Available online at www.sciencedirect.com

SCIENCE ^DIRECT8

Physics Letters B 558 (2003) 141-156

www. elsevier. com/locate/npe

On the neutrino mass spectrum and neutrinoless double-beta decay

S. Pascolia, S.T. Petcovbc1, W. Rodejohannb

a Department of Physics, University of California, Los Angeles, CA 90095-1547, USA b Scuola Internazionale Superiore di Studi Avanzati, I-34014 Trieste, Italy c Istituto Nazionale di Fisica Nucleare, Sezione di Trieste, I-34014 Trieste, Italy

Received 6 January 2003; accepted 20 February 2003

Editor: G.F. Giudice

Abstract

Assuming 3-v mixing, neutrino oscillation explanation of the solar and atmospheric neutrino data and of the first KamLAND results, massive Majorana neutrinos and neutrinoless double-beta ((PP)0v-) decay generated only by the (V-A) charged current weak interaction via the exchange of the three Majorana neutrinos, we analyze in detail the possibility of determining the type of the neutrino mass spectrum by measuring of the effective Majorana mass |(m)| in (Pfi)0v-decay. The three possible types of neutrino mass spectrum are considered: (i) normal hierarchical (NH) m\ ^ m2 ^ m3, (ii) inverted hierarchical (IH), m\ ^ m2 = m3, and (iii) quasi-degenerate (QD), m\ = m2 = m3, m1,2,3 ^ 0.20 eV. The uncertainty in the measured value of |(m)| due to the imprecise knowledge of the relevant nuclear matrix elements is taken into account in the analysis. We derive the ranges of values of tan2 0q , 0q being the mixing angle which controls the solar neutrino oscillations, and of the nuclear matrix element uncertainty factor, for which the measurement of |(m)| would allow one to discriminate between the NH and IH, NH and QD, and IH and QD spectra. © 2003 Elsevier Science B.V. All rights reserved.

1. Introduction

The solar neutrino experiments Homestake, Kamio-kande, SAGE, GALLEX/GNO, super-Kamiokande (SK) and SNO [1-4], the data on atmospheric neutrinos obtained by the super-Kamiokande (SK) experiment [5] and the results from the KamLAND reactor anti-neutrino experiment [6], provide very strong evidences for oscillations of flavour neutrinos. The ev-

E-mail address: petcov@he.sissa.it (S.T. Petcov).

1 Also at: Institute of Nuclear Research and Nuclear Energy, Bulgarian Academy of Sciences, 1784 Sofia, Bulgaria.

idences for solar v^ oscillations into active neutrinos , in particular, were spectacularly reinforced by the first data from the SNO experiment [3] when combined with the data from the SK experiment [2], by the more recent SNO data [4], and by the just published first results of the KamLAND [6] experiment. Under the rather plausible assumption of CPT-invariance, the KamLAND data practically establishes [6] the large mixing angle (LMA) MSW solution as unique solution of the solar neutrino problem. This remarkable result brings us, after more than 30 years of research, initiated by the pioneer works of B. Pontecorvo [7] and the experiment of R. Davis et al. [8], very close to a

0370-2693/03/$ - see front matter © 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0370-2693(03)00275-2

complete understanding of the true cause of the solar neutrino problem.

The interpretation of the solar and atmospheric neutrino, and of the KamLAND data in terms of neutrino oscillations requires the existence of 3-neutrino mixing in the weak charged lepton current (see, e.g., [9, 10]):

= J2 Uijvj l.

Here vlL, l = e, ¡x,x, are the three left-handed flavor neutrino fields, VjL is the left-handed field of the neutrino Vj having a mass mj and U is the Pontecorvo-Maki-Nakagawa-Sakata (PMNS) neutrino mixing matrix [11]. If the neutrinos with definite mass Vj are Majorana particles, the process of neutrinoless doublebeta ((PP)0v-) decay will be allowed (for reviews see, e.g., [12,13]). If the (PP)0v-decay is generated only by the (V-A) charged current weak interaction via the exchange of the three Majorana neutrinos Vj and the latter have masses not exceeding few MeV, the dependence of the (PP)0v -decay amplitude on the neutrino mass and mixing parameters factorizes in the effective Majorana mass |(m)| (see, e.g., [12]):

|(m)| = |m1|Ue1|2 + m2|Ue2|Va21 + m3|Ue3|Va311,

where a21 and a31 are two Majorana CP-violating phases2 [14,15]. If CP-invariance holds, one has [16]

«21 = kn, «31 = k'n, where k, k' = 0, 1,2,____In this

n2i = e

= eia21 = ±1

nsi = e1"31 =±1,

represent the relative CP-parities of the neutrinos v1 and v2, and v1 and v3, respectively. The oscillations between flavour neutrino are insensitive to the Majorana CP-violating phases a21, a31 [14,17]— information about these phases can be obtained in the (PP)ov-decay experiments [18-23] (see also [24]). Majorana CP-violating phases, and in particular, the phases a21 and/or a31, might be at the origin of the baryon asymmetry of the Universe [25].

2 We assume that mj > 0 and that the fields of the Majorana ie I

1,2,3, where C is the charge conjugation matrix.

One can express [26] (see also, e.g., [10,19,27]) the masses m2 3 and the elements of the lepton mixing matrix entering into Eq. (2) for |(m)|, in terms of the neutrino oscillation parameters measured in the solar and atmospheric neutrino and KamLAND experiments: m2 3—in terms of the neutrino mass squared differences Am© and Am A, driving the solar and atmospheric neutrino oscillations, and the mass m1, and |Uej |2, j = 1, 2, 3,—in terms of the mixing angle which controls the solar ve transitions 0©, and of the lepton mixing parameter sin2 0 limited by the data from the CHOOZ and Palo Verde experiments [28,29].

The observation of (Pfi)0v-decay will have fundamental implications for our understanding of the symmetries of the elementary particle interactions3 (see, e.g., [12]). Under the general and plausible assumptions of 3-v mixing, neutrino oscillation explanation of the solar and atmospheric neutrino data, massive Majorana neutrinos and (Pfi)0v-decay generated only by the (V-A) charged current weak interaction via the exchange of the three Majorana neutrinos, which will be assumed to hold throughout this study, the observation of (Pfi)0v-decay can provide unique information on [19,21,22,27,32,33]

(i) the type of neutrino mass spectrum which can be normal hierarchical (NH), inverted hierarchical (IH), or quasi-degenerate (QD),

(ii) on the absolute scale of neutrino masses, and [1823],

(iii) on the Majorana CP-violating phases a21 and

A measured value of |(m)| ~ few x 10-2 eV can provide, in particular, unique constraints on, or even can allow one to determine, the type of the neutrino mass spectrum in the case v12 3 are Majorana particles [33]. The solar neutrino data and the first KamLAND results,4 favor relatively large value of cos 20©, cos20© ~ 0.40 [34-38]. A value of cos20© > 0.25 would imply [33] the existence of significant lower

neutrinos vj satisfy the Majorana condition: C(vj)T = vj, j =

3 Evidences for {PP)qv-decay taking place with a rate corresponding to 0.11 eV < |(m)| < 0.56 eV (95% C.L.) are claimed to have been obtained in [30]. The results announced in [30] have been criticized in [31].

4 We assume throughout this study that CPT-invariance holds in the lepton sector.

bounds on |(m)| (exceeding 0.01 eV) in the cases of IH and QD neutrino mass spectrum, and of a stringent upper bound (smaller than 0.01 eV) if the spectrum is of the NH type. The indicated lower bounds are in the range of the sensitivity of currently operating and planned (PP)0v-decay experiments.

