PHYSICAL REVIEW D 80, 015005 (2009)

Neutrinoless double beta decay and H

! l0
l
decays in the Higgs triplet model
S. T. Petcov,1,2,* H. Sugiyama,1, and Y. Takanishi3,
1

SISSA and INFN-Sezione di Trieste, I-34014 Trieste, Italy

2
IPMU, University of Tokyo, Tokyo, Japan
3
Physik-Department, Technische Universitat Munchen, D-85747 Garching, Germany
(Received 21 April 2009; published 10 July 2009)
The connection between the neutrinoless double beta [

0 ] decay effective Majorana mass jMee j,
and the branching ratios of the decays H

! l
l0
, l, l0 e, , of the doubly charged Higgs boson
H

is analyzed within the Higgs triplet model of neutrino mass generation. We work in the version of the
model with explicit breaking of the total lepton charge conservation, in which H

! l
l0
, l, l0 e, ,
 are the dominant decay modes of H

. It is assumed also that H

are relatively light so that they can
be produced at the CERN LHC and the branching ratios of interest measured. Taking into account the
current and prospective uncertainties in the values of the neutrino mixing parameters most relevant for the
problem studiedthe atmospheric neutrino mixing angle 23 and the CHOOZ angle 13 , and allowing the
lightest neutrino mass and the CP violating Dirac and Majorana phases to vary in the intervals [0, 0.3 eV]
and 0; 2, respectively, we derive the regions of values of BRH

! e
e
and BRH

! e


for which jMee j  0:05 eV, or jMee j < 0:05 eV. This is done for the normal mass ordering case, the
inverted mass ordering case, and the case when the ordering has not been determined. The analysis is
performed both without using possible additional data on BRH

! 

, as well as using
prospective data on these branching ratios. In the latter case, results for several values of BRH

!


are presented.
DOI: 10.1103/PhysRevD.80.015005

PACS numbers: 12.60.Fr, 14.60.Pq, 14.80.Cp

I. INTRODUCTION
Determining the origin of neutrino masses and mixing is
one of the major challenges of future research in neutrino
physics. It is well known that the existence of nonzero
neutrino masses can be related to the presence of a more
complicated Higgs sector in the standard theory, involving
additional Higgs fields beyond the single doublet field.
Actually, it was realized a long time ago [13] that a
Majorana mass term for the left-handed (LH) flavor neutrino fields can be generated by SU2L  U1Y invariant
Yukawa couplings of two lepton doublet fields to a Higgs
triplet field, carrying two units of the weak hypercharge,
jYj 2. Such a Higgs field has electrically neutral, singly
charged, and doubly charged components. The Majorana
mass term for the active flavor neutrinos arises when the
neutral component of the Higgs triplet field acquires a
nonzero vacuum expectation value (vev), breaking the
SU2L  U1Y symmetry. There are several possible
realizations of this scenario. The realization in which the
global U1L symmetry associated with the conservation of
the total lepton charge L is broken only spontaneously by
the Higgs triplet vev [4], was ruled out by the LEP data on
the invisible decay width of the Z0 boson. If, however, the
U1L symmetry is broken explicitly in a manner that leads
to a nonzero vacuum expectation value of the neutral
*Also at: Institute of Nuclear Research and Nuclear Energy,
Bulgarian Academy of Sciences, 1784 Sofia, Bulgaria.

sugiyama@sissa.it

yasutaka.takanishi@ph.tum.de

1550-7998= 2009=80(1)=015005(14)

component of the Higgs triplet field (see, e.g. [5,6]) one

obtains a viable model of neutrino mass generation. This
model has been investigated in detail recently [79] (see
also, e.g. [10,11]) and was shown to have a rich and
physically interesting phenomenology owing to the fact
that (i) the couplings of the doubly and singly charged
Higgs fields to the flavor neutrinos and charged leptons are
proportional to the elements of the Majorana mass matrix
of the (flavor) neutrinos, Ml0 l , and can be relatively large,
and that (ii) the physical doubly charged and singly
charged Higgs fields, H

and H
, can have masses in
the range from 100 GeV to 1 TeV and thus can, in
principle, be produced and observed at the CERN LHC.
Point (i) implies that the indicated couplings are determined essentially by the elements of the PontecorvoMaki-Nakagawa-Sakata (PMNS) neutrino mixing matrix
[12] and by the neutrino masses. In [79], it was shown that
by studying the decays H

! l
l0
, l, l0 e, , , it
might be possible to obtain information on the absolute
neutrino mass scale, on the type of neutrino mass spectrum
[which can be, e.g. normal hierarchical (NH), inverted
hierarchical (IH), and quasidegenerate (QD)], and on the
Majorana CP violating phases [13] present in the neutrino
mixing matrix.
In the present article we investigate the possibility to use
the information on the neutrino mass spectrum and the
Majorana CP violating phases from the measurements of
the H

! l
l0
decay branching ratios, BRH

!
l
l0
, l, l0 e, , in order to obtain predictions for the
effective Majorana mass in neutrinoless double beta

015005-1

2009 The American Physical Society

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

PHYSICAL REVIEW D 80, 015005 (2009)

[

0 ] decay, jMee j (see, e.g. [14]). Our study is motivated by the fact that most probably the searches for the
doubly charged scalars H

and the decays H

! l
l0

will be carried out at LHC before the next generation of


0 -decay experiments will be operative. Among the
different decay channels H

! l
l0
, l, l0 e, , , the
easier to observe are those with two electrons (positrons),
two muons (antimuons), or an electron (positron) and a
muon (antimuon), e
e
, 

, and e

, in the final
state. If the mass of H

does not exceed approximately
400 GeV, the branching ratios of the H

decays into
e
e
, e

, and 

can be measured at LHC with
a few percent error [15]. We will show that if the doubly
charged Higgs bosons H

will be discovered at LHC and
at least the three branching ratios BRH

! e
e
,
BRH

! e

, and BRH

! 

will be
measured with a sufficient accuracy, one can obtain unique
information on the

0 -decay effective Majorana mass
jMee j. This information will be extremely important, in
particular, for the upcoming next generation of


0 -decay experiments.
II. THE HIGGS TRIPLET MODEL
In the Higgs triplet model (HTM) [13] a I 1, Y 2
complex SU2L triplet of Higgs scalar fields is added to
the standard model (SM) Lagrangian. In the 2  2 representation, the Higgs triplet field has the form


0

p

= 2

0

!

p
;

= 2

V m2 y  1 y 2 M
2 Tr
y

2 Tr
y
2 3 Det
y

4 y Tr
y
5 y i Tr
y i



1
(4)
p T i2
y  H:c: ;
2
T  0 T being the SM Higgs doublet field. In
Eq. (4), M
2 > 0 is the common mass of theptriplet
scalars.

The choice m2 < 0 ensures that h0 i v= 2
0, which

breaks spontaneously SU2  U1Y to U1Q . In the
model
of
Gelmini-Roncadelli
[4],
the
term
T i2
y  is absent, which leads for M
2 < 0 to a
spontaneous breaking of the global U1L symmetry associated with the conservation of the total lepton number.
The resulting Higgs spectrum contains a massless
Goldstone bosonthe triplet scalar Majoron, J, and another light scalar, H 0 . The decay Z0 ! H 0 J would give too
large a contribution to the invisible decay width of the Z0
boson and this model was excluded by the LEP data. The
inclusion of the term T i2
y  explicitly breaks the
lepton number conservation when
is assigned two units
of the total lepton charge L, and therefore avoids the
presence of a Goldstone bosonthe Majoron, in the model
[2,3]. Thus, the scalar potential in Eq. (4) together with the
triplet Yukawa interaction of Eq. (2) lead to a phenomenologically viable model of neutrino mass generation.
The expression for v
resulting from the minimization
of the potential V, Eq. (4), reads

