand{\capdef}{#2}%
28: \caption[#1]{{\itshape #2}}}
29: \makeatletter
30: \renewcommand{\fnum@table}{\textbf{\tablename~\thetable}}
31: \renewcommand{\fnum@figure}{\textbf{\figurename~\thefigure}}
32: \makeatother
33:
34: \newlength{\myem}
35: \settowidth{\myem}{m}
36: % \newcommand{\sep}[1]{\makebox[\myem][c]{#1}}
37: \newcommand{\sep}[1]{#1}
38: \newcounter{mysubequation}[equation]
39: \renewcommand{\themysubequation}{\alph{mysubequation}}
40: \newcommand{\mytag}{\stepcounter{mysubequation}%
41: \tag{\theequation\protect\sep{\themysubequation}}}
42: \newcommand{\globallabel}[1]{\refstepcounter{equation}\label{#1}}
43:
44: \makeatletter
45: \renewcommand{\section}{\@startsection{section}{1}{0em}%
46: {-3.5ex \@plus -1ex \@minus -.2ex}%
47: {2.3ex \@plus.2ex}%
48: {\normalfont\large\bfseries}}
49: \renewcommand{\subsection}{\@startsection{subsection}{2}{0em}%
50: {-3.25ex\@plus -1ex \@minus -.2ex}%
51: {1.5ex \@plus .2ex}%
52: {\normalfont\bfseries}}
53: \renewcommand{\subsubsection}%
54: {\@startsection{subsubsection}{3}{0em}%
55: {-3.25ex\@plus -1ex \@minus -.2ex}%
56: {1.5ex \@plus .2ex}%
57: {\normalfont\itshape}}
58: \makeatother
59:
60: \newcommand{\TeV}{\,\mathrm{TeV}}
61: \newcommand{\GeV}{\,\mathrm{GeV}}
62: \newcommand{\MeV}{\,\mathrm{MeV}}
63: \newcommand{\keV}{\,\mathrm{keV}}
64: \newcommand{\eV}{\,\mathrm{eV}}
65: \newcommand{\ecm}{e\,\mathrm{cm}}
66: \newcommand{\Fig}[1]{Fig.~\ref{fig:#1}}
67: \newcommand{\fig}[1]{fig.~\ref{fig:#1}}
68: \newcommand{\Figs}[1]{Figs.~\ref{fig:#1}}
69: \newcommand{\figs}[1]{figs.~\ref{fig:#1}}
70: \newcommand{\Eq}[1]{Eq.~(\ref{eq:#1})}
71: \newcommand{\eq}[1]{eq.~(\ref{eq:#1})}
72: \newcommand{\Eqs}[1]{Eqs.~(\ref{eq:#1})}
73: \newcommand{\eqs}[1]{eqs.~(\ref{eq:#1})}
74: \newcommand{\vev}[1]{\left\langle #1\right\rangle}
75: \newcommand{\VeV}[2][]{#1\langle #2 #1\rangle} %% \big\Big\bigg\Bigg
76: \newcommand{\ord}[1]{\mathcal{O}\left( #1 \right)}
77: \newcommand{\ordfrac}[2]{\mathcal{O}\fracwithdelims{(}{)}{#1}{#2}}
78: \newcommand{\fracwithdelims}[4]{\left#1 \frac{#3}{#4} \right#2}
79: \newcommand{\interskip}{\medskip}
80: \DeclareMathOperator{\im}{Im}
81: \DeclareMathOperator{\re}{Re}
82: \DeclareMathOperator{\tr}{Tr}
83: \DeclareMathOperator{\ph}{Ph}
84: \DeclareMathOperator{\diag}{Diag}
85: \newcommand{\com}[1]{{\color{red} \sffamily (#1)}}
86: \newcommand{\omi}[1]{{\small\ttfamily\bfseries [#1]}}
87: \newcommand{\add}[1]{{\bfseries #1}}
88:
89: \newcommand{\tilgu}{{\tilde g}_u}
90: \newcommand{\tilgd}{{\tilde g}_d}
91: \newcommand{\tilgup}{{\tilde g}^\prime_u}
92: \newcommand{\tilgdp}{{\tilde g}^\prime_d}
93: \newcommand{\tilgus}{{\tilde g}^*_u}
94: \newcommand{\tilgds}{{\tilde g}^*_d}
95: \newcommand{\tilgups}{{\tilde g}^{\prime *}_u}
96: \newcommand{\tilgdps}{{\tilde g}^{\prime *}_d}
97:
98:
99: \renewcommand{\abstractname}{\large\bfseries Abstract}
100:
101: %-----------------------------------------------------
102:
103: \newcommand{\SISSA}{SISSA/ISAS and INFN, I--34013 Trieste, Italy}
104: \newcommand{\CERN}{CERN, Department of Physics, Theory Division \\
105: CH--1211 Geneva 23, Switzerland}
106:
107:
108: %-----------------------------------------------------
109:
110: \newcommand{\preprintdate}{October 2005}
111: \newcommand{\preprintnumber}{
112: CERN--TH/2005--196 \\
113: SISSA 74/2005/EP
114: }
115: \newcommand{\hepnumber}{hep-ph/0510197}
116: \newcommand{\titletext}{Electric Dipole Moments in Split Supersymmetry}
117: \newcommand{\authortext}{\large G.F. Giudice$^{\, a}$ and A. Romanino$^{\, b}$
118: % \thanks{}
119: \medskip\\\em\normalsize
120: $\mbox{}^a$ \CERNÄ
121: \\[0.1\baselineskip]
122: $\mbox{}^b$ \SISSA}
123: \newcommand{\abstracttext}{We perform a quantitative study of the
124: neutron and electron electric dipole moments (EDM) in Supersymmetry,
125: in the limit of heavy scalars. The leading contributions arise at two
126: loops. We give the complete analytic result, including a new
127: contribution associated with $Z$--Higgs exchange, which plays an
128: important and often leading role in the neutron EDM. The predictions
129: for the EDM are typically within the sensitivities of the next
130: generation experiments. We also analyse the correlation between the
131: electron and neutron EDM, which provides a robust test of Split
132: Supersymmetry.}
133:
134:
135: %-----------------------------------------------------
136:
137: \title{
138: \normalsize
139: \begin{tabular}[t]{l}%\hepnumber\\
140: \preprintdate\end{tabular}
141: \hspace*{\fill}
142: \begin{tabular}[t]{l}\preprintnumber\end{tabular}
143: \vspace{3\baselineskip}\\\Large\bfseries\titletext\bigskip}
144: \author{\begin{minipage}[t]{0.8\textwidth}
145: \normalsize\centering\authortext
146: \end{minipage}}
147: \date{}
148:
149: \begin{document}
150:
151: \bigskip
152: \maketitle
153: \begin{abstract}\normalsize\noindent
154: \abstracttext
155: \end{abstract}\normalsize\vspace{\baselineskip}
156:
157: % \clearpage
158:
159: % \noindent
160:
161: \section{Introduction}
162:
163: The electric dipole moments (EDMs) of the Standard Model (SM) fermions
164: are powerful probes of physics beyond the SM. Once the strong CP
165: problem has been taken care of, the SM predictions for the EDMs
166: of quarks and leptons are at least 7 orders of magnitudes
167: below~\cite{Pospelov:05} the present experimental
168: limits~\cite{Harris:99a,Romalis:00a,Regan:02a}. The situation is
169: drastically different in supersymmetric extensions of the SM. The
170: supersymmetry-breaking terms involve many new sources of
171: CP-violation. Particularly worrisome are the phases associated, in the
172: universal and
173: flavour-diagonal case, to the invariants ${\rm arg}(A^* M_{\tilde g})$ and
174: ${\rm arg}(A^* B)$. Such
175: phases survive in the universal limit in which all the flavour
176: structure originates from the SM Yukawas. If these phases are of order
177: one, the electron and neutron EDMs induced at one-loop by
178: gaugino-sfermion exchange are typically (barring accidental
179: cancellations~\cite{Brhlik:98a}) a couple of orders of magnitude above
180: the limits~\cite{Abel:01a,Demir:03a,OPRS}, a difficulty which is
181: known as the supersymmetric CP problem.
182:
183: The Split limit
184: of the MSSM~\cite{AD,Giudice:04a,Arkani-Hamed:04a} does not present
185: a supersymmetric CP problem.
186: Heavy sfermions suppress the dangerous one-loop contributions to a
187: negligible level. Nevertheless, some phases survive below the
188: sfermion mass scale and, if they do not vanish for an accidental or a
189: symmetry reason, they give rise to EDMs that are
190: safely below the experimental limits, but sizeable enough to be well
191: within the sensitivity of the next generation of
192: experiments~\cite{Arkani-Hamed:04a}. Such contributions only arise at
193: the two-loop level, since the new phases appear in the
194: gaugino-Higgsino sector, which is not directly coupled to the SM
195: fermions.
196:
197: In this paper, we perform a quantitative study of the neutron and
198: electron EDMs in the limit of Split Supersymmetry. First, we compute
199: the different
200: contributions to the light quark and electron EDMs,
201: the only relevant CP-violating operators. Indeed, quark chromoelectric
202: dipoles and the gluon Weinberg operator~\cite{Weinberg} cannot be
203: generated at two loops. For the EDM,
204: the original CP-violation in the gaugino-Higgsino sector is
205: communicated to the SM fermions by gauge boson and Higgs exchanges,
206: specifically by i) $\gamma h$ , ii) $WW$, or iii) $Zh$ exchange. No
207: other possibilities are allowed at the two-loop level.
208:
209: The $\gamma h$ exchange has been widely studied in the literature in
210: several contexts~\cite{Barr:90a,Chang:98a,Chang:02a,Pilaftsis:02a}.
211: The case of Split Supersymmetry was considered
212: in ref.~\cite{Arkani-Hamed:04a}. The $WW$ exchange has also been
213: studied in different
214: limits~\cite{Marciano:86a,Kadoyoshi:96a,Deshpande:05a}. An exact
215: 2-loop computation has been performed in the context of Split
216: Supersymmetry in ref.~\cite{Chang:05a} (see also
217: ref.~\cite{Lopez-Mobilia:94a} for a computation in the context of a
218: two-Higgs doublet model). Our results, for which we give
219: explicit analytic expressions,
220: differ from those in ref.~\cite{Chang:05a}.
221: Moreover, we identify a third, important contribution due to $Zh$
222: exchange. The $Zh$ contribution is suppressed in the case
223: of the electron EDM by a $1-4\sin^2\theta_W$ factor, but it plays an
224: important role in the neutron EDM. In fact, the $Z h$ contribution is
225: always comparable and often larger than the $\gamma h$ one (which in
226: turn is tipically larger than the $WW$ contribution). We have also
227: recomputed the QCD renormalization effect, correcting a mistake present
228: in the previous literature.
229:
230: \section{General expressions for the EDMs}
231:
232: CP-violating phases can enter the effective Lagrangian below the sfermion mass
233: scale $\tilde
234: m$ through the Yukawa couplings (which are irrelevant for our study),
235: the $\mu$-parameter, the gaugino masses $M_i$,
236: $i=1,2,3$, or the Higgs-Higgsino-gaugino couplings $\tilgu$,
237: $\tilgd$, $\tilgup$, $\tilgdp$ in
238: \begin{equation}
239: \label{eq:couplings}
240: -\mathcal{L} =
241: \sqrt{2}\left( \tilgu H^\dagger \tilde W^a T_a \tilde H_u + \tilgup
242: Y_{H_u}H^\dagger \tilde B \tilde H_u + \tilgd H_c^\dagger
243: \tilde W^a T_a \tilde H_d + \tilgdp
244: Y_{H_d}H_c^\dagger \tilde B \tilde H_d \right) +{\rm h.c.},
245: \end{equation}
246: where $H_c = i\sigma_2 H^*$, $T_a$ are the $SU(2)$ generators, and
247: $Y_{H_u}=-Y_{H_d}=1/2$.
248: % We
249: %will consider a phase convention in which the
250: %$\mu$-term and the gaugino masses are real and positive.
251: The Higgs vev is in its usual form $\vev{H} = (0,v)^T$,
252: with $v\sim 174\GeV$. The gaugino and Higgsino mass parameters $M_{1,2}$ and $\mu$, and
253: the couplings $\tilgu$, $\tilgd$, $\tilgup$, $\tilgdp$ are in general complex.
254: However, only three phases are independent and
255: they are associated to the invariants $\phi_1 = \arg(\tilgups \tilgdps
256: M_1\mu)$,
257: $\phi_2 =
258: \arg(\tilgus \tilgds M_2\mu)$, $\xi
259: = \arg(\tilgu \tilgds \tilgdp \tilgups )$. The tree-level matching
260: with the full theory above $\tilde m$ gives $\arg(\tilgu) = \arg(\tilgup)$,
261: $\arg(\tilgd) = \arg(\tilgdp)$, and therefore $\xi = 0$, thus leaving only
262: two independent phases. Moreover, in
263: most models of supersymmetry breaking the phases of $M_1$ and $M_2$ are
264: equal, in which case there is actually only one CP-invariant.
265:
266: \interskip
267:
268: In terms of mass eigenstates, the relevant interactions are
269: \begin{multline}
270: \label{eq:interactions}
271: -\mathcal{L} = \frac{g}{c_W} \overline{\chi^+_i} \gamma^\mu
272: (G^R_{ij}P_R + G^L_{ij}P_L)\chi^+_j Z_\mu \\
273: + \left[ g \overline{\chi^+_i} \gamma^\mu (C^R_{ij}P_R +
274: C^L_{ij}P_L) \chi^0_j W^+_\mu +\frac{g}{\sqrt{2}} \overline{\chi^+_i}
275: (D^R_{ij}P_R +D^L_{ij}P_L) \chi^+_j h +\text{h.c.} \right] ,
276: \end{multline}
277: where
278: \globallabel{eq:mixings}
279: \begin{align}
280: &G^L_{ij} = V^{\phantom{\dagger}}_{iW^+}c_{W^+}V^\dagger_{W^+j} +
281: V^{\phantom{\dagger}}_{ih_u^+}c_{h_u^+}V^\dagger_{h_u^+j} &
282: -&{G^{R}_{ij}}^* = U^{\phantom{\dagger}}_{iW^-}c_{W^-}U^\dagger_{W^-j} +
283: U^{\phantom{\dagger}}_{ih^-_d}c_{h_d^-}U^\dagger_{h_d^-j} \mytag \\
284: &C^L_{ij} = -V^{\phantom{*}}_{iW^+} N^*_{jW_3} +\frac{1}{\sqrt{2}}
285: V^{\phantom{*}}_{ih^+_u}N^*_{jh^0_u} &
286: &C^R_{ij} = -U^*_{iW^-} N^{\phantom{*}}_{jW_3} -\frac{1}{\sqrt{2}}
287: U^*_{ih^-_d}N^{\phantom{*}}_{jh^0_d} \mytag \\
288: g&D^R_{ij} = \tilgu^* V_{ih^+_u}U_{jW^-} +\tilgd^*
289: V_{iW^+}U_{jh^-_d} & &D^L=(D^R)^\dagger . \mytag
290: \end{align}
291: In eq.~(\ref{eq:mixings}a), $c_f = T_{3f} -s_W^2 Q_f$ ($s_W^2\equiv
292: \sin^2\theta_W$) is
293: the neutral current coupling coefficient of the fermion $\tilde
294: f$ and, accordingly, $c_{W^\pm}=\pm \cos^2\theta_W$,
295: $c_{h_u^+,h_d^-}= \pm(1/2-s^2_W)$.
296: The matrices $U$, $V$, $N$ diagonalize the complex chargino and neutralino
297: mass matrices, $M_{+} = U^TM^D_{+} V$, $M_0 = N^T N^D_0 N$, where
298: $M^D_+ = \diag(M^+_1,M^+_2)\geq 0$, $M^D_0 =
299: \diag(M^0_1,\ldots,M^0_4)\geq 0$ and
300: \begin{equation}
301: \label{eq:charginos}
302: M_+ =
303: \begin{pmatrix}
304: M_2 & \tilgu v \\
305: \tilgd v & \mu
306: \end{pmatrix}, \quad M_0 =
307: \begin{pmatrix}
308: M_1 & 0 & -\tilgdp v/\sqrt{2} & \tilgup v/\sqrt{2} \\
309: 0 & M_2 & \tilgd v/\sqrt{2} & -\tilgu v/\sqrt{2} \\
310: -\tilgdp v/\sqrt{2} & \tilgd v/\sqrt{2} & 0 & -\mu \\
311: \tilgup v/\sqrt{2} & -\tilgu v/\sqrt{2} & -\mu & 0
312: \end{pmatrix}.
313: \end{equation}
314:
315: \interskip
316:
317: In Split Supersymmetry, fermion EDMs are generated only at two loops,
318: since charginos and neutralinos, which carry the information of CP violation,
319: are only coupled to gauge and Higgs bosons.
320: To identify all possible diagrams contributing to the EDM,
321: let us first consider the
322: case in which $M_{1,2},\mu \gg M_W$. After we integrate out charginos
323: and neutralinos at one-loop, we generate some effective couplings among
324: SM bosons. These can be described in terms of gauge-invariant, CP-violating
325: operators. There are 5 dimension-6 such operators: $\epsilon_{abc}
326: {\widetilde W}^a_{\mu\nu}W^{b\nu\rho}W^{c\mu}_{\rho}$, $H^\dagger H
327: {\widetilde W}^a_{\mu\nu}W^{a\mu\nu}$,
328: $H^\dagger H
329: {\widetilde B}_{\mu\nu}B^{\mu\nu}$,
330: $D_\mu H^\dagger D_\nu H{\widetilde B}
331: ^{\mu\nu}$,
332: $D_\mu H^\dagger T_a D_\nu H{\widetilde W}^{a\mu\nu}$, where $W^a_{\mu\nu}$
333: and $B_{\mu\nu}$ are the $SU(2)$ and $U(1)$ gauge strengths, and
334: ${\widetilde W}^a_{\mu\nu}$ and ${\widetilde B}_{\mu\nu}$ are their duals.
335: The effective couplings relevant to generate sizable two-loop contributions to the
336: EDM must contain 3 fields, with at least one photon and at most one Higgs
337: boson. The previously-listed operators induce only the effective couplings
338: $\gamma\gamma h$, $\gamma Z h$, and $\gamma WW$. Notice that
339: CP-violating couplings of the kind $\gamma \gamma \gamma$,
340: $\gamma \gamma Z$ and $\gamma ZZ$ are not generated (in particular,
341: the CP-violating operator $B_{\mu \nu}B^{\nu \rho}B^\mu_\rho$ identically
342: vanishes unless there are three different abelian gauge fields). The absence
343: of these couplings is also confirmed by an explicit one-loop calculation.
344: Once we insert the effective couplings in a loop, we obtain 3 different
345: diagrams contributing at the two-loop level to the EDM
346: of the light SM fermion $f$, shown in \Fig{feynman}. We therefore have
347: \begin{equation}
348: \label{eq:d}
349: d_f = d^{\gamma H}_f + d^{ZH}_f + d^{WW}_f ,
350: \end{equation}
351: \globallabel{eq:dipoles}
352: \begin{align}
353: d^{\gamma H}_f & = \frac{e Q_f\alpha^2}{4\sqrt{2}\pi^2s^2_W}
354: \im(D^R_{ii}) \frac{m_f M^+_i}{M_W m^2_H} f_{\gamma H}(r^+_{iH}) \mytag \\
355: d^{ZH}_f &= \frac{e \left(T_{3f_L} -2 s^2_W Q_f \right)
356: \alpha^2}{16\sqrt{2}\pi^2c^2_Ws^4_W} \im \left( D^R_{ij}G^R_{ji}
357: -D^L_{ij}G^L_{ji} \right) \frac{m_f M^+_i}{M_W m^2_H}
358: f_{ZH}(r^{\phantom{\dagger}}_{ZH},r^+_{iH},r^+_{jH}) \mytag \\
359: d^{WW}_f & = \frac{eT_{3f_L} \alpha^2}{8\pi^2s^4_W} \im \left(
360: C^L_{ij}C^{R*}_{ij} \right) \frac{m_f M^+_i M^0_j}{M^4_W}
361: f_{WW}(r^+_{iW},r^0_{jW}) , \mytag
362: \end{align}
363: In \eqs{dipoles}
364: the sum over the indexes $i,j$ is understood, $Q_f$
365: is the charge of the fermion $f$, $T_{3f_L}$ is the third component of
366: the weak isospin of its left-handed component. Also,
367: $r_{ZH} = (M_Z/m_H)^2$, $r^+_{iH} = (M^+_i/m_H)^2$, $r^+_{iW} =
368: (M^+_i/M_W)^2$, $r^0_{iW} = (M^0_i/M_W)^2$, where $m_H$ is the Higgs mass,
369: and the loop functions are given by
370: \globallabel{eq:loopfunctions}
371: \begin{align}
372: f_{\gamma H}(r) &= \int^1_0\frac{dx}{1-x}\,
373: j\left(0,\frac{r}{x(1-x)}\right) \mytag \\
374: f_{Z H}(r, r_1,r_2) &= \frac{1}{2}\int^1_0 \frac{dx}{x(1-x)}\,
375: j\left(r,
376: \frac{x r_1+(1-x) r_2}{x(1-x)}\right) \mytag \\
377: f_{WW}(r_1,r_2) &= \int^1_0\frac{dx}{1-x}\, j\left(0,
378: \frac{x r_1 + (1-x) r_2}{x(1-x)}\right) . \mytag
379: \end{align}
380: The symmetric loop
381: function $j(r,s)$ is defined recursively by
382: \begin{equation}
383: \label{eq:basicfunctions}
384: j(r) = \frac{r\log r}{r-1}, \quad j(r,s) = \frac{j(r)-j(s)}{r-s}.
385: \end{equation}
386: In the determination of $f_{ZH}$ we have used the symmetry of
387: $\im(D^R_{ij}G^R_{ji})M^+_i$ and $\im(D^L_{ij}G^L_{ji})M^+_i$ under
388: $i\leftrightarrow j$. \Eq{dipoles} differs from the result in
389: ref.~\cite{Chang:05a}. Analytic expressions for the functions in
390: \eqs{loopfunctions} are given in the appendix.
391:
392: \begin{figure}
393: \begin{center}
394: \includegraphics[width=\textwidth]{fey.eps}
395: \end{center}
396: \caption{Two loop contributions to the light SM fermion EDMs. The third diagram is for a down-type fermion $f$.}
397: \label{fig:feynman}
398: \end{figure}
399:
400: The parameters entering the expression of the quark EDM in \eqs{dipoles}
401: have to be evaluated at the chargino mass scale $M^+$. The renormalization to
402: the scale $\mu$
403: at which we evaluate the neutron EDM matrix element is determined
404: by the anomalous dimension of the operator $\bar q \sigma_{\mu \nu}
405: \gamma_5 q F^{\mu\nu}$. We find
406: \begin{equation}
407: \label{eq:eta}
408: d_q(\mu)=\eta_{\text{QCD}}~ d_q(M^+),\quad
409: \eta_{\text{QCD}}=\left[ \frac{\alpha_s(M^+)}{\alpha_s(\mu)}\right]^{\gamma /2b},
410: \end{equation}
411: where the $\beta$-function coefficient is $b=11-2n_q/3$ and $n_q$ is the number
412: of effective light quarks. The anomalous-dimension coefficient is
413: $\gamma =8/3$. To eliminate the quark mass dependence in the short-distance
414: contribution, it may be preferable to consider the ratio
415: $d_q/m_q$. Its renormalization is given by
416: \begin{equation}
417: \label{eq:eta2}
418: \frac{d_q}{m_q}(\mu)=\eta_{\text{QCD}}^{\gamma^\prime /\gamma}\,
419: \frac{d_q}{m_q}(M^+),
420: \end{equation}
421: with $\gamma^\prime =32/3$. For $\alpha_s (M_Z)=0.118\pm 0.004$ and
422: $\mu=1\GeV$ (the scale of the neutron mass), we find
423: $\eta_{\text{QCD}}= 0.75$ for $M^+ =1\TeV$ and $\eta_{\text{QCD}}= 0.77$ for
424: $M^+ =200\GeV$. The error in $\alpha_s (M_Z)$ gives an uncertainty
425: on $\eta_{\text{QCD}}$ of about 2\%, while we expect an uncertainty of
426: about 5\% from next-to-leading order effects.
427: Notice that the value of $\eta_{\text{QCD}}$ obtained here is different than
428: what computed in ref.~\cite{nano} and generally used in the literature.
429: Indeed, ref.~\cite{nano} incorrectly uses the opposite sign for $\gamma$.
430: Our result gives a QCD renormalization coefficient about a factor of 2 smaller
431: than usually considered, and it agrees with the recent findings of
432: ref.~\cite{degrassi}.
433:
434: To express the neutron EDM in terms of the quark EDMs, we use the
435: results of QCD sum-rule techniques~\cite{Pospelov:99a,Pospelov:00a}:
436: \begin{equation}
437: \label{eq:neutron}
438: d_n = (1\pm 0.5)\left[\frac{f_\pi^2 m_\pi^2}{(m_u+m_d) (225\MeV)^3}
439: \right] \left(\frac{4}{3} d_d -\frac{1}{3} d_u \right) ,
440: \end{equation}
441: where $f_\pi \approx 92\MeV$ and we have neglected the
442: contribution of the quark chromoelectric
443: dipoles, which does not arise at the two-loop level in the heavy-squark
444: mass limit. Note that $d_n$ depends on the light quark masses only through
445: the ratio $m_u/m_d$, for which we take the value $m_u/m_d = 0.553\pm
446: 0.043$.
447:
448: \section{Expansions in the heavy-chargino limit}
449:
450: We now discuss the result in the limit in which the $R$-symmetry
451: breaking scale, determining gaugino and Higgsino masses, is
452: larger than $M_Z$ and $m_H$. A leading-order perturbative expansion
453: of \eq{dipoles} in powers of $|M_{1,2}\mu|/M_Z^2$ and $|M_{1,2}\mu|/m_H^2$
454: (keeping all orders in $|M_{1,2}/\mu|$ and in $M_Z/m_H$) gives
455: \globallabel{eq:expansion}
456: \begin{align}
457: d^{\gamma H}_f & \simeq \frac{eQ_f\alpha m_f}{8\pi^3}
458: \frac{\tilgu \tilgd}{ M_2 \mu} \sin \phi_2 F_{\gamma H}\left( \frac{M^2_2}{\mu^2},\frac{M_2\mu}{m_H^2}\right) \mytag \\
459: d^{Z H}_f & \simeq \frac{e\left(T_{3f_L} -2s^2_WQ_f \right)
460: \alpha m_f}{16\pi^3
461: c^2_W }\frac{\tilgu \tilgd}{ M_2 \mu} \sin \phi_2
462: F_{Z H} \left(\frac{M_Z^2}{m_H^2}, \frac{M^2_2}{\mu^2},\frac{M_2\mu}{m_H^2}
463: \right) \mytag \\
464: d^{WW}_f & \simeq \frac{e T_{3f_L} \alpha m_f}{16\pi^3
465: s^2_W } \left[ \frac{\tilgu \tilgd}{ M_2 \mu} \sin \phi_2
466: F^{(2)}_{WW}\left( \frac{M^2_2}{\mu^2},\frac{M_2\mu}{m_H^2}\right) + \frac{\tilgup \tilgdp} {M_1 \mu} \sin \phi_1
467: F^{(1)}_{WW}\left( \frac{M^2_1}{\mu^2},\frac{M_1\mu}{m_H^2}\right) \right]
468: , \mytag
469: \end{align}
470: where ${\tilde g}_{u,d}$, ${\tilde g}_{u,d}^\prime$, $M_{1,2}$ and $\mu$ now
471: indicate the absolute value of the corresponding
472: quantity and the functions $F_{\gamma H}$, $F_{ZH}$, $F_{WW}$ are
473: given in the appendix. As long as $M_1/M_2<1$ (as, for instance, in
474: the case of gaugino masses
475: unifying at the GUT scale), the second term in eq.~(\ref{eq:expansion}c)
476: is suppressed with respect to the first term and numerically it is not
477: very significant.