Information on the absolute values of neutrino masses in the range of interest can also be derived in the 3H p -decay neutrino mass experiment KATRIN [39], and from cosmological and astrophysical data (see, e.g., Ref. [40]).

Rather stringent upper bounds on |(m)| have been obtained in the 76 Ge experiments by the Heidelberg-Moscow Collaboration [41], |(m)| < 0.35 eV (90% C.L.), and by the IGEX Collaboration [42], |(m)| < (0.33-1.35) eV (90% C.L.). Taking into account a factor of 3 uncertainty in the calculated value of the corresponding nuclear matrix element [13], we get for the upper limit found in [41]: |(m)| < 1.05 eV. Higher sensitivity to |(m)| is planned to be reached in several (PP)0v-decay experiments of a new generation. The NEMO3 experiment [43], which began to take data in July of 2002, and the cryogenics detector CUORICINO [44] to be operative in 2003, are expected to reach a sensitivity to values of |(m)| ~ 0.2 eV. Up to an order of magnitude better sensitivity, i.e., to |(m)| = 2.7 x 10-2 eV, 1.5 x 10-2 eV, 5.0 x 10-2 eV, 2.5 x 10-2 eV and 3.6 x 10-2 eV is planned to be achieved in the CUORE, GENIUS, EXO, MAJORANA, and MOON experiments [44],5 respectively.

In what regards the 3H p -decay experiments, the currently existing most stringent upper bounds on the electron (anti-)neutrino mass mye were obtained in the Troitzk [46] and Mainz [47] experiments and read < 2.2 eV. The KATRIN 3H p-decay experiment [39] is planned to reach a sensitivity to mye ~ 0.35 eV.

In the present article we study in detail the possibility of determining the type, or excluding one or more types, of neutrino mass spectrum by measuring of |(m)| in the next generation of (PP)0v-decay experiments. The three possible types of spectra are consid-ered:6

5 The quoted sensitivities correspond to values of the relevant nuclear matrix elements from Ref. [45].

6 We work with the convention «1 < m2 < m3 and use the

term "spectrum with normal (inverted) hierarchy" for the spectra

(i) hierarchical (NH) m1 ^ m2 ^ m3,

(ii) inverted hierarchical (IH), m1 ^ m2 = m3, and

(iii) quasi-degenerate (QD), m1 = m2 = m3 = mo,

m?,2,3 ^ Am2A.

In our analysis we take into account, in particular, the uncertainty in the determination of |(m)| due to the imprecise knowledge of the relevant nuclear matrix elements. This permits us to determine the requirements which the possibility of distinguishing between (i) the NH and IH, (ii) the NH and QD, and (iii) the IH and QD spectra, imposes on the uncertainty in the values of the (PP)0v -decay nuclear matrix elements. We derive also the maximal values of tan2 0© for which the measurement of |(m)| would allow one to discriminate between the NH and IH, NH and QD and IH and QD spectra, for different given values of the nuclear matrix element uncertainty factor. An upper limit |(m)| < few x 10-2 eV would imply a significant constraint on the type of neutrino mass spectrum in the case the massive neutrinos are Majorana particles, e.g., on the theories in which the neutrino masses are generated via the see-saw mechanism.

It should be noted that the determination of the type of neutrino mass spectrum, based on the measured value of |(m)|, would provide simultaneously unique information on the absolute neutrino mass scale [21, 22,32,33]. Similar information cannot be obtained in the neutrino oscillation experiments, in which the sign of AmA can be determined7 (see, e.g., [48, 49]) since neutrino oscillations depend on neutrino mass squared differences and are insensitive to the absolute neutrino mass scale. The sign of Am2A can be determined in very long base-line neutrino oscillation experiments at neutrino factories (see, e.g., [48]), and, e.g., using combined data from long base-line oscillation experiments at the JHF facility and with off-axis neutrino beams [49]. Under certain rather

with A«q = A«21 (Afflg = Am/332), while we call "normal hierarchical (NH)" ("inverted hierarchical (IH)") the neutrino mass spectrum with normal (inverted) hierarchy and «1 ^ «2,«3.

7 In the convention in which the sign of Am^ = Am^1 is not fixed, the latter determines the ordering of the neutrino masses: A«A > 0 corresponds to «1 < «2 < «3, while Am^ < 0 implies

«3 < «1 < «2.

special conditions it might be determined also in experiments with reactor Ve [50].

2. Neutrino oscillation data and the effective Majorana mass

The predicted value of |(m)| depends in the case of 3-v mixing on:

(i) AmA,

(ii) 00 and Amg,

(iii) the lightest neutrino mass, and on

(iv) the mixing angle 0.

Using the convention mi <m2 <m3, one has Am A = Am^j, where Amjk = m.2 — m^, and m3 = (m2 + Am2A)l/2, while either Am@ = Am21 or Am@ = Am22. The two possibilities for Am0 correspond, respectively, to the two different types of neutrino mass spectrum—with normal and with inverted hierarchy. In the first case we have m2 = (m1 + Am0)1/2, |^e1|2 = cos2 00 (1 — |Ue3|2), |Ue2|2 = sin2 0;(1 — | Ue312), and | Ue312 = sin2 0, while in the second m2 = (m2 + AmA — Am0)1/2, |Ue2|2 = cos20;(1 — |Ue1|2), | Ue3 I2 = sin2 00(1 — I Ue112), and | Ue112 = sin2 0.

Given Am0, Am2A, 00 and sin2 0, the value of |(m)| depends strongly on the type of the neutrino mass spectrum, as well as on the values of the two Majorana CP-violating phases, a21 and a31 (see Eq. (2)), present in the lepton mixing matrix. Let us note that in the case of QD spectrum, m1 = m2 = m3, m1 2 3 > AmA, Am0, |(m)| is essentially independent on Am2A and Am0, and the two possibilities, Am0 = Am21 and Am0 = Am^2, lead effectively to the same predictions for |(m)|.8

The possibility of determining the type of the neutrino mass spectrum if |(m)| is found to be non-zero in the (PP)0v -decay experiments of the next generation, depends crucially on the precision with which Am2A,

8 This statement is valid, within the convention mi < m2 < m3 we are using, as long as there are no independent constraints on the CP-violating phases «21 and a3i which enter into the expression for |(m)|. In the case ofNH spectrum, |(m)| depends primarily on a2i (|Ue312 ^ 1), while if the spectrum is with IH, |(m)| will depend essentially on a3i — a2i (|Uei |2 ^ 1).

00, Amg, sin2 0, and |(m)| will be measured. It depends also crucially on the values of 00 and of |(m)|. The precision itself of the measurement of |(m)| in the next generation of (PP)0v -decay experiments, given the latter sensitivity limits of ~ (1.5-5.0) x 10—2 eV, depends on the value of |(m)|.

The KATRIN experiment [39] can test the hypothesis of a QD spectrum,9 provided m12 3 = mVe > (0.35-0.40) eV. The KATRIN detector is designed to have a 1 s.d. error of 0.08 eV2 on a measured value of m2e. This experiment is expected to start in 2007. '

Assuming CPT-invariance, combined ve ^ vi(t) and ve ^ vIJ((t) oscillation analyzes of the solar neutrino data and of the just published first KamLAND results [6], have already been performed in [34-38]. All analyzes show that the data favor the LMA MSW solution with Am0 > 0 and tan2 00 < 1, all the other solutions (LOW, VO, etc.) being essentially ruled out. In Tables 1 and 2 we give the best-fit values and the 90% C.L. allowed ranges of Am0 and tan2 00 in the LMA solution region obtained in [34-37]. The best fit values are confined to the narrow intervals (Am0)BF = (6.9-7.3) x 10—5 eV2, (tan2 00)BF = (0.42-0.46). The latter corresponds to (cos200)BF = (0.37-0.41).

In the two-neutrino vu

Vt (VI

vt ) oscilla-

tion analysis of the SK atmospheric neutrino data performed in [5] the following best-fit value of Am2A was obtained: (Am2A)BF = 2.5 x 10—3 eV2. At 99.73% C.L., AmA was found to lie in the interval: (1.5-5.0) x10—3 eV2. According to the more recent combined analysis of the data from the SK and K2K experiments [51], one has Am\ = (2.7 ± 0.4) x 10—3 eV2. In certain cases of our analysis we will use as illustrative "best-fit" values (Am2)BF = 7.0 x 10—5 eV2 and

(AmA)BF = 3.0 x 10—3 eV2.

For the indicated allowed ranges of values of Am0 and AmA, the NH (IH) spectrum corresponds to m1 < 10—3 (2 x 10—2) eV

A3- v oscillation analysis of the CHOOZ data showed [52] that for Am0 < 10—4 eV2, the limits on sin2 0 practically coincide with those derived in the 2- v oscillation analysis in [28]. Combined 3- v

9 Given the allowed regions of values of Am0 and AmA [5], one has a QD spectrum for m1,2,3 = mye > 0.20 eV.

oscillation analysis of the solar neutrino, CHOOZ and the KamLAND data was performed in [34] under the assumption of Am© ^ AmA (see, e.g., [9,10,53]). For the best-fit value of sin2 0 the authors of [34] obtained: (sin2 0)bf = (0.00-0.01). It was also found in [34] that sin2 0 < 0.05 at 99.73% C.L.

The existing solar neutrino and KamLAND data favor values of Am© = (5.0-10.0) x 10-5 eV2 [3438]. If Am© lies in this interval, a combined analysis of the future more precise KamLAND results and of the solar neutrino data would permit to determine the values of Am© and tan2 0© with high precision: the estimated (1 s.d.) errors on Am© and on tan2 0© can be as small as ~ (3-5)% and ~ 5% (see, e.g., [51, 54]).

Similarly, if Am2A lies in the interval Am2A = (2.0-5.0) x 10-3 eV2, as is suggested by the current atmospheric neutrino data [5,51], its value will be determined with a ~ 10% error (1 s.d.) by the MINOS experiment [55]. Somewhat better limits on sin2 0 than the existing one can be obtained in the MINOS experiment [55] as well. Various options are

being currently discussed (experiments with off-axis neutrino beams, more precise reactor antineutrino and long base-line experiments, etc., see, e.g., [56]) of how to improve by at least an order of magnitude, i.e., to values of ~ 0.005 or smaller, the sensitivity to sin2 0.

The high precision measurements of Am2A, tan2 0© and Am© are expected to take place within the next ~ (6-7) years. We will assume in what follows that the problem of measuring or tightly constraining sin2 0 will also be resolved within the indicated period. Under these conditions, the largest uncertainty in the comparison of the theoretically predicted value of |(m)| with that determined in the (Pfi)ov-decay experiments would be associated with the corresponding (PP)ov-decay nuclear matrix elements. We will also assume in what follows that by the time one or more (PP)ov-decay experiments of the next generation will be operative (2009-2010) at least the physical range of variation of the values of the relevant (Pfi)ov -decay nuclear matrix elements will be unambiguously determined.

Table 1

The best-fit values of tan2 0© and Am© (in units of 10-5 eV2) in the LMA solution region, as reported by different authors. Given are also the calculated maximal values of |(m)| (in units of 10-3 eV) for the NH spectrum and the minimal values of |(m)| (in units of 10-3 eV) for the IH and QD spectra. The results for |(m)| in the cases of NH and IH spectra are obtained for m\ = 10-3 eV and the best-fit value of AmA, AmA = 2.7 x 10-3 eV2 [51], while those for the QD spectrum are derived for m0 = 0.2 eV. In all cases sin2 0 = 0.05 has been used. For |(m)|Sax we included in brackets also the values for sin2 0 = 0.01. The chosen value of AmA corresponds to |(m)|mHx = 52.0 x 10-3 eV

Reference_(tan290)BF_(A«|)bf_IWlS_l(«>Cm_

[34] 0.46 7.3 5.9(3.9) 18.4 59.9

[35] 0.46 6.9 5.8(3.8) 18.4 59.9

[36] 0.42 7.2 5.7(3.7) 20.3 67.2

[37] 0.43 7.0 5.7(3.7) 19.8 65.3

Table 2

The ranges of allowed values of tan2 0© and Am© (in units of 10-5 eV2) in the LMA solution region, obtained at 90% C.L. by different authors. Given are also the corresponding maximal values of |(m)| (in units of 10-3 eV) for the NH spectrum, and the minimal values of |(m)| (in units of 10-3 eV) for the IH and QD spectra. The results for the NH and IH spectra are obtained for m\ = 10-3 eV, while those for the QD spectrum correspond to m0 = 0.2 eV. Ami was assumed to lie in the interval [51] (2.3-3.1) x10-3 eV2. This implies |(m)|®x = 55.7 x 10-3 eV. As in Table 1, in all cases sin2 0 = 0.05 has been used. For Km)!^®; we included in brackets also the values for sin2 0 = 0.01

Reference_tan290_Am|_IWlS_l(«>Cm_I Wig"

[34] 0.32-0.72 5.6-17 8.6(6.6) 7.6 20.6

[35] 0.31-0.68 5.7-15 8.1 (6.1) 8.9 25.8

[36] 0.31-0.56 6.0-8.7 6.6(4.5) 13.0 43.2

[37] 0.31-0.66 5.9-8.9 7.0(4.9) 9.5 28.6

3. Determining the type of neutrino mass spectrum

The possibility to distinguish between the three different types of neutrino mass spectrum in the 3-neutrino mixing case under discussion depends on the allowed ranges of values of |(m)| for the three spectra. More specifically, it is determined by the maximal values of |(m)| in the cases of NH and IH spectra and by the minimal values of |(m)| for the IH and QD spectra. For the NH neutrino mass spectrum (m1 ^ m2 ^ m3), the maximal value of |(m)| is obtained in the case of CP-conservation and equal CP-parities of v123:

I , , |NH ^ 1 — S2

1 + tan2 eQ

x mi + tan OqJ AmQ

where s2 = sin2 0 and we have neglected m21 with respect to Am0 and Am2A.

In the case of IH neutrino mass spectrum (m1 ^ m2 = m3), the effective Majorana mass lies in the interval [18,19]

< 1 <m

I = (! - cos20oyz^ï,

I min | IH

= (1 -,

where we have neglected m1. The minimal (maximal) value of |(m)|, |(m)|JHn corresponds to CP-conservation and opposite (equal) CP-parities of the neutrinos 2 and 3.