(1)
v

where
,
, and

are neutral, singly charged, and
doubly charged scalar fields. In the flavor basis in which
the charged lepton mass matrix is diagonal and which we
are going to use throughout this article, a Majorana mass
term for the LH flavor neutrinos can be generated (without
the introduction of a right-handed neutrino field) by the
SU2L  U1Y gauge invariant Yukawa interaction:
L hl0 l c Tl0 L Ci2
c lL H:c:

(2)

Here hl0 l hll0 , l0 , l e, ,  are complex Yukawa couplings forming a symmetric matrix h, C is the charge
conjugation matrix, 2 is a Pauli matrix for SU2L indices,
and c TlL lL lL T , l e, ,  is the LH lepton doublet
field. A nonzero triplet vacuum expectation value, h
0 i 
v

0, gives rise to a Majorana mass matrix M for the LH
flavor neutrino fields lL :
p
Ml0 l 2hl0 l h
0 i 2hl0 l v
;

l0 ; l e; ; :

(3)

The requisite v

0 arises from the minimization of the

most general SU2  U1Y invariant Higgs potential
[5,6]:

2M
2

v2
;
4 5 v2

for v
 v:

(5)

In the scenario of relatively light triplet scalars within the

discovery reach of the LHC we will be interested in, one
has M
v and Eq. (5) leads to v
. In extensions of
the HTM, the term T i2
y  can arise in various
ways: (i) through the vev of a Higgs singlet field [16,17];
(ii) generated at higher orders in perturbation theory [6]; or
(iii) appearing in the context of theories with extra dimensions [5,18].
An upper limit on v
can be obtained from considering
2
its effect on the parameter MW
=MZ2 cos2 W . In the
SM, 1 at tree level, while in the HTM one has
 1

1 2x2
;
1 4x2

x  v
=v:

(6)

The measurement 1 leads to the bound v
=v & 0:03,

or v
< 8 GeV. At the 1-loop level v
must be renormalized and explicit analyses lead to bounds on its magnitude
similar to those derived from the tree-level analysis [19].
The HTM has seven physical Higgs scalar particles
H ; H ; H ; H  ; H 0 ; A0 ; h0 . The doubly charged
Higgs field H coincides with the triplet scalar field

. The remaining Higgs mass-eigenstate fields are, in

015005-2

NEUTRINOLESS DOUBLE BETA DECAY AND . . .

PHYSICAL REVIEW D 80, 015005 (2009)

general, mixtures of the doublet and triplet fields. The

corresponding mixing parameter is proportional to the ratio
of triplet and doublet vevs, v
=v, and hence is small even if
v
assumes its largest value of a few GeV. Therefore H ,
H 0 , A0 are predominantly composed of the triplet fields,
while h0 is predominantly composed of the doublet field
and plays the role of the SM Higgs boson. The masses of
H

, H
, H0 , A0 are of order M
with splittings of order
5 v. For M
< 1 TeV of interest for direct searches for the
Higgs bosons at the LHC, the couplings hl0 l are constrained
to be O0:1 or less by a variety of processes such as  !
eee,  ! lll, etc. These constraints are reviewed in
[20,21].
In this article we will be interested, in particular, in the
decays of H

into a pair of same-sign charged leptons,
H

! l0
l
, l0 , l e, , , which give clear signals
even in hadron colliders like the LHC. These decays are
important not only for the searches of H

, but also
because of the very interesting possibility that their lepton
flavor dependence can be directly related to the Majorana
mass matrix of the LH flavor neutrinos.
In our analysis, we will assume that MH


MH
and
v
& 1 MeV. Under these conditions the decay H

!
H
W
is forbidden, while the decay H

! W
W
is
sufficiently strongly suppressed. In this case, the branching
ratios of the decays H

! l0
l
are given by the following simple expressions (see, e.g. [79]):
BR l0 l  BRH

! l0
l

2
jh 0 j2
P ll 2
1
l0 l jhl0 l j

(7)

l0 l

2
jMl0 l j2
;
P
1
l0 l m2i

(8)

scale (i.e. lightest neutrino mass), type of neutrino mass

spectrum, Majorana CP violating phases in the neutrino
mixing matrix, etc. This simple relation does not hold in a
type of the left-right symmetric models [23] with Higgs
triplets [24], in which lepton number is spontaneously
broken and Majoron is eaten by new gauge bosons, because of contaminations from the right-handed neutrino
sector.
Among six channels of H

! l0
l
we use BRee ,
BRe , and BR in our analysis. For measurements of
those three branching ratios in pp ! H H  (whose
p
cross section is about 20 fb at MH

300 GeV for s
14 TeV), SM backgrounds come from tt, ttZ, and ZZ.
Those backgrounds can be eliminated easily by requiring
high transverse momentum, exclusion of invariant masses
of l0 l around mZ , and coincidence of two invariant
masses of l0 l and l0 l . The background from W for
pp ! H

H  (whose cross section is about twice of that
of pp ! H H for MH
MH

[25]) can be cut by
taking high transverse invariant mass given by transverse
energy-momentums of l and missing one. Some details on
cuts of SM backgrounds can be seen in [15] for also some
channels involving s.
III. THE NEUTRINO MASSES, MIXING, AND THE


0 DECAY
We work in the flavor basis in which the mass matrix of
the charged leptons is diagonal. As we have shown, in the
Higgs triplet model of interest, the LH flavor neutrino
fields lL acquire a Majorana mass term. The corresponding Majorana mass matrix M is diagonalized with the help
of the PMNS neutrino mixing matrix [12]:

where
is the Kronecker delta. Note that the branching
ratios depend only on the parameters of neutrino mass
matrix. The measurement of BRl0 l can give significant
information on the elements of the neutrino mass matrix1
jMl0 l j, and therefore, e.g. on the absolute neutrino mass
l0 l

UPMNS

c12 c13
@ s12 c23  c12 s23 s13 ei

B
s12 s23  c12 c23 s13 ei

T
Mll0 UPMNS diagm1 ; m2 ; m3 UPMNS
ll0 ;

where mj , j 1, 2, 3, are the real positive eigenvalues of

Mll0 the masses of the Majorana neutrinos j with definite mass. In what follows, we will use the standard parametrization of the PMNS matrix (see, e.g. [26,27]):

s12 c13
c12 c23  s12 s23 s13 ei

c12 s23  s12 c23 s13 ei

where cij  cosij , sij  sinij , the angles ij 0; =2

(i < j 1, 2, 3),
0; 2 is the Dirac CP-violating
phase, and 21 and 31 are two Majorana CP-violation
phases [13,28]. The phases 21 and 31 can vary in the
interval 0; 2. It proves useful for our further discussion
to define also the difference of the two Majorana phases:
32  31  21 . Let us add that at present we do not have
experimental information on
, 21 , and 31 .
The existing neutrino oscillation data [2933] allow us
to determine with a rather good precision the mixing angles

(9)

1
s13 ei

C
s23 c13 Adiag1; ei21 =2 ; ei31 =2 ;
c23 c13

(10)

and neutrino mass squared differences which drive the

solar neutrino and the dominant atmospheric neutrino oscillations, sin2 212 ,
m221 and sin2 223 , j
m231 j
j
m232 j, and to obtain a rather stringent limit on the
CHOOZ angle 13 . In our analysis, we will use the following best fit values of sin2 212 ,
m221 , sin2 223 , and j
m231 j
[3436]:

015005-3

For the current experimental information on jMll0 j, see [22].

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

m221 7:6  105 eV2 ;

j
m231 j 2:4  103 eV2 ;

PHYSICAL REVIEW D 80, 015005 (2009)

sin2 212 0:87;

(11)

sin2 223 1:

(12)

The upper limit on sin2 213 obtained in CHOOZ reactor

antineutrino experiment [32] reads
sin 2 213 < 0:14:

(13)

From the global analyses of the neutrino oscillation data

one finds (see, e.g. [36])
sin 2 13 < 0:056;

99:73%C:L:

(14)

The next generation of experiments with reactor  e , which

are under preparation, Double CHOOZ [37], Daya Bay
[38], RENO [39], can improve the currently reached sensitivity to the value of sin2 213 by a factor of (5)(10) (see,
e.g. [40]), while future long baseline experiments aim at
measuring values of sin2 213 as small as 104 103 (see,
e.g. [41]).
Let us note that the uncertainty in the experimental
determination of sin2 223 corresponds to a rather large
interval of allowed values of s223 [30]: 0:38
s223
0:62.
We will take into account this uncertainty in our numerical
analysis.2 It should be added that the accuracy on sin2 223
is planned to be improved considerably in future long
baseline experiments. The uncertainty in sin2 223 is foreseen to be reduced in the T2K experiment [42], for instance, to sin2 223 > 0:99 (0:45 < s223 < 0:55), if the true
value of sin2 223 1. As we will see, the correlations
between the branching ratios of the decays H

!
l
l0
, BRll0 , l, l0 e, , , and the effective Majorana
mass in neutrinoless double beta [

0 ] decay, jhmij 
Mee , which is the main subject of our study, depend not
only on the elements of the neutrino mixing matrix, but
also on the type of neutrino mass spectrum and on the
absolute scale of neutrino masses.
As is well known, owing to the fact that the sign of
m231
cannot be determined from the existing data, there are two
possible types of neutrino mass spectrum compatible with
the datawith normal ordering and with inverted ordering.
In the standard convention we are going to employ, the two
spectra correspond to
(i) m1 < m2 < m3 ,
m231 > 0, normal ordering (NO),
(ii) m3 < m1 < m2 ,
m231 < 0, inverted ordering (IO).
The -mass spectrum can be (i) normal hierarchical,
q
m1  m2 < m3 , with m2
m221 8:8  103 eV,
q
m3
m231 4:9  102 eV; (ii) inverted hierarchical,
q
m3  m1 < m2 , with m2
m223 4:9  102 eV,
q
m1
m223 
m221 4:8  102 eV;
and
2
Varying sin2 12 in the 3 interval of allowed values of
sin2 12 [34,36], 0:25 & sin2 12 & 0:37, has an essentially negligible effect on the results of our analysis.

(iii) quasidegenerate), m1 m2 m3 , m21;2;3 j
m231 j.

In the latter case one has mj * 0:10 eV.
The type of neutrino mass spectrum, i.e. sgn
m231 , can
be determined by studying oscillations of neutrinos and
antineutrinos, say,  $ e and   $  e , in which matter
effects are sufficiently large. This can be done in long
baseline -oscillation experiments (see, e.g. [41]). If
sin2 213 * 0:05 and sin2 23 * 0:50, information on
sgn
m231 might be obtained in atmospheric neutrino experiments by investigating the effects of the subdominant
transitions e ! e and  e !  e of atmospheric
neutrinos which traverse the Earth [43]. For e [or  e ]
crossing the Earth core, new type of resonancelike enhancement of the indicated transitions takes place due to
the (Earth) mantle-core constructive interference effect
[neutrino oscillation length resonance (NOLR)] [44].3
For
m231 > 0, the neutrino transitions e ! e are
enhanced, while for
m231 < 0 the enhancement of antineutrino transitions  e !  e takes place, which might
allow to determine sgn
m231 . If sin2 13 is sufficiently
large, the sign of
m231 can also be determined by studying
the oscillations of reactor  e on distances of 2040 km
[48]. An experiment with reactor  e , which, in particular,
might have the capabilities to measure sgn
m231 , was
proposed recently in [49] (see also [50]). Information on
the type of neutrino mass spectrum can also be obtained in

-decay experiments having a sensitivity to neutrino
q
masses  j
m2A j 5  102 eV [51] (i.e. by a factor
of 4 better than the planned sensitivity of the KATRIN
experiment [52]; see below).
Direct information on the absolute neutrino mass scale
can be derived in 3 H
-decay experiments [5254]. The
most stringent upper bounds on the  e mass were obtained
in the Troitzk [54] and Mainz [52] experiments:
m e < 2:3 eV;

95%C:L:

(15)

We have m e m1;2;3 in the case of the QD -mass spectrum. The KATRIN experiment [52], which is under preparation, is planned to reach a sensitivity of m e  0:20 eV,
i.e. it will probe the region of the QD spectrum.
The CMB data of the WMAP experiment [55], combined with data from large scale structure surveys
(2dFGRS, SDSS), lead to the following upper limit on
the sum of neutrino masses (see, e.g. [56]):
3
As a consequence of this effect, the corresponding e [or
 e ] transition probabilities can be maximal [45] (for the
precise conditions of the mantle-core (NOLR) enhancement,
see [44,45]). Let us note that the Earth mantle-core (NOLR)
enhancement of neutrino transitions differs [44] from the
Mikheev-Smirnov-Wolfenstein one. The conditions of the
Earth mantle-core enhancement [44,45] also differ [46] from
the conditions of the parametric resonance enhancement of the
neutrino transitions discussed in the articles [47].

015005-4

NEUTRINOLESS DOUBLE BETA DECAY AND . . .

X
j

mj   < 0:41:7 eV;

95%C:L:

PHYSICAL REVIEW D 80, 015005 (2009)

(16)

Data on weak lensing of galaxies, combined with data from

the WMAP and PLANCK experiments, may allow  to be
determined with an uncertainty of 0:04 eV [56,57].
In our analysis, we will consider both types of neutrino
mass spectrumwith normal and with inverted ordering,
as well as the specific cases of NH, IH, and QD spectra.
Correspondingly, the lightest neutrino mass minmj 
m0 , which determines the absolute neutrino mass scale,
will be varied in the interval
0
m0
0:3 eV;

m0  minmj ;

j 1; 2; 3:
(17)

As we will show, the results we obtain essentially do not

depend on the maximal value m0 as long as the latter is not
smaller than 0.3 eV. The reason is that for m0 * 0:3 eV
(i.e. in the QD region), the branching ratios BRl0 l we are
interested in practically do not depend on the neutrino
masses:
BR l0 l





X

2
2




Ul0 j Ulj 
;





31
l0 l  j

m0 * 0:3 eV:
(18)

A. Neutrinoless Double Beta decay

In the Higgs triplet model the massive neutrinos are
Majorana particles. Determining the nature of massive
neutrinos is of fundamental importance for understanding
the origin of neutrino masses and, more generally, for
understanding the symmetries governing the particle interactions. The existence of massive Majorana neutrinos is
associated with nonconservation of the total lepton charge.
In this case, the neutrinoless double beta decay A; Z !
A; Z 2 2e is allowed (see, e.g. [14,58,59]).
Assuming that the dominant mechanism for the decay is
the exchange of light Majorana neutrinos (Fig. 1(a)), the
0
half-life T1=2
for the decay is given by
0
G0 jM0 j2 jMee j2 1 ;
T1=2

(19)

where G0 is a phase space factor and M0 is the nuclear

matrix element (NME) of the process. All the dependence
0
of T1=2
on the neutrino masses and mixing parameters
factorizes into the

0 -decay effective Majorana mass
Mee :
jMee j jc212 c213 m1 s212 c213 m2 ei21 s213 m3 ei31 2
j:
(20)

FIG. 1. Diagrams for the neutrinoless double beta decay. The

diagram (a) is the standard and dominant one, and (b)-(f) are
possible but negligible in the HTM.

The most stringent upper bound on jMee j, jMee j <

0:351:05 eV was obtained by using the lower limit
0
> 1:9  1025 yr (90% C.L.) found4 in the
T1=2
Heidelberg-Moscow 76 Ge experiment [61]. The IGEX col0
> 1:57  1025 yr
laboration has obtained the result T1=2
(90% C.L.), from which the limit jMee j <
0:331:35 eV was derived [62]. A positive
0

0 -decay signal at >3 , corresponding to T1=2

25
0:694:18  10 yr (99.73% C.L.) and implying
jMee j 0:10:9 eV, is claimed to be observed in [63],
while a recent analysis reports evidence at 6 of

0
decay with jMee j 0:32
0:03 eV [64]. Two experiments, NEMO3 (with 100 Mo, 82 Se, etc.) [65] and
CUORICINO (with 130 Te) [66], designed to reach a sensitivity to jMee j  0:20:3 eV, set the limits, jMee j <
0:611:26 eV [65] and jMee j < 0:190:68 eV [66]
(90% C.L.), where estimated uncertainties in the NME
are accounted for. The two upper limits were derived
from the experimental lower limits on the half-lives of
100 Mo and 130 Te, T 0 > 5:8  1023 yr (90% C.L.) [65]
1=2
0
> 3:0  1024 yr (90% C.L.) [66]. Most imporand T1=2
tantly, a large number of projects aim at a sensitivity to
jMee j  0:010:05 eV [67]: CUORE (130 Te), GERDA
(76 Ge), SuperNEMO, EXO (136 Xe), MAJORANA (76 Ge),
MOON (100 Mo), COBRA (116 Cd), XMASS (136 Xe),
CANDLES (48 Ca), etc. These experiments, in particular,
will test the positive result claimed in [63].
The predicted value of jMee j depends strongly on the
type of -mass spectrum [26,68], more precisely, on the
In the quoted upper bound for jMee j a factor of 3 uncertainty
in the relevant NME (see, e.g. [60]) is taken into account.

015005-5

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

PHYSICAL REVIEW D 80, 015005 (2009)

type of hierarchy neutrino masses obey. The existence of

significant and robust lower bounds on jMee j in the cases of
IH and QD spectra [68] (see also [69]), given, respectively,5 by jMee j * 0:01 eV and jMee j * 0:03 eV, which
lie either partially (IH spectrum) or completely (QD spectrum) within the range of sensitivity of the next generation
of

0 -decay experiments, is one of the most important
features of the predictions of jMee j. At the same time we
have jMee j & 5  103 eV in the case of NH spectrum
[70]. The fact that maxjMee j in the case of NH spectrum
is considerably smaller than minjMee j for the IH and QD
spectrum opens the possibility of obtaining information
about the type of -mass spectrum from a measurement
of jMee j
0 [68]. More specifically, a positive result in the
future generation of

0 -decay experiments with
jMee j > 0:01 eV would imply that the NH spectrum is
strongly disfavored (if not excluded). For
m231 > 0, such
a result would mean that the neutrino mass spectrum is
with normal ordering, but is not hierarchical. If
m231 < 0,
the neutrino mass spectrum should be either IH or QD.
IV. PREDICTION FOR jMee j FROM
MEASUREMENTS OF BRH

! l0
l

In this section, we investigate within the HTM the
predictions one can obtain for the

0 -decay effective
Majorana mass jMee j by using data on BRl0 l . The dominant
mechanism of

0 decaythe light Majorana neutrino
exchange corresponds to the diagram in Fig. 1(a), the
contributions from the diagrams in Figs. 1(b)1(f) being
negligible [71,72].
We use three branching ratios, BRee , BRe , and BR ,
in our analysis. The expressions for these branching ratios
in terms of neutrino masses, neutrino mixing angles, and
CP violating phases read (see [79])
X
i


m2i BRee jMee j2
jc212 c213 m1 s212 c213 m2 ei21
s213 m3 ei31 2
j2 ;

s12 c13 c12 c23  s12 s23 s13 ei

m2 ei21

c12 c23  s12 s23 s13 ei

2 m2 ei21
s223 c213 m3 ei31 j2 :

(22)

Up to small corrections we have in the cases of two spectra

[68]: jMee j * j
m232 cos212 j (IH) and jMee j * m0 cos212
(QD). The possibility of cos212 0 is ruled out at 6 by
the existing data which also imply that cos212 * 0:30 (0.26) at
2 (3 ) [34,36]. We also have 2:07  103 eV2 & j
m232 j &
2:75  103 eV2 at 3 .

(23)

Given the solar and atmospheric neutrino oscillation

parameters and the CHOOZ angle, sin2 12 ,
m221 ,
sin2 23 , j
m231 j, and 13 , jMee j depends on m0 
minmj , 21 , 31 ,
and on the type of neutrino mass
spectrum (NO or IO). In the case of spectrum with IO or of
QD type, the dependence of jMee j on 13 is relatively weak
and can be neglected, as long as6 cos212 sin2 13 . In
this case jMee j does not depend on the Majorana phase 31
and on the Dirac phase
. From the measurement of the
three observables, BRee , BRe , and BR , three parameters, say, m0 , 21 , and 31 , can, in principle, be determined
and information on the type of neutrino mass spectrum
with NH, IH or QDcan be obtained [79]. This would
allow us to tightly constrain jMee j. Let us review briefly the
predictions for BRee , BRe , and BR in the cases of NH,
IH, and QD spectrum (see also [79]).
A. NH spectrum, m1  m2 < m3
Using the best fit values of the neutrino oscillation
parameters one finds7 that in this case BRee , BRe , and
BR can take values in the following intervals: 0 &
2
NH
NH
BRNH
ee & 10 , 0 & BRe & 0:08, 0:16 & BR & 0:31.
NH
We get BRee 0 for 32  2
 and s213
q
s212
m221 =
m231 0:05, while BRNH
e 0 for 32 


and
s213 s212 c212 cot2 23
m221 =
m231
6:9  103 . The minimal and maximal values of BRNH

depend weakly on s213 . Neglecting this dependence, one
obtains a simple expression for the Majorana phase (difference) 32 in terms of BRNH
 :


4
1
m231 1=2 BRNH
  s23
:
(24)
cos32
2
2
2
2
2
m21
c12 c23 s23

(21)

X 
m2i BRe 2jc12 c13 s12 c23  c12 s23 s13 ei
m1

s23 c13 s13 m3 ei31 

j2 ;

X 
m2i BR js12 c23  c12 s23 s13 ei
2 m1

B. IH spectrum, m3  m1 < m2
We find very different results in this case:
4
IH
0:5c413 cos2 212 & BRIH
ee & 0:5c13 or 0:06 & BRee & 0:5,
IH
IH
0 & BRe & 0:48, 0 & BR & 0:14. One has BRIH

2
IH
0 for
0, 21 , and s13 0:036. Now BRee exhibits a very weak dependence on s213 . For the Majorana phase
21 we obtain in terms of BRIH
ee :
6
The inequality cos212 sin2 13 is fulfilled for the 2
experimentally allowed ranges of values of cos212 and
sin2 13 ; see, e.g. [34,36]. If one uses the 3 ranges, one obtains
sin2 13 = cos212 & 0:22.
7
The limiting values quoted in this paragraph are obtained for
the best fit values of the neutrinos oscillation parameters and for
sin2 213
0:14.

015005-6

NEUTRINOLESS DOUBLE BETA DECAY AND . . .

cos21 1 

c413  2BRIH
ee
:
2c413 c221 s221

PHYSICAL REVIEW D 80, 015005 (2009)

(25)

C. QD spectrum, m1;2;3 * 0:1 eV

It is not difficult to convince oneself that the branching
ratios of interest for the QD spectrum to a good approximation can take values in the following intervals:
QD
cos2 212 =3 & BRQD
ee & 1=3 or 0:03 & BRee & 0:33, 0 &
QD
QD
2
BRe & 0:46, cos 223 =3 & BR & 0:33. Actually, we
IH
have up to small corrections BRQD
ee 2=3BRee . For the
Majorana phase 21 in this case we get
cos21 1 

c413  3BRQD
ee
:
2c413 c221 s221

(26)

Given 21 and a sufficiently large s13 , information about

the Dirac phase
and the Majorana phase 31 can be
QD
obtained from the knowledge of BRQD
e and BR . If,
however, a stringent limit on s13 will be obtained, 31
can be determined using BRQD
 and the knowledge of 21 .
It is clear from the above simple analysis that the measurement of the three branching ratios BRee , BRe , and
BR would provide information about the type of neutrino mass spectrum and the Majorana phases. If, for
instance, it is experimentally established that BRee >
0:01, the neutrino mass spectrum of NH type will be
excluded. The spectrum can either be of IH or QD type.
If in addition BR is determined to be BR > 0:14, the
IH spectrum will be ruled out. If, however, the neutrino
mass spectrum will turn out to be QD, it will be very
m23

cos21 1 

difficult (if not practically impossible) to distinguish between the spectrum with NO and that with IO, i.e. to get
information about the sign of
m231 .
Consider next the more general case of m0 having an
arbitrary value. First, let us set 13 0 for simplicity. We
will consider the case of 13
0 later. For 13 0, the
main uncertainty in the prediction of jMee j comes from the
lack of knowledge of m0 and 21 . Note that in this case
BRee and BRe are independent of 31 , similar to Mee .8
Knowing these two branching ratios allows us to determine
m0 and cos21 . In the case of spectrum with NO we have
for the lightest neutrino mass
m21

(27)
m1
cos21 q
m21
m221

m221 2c212 c223 BRee  s212 BRe  m21 BRe

:
q
2c212 s212 m1 m21
m221 BRe 2BRee c223
(28)

Equation (27) is valid also for the second to lightest neutrino mass m1 in the case of spectrum with IO. The expression for cos21 obviously cannot be used to determine
cos21 for m1 0: for 13 0 and m1 0, neither BRee
nor BRe depend on 21 . In the case of spectrum with IO
we obtain

2
m223 
m221 2c223 BRee BRe  2c223
m223  c212
m221

;
2c223  6c223 BRee  3BRe



BRe

m221

O
:
2c212 s212 BRe 2BRee c223

m223
(30)

2c223 BRee BRe

sgn
m231  BRee
 2:4  103 eV2 ;
1  3BRee  3BRe

2c223 m21 s212
m221

;
3m21
m221
m231

(33)

2c223 m23
m223

;
3m23 2
m223

(34)

NO spectrum;
2c223 BRee BRe

2c2
m2  2c2 3s2  1
m221

23 2 31 2 23 12
BRee
2c23  6c23 BRee  3BRe

(29)

where in the last equation we have used the best fit value of
23 and have neglected the term 3s212  1
m221 =
m231 .
Note that the denominator in the expression for jMee j,
Eq. (31), does not go through zero since we have

As can be expected, for m21
m221 , the expression for

cos21 in the case of spectrum with NO coincides with that
for spectrum with IO. Using Eq. (27), we get a universal
expression for jMee j valid for both types of spectrum
with NO and IO and any hierarchy between neutrino
masses:
X 
jMee j2
m2i BRee

m221
m231 2c223 BRee BRe  2s212 c223
m221

;
2c223  6c223 BRee  3BRe

IO spectrum;
(31)

where we have neglected terms 
m221 =
m223 in the

second equation. We see that in the QD region, where

(32)

BRe is independent of 31 as well, but this mode is more

difficult to measure than the two modes we are discussing.

015005-7

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

PHYSICAL REVIEW D 80, 015005 (2009)

NO
0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.1

0.2

BRee

0.3

0.4

BRee

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe=1

0.6

0.5

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe=1

0.6

unknown sign of m31

IO

0.6

0.7

0.1

0.2

0.3

0.4

0.5

0.6

0.7

BRee

FIG. 2 (color online). Values of BRee and BRe , for which jMee j > 0:05 eV (black areas limited by the dashed blue lines) or jMee j <
0:05 eV (grey areas limited by the dash-dotted green lines) in the HTM. The solid (red) line shows the entire region of allowed values
of BRee and BRe in the HTM (black and grey areas and the white area between the colored one). The results shown are obtained by
varying m0 , sin2 223 , sin2 213 ,
, 21 , and 31 , in the ranges given in Eqs. (17) and (36). The left and middle panels correspond to NO
and IO spectrum, respectively, while the right panel was obtained assuming that sgn
m231 is unknown. The dotted line corresponds to
BRee BRe 1. The region above this line is unphysical. See text for further details.

jMee j has a relatively large value, one has 2c223 BRee

BRe 2=3c223 1
m223 =3m20 .
It follows from Eq. (32) that jMee j > 0:05 eV
q
j
m231 j can be predicted without the knowledge of
sgn
m231 , if the collider experiments show that the
branching ratios BRee and BRe satisfy
4c223
2c2
8c2
2c2
BRee 23 * BRe *  23 BRee 23 : (35)
3
3
3
3
2
If indeed sin 13 is negligibly small and these conditions
are satisfied by the measured BRee and BRe , a positive
result can be expected in the next generation of


0 -decay experiments having a sensitivity to jMee j 
0:05 eV. Note that the magnitude of the left and right sides
of the inequality is very sensitive to the value of c223 . Note
also that these conditions do not depend explicitly on
BR . For this reason we will first obtain constraints on
jMee j using only the branching ratios BRee and BRe .
Next, we analyze the case of 13
0 numerically. We
calculated BRee , BRe , and Mee by using j
m231 j,
m221 ,
and sin2 212 given in Eq. (11) and (12). We allow m0 to
vary in the interval in Eq. (17), while the other parameters
are varied in the following ranges reflecting the uncertainties in their knowledge or lack of any constraints:


sin 2 223 > 0:94;

sin2 213 < 0:14;

; 21 ; 31 0  2:

(36)

Later we will present results for the prospective smaller

uncertainties in sin2 223 and sin2 213 , corresponding to
sin2 223 > 0:99 and sin2 213 < 0:04.
In Fig. 2, we show the regions in the BRee  BRe plane
where we definitely have jMee j  0:05 eV or jMee j <

0:05 eV. More specifically, the solid (red) line determines

the complete allowed region in the HTM, corresponding to

m221 , j
m231 j, and sin2 212 given in Eqs. (11) and (12),
and values of m0 , sin2 223 , sin2 213 ,
, 21 , and 31 ,
which were allowed to vary in the ranges specified in
Eqs. (17) and (36). The dashed blue (dash-dotted green)
lines determine the black (grey) regions where jMee j is
definitely larger (smaller) than 0.05 eV in the HTM when
m0 , sin2 223 , sin2 213 ,
, 21 , and 31 , are varied within
the indicated intervals [i.e. Eqs. (17) and (36)]. For values
of BRee and BRe from the region depicted in white (and
located between those shown in black and in grey), the
determination of jMee j is not unambiguous: both values of
jMee j  0:05 eV and jMee j < 0:05 eV are possible. This
degeneracy can be lifted to certain extent, but not completely, by using additional information on BR (see
further). The dotted black line in Fig. 2 corresponds to
BRee BRe 1. We show it only to indicate the boundary of the region of possible values of BRee and BRe : the
area above this line is unphysical. The results in Fig. 2 are
obtained without using the possible additional data on9
BR . The left and middle panels correspond to NO and
IO spectrum, respectively, while the results shown in the
right panel were obtained assuming that the sgn
m231 (i.e.
the type of the neutrino mass spectrum) is unknown. The
black area where jMee j is, e.g. definitely larger than 0.05 eV
in the right panel corresponds to the intersection of the
black areas in left and middle panels. Note that we can
have jMee j * 0:05 eV also in the region shown in white
and located between the grey areas in the right panel of
9

We recall that in this analysis we do not use possible data on

BR , BR , and BRe .

015005-8

NEUTRINOLESS DOUBLE BETA DECAY AND . . .

PHYSICAL REVIEW D 80, 015005 (2009)

NO, BR=0
0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

unknown sign of m31, BR=0

IO, BR=0

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

0.6

0.7

BRee

FIG. 3 (color online). The same as in Fig. 2, but assuming that the experimentally determined BR 0. The dotted line
corresponds to BRee BRe BR 1; the region above the line is unphysical. See text for further details.