478: Notice that eqs.~(\ref{eq:expansion}) explicitly exhibit
479: the dependence on the two CP-violating invariants $|\tilgu \tilgd / M_2 \mu|
480: \sin \phi_2$ and $|\tilgup \tilgdp / M_1 \mu|
481: \sin \phi_1$. Because of the suppression of the second term in
482: eq.~(\ref{eq:expansion}c), both the electron and neutron EDM are mostly
483: characterized by a single invariant.
484:
485: While eqs.~(\ref{eq:expansion}a,b) can be obtained from an expansion
486: at the first order in $v/M_{1,2}$, eq.~(\ref{eq:expansion}c) arises only at
487: the second order. This is because the origin of $d^{WW}_f$ can be
488: traced back to the vertices $\tilde W^{\mu\nu} D_\mu H D_\nu H$,
489: $\tilde B^{\mu\nu} D_\mu H D_\nu H$ in the unbroken electroweak symmetry
490: phase, requiring two insertions of the Higgs vev.
491: The additional $M/v$ factor in eq.~(\ref{eq:dipoles}c) leads to
492: a contribution to the EDM which is parametrically of the same order
493: of $d^{\gamma H}_f$ and $d^{ZH}_f$ in the $v/M$ expansion. Notice also that
494: the coefficients of the potentially large logarithms of $|M_{1,2}\mu|
495: /m_H^2$ correspond to the anomalous dimensions that mix the EDM operator
496: with the CP-violating dimension-6 operators obtained from integrating out the
497: supersymmetric particles well above the weak scale.
498:
499: The relative importance of the three contributions to $d_f$ in \eq{d}
500: can be estimated from \eqs{expansion}. Let us consider for definitess
501: the case $M_2=\mu$. By keeping only terms enhanced by a large
502: $\log(M_2\mu/m^2_H)$ in the expressions~(\ref{eq:Fexpansion}) for
503: $F_{\gamma H}$, $F_{ZH}$, $F^{(2)}_{WW}$, we obtain $F_{ZH} \approx
504: F_{\gamma H}(-s_W^2 +3/4)$ and $F^{(2)}_{WW} \approx -F_{\gamma
505: H}/4$. As a consequence, we find
506: \begin{equation}
507: \label{eq:largelog2}
508: \frac{d_f^{ZH}}{d_f^{\gamma H}} \approx \frac{(T_{3f_L}-2s^2_WQ_f)(3-4s_W^2)}
509: {8c^2_WQ_f} \quad \frac{d_f^{WW}}{d_f^{\gamma H}} \approx
510: -\frac{T_{3f_L}}{8s^2_WQ_f} \quad \text{($M_2=\mu$)},
511: \end{equation}
512: where in the expression for $d_f^{WW}$ we have neglected the subleading
513: second term in
514: eq.~(\ref{eq:expansion}c).
515: Numerically, \eq{largelog2} gives $d_e^{ZH} \approx 0.05\,d_e^{\gamma
516: H}$, $d_e^{WW} \approx -0.3 \,d_e^{\gamma H}$ and $d_n^{ZH} \approx
517: d_n^{\gamma H}$, $d_n^{WW} \approx -0.7 \,d_n^{\gamma H}$. These
518: simple estimates show the importance of the $ZH$ contribution to the
519: neutron EDM. A detailed numerical analysis of the relative
520: importance of the different contributions to the electron and neutron
521: EDMs is given in the next Section. The qualitative estimates above are
522: in a remarkably good agreement in the large $M_2=\mu$ limit.
523:
524:
525: \section{Numerical analysis}
526:
527: We now perform a numerical analysis of the full results for the EDM in
528: \eqs{dipoles}. We consider a standard unified framework for the
529: gaugino masses at the GUT scale. By using the RGEs given
530: in refs.~\cite{Giudice:04a,Arvanitaki:04a}, the parameters in \eqs{dipoles}
531: can be expressed in terms of the single phase $\phi\equiv\phi_2$ and
532: the four following positive parameters: $M_2$, $\mu$ (evaluated at
533: the low-energy scale), $\tan\beta$, and
534: the sfermion mass scale $\tilde m$.
535: In first approximation, the dipoles depend on
536: $\beta$ and $\phi$ through an overall factor $\sin2\beta\sin\phi$.
537: Therefore, in order to maximize the effect,
538: we choose $\tan\beta = 1$ and the phase $\phi$ evaluated at the low-energy
539: scale such that $\sin\phi = 1$. Notice that if the ratio $|\mu /M_2|$
540: is much larger or much smaller than one, the renormalization effects
541: which mixes $\mu$ and $M_2$ (peculiar of Split Supersymmetry) tend to
542: suppress the effective phase. In this case, a maximal CP violation can only
543: be achieved with very particular choices of the initial values for the higgsino
544: and gaugino masses. Once we have fixed $\sin 2\beta \sin \phi =1$,
545: we are then left with
546: the three dimensionful parameters $M_2$, $\mu$,
547: $\tilde m$. The overall sfermion scale $\tilde m$ enters only
548: logarithmically through the RGE equations for ${\tilde g}_{u,d}$,
549: ${\tilde g}_{u,d}^\prime$. We choose to present the
550: results as contour plots in the $M_2$--$\mu$ plane and set $\tilde m =
551: 10^9\GeV$, which is consistent with the cosmological bounds given in
552: ref.~\cite{Pierce}.
553:
554: \Fig{edm} shows the prediction for the electron EDM, the neutron EDM,
555: and their ratio $d_n/d_e$. The red thick line corresponds to the
556: present experimental limits $d_e< 1.6\times
557: 10^{-27}\ecm$~\cite{Regan:02a}, while the limit $d_n<0.63\times
558: 10^{-27}\ecm$~\cite{Harris:99a} does not pose a constraint on the
559: parameters shown in \Fig{edm}. In the Split limit, and assuming
560: gaugino mass unification, all EDMs are controlled by a single
561: phase. The results for $d_e$ and $d_n$ shown in \Fig{edm} scale approximately
562: linearly with $\sin 2\beta \sin \phi$.
563:
564: \begin{figure}
565: \begin{center}
566: \includegraphics[width=0.8\textwidth]{edm.eps}
567: \end{center}
568: \caption{Prediction for $d_n$, $d_e$, and their ratio $d_n/d_e$. We
569: have chosen $\tan\beta = 1$, $\sin\phi = 1$, and $\tilde m =
570: 10^9\GeV$. The results for $d_n$ and $d_e$ scale approximately linearly with
571: $\sin 2\beta \sin\phi$, while the ratio is fairly independent of $\tan\beta$,
572: $\sin\phi$ and $\tilde m$.
573: The red thick line corresponds to the present experimental
574: limit $d_e< 1.6\times 10^{-27}\ecm$~\cite{Regan:02a}.}
575: \label{fig:edm}
576: \end{figure}
577:
578: A robust test of Split Supersymmetry can be performed if both the
579: electron and the neutron EDM are measured. Indeed,
580: in the ratio $d_n/d_e$ the dependence on $\sin\phi$,
581: $\tan\beta$ and $\tilde m$ approximately cancels out.
582: This can be easily understood from eqs.~(\ref{eq:expansion}) which
583: show that, as long as the chargino and neutralino masses are sufficiently
584: larger than $M_Z$, the only dependence of $d_n/d_e$ on ${\tilde g}_{u,d}$,
585: ${\tilde g}_{u,d}^\prime$, and $\sin \phi_{1,2}$ comes from the
586: existence of the $M_1$-dependent term in $d_f^{WW}$, see
587: eq.~(\ref{eq:expansion}c). This term, as previously discussed, is numerically
588: small.
589: Nevertheless, because of the different loop functions associated to
590: the different contributions, the ratio $d_n/d_e$ varies by a
591: $\ord{100\%}$ factor when the $M_2$ and $\mu$ are varied in the range
592: spanned in the Figures. Still, the variation of $d_n/d_e$
593: is comparable with the
594: theoretical uncertainty on the determination of $d_n$ in terms of
595: quark EDMs in \eq{neutron} due to the hadronic matrix element, and is
596: significantly smaller than the variation in the ordinary MSSM prediction,
597: even in the case of universal phases (see
598: e.g.~ref.~\cite{AL}\footnote{Note that the $ZH$ contribution is missing in
599: the analysis of the Split Supersymmetry case in ref.~\cite{AL}, which leads
600: to a stronger correlation between $d_e$ and $d_n$.}). On the other
601: hand, the usual tight correlation between the
602: electron and muon EDMs, $d_\mu/d_e = m_\mu/m_e$ persists.
603:
604: \begin{figure}
605: \begin{center}
606: \includegraphics[width=0.8\textwidth]{con.eps}
607: \end{center}
608: \caption{Relative importance of the different contributions to the
609: EDMs. We have chosen $\tan\beta = 1$, $\sin\phi = 1$ and $\tilde m =
610: 10^9\GeV$, but the result depends only very weakly on this choice.}
611: \label{fig:con}
612: \end{figure}
613:
614: \Fig{con} shows the relative importance of the different contributions
615: to the EDMs. For the same reason explained above, the results shown in
616: \Fig{con} are fairly independent of $\sin\phi$,
617: $\tan\beta$ and $\tilde m$.
618: As anticipated, the $ZH$ contribution to the electron EDM
619: is suppressed by the $T_{3f_L}-2s^2_WQ_f$ factor. On the other hand,
620: the corresponding contribution to $d_n$ is always important and
621: represents the largest contribution in a significant portion of the
622: parameter space shown in \Fig{con}. The $WW$ contribution is also
623: sizable, especially in the case of the neutron EDM, but is typically
624: smaller than the $ZH$ or $\gamma H$ ones. While the $ZH$ and $\gamma
625: H$ contributions always add constructively, the $WW$ contribution has
626: an opposite sign. However, also due to the $ZH$ contribution, its
627: size is not large enough to flip the
628: sign of the overall electron or neutron EDM.
629:
630: \section{Conclusions}
631:
632: In summary, we performed a quantitative study of the neutron and
633: electron EDMs in Split Supersymmetry. Clearly, our results also apply
634: to the MSSM in any limit in which the contributions involving sfermion
635: exchange are suppressed. Our result for the $WW$ exchange differs from
636: the one in the literature. Moreover, we find a new contribution
637: associated to $Zh$ exchange, which plays an important and often
638: leading role in the neutron EDM. We have also given the correct value
639: for the QCD renormalization of the quark EDM.
640: We performed an analytical and
641: numerical analysis of our results, summarized in \Figs{edm}
642: and~\ref{fig:con}. The correlation between the electron and neutron
643: EDMs is found to be stronger than in standard supersymmetric
644: scenarios, and it may become a crucial experimental test for
645: Split Supersymmetry. Still, we find an $\ord{100\%}$ variation of $d_n/d_e$ in
646: the parameter space we considered, which is comparable with the
647: hadronic uncertainty on the determination of $d_n$ in terms of quark
648: EDMs. The results summarized in \Fig{edm} are quite promising in the
649: light of the expected impressive improvement of the experimental
650: sensitivities in the years to come and represent one of the most
651: relevant windows on Split Supersymmetry.
652:
653: \section*{Acknowledgments}
654:
655: We thank A. Ritz for useful discussions and the authors
656: of ref.~\cite{Chang:05a} for communications. A.R. wishes to thank
657: the CERN Theory Division for the kind hospitality during the first
658: stage of this work.
659:
660: \section*{Appendix}
661:
662: We write below the analytic expression of the loop functions in \eqs{dipoles} and~(\ref{eq:loopfunctions}).
663: \begin{equation}
664: f_{\gamma H}(r) =\frac{1}{\sqrt{1-4 r}}\left[\log r \; \log
665: \frac{\sqrt{1-4 r}-1}{\sqrt{1-4 r}+1}
666: +\text{Li}_2\left(\frac{2}{1- \sqrt{1-4r}}
667: \right)
668: -\text{Li}_2\left(\frac{2}{1+ \sqrt{1-4r}}\right) \right]
669: \end{equation}
670: \begin{multline}
671: f_{ZH}(r,r_1,r_2) =\frac{1}{r-1}\left\{
672: g(r,r_1,r_2)-g(1,r_1,r_2) \right. \\
673: \left. -\frac{\log r}{2(x_1-x_2)} \left[\log\left(1-\frac{1}{x_1}\right) -\log\left(1-\frac{1}{x_2}\right)\right]\right\}
674: \end{multline}
675: \begin{multline}
676: f_{WW}(r_1,r_2) = \frac{1}{\hat y_1 -\hat y_2}\left\{
677: (r_2-\hat y_1) \left[ \log r_2 \log \left(1-\frac{r_2}{\hat y_1}\right)
678: - \log r_1 \log \left(1-\frac{r_1}{\hat y_1}\right)
679: \right. \right.\\
680: \left. +
681: \text{Li}_2\left( \frac{r_2}{\hat y_1}\right)-
682: \text{Li}_2\left( \frac{r_1}{\hat y_1}\right) \right] +
683: (r_2-\hat y_2) \left[ \log r_1 \log \left(1-\frac{r_1}{\hat y_2}\right) \right. \\
684: \left.\left. - \log r_2 \log \left(1-\frac{r_2}{\hat y_2}\right) +
685: \text{Li}_2\left( \frac{r_1}{\hat y_2}\right)-
686: \text{Li}_2\left( \frac{r_2}{\hat y_2}\right) \right] \right\}
687: \\
688: +\frac{1}{\hat x_1-\hat x_2}\left\{ \hat x_1 \left[
689: \text{Li}_2\left(\frac{1}{1-\hat x_1}\right) -
690: \text{Li}_2\left(\frac{1}{\hat x_1}\right) \right] -\hat x_2 \left[
691: \text{Li}_2\left(\frac{1}{1-\hat x_2}\right) -
692: \text{Li}_2\left(\frac{1}{\hat x_2}\right) \right] \right\}
693: \end{multline}
694: \begin{multline}
695: g(r,r_1,r_2) = \frac{r_1-r_2}{2 \left( y_1-y_2\right)} \left\{ \log
696: r_1 \left[ \log \left( 1- \frac{r_1}{y_1}\right) - \log \left( 1-
697: \frac{r_1}{y_2}\right) \right] \right. \\
698: \left.
699: -\log r_2 \left[ \log
700: \left( 1- \frac{r_2}{y_1}\right) - \log \left( 1-
701: \frac{r_2}{y_2}\right) \right]
702: + \text{Li}_2\left(\frac{r_1}{y_1}\right)-\text{Li}_2
703: \left(\frac{r_2}{y_1}\right)-\text{Li}_2\left(\frac{r_1}
704: {y_2}\right)+\text{Li}_2\left(\frac{r_2}{y_2}\right) \right\} \\
705: +\frac{1}{2(x_1-x_2)}\left[ \text{Li}_2\left(\frac{1}{1-x_1}\right) -
706: \text{Li}_2\left(\frac{1}{1-x_2}\right)
707: -\text{Li}_2\left(\frac{1}{x_1}\right) +
708: \text{Li}_2\left(\frac{1}{x_2}\right)\right]
709: \end{multline}
710: where
711: \globallabel{eq:roots}
712: \begin{align}
713: x_{1,2} &= \frac{1}{2r} \left[ r-r_1+r_2 \mp\sqrt{(r-r_1+r_2)^2-4rr_2}\right] &
714: \hat x_i &= x_i|_{r=1} \mytag \\
715: y_{1,2} &=r_2+(r_1-r_2)x_{1,2} & \hat y_i &= y_i|_{r=1}. \mytag
716: \end{align}
717: In the physical region, the functions $f_{\gamma H}$, $f_{ZH}$ and $f_{WW}$
718: do not develop imaginary parts.
719:
720: In order to study the case of chargino masses larger than $M_Z$, or
721: $r_1,r_2\gg 1$, it is useful to switch to the variables $R =
722: \sqrt{r_1r_2}$, $\rho=r_1/r_2$ and expand in $1/R$. We then get
723: \globallabel{eq:fexpansion}
724: \begin{gather}
725: f_{\gamma H}(R) = \frac{2+\log R}{2R} + \ord{\frac{\log R}{R^2}}
726: \mytag \\
727: f_{ZH}(r,r_1,r_2) = a_{ZH}(\rho)\frac{\log R}{R}
728: +\frac{b_{ZH}(r,\rho)}{R}+ \ord{\frac{\log R}{R^2}} \mytag \\
729: f_{WW}(r_1,r_2) = a_{WW}(\rho)\frac{\log R}{R}
730: +\frac{b_{WW}(\rho)}{R}+ \ord{\frac{\log R}{R^2}} \mytag\end{gather}
731: where \globallabel{eq:fexpansioncoe}
732:
733: \begin{align}
734: a_{ZH}(\rho) &= \frac{\sqrt{\rho}}{2(\rho-1)}\;\log \rho \mytag \\
735: b_{ZH}(r,\rho) &= \frac{\sqrt{\rho}}{2(\rho-1)}\; \left[ \frac{r\log
736: r\log \rho}{1-r}-\text{Li}_2(1-\rho)+\text{Li}_2(1-1/\rho)\right]
737: \mytag \\
738: a_{WW}(\rho) &= \frac{\sqrt{\rho}}{(\rho-1)^2}\;\left(
739: \rho-1-\log \rho \right) \mytag \\
740: b_{WW}(\rho) &=
741: \frac{\sqrt{\rho}}{(\rho-1)^2} \;\left[ \rho-1+\frac{(\rho+1)}{2}\log
742: \rho +\text{Li}_2(1-\rho)- \text{Li}_2(1-1/\rho)\right] . \mytag
743: \end{align}
744:
745: From these expansions, the expressions for the
746: functions $F_{\gamma H}$, $F_{ZH}$,
747: $F^{(1)}_{WW}$, $F^{(2)}_{WW}$ in \eqs{expansion} follow:
748: \globallabel{eq:Fexpansion}
749: \begin{align}
750: F_{\gamma H}(\rho, R) &= -\frac{1}{2}\log R -1 +\frac{(\rho+1)\log
751: \rho} {4(\rho-1)} +\ord{\frac{\log R}{R}} \mytag \\
752: F_{ZH}(r,\rho,R)
753: &= A_{ZH}(\rho)\log R +B_{ZH}(r,\rho) + \ord{\frac{\log R}{R}} \mytag
754: \\
755: F^{(1)}_{WW}(\rho,R)& = A^{(1)}_{WW}(\rho)\log R
756: +B^{(1)}_{WW}(\rho)+\ord{\frac{\log R}{R}} \mytag \\
757: F^{(2)}_{WW}(\rho,R) &= A^{(2)}_{WW}(\rho)\log R +B^{(2)}_{WW}(\rho)
758: +\ord{\frac{\log R}{R}} , \mytag
759: \end{align}
760: where \globallabel{eq:Fexpansioncoe}
761: \begin{align}
762: A_{ZH}(\rho) &= \frac{(\rho-1) (2- \rho) - \rho \log
763: \rho}{4(\rho-1)^2} +\frac{s_W^2}{2} \mytag \\
764: \begin{split}
765: B_{ZH}(r,\rho) &= \frac{1}{{4(r-1)(\rho-1)^2}}\left\{
766: \left( 2-2r+r\log r \right) \left( \rho -1 \right)
767: \left[ \rho -2 -2 s_W^2 (\rho -1)\right] \right. \\ &
768: +(r-1) (\rho -1)
769: \left[ \frac{\rho}{2}+1-s_W^2 (\rho +1)\right] \log \rho
770: + r\rho \log r \log \rho \\ &
771: \left. +(r-1)
772: \rho\left[ \text{Li}_2(1-\rho)-\text{Li}_2(1-1/\rho)\right] \right\}
773: \end{split} \mytag \\
774: A^{(1)}_{WW}(\rho) &= \rho \; \frac{-\rho^2+1+2\rho\log
775: \rho}{8(\rho-1)^3}\mytag \\
776: B^{(1)}_{WW}(\rho) &= \rho\;
777: \frac{-4\rho(\rho-1)+(\rho^2-4\rho-1)\log \rho-4\rho
778: \left[\text{Li}_2(1-\rho)-\text{Li}_2(1-1/\rho)\right]}{16(\rho-1)^3}\mytag \\
779: A^{(2)}_{WW}(\rho) &= \rho\;\frac{(\rho-7)(\rho-1)+2(\rho+2)\log
780: \rho}{8(\rho-1)^3} \mytag \\
781: B^{(2)}_{WW}(\rho) &=
782: \rho\;\frac{4(\rho-4)(\rho-1)-(\rho^2+4\rho+7)\log \rho
783: -4(\rho+2)\left[ \text{Li}_2(1-\rho)-\text{Li}_2(1-1/\rho)
784: \right]}{16(\rho-1)^3}
785: . \mytag
786: \end{align}
787:
788:
789: \begin{thebibliography}{10}
790:
791: \bibitem{Pospelov:05}
792: M. Pospelov and A. Ritz,
793: \newblock Annals Phys. 318 (2005) 119, hep-ph/0504231.
794: %%CITATION = HEP-PH 0504231;%%
795:
796: \bibitem{Harris:99a}
797: P.G. Harris et~al.,
798: \newblock Phys. Rev. Lett. 82 (1999) 904.
799: %%CITATION = PRLTA,82,904;%%
800:
801: \bibitem{Romalis:00a}
802: M.V. Romalis, W.C. Griffith and E.N. Fortson,
803: \newblock Phys. Rev. Lett. 86 (2001) 2505, hep-ex/0012001.
804: %%CITATION = HEP-EX 0012001;%%
805:
806: \bibitem{Regan:02a}
807: B.C. Regan et~al.,
808: \newblock Phys. Rev. Lett. 88 (2002) 071805.
809: %%CITATION = PRLTA,88,071805;%%
810:
811: \bibitem{Brhlik:98a}
812: M. Brhlik, G.J. Good and G.L. Kane,
813: \newblock Phys. Rev. D59 (1999) 115004, hep-ph/9810457.
814: %%CITATION = HEP-PH 9810457;%%
815:
816: \bibitem{Abel:01a}
817: S. Abel, S. Khalil and O. Lebedev,
818: \newblock Nucl. Phys. B606 (2001) 151, hep-ph/0103320.
819: %%CITATION = HEP-PH 0103320;%%
820:
821: \bibitem{Demir:03a}
822: D.A. Demir et~al.,
823: \newblock Nucl. Phys. B680 (2004) 339, hep-ph/0311314.
824: %%CITATION = HEP-PH 0311314;%%
825:
826: \bibitem{OPRS}
827: K.A. Olive et~al.,
828: \newblock (2005), hep-ph/0506106.
829: %%CITATION = HEP-PH 0506106;%%
830:
831: \bibitem{AD}
832: N. Arkani-Hamed and S. Dimopoulos,
833: \newblock (2004), hep-th/0405159.
834: %%CITATION = HEP-TH 0405159;%%
835:
836: \bibitem{Giudice:04a}
837: G.F. Giudice and A. Romanino,
838: \newblock Nucl. Phys. B699 (2004) 65, hep-ph/0406088.
839: %%CITATION = HEP-PH 0406088;%%
840:
841: \bibitem{Arkani-Hamed:04a}
842: N. Arkani-Hamed et~al.,
843: \newblock Nucl. Phys. B709 (2005) 3, hep-ph/0409232.
844: %%CITATION = HEP-PH 0409232;%%
845:
846: \bibitem{Weinberg}
847: S.~Weinberg,
848: \newblock Phys. Rev. Lett. 63 (1989) 2333.
849: %%CITATION = PRLTA,63,2333;%%
850:
851: \bibitem{Barr:90a}
852: S.M. Barr and A. Zee,
853: \newblock Phys. Rev. Lett. 65 (1990) 21.
854: %%CITATION = PRLTA,65,21;%%
855:
856: \bibitem{Chang:98a}
857: D. Chang, W.Y. Keung and A. Pilaftsis,
858: \newblock Phys. Rev. Lett. 82 (1999) 900, hep-ph/9811202.
859: %%CITATION = HEP-PH 9811202;%%
860:
861: \bibitem{Chang:02a}
862: D. Chang, W.F. Chang and W.Y. Keung,
863: \newblock Phys. Rev. D66 (2002) 116008, hep-ph/0205084.
864: %%CITATION = HEP-PH 0205084;%%
865:
866: \bibitem{Pilaftsis:02a}
867: A. Pilaftsis,
868: \newblock Nucl. Phys. B644 (2002) 263, hep-ph/0207277.
869: %%CITATION = HEP-PH 0207277;%%
870:
871: \bibitem{Marciano:86a}
872: W.J. Marciano and A. Queijeiro,
873: \newblock Phys. Rev. D33 (1986) 3449.
874: %%CITATION = PHRVA,D33,3449;%%
875:
876: \bibitem{Kadoyoshi:96a}
877: T. Kadoyoshi and N. Oshimo,
878: \newblock Phys. Rev. D55 (1997) 1481, hep-ph/9607301.
879: %%CITATION = HEP-PH 9607301;%%
880:
881: \bibitem{Deshpande:05a}
882: N.G. Deshpande and J. Jiang,
883: \newblock Phys. Lett. B615 (2005) 111, hep-ph/0503116.
884: %%CITATION = HEP-PH 0503116;%%
885:
886: \bibitem{Chang:05a}
887: D. Chang, W.F. Chang and W.Y. Keung,
888: \newblock Phys. Rev. D71 (2005) 076006, hep-ph/0503055.
889: %%CITATION = HEP-PH 0503055;%%
890:
891: \bibitem{Lopez-Mobilia:94a}
892: R.~Lopez-Mobilia and T.~H.~West,
893: \newblock Phys.\ Rev.\ D {\bf 51} (1995) 6495.
894: %%CITATION = PHRVA,D51,6495;%%
895:
896: \bibitem{nano}
897: R.~Arnowitt, J.~L.~Lopez and D.~V.~Nanopoulos,
898: \newblock Phys. Rev. D42 (1990) 2423.
899: %%CITATION = PHRVA,D42,2423;%%
900:
901: \bibitem{degrassi}
902: G.~Degrassi, E.~Franco, S.~Marchetti and L.~Silvestrini,
903: \newblock (2005), hep-ph/0510137.
904: %%CITATION = HEP-PH 0510137;%%
905:
906: \bibitem{Pospelov:99a}
907: M. Pospelov and A. Ritz,
908: \newblock Phys. Rev. Lett. 83 (1999) 2526, hep-ph/9904483.
909: %%CITATION = HEP-PH 9904483;%%
910:
911: \bibitem{Pospelov:00a}
912: M. Pospelov and A. Ritz,
913: \newblock Phys. Rev. D63 (2001) 073015, hep-ph/0010037.
914: %%CITATION = HEP-PH 0010037;%%
915:
916: \bibitem{Arvanitaki:04a}
917: A.~Arvanitaki, C.~Davis, P.~W.~Graham and J.~G.~Wacker,
918: Phys.\ Rev.\ D {\bf 70} (2004) 117703, hep-ph/0406034.
919: %%CITATION = HEP-PH 0406034;%%
920:
921: \bibitem{Pierce}
922: A. Arvanitaki et~al.,
923: \newblock (2005), hep-ph/0504210.
924: %%CITATION = HEP-PH 0504210;%%
925:
926: \bibitem{AL}
927: S. Abel and O. Lebedev,
928: \newblock (2005), hep-ph/0508135.
929: %%CITATION = HEP-PH 0508135;%%
930:
931: \end{thebibliography}
932:
933: \end{document}