The minimal value of |(m)| for the quasi-degenerate (QD) neutrino mass spectrum (m1 = m2 = m3 = mo, m0 > Am0,Am2A), for fixed value of m0 is given by

1/ nIQD ~ l~g2 \{m)\ . =-^—

1 min 1 + tan2 00

x I 1 - tan2 eG

(1 + tan2Og) mo, (7)

where m0 > 0.20 eV and we have neglected Am20 and Am A with respect to m0. As Eq. (7) shows, |(m)||mn scales to a good approximation with m0. Correspondingly, the minimal allowed value of |(m)| for the QD mass spectrum is obtained for m0 = 0.2eV.

In Tables 1 and 2 we show the calculated (i) maximal predicted value of |(m)| in the case of NH neutrino mass spectrum, (ii) the minimal value of |(m)| for the IH spectrum, and (iii) the minimal value of |(m)| for the QD spectrum (m0 = 0.2 eV), for the best-fit and the 90% C.L. allowed ranges of values of tan2 00 and Am0 in the LMA solution region. In Table 3 we give the same quantities, |(m)|mx, |(m)|JHn and |(m)||mn, calculated using the best-fit values of the neutrino oscillation parameters, including 1 s.d. (3 s.d.) uncertainties of 5% (15%) on tan2 00 and Am0 and of 10% (30%) on Am2A.

The maximal predicted value of (in ) for the IH

spectrum is given by |(»z)|£L = Am2)max. For the best-fit value [5,51] and the 99.73% C.L. allowed range [5] of Am2A we have, respectively, |(m)|mx = 0.05 and 0.07 eV.

On the basis of the results shown in Tables 1-3, we can conclude, in particular, that the NH spectrum could be ruled out if the measured value of |(m)|

Table 3

The values of Km)!^®;, |(m)|JHn and Km^^ (in units of 10 3 eV), calculated using the best-fit values of solar and atmospheric neutrino oscillation parameters from Table 1 and including 1 s.d. (3 s.d.) uncertainties of 5% (15%) on tan2 00 and Am0, and of 10% (30%) on Ami .

In this case one has:

I<m)lmax

= 54.5 (59.2) x 10-3 eV

Reference |{m)|NH (Î2=0.05) |{m)|NH (i2 = 0.01) IM Cm

[34] 6.1 (6.7) 4.1 (4.4) 16.5 (12.9) 55.9 (48.2)

[35] 6.1 (6.6) 4.0 (4.3) 16.5 (12.9) 55.9 (48.2)

[36] 6.0(6.5) 3.9 (4.2) 18.3 (14.6) 63.3 (55.9)

[37] 6.0(6.5) 3.9 (4.2) 17.9(14.1) 61.4(54.0)

exceeds approximately 0.9 x 10-2 eV, where we have been rather conservative in choosing the maximal value.

3.1. Theoretical and experimental uncertainties in |(m)|

Following the notation in Ref. [23], we will parametrize the uncertainty in |(m)| due to the imprecise knowledge of the relevant nuclear matrix elements— we will use the term "theoretical uncertainty" for the latter—through a parameter Z, Z ^ 1, defined as:

= í( (MLp)min ± ^

where (|(m)|exp )MIN is the value of |(m)| obtained from the measured (PP)0v -decay half life-time of a given nucleus using the largest nuclear matrix element and A is the experimental error. An experiment measuring a (PP)0v-decay half-life time will thus determine a range of |(m)| corresponding to

3.2. Requirements on the solar neutrino mixing angle

We shall derive next the constraints tan2 0© must satisfy in order to be possible to distinguish between the three types of neutrino mass spectrum NH, IH and QD.

Case (i). Normal hierarchical and inverted hierarchical spectra In order to be possible to distinguish between the NH and IH spectra, the following inequality must hold:

ZMf® <|(m)|IIi, Z > 1. (11)

'I 1 ' I max I 1 ' Imin' ' ^ v '

From this inequality, using Eqs. (4) and (6), we get:

2 WO3 + t2) tan 0G <-—,

1 + ¡;(a + t2Y

(MLp)min - A ^ Ml ^ ?((l<m)lexp)min + zi)- where t2 = s2/{ 1 - i2), a = JAm2Q/Am2A and /3 =

y»i2/Am2A. For our illustrative "best-fit" values

The currently estimated range of Z for experimentally interesting nuclei varies from 3.5 for 48Ca to 38.7 for 130Te, see, e.g., Table 2 in Ref. [13] and Ref. [57].

We estimate, following again [23], the 1 s.d. error on the experimentally measured value of |(m)| by using the standard expression

= V(£i)2 + (£2)2,

a(\(m)\)

where E1 and E2 are the statistical and systematic errors. We choose E2 = const = 0.05 and take E1 = //|(m)|, where we assume f = 0.028 eV. This gives a total relative error a(|(m)|)/|(m)| = 15% at |(m)| = 0.20 eV. The above choices were motivated by the fact that the sensitivities of the next generation of (PP)0v-decay experiments in the measurement of |(m)| are estimated to be in the range of ~ (1.5-5.0) x 10-2 eV and if, e.g, |(m)| > 0.20 eV, a precision in the determination of |(m)| corresponding to an error of ~ 15% could be reached in these experiments. Moreover, for values of |(m)| which are sufficiently bigger than the quoted sensitivity limits of the future experiments, the statistical error scales as |(m)| increases like E1 ~ const/|(m)|.

(Am©)bf = 7.0 x 10-5 eV2 and (Am2A)BF = 3.0 x 10-3 eV2, one has a ~ 0.153; with m1 < 0.001 eV one also finds p < 0.018. For Z = 1, the indicated values of a and p and s2 = 0.05 (0), Eq. (12) is fulfilled for tan20© < 0.77 (0.85). Taking Z = 2, one finds tan2 0© < 0.61 (0.74), while for Z = 3 the result is tan2 0© < 0.49 (0.65).

The smaller m1 and/or Am©, the closer the upper bound on tan2 0© of interest becomes to 1. The above analysis shows also that the upper bound on tan2 0© under discussion exhibits relatively strong dependence on the value of s2 < 0.05: it increases by a factor of ~ (1.2-1.5) when s2 decreases from 0.05 to 0.

Case (ii). Normal hierarchical and quasi-degenerate spectra Distinguishing between the NH and QD spectra requires that the following inequality is satisfied:

Z |m>r < |<m>imD, z > 1.

From this inequality using Eqs. (4) and (7) we get:

2 1 — Z P — t (1 + Zy)

tan 0© < -----—,

l + Ça + t2(l + Çy)'

where a = yAm©/m0, P = m1/m0 and y =

JAni2^/ni^. For our illustrative "best-fit" values of Am© and Am2A and m0 > 0.2 eV, we have: a < 0.042, P < 0.000025, and y < 0.274. Using these upper limits we find that for s2 = 0.05 (0) and Z = 1, Eq. (14) is satisfied if tan20© < 0.9 (1.0). Taking Z = 2, 3 we get tan2 0© < 0.8 (1.0) for s2 = 0.05 (0). If, e.g., m0 = 2.0 eV, one finds y < 0.035, i.e., the larger the value of mo, the smaller y and the closer is the upper bound on tan2 0© to 1, i.e., the less constraining it is. Since y enters into Eq. (14) multiplied by the relatively small quantity t2, the deviation of the upper bound on tan2 0© under discussion from 1 is determined essentially by the value of a. Correspondingly, the maximal value of tan2 0© permitting to distinguish between the NH and QD neutrino mass spectra decreases with increasing of Am©.

Case (iii). Inverted hierarchical and quasi-degenerate spectrum One could distinguish between these two types of spectra if the following inequality is fulfilled:

Z > 1.

This condition together with Eqs. (6) and (7) leads to the constraint

tan2 0© <

1 - Zi>- t2

0 ^ l + ZY + t2' where y was defined earlier. Using our illustrative Am© and Am\ "best-fit" values and m0 > 0.2 eV, one finds Y < 0.274. For s2 = 0.05 (0) and Z = 1, the above limit on y together with Eq. (16) leads to tan2 0© < 0.5 (0.6). Larger values of Z lead to stringent restrictions on tan2 0©: for Z = 2, for instance, we find tan20© < 0.2 (0.3) for s2 = 0.05 (0). The requirement that the two spectra could be distinguished is less restrictive for larger values of m0 in this case as well.

These simple quantitative analyses show that if |(m)| is found to be non-zero in the future (PP)0v-decay experiments, it would be easier, in general, to distinguish between the spectrum with NH and those with IH or of QD type using the data on |(m)| = 0, than to distinguish between the IH and the

QD spectra. Discriminating between the latter would be less demanding if m0 is sufficiently large. The requirement of distinguishing between the NH and the QD spectra leads to the least stringent conditions.

The above analyses also show that the possibility to distinguish between the IH and QD, and NH and QD, spectra depends rather weakly on the value s2, satisfying the existing upper limits [28,29,51]: the relevant upper bounds on tan2 0© decrease somewhat with decreasing of s2. This is not so in the case of NH versus IH spectra: the upper bound of interest can increase noticeably (e.g., by a factor of ~ (1.2-1.5)) when s2 decreases from 0.05 to 0.

It is worth noting that in contrast to the conditions which would allow one to establish on the basis of a measurement of |(m)| = 0 the presence of CP violation due to the Majorana CP-violating phases [23], the conditions permitting to distinguish between the three types of neutrino mass spectrum imply an upper limit on tan2 0©.

In Fig. 1 we show the upper bounds on tan2 0©, for which one can distinguish the NH spectrum from the IH spectrum and from that of QD type, as a function of Am© for different values of Z. As is seen from the figure, the dependence of the maximal value of tan2 0© of interest on m0 in both cases is modest. Obviously, with the increasing of Am© and/or s2, l(m)C also increases. As a consequence, the maximal tan2 0© under discussion decreases, which means that the corresponding spectra become harder to distinguish.

As we have seen, in order to be possible to distinguish between the IH and the QD spectra Eq. (16) should be fulfilled. Fig. 2 shows the upper bound on tan2 0© as implied by Eq. (16), for s2 = 0.05 and 0.0 as a function of Am A. The upper bound on tan2 0© of interest depends strongly on the value of m0. It decreases with the increasing of Am2A, the dependence on Am A being noticeable for m0 = 0.20 eV and rather mild for m0 > 0.40 eV. As it follows from Fig. 2, for the values of Am2A favored by the neutrino oscillation data and for Z > 2, distinguishing between the IH and QD spectra in the case of m0 = 0.20 eV requires too small, from the point of view of the existing data, values of tan2 0©. For m0 > 0.40 eV, the values of tan2 0© of interest fall in the ranges favored by the existing solar neutrino and KamLAND data even for Z = 3.

Fig. 1. The upper bound on tan2 0;, for which one can distinguish the NH spectrum from the IH spectrum and from that of QD type, as a

function of Am0 for AmA = 3 x 10 3 eV2 and different values of Ç (see Eqs. (12) and (14)). The lower (upper) line corresponds to s2 = 0.05 (0). For NH vs. IH there is a third (middle) line corresponding to s2 = 0.01.

3.3. Requirements on A and Z

We will investigate now the requirements the experimental and theoretical uncertainties A and Z should

satisfy in order to allow one to discriminate between the three different neutrino mass spectra if |(m)| is measured, or a significantly improved bound on |(m)| is obtained.

Fig. 2. The upper bound on tan2 0g allowing one to discriminate between the IH and the QD neutrino mass spectra, as a function of AmA for different values of Z (see Eq. (16)). The lower (upper) line corresponds to s2 = 0.05 (0).

3.3.1. Testing the quasi-degenerate neutrino mass spectrum

In order to rule out the QD spectrum it is necessary that

f( (mu)min + *)< imcn.

which translates into a condition on the nuclear matrix element uncertainty Z

(l(wî)|exp)MIN V \{m)\ )

For the illustrative sensitivities of the future (PP)0v-decay experiments, (|(m)|exp)MIN = 0.01, 0.02, 0.04 eV, negligible a(|(m)|)/|(m)|, and the predicted values of Km)^ = (0.048-0.056) eV reported in Table 3 (the 3 s.d. case), we have Z < (4.8-5.6), (2.42.8), (1.2-1.4). The better the sensitivity of the future experiments, the larger is the allowed nuclear matrix element uncertainty. Including a non-negligible a(l(m)\)/|(m)| makes even more restrictive the condition on Z .

Proving that the neutrino mass spectrum is of the QD type requires that (|(m)|exp )min — A > |(m)| which implies an upper bound on A :

)min — i(m

exp MIN

Using the values of |(m)|min reported in Table 3 (the 3 s.d. case), we find the corresponding upper bounds A < (52, 52, 44, 46) x 10—3 eV, for (|(m)|exp)MiN = 0.1 eV, and A < (152, 152,144, 146) x 10—3 eV, for (|(m)|exp)MIN = 0.2 eV. If the above condition is fulfilled, condition (15) with Z = 1, which guarantees that the QD spectrum is distinguishable from the NH and IH ones, should also be satisfied. For the illustrative "best-fit" values Am0 = 7.0 x 10—5 eV2 and Am A = 3.0 x 10—3 eV2, and for s2 = 0.05 (0), Eq. (15) (Z = 1) holds if tan2 0g < 0.5 (0.6).

3.3.2. Testing the normal hierarchical spectrum The NH neutrino mass spectrum would be ruled out

(MexpW — A>

Parametrizing (|(m)|exp)MiN as (|(m)|exp)min = y x

Km)C, yNH > 1, we get

A<(yNH — 1)

II NH . max.

How restrictive this condition is depends on the value of |(m)|mHx. Assuming that the more precise measurements of tan2 00, Am0, and Am A will not produce results very different from their current bestfit values, we can use the predictions for Km^m^ given in Table 3 (3 s.d. case)for s2 = 0.05: Km^m^ = 0.0066 eV With this value one finds that for yNH = 40, 30, 20, 10, 7, 5, condition (21) is satisfied if A < 25.7,19.1,12.5, 5.9,4.0,2.6 x 10—2 eV. Alternatively, if experimentally A = 0.30, 0.20, 0.10, 0.05 eV,

condition (21) will hold provided yNH > 46.5, 31.3, 16.2,8.6.

3.3.3. Probing the inverted hierarchical spectrum

For the IH neutrino mass spectrum, |(m)|IH is constrained to lie in the interval given by Eqs. (5) and (6). The IH spectrum can be ruled out if the experimentally measured value of |(m)|, with both the experimental error A and the nuclear matrix element uncertainty factor Z taken into account, lies outside the range given in Eq. (5). There are two possibilities.

Case (i)

(|(m)|exp)MIN — A> IH

II IH , max,

where |(m)|mx depends on the allowed values of and s2 and is given in the captions of Tables 1-3. Using the parametrization (|(m)|exp)MIN =

y" |(m)|max>

1, we are lead to the condition

For yIH = 5, 4, 3, 2, 1.5 this condition is fulfilled if A < (0.28, 0.21, 0.14, 0.07,0.03) (1 — s2) eV The larger the measured value of |(m)|, the larger is the maximal experimental error which still permits to rule out the IH spectrum. Alternatively, for a value of the experimental error A = (0.2, 0.1, 0.05, 0.03) eV it would be possible to rule out the IH spectrum provided yIH > 3.9,2.4,1.7,1.4, respectively.

Case (ii) The spectra with inverted hierarchy can be ruled out also if:

exp MIN

II IH )I

~ (1 - i2)(cos200)minyAm2.

Since |(m)||Hn is of the order of 0.01 eV, the experimental uncertainty will be required to be even below this value, making it not within reach of the currently planed experiments, except possibly for the 10t version of GENIUS. For instance, for (|(m)|exp)MIN = 0.01 eV and A = 0.01 eV one finds, e.g., Z < 1.1 if |(m)|JHn = 0.022 eV.

Probing the IH neutrino mass spectrum requires that the following conditions be fulfilled:

(mlp)min - a > z ((MLJmin + a) < l<m

|IH Imin'

Using the fact that |(m>|mn = cos200|(m>|

the parametrization (|<m>|exp)MiN = ym|(m>|

| max |IH

and the

necessary conditions on Z and yIH read

yIH > cos

\ \(m)\ J

ZyIH <

KIWDV1

\{m)\ J

In the most favorable situation in which a( | (m )|)/ |(m)| ^ 1 and yIH = cos20©, Z is required to be Z < 1/ cos 20©. For the present best-fit values of tan2 0© reported in Table 1, we obtain Z < 2.7 , 2.7 , 2.4 , 2.5. Let us note, however, that from experimental point of view this possibility is rather demanding: as a first approximation, A has to be of the order of, or smaller than, the difference between the maximal and minimal values of |(m)| in the IH case. This difference is typically of the order of ~ (0.02-0.04) eV, and does not exceed ~ 0.06 eV.

If conditions (26) are satisfied, in order to establish the IH spectrum both Eqs. (11) and (15) with Z = 1 should also be valid. Taking as illustrative values Am© = 7.0 x 10-5 eV2 and Am2A = 3.0 x 10-3 eV2, both conditions are satisfied for tan2 0© < 0.5 (0.6) if s2 = 0.05 (0).

3.3.4. The inverted hierarchical versus the quasi-degenerate spectrum

Let us assume that a value of (|(m)|exp )MIN of a few 10 meV has been found, thus ruling out the NH spectrum. The remaining question to ask in this situation would be whether the neutrino mass spectrum is of the IH or QD type. Distinguishing between the two types of spectra might be possible provided

|QD min.

Obviously, one can reach a definite conclusion concerning the type of the spectrum only if the value

or is smaller than

-----r ^ . , .mm

|( m )| Im

of |<m>|exp is larger than |<m>|

(i) |(m)|exp > |(m)|mm: this is equivalent to ruling out the IH spectrum and thus to the case (i) analyzed in Section 3.3.3, see Eq. (23) and the discussion thereafter.

(ii) |(m)|exp < Km)!^: using Eqs. (6) and (8), we find that |(m)|exp < |(m)|JHaxif

1 < Z <

(1 - i2).y(Am2 )max

(l(»î}|exp)MIN + ^

This inequality practically coincides with the second condition in (25). It is more restrictive for smaller values of (AmA)max and larger values of A. Eq. (28) can hold only for a rather limited range of parameters, since the sum of (|(m>|exp)MIN and A has to be smaller than

y (Am2 )max ~ 0.07 eV.

Let us note that the various conditions discussed in this section do not require any additional input from 3H j-decay experiments or from cosmological and astrophysical measurements.

4. Distinguishing between different neutrino CP-parity configurations

In this section we will discuss whether a measurement of |(m>| = 0 might allow one to distinguish between some of the possible neutrino CP-parity configurations when the Majorana phases take CP-conserving values, a21,a31 = 0, . We will denote these configurations by i-1(ncp(v1) vcpV) ncp (V3)), where nCP(vj) is the CP-parity of the neutrino Vj, nCP(vj) = ±i. The possibility of determining the values of the Majorana CP-violating phases in the general case of CP-non-conservation has been discussed in detail in Ref. [23].

Inspecting Tables 1-3 leads to the conclusion that it might be relatively easy to distinguish between the

(+---) and (—1—) configurations in the case of

the IH spectrum (i.e., |(m>|JHn) and the (+---) and

(—I—) configurations for the QD spectrum (i.e., |(m>|2^). The more interesting question is whether it might be possible to distinguish between the different CP-parity configurations for a given type of neutrino

mass spectrum. We will study it briefly in what follows in the cases of IH and QD spectra.10

4.1. Inverted hierarchical spectrum

Due to the smallness of m1|Ue1|2, one cannot

distinguish the (+ + +) from the (---+), as well

as the (+---) from the (— + —), configurations [22].

The first pair of CP-parity configurations corresponds to |(m)|max, while the second corresponds to |(m)|mn.

The CP-parity patterns (+ ++), (---+) and (+---),

(—|—) would be distinguishable if the following condition holds:

(m) |IH > Z I (m)

( ' miiv ~ ( '

iIH Imin'

This can be translated into a condition on Z, which reads

min 1 + (tan2 0q)„

(Am2)max

1 - (tan2 eQ)m

The first ratio in the right-hand side of Eq. (30) is, for an assumed error on Am2A of 10%, approximately 0.8, while for (tan2 00)min = 0 . 2,0.3,0.4,0.5,0.6, the second ratio reads 1.5,1.9,2.3,3.0,4.0, respectively. If Z < 1. 5,2.0, 3.0,4.0, values of tan2 00 > 0 . 3,0.4,0.5, 0.7 are required in order to be possible to distinguish between the two cases under study.

4.2. Quasi-degenerate spectrum

In the case of QD spectrum, the (+---) and

(—|—), and the (+ + +) and (---+), configurations are difficult to distinguish due to the smallness of the mixing parameter s2 limited by the reactor antineutrino experiments [28,29]: the corresponding differences in the predicted values of |(m)| do not exceed ~ 10%. Therefore we shall analyze again the possibility to discriminate between these two pairs. Tak-

ing into account that |(m)|QD__-) > |(m)|Q-+

|(m)|QD++) > |(m)|Q——+), the indicated two pairs of CP-parity configurations can be distinguished if the

10 For an analysis ofthis possibility for the NH spectrum without taking into account the nuclear matrix element uncertainty, see [21, 22].

following inequality holds:

U \IQD >-U \IQD

|(m)|(__+) > Z |(m)|(—+—y (31)

The above inequality leads to the condition

(m0)min 1 + (tan2 00) max

(1 — s2)

Z <--~-(32)

(m0) max 1 — (tan2 00)min(1 — s2)

which is very similar to Eq. (30). Assuming a KATRIN inspired error of 0.28 eV on mo, the first fraction in the right-hand side of Eq. (32) is 0.1, 0.3, 0.6, 0.7 for m0 = (0.35, 0.5, 1.0, 1.5) eV, respectively. If a(3 m0) = 0.10 eV, as is expected from combined astrophysical and cosmological measurements, then for the same fraction one gets 0.83, 0.93, 0.94, 0.96 for m0 = (0.35,0.5,1.0,1.5) eV, respectively. In what regards the second fraction, for s2 = 0 the values from Section 4.1 are valid, while for s2 = 0.05 they read 1.7, 2.1, 2.6, 3.3 for (tan2 00)min = 0.3, 0.4, 0.5, 0.6.