Fig. 2. This cannot be unambiguously predicted, however,

knowing only the values of BRee and BRe which lie in the
white area.
Next we show how the results discussed above change
when we add information on BR . The 

decay
mode of H

is relatively easy to measure at LHC by
virtue of the two same-sign muons in the final state. We
present results for BR 0, 0.1, 0.2, 0.3 in Figs. 36,
respectively, where BR 0 in practice corresponds to
BR < 0:01. When we quote a specific value of BR
x, we include an uncertainty of
0:01 in x, i.e. we use
BR x
0:01 in the numerical calculations. The dotted lines in Figs. 36 correspond to BRee BRe
BR 1. We do not use this constraint: the line represents the boundary of the physical region of values of BRee ,
BRe , and BR .
It is clear from Fig. 3 that the measurement of BR can
improve the predictability of jMee j: the relative magnitude
of the white degeneracy region, in general, is smaller
NO, BR=0.1
0.7

0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

BRee

0.5

0.6

0.7

0.1

0.2

0.3

0.4

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

BRe

unknown sign of m31, BR=0.1

IO, BR=0.1

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

than in the case when no information on BR is available.

This is not the case, however, for IO spectrum and values of
BR 0:1, 0.2 (Figs. 4 and 5, middle panels).
Figures 26 show also the regions (dash-dotted line)
where one definitely has jMee j < 0:05 eV. If the measured
values of BRee , BRe , and BR lie in one of these
regions, the observation of

0 decay in the next generation of experiments can be extremely challenging. Even
in such a case, however, searches for the

0 decay are
important and necessary also as a test of the HTM itself. If
the

0 decay is observed while the measured values of
the H

leptonic decay branching ratios imply, e.g. a
negative result of the searches for

0 decay, we will
be led topconclude
that Ml0 l and hl0 l are not directly related:

Ml0 l
2hl0 l v
. Such a situation can arise, for instance, if

v
0, or in models with H which is not an SU2L
triplet, but, e.g. is a Y 4 singlet [73] with couplings to the
charged leptons given by hl0 l l0R C lR H .

0.6

0.7

BRee
FIG. 4 (color online). The same as in Fig. 3, but for BR 0:1.

015005-9

0.1

0.2

0.3

0.4

BRee

0.5

0.6

0.7

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

PHYSICAL REVIEW D 80, 015005 (2009)

NO, BR=0.2
0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

unknown sign of m31, BR=0.2

IO, BR=0.2

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

0.6

0.7

BRee

FIG. 5 (color online). The same as in Fig. 3, but for BR 0:2.

NO, BR=0.3
0.7

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.1

0.2

BRee

0.5

0.6

0.7

0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe=1

0.6

0.4

0.4

BRe

0.4

BRe

0.5

0.3
0.2

0.2

0.1

0.1

0.1

0.2

0.3

0.4

BRee

0.5

0.6

0.7

0.5

0.6

0.7

0.3

0.2

0.1

0.4

0.1

0.2

0.3

0.4

BRee

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe=1

0.6

0.5

0.3

0.3

unknown sign of m231

IO

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe=1

0.2

BRee

0.5

0.1

The same as in Fig. 3, but for BR 0:3.

NO
0.6

0.4

BRee
FIG. 6 (color online).

0.7

0.3

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
BRee+BRe+BR=1

0.6

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

unknown sign of m231, BR=0.3

IO, BR=0.3

0.6

0.7

0.1

0.2

0.3

0.4

0.5

0.6

0.7

BRee

FIG. 7 (color online). The same as in Fig. 2, but allowing sin2 223 and sin2 213 to vary in the following more narrow intervals:
sin2 223 > 0:99 and sin2 213 < 0:04. No information on BR was used in deriving the results shown in the figure. See text for further
details.

015005-10

NEUTRINOLESS DOUBLE BETA DECAY AND . . .

PHYSICAL REVIEW D 80, 015005 (2009)

NO, BR=0
0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

unknown sign of m31, BR=0

IO, BR=0

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

0.6

0.7

BRee

FIG. 8 (color online). The same as in Fig. 7, but assuming that the experimentally determined BR 0. The dotted line
corresponds to BRee BRe BR 1; the region above the line is unphysical.

NO, BR=0.2
0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

unknown sign of m31, BR=0.2

IO, BR=0.2

0.6

0.7

0.1

0.2

BRee

0.3

0.4

0.5

0.6

0.7

BRee

FIG. 9 (color online). The same as in Fig. 8, but for BR 0:2.

NO, BR=0.3
0.7

0.7

possible in HTM
Mee > 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

0.4

0.4

0.4

BRe

0.5

0.3

0.3

0.3

0.2

0.2

0.2

0.1

0.1

0.1

0.1

0.2

0.3

0.4

BRee

0.5

0.6

0.7

0.1

0.2

0.3

0.4

0.5

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

0.5

BRe

BRe

0.7

possible in HTM
Mee > 0.05eV is predicted
Mee < 0.05eV is predicted
BRee+BRe+BR=1

0.6

unknown sign of m31, BR=0.3

IO, BR=0.3

0.6

0.7

BRee
FIG. 10 (color online). The same as in Fig. 8, but for BR 0:3.

015005-11

0.1

0.2

0.3

0.4

BRee

0.5

0.6

0.7

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

PHYSICAL REVIEW D 80, 015005 (2009)

We have performed the same analysis, but with reduced

uncertainties in sin2 223 and sin2 213 : sin2 223 > 0:99 and
sin2 213 < 0:04. The indicated precisions (or better ones)
in the determination of sin2 223 and sin2 213 are expected
to be achieved in the upcoming T2K [42] and reactor
antineutrino experiments Double CHOOZ [37], Daya
Bay [38], and RENO [39], respectively. The results of
the analysis are shown graphically in Figs. 710. The
notations in Figs. 710 are the same as in Figs. 26. We
see from Figs. 710 that improving the precision on
sin2 223 and sin2 213 leads to a noticeable reduction of
the regions of values of BRee and BRe , for which it is
impossible to determine whether jMee j  0:05 eV or
jMee j < 0:05 eV. As Fig. 8 demonstrates, the reduction
will be particularly significant if the measured BR <
102 (which we remind the reader corresponds to the case
denoted by us as BR 0).
V. CONCLUSIONS
We have investigated the connection between the


0 -decay effective Majorana mass jMee j, and the
branching ratios of the decays H

! l
l0
, l, l0 e,
, of the doubly charged Higgs boson H

within the
HTM of neutrino mass generation. Our analysis was performed within the version of the model with explicit breaking of the total lepton charge conservation, in which
H

! l
l0
, l, l0 e, , , are the dominant decay
modes of H

. In this model, the couplings of the doubly
charged Higgs field H to the flavor neutrinos and
charged leptons are proportional to the elements of the
Majorana mass matrix of the (flavor) neutrinos Ml0 l and
the branching ratios BRH

! l
l0
are entirely determined by the elements of the PMNS matrix and neutrino
masses. The latter possibility is realized if the mass of the
doubly charged Higgs scalar does not exceed the mass of
the singly charged one, MH


MH
, and if the vacuum
expectation value of the neutral component of the Higgs
triplet field satisfies v
& 1 MeV. The model under discussion was shown [79] to have a rich and physically
interesting phenomenology owing to the fact that the
physical doubly charged and singly charged Higgs fields,
H

and H
, can have masses in the range from
100 GeV to 1 TeV and thus can, in principle, be
produced and observed at LHC. More importantly, by
studying the decays H

! l
l0
, l, l0 e, , , it might
be possible to obtain information on the absolute neutrino
mass scale, on the type of neutrino mass spectrum and on
the Majorana CP violating phases present in the neutrino
mixing matrix [79].
In the present article we have investigated the possibility
to use the information on the neutrino mass spectrum and
the Majorana CP violating phases from the measurements
of the H

! l
l0
decay branching ratios, BRH

!
l
l0
, l, l0 e, , in order to obtain predictions for the
effective Majorana mass in neutrinoless double beta

[

0 ] decay, jMee j. Among the different decay channels

H

! l
l0
, l, l0 e, , , the easier to observe and
measure the corresponding branching ratios with high
precision are those with two electrons (positrons), two
muons (antimuons), or an electron (positron) and a muon
(antimuon), e
e
, e

, and 

, in the final state. If
the mass of H

does not exceed approximately 400 GeV,
the branching ratios of the H

decays into e
e
, e

,
and 

can be measured at LHC with a few percent
error [15].
Taking into account the current and prospective uncertainties in the values of the neutrino mixing parameters
most relevant for the problem studiedthe atmospheric
neutrino mixing angle 23 and the CHOOZ angle 13 and
allowing the lightest neutrino mass and the CP violating
Dirac and Majorana phases to vary in the intervals [0,
0.3 eV] and 0; 2, respectively, we have derived the
regions of values of BRH

! e
e
and BRH

!
e

for which we definitely have jMee j  0:05 eV, or
jMee j < 0:05 eV. This is done for neutrino mass spectrum
with NO, IO, and in the case when the type of the spectrum
is not known. In what concerns the branching ratio
BRH

! 

