1: %\documentstyle[12pt,epsfig]{article}

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

3: \usepackage{amssymb,amsmath,amscd,epsfig}

4: %\usepackage{Emlines2,Emlines,Eepic,Epic}

5: \textwidth=160mm \textheight=230mm

6: \voffset=0mm \topmargin=-10mm

7: \hoffset=-5mm

8: 

9: \title{Statistics of  neutrinos 

10: and the double beta decay}

11: \author{

12: A.S. Barabash$^{\rm a}$ , A.D. Dolgov$^{\rm a,b,c,d}$,  R. Dvornick\'y$^{\rm 

13: e}$,\\  

14: F. \v Simkovic$^{\rm e}$,  A.Yu. Smirnov$^{\rm d,f}$, 

15: \\[5mm]

16: ${\rm ^a}$ {\small\it Institute of Theoretical and Experimental Physics, 

17: %B.Cheremushkinskaya 25,

18: %}\\

19: %{

20: \small\it 117259 Moscow, Russia} \\

21: ${\rm ^b}$\small\it{Dipartimento di Fisica, 

22: Universit\`a degli Studi di Ferrara, I-44100 Ferrara, Italy}

23: \\

24: ${\rm ^c}$ {\small\it Istituto Nazionale di Fisica Nucleare,

25: Ferrara 44100, Italy} \\

26: %${\rm ^c}$ {\small\it Institute of Theoretical and Experimental Physics, 

27: %117218, Moscow, Russia}\\

28: ${\rm ^d}$ {\small\it 

29: The Abdus Salam International Centre for Theoretical Physics, 

30: I-34100 Trieste, Italy}\\

31: ${\rm ^e}$ {\small \it 

32: Comenius University, Dept. of Nuclear Physics and Biophysics, 

33: Mlynsk\'a dolina,}\\

34: {\small\it SK-84248 Bratislava, Slovakia}\\

35: ${\rm ^f}$ {\small\it Institute for Nuclear Research, Russian Academy of

36: Sciences, Moscow, Russia} 

37: }

38: \date{}

39: 

40: \begin{document}

41: 

42: \special{papersize=210mm 297mm}

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

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

45: \newcommand{\be}{\begin{eqnarray}}

46: \newcommand{\ee}{\end{eqnarray}}

47: \newcommand{\bi}{\bibitem}

48: \newcommand{\lar}{\leftarrow}

49: \newcommand{\rar}{\rightarrow}

50: \newcommand{\lrar}{\leftrightarrow}

51: \newcommand{\mplq}{m_{Pl}^2}

52: \newcommand{\mnu}{m_\nu}

53: \newcommand{\nnu}{n_\nu}

54: \newcommand{\ngam}{n_\gamma}

55: \newcommand{\ms}{m_S}

56: \newcommand{\taus}{\tau_S}

57: \newcommand{\ns}{n_S} 

58: \newcommand{\me}{m_e}

59: \newcommand{\dnnu}{\Delta N_\nu}

60: \newcommand{\Tbbn}{T_{BBN}}

61: \newcommand{\nue}{\nu_e} 

62: \newcommand{\hf}{\hat f}

63: \newcommand{\hb}{\hat b}

64: \newcommand{\hfc}{\hat f^+}

65: \newcommand{\hbc}{\hat b^+}

66: 

67: 

68: 

69: \maketitle

70: 

71: \begin{abstract}

72: We assume that the Pauli exclusion principle is violated for neutrinos, 

73: and thus, neutrinos obey at least partly the Bose-Einstein statistics. 

74: The parameter $\sin^2 \chi$ is introduced that characterizes the 

75: bosonic (symmetric) 

76: fraction of the neutrino wave function. Consequences of the violation of 

77: the exclusion principle for the two-neutrino

78: double beta decays ($2\nu\beta\beta$-decays) are considered. This 

79: violation strongly changes the rates of the decays and modifies the energy 

80: and angular distributions of the emitted electrons. 

81: Pure bosonic neutrinos are excluded by the present data. In the case of partly 

82: bosonic (or mixed-statistics) neutrinos the analysis of the existing 

83: data allows 

84: to put the conservative upper bound $\sin^2 \chi < 0.6$. 

85: The sensitivity of future measurements of the 

86: $2\nu\beta\beta$-decay to $\sin^2 \chi$ is evaluated. 

87: \end{abstract} 

88: 

89: 

90: \section{Introduction \label{s-intro}}

91: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

92: 

93: 

94: Does neutrino respect the exclusion principle of it's inventor?

95: In this paper we assume  that Pauli exclusion principle is 

96: violated for neutrinos and therefore neutrinos obey 

97: (at least partly) the Bose-Einstein  statistics.   

98: Possible violation of the exclusion principle was discussed 

99: in a series of  papers~\cite{ign-kuz} though no satisfactory 

100: and consistent mechanism of the violation  has been proposed so far. 

101: The assumption of violation of the Pauli exclusion principle 

102: leads to a number of fundamental problems which include   

103: loss of a positive definiteness of energy, violation of 

104: the CPT invariance, and possibly, of the  Lorentz invariance as well 

105: as of the unitarity of S-matrix.  

106: (For a critical review see ref.~\cite{lbo-rev}.) 

107: Experimental searches of the effects of the 

108: Pauli principle violation for electrons~\cite{exp-viol}  

109: and nucleons~\cite{exp-bar} have given negative results,  

110: leading to extremely strong bounds on the magnitude of 

111: violation.  

112:  

113: It may happen however that due to unique properties of neutrinos 

114: (neutrality, smallness of mass associated to some  

115: high mass scales), 

116: a violation of the Pauli principle in the  neutrino sector is  

117: much stronger than in other particle sectors. Therefore one 

118: may expect that  effects of its

119: violation can be first seen in neutrino physics.   

120: 

121: A possibility of the Bose statistics for neutrinos has been first 

122: considered in ref.~\cite{gri} where its effects on 

123: the Big Bang Nucleosynthesis (BBN) have been 

124: studied. According to \cite{gri} the change of neutrino 

125: statistics from pure fermionic to pure bosonic diminishes the 

126: primordial $^4{\rm He}$ abundance by $\sim 4\%$.   

127: %is equivalent to the 

128: %decrease of  number of the effective neutrino species 

129: %$\Delta N_{\nu} = -0.74$. 

130: 

131: 

132: The idea of bosonic neutrinos has been proposed independently 

133: in ref.~\cite{dosm}, where cosmological and astrophysical 

134: consequences of this hypothesis have been studied.  

135: Bosonic neutrinos might form a cosmological Bose condensate which 

136: could account for all (or a part of) the dark matter in the 

137: universe. ``Wrong'' statistics

138: of neutrinos modifies the BBN, leading to the

139: effective number of neutrino species smaller than three.

140: The conclusion in~\cite{dosm}  agrees qualitatively with  results of 

141: ~\cite{gri} though quantitatively a smaller decrease of 

142: $N_{\nu}$ is found~\cite{hansen}. 

143: 

144: As far as the astrophysical consequences are concerned, 

145: dynamics of the  supernova collapse would be influenced and

146: spectra of the supernova neutrinos may change~\cite{dosm,kar}. 

147: The presence of neutrino condensate  would enhance contributions of the Z-bursts

148: to the flux of the UHE cosmic rays and lead to substantial

149: refraction effects for neutrinos from remote sources \cite{dosm}.\\

150: 

151: We assume that the Pauli principle is violated substantially 

152: for neutrinos, while the  violation is negligible for other particles. 

153: In particular, for electrons we will assume the usual Fermi-Dirac (FD) 

154: statistics. How to reconcile this pattern of the violation  

155: with the fact that in the standard model the left-handed

156: neutrino and electron belong to the same doublet?   

157: %Notice that being the only neutral leptons, 

158: The answer may be connected to  the fact that neutrinos are the only

159: known neutral leptons and thus they can 

160: have substantially different properties from those of the charged 

161: leptons. In particular, neutrinos can be the  Majorana particles and 

162: violate lepton number conservation. 

163: The  difference between  charged leptons and neutrinos 

164: should be  related to  breaking of the electro-weak (EW) symmetry, and 

165: it can originate from  some high 

166: mass  scale of nature.  

167: One may consider  scenario where  violation of the Pauli 

168: principle occurs in  a hidden sector of  theory related to 

169: the Planck scale physics, or strings physics. 

170: It could be mediated by some singlets of the Standard model - (heavy) neutral 

171: fermions which mix with neutrinos when the EW symmetry is broken. 

172: Since only neutrinos can mix with the singlets, effects of the Pauli 

173: principle violation would show up first in the neutrino sector  and then 

174: communicate to other particles. 

175: %Also one can consider a possibility that the messenger of 

176: %the Pauli principle violation is the light sterile neutrino. 

177: %It has a small mixing with the active components, and  this small mixing  

178: %quantifies the degree of violation in the observable sector.  

179: In this way a small or partial violation of the relation between spin and 

180: statistics  might occur. 

181: A violation of the  spin-statistics theorem for other particles can be 

182: suppressed by  an additional power of a small parameter relevant for the 

183: violation in the neutrino sector and due to weak coupling of neutrino 

184: to other particle sector.\\

185: 

186: 

187: A violation of the Pauli principle for neutrinos

188: should show up in the elementary processes where identical

189: neutrinos are involved. A realistic process for  this test  

190: is the two-neutrino  double beta decay ($2\nu\beta\beta$-decay), 

191: \begin{equation}

192: A\rar A'+ 2\bar{\nu} + 2e^- 

193: \end{equation}

194: (or similar with neutrinos and positrons). 

195: It was shown in~\cite{dosm} that 

196: the  probability of the decay as well as the energy spectrum

197: and angular distribution of electrons should be affected. 

198: Qualitative conclusions were that the pure bosonic neutrino is excluded,  

199: whereas large fraction of the bosonic component 

200: in a neutrino state is still allowed by the present data. 

201: In this connection, a possibility of partly bosonic (mixed-statistics) 

202: neutrinos can be considered.\\ 

203: 

204: 

205: In this paper we perform 

206: a detailed study of the effects of bosonic 

207: neutrinos on the double beta decay. In sect. 2 we 

208: consider the  general case of partly bosonic neutrinos. 

209: We introduce a phenomenological parameter 

210: $\sin^2\chi$ which describes the fraction of bosonic 

211: neutrinos in such a way that a smooth change of 

212: $\sin^2\chi$ from 0 to  1 

213: transforms fermionic neutrinos into bosonic ones. 

214: So, in general,  neutrinos may  possess a kind of mixed or 

215: more general statistics than Bose or Fermi ones~\cite{para,ign-kuzm}.

216: In sect. 3 we present an analytic study of the double beta decay probabilities.

217: The exact expressions for the  $2\nu\beta\beta$-decay rates 

218: to ground and excited $0^+$  and

219: $2^+$ states with corresponding nuclear matrix elements (NME's) 

220: are given in sect. 4.

221: The results of numerical calculations of the total rates and various 

222: distributions for the $2\nu\beta\beta$-decays of $^{76}{\rm Ge}$ and 

223: $^{100}{\rm Mo}$ 

224: are presented in sect 5. In sect. 6. we obtain the bounds on  

225: $\sin^2 \chi$ from the existing data and  evaluate the sensitivities

226: of future double beta decay experiments. 

227: Discussion and conclusions are given in sect. 6. 

228: 

229: 

230: %It is not clear if effects considered in this paper are 

231: %consistent with these bounds, which depends on particular mechanism 

232: %of the violation.   

233: 

234: 

235: %%\section{Bosonic and partially bosonic neutrinos}

236: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

237: 

238:  

239: %In fact, the high accuracy of the validity of Fermi statistics for 

240: %electrons~\cite{exp-viol} could put a strong bound on a possible 

241: %``transfer'' of wrong statistics 

242: %from neutrinos  to electrons. 

243: 

244: \section{The $2\nu\beta\beta$-decay for  bosonic and partly bosonic neutrinos}

245: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

246: 

247: In the case of mixed statistics the operator of neutrino state can be 

248: written as 

249: \be

250: |\nu\rangle =  \hat a^+ |0\rangle  \equiv

251: c_\delta \hat f^+ |0\rangle + s_\delta \hat b^+ |0\rangle

252: = c_\delta | f\rangle + s_\delta | b\rangle

253: \label{nu-state}

254: \ee

255: where $| f\rangle$ and $| b\rangle$ are respectively 

256: one particle fermionic and bosonic states. 