Thus, if m0 is measured in tritium j -decay experiments, relatively large m0 > 1.5 eV and tan2 00 > 0.5 are required in order to distinguish between the (+ — —), (— + —), and the (+ + +), (— — +) CP-parity configurations. If astrophysical and cosmologi-cal measurements provide m 0, then a value of Z < 1.5, 2.0, 3.0, 4.0 would require tan2 00 > 0.2, 0.4, 0.5, 0.6.

5. Conclusions

Assuming 3-neutrino mixing and massive Majorana neutrinos, (PP)0v-decay induced only by the ( V-A) charged current weak interactions, LMA MSW solution of the solar neutrino problem and neutrino oscillation explanation of the atmospheric neutrino data, we have studied the requirements on the "solar" mixing angle 0q, the nuclear matrix element uncertainty factor Z and the experimental error on the effective Majorana mass |(m)|, A, which allow one to distinguish between, and/or test, the normal hierarchical (NH), inverted hierarchical (IH) and quasi-degenerate (QD) neutrino mass spectra if |(m)| = 0 is measured, or a stringent upper bound on |(m)| is obtained. The possibility to discriminate between the three types of spectra depends on the allowed ranges of values of |(m)| for the three spectra: it is determined by the maximal values of |(m)| in the cases of NH and IH spectra,

|(m)|max , and by the minimal values of |(m)| for the IH and QD spectra, |(m)|mjnQD. These are reported in

Tables 1-3. In deriving them we have used the values of the solar and atmospheric neutrino oscillation parameters, 0©, Am©, Am2A and sin2 0, favored by the existing data [1,2,4-6,28,29] (Tables 1 and 2) and assumed prospected precisions of future measurements (Table 3).

For the currently favored values of the neutrino oscillation parameters and sin2 0 = 0, the upper bound on tan2 0© permitting to distinguish the NH from the IH spectrum is satisfied even for Z = 3. If sin2 0 lies close to its present 99.73% C.L. upper limit of 0.05 [34] (see also [28,29]), the upper bound on tan2 0© of interest decreases by up to 50% and values of Z slightly lower than 3 might be required (Fig. 1). For sin2 0 = 0.01, distinguishing between NH and IH spectra would be possible for Z = 3 (2), provided tan2 0© < 0.50 (0.60). The possibility to discriminate between the NH and the QD spectra depends weakly on sin2 0 and on the neutrino mass m12,3 = m0, and the respective conditions are satisfied even for values of Z exceeding 3 (Fig. 1). Without any additional input from 3H beta-decay experiments and/or cosmological and astrophysical measurements, and given the values of tan2 0© and Am2A favored by the data, the IH and QD spectra can be distinguished only if Z < 15 (Fig. 2).

Let us emphasize that the conditions which would allow one to establish the presence of CP violation due to the Majorana CP-violating phases using a measurement of |(m>| = 0 lead to a lower bound on tan2 0© and, in general, require Z < 2. In contrast, the conditions permitting to distinguish between the three types of neutrino mass spectrum imply an upper limit on tan2 0© and in most of the cases can be satisfied even for Z — 3.

We have studied also the conditions on Z and A which would permit to rule out, or establish, the NH, IH and the QD mass spectra. Typically, the next generation of (jj)0v-decay experiments will be able to rule out the QD mass spectrum if Z < 3, and establish it if, e.g., the measured (|(m>|exp)MIN ~ 0.2 (0.1) eV (see Eq. (8)) and the experimental error is A ~ 0.15 (0.05) eV. The NH spectrum can be excluded provided the measured value of |(m>| is, e.g., ~ 10 (7) times larger than |(m>|mx and the experimental error is A < 0.12 (0.06) eV. The IH spectrum can be ruled out for A < 0.07 (0.10) eV provided (|(m>|exp )MIN is at least by a factor of

~ 2.0 (2.5) larger than |(m>|max. Establishing the IH mass spectrum is quite demanding and requires a measurement of |(m>| with an error A < 0.02-0.04 eV.

Finally, we have studied the possibility to distinguish between certain neutrino CP-parity configurations in the case of CP-conservation. Due to the small-ness of sin2 0, there are two pairs of CP-parities in the cases of QD and IH spectra, the two different CP-parity patterns within each pair being indistinguishable. Given the best-fit values of tan2 0©, one can discriminate between these two pairs for the IH mass spectrum if Z < 2. For the QD mass spectrum and if mo is measured in tritium j -decay experiments, relatively large m0 > 1.5 eV and tan2 0© > 0.5 are necessary. If astrophysical and cosmological measurements provide m0, values of Z < 2 are required.

Acknowledgements

Part of the present work was done during the Workshop on (jj)0v -decay at the Institute of Nuclear Theory (ITP) at the University of Washington, Seattle. S.P. and S.T.P. would like to thank the organizers of the Workshop and the members of ITP for kind hospitality and L. Wolfenstein, W. Haxton, P. Vogel, R. Mohapa-tra and F. Avignone for useful discussions. This work was supported in part by the EC network HPRN-CT-2000-00152 (S.T.P. and W.R.), the ItalianMIUR under the program "Fenomenologia delle Interazioni Fonda-mentali" (S.T.P.) and by the US Department of Energy (S.P.).

References

[1] B.T. Cleveland, et al., Astrophys. J. 496 (1998) 505;

Y. Fukuda, et al., Phys. Rev. Lett. 77 (1996) 1683;

V. Gavrin, Nucl. Phys. (Proc. Suppl.) 91 (2001) 36;

W. Hampel, et al., Phys. Lett. B 447 (1999) 127;

M. Altmann, et al., Phys. Lett. B 490 (2000) 16.

[2] Super-Kamiokande Collaboration, Y. Fukuda, et al., Phys. Rev.

Lett. 86 (2001) 5651;

Super-Kamiokande Collaboration, Y. Fukuda, et al., Phys. Rev.

Lett. 86 (2001) 5656.

[3] SNO Collaboration, Q.R. Ahmad, et al., Phys. Rev. Lett. 87

(2001)071301.

[4] SNO Collaboration, Q.R. Ahmad, et al., Phys. Rev. Lett. 89

(2002)011302;

SNO Collaboration, Q.R. Ahmad, et al., Phys. Rev. Lett. 89

(2002)011301.

[5] Super-Kamiokande Collaboration, M. Shiozawa, Talk at the International Conference on Neutrino Physics and Astrophysics "Neutrino '02", May 25-30, 2002, Munich, Germany.

[6] KamLAND Collaboration, K. Eguchi, et al., hep-ex/0212021.

[7] B. Pontecorvo, Chalk River Lab. Report PD-205, 1946; B. Pontecorvo, Zh. Eksp. Teor. Fiz. 53 (1967) 1717.

[8] R. Davis, D.S. Harmer, K.C. Hoffman, Phys. Rev. Lett. 20 (1968) 1205;

R. Davis, D.S. Harmer, K.C. Hoffman, Acta Phys. Acad. Sci. Hung. Suppl. 29 (4) (1970) 371;

R. Davis, in: A. Frenkel, G. Marx (Eds.), Proceedings of the "Neutrino '72" International Conference, Balatonfured, Hungary, June 1972, OMKDK-TECHNOINFORM, Budapest, 1972, p. 5.

[9] S.M. Bilenky, C. Giunti, W. Grimus, Prog. Part. Nucl. Phys. 43 (1999) 1.

[10] S.T. Petcov, hep-ph/9907216.

[11] B. Pontecorvo, Zh. Eksp. Teor. Fiz. 33 (1957) 549; B. Pontecorvo, Zh. Eksp. Teor. Fiz. 34 (1958) 247;

Z. Maki, M. Nakagawa, S. Sakata, Prog. Theor. Phys. 28 (1962) 870.

[12] S.M. Bilenky, S.T. Petcov, Rev. Mod. Phys. 59 (1987) 671.

[13] S.R. Elliot, P. Vogel, Annu. Rev. Nucl. Part. Sci. 52 (2002).

[14] S.M. Bilenky, et al., Phys. Lett. B 94 (1980) 495.

[15] M. Doi, et al., Phys. Lett. B 102 (1981) 323.

[16] L. Wolfenstein, Phys. Lett. B 107 (1981) 77;

S.M. Bilenky, N.P. Nedelcheva, S.T. Petcov, Nucl. Phys. B 247 (1984)61;

B. Kayser, Phys. Rev. D 30 (1984) 1023.

[17] P. Langacker, et al., Nucl. Phys. B 282 (1987) 589.

[18] S.M. Bilenky, et al., Phys. Rev. D 54 (1996) 4432.

[19] S.M. Bilenky, S. Pascoli, S.T. Petcov, Phys. Rev. D 64 (2001) 053010;

S.M. Bilenky, S. Pascoli, S.T. Petcov, Phys. Rev. D 64 (2001) 113003.

[20] W. Rodejohann, Nucl. Phys. B 597 (2001) 110.

[21] S. Pascoli, S.T. Petcov, L. Wolfenstein, Phys. Lett. B 524 (2002)319;

S. Pascoli, S.T. Petcov, hep-ph/0111203.

[22] W. Rodejohann, hep-ph/0203214.

[23] S. Pascoli, S.T. Petcov, W. Rodejohann, Phys. Lett. B 549 (2002) 177.

[24] K. Matsuda, et al., Phys. Rev. D 62 (2000) 093001; K. Matsuda, et al., Phys. Rev. D 63 (2001) 077301;

H. Nunokawa, W.J.C. Teves, R. Zukanovich Funchal, hep-ph/0206137.

[25] M. Fukugita, T. Yanagida, Phys. Lett. B 174 (1986) 45;

For a recent review see, e.g., W. Buchmüller, M. Plümacher, Int. J. Mod. Phys. A 15 (2000) 5047.

[26] S.T. Petcov, A.Yu. Smirnov, Phys. Lett. B 322 (1994) 109.

[27] S.M. Bilenky, et al., Phys. Lett. B 465 (1999) 193; S.M. Bilenky, C. Giunti, hep-ph/9904328;

F. Vissani, JHEP 06 (1999) 022; M. Czakon, etal., hep-ph/0003161;

H.V. Klapdor-Kleingrothaus, H. Päs, A.Yu Smirnov, Phys. Rev. D 63 (2001) 073005.

[28] M. Apollonio, et al., Phys. Lett. B 466 (1999) 415.

[29] F. Boehm, et al., Phys. Rev. Lett. 84 (2000) 3764; F. Boehm, et al., Phys. Rev. D 62 (2000) 072002.

[30] H.V. Klapdor-Kleingrothaus, et al., Mod. Phys. Lett. 16 (2001) 2409.

[31] C.E. Aalseth, et al., Mod. Phys. Lett. 17 (2002) 1475.

[32] V. Barger, K. Whisnant, Phys. Lett. B 456 (1999) 194; T. Fukuyama, et al., Phys. Rev. D 64 (2001) 013001;

D. Falcone, F. Tramontano, Phys. Rev. D 64 (2001) 077302; H. Minakata, H. Sugiyama, Phys. Lett. B 526 (2002) 335; F. Feruglio, A. Strumia, F. Vissani, Nucl. Phys. B 637 (2002) 345.

[33] S. Pascoli, S.T. Petcov, Phys. Lett. B 544 (2002) 239.

[34] G.L. Fogli, et al., hep-ph/0212127.

[35] M. Maltoni, T. Schwetz, J.W.F. Valle, hep-ph/0212129.

[36] A. Bandyopadhyay, et al., hep-ph/0212146.

[37] J.N. Bahcall, M.C. Gonzalez-Garcia, C. Pena-Garay, hep-ph/0212147.

[38] V. Barger, D. Marfatia, hep-ph/0212126;

H. Nunokawa, W.J.C. Teves, R. Zukanovich Funchal, hep-ph/0212202;

P. Aliani, etal., hep-ph/0212212;

P. de Holanda, A.Yu. Smirnov, hep-ph/0212270.

[39] KATRIN Project, A. Osipowicz, et al., hep-ex/0109033.

[40] W. Hu, M. Tegmark, Astrophys. J. Lett. 514 (1999) 65; S. Hannestad, astro-ph/0211106.

[41] H.V. Klapdor-Kleingrothaus, et al., Nucl. Phys. (Proc. Suppl.) 100 (2001) 309.

[42] C.E. Aalseth, F.T. Avignone III, et al., Phys. At. Nucl. 63 (2000) 1225.

[43] NEMO3 Collaboration, C. Marquet, et al., Nucl. Phys. (Proc. Suppl.) B 87 (2000) 298.

[44] E. Fiorini, Phys. Rep. 307 (1998) 309;

GENIUS Collaboration, H.V. Klapdor-Kleingrothaus, et al., J. Phys. G 24 (1998) 483;

EXO Collaboration, M. Danilov, et al., Phys. Lett. B 480 (2000) 12;

Majorana Collaboration, L. De Braeckeleer, in: Proceedings of the Carolina Conference on Neutrino Physics, Columbia, SC, March 2000;

MOON Collaboration, H. Ejiri, et al., Phys. Rev. Lett. 85 (2000) 2917.

[45] A. Staudt, K. Muto, H.V. Klapdor-Kleingrothaus, Europhys. Lett. 13 (1990)31.

[46] V. Lobashev, et al., Nucl. Phys. (Proc. Suppl.) 91 (2001) 280.

[47] C. Weinheimer, et al., Talk at the International Conference on Neutrino Physics and Astrophysics "Neutrino '02", May 2530, 2002, Munich, Germany.

[48] M. Freund, et al., Nucl. Phys. B 578 (2000) 27.

[49] V. Barger, D. Marfatia, K. Whisnant, hep-ph/0210428; P. Huber, M. Lindner, W. Winter, hep-ph/0211300.

[50] S.T. Petcov, M. Piai, Phys. Lett. B 533 (2002) 94.

[51] G.L. Fogli, et al., hep-ph/0208026.

[52] S.M. Bilenky, D. Nicolo, S.T. Petcov, Phys. Lett. B 538 (2002) 77.

[53] A. De Rujula, et al., Nucl. Phys. B 168 (1980) 54.

[54] A. de Gouvea, C. Pena-Garay, Phys. Rev. D 64 (2001) 113011; P. Aliani, et al., hep-ph/0207348;

A. Bandyopadhyay, et al., hep-ph/0211266. [56] M. Spiro, Summary talk at the International Conference on [55] MINOS Collaboration, D. Michael, Talk at the International Neutrino Physics and Astrophysics "Neutrino '02", May 25-Conference on Neutrino Physics and Astrophysics "Neutrino 30, 2002, Munich, Germany.

'02", May 25-30, 2002, Munich, Germany. [57] S.M. Bilenky, J.A. Grifols, Phys. Lett. B 550 (2002) 154.