, we have considered two cases:
(i) the possible data on BRH

! 

is not used
as an additional constraint in the analysis, (ii) the possible
data on BRH

! 

is included in the analysis. In
the latter case, results for several values of BRH

!


have been obtained.
Our results are presented graphically in Figs. 210. They
show that if the doubly charged Higgs bosons H

will be
discovered at LHC and at least the two branching ratios
BRH

! e
e
and BRH

! e

will be measured with a sufficient accuracy, one can obtain important
information on the

0 -decay effective Majorana mass
jMee j. In the various cases considered, we have identified
the regions values of BRH

! e
e
and BRH

!
e

, for which jMee j is definitely bigger or smaller than
0.05 eV (Fig. 2). We have shown also that due to (i) the
uncertainties in the determination of sin2 223 and sin2 213 ,
(ii) the absence of data on the CP violating phases in the
neutrino mixing matrix, and (iii) the existing rather loose
upper bound on the absolute neutrino mass scale, there
exist also noticeable regions of values of BRH

!
e
e
and BRH

! e

for which it is impossible
to determine unambiguously whether jMee j  0:05 eV or
jMee j < 0:05 eV (Fig. 2). This degeneracy can be partially
lifted by using the additional information from a measurement of BR (Figs. 36).
The same analysis was performed with reduced uncertainties in sin2 223 and sin2 213 corresponding to
sin2 223 > 0:99 and sin2 213 < 0:04. The results are presented graphically in Figs. 710. They show that improving the precision on sin2 223 and sin2 213 leads to a
noticeable reduction of the regions of values of BRee and
BRe for which it is impossible to determine whether

015005-12

NEUTRINOLESS DOUBLE BETA DECAY AND . . .

PHYSICAL REVIEW D 80, 015005 (2009)

jMee j  0:05 eV or jMee j < 0:05 eV. The reduction will

be particularly significant if the measured BR < 102 .
ACKNOWLEDGMENTS
This work was supported in part by the INFN under the
program Fisica Astroparticellare, by the European
Network of Theoretical Astroparticle Physics ILIAS/N6

[1] W. Konetschny and W. Kummer, Phys. Lett. 70B, 433

(1977).
[2] T. P. Cheng and L. F. Li, Phys. Rev. D 22, 2860 (1980).
[3] J. Schechter and J. W. F. Valle, Phys. Rev. D 22, 2227
(1980).
[4] G. B. Gelmini and M. Roncadelli, Phys. Lett. 99B, 411
(1981).
[5] E. Ma, M. Raidal, and U. Sarkar, Phys. Rev. Lett. 85, 3769
(2000); Nucl. Phys. B615, 313 (2001).
[6] E. J. Chun, K. Y. Lee, and S. C. Park, Phys. Lett. B 566,
142 (2003).
[7] A. G. Akeroyd, M. Aoki, and H. Sugiyama, Phys. Rev. D
77, 075010 (2008).
[8] J. Garayoa and T. Schwetz, J. High Energy Phys. 03
(2008) 009.
[9] M. Kadastik, M. Raidal, and L. Rebane, Phys. Rev. D 77,
115023 (2008).
[10] F. del Aguila and J. A. Aguilar-Saavedra, Nucl. Phys.
B813, 22 (2009).
[11] A. Aranda, J. Hernandez-Sanchez, and P. Q. Hung, J. High
Energy Phys. 11 (2008) 092.
[12] B. Pontecorvo, Sov. Phys. JETP 6, 429 (1957) [Zh. Eksp.
Teor. Fiz. 33, 549 (1957)]; Sov. Phys. JETP 7, 172 (1958)
[Zh. Eksp. Teor. Fiz. 34, 247 (1957)]; Sov. Phys. JETP 26,
984 (1968) [Zh. Eksp. Teor. Fiz. 53, 1717 (1967)]; Z.
Maki, M. Nakagawa, and S. Sakata, Prog. Theor. Phys. 28,
870 (1962).
[13] S. M. Bilenky, J. Hosek, and S. T. Petcov, Phys. Lett. 94B,
495 (1980).
[14] S. M. Bilenky and S. T. Petcov, Rev. Mod. Phys. 59, 671
(1987).
[15] A. Hektor, M. Kadastik, M. Muntel, M. Raidal, and L.
Rebane, Nucl. Phys. B787, 198 (2007); P. Fileviez Perez,
T. Han, G. y. Huang, T. Li, and K. Wang, Phys. Rev. D 78,
015018 (2008).
[16] J. Schechter and J. W. F. Valle, Phys. Rev. D 25, 774
(1982).
[17] M. A. Diaz, M. A. Garcia-Jareno, D. A. Restrepo, and
J. W. F. Valle, Nucl. Phys. B527, 44 (1998).
[18] M. C. Chen, Phys. Rev. D 71, 113010 (2005).
[19] T. Blank and W. Hollik, Nucl. Phys. B514, 113 (1998); M.
Czakon, M. Zralek, and J. Gluza, Nucl. Phys. B573, 57
(2000); M. Czakon, J. Gluza, F. Jegerlehner, and M.
Zralek, Eur. Phys. J. C 13, 275 (2000); J. R. Forshaw,
D. A. Ross, and B. E. White, J. High Energy Phys. 10
(2001) 007; M. C. Chen and S. Dawson, Phys. Rev. D 70,
015003 (2004);M. C. Chen, S. Dawson, and T. Krupov-

(Contract No. RII3-CT-2004-506222) and by World

Premier International Research Center Initiative (WPI
Initiative), MEXT, Japan. S. T. P. acknowledges with gratefulness the hospitality and support of IPMU, University of
Tokyo, where part of the work on the present article was
done.

[20]
[21]
[22]
[23]

[24]
[25]
[26]
[27]
[28]
[29]

[30]

[31]

[32]
[33]
[34]
[35]
[36]
[37]
[38]

015005-13

nickas, Int. J. Mod. Phys. A 21, 4045 (2006); Phys. Rev. D

74, 035001 (2006).
J. F. Gunion, J. Grifols, A. Mendez, B. Kayser, and F. I.
Olness, Phys. Rev. D 40, 1546 (1989).
F. Cuypers and S. Davidson, Eur. Phys. J. C 2, 503 (1998).
A. Merle and W. Rodejohann, Phys. Rev. D 73, 073012
(2006).
J. C. Pati and A. Salam, Phys. Rev. D 10, 275 (1974); 11,
703(E) (1975); R. N. Mohapatra and J. C. Pati, Phys. Rev.
D 11, 2558 (1975); G. Senjanovic and R. N. Mohapatra,
Phys. Rev. D 12, 1502 (1975).
R. N. Mohapatra and G. Senjanovic, Phys. Rev. Lett. 44,
912 (1980).
A. G. Akeroyd and M. Aoki, Phys. Rev. D 72, 035011
(2005).
S. M. Bilenky, S. Pascoli, and S. T. Petcov, Phys. Rev.
D64, 053010 (2001); 64, 113003 (2001).
S. T. Petcov, Nucl. Phys. B, Proc. Suppl. 143, 159 (2005).
J. Schechter and J. W. F. Valle, Phys. Rev. D 22, 2227
(1980); M. Doi et al., Phys. Lett. 102B, 323 (1981).
B. T. Cleveland et al., Astrophys. J. 496, 505 (1998); W.
Hampel et al. (GALLEX Collaboration), Phys. Lett. B
447, 127 (1999); J. N. Abdurashitov et al. (SAGE
Collaboration), Zh. Eksp. Teor. Fiz. 122, 211 (2002) [J.
Exp. Theor. Phys. 95, 181 (2002)]; J. Hosaka et al. (SuperKamkiokande Collaboration), Phys. Rev. D 73, 112001
(2006); B. Aharmim et al. (SNO Collaboration), Phys.
Rev. Lett. 101, 111301 (2008); C. Arpesella et al.
(Borexino Collaboration), Phys. Rev. Lett. 101, 091302
(2008).
Y. Ashie et al. (Super-Kamiokande Collaboration), Phys.
Rev. D 71, 112005 (2005); J. L. Raaf (Super-Kamiokande
Collaboration), J. Phys. Conf. Ser. 136, 022013 (2008).
M. H. Ahn et al. (K2K Collaboration), Phys. Rev. D 74,
072003 (2006); P. Adamson et al. (MINOS
Collaboration), Phys. Rev. Lett. 101, 131802 (2008).
M. Apollonio et al. (CHOOZ Collaboration), Eur. Phys. J.
C 27, 331 (2003).
S. Abe et al. (KamLAND Collaboration), Phys. Rev. Lett.
100, 221803 (2008).
A. Bandyopadhyay et al., Phys. Lett. B 608, 115 (2005).
G. L. Fogli et al., Phys. Rev. D 78, 033010 (2008).
T. Schwetz, M. Tortola, and J. W. F. Valle, New J. Phys.
10, 113011 (2008).
F. Ardellier et al. (Double Chooz Collaboration), arXiv:
hep-ex/0606025.
See, e.g., K. M. Heeger, 22nd International Conference on

S. T. PETCOV, H. SUGIYAMA, AND Y. TAKANISHI

[39]
[40]
[41]
[42]

[43]

[44]
[45]
[46]
[47]

[48]

[49]
[50]
[51]
[52]
[53]
[54]
[55]

PHYSICAL REVIEW D 80, 015005 (2009)

Neutrino Physics and Astrophysics (Neutrino 2006), Santa

Fe, New Mexico, 2006 (unpublished), and the Daya Bay
homepage http://dayawane.ihep.ac.cn/.
S.-B. Kim et al. (RENO Collaboration), J. Phys. Conf. Ser.
120, 052025 (2008).
K. Anderson et al., arXiv:hep-ex/0402041, and the references quoted therein.
A. Bandyopadhyay et al. (ISS Physics Working Group),
arXiv:0710.4947, and the references quoted therein.
Y. Itow et al. (T2K Collaboration), arXiv:hep-ex/0106019.
For an updated version, see http://neutrino.kek.jp/jhfnu/
loi/loi.v2.030528.pdf.
M. V. Chizhov, M. Maris, and S. T. Petcov, arXiv:hep-ph/
9810501; J. Bernabeu, S. Palomares-Ruiz, and S. T.
Petcov, Nucl. Phys. B669, 255 (2003); S. PalomaresRuiz and S. T. Petcov, Nucl. Phys. B712, 392 (2005).
S. T. Petcov, Phys. Lett. B 434, 321 (1998); 444, 584(E)
(1998).
M. V. Chizhov and S. T. Petcov, Phys. Rev. Lett. 83, 1096
(1999); Phys. Rev. D 63, 073003 (2001).
M. V. Chizhov and S. T. Petcov, Phys. Rev. Lett. 85, 3979
(2000).
V. K. Ermilova et al., Short Notices of the Lebedev
Institute 5, 26 (1986); E. K. Akhmedov, Sov. J. Nucl.
Phys. 47, 301 (1988) [Yad. Fiz. 47, 475 (1988)]; P. I.
Krastev and A. Yu. Smirnov, Phys. Lett. B 226, 341
(1989).
S. T. Petcov and M. Piai, Phys. Lett. B 533, 94 (2002); S.
Choubey, S. T. Petcov, and M. Piai, Phys. Rev. D 68,
113006 (2003).
J. Learned et al., Phys. Rev. D 78, 071302 (2008); M.
Batygov et al., arXiv:0810.2580.
L. Zhan et al., Phys. Rev. D 78, 111103 (2008); 79,
073007 (2009).
S. M. Bilenky, M. D. Mateev, and S. T. Petcov, Phys. Lett.
B 639, 312 (2006).
K. Eitel et al., Nucl. Phys. B, Proc. Suppl. 143, 197
(2005).
F. Perrin, Comptes Rendus 197, 868 (1933); E. Fermi,
Nuovo Cimento 11, 1 (1934).
V. Lobashev et al., Nucl. Phys. A719, C153 (2003).
D. N. Spergel et al. (WMAP Collaboration), Astrophys. J.

Suppl. Ser.. 148, 175 (2003).

[56] M. Tegmark, Phys. Scr. T121, 153 (2005); S. Hannestad,
H. Tu, and Y. Y. Y. Wong, J. Cosmol. Astropart. Phys. 06
(2006) 025.
[57] W. Hu and M. Tegmark, Astrophys. J. Lett. 514, L65
(1999).
[58] C. Aalseth et al., arXiv:hep-ph/0412300.
[59] S. T. Petcov, New J. Phys. 6, 109 (2004); Phys. Scr. T121,
94 (2005); S. Pascoli and S. T. Petcov, arXiv:hep-ph/
0308034.
[60] V. A. Rodin et al., Nucl. Phys. A766, 107 (2006); A793,
213(E) (2007); A. Poves, NDM06 International
Symposium, Paris, 2006 (unpublished); E. Caurier et al.,
Phys. Rev. Lett. 100, 052503 (2008); Eur. Phys. J. A 36,
195 (2008).
[61] H. V. Klapdor-Kleingrothaus et al., Nucl. Phys. B, Proc.
Suppl. 100, 309 (2001).
[62] C. E. Aalseth et al., Phys. At. Nucl. 63, 1225 (2000).
[63] H. V. Klapdor-Kleingrothaus et al., Phys. Lett. B 586, 198
(2004).
[64] H. V. Klapdor-Kleingrothaus et al., Mod. Phys. Lett. A 16,
2409 (2001).
[65] A. S. Barabash (NEMO Collaboration), arXiv:0807.2336.
[66] C. Arnaboldi et al. (CUORICINO Collaboration), Phys.
Rev. C 78, 035502 (2008).
[67] F. Avignone, Nucl. Phys. B, Proc. Suppl. 143, 233 (2005).
[68] S. Pascoli and S. T. Petcov, Phys. Lett. B 544, 239 (2002);
580, 280 (2004).
[69] S. Pascoli, S. T. Petcov, and L. Wolfenstein, Phys. Lett. B
524, 319 (2002); S. Pascoli and S. T. Petcov, Yad. Fiz. 66,
472 (2003) [Phys. At. Nucl. 66, 444 (2003)].
[70] S. Pascoli, S. T. Petcov, and T. Schwetz, Nucl. Phys. B734,
24 (2006); S. Pascoli and S. T. Petcov, Phys. Rev. D 77,
113003 (2008).
[71] J. Schechter and J. W. F. Valle, Phys. Rev. D 25, 2951
(1982).
[72] R. N. Mohapatra and J. D. Vergados, Phys. Rev. Lett. 47,
1713 (1981); W. C. Haxton, S. P. Rosen, and G. J.
Stephenson, Phys. Rev. D 26, 1805 (1982); L.
Wolfenstein, Phys. Rev. D 26, 2507 (1982).
[73] A. Zee, Nucl. Phys. B264, 99 (1986); K. S. Babu, Phys.
Lett. B 203, 132 (1988).

015005-14