257: The normalization of $|\nu \rangle$ implies $c^2_\delta + s^2_\delta =1$

258: ($c_\delta \equiv \cos \delta$ and {$s_\delta \equiv \sin \delta$}).

259: $\hf$  ($\hf^+$) and $\hb$ ($\hb^+$) denote  fermionic, 

260: and bosonic annihilation (creation) operators.

261: 

262: To develop  a formalism for description of identical neutrinos 

263: one needs to specify  commutation/anti-commutation relations.

264: We assume that they have the following form:

265: \be

266: \hf \hb = e^{ i \phi} \hb \hf,\,\,\,

267: \hfc \hbc = e^{ i \phi} \hbc \hfc, \,\,\,

268: \hf \hbc = e^{ - i \phi} \hbc \hf,\,\,\,

269: \hfc \hb = e^{- i \phi} \hb \hfc,

270: \label{ab-af}

271: \ee

272: where $\phi$ is an arbitrary phase. Then the two-neutrino state can be  

273: defined as

274: \be

275: |k_1,k_2\rangle = \hat a_1^+ \hat a_2^+ |0\rangle. 

276: \label{two-nu}

277: \ee

278: 

279: %%{\bf (Add more. We also need to compare our description with 

280: %%what people did for Pauli violation for electrons)}

281: 

282: For the pure bosonic neutrino one cannot introduce the Majorana mass term. 

283: So,  the neutrinoless double beta decay should be absent. In the case of 

284: partly bosonic neutrino, the neutrino mass would appear due to its fermionic 

285: component. This means that the kinematical mass measured,  {\it e.g.} in the 

286: tritium beta decay, would not be the same as the mass found from 

287: the neutrinoless 

288: beta decay. Such a situation, however, can be realized in the case of the usual 

289: fermionic neutrinos too.  

290: 

291: 

292: The amplitude of the decay of nucleus $A \rightarrow 2\nu+2e+A'$ 

293: can be written as

294: \be

295: A_{2\beta} = 

296: \langle e(p_{e1}), e(p_{e2}), 

297: \overline\nu (p_{\nu 1}), \overline\nu (p_{\nu 2}),A'|

298: \int d^4 x_1 d^4 x_2

299: \psi_\nu (x_1) \psi_\nu (x_2) {\cal M}(x_1,x_2) 

300: | A \rangle.

301: \label{A-2beta}

302: \ee

303: After making the necessary commutation, according to  eq. (\ref{ab-af}), 

304: we obtain 

305: \be

306: A_{2\beta} = A_f \left[ c_\delta^4 + c_\delta^2 s_\delta^2

307: \left( 1-\cos \phi \right)\right]

308: + A_b \left[ s_\delta^4 + c_\delta^2 s_\delta^2 \left( 1+\cos \phi \right)\right], 

309: \label{A-2beta-2}

310: \ee

311: where $A_f$ and $A_b$ are respectively fermionic (antisymmetric) 

312: and bosonic  (symmetric) parts of two antineutrino emission.

313: The amplitude can be parametrized as

314: \be

315: A_{2\beta} =\cos^2\chi\, A_f + \sin^2\chi\, A_b,

316: \label{A-2beta-3}

317: \ee

318: where $\cos^2\chi =  c_\delta^4 + c_\delta^2 s_\delta^2 \left( 1-\cos \phi \right)$

319: and  $\sin^2\chi =  s_\delta^4 + c_\delta^2 s_\delta^2 \left( 1+\cos \phi \right)$.

320: 

321: After integration over the neutrino phase space an interference between fermionic 

322: $A_f$ and bosonic $A_b$  parts of the amplitude $A_{2\beta}$ 

323: vanishes because the fermionic part is 

324: antisymmetric with respect to neutrino interchange, while bosonic is symmetric. 

325: The probability of the $2\nu\beta\beta$-decay is equal to:

326: \be

327: W_{tot} = \cos^4\chi\, W_f + \sin^4\chi\, W_b,  

328: %\nonumber\\

329: %&=& (1-b^2)~W_f~+~b^2~W_b,

330: \label{W-tot}

331: \ee

332: where $W_{f,b}$ are proportional to $|A_{f,b}|^2$. The expressions

333: for $W_{f,b}$ will be given in the next section.

334: 

335: Qualitative features of the $\beta\beta-$ decay in the presence of  the 

336: bosonic or partly bosonic neutrinos can be understood using the following 

337: consideration.  

338: Essentially, the effect of neutrino ``bosonization''  is that 

339: two contributions to the amplitude of the decay from diagrams with 

340: permuted neutrino momenta $p_{\nu 1} \leftrightarrow p_{\nu 2}$ 

341: should have relative plus sign instead of minus in the FD-case. 

342: 

343: The decay probability, $W_b$,  is proportional 

344: to  the bilinear combinations  of the type

345: $K^b_m K^b_n$,  $K^b_m L^b_n$, $L^b_m L^b_n$ (see the next section), where 

346: \begin{eqnarray}

347: K^b_m \equiv [E_m  - E_i + E_{e1} + E_{\nu 1}]^{-1} - [E_m  - E_i + E_{e2} + 

348: E_{\nu 2}]^{-1},

349: \nonumber\\

350: L^b_m \equiv [E_m  - E_i + E_{e2} + E_{\nu 1}]^{-1} - [E_m  - E_i + E_{e1} + 

351: E_{\nu2}]^{-1}. 

352: \label{prop}

353: \end{eqnarray}

354: Here $E_i$ is the energy of the initial nuclei, $E_m$ is the energy 

355: of the intermediate nuclei,  

356: %state $m$ 

357: $E_{ej}$, and $E_{\nu j}$ are the 

358: energies  of electrons and neutrinos respectively. 

359: The factors (\ref{prop}) correspond to the propagators of the 

360: intermediate nucleus. The key difference between

361: the bosonic and fermionic cases 

362: is the opposite signs of the two terms in the expressions (\ref{prop}).   

363: In the case of fermionic neutrinos 

364: they enter with the same signs (see, {\it e.g.}  \cite{boehm}):

365: \begin{eqnarray}

366: K^f_m \equiv [E_m  - E_i + E_{e1} + E_{\nu 1}]^{-1} + [E_m  - E_i + E_{e2} + 

367: E_{\nu 2}]^{-1},

368: \nonumber\\

369: L^f_m \equiv [E_m  - E_i + E_{e2} + E_{\nu 1}]^{-1} + [E_m  - E_i + E_{e1} + 

370: E_{\nu2}]^{-1}. 

371: \label{propf}

372: \end{eqnarray}

373: (Remember that for electrons we assume the normal 

374: Fermi statistics.) The terms in (\ref{prop}) correspond to 

375: the amplitudes with permuted momenta of both neutrinos and  

376: electrons. 

377: In the case of fermionic neutrinos such an interchange  

378: flips the sign twice (due to neutrinos and electrons), so that 

379: the overall sigh turns out to be plus.  

380: In the case of bosonic neutrinos the permutation of electrons only 

381: changes the sign, and the overall sign is minus. \\

382: 

383: Experimentally interesting are the 

384: $2\nu\beta\beta$-decays to  

385: the ground states  $0^+_{g.s.}$ and to excited states 

386: $0^+_1$ and $2^+_1$.  

387: %excited states are subject of interest.  

388: The effect of 

389: bosonic neutrinos on the $2\nu\beta\beta$-decay half-life is different

390: for $J^\pi = 2^+$ and $J^\pi = 0^+$. This can be understood qualitatively, 

391: approximating the combinations  $K^b_m$ and $L^b_m$ for bosonic neutrinos  by 

392: \begin{equation}

393: K^b_m \approx \frac{E_{e2} - E_{e1} + E_{\nu 2} - E_{\nu 1}}{( E_m  - E_i + E_0/2 )^2}, 

394: ~~~~L^b_m \approx \frac{E_{e1} - E_{e2} + E_{\nu 2} - E_{\nu 1}}{( E_m - E_i + E_0/2 )^2}, 

395: \label{KL}

396: \end{equation}

397: and the corresponding combinations  for the  fermionic neutrinos by  

398: \begin{equation}

399: K^f_m \approx L^f_m \approx  \frac{2}{E_m  - E_i + E_0/2 }.

400: \end{equation}

401: Here $E_0/2 \equiv \langle E_e + E_{\nu}\rangle $ is the average energy of 

402: the leptonic pair,  $E_0 \equiv E_i - E_f$ is the energy release in the decay, and $E_f$ 

403: is the energy of the final nucleus. 

404: 

405: For the $0^+ \rightarrow 0^+$ transitions 

406: an appearance of the differences of the electron 

407: and neutrino energies in the numerators of (\ref{KL}) 

408: leads to substantial (1-3 orders of magnitude) 

409: suppression of the total probability.  

410: It also modifies the energy distributions of electrons.

411: 

412: The effect of bosonic neutrinos on $0^+ \rightarrow 2^+$ 

413: transitions is opposite: The probabilities of transitions 

414: are proportional to the combinations 

415: $(K^{b}_m - L^{b}_m)(K^{b}_n - L^{b}_n)$,

416: where

417: \begin{equation}

418: (K^b_m - L^b_m) \approx  \frac{2(E_{e2} - E_{e1})}{(E_m  - E_i + 

419: E_0/2)^2}.

420: \end{equation}

421: In the case of fermionic neutrinos  

422: the combination $(K^f_m - L^f_m)$ has an additional factor 

423: $(E_{\nu2} - E_{\nu1})/(E_m  - E_i + E_0/2)$ and the suppression 

424: is stronger.  

425: Parametrically the probabilities of the $0^+ \rightarrow 2^+$ and 

426: $0^+ \rightarrow 0^+$ transitions become of the same order for  

427: bosonic neutrinos. 

428: 

429: 

430: In the decay rates,  the kinematical factors $K^{f,b}_m$ and 

431: $L^{f,b}_n$ are weighted with the corresponding 

432: nuclear matrix elements (NME's).  

433: %%(see the next section). 

434: Let us introduce the ratio 

435: \begin{equation}

436: r_0 (J^\pi) \equiv \frac{W_b (J^\pi)}{W_f (J^\pi)},

437: \label{ratiow}

438: \end{equation}

439: of the decay probabilities to ground ($J^\pi= 0^+_{g.s.}$) 

440: and excited ($J^\pi= 0^+_{1},~ 2^+_{1}$) states

441: in pure bosonic $W_b(J^\pi)$  and pure fermionic

442: cases $W_f(J^\pi)$. In general, to find  $r_0(J^\pi)$ one needs to  

443: calculate the  NME 

444: for a given transition within an appropriate nuclear model.

445: The situation is simplified for those nuclear systems, where 

446: the transition via solely the ground state of the  intermediate 

447: nuclei $m=1$ dominates \cite{ABA84,SDS,DKSS}. 

448: For those nuclei the single state dominance 

449: (SSD) approximation (hypothesis) can be used.  

450: In this case the NME's can be factored out in  the rates  

451: and therefore  cancel in the ratio   $r_0(J^\pi)$. 

452: 

453: Let us consider the characteristics of the $\beta\beta$ decay to 

454: the ground and excited states $J^\pi$ in the mixed-statistic case 

455: of partly bosonic neutrinos. According to 

456: our considerations the total decay probability 

457: and the normalized total differential rate can be written as 

458: \begin{eqnarray}

459: W_{tot}(J^\pi) &=& \cos^4\chi W_f(J^\pi) + \sin^4\chi W_b(J^\pi),

460: \label{totpro}\\

461: \nonumber\\

462: P (J^\pi) &=& \frac{dW_{tot}(J^\pi)}{W_{tot}(J^\pi)} %\nonumber\\ 

463: = \frac{\cos^4\chi\,

464: d\omega_f(J^\pi) + \sin^4\chi\, r_0(J^\pi) d\omega_b (J^\pi)}

465: { \cos^4\chi + \sin^4\chi\, r_0(J^\pi)},   

466: \label{distr}

467: \end{eqnarray}

468: where 

469: \begin{equation}

470: d\omega_f(J^\pi) \equiv \frac{dW_f(J^\pi)}{W_f(J^\pi)}, 

471: ~~~d\omega_b(J^\pi) \equiv \frac{dW_b(J^\pi)}{W_b(J^\pi)}

472: \label{normd}

473: \end{equation}

474: are the normalized distributions.  

475: Here $dW_f(J^\pi)$ and $dW_b(J^\pi)$  are the differential rates 

476: of the $2\nu\beta\beta$-decay for the pure fermionic and bosonic  

477: neutrinos. In the case of single state dominance due to factorization,  

478: the normalized  distributions do not depend on the uncertainties 

479: of the matrix elements \cite{SDS,DKSS}. 

480: In general, the factorization does not occur and the 

481: uncertainties of nuclear matrix elements restrict substantially the 

482: sensitivity of the $\beta\beta$-decay to statistics of neutrinos. \\

483: 

484: 

485: 

486: \section{Rates and nuclear matrix elements}

487: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

488: 

489: 

490: For the cases of pure fermionic and bosonic neutrinos we outline

491: the derivation of $2\nu\beta\beta$-decay rates. The relevant nuclear matrix

492: elements will be evaluated and discussed

493: using the  SSD and HSD (higher states dominance) hypothesis 

494: \cite{SDS,DKSS}. 

495: 

496: The matrix element of the $2\nu\beta\beta$-decay process takes the form

497: \begin{eqnarray}

498: <{f}|S^{{(2)}}|{i}> = 

499: \hspace{5cm}\nonumber \\

500: \frac{(-i)^2}{2} 

501: \int {<}e(p_{e1}), e(p_{e2}), 

502: \overline\nu (p_{\nu 1}), \overline\nu (p_{\nu 2}),A'|

503: T \left[ {\cal H}^\beta_{} (x_1) {\cal H}^\beta_{} (x_2)

504: \right] |A{>} dx_1 dx_2, 

505: \label{eq.7} 

506: \end{eqnarray}

507: where the weak $\beta$-decay Hamiltonian is  

508: \begin{equation}

509: {\cal H}^\beta_{} (x) = 

510: \frac{G_{F}}{\sqrt{2}} 

511: \left[\bar{e} (x)\gamma^\mu (1+\gamma_5) \nu_{e}(x)\right]

512: J_\mu(x) + {h.c.}.

513: \label{eq.8} 

514: \end{equation}

515: Here, $J_\mu(x)$ is the

516: weak charged (nuclear) hadron current in the Heisenberg 

517: representation. 

518: The $T$-product of the two hadron currents can be written as

519: \begin{eqnarray}

520: T(H^\beta_{} (x_{{1}}) H^\beta_{}(x_{{2}}))= 

521: \hspace{4cm}\nonumber \\

522: \Theta(x_{{10}} - x_{{20}})H^\beta_{}(x_{{1}})H^\beta_{}(x_{{2}}) + 

523: \Theta(x_{{20}} - x_{{10}})H^\beta_{}(x_{{2}}) H^\beta_{}(x_{{1}}). 

524: \end{eqnarray}

525: 

526: In the derivation of the $2\nu\beta\beta$-decay rate a number of 

527: conventional approximations have been used: i) Only the $s_{1/2}$ 

528: wave states of the outgoing leptons are taken into account. 

529: ii) The contribution of the double Fermi matrix element to the decay rate

530: is neglected as the initial and final nuclei belong to different

531: isospin multiplets. iii) Only the leading order 

532: ($1/m_p$)  Gamow-Teller operators in the non-relativistic reduction

533: of the hadron current are retained. 

534: 

535: For the differential $2\nu\beta\beta$-decay rates 

536: to $0^+$ ground state and $2^+$ excited state we obtain 

537: \begin{eqnarray}

538: dW_{f,b}(J^+) = a_{2\nu} F(Z_f,E_{e1}) F(Z_f,E_{e2})

539: ~{\cal M}^{f,b}_{J^\pi}~

540: d\Omega,   

541: \end{eqnarray}

542: where $a_{2\nu}=(G^4_\beta g_A)^4 m_e^9 /(64 \pi^7)$ 

543: and $G_\beta=G_F \cos\theta_c$ ($G_F$ is Fermi constant,

544: $\theta_c$ is Cabbibo angle). $F(Z_f,E_e)$ denotes the 

545: relativistic Coulomb factor and $g_A$ is the axial-vector

546: coupling constant. 

547: The upper index $f$ ($b$) stands for fermionic 

548: (bosonic) neutrinos. 

549: 

550: The phase space factor equals  

551: \begin{eqnarray}

552: d\Omega &=& \frac{1}{m^{11}_e}

553: E_{e1} p_{e1}~ E_{e2} p_{e2}~ E^2_{\nu 1}~ E^2_{\nu 2}

554: ~\delta (E_{e1} + E_{e2} + E_{\nu 1} + E_{\nu 2} + E_{f} - E_{i})

555: \times \nonumber \\

556: && ~~~~~~~~~~d E_{e1}~d E_{e2}~d E_{\nu 1}~d E_{\nu 2}~ d\cos\theta . 

557: \end{eqnarray}

558: Here, $\theta$ is the angle between the outgoing electrons. 

559: ${\cal M}^{f,b}_{J^\pi}$ ($J^\pi = 0^+,~2^+$)

560: consists of the products of nuclear matrix elements: 

561: \begin{eqnarray}

562: {\cal M}^{f,b}_{0^+} &=&

563: \frac{m^2_e}{4} \left[ |{\cal K}^{f,b}_{0^+}+{\cal L}^{f,b}_{0^+}|^2 

564: + \frac{1}{3}|{\cal K}^{f,b}_{0^+}-{\cal L}^{f,b}_{0^+}|^2 

565: \right] \nonumber\\

566: &&-\frac{m^2_e}{4} 

567: \left[ |{\cal K}^{f,b}_{0^+}+{\cal L}^{f,b}_{0^+}|^2 

568: - \frac{1}{9}|{\cal K}^{f,b}_{0^+}-{\cal L}^{f,b}_{0^+}|^2 \right]

569: ~\frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}, \nonumber\\

570: {\cal M}^{f,b}_{2^+} &=& m^2_e~

571: |{\cal K}^{f,b}_{2^+} - {\cal L}^{f,b}_{2^+}|^2 ~

572: \left(1+\frac{1}{3}

573: \frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}\right)

574: \end{eqnarray}

575: with

576: \begin{eqnarray}

577: {\cal K}^{f,b}_{J^+} &=&

578: \frac{m_e}{\sqrt{s}}  

579: \sum_m <J^\pi_f||\sum_j \tau^+_j \sigma_j || 1^+_m>               

580:               <1^+_m||\sum_k \tau^+_k \sigma_k || 0^+_i> ~K^{f,b}_m \nonumber\\

581: {\cal L}^{f,b}_{J^+} &=&

582: \frac{m_e}{\sqrt{s}}  

583:  \sum_m <J^\pi_f||\sum_j \tau^+_j \sigma_j || 1^+_m>

584:                <1^+_m||\sum_k \tau^+_k \sigma_k || 0^+_i> ~L^{f,b}_m. 

585: \end{eqnarray}

586: Here, $s=1$ for $J=0$ and $s=3$ for $J=2$. $|0^+_i>$, $|0^+_f>$ ($|2^+_f>$)

587: and $|1^+_m>$ are, respectively, the states of the initial, final

588: and intermediate nuclei with corresponding energies $E_i$, $E_f$ and

589: $E_m$. The energy denominators  $K^{f,b}_m$ and $L^{f,b}_m$  

590: were introduced in Eqs. (\ref{prop}) and (\ref{propf}). 

591: 

592: 

593: 

594: \subsection{Higher states dominance}

595:   

596: The $2\nu\beta\beta$-decay rates are usually evaluated in 

597: the  approximation in which the sum of the two lepton energies

598: in the denominator of the nuclear matrix element is replaced

599: with their average value $E_0/2$ 

600: \begin{equation}  

601:  E_m - E_i + E_{e j}+E_{\nu k} \approx E_m - E_i + E_0/2 

602: \end{equation}  

603: ($j,k=1,2$). The main purpose of this approximation is to

604: factorize the lepton and nuclear parts in the calculation 

605: of the $2\nu\beta\beta$-decay half-life. This approximation

606: is justified if the transitions through the higher-lying 

607: states of the intermediate nucleus 

608: (at least few MeV above the ground state of (A,Z+1)

609: nucleus) give the dominant 

610: contribution to the $2\nu\beta\beta$-decay amplitude. This

611: assumption is called the higher states dominance (HSD)

612: hypothesis. It is expected to be realized for

613: A= 48, 76, 82, 130, 136 nuclear systems.

614: 

615: Assuming the HSD hypothesis we obtain for fermionic neutrinos  

616: \begin{eqnarray}

617: {\cal M}^{f}_{0^+} &\simeq& |M_{GT}^{(1)}(0^+)|^2

618: ~\left(1-\frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}\right), 

619: \nonumber\\

620: {\cal M}^{f}_{2^+} &=& |M_{GT}^{(3)}(2^+)|^2

621: \frac{(E_{e1}-E_{e2})^2~(E_{\nu 1}-E_{\nu 2})^2}{2 m^6_e}

622: \left(1+\frac{1}{3}

623: \frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}\right).

624: \label{nmef}

625: \end{eqnarray}

626: In the case of bosonic neutrinos we end up with

627: \begin{eqnarray}

628: {\cal M}^{b}_{0^+} &=& |M_{GT}^{(2)}(0^+)|^2

629: ~\left[ \frac{3(E_{\nu 2}-E_{\nu 1})^2+(E_{e 2}-E_{e 1})^2}{48 m_e^2}-

630: \right.\nonumber \\

631: &&~~~~~~~~~~~~~~~~~~

632: \left. \frac{9(E_{\nu 2}-E_{\nu 1})^2-(E_{e 2}-E_{e 1})^2}{144 m_e^2}

633: ~\frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}\right], 

634: \nonumber\\

635: {\cal M}^{b}_{2^+} &=& |M_{GT}^{(2)}(2^+)|^2

636: \frac{(E_{e1}-E_{e2})^2}{4 m^2_e}

637: \left(1+\frac{1}{3}

638: \frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}\right).

639: \label{nmeb}

640: \end{eqnarray}

641: The Gamow-Teller matrix elements are given by 

642: \begin{equation}

643: M_{GT}^{(r)}(J^\pi) =

644: \frac{(2 m_e)^r}{\sqrt{s}}  

645: \sum_m \frac{<J^\pi_f||\sum_j \tau^+_j \sigma_j || 1^+_m>               

646:  <1^+_m||\sum_k \tau^+_k \sigma_k || 0^+_i>}

647: {(E_m~-~E_i~+~E_0/2)^r} 

648: \end{equation}

649: ($r=1,2,3$). 

650: 

651: The full decay probabilities in pure bosonic $W_b$ and pure 

652: fermionic $W_f$ cases can be written as

653: \begin{eqnarray}

654: W_{f}(0^+) &=& |M_{GT}^{(1)}(0^+)|^2 {\cal I}^{f}_{HSD}(0^+), 

655: \nonumber\\

656: W_{f}(2^+) &=& |M_{GT}^{(3)}(2^+)|^2 {\cal I}^{f}_{HSD}(2^+) 

657: \end{eqnarray}

658: and

659: \begin{eqnarray}

660: W_{b}(J^\pi) = |M_{GT}^{(2)}(J^\pi)|^2 {\cal I}^{f}_{HSD}(J^\pi),  

661: \end{eqnarray}

662: where the phase space integrals are given by

663: \begin{eqnarray}

664: {\cal I}^{f,b}_{HSD}(J^\pi) = \frac{2 a_{2\nu}}{m^{11}_e}

665: \int_{m_e}^{E_i-E_f-m_e} 

666: f^{f,b}_{J^\pi}(E_{e1},E_{e2},E_{\nu 1},E_{\nu 2})

667: F_0(Z_f,E_{e1}) p_{e1} E_{e1} dE_{e1}\times

668: \nonumber\\ 

669: \int_{m_e}^{E_i-E_f-E_{e1}} F_0(Z_f,p_{e2}) p_{e2} E_{e2} dE_{e2}

670: \int_{0}^{E_i-E_f-E_{e1}-E_{e2}}  E_{\nu 2}^2  E_{\nu 1}^2  

671: d E_{\nu 1}

672: \end{eqnarray}

673: with $E_{\nu 2}= E_i -E_f -E_{e1}-E_{e2}-E_{\nu 1}$ and 

674: \begin{eqnarray}

675: f^{f}_{J^\pi}(E_{e1},E_{e2},E_{\nu 1},E_{\nu 2})

676: &=& 1~~~~~~~~~~~~~~~~~~~~

677: ~~~~~~~~~~~~~~~~~~~~~~~ (J^\pi = 0^+),\nonumber\\

678: &=& \frac{(E_{e1}-E_{e2})^2~(E_{\nu 1}-E_{\nu 2})^2}{2 m^6_e}

679: ~~~~~~~~~~ (J^\pi = 2^+),\nonumber\\

680: f^{b}_{J^\pi}(E_{e1},E_{e2},E_{\nu 1},E_{\nu 2})

681: &=& 

682: \frac{3(E_{\nu 2}-E_{\nu 1})^2+(E_{e 2}-E_{e 1})^2}{48 m_e^2}

683: ~~~~~~ (J^\pi = 0^+),\nonumber\\

684: &=&

685: \frac{(E_{e1}-E_{e2})^2}{4 m^2_e}

686: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ (J^\pi = 2^+).\nonumber\\

687: \end{eqnarray}

688: The $2\nu\beta\beta$-decay half-life is 

689: \begin{equation}

690: T^{f,b}_{1/2}(J^\pi) = \frac{\ln{2}}{W_{f,b}(J^\pi)}.

691: \end{equation}

692: 

693: 

694: \subsection{Single state dominance}

695: 

696: The single state dominance hypothesis assumes that the

697: $2\nu\beta\beta$-decays with $1^+$ ground state of the

698: intermediate nucleus (e.g., A=100, 116 and 128 nuclear systems)

699: are only governed by the two 

700: virtual $\beta$-transitions: i) the first one connects

701: the ground state of the initial nucleus with $1^+_1$

702: intermediate state; ii) the second one proceeds from

703: $1^+_1$ state to the final ground state. In this case

704: we find 

705: \begin{eqnarray}

706: {\cal M}^{f,b}_{0^+} &=& |M_{g.s.}(0^+)|^2

707: {m^2_e} \left[\frac{1}{3}(K^{f,b}K^{f,b}+{L}^{f,b}{L}^{f,b}+

708: {K}^{f,b}{L}^{f,b}) - \right.\nonumber \\

709: &&

710: \left. \frac{1}{9}(2 K^{f,b}K^{f,b}+2 {L}^{f,b}{L}^{f,b}+

711: 5 {K}^{f,b}{L}^{f,b})

712: ~\frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}

713: \right], \nonumber\\

714: {\cal M}^{f,b}_{2^+} &=& m^2_e~|M_{g.s.}(2^+)|^2~

715: ({K}^{f,b} - {L}^{f,b})^2 ~

716: \left(1+\frac{1}{3}

717: \frac{{\vec p}_{e1}\cdot {\vec p}_{e2}}{E_{e1}E_{e2}}\right)

718: \end{eqnarray}

719: with $K^{f,b}\equiv K^{f,b}_{m=1}$,

720: $L^{f,b}\equiv L^{f,b}_{m=1}$ and 

721: \begin{equation}

722: M_{g.s.}(J^\pi) =

723: \frac{1}{\sqrt{s}}  

724: <J^\pi_f||\sum_j \tau^+_j \sigma_j || 1^+_1>               

725: <1^+_1||\sum_k \tau^+_k \sigma_k || 0^+_i>.

726: \end{equation}

727: 

728: The value of the matrix element $M_{g.s.}(J^\pi)$

729: can be determined in a model independent way from the

730: single $\beta$-decay and electron capture measurements.

731: From the experimental values of 

732: $\log ~ft$ {\footnote {Because of wide range of $\beta$-lifetimes,

733: transitions are classified by $\log_{10} f t$ values (see e.g. \cite{behr}). 

734: $t$ and $f$ 

735: denote the measured half-life and the Fermi integral, respectively.}} 

736: for the electron

737: capture and the single $\beta$ decay of the ground state 

738: of the intermediate nucleus with $J^\pi= 1^+$ we obtain 

739: \begin{eqnarray}

740: |<1^+_1||\sum_k \tau^+_k \sigma_k || 0^+_i>|

741: &=& \frac{1}{g_A}\sqrt{\frac{3 D}{ft_{EC}}},

742: \nonumber\\

743: |<J^\pi_f||\sum_j \tau^+_j \sigma_j || 1^+_1>|               

744: &=& \frac{1}{g_A}\sqrt{\frac{3 D}{ft_{\beta^-}}}.

745: \end{eqnarray}

746: Here $D=G^4_\beta g^4_A/(8\pi^7)$.

747: 

748: Within the SSD approach for the full decay probabilities

749: we find

750: \begin{eqnarray}

751: W_{f,b}(J^\pi) &=& |M_{g.s.}(J^\pi)|^2 

752: {\cal I}^{f,b}_{SSD}(J^\pi),  

753: \label{ssdpi}

754: \end{eqnarray}

755: where 

756: \begin{eqnarray}

757: {\cal I}^{f,b}_{SSD}(J^\pi)

758:  = \frac{2 a_{2\nu}}{m^{11}_e}

759: \int_{m_e}^{E_i-E_f-m_e} 

760: g^{f,b}_{J^\pi}(E_{e1},E_{e2},E_{\nu 1},E_{\nu 2})

761: F_0(Z_f,E_{e1}) p_{e1} E_{e1} dE_{e1}\times

762: \nonumber\\ 

763: \int_{m_e}^{E_i-E_f-E_{e1}} F_0(Z_f,p_{e2}) p_{e2} E_{e2} dE_{e2}

764: \int_{0}^{E_i-E_f-E_{e1}-E_{e2}}  E_{\nu 2}^2  E_{\nu 1}^2  

765: d E_{\nu 1}

766: \end{eqnarray}

767: with

768: \begin{eqnarray}

769: g^{f,b}_{0^+}(E_{e1},E_{e2},E_{\nu 1},E_{\nu 2})

770: &=& 

771: {m^2_e} \left[\frac{1}{3}(K^{f,b}K^{f,b}+{L}^{f,b}{L}^{f,b}+

772: {K}^{f,b}{L}^{f,b})\right]\nonumber \\

773: g^{f,b}_{2^+}(E_{e1},E_{e2},E_{\nu 1},E_{\nu 2})

774: &=& 

775: {m^2_e} ~\left({K}^{f,b} - {L}^{f,b}\right)^2. 

776: \end{eqnarray}

777: 

778: 

779: 

780: 

781: \section{Characteristics of double beta decays}

782: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

783: 

784: 

785: In what follows we calculate the characteristics of the 

786: double beta decay mainly for two nuclei 

787: $^{100}{\rm Mo}$ and $^{76}{\rm Ge}$ for which the highest number of

788: events has been collected in experiment (see Ref. \cite{nemo}

789: and \cite{klapdor} respectively).

790: 

791: 

792: 

793: \subsection{Double beta decay of $^{100}{\rm Mo}$}

794: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

795: 

796: 

797: The NEMO-3 collaboration has detected about 

798: 219 000  $(0^+ \rightarrow 0^+)$-decays of $^{100}{\rm Mo}$ 

799: \cite{nemo}. 

800: The signal to background ratio is very high  S/B =  

801: 44 and the background is at the level of ~2.5\% only. 

802: All parameters of the decay: the sum of the electron energies, the energy of

803: each electron and the angular distribution (angular correlation of electrons) 

804: have been  measured. 

805: 

806: In the case of $^{100}{\rm Mo}$ the decay proceeds mainly 

807: through the $1^+$ intermediate nucleus and  

808: the single state dominance (SSD) hypothesis should give  a good 

809: approximation. This is also confirmed by spectra measurements in NEMO-3 

810: experiment~\cite{ARN04,SHI06}.  

811: Since $E_m - E_i \sim E_i - E_f $,  the lepton energies are important 

812: in the energy-denominators (\ref{prop}),  and consequently, in the rates.  

813: 

814: In the SSD approximation one can calculate the probability (NME)

815: using existing  experimental data for the beta-decay and the electron 

816: capture  of  $^{100}{\rm Tc}$  which is the intermediate dominating state. 

817: Accuracy of this  ``phenomenological''  calculation  is about  50\%, 

818: mainly because of poor experimental

819: accuracy for the electron capture process. 

820: 

821: Using the SSD approximation we calculated the 

822: $2\nu\beta\beta$-decay half-life of $^{100}{\rm Mo}$ to ground state  

823: for fermionic \cite{DKSS} and bosonic neutrinos (see sect. 3)  

824: \begin{equation} 

825: T_{1/2}^{f}(0^+_{g.s.}) = 6.8~10^{18} {\rm years}, ~~~

826: T_{1/2}^{b}(0^+_{g.s.}) = 8.9~10^{19}  {\rm years}, 

827: \end{equation}

828: so that the ratio of probabilities equals

829: \begin{equation}

830: r_0(0^+_{g.s.}) = 0.076.

831: \label{r0gs}

832: \end{equation}

833: The ratio $r_0(0^+_{g.s.})$ determines the weight with which  the bosonic

834: component  enters the total rate and differential distribution [see Eq.(\ref{totpro}]. 

835: For small $r_0$, a substantial modification of the distribution is expected for 

836:  $\sin^2 \chi$ being close to 1. 

837: 

838: The higher intermediate levels can give some (basically unknown) 

839: contribution  and this produces a systematic error in our analysis.  

840: To evaluate effect of the higher states, one can consider 

841: the extreme case described by the  higher states dominance (HSD) approximation,

842: which allows one to factorize the nuclear matrix element and 

843: integration over the phase space of outgoing leptons. In this case

844: the main contribution to the $2\nu\beta\beta$-decay matrix element comes

845: from the transition through higher energy states (including the region of 

846: the Gamow-Teller resonance)  of the intermediate nucleus.

847: Thus, the lepton energies in the denominators (\ref{prop}) can be neglected

848: (or approximated by $(E_f - E_i)/2$ ) due to a large value of $E_n - E_i$.

849: The fermionic and bosonic $2\nu\beta\beta$-decay rates are 

850: associated with different nuclear matrix elements [see Eq. (\ref{nmef})

851: and (\ref{nmeb})].

852: They can be evaluated within an appropriate nuclear model like

853: Quasiparticle Random Phase approximation (QRPA) or Nuclear Shell Model

854: (NSM). Then, the evaluated values of $2\nu\beta\beta$-decay half-life  and

855: ratio $r_0(0^+_{g.s.})$ are model dependent. Contrary, the normalized

856: differential characteristics are model independent for cases of pure fermionic

857: and bosonic neutrinos. 

858: 

859:  

860: %%%%%%%%ffff1%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

861: \begin{figure}[tb]

862: \begin{center}

863: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig1.eps}

864: \caption{The differential decay rates normalized to the total decay rate 

865: vs. the sum of the kinetic energy of outgoing electrons $T$ for 

866: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ 

867: to the ground state of final nucleus.

868: The results are presented for the cases of pure fermionic and pure bosonic neutrinos.

869: The calculations have been performed within the single-state dominance hypothesis

870: (SSD) and with the assumption of dominance of higher lying states (HSD).

871: }

872: \label{mototapp}

873: \end{center}

874: \end{figure}

875: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

876: 

877: 

878: %%%%%%%%ffff2%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

879: \begin{figure}[tb]

880: \begin{center}

881: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig2.eps}

882: \caption{The single electron differential decay rate normalized 

883: to the total decay rate vs. the electron energy for 

884: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ 

885: to the ground state of final nucleus.

886: $E$ and $m_e$ represent the energy and mass of the electron, respectively.

887: The results are presented for the cases of pure fermionic and pure bosonic neutrinos.

888: The conventions are the same as in Fig. \protect\ref{mototapp}.

889: }

890: \label{mosinapp}

891: \end{center}

892: \end{figure}

893: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

894: 

895: 

896: The energy spectra of electrons calculated in the 

897: SSD and  HSD  approximations are presented 

898: in the figs. (\ref{mototapp}) and (\ref{mosinapp}). 

899: The SSD approximation gives slightly wider spectra of  

900: two electrons both for fermionic and bosonic neutrinos.  

901: The spectra for bosonic neutrinos are softer in both approximations. 

902: In particular, the  maxima  of SSD and HSD spectra are shifted to 

903: low energies for bosonic neutrinos by about 15 \% with respect 

904: to fermionic-neutrino  spectra. This shift does not 

905: depend on the approximation and therefore can be considered  as the solid 

906: signature of bosonic neutrino. 

907: Also the energy spectrum for single electron becomes softer in the bosonic 

908: case (Fig. \ref{mosinapp}).  

909: 

910: In Fig.~\ref{mosum} we show the energy spectra of two electrons for 

911: different values of the bosonic-fraction $\sin^2 \chi$. 

912: With increase of  $\sin^2 \chi$ the spectra shift to smaller energies. 

913: Due to smallness of $r_0$ substantial shift occurs only when 

914: $\sin^2 \chi$ is close to 1.0 

915: 

916: %%%%%%%%ffff3%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

917: \begin{figure}[tb]

918: \begin{center}

919: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig3.eps}

920: \caption{The differential decay rates normalized to the total decay rate 

921: vs. the sum of the kinetic energy of outgoing electrons $T$ for 

922: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ 

923: to the ground state of final nucleus.

924: The results are presented for different values of 

925: the squared admixture $\sin^2\chi$ of the bosonic component.

926: The spectra have been calculated in the SSD approximation.

927: } 

928: \label{mosum}

929: \end{center}

930: \end{figure}

931: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

932: 

933: In Fig.~\ref{mosingle} we show the energy spectra of 

934: single electrons for different values of  $\sin^2 \chi$. 

935: A substantial change occurs at very low energies, with 

936: $E_{kin} = 0.3$ MeV being a fixed point.  

937: For $E_{kin} < 0.3$ MeV the distribution increase with  $\sin^2 \chi$, 

938: whereas for $E_{kin} =  0.3 - 1.4$ MeV it decreases.\\ 

939: 

940: %%%%%%%%ffff4%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

941: \begin{figure}[tb]

942: \begin{center}

943: \includegraphics[width=14.0cm, height=10.cm, angle=0]{bbb_fig4.eps}

944: \caption{

945: The single electron differential decay rate normalized 

946: to the total decay rate vs. the electron energy for 

947: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ 

948: to the ground state of final nucleus.

949: The results are presented for different values of 

950: the squared admixture $\sin^2\chi$ of the bosonic component.

951: The spectra have been calculated in the

952: SSD approximation.

953: The conventions are the same as in Fig. \protect\ref{mosinapp}.

954: }

955: \label{mosingle}

956: \end{center}

957: \end{figure}

958: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

959: 

960: As we mentioned before,  the rates of transitions to first excited

961: $2^+_1$ state are affected by the presence of bosonic neutrino component  

962: in the opposite (to $0^+$) way.  

963: Furthermore, in the SSD approximation the ratio of decay rates to the excited

964: $2^+$ state and to the $0^+_{g.s.}$ ground state does not depend on

965: the $\log ft_{EC}$ value, which is not measured accurately enough. 

966: For the $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ 

967: within the SSD approximation we obtain 

968: \be

969: T^{}_{1/2}(2^+_1) &=& 1.7~ 10^{23}~{\rm years} ~~~~~~~

970: ({\rm fermionic}~\nu) \nonumber\\

971: &=& 2.4~ 10^{22}~{\rm years} ~~~~~~~

972: ({\rm bosonic}~\nu).

973: \ee

974: Then the  ratio of the bosonic and fermionic half-lives equals

975: \be

976: r_0 (2^+_1) = 7.1.   

977: \ee

978: The bosonic rate is larger in agreement with our 

979: qualitative consideration in sect. 2.

980: 

981: 

982: The best lower bound on the $2\nu\beta\beta$-decay half-life to 

983: excited $2^+_1$ state is $1.6~10^{21}$ years \cite{BAR95}. 

984: %%Thus a small progress

985: %%in the measurement of this decay mode will practically exclude 

986: %%the possibility of pure bosonic neutrinos. 

987: The current limit of NEMO-3 experiment is $1.1~10^{21}$ years \cite{ARN07} 

988: (for 1 year of measurements). After 5 years of 

989: measurements with the present low-radon background conditions sensitivity will 

990: increase up to $\sim 10^{22}$ years thus approaching the 

991: prediction in the case of bosonic neutrinos. 

992: Due to the large value of $r_0$ even a small fraction of bosonic neutrinos 

993: can produce significant distortion of the standard (fermionic) 

994: spectra. 

995: 

996: Modifications of the spectra are opposite for the decay of  $^{100}{\rm Mo}$

997: into $2^+$ excited state: the spectra become harder with increase of 

998: $\sin^2\chi $ (see  Fig.~\ref{mosumex} and \ref{mosinex}). 

999: This is apparently related to the change of the spin of the nuclei. 

1000: In the case of $0^+ -2^+_1$ transition the leptonic system should 

1001: take spin 2 and therefore due to polarization of leptons 

1002: (determined by V - A character 

1003: of interactions) both electrons  move preferably in 

1004: the same direction (hemisphere)  

1005: and two antineutrinos in the opposite direction with the corresponding 

1006: Pauli blocking factor. In the case of  bosonic neutrinos the Pauli blocking  

1007: effect is reduced and therefore the electrons can be more aligned and consequently 

1008: have higher energies. Correspondingly the spectrum becomes harder. 

1009: In the case of $0^+ - 0^+$ transition the total leptonic momentum is zero, 

1010: so that the electrons move in the opposite directions. 

1011: 

1012: According to Fig.~\ref{mosumex} even 10 $\%$ of "bosonic" admixture 

1013: gives  substantial distortion effect and this fact can be used in the future 

1014: experiments.   

1015: 

1016: %%%%%%%%ffff5%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1017: \begin{figure}[tb]

1018: \begin{center}

1019: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig5.eps}

1020: \caption{

1021: The differential decay rates normalized to the total decay rate 

1022: vs. the sum of the kinetic energy of outgoing electrons $T$ for 

1023: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ to the excited $2^+_1$ state 

1024: of final nucleus. 

1025: The results are presented for different values of 

1026: the squared admixture $\sin^2\chi$ of the bosonic component.

1027: The spectra have been calculated in the

1028: SSD approximation.

1029: }

1030: \label{mosumex}

1031: \end{center}

1032: \end{figure}

1033: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1034: 

1035: 

1036: %%%%%%%%ffff6%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1037: \begin{figure}[tb]

1038: \begin{center}

1039: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig6.eps}

1040: \caption{

1041: The single electron differential decay rate normalized 

1042: to the total decay rate vs. the electron energy for 

1043: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ to the excited $2^+_1$ state 

1044: of final nucleus.

1045: The results are presented for different values of 

1046: the squared admixture $\sin^2\chi$ of the bosonic component.

1047: The spectra have been calculated in the SSD approximation.

1048: The conventions are the same as in Fig. \protect\ref{mosinapp}.

1049: }

1050: \label{mosinex}

1051: \end{center}

1052: \end{figure}

1053: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1054: 

1055: The angular distribution of outgoing electrons \cite{SDS} can be 

1056: written as 

1057: \be

1058: \frac{d W_{f,b}(J^\pi)}{d \cos \theta } =  

1059: \frac{W_{f,b}(J^\pi)}{2} 

1060:  (1 + \kappa^{f,b}(J^\pi) \cos \theta), 

1061: \label{angular}

1062: \ee

1063: where $\theta$ is the angle between two electrons. 

1064: For  $0^+ - 0^+$ transition and fermionic 

1065: neutrinos in the SSD approximation  

1066: \be

1067: \kappa^f(0^+_{g.s.}) = -0.627 ~~~{\rm (fermionic~~ neutrino)}. 

1068: \ee 

1069: (The HSD approximation gives similar number: $ -0.646 $.)  

1070: Notice that the preferable direction is $\theta = 180^{\circ}$ when electrons move in the 

1071: opposite directions. The configuration with 

1072:  the same direction of two electrons is suppressed.  

1073: For bosonic neutrinos we find 

1074: \begin{equation} 

1075: \kappa^b (0^+_{g.s.}) = -0.344~~~   {\rm (bosonic~~ neutrino)}.  

1076: \end{equation}

1077: (The HSD approximation gives $-0.422$.)  

1078: So,  the configuration with the same direction of electrons is less suppressed 

1079: and the distribution is more isotropic (flatter) than in the fermionic case. 

1080: 

1081: 

1082: \subsection{$^{76}{\rm Ge}$ double beta decay}

1083: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1084: 

1085: %%In comparison with $^{100}{\rm Mo}$,  

1086: The statistics of $^{76}{\rm Ge}$ decays

1087: is  about  113000 events, the background is rather high, S/B =1.3,  and  only

1088: the  sum of two electron energies is measured~\cite{klapdor}.

1089: The  systematic error can be as large as 

1090: 10\%  and the main source of the error is the background. 

1091: One has to estimate this background  independently and

1092: make subtraction. 

1093: %%Since the background in not very well known it  gives a large error. 

1094: So, one  can shift  the  spectrum

1095: and its maximum within  the error. Furthermore, the energy spectrum of two

1096: electrons  

1097: %the $\beta\beta$ decay  

1098: starts to dominate over the background above  0.7 MeV  which means that 

1099: the maximum of the spectrum is not observed. 

1100: The advantage of $^{76}{\rm Ge}$ is that there is practically no difference 

1101: between the results of HSD and SSD approximations for 

1102: the energy distributions 

1103: because the nearest $1^+_1$ state of the intermediate nucleus is 

1104: lying high enough. Thus, one does not need to make assumptions 

1105: about SSD or HSD.

1106: In this way the conclusion does not depend on the nuclear structure details. 

1107: 

1108: %In the case of $^{76}{\rm Ge}$  the situation may be more favorable in the 

1109: %sense 

1110: %SSD  and HSD cases are practically identical due to a high-lying first 

1111: %$1^+$ state of the intermediate nucleus.

1112: 

1113: 

1114: In the  HSD approximation,  evaluating  the 

1115: phase space integrals and nuclear matrix elements

1116: within the proton-neutron QRPA we find 

1117: \begin{equation} 

1118: r_0 (0^+_{g.s.}) = 0.0014.

1119: \end{equation}

1120: This smallness is related to a large extend to  

1121: high energies of the intermediate states, $E_m - E_i$ 

1122: in comparison with leptonic energies restricted by the energy release 

1123: $ E_l < (E_i - E_f)/2$: 

1124: $E_l \ll E_m - E_i$. 

1125: According to (\ref{prop}) the factors $K^b_m$, $L^b_m$ and consequently 

1126: the rate are zero in the limit $ E_l = 0$. 

1127: In the lowest approximation we obtain 

1128: \begin{equation}

1129: K^b_m, L^b_m \sim 

1130: \frac{[(E_{\nu 2} - E_{\nu 1}) \pm (E_{e2} - E_{e1}) ]}{(E_m - E_i)^2}, 

1131: \end{equation}

1132: (where plus sign is for $K$-factors). 

1133: Then the ratio of the rates can be estimated as 

1134: \begin{equation}

1135: r_0(0^+_{g.s.}) \sim \frac{\epsilon_l^2}{4(E_m - E_i)^2},  

1136: \end{equation}

1137: where $\epsilon_l$ is the  average energy of the lepton. 

1138: Taking parameters of the  $^{76}{\rm Ge}$ -decay we find $r_0 \approx 10^{-3}$ 

1139: in a good agreement with the calculations in QRPA. 

1140: 

1141: 

1142: In Fig.~\ref{gesum} we show the normalized distributions of the total 

1143: energy of two electrons for pure fermionic and bosonic neutrinos. As in 

1144: the case of $^{100}{\rm Mo}$, the  decay  with bosonic neutrinos 

1145: has softer spectrum.  

1146: The energy distribution of single electron is shown in Fig.~\ref{gesin}

1147: 

1148: Due to a small value of $r_0 (0^+_{g.s.})$ 

1149: a substantial effect of the bosonic component 

1150: should show up only for $\sin^2 \chi$ being very close to 1: 

1151: $(1 - \sin^2 \chi)^2 \sim 10 r_0(0^+_{g.s.})$. 

1152: So studies of the spectra are not sensitive to $\sin^2\chi$. 

1153: In contrast, the total rate of the $^{76}{\rm Ge}$ decay gives 

1154: a strong bound on $\sin^2\chi$. 

1155: 

1156: %%%%%%%%ffff7%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1157: \begin{figure}[tb]

1158: \begin{center}

1159: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig7.eps}

1160: \caption{

1161: The differential decay rates normalized to the total decay rate 

1162: vs. the sum of the kinetic energy of outgoing electrons $T$ for 

1163: $2\nu\beta\beta$-decay of $^{76}{\rm Ge}$ to the ground state of final nucleus.

1164: The results are presented for the cases of pure fermionic and pure bosonic neutrinos.

1165: The calculations have been performed with the HSD assumption.

1166: }

1167: \label{gesum}

1168: \end{center}

1169: \end{figure}

1170: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1171: 

1172: %%%%%%%%ffff8%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1173: \begin{figure}[tb]

1174: \begin{center}

1175: \includegraphics[width=14.0cm, height=10.0cm, angle=0]{bbb_fig8.eps}

1176: \caption{

1177: The single electron differential decay rate normalized 

1178: to the total decay rate vs. the electron energy for 

1179: $2\nu\beta\beta$-decay of $^{76}{\rm Ge}$ to the ground state of final nucleus.

1180: $E$ and $m_e$ represent the energy and mass of the electron, respectively.

1181: The results are presented for the cases of pure fermionic and 

1182: pure bosonic neutrinos.

1183: The calculations have been performed with the HSD assumption.

1184: The conventions are the same as in Fig. \protect\ref{mosinapp}.

1185: }

1186: \label{gesin}

1187: \end{center}

1188: \end{figure}

1189: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1190: 

1191: 

1192: 

1193: \section{Bounds on bosonic neutrinos}

1194: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1195: 

1196: One can  search for/restrict the 

1197: bosonic  or partly bosonic neutrino

1198: using total rates,  ratios of rates of the transitions to the 

1199: excited and ground states, 

1200: energy spectra, and  angular distributions.   

1201: Let us evaluate the bounds on $\sin^2\chi$ that  can be obtained 

1202: from the existing data using these methods. 

1203: 

1204: 

1205: As follows from our general discussion in sec.~3,  

1206: for $0^+ \rightarrow 0^+$ transitions: $r_0 \ll 1$. 

1207: For nuclei with small $r_0$ the best bound on bosonic neutrino fraction  

1208: can be obtained from the total rates.  A modification of the spectrum 

1209: due to presence of bosonic component is small. In contrast, the 

1210: strongest modification of the spectrum  is expected 

1211: for the nuclei with large $r_0$. 

1212: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1213: This is true,   e.g., for $0^+ \rightarrow 2^+$ transition, 

1214: where $r_0 \gg  1$.\\ 

1215: 

1216: 

1217: 

1218: {\it 1) Method 1:} Comparison of the predicted and measured  

1219: half-life times. Using (\ref{totpro}) we can write 

1220: \begin{equation}

1221: \sin^2\chi = \frac{1}{1 + r_0}\left[1 - 

1222: \sqrt{\frac{T^f_{1/2}}{T^{exp}_{1/2}} 

1223: - r_0 \left(1 -\frac{T^f_{1/2}}{T^{exp}_{1/2}}\right)}

1224: \right], 

1225: \label{bound}

1226: \end{equation}

1227: where $r_0 = T^f_{1/2}/T^b_{1/2}$, 

1228: $T^f_{1/2}$ ($T^b_{1/2}$)  are the theoretically predicted 

1229: life-times for 

1230: fermionic (bosonic) neutrinos and  $T^{exp}_{1/2}$ is the experimentally 

1231: measured life-time. In the case of agreement between  the  measured  

1232: and the predicted (for fermionic neutrinos) life-times, we can 

1233: use (\ref{bound}) to establish the bound on parameter $\sin^2\chi$: 

1234: \begin{equation}

1235: \sin^2\chi < \frac{1}{1 + r_0}\left[1 - 

1236: \sqrt{\frac{T^{f-min}_{1/2}}{T^{exp-max}_{1/2}} 

1237: - r_0 \left(1 -\frac{T^{f-min}_{1/2}}{T^{exp-max}_{1/2}}\right)}

1238: \right].

1239: \label{bound1}

1240: \end{equation}

1241: Here, $T^{f-min}_{1/2}$ and $T^{exp-max}_{1/2}$ are, respectively, minimal theoretical 

1242: value within a considered nuclear model (e.g., QRPA and its modification, NSM)

1243: and maximal experimental value of the permitted experimental range

1244: of the $2\nu\beta\beta$-decay half-life.

1245: For  $r_0 \ll 1$ and  $r_0$ smaller the relative accuracy of determination of 

1246: $T^f_{1/2}/T^{exp}_{1/2}$ the terms proportional to $r_0$   

1247: in (\ref{bound1}) can be omitted. Then we get 

1248: $\sin^2\chi < (1 - \sqrt{{T^{f-min}_{1/2}}/{T^{exp-max}_{1/2}}})$.

1249: 

1250: 

1251: Apparently, this method  requires knowledge of the nuclear 

1252: matrix element, and as 

1253: we mentioned above,  reliable estimations  can be done for some nuclei 

1254: e.g., $^{100}{\rm Mo}$ and $^{116}{\rm Cd}$  assuming SSD hypothesis. For some other 

1255: nuclear systems nuclear models have to be considered. The two basic 

1256: approaches used so far for the evaluation of the double beta decay 

1257: matrix elements are the QRPA and the NSM. For the $2\nu\beta\beta$-decay 

1258: of $^{76}{\rm Ge}$ the predicted half-lives 

1259: are $7.7~10^{20}-1.4~10^{21}$ years (QRPA) \cite{SUH98} and 

1260: %$2.6~10^{21}$ years (NSM) \cite{NSM}. 

1261: $1.15~10^{21}$ years (NSM)~\cite{NSM}.

1262: The experimental half-life (average half-life value is 

1263: $(1.5\pm 0.1)~ 10^{21}$ years \cite{BAR06}) 

1264: is in rather good agreement with the theoretical ones 

1265: for fermionic neutrino within 

1266: uncertainty characterized by the factor $\sim  2$  (see \cite{SUH98}).

1267: For pure bosonic neutrinos $r_0(0^+_{g.s.}) \approx 10^{-3}$ 

1268: (QRPA) and therefore 

1269: for the  half-life time we would have 

1270: $T^{b}_{1/2} \approx 1.5~ 10^{24} $ years, 

1271: which is in contradiction with the experimental value. 

1272: So, purely bosonic neutrino is certainly excluded. 

1273:  

1274: The axial-vector coupling constant $g_A$ is a significant source of 

1275: uncertainty in the theoretical calculation of the $2\nu\beta\beta$-decay

1276: rate, which is proportional to $g_A^4$. The commonly adopted values are 

1277: $g_A=1.0$ (by assuming quenching in nuclear medium) and $g_A = 1.25$ 

1278: (as for free nucleon). This gives about 1.5 uncertainty in NME's. 

1279: 

1280: For factor 2 uncertainty in NME we obtain factor 4 uncertainty 

1281: in $T^f_{1/2}$. Therefore taking $T^f_{1/2} \sim T^{exp}_{1/2}$,  

1282: we can put the  bound 

1283: \begin{equation}

1284: \frac{T^{f-min}_{1/2}}{T^{exp-max}_{1/2}} > \frac{1}{4}. 

1285: \end{equation}

1286: Then, eq. (\ref{bound1}) gives 

1287: \begin{equation} 

1288: \sin^2\chi <  0.50. 

1289: \end{equation}

1290: Notice that uncertainty in $T^f_{1/2}$ (and not $r_0$) dominates in this bound. 

1291: 

1292: 

1293: We can also use the half-life time of $^{100}{\rm Mo}$. 

1294: Here $r_0(0^+_{g.s.})$ is much larger (\ref{r0gs})  but the accuracy of calculations of 

1295: NME is better. 

1296: Taking  SSD approximation we  can calculate the half life with 50\% accuracy:

1297: $T^f_{1/2} = (6.84 \pm 3.42)~10^{18}$ years \cite{DKSS}. This value is in 

1298: agreement with NEMO-3 value, 

1299: $T^{exp}_{1/2} = (7.11 \pm 0.54)~ 10^{18}$ years \cite{nemo}. 

1300: Plugging these numbers into (\ref{bound1}) we 

1301: obtain for $r_0(0^+_{g.s.}) = 0.086$ 

1302: \begin{equation}

1303: \sin^2 \chi < 0.34.\,\,\,

1304: \end{equation}

1305: %%that $b^2$ = 0.65. In this case the prediction will be

1306: %%$(16.8 \pm 8.4)~ 10^{18}$ years, because "bosonic" contribution to 

1307: %%half-life is very small (suppressed by factor 0.08642).    So, in this case

1308: %%it would be in contradiction 

1309: %%with NEMO-3 result. A small admixture of HSD cannot change this 

1310: %%conclusion.

1311: Notice that the accuracy of predicted half-life value is connected 

1312: with experimental  accuracy for EC (electron capture)

1313: half-life of $^{100}{\rm Tc}$ \cite{GAR93}. This accuracy can be 

1314: improved in the future experiments{\footnote {In ref.~\cite{GAR93} 

1315: Mo enriched to 97.4\% 

1316: was used and the main background was 

1317: connected with X-rays from different Tc isotopes which were produced in 

1318: the sample due to (p,n) and (p,$\alpha$) reactions on different Mo isotopes, 

1319: from $^{92}{\rm Mo}$ to $^{98}{\rm Mo}$; see Table II in~\cite{GAR93}. 

1320: If one uses Mo enriched 

1321: to 99\% (or more) then the mentioned above background would be much lower 

1322: and the accuracy of the measurement would be several times better.}}

1323: %in a few time more precisely.}

1324: %{\bf [[how, when, give ref ]]} 

1325: down to $\sim 10\%$ and 

1326: correspondingly, the sensitivity to  $\sin^2\chi$

1327: can reach   $\sim 0.1$. Unfortunately, there is only one (not very

1328: precise) EC measurement for $^{100}{\rm Tc}$ and thus the above limit on 

1329: $sin^2 \chi$ is not reliable enough.   

1330: 

1331: 

1332: Even stronger bound can be obtained from studies of $^{116}{\rm Cd}$ -decay. 

1333: Recently a precise estimation of half-life 

1334: value based on the  SSD approximation  and  information from the 

1335: $^{116}{\rm Cd(p,n)}$ reaction was obtained: $T^f_{1/2} = (2.76\pm 0.12)~ 10^{19}$ years 

1336: \cite{SAS07}. 

1337: This prediction is in a very good agreement with experimental value 

1338: (The experimental average is $(2.8 \pm 0.2)~ 10^{19}$ years 

1339: \cite{BAR06}). Using these results we obtain from (\ref{bound1}) 

1340: \begin{equation}

1341: \sin^2 \chi < 0.06. \,\,\, 

1342: \end{equation}

1343: It should be noticed that the result of ref. \cite{SAS07} substantially differs 

1344: from the earlier estimation $T^f_{1/2} =   (1.1 \pm 0.3)~ 10^{19}$ years

1345: \cite{DKSS} (also based on SSD and measured value of electron capture rate

1346: of $^{116}{\rm In}$ \cite{BHA98}). This result 

1347: disagrees with the experimental value and could be interpreted 

1348: as the effect of partly bosonic neutrino

1349: with  $\sin^2 \chi   \sim 0.4$. \\

1350: 

1351: %The accuracy  of estimation in \cite{DKSS} is limited by the 

1352: %experimental accuracy 

1353: %for EC in $^{116}$In \cite{BHA98}. 

1354: %{\bf [[how this problem is resolved in  \cite{SAS07}???]]}\\

1355: 

1356: %%the conservative bound $b^2 < 0.7$ can be imposed in this 

1357:  

1358: 

1359: {\it 2) Method 2:}  Measurements of the differential characteristics of the decays: 

1360: %In the case of  the $^{100}Mo$  an uncertainty is related to the 

1361: %assumption of SSD.  

1362: shapes of the energy spectra (sum energy 

1363: and single

1364: electron energy) and  angular distribution. Such information is provided  now 

1365: by  NEMO-3 for

1366: $^{100}{\rm Mo}$, $^{82}{\rm Se}$, 

1367: $^{116}{\rm Cd}$, $^{150}{\rm Nd}$, $^{96}{\rm Zr}$ and $^{48}{\rm Ca}$. 

1368: In the future the results for $^{130}{\rm Te}$ will be also available 

1369: \cite{nemo,ARN04,SHI06,baranew}. 

1370: In this method one 

1371: compares the experimental  and theoretical energy spectra   

1372: as well as the   angular distribution.

1373: In practice one should perform the statistical fit of the spectra by  

1374: a general distribution (\ref{distr}) with  $\sin^2 \chi$

1375: being a free parameter.  

1376: As we have seen the spectral method  has substantial sensitivity 

1377: to $\sin^2\chi$ 

1378: for nuclei and transitions with large $r_0$.  That includes  

1379: $^{100}{\rm Mo}$, as well as transitions to the excited states. 

1380: $^{76}{\rm Ge}$  with very small $r_0$ has no high sensitivity.  

1381: 

1382: 

1383: 

1384: a) Let us consider first the energy spectra of  

1385: $0^+_{g.s.} \rightarrow 0^+_{g.s.}$ decay of $^{100}{\rm Mo}$~\cite{nemo}. 

1386: In the present paper we will not perform detailed statistical analysis 

1387: of the spectra,  postponing this to the time when measurements  

1388: will be finished and all

1389: careful calibrations will be done. Instead, we give some 

1390: qualitative estimates. There is a reasonable agreement between 

1391: the predicted energy spectrum of  two 

1392: electrons and  the experimental points. 

1393: Therefore  we can certainly exclude the 

1394: pure bosonic case ($\sin^2\chi  = 1$). 

1395: Furthermore,  comparing the results of Fig.~\ref{mosum}

1396: (essentially, the relative shift of the maximum of spectrum)

1397: with the experimental spectrum  we can put the conservative bound  

1398: $\sin^2 \chi < 0.6$.

1399: In fact,  there is no ideal agreement between data and theoretical 

1400: spectrum.  A better fit can be obtained  for $\sin^2\chi \sim 0.4-0.5$.  

1401: 

1402: b) Let us comment on the   

1403: single-electron energy spectrum from $^{100}{\rm Mo}$ decay. 

1404: The data reasonably well agree with the predictions from the fermionic 

1405: SSD mechanism, but  

1406: some difference exists between the data 

1407: and the fermionic HSD-mechanism predictions. From this 

1408: it was concluded that the  SSD mechanism is more relevant 

1409: here \cite{ARN04,SHI06}. Comparing  the experimental 

1410: data and spectra for partly bosonic neutrinos 

1411: (Fig.~\ref{mosingle}) we  obtain:  $\sin^2\chi < 0.7$.  

1412: 

1413: Notice that the SSD spectrum does not show an

1414: ideal agreement with data either. 

1415: There is some discrepancy, especially in the low

1416: energy region ($E = 0.2-0.4$ MeV). 

1417: That could be explained by the effect of 

1418: partly bosonic neutrinos with $\sin^2 \chi \sim$ 0.5 - 0.6.\\

1419: 

1420: 

1421: Complete analysis 

1422: %%(using maximal likelihood methods) 

1423: of all existing NEMO-3 information (energy and angular distributions) 

1424: using e.g. maximal likelihood methods, 

1425: will have a higher sensitivity to $\sin^2 \chi$. However, 

1426: it is difficult to expect a better bound than 

1427:  $\sin^2 \chi \sim 0.4-0.5$, mainly because of the

1428: existing disagreement between 

1429: the data and Monte Carlo (MC) simulations. In fact,  

1430: it can be just some systematic effect connected 

1431: to the  present poor understanding of response function of 

1432: the detector. If in future 

1433: the NEMO experimental data turn out to be  in much better agreement with 

1434: the MC-simulated spectrum, 

1435: the  sensitivity to partly bosonic neutrino will be  improved 

1436: down to $\sin^2 \chi= 0.2 - 0.3$.\\ 

1437: %% provided that we will not see the positive effect.\\

1438: 

1439: 

1440: 

1441: 

1442: {\it 3) Method 3:} Determination of the ratios of half-lives to excited and ground state, 

1443: \begin{equation}

1444: r^*_{f,b} (J^\pi) \equiv 

1445: \frac{T^{f,b}_{1/2}(J^\pi)} 

1446: {T^{f,b}_{1/2}(0^+_{g.s.})},

1447: \label{ratio-ex}

1448: \end{equation}

1449: separately for fermionic and bosonic neutrinos.

1450: For $2\nu\beta\beta$-decay of  $^{100}{\rm Mo}$ the ratio can be calculated 

1451: rather reliably using 

1452: the SSD-approximation.  The  advantage of this quantity

1453: is that the EC amplitude, 

1454: [(A,Z) $\rightarrow$ (A,Z+1) transition], which is not well determined, 

1455: cancels in  the ratio (\ref{ratio-ex}). 

1456: 

1457: For $^{100}{\rm Mo}$  the transitions to the ground $0^+_{g.s.}$ 

1458: and excited $0^+_1$ states 

1459: were detected,  and in fact, some  discrepancy has been observed.  

1460: The corresponding experimental ratio $r^*$ equals 

1461: \begin{equation}

1462: r^*_{exp.} (0^+_1) \simeq 80 

1463: \end{equation}

1464: (NEMO-3 results \cite{nemo,ARN07}),  

1465: whereas within the SSD approach the  calculated ones are

1466: \begin{eqnarray}

1467: r^* (0^+_1) &\simeq& 61 ~~~~~~~({\rm fermionic}~\nu) \nonumber\\

1468: &\simeq& 73 ~~~~~~~({\rm bosonic}~\nu).

1469: \end{eqnarray}

1470: A bosonic neutrino fits the data slightly better but the differences are 

1471: probably beyond the accuracy of the SSD assumption. Still it is also 

1472: necessary to  improve statistics in measurements of the transition to 

1473: excited $0^+_1$ state.

1474: 

1475: Contrary to the case of $0^+$ excited state, the ratio 

1476: of $2\nu\beta\beta$-decay 

1477: half-lives to excited $2^+$ and ground state  is expected to be 

1478: strongly different for bosonic and fermionic neutrinos. 

1479: Using the SSD approximation for the 

1480: $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ we found  

1481: \begin{eqnarray}

1482: r^* (2^+_1) &\simeq& 2.5~10^{4} ~~~~~~~({\rm fermionic}~\nu) \nonumber\\

1483: &\simeq& 2.7~10^{2} ~~~~~~~({\rm bosonic}~\nu).

1484: \end{eqnarray}

1485: The $2\nu\beta\beta$-decay of $^{100}{\rm Mo}$ 

1486: to excited $2^+_1$ state has been

1487: not measured yet. Using the best experimental limit on the half-life

1488: found in \cite{BAR95} we get

1489: \begin{equation}

1490: r^*_{exp} (2^+_1) > 2.2~10^{2}. 

1491: \end{equation}

1492: This bound is close to the bosonic prediction. A further experimental 

1493: progress in measuring this nuclear transition will allow one to analyze also

1494: the case of partially bosonic neutrino, and therefore is highly required.  

1495: 

1496: 

1497: %Let us consider possible future tests of bosonic neutrino 

1498: %and improvements of bounds on the parameter $b$. 

1499: %Perhaps, it might be interesting to discuss also ECEC capture of 

1500: %$^{106}$Cd. In 

1501: %this case instead of electron energies $E_1$, $E_2$ in the denominators

1502: %(\ref{}) the mass of electron $-m_e$  appears. Here, the suppression of 

1503: %bosonic mode can be weak (SSD case)?\\

1504: 

1505: 

1506: 

1507:  

1508: \section{Conclusions}

1509: %%%%%%%%%%%%%%%%%%%

1510: 

1511: 

1512: \noindent

1513: A study of the double beta decay  can provide a sensitive test of the Pauli 

1514: exclusion principle and statistics of neutrinos. 

1515: (Notice, that relation between the statistics of neutrinos and 

1516: possible (small) violation of the Pauli principle is an open issue.) 

1517: Appearance of the bosonic component in the neutrino 

1518: states changes substantially 

1519: the total rates of the decays as well as  the energy and angular distributions. 

1520: We find, in particular,  that the ratio $r_0(0^+_{g.s.})$ of 

1521: the rates to ground state 

1522: for bosonic and fermionic neutrinos, is $< 10^{-3}$ 

1523: for $^{76}{\rm Ge}$ and $0.076$ for $^{100}{\rm Mo}$, 

1524: which excludes pure bosonic neutrinos. For transitions to $2^+$

1525: excited states $r_0 (2^+) \gg 1$, in particular  

1526: $r_0 (2^+_1)\simeq 7$. However, this 

1527: $2\nu\beta\beta$-decay channel has been not measured yet.\\

1528: 

1529: \noindent

1530: We have introduced phenomenological parameter 

1531: $\sin^2\chi$ that describes the mixed 

1532: statistics case of partly bosonic neutrinos. 

1533: The dependence of the energy spectra and 

1534: angular correlation of electrons on   $\sin^2\chi$ has been studied.

1535: The bound on  $\sin^2 \chi$  can be obtained by comparison of the 

1536: predicted and measured total rates of the decays. In spite of the 

1537: big difference of the rates for fermionic and bosonic neutrinos, 

1538: this method does not give strong and very reliable bound on  $\sin^2 \chi$

1539: due to uncertainties  

1540: in NME's.  The conservative upper bound  $\sin^2 \chi < 0.5$

1541:  is found using  the $^{100}{\rm Mo}$ and 

1542: $^{76}{\rm Ge}$ results. Much stronger bound,  

1543:  $\sin^2 \chi< 0.06$,  is obtained  from recent studies of $^{116}{\rm Cd}$, 

1544: however this bound requires further checks.\\ 

1545: 

1546: \noindent

1547: The method based on the study of the normalized energy and angular spectra

1548: is less affected by uncertainties in the NME's. 

1549: The transitions with 

1550: large $r_0(J^\pi)$ have the highest sensitivity to 

1551: spectrum distortions and therefore  $\sin^2 \chi$. 

1552: Using the data on 

1553: the $0^+_{g.s.} \rightarrow 0^+_{g.s.}$ transition 

1554: of  $^{100}{\rm Mo}$ we obtain the bound  $\sin^2 \chi<0.6$. 

1555: In the future this bound can be improved down to  

1556: $\sin^2 \chi\sim 0.2$.

1557: The $0^+_{g.s.} \rightarrow 2^+_1$ transition with $r_0 (2^+_1) \simeq 7$ 

1558: can give much stronger bound, but  here new, more sensitive experimental 

1559: results are needed. 

1560: We find that modification of the energy spectra due the presence of  

1561: the bosonic components is opposite for $0^+_{g.s.} \rightarrow 0^+_{g.s.}$ 

1562: and $0^+_{g.s.} \rightarrow 2^+_1$ transitions: 

1563: for $0^+_{g.s.} \rightarrow 0^+_{g.s.}$  

1564: the bosonic component leads to softer spectrum whereas for 

1565: $0^+_{g.s.} \rightarrow 2^+_1$ transitions to harder spectrum of electrons. 

1566: Also the presence of bosonic component 

1567: leads to flatter angular ($\cos \theta$) 

1568: distribution. \\

1569: 

1570: 

1571: \noindent

1572: Strong bound  (potentially down to  $\sin^2 \chi \sim 0.1-0.05$) 

1573: might be obtained 

1574: from measurements of ratios of the decay rates to the $2^+_1$ 

1575: excited and ground state. However, this requires further experimental progress. \\

1576: 

1577: \noindent

1578: We note that currently there are no restrictions on the admixture of bosonic component

1579: from the BBN. However, as it was indicated in \cite{hansen} the future BBN studies 

1580: will be able to constrain the fermi-bose parameter to $\kappa~ >~ 0.5$.    

1581: The bound on parameter $\sin^2\chi~<~0.6$  from the $2\nu\beta\beta$-decay results in

1582: $\kappa~ > - 0.2$.\\ 

1583: 

1584: 

1585: \noindent 

1586: In conclusion, the present data allow to put the conservative

1587: upper bound on the admixture of the bosonic component 

1588:  $\sin^2 \chi < 0.6$. With the 

1589: presently operating experiments this bound might be improved down 

1590: to $0.2$.  In future  one order of magnitude improvement seems feasible.   

1591: 

1592: 

1593: %- SSD: It is suppose that the dominant transition is through the

1594: %  $1^+$ ground state of the intermediate nucleus (e.g. Mo100, Cd116, 

1595: %Te128).

1596: %  The half-life one can calculate from the log ft values of

1597: %  corresponding EC capture and beta- decay of the ground state

1598: %  of the intermediate nucleus.

1599: 

1600: 

1601: 

1602: \section{Acknowledgments}

1603: 

1604: We are grateful to  L.B. Okun for helpful discussions.

1605: F. \v S and A Yu. S. acknowledge the support of  the EU ILIAS project 

1606: under the contract RII3-CT-2004-506222 and the VEGA Grant agency of the Slovak Republic 

1607: under the contract No.~1/0249/03. 

1608: A. Yu. S. is also grateful for support to the Alexander von Humboldt Foundation.  

1609: This work was supported by Russian Federal Agency 

1610: for Atomic Energy and by RFBR (grant 06-02-72553).

1611: 

1612: 

1613: 

1614: \bigskip

1615: 

1616: 

1617: \begin{thebibliography}{99}

1618: 

1619: %1

1620: \bi{ign-kuz}

1621: A.Yu.~Ignatiev, V.A.~Kuzmin, Yad. Fiz. {\bf 46} (1987) 786 

1622: [Sov.~J.~Nucl.~Phys. {\bf 46} (1987) 786];

1623: JETP Lett. {\bf 47} (1988) 4; \\

1624: L.B.~Okun, Pis'ma ZhETF, {\bf 46} (1987) 420 [JETP Lett. {\bf 46} (1987) 529];

1625: Yad. Fiz. {\bf 47} (1988) 1192;\\

1626: O.W. Greenberg, R.N. Mohapatra, Phys.~Rev.~Lett. {\bf 59} (1987) 2507, 

1627: {\bf 62} (1989) 712, Phys.~Rev. D{\bf 39} (1989) 2032;\\

1628: A.B.~Govorkov, Phys.~Lett. A{\bf 137} (1989) 7.

1629: 

1630: %2

1631: \bi{lbo-rev}

1632: L.B.~Okun, Uspekhi Fiz. Nauk {\bf 158} (1989) 293 [Sov.~Phys.~Usp. {\bf 32} (1989) 543],

1633: L.B.~Okun, Comments Nucl.~Part.~Phys., {\bf 19} (1989) 99.

1634: 

1635: %3

1636: \bi{exp-viol}

1637: M. Goldhaber, G. Scharff-Goldhaber, Phys. Rev. {\bf 73} (1948) 1472;\\ 

1638: E. Fishbach, T. Kirsten and O. Shaeffer, Phys. Rev. Lett. {\bf 20} (1968) 

1639: 1012;\\

1640: F. Reines, H.W. Sobel, Phys. Rev. Lett. {\bf 32} (1974) 954;\\

1641: B.A. Logan, A. Ljubicic, Phys. Rev. C {\bf 20} (1979) 1957;\\

1642: R.D. Amado, H. Primakoff, Phys. Rev. C {\bf 22} (1980) 1338;\\

1643: E. Ramberg and G. Snow, Phys. Lett. B {\bf 238} (1990) 438;\\

1644: K. Deilamian, J. D. Gillaspy and D. Kelleher, Phys. Rev. Lett. {\bf 74} (1995) 

1645: 4787;\\

1646: V. M. Novikov, A. A. Pomansky, E. Nolte, 

1647: JETP Lett. {\bf 51} (1990) 1;\\

1648: V.M. Novikov, et al., 

1649: Phys. Lett. B{\bf 240} (1990)  227;\\

1650: A.S. Barabash et al., JETP Lett. {\bf 68} (1998) 112;\\

1651: D. Javorsek II et al., Phys. Rev. Lett. {\bf 85} (2000) 2701;\\

1652: S. Bartalucci et al., Phys. Lett. B {\bf 641} (2006) 18.

1653: 

1654: %4

1655: \bi{exp-bar}

1656: R. Plaga, Z. Phys. A {\bf 333} (1990) 397;\\

1657: R. Bernabei et al., Phys. Lett. B {\bf 408} (1997) 439;\\

1658: E. Baron, R.N. Mohapatra, V.L. Teplitz, Phys. Rev. D{\bf 59} (1999) 036003;\\

1659: R. Arnold et al., Eur. Phys. J. A {\bf 6}  (1999) 361;\\

1660: Borexino collaboration, H.O.Back, et al, hep-ph/0406252,

1661: European Physical Journal C {\bf 37} (2004) 421.

1662: 

1663: %5

1664: \bi{gri} L. Cucurull, J.A. Grifols, R. Toldra, 

1665: %SPIN STATISTICS THEOREM, NEUTRINOS, AND BIG BANG NUCLEOSYNTHESIS.

1666: Astropart. Phys. {\bf 4}, (1996) 391.

1667: 

1668: %6

1669: \bi{dosm}

1670: A.D. Dolgov, A.Yu. Smirnov, Phys. Lett. B {\bf 621} (2005) 1.

1671: 

1672: %7

1673: \bi{hansen}

1674: A.D. Dolgov, S.H. Hansen, A.Yu. Smirnov, JCAP {\bf 0506} (2005) 004. 

1675: 

1676: %8

1677: \bi{kar}

1678: S. Choubey, K. Kar, Phys. Lett. B {\bf 634} (2006) 14.

1679: 

1680: %9

1681: \bi{para}

1682: H.S. Green, Phys. Rev. {\bf 90} (1953) 270;\\

1683: O.W. Greenberg, Phys. Rev. Lett. {\bf 13} (1964) 598;\\

1684: O.W. Greenberg, A.M.L. Messiah, Phys. Rev. {\bf 138} (1965) B1155;\\

1685: for a recent discussion see O.W. Greenberg, A.K. Mishra, math-ph/0406011.

1686: 

1687: 

1688: %10

1689: \bi{ign-kuzm}

1690: A.Yu. Ignatiev, V.A. Kuzmin,  hep-ph/0510209.

1691: 

1692: %11

1693: \bibitem{boehm}

1694: W. C. Haxton, G. J. Stephenson Jr., 

1695: Prog. in Part. and Nucl. Phys., {\bf 12} (1984) 409;\\

1696: M. Doi et al., Prog. of Theor. Phys. Suppl., {\bf 83} (1985);\\

1697: J. D. Vergados Phys. Rep.  {\bf 133} (1986) 1;\\

1698: F. Boehm,  P. Vogel, {\it Physics of massive neutrinos}, Cambridge

1699: Univ. Press. 1987;\\

1700: A. Faessler, F. \v Simkovic, J. Phys. G {\bf 24} (1998) 2139.

1701: 

1702: \bibitem{ABA84}

1703: J. Abad, A. Morales, R. Nunez-Lagos and A. Pacheo, 

1704: Ann. Fis. {\bf 80} (1984) 9;\\

1705: A. Griffiths, P. Vogel, Phys. Rev. C {\bf 46} (1992) 181.

1706: 

1707: %20

1708: %

1709: \bi{SDS}

1710: %[2] THE SINGLE STATE DOMINANCE HYPOTHESIS AND THE TWO NEUTRINO DOUBLE

1711: %    BETA DECAY OF MO-100.

1712: F. \v Simkovic, P. Domin, S.V. Semenov, J. Phys. G {\bf 27} (2001) 2233. 

1713: 

1714: 

1715: %21

1716: \bi{DKSS} 

1717: %NEUTRINO ACCOMPANIED BETA+-BETA+-, BETA+/EC AND EC/EC PROCESSES 

1718: %WITHIN   SINGLE STATE DOMINANCE HYPOTHESIS.

1719: 

1720: P. Domin, S. Kovalenko, F. \v Simkovic, S.V. Semenov, 

1721: Nucl. Phys. A {\bf 753} (2005) 337.

1722: 

1723: \bibitem{behr} H. Behrens and W. B\"uhring, ``Electron radial 

1724: wave functions and nuclear $\beta$-decay'', Clarendon press - Oxford,

1725: 1982, p. 230.

1726: 

1727: %14

1728: \bibitem{nemo}

1729: The NEMO Collaboration (R. Arnold et al.)  

1730: Phys. Rev. Lett. {\bf 95} (2005) 182302.

1731: 

1732: 

1733: %15

1734: \bi{klapdor}

1735: H.V. Klapdor-Kleingrothaus, {\it et al}, Eur. Phys. J. A {\bf 12} (2001) 

1736: 147;

1737: H.V. Klapdor-Kleingrothaus, A. Dietz, I.V. Krivosheina, O. Chkvorets,  

1738: Nucl. Instrum. Meth. A {\bf 522} (2004)  371.

1739: %DATA ACQUISITION AND ANALYSIS OF THE GE-76 DOUBLE BETA EXPERIMENT IN 

1740: %GRAN SASSO 1990-2003.

1741: 

1742: %16

1743: \bibitem{ARN04}

1744: R. Arnold et al. JETP Lett. {\bf 80} (2004) 377.

1745: 

1746: %17

1747: \bibitem{SHI06}

1748: The NEMO Collaboration (Yu. Shitov et al.)  

1749: Phys. At. Nucl. {\bf 69} (2006) 2090.

1750: 

1751: %12

1752: \bibitem{BAR95}

1753: A.S. Barabash et al., Phys. Lett. B {\bf 345} (1995) 408.

1754: 

1755: %13

1756: \bibitem{ARN07}

1757: R. Arnold et al., Nucl. Phys. A {\bf 781} (2007) 209.

1758: 

1759: %18

1760: \bibitem{SUH98}

1761: J. Suhonen and O. Civitarese, Phys. Rep. {\bf 300} (1998) 123.

1762: 

1763: \bibitem{NSM}

1764: E.Caurier, F. Nowacki and A. Poves, Int. J. Mod. Phys. E {\bf 16} (2007) 552.

1765: %E. Caurier, G. Martinez-Pinedo, F. Nowacki, A. Poves, and A.P. Zuker, 

1766: %Rev. Mod. Phys. {\bf 77} (2005) 427.

1767: 

1768: 

1769: %19

1770: \bibitem{BAR06}

1771: A.S. Barabash, Czech. J. Phys. {\bf 56} (2006) 437.

1772: 

1773: 

1774: \bibitem{GAR93}

1775: A. Garcia et al., Phys. Rev. C {\bf 47} (1993) 2910.

1776: 

1777: %23

1778: \bibitem{SAS07}

1779: M. Sasano et al., Nucl. Phys. A (2007) (to be published).

1780: 

1781: 

1782: % 22

1783: \bibitem{BHA98}

1784: M. Bhattacharya et al., Phys. Rev. C {\bf 58} (1998) 1247.

1785: 

1786: 

1787: \bibitem{baranew} A.S. Barabash, hep-ex/0610025.

1788: 

1789: %

1790: %\bi{barabash}

1791: %A.S. Barabash, Czech. J. Phys. {\bf 52} (2002) 567;

1792: %Czech. J. Phys. {\bf 56} (2006) 437.

1793: 

1794: % 

1795: %\bibitem{ZAR04}

1796: %X. Sarazin (NEMO Collaboration), hep-ex/0412012; talk at NEUTRINO'04.

1797: 

1798: % 

1799: %\bibitem{DAN03}

1800: %F.A. Danevich et al., Phys. Rev. C {\bf 68} (2003) 035501.

1801: 

1802: %\bi{trit} ....

1803: 

1804: \end{thebibliography}

1805: 

1806: 

1807: \end{document}

1808: 

1809: 

1810: 

1811: