1: %X!TEX TS-program = pdflatex

2: \documentclass[a4paper,11pt]{article}

3: \usepackage{graphicx, rotating,amsmath}

4: \usepackage{ifpdf}

5: \ifpdf

6: \usepackage{hyperref, pdfsync, epstopdf}	% This is for pdftex

7: \else

8: \usepackage[dvips,bookmarks]{hyperref}	% This is for arXiv.org

9: \fi

10: \hypersetup{colorlinks,bookmarksopen,bookmarksnumbered,citecolor=verdes,

11: linkcolor=blus,pdfstartview=FitH,urlcolor=rossos}

12: \def\hhref#1{\href{http://arxiv.org/abs/#1}{#1}} % in bibliography

13: \def\mhref#1{\href{mailto:#1}{#1}}		% email on title page

14: \oddsidemargin 0cm  \evensidemargin 0cm

15: \topmargin -1cm  \textwidth 16cm  \textheight 22.9cm

16: \newcommand{\riga}[1]{\noalign{\hbox{\parbox{\textwidth}{#1}}}\nonumber}

17: \newcommand{\gE}{\gamma_{\rm E}}

18: \newcommand{\mub}{\bar{\mu}}

19: \newcommand{\Ord}{{\cal O}}

20: \newcommand{\eq}[1]{~{\rm (\ref{eq:#1})}}

21: \newcommand{\sys}[1]{~{\rm (\ref{sys:#1})}}

22: \newcommand{\scatola}[1]{\fbox{$\displaystyle #1$}}

23: \newcommand{\GeV}{\,{\rm GeV}}

24: \newcommand{\TeV}{\,{\rm TeV}}

25: \newcommand{\cm}{\,{\rm cm}}

26: \newcommand{\kmwe}{\,{\rm kmwe}}

27: \def\circa#1{\,\raise.3ex\hbox{$#1$\kern-.75em\lower1ex\hbox{$\sim$}}\,}

28: \newcommand{\pl}{p\hspace{-4.2pt}{\scriptstyle/}}

29: \newcommand{\qsl}{q\hspace{-4.9pt}{\scriptstyle /}}

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

31: 

32: \newcommand{\xxx}[1]{{\color{red}\bf[#1]}}

33: 

34: \newcommand{\NP}{Nucl. Phys.}

35: \newcommand{\PRL}{Phys. Rev. Lett.}

36: \newcommand{\PL}{Phys. Lett.}

37: \newcommand{\PR}{Phys. Rev.}

38: 

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

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

41: \newcommand{\beqa}{\begin{eqnarray}}

42: \newcommand{\eeqa}{\end{eqnarray}}

43: \newcommand{\no}{\nonumber}

44: 

45: \newcommand{\mb}[1]{\mbox{\boldmath $#1$}}

46: \newcommand{\DM}{{\rm DM}}

47: \newcommand{\AL}{{\cal L}}

48: \newcommand{\YBL}{{\cal B}-{\cal L}}

49: \newcommand{\YB}{{\cal B}}

50: \newcommand{\YL}{{\cal L}}

51: \font\tenrsfs=rsfs10 at 11pt

52: \font\sevenrsfs=rsfs7

53: \font\fiversfs=rsfs5

54: \newfam\rsfsfam

55: \textfont\rsfsfam=\tenrsfs

56: \scriptfont\rsfsfam=\sevenrsfs

57: \scriptscriptfont\rsfsfam=\fiversfs

58: \def\mathscr#1{{\fam\rsfsfam\relax#1}}

59: \def\Lag{\mathscr{L}}

60: \def\Amp{\mathscr{A}}

61: \def\Ham{\mathscr{H}}

62: \def\circa#1{\,\raise.3ex\hbox{$#1$\kern-.75em\lower1ex\hbox{$\sim$}}\,}

63: \makeatletter

64: \newcommand{\fig}[1]{~\ref{fig:#1}}

65: %

66: % formato bibliografico standard

67: %

68: %\art[hep-ph/yymmnnn]{autori}{rivista}{numero}{pagina}{anno}

69: \def\art{\@ifnextchar[{\eart}{\oart}}

70: \def\eart[#1]#2#3#4#5#6{{\rm #2}, {#3 #4} {\rm (#6) #5} [{\hhref{#1}}]}

71: %\def\eart[#1]#2#3#4#5#6{{\rm #2}, {\em #1}}

72: \def\hepart[#1]#2{{\rm #2, \hhref{#1}}}

73: \newcommand{\oart}[5]{{\rm #1}, {\em #2 \rm #3} {\rm (#5) #4}}

74: \newcommand{\y}{{\rm and} }

75: %

76: % definizione della macro EQNSYSTEM

77: %

78: \newcounter{alphaequation}[equation]

79: %\def\thealphaequation{\theequation\alph{alphaequation}}

80: \def\thealphaequation{\theequation\hbox to

81: 0.6em{\hfil\alph{alphaequation}\hfil}}

82: % MODIFICATA PER DARE UNA DIMENSIONE UGUALE AD UN 1em AD OGNI LETTERA

83: \def\eqnsystem#1{

84: \def\@eqnnum{{\rm (\thealphaequation)}}

85: %

86: \def\@@eqncr{\let\@tempa\relax \ifcase\@eqcnt \def\@tempa{& & &} \or

87:   \def\@tempa{& &}\or \def\@tempa{&}\fi\@tempa

88:   \if@eqnsw\@eqnnum\refstepcounter{alphaequation}\fi

89: \global\@eqnswtrue\global\@eqcnt=0\cr}

90: %

91: \refstepcounter{equation} \let\@currentlabel\theequation \def\@tempb{#1}

92: \ifx\@tempb\empty\else\label{#1}\fi

93: %

94: \refstepcounter{alphaequation}

95: \let\@currentlabel\thealphaequation

96: %

97: \global\@eqnswtrue\global\@eqcnt=0 \tabskip\@centering\let\\=\@eqncr

98: $$\halign to \displaywidth\bgroup \@eqnsel\hskip\@centering

99: $\displaystyle\tabskip\z@{##}$&\global\@eqcnt\@ne

100: \hskip2\arraycolsep\hfil${##}$\hfil& \global\@eqcnt\tw@\hskip2\arraycolsep

101: $\displaystyle\tabskip\z@{##}$\hfil

102: \tabskip\@centering&\llap{##}\tabskip\z@\cr}

103: %

104: \def\endeqnsystem{\@@eqncr\egroup$$\global\@ignoretrue} \makeatother

105: 

106: \newcommand{\eV}{\,{\rm eV}}

107: 

108: \newcommand{\SU}{\,{\rm SU}}

109: 

110: \def\baselinestretch{1.08}

111: 

112: 

113: \usepackage{multicol}

114: \usepackage{color}

115: \definecolor{rosso}{cmyk}{0,1,1,0.4}

116: \definecolor{rossos}{cmyk}{0,1,1,0.55}

117: \definecolor{rossoc}{cmyk}{0,1,1,0.2}

118: \definecolor{blu}{cmyk}{1,1,0,0.3}

119: \definecolor{blus}{cmyk}{1,1,0,0.6}

120: \definecolor{bluc}{cmyk}{1,1,0,0.1}

121: \definecolor{verde}{cmyk}{0.92,0,0.59,0.25}

122: \definecolor{verdec}{cmyk}{0.92,0,0.59,0.15}

123: \definecolor{verdes}{cmyk}{0.92,0,0.59,0.4}

124: 

125: 

126: \begin{document}

127: \begin{center}

128: %FUP-TH/06-23

129: 

130: \bigskip\bigskip

131: 

132: \color{black}

133: %\vspace{0.2cm}

134: {\Huge\bf\color{rossos} New bounds on millicharged particles from cosmology}

135: \medskip

136: \bigskip\color{black}\vspace{0.5cm}

137: 

138: {{\large\bf A.\ Melchiorri}$^a$,

139: {\large\bf A.D.\ Polosa}$^b$,

140: {\large\bf A.\ Strumia}$^c$

141: }

142: \\[7mm]

143: {\it $^a$  Dipartimento di Fisica, Universit\`a di Roma I

144: Piazzale A.\ Moro 2, I-00185 Roma Italy}\\

145: {\it $^b$  INFN Roma I,

146: Piazzale A.\ Moro 2, I-00185 Roma Italy}\\

147: {\it $^c$ Dipartimento di Fisica dell'Universit{\`a} di Pisa and INFN, Pisa, Italia}\\

148: \end{center}

149: 

150: \bigskip

151: 

152: \centerline{\large\bf\color{blus} Abstract}

153: 

154: 

155: \begin{quote}\large\color{blus}

156: Particles with millicharge $q$ and sub-eV mass can be produced

157: in photon-photon collisions, 

158: distorting the energy spectrum of the Cosmic Microwave Background. 

159: We derive the conservative bound $q\circa{<} 10^{-7}e$

160: (as well as model-dependent bounds two orders of magnitude stronger),

161: incompatible with proposed interpretations of the PVLAS anomaly based

162: on millicharged production or on millicharged-mediated axion-like couplings.

163: 

164: 

165: %Particles with millicharge $q$ and sub-eV mass

166: %affect cosmology at sub-eV temperatures:

167: %photons no longer travel freely and this implies

168: %a safe and model-independent bound

169: %$q\circa{<} 10^{-7}e$,

170: %as well as model-dependent bounds two orders of magnitude stronger.

171: %This excludes interpretations of the PVLAS anomaly based

172: %on millicharged production or on millicharged-mediated axion-like couplings.

173: 

174: 

175: \color{black}

176: \end{quote}

177: 

178: 

179: 

180: 

181: %% because of the external field, in couples of fermions, or scalars, having a very small mass and electric charge. The produced pairs eventually get out of the apparatus thus generating dichroism (consider that pair production in magnetic fields depends on the polarization of the incoming photon). Millicharge particles in loops  could induce birefringence. 

182: % 

183: 

184: %%%Some explanations of dichroism in terms of new particles have been suggested.

185: %%Classically speaking, the rotation of the polarization plane could be due to a loss of power in the  polarization component of the beam parallel (orthogonal) to the external field, the orthogonal (parallel) one remaining unaltered. This would  generate a (counter)-clockwise rotation of the beam polarization vector.

186: %From a microscopical point of view, such a power loss has been interpreted, in Axion-Like-Particle (ALP) models~\cite{mpz}, as being due to the oscillation, induced by the magnetic field,  of some of the beam photons in a very weakly interacting particle, escaping from the apparatus.

187: %An effective interaction of the kind  $(\phi/M) F\cdot F$ or $(\phi/M) F\cdot\tilde{F}$ can explain the dichroism effect via the introduction of a scalar or

188: %pseudoscalar ALP field $\phi$:

189: %the $\phi$ particle is most effectively produced when the incoming photon has a polarization, relative to the external magnetic field, which maximizes the product $(F\cdot F)$ or $(F\cdot\tilde{F})$.

190: 

191: %

192: %Thorough investigations of the photon-splitting in the PVLAS magnetic field can be found in~\cite{mimmo}.

193: 

194: %In this paper we intend to explore the consequences  of one of the pictures proposed  in the literature, namely the one explaining the PVLAS effects introducing millicharged particles. Some of the eV photons in the beam can convert, because of the external field, in couples of fermions, or scalars, having a very small mass and electric charge. The produced pairs eventually get out of the apparatus thus generating dichroism (consider that pair production in magnetic fields depends on the polarization of the incoming photon). Millicharge particles in loops  could induce birefringence. 

195: 

196: 

197: 

198: %Many models  employ new millicharged particles that can

199: %either mediate  a coupling to a neutral axion-like

200: %scalar~\cite{Redondo} at loop level, or can be directly produced~\cite{Ringwald}.

201: 

202: 

203: 

204: \section{Introduction}

205: 

206: New particles $q$ with mass $m_q$ and charge $q\ll e$ (thereby called `millicharged')

207: can naturally arise in field-theory models, e.g.\

208: via small kinetic mixing of the photon with a new vector.

209: Distinctive aspects and constraints on millicharged particles are reviewed 

210: in~\cite{Davidson}.

211: 

212: Renewed interest in millicharged particles was prompted by an experimental result

213: claimed by the PVLAS collaboration~\cite{pvlas}: 

214: the polarization of a linearly polarized

215: laser beam  (photons with $E_\gamma\sim 1\eV$) rotates (`dichroism') and develops

216: a tiny elliptical component (`birefringence') after multiple passages 

217: trough a vacuum Fabry-Perot cavity containing a (rotating) $5.5$~Tesla magnetic field orthogonal to the beam direction.

218: %The statistical significance of the anomaly is tens of standard deviations:

219: %if it is not due to some unidentified systematics

220: %If the result will be confirmed, excluding systematics 

221: If confirmed, PVLAS calls for new physics, possibly pointing to the existence of new light particles that, interacting with photons, drain part of the beam energy in a polarization-dependent way.

222: 

223: Thanks to the external field, photons in the beam can non-perturbatively

224: convert into couples of new

225: light millicharged particles: one can fit PVLAS for 

226: charge $q\sim 3 \cdot 10^{-6}e$ and mass $m_q\sim 0.1\eV$~\cite{Ringwald,Foot}.

227: Alternatively, the PVLAS birefringence could be due to mixing, induced by the magnetic field,  

228: of laser photons with a new axion-like scalar $a$, coupled to photons by an effective interaction term 

229: $(a/4M)\, F_{\mu\nu} F^{\mu\nu}$

230: (a pseudo-scalar axion coupled as $(a/4M)\, F_{\mu\nu}\tilde{F}^{\mu\nu}$

231: produces birefringence with sign opposite to the one observed by PVLAS)~\cite{mpz},

232: and the PVLAS dichroism could be due to conversion of some photons in axions escaping from the apparatus.

233: The needed value of $M \sim 2\cdot 10^5 \GeV$ is excluded by $a$ production in stars~\cite{cast},

234: unless the effective coupling is mediated by a loop of light enough particles,

235: $1/M \sim \alpha q^2/v$, that therefore need to have 

236: a small millicharge $q \approx 10^{-6} e\cdot \sqrt{v/\eV}$, 

237: where $v$ is a loop function, related and comparable to $m_q$~\cite{Redondo}.\footnote{Other interpretations not involving millicharged particles have been put forward. 

238: %Ref.~\cite{Moha} proposes an axion-like coupling proportional to the vev of an extra scalar field,

239: %that should be non zero only at temperature $T\ll {\rm keV}$.

240: %However, to feel the effects of the thermal bath the extra field should be coupled to it,

241: %giving rise to new potential problems.

242: Ref.~\cite{antoniadis} suggested that the axion-like scalar might be replaced with

243: a massive vector $B_\mu$ with a Chern-Simons-like coupling to photons $A_\mu$.

244: However we do not see how the $A_\mu \leftrightarrow B_\mu$ oscillations proposed in~\cite{antoniadis}

245: can produce

246: the dichroism claimed by PVLAS.

247: It could be produced by emission of 

248: the longitudinal component of the extra vector, with couplings enhanced

249: by the polarization vector

250: $\epsilon_\mu \sim {q_\mu}/{m}$, but (just like in the axion case) this

251: possibility is not compatible with bounds from star cooling. 

252: Extra observables related to axion-like physics are discussed in~\cite{mimmo}.}

253: 

254: Light millicharged particles therefore are a key ingredient of 

255: proposed new physics that can fit the PVLAS anomaly.

256: Still, one needs to  circumvent the too strong constraints on millicharged  particles in~\cite{Davidson,} 

257: by designing models that exploit the fact that PVLAS involves ${\cal O}(\eV)$ energies,

258: while the constraints in~\cite{Davidson} come from physics at

259: ${\cal O}({\rm keV})$ energies (the temperature in the core of stars) or higher.

260: An energy-dependent form factor 

261: that suppresses millicharged couplings at high energies can be obtained in the following way.

262: (Although we use a basis that makes its presentation simpler, we are just outlining

263: the model of~\cite{Redondo}).

264: The gauge group is ${\rm U}(1)_{\rm em}\otimes{\rm U}(1) \otimes{\rm U}(1)'$

265: where the first U(1)$_{\rm em}$ is the usual electromagnetism, and new light particles

266: (either scalars or fermions)  are charged

267: under the last U(1)$'$, with charge $q'$, assumed to be comparable to $e$~\cite{Redondo}.

268: The propagator matrix of the three vectors is assumed to be~\cite{Redondo}

269: \beq\label{eq:matrix}

270: \begin{pmatrix}

271: & q^2 & \epsilon q^2 & 0 \\

272: &\epsilon q^2 & q^2 + m^2 & m^2 \cr

273: &0 & m^2 & q^2 + m^2

274: \end{pmatrix}\eeq

275: i.e.\ canonical kinetic terms, plus a small kinetic mixing between the first two U(1),

276: plus a vev that breaks the last two U(1) to their diagonal sub-group, 

277: giving a mass $\sim m$ to their symmetric combination.

278: At $q^2 \ll m^2$, the new light particles get a millicharge $q \sim \epsilon q'$ under the photon.

279: At $q^2 \gg m^2$ one can neglect $m^2$ and this millicharge disappears:

280: choosing $m\circa{<} \eV$ allows to avoid the bounds from higher-energy physics discussed in~\cite{Davidson}.\footnote{The Higgs $h^\prime$ that breaks ${\rm U}(1) \otimes{\rm U}(1)'$ 

281: to its diagonal subgroup

282: providing the vector mass $m$

283: gets a millicharge $q\sim \epsilon q'$ with no form factor that suppresses it at high energy:

284: therefore the model is excluded if $h^\prime$ is light enough to be produced in stars or during BBN.

285: Since $m_{h^\prime} \sim \lambda m/g^\prime$, a heavy enough $h^\prime$ needs a quartic scalar coupling $\lambda |h'|^4$ with $\lambda \gg 1 $.

286: This may appear crazy, and indeed such a non-perturbative Higgs coupling would be excluded

287: if $h'$ would break a non-abelian gauge group, because non-abelian vectors feel it

288: (for example the SM Higgs must be lighter than $4\pi M_Z/g\sim {\rm TeV}$).

289: However the model only involves abelian vectors, and

290: in the abelian case a big $\lambda$ remains confined to the Higgs sector, 

291: so that decoupling the physical Higgs is not impossible.

292: Indeed this is well known~\cite{Stuckelberg}: an abelian theory remains sensible if

293: gauge invariance is broken by

294: adding a vector mass term (i.e.\ no Higgs or infinitely heavy Higgs).

295: In conclusion, the models of~\cite{Redondo,Ringwald} an unusual but 

296: acceptable ingredient.}

297: 

298: Searches for this kind of millicharged particles have been performed using  reactor neutrino experiments,

299: obtaining the bound $q\circa{<} 10^{-5}e$~\cite{Rubbia}.

300: New reactor experiments have been discussed as a way to test interpretations of the PVLAS anomaly that involve millicharged particles~\cite{Rubbia}.

301: 

302: 

303: 

304: \bigskip

305: 

306: We show, trying to be as model-independent as possible, that 

307: existing models are not compatible with known CMB physics after recombination,

308: which probes the same sub-eV energies relevant for PVLAS,

309: and we obtain new bounds on millicharged particles.

310: 

311: %We find a model-independent bound

312: %$q\circa{<}10^{-7} e$ for $m_q\sim 0.1\eV$

313: %\xxx{as well as much stronger model-dependent bounds?}, excluding proposed 

314: 

315: \section{Model-independent cosmological constraints}

316: 

317: Millicharged particles affect CMB cosmology in many ways.

318: We list the relevant processes, discuss their rate, their effect,

319: and their model-independence.

320: We restrict our attention to light millicharged particles, $m_q\circa{<}\eV$,

321: and on cosmology after decoupling, $T\circa{<}\eV$:

322: according to the SM, photons behave as free particles with negligible thermal mass:

323: their dispersion relation is $q^2 =0$.

324: We conservatively assume that the universe initially contains only ordinary SM particles. 

325: \begin{itemize}

326: \item[a)] $\gamma\gamma \to q \bar{q}$ is the main process for production of millicharged particles.

327: Being induced by a small adimensional gauge coupling, this process is maximally effective at low

328: temperatures, $T_* \sim \max(m_q,T_0)$

329: (where $T_0 =2.7\,{\rm K} = 0.23\,{\rm meV}$ is the present temperature), 

330: so that its rate is dominated by

331: model-independent gauge interactions, controlled only by $m_q$ and $q$.\end{itemize}

332: We can neglect the process $\gamma e\to e q\bar{q}$,

333: that contains Bethe-Heitler-like contributions $\sigma \sim e^2 q^4/(4\pi)^3T_*^2$

334: (suppressed with respect to $\gamma\gamma \to q\bar{q}$ by a

335: factor $\sim e^2/(4\pi)^2$) and

336: Compton-like contributions $\sigma \sim e^4 q^2/(4\pi)^3 m_e^2$

337: (suppressed by the electron mass).

338: Both contributions give rates additionally suppressed by the small number

339: density of free electrons, $n_e/n_\gamma\ll 1$.

340: 

341: 

342: 

343: We can estimate the number density $n_q$ of millicharged

344: particles plus their anti-particles produced by process a) as

345: \beq Y\equiv \frac{n_q}{n_\gamma}

346:  \sim \min\left(1,\sigma(\gamma\gamma \to q \bar{q}) \frac{ n_\gamma(T_*)}{H(T_*)}\right)\qquad \hbox{where}\qquad

347:  \sigma(\gamma\gamma \to q \bar{q}) \sim \frac{q^4}{4\pi T_*^2}\eeq

348:  and  $n_\gamma(T) = 2\zeta(3) T^3/\pi^2$ is the photon number density,

349:  used to normalize $n_q$ such that  their relative density $Y= n_q/n_\gamma$ is not affected by the expansion, with Hubble rate $H(T) = \dot R/R$.

350: 

351: \begin{figure}[t]

352: \begin{center}

353: \includegraphics[width=0.9\textwidth]{nq}

354: \caption{\label{fig:nq}\em

355: Isocurves of the cosmological abundance $n_q/n_\gamma$ 

356: of fermionic (left plot)

357: or scalar (right plot) millicharged particles as function of their mass $m_q$ and charge $q$.

358: The shaded region is excluded by the CMB energy spectrum.

359: %The dashed line is the model-dependent constraint coming from the photon plasma mass.

360: The dot-dashed 

361: curve is the (robust, but not fully model-independent) constraint estimated in  eq.\eq{Yd}, for

362: $q'\sim e$.

363: For comparison, the PVLAS anomaly can be interpreted as production of millicharges with

364: $q\sim \hbox{\rm few}\times10^{-6}e$

365: and maybe $m_q\sim 0.1\eV$~\cite{Ringwald}.

366: }

367: \end{center}

368: \end{figure}

369: 

370: 

371: 

372: 

373: Fig.\fig{nq} shows isocurves of $Y$, more precisely computed solving the Boltzmann

374: equations described in the Appendix.

375: If $Y\ll 1$ the process

376:  $\gamma\gamma\to q\bar{q}$ leads to an energy-dependent depletion of the CMB,

377:  whose energy spectrum has been measured with $\sim 10^{-5}$ accuracy by FIRAS~\cite{FIRAS}. 

378:  As  illustrated in fig.\fig{spettro}, by fitting FIRAS data we find

379: \beq\label{eq:Ybound}

380: Y = n_q/n_\gamma \circa{<}

381: 6\cdot 10^{-5} \qquad\hbox{at $3\sigma$ confidence level ($\Delta \chi^2 = 9$)}.\eeq

382: for $m_q\sim 0.1\eV$, and a slightly weaker bound at smaller $m_q$.

383: Fig.\fig{nq} shows the precise constraint in the $(q,m_q)$ plane.

384: We  ignored effects possibly related to the formation of bound states at recombination, 

385: as they would appear at the frequencies of the H and $^4$He recombination:

386: a region of the CMB spectrum not tested by FIRAS and

387: dominated by the galactic signal of thermal dust.

388: 

389: 

390: 

391: Before concluding that we can safely exclude the interpretation of the PVLAS anomaly based on~\cite{Ringwald}

392: (the interpretation of~\cite{Redondo} employs values of ($q,m_q$) that can be  

393: not far from the allowed region) we need to assess if this bound is model-independent,

394: or if one can add some other new physics and use it to restore a thermal CMB spectrum.

395: 

396: We do not see how additional interactions, slower than the expansion rate, could 

397: provide such a restoration 

398: since the CMB spectrum has been measured in a sizable range of energies.

399: New interactions,   enough faster  than the interaction rate, would re-thermalize photons, but at the expense of thermalizing also some new particle (this happens e.g.\ in the model of~\cite{Foot}):

400: data on  cosmological anisotropies disfavors the resulting significant depletion

401: of the photon energy density with respect to other components,

402: and the addition of new quasi-relativistic interacting particles, 

403: that behave as a non freely streaming fluid.

404: 

405: %the resulting $\sim 50\%$ reduction in $n_\gamma$ with observations on cosmological anisotropies.

406: 

407: This problem is avoided if the new particle, added to keep photons in thermal equilibrium, has a mass $m\gg T_0$, such that

408: it decouples at $T\circa{<} m$.

409: In the simplest scenario the millicharged particles themselves could do the job.

410: 

411: Indeed the process a) is much faster than the expansion rate if $q \circa{>}10^{-5}e$; however such a large millicharge cannot fit PVLAS (that actually excludes a too large millicharge) and leads to a thermalized rate also at higher temperatures, thereby 

412: significantly distorting CMB anisotropies imprinted 

413: around last scatterings, at $T\approx 0.25\eV$.

414: 

415: Without performing the necessary dedicated analysis, it seems unlikely that

416: such a big  modification of cosmology could be compatible with observations:

417: in the tight coupling limit the photon/baryon/millicharged fluid would

418: have a sound speed different than the usual photon/baryon fluid

419: (whose acoustic oscillations have been observed as peaks in the 

420: CMB anisotropy angular spectrum).

421: 

422: 

423: Similar problems arise if one instead

424: assumes that millicharged particles are thermalized at the same

425: temperature as photons: extra non-freely streaming particles,

426: photons that remain coupled after recombination,

427: a non standard amount of relativistic energy density at recombination.

428: Since interaction rates mediated by adimensional couplings 

429: become slower than the expansion rate at large temperatures,

430: this alternative scenario is unnatural and we see

431: no preferred way that realizes it in a clean way.

432: One can imagine many different possibilities;

433: for example the effective number of

434: neutrinos (commonly used to parameterize the relativistic energy density

435: at recombination, probed by CMB data)

436: typically deviates from $3$ by some ${\cal O}({\rm few})$ factor

437: and can even be smaller than 3,

438: in scenarios where millicharged particles reheat photons

439: below the decoupling temperature of neutrinos.

440: 

441: 

442: 

443: 

444: %Assuming $n_q \ll n_\gamma$, the Bolztamnn equation describing

445: %millicharged production is

446: %$$zHs \frac{dn_\gamma}{dz} = \gamma$$

447: %where $2/2!=1$.

448: 

449: \begin{figure}[t]

450: \begin{center}

451: \includegraphics[width=0.8\textwidth]{spettro}

452: \caption{\label{fig:spettro}\em FIRAS data compared to

453: the energy-dependent depletion  of the CMB spectrum due to

454: $\gamma\gamma\to q\bar{q}$.

455: We plot $1-r$, with $r$ given in eq.\eq{r}, 

456: computed for fermion (scalar) millicharges with

457: $m_q = 0.1\eV$ and

458: $q=10^{-7}e$ ($q=1.7~10^{-7}e$), chosen such that the two cases give roughly equal effects,

459: excluded at about 3 standard deviations.}

460: \end{center}

461: \end{figure}

462: 

463: %\begin{itemize}

464: 

465: 

466: 

467: %

468: %\item[b)]  contains a  contribution to the millicharges production rate

469: %that can be important at low $q$, 

470: %being enhanced by  $e^2/q^2$ with respect to $\gamma\gamma\to q \bar q$. 

471: %However it is suppressed by a 3-body factor $e^2/(4\pi)^2$ and by the small number

472: %density of free electrons.\footnote{We recall that

473: %today  $Y_e=n_e/n_\gamma =(1-Y_p/2)\eta

474: %\approx 5~10^{-10}$, where $\eta$ is the baryon asymmetry and $Y_p$

475: %is the Helium mass fraction, 

476: %and about $99.9\%$ of the electrons got

477: %bound in neutral hydrogen at  $T\circa{<}\eV$.}

478: %Neglecting the Bethe-Heitler contribution (not enhanced by $e^2/q^2$)

479: %we find

480: %%Actually we are concerned with the ratio $\sigma(\gamma e\to e q\bar{q})/\sigma(\gamma\gamma \to q \bar{q})$  weighted by a factor  $Y_e =n_e/n_\gamma$ 

481: %%which today is  $n_e/n_\gamma \approx 6\times 10^{-10}$. 

482: %%In the limit in which the incoming photon is very soft, $~1$~MeV, we can consider the electron as a static source of electric field. This allows us to use the Bethe-Heitler

483: %%formula for pair production by a photon in an external (nucleus) field. In our case we write:

484: %%\begin{equation}

485: %%\sigma(\gamma e\to e q\bar{q})=\frac{28}{9}\alpha^3 \frac{q^4}{m_q^2}

486: %%\left(\ln\left(\frac{2 E_\gamma}{m_q}\right)-\frac{109}{42}\right).

487: %%\end{equation}

488: %\begin{equation}

489: %Y_e \frac{\sigma(\gamma e\to e q\bar{q})}{\sigma(\gamma\gamma \to q \bar{q})}\approx 1.6\times 10^{-3}

490: %(\frac{10^{-6}e}{q})^2XXX

491: %\end{equation}

492: %at  $E_\gamma \sim  1$~eV.

493: %% and using the fermion millicharged particle parameters fitting PVLAS, we get

494: %%As  the photon energy is increased, the target electron cannot be considered any more as a static electric source, and a full computation of the process $\gamma e\to e q\bar{q}$ is in order. At a value of  $s=(p_\gamma+p_e)=1$~MeV, we find

495: %%\begin{equation}

496: %%\Phi\frac{\sigma(\gamma e\to e q\bar{q})}{\sigma(\gamma\gamma \to q \bar{q})}=1.6\times 10^{-3},

497: %%\end{equation}

498: %%which still appears small enough to give confidence that we can safely neglect this channel in our next considerations.

499: %%Notably one should also consider that most electrons  got

500: %%bound in neutral hydrogen at  $T\circa{<}\eV$, so these are basically overestimated ratios. \xxx{cosmologomelchiorri, questa storia dell'idrogeno va bene detta cos\`i ?}

501: 

502: %\item[c)]

503: %We can also neglect $\gamma^* \to q\bar{q}$,  that can only proceed 

504: % in a medium dense enough that the photon gets a mass $m_\gamma > 2 m_q$.

505: % Moreover average galactic magnetic fields are not sufficient to open this channel, (whereas string magnetic fields in the proximity of some astrophysical objects could).

506: 

507: %\end{itemize}

508: %In conclusion, low-energy millicharged production is dominated by process a).

509: 

510: \medskip

511: 

512: Once that enough millicharged particles have been produced, photons start having significant interactions with them.  

513: Let us now turn to examine the potentially most remarkable of such scattering processes.

514: 

515: \begin{itemize}

516: 

517: \item[b)] $\gamma q \to \gamma q$ leads to a photon attenuation length.

518: Its cross section is $\sigma =q^4/6\pi m_q^2$ in the non-relativistic Thompson  limit.

519: %Furthermore, photons with frequency $\omega$

520: %propagating in a plasma of millicharged particles develop a refraction index $n$.

521: %In the a non-relativistic plasma it is given by

522: %$n = (1 - \omega^2_p/\omega^2)^{1/2}$ where

523: %$\omega$ is the photon  frequency and

524: %$\omega_p = q \sqrt{n_q/m_q} = 800\,{\rm Hz} (q/10^{-6}e) \sqrt{Y\cdot \eV/m_q}$.

525: %The numerical result holds for the present value of $n_\gamma$.

526: %%It modifies the photon dispersion relations

527: %%in the same way as a photon mass $m_\gamma = \omega_p$.

528: This process does not lead to new significant constraints and

529: can be suppressed in a model-dependent way,

530: as discussed at point d).

531: 

532: 

533: 

534: %The dashed blue line in fig.\fig{nq} corresponds to  $\omega_p =1\,{\rm Hz}=0.6~10^{-15}\eV$

535: %which roughly is the present bound~\cite{mgamma}.\footnote{This bound comes

536: %from planetary-scale observations, that presumably are sensitive to

537: %the local galactic millicharge density, that might 

538: %higher than the average cosmological density.

539: %We conservatively ignore this effect; studies have been performed in the  case of CMB neutrinos

540: %(which have mass range comparable to millicharged particles), finding 

541: %over-densities of at most a factor of few~\cite{nulocal}.

542: 

543: %Notice also that a much stronger bound, 

544: %$m_\gamma\circa{<}1/R\sim 10^{-26}\eV$, holds in models where a photon mass

545: %prevents long-range magnetic fields, that are known to extend up to a galactic size $R$.

546: %Ref.~\cite{Dvali} provided an example of a mechanism that 

547: %gives a photon mass

548: %%distorts the photon dispersion relation giving a photon mass 

549: %without erasing magnetic fields:

550: %the usual Higgs mechanisms admits in a medium non trivial `vortex' configurations~\cite{Dvali}.

551: %The case considered here, a photon plasma mass, is another example

552: %of how new physics might show up in experiments that search for

553: %deviations from Maxwell equations.

554: 

555: %}

556: 

557: 

558: \end{itemize}

559: Furthermore, there can be model-dependent scatterings involving new-physics particles.

560: The presence of at least one new extra new light vector $\gamma'$

561: seems almost model independent;  e.g.\ in the model outlined in the introduction

562: the millicharged particle has a sizable  gauge coupling $q' \sim e$ to two

563: extra vectors $\gamma'$ with masses $m'=0$ and $m'\sim m$.

564: Therefore, unless all $\gamma'$ are too heavy for being relevant, they have sizable effects.

565: 

566: \begin{itemize}

567: \item[c)]  $\gamma q \to \gamma' q$, with cross section

568: $\sigma=q^2 q^{\prime 2}/6\pi m_q^2$ in

569: the limit of non-relativistic millicharge $q$ and of ultra-relativistic $\gamma'$.

570: If $q'\sim e$ one gets a constraint  orders of magnitude stronger than previous bounds

571: by demanding that $\gamma q \to \gamma' q$ does not distort the CMB energy spectrum.

572: (This process can easily be in thermal equilibrium, again restoring a thermal CMB spectrum

573: but also generating the other problems previously discussed).

574: Coherent forward scattering, i.e.\ $\gamma \leftrightarrow \gamma'$ oscillations, are

575: suppressed by a mixing angle $\theta\sim q/q'$

576: which is small in the most plausible part of the parameter space, $q'\sim e$.

577: 

578: 

579: %\xxx{estimate: this implies $q\circa{<}10^{-9}$. 

580: %Or less

581: %Or much stronger if collective effect generate

582: %a $\gamma\gamma'$ mixing mass??}

583: 

584: 

585: %If instead it goes in thermal equilibrium, it

586: 

587: 

588: %

589: 

590: % can lead to 

591: %additional distortion of the CMB energy spectrum, and to a reduced attnuation length for $\gamma$.

592: %t is $\propto q^2 q^{\prime2}$ and $q'$ is (?) big.

593: %For $q'=e$, it implies $(q/e)^2 Y < 10^{-22}$, which is STRONG.

594: 

595: 

596: 

597: \item[d)] Scatterings b) and c) involve an initial state millicharge $q$: 

598: its density can be suppressed by 

599: $q\bar{q}\to \gamma'\gamma'$. 

600: A sizable suppression of $n_q$ is obtained only if  $m_{\gamma'} \circa{<} T \ll m_q$ and if

601: this process is in thermal equilibrium

602: (this is indeed the most natural situation, 

603: since $\sigma(q\bar{q}\to \gamma'\gamma' )\sim q^{\prime 4}/m_q^2$ is not suppressed by any millicharged coupling),

604: while $n_q$ is only reduced by an ${\cal O}(1)$ factor at $T\sim m_q$.

605: \end{itemize}

606: The above considerations suggest another robust (but not fully model-independent) constraint,

607: coming from the distortion of the CMB energy spectrum due to

608: $\gamma q\to \gamma' q$ at temperatures $T\sim T_*$.

609: The resulting total deficit of photons is estimated to be 

610: \beq   Y_\gamma \sim\min\left(1, \sigma(\gamma q \to \gamma' q) \frac{ Y n_\gamma(T_*)}{H(T_*)}\right)\qquad

611: \hbox{where}\qquad \sigma(\gamma q \to \gamma' q)\sim \frac{q^2 q^{\prime 2}}{4\pi T_*^2}\eeq

612: Since this effect is energy-dependent, FIRAS data imply

613: $Y_\gamma \circa{<} 10^{-4}$ which translates into 

614: \beq \label{eq:Yd}q'\circa{<} 10^{-9}e ~ \bigg(\frac{q'}{e}\bigg)^{1/3} \max \left(1, \frac{m_q}{T_0} \right) ^{1/6}  .  \eeq

615: This strong bound depends weakly on the model-dependent parameter $q'$, which is

616: comparable to $e$ if one wants to get millicharged particles from a small $\gamma/\gamma'$ kinetic

617: mixing, rather than putting by hand a small gauge coupling.

618: This bound is grossly incompatible with the values of $q,q'$ used in the model of~\cite{Redondo}

619: to mediate at one loop the effective operator $aF_{\mu\nu}F^{\mu\nu}/4M$.

620: Indeed this operator  generates itself a $Y_\gamma \sim 10^{-(7 \hbox{--} 8)}$,

621: a few orders of magnitude below the sensitivity of FIRAS, 

622: for the values of  $M\sim \hbox{few} \cdot 10^5\GeV$ and of the axion-like mass $m_a \sim$ few meV

623: suggested by PVLAS.

624: 

625: 

626: 

627: 

628: %We have seen that cosmology can be affected in many different ways.

629: %The bound that we consider comes from the model-independent

630: % process a): assuming that millicharged particles have a small abundance,

631: % $\gamma\gamma\to q\bar{q}$ leads to a small distortion of the CMB energy spectrum,

632: % measured with $10^{-5}$ accuracy by FIRAS~\cite{FIRAS}:

633: % we will therefore obtain a bound $N_q \circa{<} 10^{-5} N_\gamma$.

634: %This spectral distortion can be suppressed if process a) or some other model-dependent

635: % process such as e) go in thermal equilibrium.

636: %Without performing detailed studies of such scenarios, we will argue that

637: %such big modifications of cosmology are not compatible with observations.

638: %We will not consider bounds from model-dependent processes such as e),

639: %although (in regions of the parameter space) they are much stronger than the

640: %model-independent bound.

641: 

642: 

643: \section{Conclusions}

644: Motivated by the PVLAS anomaly~\cite{pvlas}, the authors of~\cite{Redondo,Ringwald} proposed 

645: new models containing millicharged particles whose charge only appears at low energy

646: (or more precisely at small momentum transfer),

647: avoiding the  bounds  on millicharged particles from higher-energy probes, such as star cooling and BBN~\cite{Davidson}.

648: %\footnote{According to~\cite{Ringwald}, a second motivation for this physics is that it is predicted by most models coming from string theory. }

649: Millicharged particles are used to directly fit PVLAS in~\cite{Ringwald}, and to mediate an axion-like coupling to a

650: new light scalar in~\cite{Redondo}.

651: 

652: 

653: Cosmology at low energies (namely at $T\circa{<}\eV$ after recombination)

654: provides a direct and sensitive probe to this sort of new physics.

655: Within standard cosmology photons behave as free particles with negligible thermal mass.

656: New physics processes like $\gamma \gamma \to q \bar{q}$ 

657: %(computable in terms of gauge interactions)

658: %(and $\gamma q \to q\gamma'$ in models with light extra vectors $\gamma'$)

659: distort the CMB energy spectrum,

660: potentially conflicting with FIRAS that measured a black-body energy spectrum

661: up to $\sim10^{-5}$ accuracy~\cite{FIRAS}.

662: The resulting constraints on millicharged particles, summarized in fig.\fig{nq},

663:  are therefore based on simple and fully safe cosmology,

664:  and exclude the interpretations of the PVLAS anomaly proposed in~\cite{Redondo,Ringwald}.

665: We cannot propose modifications of the models that

666: avoid the conflict with FIRAS data.

667: 

668: %%The model-dependent form factor of eq.\eq{matrix} was used in~\cite{Redondo,Ringwald}

669: %%as well as in the present analysis only as a motivation to ignore

670: %%higher-energy constraints:

671: %%processes relevant for PVLAS (in~\cite{Redondo,Ringwald})

672: %%and for cosmology (in the present paper) are computed in terms

673: %%of gauge interactions only. 

674: %A dedicated analysis is needed to see if some  modification of the model

675: %%(or possibly the form factor of eq.\eq{matrix}, that plays no direct r\^ole

676: %%in our cosmological analysis nor in the PVLAS  interpretation of~\cite{Redondo,Ringwald})

677: %can be used to avoid the conflict with FIRAS data:

678: %this would require selectively suppressing  cosmological effects a few orders of magnitude

679: %more than PVLAS effects.

680: 

681: %In cosmology $q \sim eT$ while in PVALS $q \sim 2 m_q$,

682: %and lowering $m_q$ is impossible both because ruins the PVLAS fit

683: %and because having  Landau levels below the laser energy,

684: % $E_q = \hbar \omega_q =qB/m \sim qB/m < E_\gamma$,

685: % requires $m_q > qB/E_\gamma = 0.3 meV$

686: %($qB = 18 meV$) for q = 10^-6 e.

687: % 1Tesla = 192 eV^2

688: 

689: 

690: 

691: \paragraph{Acknowledgements} 

692: We thank R. Barbieri, A. Macchi, L. Maiani,  E. Gabrielli, U. Gastaldi, M. Roncadelli  

693: and especially  S. Davidson for many useful discussions and suggestions.

694: 

695: 

696: \small\appendix

697: \section*{Appendix}

698: 

699: 

700: 

701: The evolution of $Y = n_q/n_\gamma $ is described by the Boltzmann equation

702: \beq sHz \frac{dY}{dz} =

703:   -2\bigg(\frac{Y^2}{Y^{2}_{\rm eq}}-1\bigg)\gamma_A  \,, \eeq

704: where $z=m_q/T$, $H$ is the expansion rate.

705: The thermally averaged interaction rate for $\gamma\gamma\leftrightarrow q\bar{q}$ is given by

706: \beq \label{eq:gammaA}

707: \gamma_A =  \frac{T}{64 \pi^4} \int_{s_{\rm min}}^{\infty} ds~ s^{1/2}

708:  {\rm K}_1\bigg(\frac{\sqrt{s}}{T}\bigg)

709:   \hat{\sigma}(s)\eeq

710: where the reduced cross section,

711: defined as $d\hat{\sigma}/dt = \sum |A|^2/8\pi s$ where the sum runs over all initial- and final-state indices, is

712: \beq \hat\sigma(s) = \frac{q^4}{\pi   x^2}

713:  \left[-\beta x(4+x)+(x^2+4x-8) \ln\frac{1+\beta}{1-\beta }\right]

714:  \eeq

715:  if millicharged particles are a Dirac fermion

716:  and

717:  \beq \hat\sigma(s) = \frac{q^4}{2\pi   x^2}

718:  \left[\beta x(4+x)+4(2-x) \ln\frac{1+\beta}{1-\beta }\right]\eeq

719:  if millicharged particles have spin 0.

720:  Here $x\equiv s/M^2$ and $\beta=\sqrt{1-4/x}$ is the $q,\bar{q}$ velocity with respect to their center-of-mass frame. 

721: 

722:  Assuming $Y\ll1$

723: the Bolztmann equation for the distortion in the photon energy spectrum, 

724: $r(\epsilon) = f_\gamma(\epsilon)/f_{\rm BE}(\epsilon)$, is

725:  \beq \label{eq:r}

726:  Hz \frac{dr(\epsilon)}{dz} =  \frac{1}{32 \pi^2 \epsilon}\int{dc\, d\epsilon^\prime ~ \epsilon^\prime} f_{BE}(\epsilon^\prime)\cdot \hat\sigma(s = 2\epsilon \epsilon^\prime(1-c)T^2)\eeq

727:  where $\epsilon = E/T$ is the comoving photon energy, $f_{\rm BE}(\epsilon)=1/(e^\epsilon-1)$ is the Bose-Einstein distribution function and $c$ is the cosine of the scattering angle between the two photons.

728: ($Y$ is computed using eq.\eq{gammaA}, 

729: where, as usual, all thermal distributions are approximated

730: with the Maxwell distribution: 

731: this approximation makes almost a factor of 2 difference in $Y$,

732: that was compensated in eq.\eq{Ybound} by properly 

733: rescaling the right-handed side).

734: 

735: 

736: 

737: %\begin{multicols}{2}

738: \footnotesize

739: 

740:  \begin{thebibliography}{nn}

741: 

742:  \bibitem{Davidson}

743: \art[hep-ph/0001179]{S. Davidson, S. Hannestad, G. Raffelt}{JHEP}{05}{003}{2000}.

744: See also \art[hep-ph/0311189]{S. Dubovsky, D. Gorbunov, G. Rubtsov}{JETP Lett.}{79}{1}{2004}. For an earlier accounts on the theoretical possibility of millicharged particles see: 

745: %\cite{Holdom:1985ag}

746: %\bibitem{Holdom:1985ag}

747:   B.~Holdom,

748:   %``Two U(1)'S And Epsilon Charge Shifts,''

749:   Phys.\ Lett.\  B {166}, 196 (1986);

750:   R.~N.~Mohapatra and I.~Z.~Rothstein,

751:   %``Astrophysical Constraints On Minicharged Particles,''

752:   Phys.\ Lett.\  B {247} (1990) 593   

753:  and references therein.

754: 

755: \bibitem{pvlas}

756: \art[hep-ex/0507107]{E.~Zavattini {\it et al.}  [PVLAS Collaboration]}{Phys.\ Rev.\ Lett.}{96}{110406}{2006}. For further details on the PVLAS effect see: 

757: \art[hep-ph/0611267]{S.~L.~Adler}{J.\ Phys.}{A  {40}}{F143}{2007}.

758: 

759: 

760:   

761: \bibitem{Ringwald}

762: \hepart[hep-ph/0608248]{S. Abel, J. Jaeckel, V. Khoze, A. Ringwald}.

763: \hepart[hep-ph/0612098]{M. Ahlers, H. Gies, J. Jaeckel, A. Ringwald}.

764: 

765: \bibitem{Foot}

766: \hepart[hep-ph/0702125]{R. Foot, A. Kobakhidze}.

767: 

768: \bibitem{mpz}

769:   L.~Maiani, R.~Petronzio and E.~Zavattini,

770:    Phys.\ Lett.\  B {175}, 359 (1986).

771: 

772:   \bibitem{cast} 

773:   See e.g.\

774:   \hepart[hep-ph/0611350]{G.~G.~Raffelt} and

775:   \hepart[hep-ex/0702006]{CAST Collaboration}.

776:   %``An improved limit on the axion - photon coupling from the CAST

777:   %experiment,''

778: 

779: 

780: \bibitem{Redondo}

781: \art[hep-ph/0606163]{E. Masso, J. Redondo}{Phys. Rev. Lett.}{97}{151802}{2006}.

782: 

783: %\bibitem{Moha}

784: %\art[hep-ph/0610068]{R. Mohapatra, S. Nasri}{Phys. Rev. Lett.}{98}{050402}{2007}.

785: 

786: \bibitem{antoniadis}

787: \hepart[hep-ph/0606306]{I.~Antoniadis, A.~Boyarsky and O.~Ruchayskiy}.  

788: 

789: 

790: 

791: 

792:   \bibitem{mimmo}

793:   \hepart[hep-ph/0702197]{E.~Gabrielli and M.~Giovannini}, 

794:   see also E.~Gabrielli, K.~Huitu and S.~Roy,

795:   %``Photon propagation in magnetic and electric fields with scalar /

796:   %pseudoscalar couplings: A new look,''

797:   \art[hep-ph/0604143]{E.~Gabrielli, K.~Huitu, S.~Roy}{Phys.\ Rev.}{D74}{073002}{2006}.

798:   \hepart[astro-ph/0612227]{A. Dupays, M. Roncadelli}.

799:  

800: 

801: 

802: 

803: \bibitem{Rubbia}

804: \hepart[hep-ph/0612203]{S.N. Gninenko, N.V. Krasnikov, A. Rubbia}.

805: 

806: 

807: \bibitem{FIRAS}

808: J.~C.~Mather {\it et al.},

809:   %``Measurement of the Cosmic Microwave Background spectrum by the COBE FIRAS

810:   %instrument,''

811:   Astrophys.\ J.\  {420} (1994) 439.

812:   Data available at \url{lambda.gsfc.nasa.gov/product/cobe/firas_overview.cfm}.

813: 

814: 

815: \bibitem{Stuckelberg}

816: \art{E.C.G. Stueckelberg}{Helv. Phys. Acta}{11}{225}{1938}.

817: 

818: 

819: %\bibitem{Nnu} Effects after decoupling can be

820: %approximatively parameterized as an effective number of neutrinos.

821: 

822: %

823: 

824: %

825: %\bibitem{mgamma}

826: %For reviews see

827: %\art{A.S. Goldhaber et al.}{Rev. Mod. Phys.}{43}{277}{1971};

828: % \art{J.D. Barrow, R.R. Burman}{Nature}{307}{14}{1984}.

829: 

830: %

831: %\bibitem{nulocal}

832: %The local density of CMB neutrinos has been studied in

833: %\art[astro-ph/0208419]{S.~Singh and C.~P.~Ma}{\PR}{D67}{023506}{2003};

834: %\art[hep-ph/0408241]{A. Ringwald, Y.Y.Y. Wong}{JCAP}{0412}{005}{2004}.

835: 

836: %

837: 

838: % \bibitem{Dvali}

839: % \art[hep-ph/0306245]{E. Adelberger, G. Dvali, A. Gruzinov}{Phys. Rev. Lett.}{98}{010402}{2007}.

840: 

841: 

842: \end{thebibliography}

843: 

844: 

845: \end{document}

846: 

847: 

848: 

849: