Neutrinoless double beta decay with and without Majoron-like boson
emission in a 3-3-1 model
J. C. Montero1 ∗ , C. A. de S. Pires2 † , V. Pleitez1
‡
1
2
Instituto de Fı́sica Teórica
Universidade Estadual Paulista
Rua Pamplona 145
01405-900–São Paulo, SP
Brazil
Department of Physics, University of Maryland
College Park, MD 20742-4111, USA
(July 2, 2001)
We consider the contributions to the neutrinoless double beta decays in a SU (3)L ⊗ U (1)N electroweak model. We show that for a range of the parameters in the model there are diagrams
involving vector-vector-scalar and trilinear scalar couplings which can be potentially as contributing
as the light massive Majorana neutrino exchange one. We use these contributions to obtain constraints upon some mass scales of the model, like the masses of the new charged vector and scalar
bosons. We also consider briefly the decay in which besides the two electrons a Majoron-like boson
is emitted.
PACS numbers: 23.40.-s; 12.60. Cn; 14.60.St
mass is required for the decay to take place [8]. However,
this does not mean that the neutrino mass is necessarily
the main factor triggering this decay. In some models
the (ββ)0ν decay can proceed with arbitrary small neutrino mass via scalar boson exchange [9]. The mechanism involving a trilinear interaction of the scalar bosons
was proposed in Ref. [10] in the context of a model with
SU (2) ⊗ U (1) symmetry with doublets and a triplet of
scalar bosons. However, since in this type of models there
is no large mass scale, it was shown in Refs. [11] that the
contribution of the trilinear interactions are in fact negligible. In general, in models with that symmetry, a fine
tuning is needed if we want that the trilinear terms give
important contributions to the (ββ)0ν decay [8,12].
Here we will show that in a model with gauge symmetry SU (3)c ⊗SU (3)L ⊗U (1)N (3-3-1 model by short) [13],
which has a rich Higgs bosons sector as in the multi-Higgs
extensions of the standard model, there are new contributions to the (ββ)0ν decay. However, unlike the latest
sort of models, a fine tuning of the parameters of the 33-1 model is not necessary since some trilinear couplings,
which have mass dimension, could imply an enhancement
of the respective amplitudes (See. Sec. III).
We will use the following strategy: First, we consider
the several new contributions to the (ββ)0ν decay introduced by the 3-3-1 model. Next, once this decay has
not experimentally been seen, we will consider the usual
standard model amplitude (that would arise with massive Majorana neutrinos) as the reference one and make
the assumption that all the new amplitudes are at most
as contributing as this one. Hence, we can obtain constraints on some typical mass scale 3-3-1 parameters. The
new contributions to the (ββ)0ν decay are of the short
I. INTRODUCTION
The issue of neutrino masses continues to be a golden
plate in elementary particle physics. Although data coming from solar [1], atmospheric [2], and the accelerator LSND [3] neutrino experiments strongly suggest that
neutrinos must be massive particles, direct measurements
did not obtain any positive result [4].
It is a very well known fact that if neutrinos are massive Majorana particles it should exist the neutrinoless
double beta (ββ)0ν decay [5,6]. If the neutrino mass is
the main effect that triggers this decay, the decay lifetime
is proportional to (for the case of light neutrinos)
X
2
(1)
Uei
mνi ,
hMν i =
i
where Uei ; i = 1, 2, 3 denote the elements of a mixing
µ, τ and mass
matrix that relates symmetry να ; α = e, P
eigenstates νi through the relation να = i Uαi νi ; and
mνi are the neutrino masses. Experimentally a half life
0ν
> 1.8 × 1025 yr implies [7]
limit T1/2
hMν i < 0.2 eV.
(2)
The important point is that the (ββ)0ν decay probes
the physics beyond the standard model. In particular
the observation of this decay would be an evidence for a
massive Majorana neutrino although it could say nothing about the value of the mass. This is because although right-handed currents and/or scalar-bosons may
affect the decay rate, it has been shown that whatever
the mechanism of this decay is, a nonvanishing neutrino
1
+
and Vµ+ are related to the new mass eigenstates W1µ
and
+
as
W2µ
+
+
cθ s θ
W1
W
(6)
=
−sθ cθ
V+
W2+
range type [14]. Since the respective matrix elements are
different from those of the long range contributions (the
exchange of a light-Majorana neutrino) our results should
be considered only as an indication of the possible large
contributions to this decay in the context of the 3-3-1
model.
The outline of the paper is the following. In Sec. II
we introduce the interactions which are relevant to the
6 0, which is some
present study. The model with hσ10 i =
cases it has a Majoron-like Goldstone boson, is also discussed. In Sec. III we consider the more important contributions to the (ββ)0ν decay and the constraints upon
some masses of the model. In Sec. IV we show that if
we add a neutral scalar singlet to the minimal model
a Majoron-like Goldstone boson is consistent with the
Z 0 invisible width and we also discuss briefly the Majoron emission process (ββ)0νM comparing the relative
strength of two amplitudes. Our conclusions remain in
the last section.
2
−MV2 ). We can obtain an upper
with tan 2θ = −2δ/(MW
bound on δ by assuming that the main contribution to
2
mass is given by vσ2 ≈ 246 GeV and that vσ1
the MW
has its maximum value 3.89 GeV allowed by the value
of the ρ-parameter [16]. In fact if vσ2 were the main
2
≈
contribution to the MW mass we would have δ/MW
(2vσ1 /vσ2 ) < 0.032. The constraint on the mixing angle
θ is:
2
MV2 − MW
1
1
1−
< . (7)
0 ≤ s2θ =
2 )2 ]1/2
2
2
[4δ 2 + (MV2 − MW
Some illustrative values for sθ are obtained by using
typical values for the parameters. For instance, for vσ1 =
3.89 GeV; vσ2 = 10 GeV, MW = 80.41 GeV and MV =
100 (300) GeV we get s2θ = 1.9 × 10−5 (3.4 × 10−8 ); or if
MV = 100 GeV and if vσ2 has its maximal value vσ2 =
246 GeV we have s2θ = 1.1 × 10−2 . We see that only
for values of MV ≈ MW the s2θ is almost 0.5 but this
light vector boson may be not phenomenologically safe.
However if vσ1 is of the same order of magnitude of the
neutrino mass smaller values for the mixing angle are
obtained. Hence, it may be no relevant for the collider
physics and low energy processes like the (ββ)0ν decay
at all and in practice W1+ ≈ W + , W2+ ≈ V + ; but this
could not be the case in astrophysical processes [15].
Next, we consider the several interactions that are
present in this model. The scalar-quark interactions are
√
√
2
2
†
u−d
u
−
DL VCKM M UR ρ +
U L VCKM M d DR η1+
−LY =
|vρ |
|vη |
"√
#
√
2
2
T
η− −
ρ−
+ DL VLd △ VLu M u UR
|vη | 1
|vρ |
#
"√
√
2 +
2 +
u T
d
d
ρ −
η
+ U L (VL ) △ VL M DR
|vρ |
|vη | 1
II. THE MODEL
Here we will consider the 3-3-1 model with the leptons
belonging to triplets (νl l lc )T ; l = e, µ, τ and in which a
sextet of scalar bosons
h−
h+
√1
σ10 √22
2
h−
σ0
√2 H −−
√2 ∼ (6, 0) .
S=
(3)
1
2
2
h+
σ20
++
1
√
√
H2
2
2
is necessary to give to the charged leptons a mass if
hσ20 i ≡ vσ2 6= 0 [13].
Most of the phenomenological studies of the model has
been done by considering hσ10 i ≡ vσ1 = 0. The case when
6 0 it was considered in Ref. [15], where the other
hσ10 i =
scalar multiplets are explicitly given. The main difference
in the latter case with respect to the former one is that
there is a mixing between the vector bosons W + and V + :
2
−
MW
Wµ
δ
+
+
Wµ Vµ
,
(4)
Vµ−
δ MV2
2
2
MW
+ H.c.,
MV2
and
are the mass
where δ = (g /2)(2vσ1 vσ2 ),
eigenvalues when δ = 0; if δ 6= 0 (i.e. when vσ1 6= 0) the
mass of the physical fields are now
2
1/2
2
+ MV2 ) ± (MW
− MV2 )2 + 4δ 2
2M1,2 = (MW
(8)
with
0 0 0
∆ ≡ 0 0 0 ,
0 0 1
(5)
(9)
and VLu,d are unitary mixing matrices, and M u,d are the
diagonal mass matrices of the u-like and d-like quark
sectors, and VCKM denotes the usual mixing matrix of
Cabibbo-Kobayashi-Maskawa.
The Yukawa interactions in the lepton sector are
2
= (g 2 /2)(vη2 + vρ2 + vσ22 + vσ21 ),
and we have defined MW
2
2
2
2
2
MV = (g /2)(vη +vχ +vσ2 +vσ21 ) and g is the SU (3)L coupling constant which is numerically
√ equal to the coupling
2
GF / 2. We have denoted by
of SU (2)L i.e., g 2 = 8MW
vη , vρ and vχ the vacuum expectation values of the neutral components of the triplets. Notice that M1 → MW
and M2 → MV when δ → 0. The vector bosons Wµ+
1
1
c
H2−
−LlY = √ νL K1 lR H1+ + √ l̄L K2 νR
2
2
2
1
1
lL K3 (lL )c H1−− + (lc )L K4 lR H2++
2
2
c −
η2 + H.c.,
+ 2 νL K1′ lR η1+ − 2 lL K2′ νR
There are also trilinear interactions involving two
vector- and one scalar-bosons (the χ− scalar couples to
ordinary and exotic quarks and for this reason it is not
of our concern here). They are given by (up to a g 2 /2
factor)
+
(10)
where K1 = ELν† GER , K2 = EL† GELν∗ ; K3 = EL† GEL ,
T
GER ; K1′ = ELν† G′ ER , K2′ = EL† G′ ELν∗ ; G and
K4 = ER
′
G are symmetric and antisymmetric (they can be complex) matrices, respectively. ER , EL , ELν are the rightand left-handed mixing unitary matrices in the lepton
sector relating symmetry eigenstates (primed fields) with
mass eigenstates (unprimed fields) [17]:
′
= ER lR ,
lR
′
lL
= EL lL ,
νL′ = ELν νL ,
vσ
vσ1 ++ −
H1 W · W − + 1 H2++ V − · V −
2
2
vσ2
++
++
(H1++ + H2++ )
+ vρ ρ + vχ χ
+
2
−
−
(15)
W ·V .
L2V S =
Notice that there is a coupling which is proportional to
vχ and hence it will be the dominant one.
Next, we write down the trilinear interactions among
three vector bosons
g
+ ν+ ν−−
+ ν+ µ−−
V U
+ Vµν
V U
L3V = i √ Wµν
2
−−
+ W ν+ V µ+ Uµν
(16)
(11)
Some of the couplings in Eq. (10) do not depend on the
charged lepton masses and since all matrices in Eq. (10)
are not unitary, the model breaks the lepton universality
but it can be shown that, for the massless neutrino case
no strong constraints arise from exotic muon and tau
decays [18].
In the scalar sector we have also mixing
In the
P angles.
−
=
O
H
,
where
singly charged sector we have φ−
ij
i
j
l
− − −
−
−
−
−
φ−
=
η
,
η
,
ρ
,
χ
,
h
,
h
and
H
,
j
=
1,
...,
6
denotes
1
2
1
2
i
j
the respective mass eigenstate field; similarly
in the douP
−−
O
Ψ
=
bly charged sector we have Φ−−
ij
j , with
i
l
−−
−−
−−
−−
Ψi = ρ−− , χ−− , H1 , H2 and Ψj , j = 1, .., 4 the
respective mass eigenstate fields. However, in the following we will use H − and H −− as typical mass eigenstates
of the respective charged fields and omit the scalar mixing parameters.
We recall that the model conserves the F = L + B
quantum number; L is the total lepton number and B is
the baryon number. The assignments are:
where Wµν = ∂µ Wν − ∂ν Wµ and so on.
Finally, we have trilinear interactions among scalarbosons only
L3S =
f1 ijk
f2 T †
ǫ ηi ρj χk +
χ S ρ + H.c..
2
2
(17)
The couplings f1,2 have dimension of mass but are both
of them arbitrary parameters (see next section). Other
terms like the trilinears f3 η T S † η and f4 ǫSSS and the
quartic interactions ǫχ(Sη ∗ )ρ, χ† ηρ† η and ǫχρSS violate
the conservation of F .
However, as we will show in Sec. IV, when discussing
the Majoron emission, the model must be modified by
adding a scalar singlet in order to be consistent with the
LEP data.
F (U −− ) = F (V − ) = F (ρ−− ) = F (χ−− ) = F (η2− )
−−
= F (χ− ) = F (σ10 ) = F (h−
2 ) = F (H1 )
−−
(12)
= F (H2 ) = 2,
and all other scalar fields with F = 0.
The charged currents coupled to the vector bosons are
given by
g
LCC = − √ U L γ u VCKM DL Wµ+ − ν̄L γ µ VW lL Wµ+
2
c γ µ V V † ν V + − l c γ µ V l U ++ + H.c, (13)
+ lL
U W L µ
U L µ
L
III. THE NEUTRINOLESS DOUBLE BETA
DECAY
Some of the more relevant diagrams of the (ββ)0ν decay in the present model are shown in Figs. 1-6. Our goal
is to analyze the order of magnitude of each diagram and
to obtain constraints on some mass scales of the model.
We will consider the diagram in Fig. 1 as the reference
one, i.e., it is the diagram that already exist in the standard model framework with massive Majorana neutrinos
and which is parameterized by two effective four-fermion
interactions. The other contributions will be considered
as been at most equally important than the standard one.
The strength of the diagram in Fig. 1 is given by
†
with the mixing matrices defined as VCKM = (VLu ) VLd in
T ν
EL in the
the quark sector; and VW = ELν † EL , VU = ER
leptonic sectors.
We have the trilinear interactions involving one vectorand√two scalar-bosons which are of the form (up to a
ig/ 2 factor):
−−
+
LV 2S = ∂ µ χ+ χ−− Wµ+ + χ− ∂ µ χ++ Wµ− + ∂ µ h+
2 H1 Wµ
A(1) ∝
−
µ − ++ −
+ ∂ µ H1++ h−
Vµ + ∂ µ ρ−− ρ+ Vµ+
2 Wµ + ∂ ρ ρ
32G2F hMν i 4
g 4 hMν i 4
cθ ,
cθ =
4
2
MW hp i
hp2 i
(18)
+ µ +
µ −− +
µ − ++ −
+ h+
where hMν i is the effective mass defined in Eq. (1)
1 ∂ H2 Vµ + ∂ h1 H2 Vµ + η2 ∂ η1
2
+ µ +
+ µ +
+ µ +
−−
+ η1 ∂ η2 + h2 ∂ h1 + h1 ∂ h2 Uµ + H.c. . (14) and hp i is the average of the four-momentum transfer
3
uL
dL
W
R
eL
L
eL
We recall that a lower limit of 440 GeV is obtained for
MV from the muon decays but when only the bilepton
contributions to those decays are considered [19]. However, in the minimal 3-3-1 model the scalar-boson contributions cannot be negligible since some of the charged
scalar-bosons can be lighter than the vector bilepton boson V − . Hence, a lighter vector boson V may still be
possible but this subject deserves a more detailed study
of the muon decay considering both vector and scalar
contributions. A contribution similar to that in Fig. 1
but with two V − bosons instead of two W − bosons may
be not negligible but it does not constraint the mass MV
as much as those in Eq. (21) since the condition that its
ratio to the A(1) amplitude
be less than one gives the
√
condition MV > MW tan θ.
All the Lagrangian interactions in Eqs. (8), (10), (13),
(14), (15) and (16) are written in terms of symmetry
eigenstates. We have assumed Yukawa couplings of the
order of unity. As we are not considering the mixing
among the scalar fields our constraints are valid only for
the main component of the symmetry eigenstate scalar
fields. It means that H − and H −− denote the dominant
mass eigenstates of the singly and doubly charged scalar
fields, respectively.
The amplitude of the diagram in Fig. 3 is
W
uL
dL
FIG. 1. (ββ)0ν decay via light massive Majorana neutrinos.
squared, which is of the order of (100 MeV)2 . Below we
will use a small δ so that M1 → MW and M2 → MV .
In Eq. (18) and hereafter we will omit for simplicity the
mixing parameters. Only in the vertices we will take care
about the mixing between W and V defined in Eq. (6)
but in the propagator we will use the masses of W and
V.
A(3) ∝
uL
dL
eL
V
eR
L
eR
H
uL
dL
H
R
eR
(22)
uR
dL
W
R
hMν i
4 .
hp2 iMH
−
dL
uR
FIG. 2. The same as in Fig. 1 but with one bilepton vector
boson V − replaced by a vector boson W − .
FIG. 3. Charged scalars contribution to the (ββ)0ν decay.
Next, let us consider the diagram in Fig. 2 which has
the strength given by
The scalar contribution in Fig. 3 can be as important
as the standard one in Fig. 1. We have
A(2) ∝ 32G2F
MW
MV
2
c3 sθ
pθ
,
hp2 i
and we have the ratio
2 p 2
hp i
MW
A(2)
=
tan θ,
A(1)
MV
hMν i
1
A(3)
=
4 c4 ,
A(1)
32G2F MH
− θ
(19)
(23)
and assuming that A(3)/A(1) < 1 and cθ = 1 we get
MH − > 124 GeV.
(20)
(24)
From Eq. (15) we see that the contribution
vχ χ++ Wµ− V µ− is the dominant one in diagrams like that
in Fig. 4. As we said before we will omit the mixing angles, i.e., assuming χ−− ≈ H −− . Hence we have
and if A(2)/A(1) < 1 we have that
√
√
MV > 2.2 × 104 MW tanθ = 1.79 × 106 tanθ GeV.
(21)
A(4) ∝
4
vχ
c2θ s2θ .
2 M2 M2
MW
V
H −−
(25)
uL
dL
V
If A(6)/A(1) < 1, and assuming cθ = 1 and MH − =
124 GeV, we obtain the constraint
eR
H
W
dL
f
2
MH
−−
eR
uL
Next, we note that
vχ
hp2 i
A(4)
∝
tan2 θ,
2
2
2
A(1)
hMν i 32GF MW MV2 MH
−−
(1GeV)4
,
2
MV2 MH
−−
(26)
where we used hMν i = 0.2 eV [4], vχ = 3 TeV and hp2 i =
(100 MeV)2 . If A(4)/A(1) < 1 it implies
(1GeV)2
.
MV > 7.3 × 10 tan θ
MH −−
7
(27)
W
dL
H
dL
uL
V
uL
FIG. 5. The same as in Fig. 4 but only involving vector
bosons.
More interesting are the contributions involving trilinear scalar interactions given in Eq. (17) like that of the
diagram in Fig. 6. We have in this case
eR
uR
IV. MAJORON EMISSION
If the F quantum number is spontaneously broken as
in the present model, it means that a Majoron-like bosons
does exist. Since the scalar field that is responsible for
the breakdown of that continuous symmetry is σ10 , and
it belongs to a triplet of the subgroup SU (2) ⊗ U (1),
this Majoron-like Goldstone has similar couplings that
the triplet Majoron model of Ref. [21]. It is well known
that this sort of Majoron model has been ruled out by
the LEP data [22]. Apparently, since the Higgs sector of
the present model is rather complicated having a neutral
scalar singlet (under SU (2) ⊗ U (1)), χ0 , it seems that
the Majoron-like Goldstone in this case it will be able
to avoid the LEP constraints as claimed in Refs. [15,23].
However, we will show that this is not indeed the case.
(28)
where MH − represents a typical mass of the singly
charged scalar bosons, say 124 GeV; MH −− is the mass
of the doubly charged scalar boson and f is the trilinear
coupling f1 or f2 in Eq. (17) with dimension of mass.
The ratio of these amplitudes is:
f hp2 i
A(6)
∝
4 M2
4
A(1)
32G2F MH
−
H −− hMν icθ
!
f /GeV
4.8 × 107
·
,
≈
2
2
c4θ
MH
−− /GeV
H
There is also a diagram in which the doubly charged
scalar field in Fig. 6 is substituted by a vector boson
U −− . Although the interactions in Eq. (14) are proportional to g they are p
also derivative and proportional to
the momentum p ∼ hp2 i; hence it is suppressed with
respect to the diagram in Fig. 6.
eL
f
,
A(6) ∝ 4
2
MH − MH
−−
eR
FIG. 6. Pure scalar Higgs bosons contribution the (ββ)0ν
decay in the 3-3-1 model.
eR
U
uR
dL
H
Similar analysis arises by considering Fig. 5, however it
is less enhanced than the contribution of Fig. 4 because
the momentum of one of the
instead of vχ it appears
p
2
vector bosons, p ∼ hp i.
dL
(30)
For arbitrary U (1)N charge for the scalar multiplets
the symmetry of the potential is SU (3)L ⊗ [U (1)]2 . If
the triplet η and the sextet S have both N = 0, as it
is the case for the present model, the trilinear couplings
f1,2 break the extra U (1) symmetry. We have verify that
if both f1,2 = 0 there is indeed a pseudo-Goldstone boson [20]. It means that f1,2 are arbitrary parameters and
in principle they can be small (say 1 GeV), or large (say
1 TeV) mass scales.
We see that if f = 1 (10−3 ) TeV then MH −− is greater
or of the order of 300 (10) TeV. For this value for the mass
of the doubly charged scalar field and θ small the constraint given in Eq. (21) is stronger than that of Eq. (27).
For instance if tan θ = 10−8 we have from Eq. (21) that
MV ≥ 179 GeV.
FIG. 4. New contribution to the (ββ)0ν decay in the 3-3-1
model.
≈ 5.33 × 1015 tan2 θ
< 2.1 × 10−8 GeV−1 .
(29)
5
V (Xi , Σ) = µ25 Σ2 + λ21 Σ4 +
The mass matrices of the scalar and pseudoscalar in
this model have been given in Ref. [15]. Here we will
only give the results of the mass eigenvalues and the respective mixing matrix in the CP-even scalar sector. The
argument in Ref. [15] was
following. Let us begin
P the
o
Hj0 , where Hj0 , j = 1, ...5,
with the relation R40 = j O4j
denotes the mass scalar eigenstates and R4 the real component of the scalar field σ10 according to the general
shifting of the neutral√
scalar fields in the scalar potential
of the form Xi0 → (1/ 2)(vXi + Ri + iIi ), i = 1, 2, 3, 4, 5
where Xi0 = η, ρ, χ, σ1 , σ2 respectively. In this case if
H10 denotes the lightest scalar boson (MH1 < MZ , we
are assuming a mass spectrum where MHi < MHj if
i < j), the contribution to the decay mode Z 0 → H10 M 0
o
o 2 Z
| < 10−2 the
| Γν ν̄ . Hence, if |O41
is ΓZ
= 2|O41
H10 M 0
model would be consistent with the LEP data i.e., now
would be reduced to an acceptable level.
ΓZ
H10 M 0
First of all recall that as shown in Ref. [15] the
Majoron-like boson decouples from the other pseudoscalar fields i.e., Imσ10 ≡ I4 = M 0 . For instance, using
the same values of the VEV and f1,2 = −1 TeV and with
the dimensionless constant of the scalar potential given in
Ref. [15] with λk = 0.1 for k = 1, 2, 3, 4, 5, 6, 7, 8, 9, 18, 20,
λm = 0.01 for m = 13, 14, 16, 17, λn = 0.001 for
n = 10, 11, 12, 19 and λ15 = 0.05 we obtain the following
masses in the scalar sector (0.056, 102, 1342, 3626, 4325)
GeV and the mixing matrix (up to three decimal places)
0.0 0.081 −0.010 0.996 0.021
0.0 0.995 −0.029 −0.082 0.039
Oo = 0.0 −0.030 −0.999 −0.008 0.004 (31)
1.0 0.0
0.0
0.0
0.0
0.0 0.040 −0.005 0.017 −0.999
Xh
λXi T r (Xi† Xi )Σ2
i
− κ η T S † ηΣ + H.c.
(32)
where Xi denotes any triplet η, ρ, χ or the sextet S, and
we will denote λXi as λ22,23,24,25 respectively and κ > 0.
The neutral Higgs sector contains six CP-even scalars and
three massive CP-odd pseudoscalar beside the massless
CP-odd Majoron. The neutral scalar
singlet also gains a
√
VEV, i.e., Σ = (vΣ + R6 + iI6 )/ 2, and the mass term is
given by M 2 /2, where M 2 in the pseudoscalar sector in
the basis I1 , I2 , I3 , I4 , I5 , I6 is given by (the constraints
equation appear in the Appendix)
λ16 vρ2 vσ2
1
+ 2λ17 vσ22 + √ (λ15 vχ vσ2
M11 = − √
2 2 vη
2 2
tη
vχ
+ 2κvσ1 vΣ + ,
− f1 vρ )
vη
vη
√
1
vχ
tρ
+ ,
M22 = − ( 2f1 vη + f2 vσ2 )
4
vρ
vρ
vρ
tχ
1 √
+ ,
M33 = − ( 2f1 vη + f2 vσ2 )
4
vχ
vχ
vη2 vΣ
tσ
+ 1,
M44 = κ
2vσ1
vσ1
f2 vρ vχ
vη
1
+ 2λ17 vη2 −
M55 = √ (λ15 vχ2 − λ16 vρ2 )
vσ2
4 vσ2
2 2
2
v
v
tΣ
tσ
η σ1
+
,
+ 2 , M66 = κ
vσ2
2vΣ
vΣ
f1
f1
M12 = − √ vχ , M13 = − √ vρ , M14 = −κvη vΣ ,
2 2
2 2
1
M15 = √ (λ15 vχ2 − λ16 vρ2 ) + 2λ17 vη vσ2 , M16 = κvη vσ1 ,
2 2
1 √
f2
M23 = − ( 2f1 vη + f2 vσ2 ), M24 = 0, M25 = vχ ,
4
4
f2
M26 = 0, M34 = 0, M35 = vρ , M36 = 0,
4
vη2
(33)
M45 = 0, M46 = −κ , M56 = 0.
2
This pattern of mixing matrix remains the same for several values of the parameters provided that vσ1 is a small
VEV restricted to the condition that it has to be smaller
than 3.89 GeV [16]. From Eq. (31) it can be seen that
the scalar partner of the Majoron is always mainly the
o
| ∼ 1 and it would be always prolightest scalar i.e., |O41
duced at LEP. We see that the Majoron in the minimal
3-3-1 model has been also ruled out by the LEP data.
One possibility to recover consistency with the LEP
data is to break explicitly the F symmetry by adding
trilinear terms like f3 η T Sη, f4 χT S † ρ in the scalar potential, see Eq. (17). In this case there is no Majoron at
all and although vσ1 still has a small value, due to f3 all
scalars are heavy enough to not be produced at the LEP
energies [24,25]. Of course, in this case there is no contribution to Majoron emission in the neutrinoless double
beta decay. However, our results in Sec. III are still valid
since they depend only on the small value of vσ1 .
Another possibility which we will consider here is to
modify the model by introducing a scalar singlet Σ0 ,
which carries F = 2 (or L = 2), in the same way as considered in Ref. [26,27] in the context of a SU (2) ⊗ U (1)
model. In this case we have to add the following terms
to the scalar potential in Ref. [15]
The mass matrix above has two true Goldstone bosons
G01,2 and the Majoron-like one, M 0 , and three massive
CP-odd pseudoscalar bosons. The massless ones are
given by
G01 = (0, vρ /vχ , −1, 0, 0, 0)/(1 + vρ2 /vχ2 )1/2
G02 = (
2
vρ2 vχ 2vσ1 vΣ
vρ vχ2
2v 2 vΣ
vη
,−
,
, −1, − σ1 )/N
,−
vσ2
V1
V1
V2
V2
2 1/2
)
M 0 = (0, 0, 0, vσ1 /vΣ , 0, 1)/(1 + vσ21 /vΣ
(34)
2
); N is the
where V1 = vσ2 (vρ2 + vχ2 ), V2 = vσ2 (vσ21 + vΣ
normalization factor that we will omit here. We have verify that M 0 in Eq. (34) is in fact the Majoron: by adding
an explicit F -violating term, like f3 η T Sη, it gets a mass
6
with f can be f1 or f2 .
Let us consider the ratio
while the other two G1,2 remain massless. The massive
pseudoscalars, for the parameters used before have the
following masses (in GeV): 174, 3625 and 4325. On the
other hand, if vσ1 = 0, which forces κ = 0, the Majoron
is purely singlet and the real and imaginary parts of σ10
are mass degenerate, i.e., form a complex field, with mass
mσ1 = µ24 + λ10 vσ22 + (λ12 + λ19 )
vρ2
v2
+ λ14 + λ25 Σ .
2
2
vη2
2
+ λ13
mν Q
B(7)
∝
4 hM ic4 v ,
A(1)
32G2F MX
ν θ Σ
where we have introduced the factor Q which denotes the
available energy. It implies that the diagram in Fig. 7
is a potentially important contribution when X is the
W vector boson since for Q ∼ 3 MeV [30] the suppression of B(7) will depend mainly on the value of vΣ . If
B(7)/A(1) < 1 we obtain that vΣ > 1.65 × 10−2 GeV
which is automatically satisfied.
vχ2
2
(35)
We see that in this case the Majoron has not doublet components at all and it is mainly singlet. Hence
it is possible to keep consistence with LEP data. Although there are astrophysical constraints (the Majoron
emission implies a different rate for the stellar cooling)
that have to be taken into account [28], in the basis we
have chosen they are less severe since we have avoided the
doublet component of the Majoron. Any way, since these
constraints have been already considered in Ref. [27] and
they imply that vσ1 < 0.33 GeV if vΣ = 1 TeV, we will
use these values for vσ1 , vΣ .
Once we have shown in what situation there is a safe
Majoron-like boson in the present model we can consider
the emission of this Goldstone boson in the neutrinoless
double beta decay. In fact, as in the triplet Majoron
model, in the present model it is possible to have the
neutrinoless double beta decay with Majoron emission:
2n → 2p + 2e− + M 0 [29], denoted here by (ββ)0νM . We
will denote the strength of the amplitude of the diagram
i of the (ββ)0νM decay by B(i). This decay proceeds via
the diagram in Fig. 7 and it has an strength proportional
to
B(7) ∝
mν (vσ1 /vΣ )
4 hp2 iv ,
MX
σ1
−
f f2
,
6 M2
MH
−
H −−
u
d
X
e
e
M0
X
u
d
FIG. 7. Contribution to the Majoron emission (ββ)0νM decay. X − can be a scalar or vector boson.
On the other hand, comparing the amplitudes of the
diagrams in Figs. 7 and 8 we have
4
f f2 hp2 iMX
vΣ
B(8)
∝ 6
,
2
B(7)
MH − MH −− mν
(39)
and for MX = MV = 400 GeV and vΣ = 1 TeV, using
typical values as f = f1 = f2 = −1 TeV, MH − = 124
GeV and MH −− = 500 GeV, and the other parameters
in Eq. (39), we have that B(8)/B(7) ≈ 1.3 × 109 or;
B(8)/B(7) ≈ 1.5 × 103 if f1 = f2 = f = −10−3 TeV. The
relative importance of the processes in Fig. 7 and 8 will
depend on the values of the trilinear parameters and on
the value of vΣ .
(36)
where X − can be an scalar or a vector boson, i.e., the
diagram in Fig. 7 can be formed with anyone of Figs.1,
2 or 3 with a Majoron attached to the neutrinos. The
couplings between neutrinos and the Majoron are diagonal and given by mν /vσ1 . Notice that in Eq. (36) it
appears the truly neutrino mass instead of the effective
mass hMν i defined in Eq. (1). However we still can assume that neutrinos have small masses and numerically
mν ≈ hMν i. We will assume also that the contribution
to the (ββ)0νM -decay in Fig. 7 with X − = W − is the reference one. This diagram depends only on the neutrino
masses and mixing angles, and we will compare it with
other contributions like the one in Fig. 8. The couplings
of the Majoron to the vector bosons are proportional to
vσ1 and so they are negligible. We will consider only the
diagram with the Majoron coupled to the scalar H − since
it is proportional to the trilinear f2 shown in Eq. (17).
We have
B(8) ∝
(38)
uR
dL
H
H
dL
M0
eR
eR
H
uR
FIG. 8. Trilinear scalar coupling contributing to the
(ββ)0νM decay.
V. CONCLUSIONS
We see that in the 3-3-1 model, like in other models with complicated Higgs sector [8,12], besides the well
known mechanism of exchanging massive Majorana neutrinos between two standard model V − A vertices, there
(37)
7
box of Ref. [8] may induce a negligible Majorana mass to
the neutrinos and in the context of the present model we
must interpret this situation as an indication of the fact
that the main contribution to the (ββ)0ν decay is not the
diagram in Fig. 1. In this case the neutrinos would be
almost Dirac particles and the constraints on the several
mass scales of the model should be obtained by comparing directly these contributions with the lower bound on
0ν
> 1.8 × 1025 yr [7].
the half life T1/2
iii) In the basis we have chosen, see Eq. (34), the Majoron couples at the tree level only to neutrinos, and
hence the constraints on the ν-ν-M vertex, coming from
muon (µ → eννM ), pion and kaon π + (K + ) → lν̄M
decays, are the same as in Ref. [33]. The existence of
the vertex νe− V + , which is proportional to sin θ and for
this reason is not relevant for laboratory processes, may
have, as we said before, important astrophysical consequences [15].
iv) phenomenology of non-zero initial electric charge
processes, like e− e− and hadronic ones, will furnish constraints on the trilinear vertices appearing in Figs. 4, 5
and 6 but this will be considered elsewhere. The 3-3-1
model has a rich scalar sector indeed. This implies that it
may be, in principle, difficult to separate, in a given process, the contributions of all fields belonging to a charged
sector. However, it has been shown that in lepton-lepton
colliders the left-right asymmetries are not sensible to the
scalar contributions. It means that those asymmetries
are the appropriate observable for the doubly charged
vector bilepton discovery [36]. We see that the opposite
occurs in the (ββ)0ν decay: it is possible that the main
contribution comes from the doubly charged scalar boson, through the diagram in Fig. 6, while the respective
vector boson contribution seems to be negligible.
Finally we would like to compare our Majoron model
with that of Schechter and Valle [26]. Firstly, we notice
that although our model has two singlet, three doublets
and a triplet of scalars under the subgroup SU (2)⊗U (1),
the respective scalar potential is not reduced to the scalar
potential invariant under the standard SU (2)⊗U (1) symmetry, involving the same multiplets. For instance in our
model there are cubic invariants which are not present
in the former. Secondly, we have not introduced righthanded neutrinos and for this reason we have only light
neutrinos. It means that the singlet Σ does not couple to
the leptons and that the coupling of neutrinos to Majoron
and Z 0 are diagonal. Thus, the decays νH → νL + M 0
and νH → νL + νL′ + νL′ , are not induced at the tree level,
where νH , although light, is heavier that νL . The decay
νH → νLc + M 0 is produced at the one loop level due
to the mixing√between W and V . The vertex is proportional to (g 2 / 2)(vσ2 vσ1 /vΣ ); then even with vσ2 of the
order of 10 GeV and vΣ of the order of 1 TeV the lifetime is of the order of the age of the universe. (The decay
νH → νL + M 0 also
√ occurs but the vertex involved is proportional to (g 2 / 2)(vσ21 /vΣ ).) Notice also that in the
basis given in Eq. (34) the Majoron does not couple to the
are new contributions involving the exchange of scalar
bosons. However, unlike similar mechanism in the context of extensions of the standard model there is no need
of fine tuning in order to have trilinear scalar couplings
giving large contributions to the several neutrinoless double beta decay modes. Notice that effective interactions
from diagrams like those in Fig. 3 are still parameterized
in the form of two general four-fermion effective interactions (they are point-like at the Fermi scale) exchanging
a light neutrino in between [31]. However, contributions
involving trilinear interactions like those in Figs. 4, 5 and
6 necessarily need a six-fermion effective interaction parameterization.
Another important point to be stressed here is that in
the present model the double Majoron emission: 2d →
2p + 2e− + 2M 0 may be as contributing as the decay
with only one Majoron boson. This decay is expected to
be important in supersymmetric models [32,33]. In the
present model it can occur because in diagrams like that
in Fig. 8 a second Majoron can be attached to the scalar
lines. Since this coupling is proportional to the trilinear
f2 it is still possible that the suppression coming from
the mass square in the denominator do not sufficiently
suppresses this process (there is also an important contribution coming from the vertex vχ χ− η1+ M 0 ). There are
also contributions similar to the one in Fig. 8 but now
with the scalar-Majoron vertex above being substituted
by the vertex W − V + M 0 which is proportional to g 2 vσ2
and for this reason it is not necessarily suppressed. It is
interesting to note that this sort of contribution to the
(ββ)0νMM decay, coming from adding another trilinear
coupling in the diagram in Fig. 8, which is not derivatively suppressed, was not considered in Ref. [33]. Recent
experimental data on Majoron emission decays have been
constrained only the effective Majoron-neutrino coupling
constant [30]. Other processes like the double K capture [29] can also be important in the present model.
Some comments are now in order. i) We have not
considered possible cancellations among several contributions to each diagram. It means that our constraints
are valid, as we said in Sec. III, for the main component
of each scalar field of the singly and doubly charged scalar
sectors.
ii) Our results were obtained assuming that all new
contributions, say to the (ββ)0ν decay, are at most as
important as the contribution due to a light massive Majorana neutrino exchange which is proportional to hMν i.
However, we can wonder what would be the value of the
effective mass hMν i if we use the oscillation data and direct measurements on neutrino masses. Recent analysis
shown that assuming a normal mass hierarchy the effective mass parameter can take any value from zero to the
present upper limit [34]. In fact, if the data on oscillation is put together with that from (ββ)0ν decay and
tritium beta decay, it was shown that if the minimum of
hMν i with respect to the mixing angles is greater than
the present bound of 0.2 eV, then neutrinos are quasiDirac particles [35]. As discussed previously, the black
8
charged leptons so there is not the process γ +e → M 0 +e
at the tree level which imposes severe astrophysical constraints in vσ1 as has been noted in Ref. [27].
[1] K. Lande (Homestake Collaboration), in Neutrinos ’98,
Proceedings of the XVIII International Conference on
Neutrino Physics and Astrophysics, Japan, 4-9 June
1998, edited by Y. Suzuki and T. Tosuka, to be published in Nucl. Phys. B (Proc. Suppl.); Y. Fukuda et al.
(Kamiokande Collaboration), Phys. Rev. Lett. 77, 1683
ACKNOWLEDGMENTS
(1996); T. Kirsten (GALLEX Collaboration) in Neutrinos ’98; V. Gavrin (SAGE Collaboration) in Neutrinos
This work was supported by Fundação de Amparo à
’98; Y. Suzuki (SuperKamiokande Collaboration) in NeuPesquisa do Estado de São Paulo (FAPESP), Conselho
trinos ’98.
Nacional de Ciência e Tecnologia (CNPq) and by Pro[2] Y. Fukuda et al. (SuperKamiokande Collaboration ),
grama de Apoio a Núcleos de Excelência (PRONEX).
Phys. Lett. B433, 9 (1998); Phys. Rev. Lett. 81, 1562
(1998); Phys. Lett. B436, 33 (1988).
[3] C. Athanassopoulos et al., Phys. Rev. Lett. 77, 3082
APPENDIX A: CONTRAINTS EQUATION OF
(1996); ibid, 81, 1774 (1998).
THE SCALAR POTENTIAL
[4] D. E. Groom et al., The European Physical Journal,
C15, 1 (2000).
[5] H. Primakov and S. P. Rosen, Rep. Prog. Phys. 22, 111
Here we show the constraints equation that must be
(1969).
satisfied by the scalar potential
[6] For a recent review see H. V. Klapdor-Kleingrothaus, Int.
λ4
λ5
λ12 2
J. Mod. Phys. A13, 3953 (1998); P. Fisher, B. Kayser
(v + vσ22 )vη
tη = µ21 vη + λ1 vη3 + vρ2 vη + vχ2 vη +
and K. S. McFarland, Ann. Rev. Nucl. Part. Sci. 49
2
2
2 σ1
(1999); hep-ph/9906244. For early references see M. Doi,
λ
λ16 2
λ15 2
19 2
vσ1 vη
− √ vχ vσ2 + √ vρ vσ2 − λ17 vσ22 vη +
T. Kotani and E. Takasugi, Progr. Theor. Phys. Suppl.
2
2 2
2 2
83, 1 (1985).
f1
[7]
L.
Baudis et al., Phys. Rev. Lett. 83, 41 (1999); hep− κvη vσ1 vΣ + √ vρ vχ ,
2 2
ex/9902014.
[8] J. Schechter and J. W. F. Valle, Phys. Rev. D 25, 2951
λ6
λ14 2
λ4
(v + vσ22 )vρ
tρ = µ22 vρ + λ2 vρ3 + vη2 vρ + vχ2 vρ +
(1982).
2
2
2 σ1
[9]
V.
Pleitez and M. D. Tonasse, Phys. Rev. D 48, 527
2
λ20 2
λ16
vρ vΣ
(1993).
vσ2 vρ + λ23
+ √ vσ2 vη vρ +
4
2
2
[10] R. N. Mohapatra and J. D. Vergados, Phys. Rev. Lett.
f2
f1
47, 1713 (1981).
+ √ vη vχ + vσ2 vχ ,
[11] W. C. Haxton, S. P. Rosen and G. J. Stephenso, Jr.,
4
2 2
Phys. Rev. D 26 1805 (1982); L. Wolfenstein, ibid, D 26,
λ6
λ13 2
λ5
(vσ1 + vσ22 )vχ
tχ = µ23 vχ + λ3 vχ3 + vη2 vχ + vρ2 vχ +
2507 (1982).
2
2
2
[12]
C. O. Escobar and V. Pleitez, Phys. Rev. D 28, 1166
2
λ15
λ18 2
vχ vΣ
(1983).
v vχ + λ24
− √ vσ2 vη vχ +
4 σ2
2
2
[13] F. Pisano and V. Pleitez, Phys. Rev. D 46, 410 (1992);
P. Frampton, Phys. Rev. Lett. 69, 2889 (1992); R. Foot,
f2
f1
+ √ vη vρ + vσ2 vρ ,
O. F. Hernandez, F. Pisano and V. Pleitez, Phys. Rev.
4
2 2
D 47, 4158 (1993).
λ12 2
[14] M. Hirsch, H. V. Kkapdov-Kleingrothaus and S. G. Kovη vσ1
tσ1 = µ24 vσ1 + λ10 (vσ21 + vσ22 )vσ1 + λ11 vσ31 +
2
valenko, Phys. Rev. Lett. 75, 17 (1995); Phys. Rev. D
λ14 2
λ19 2
λ25
κ 2
λ13 2
53, 1329 (1996); H. V. Klapdor-Kleingrothaus and H.
2
v vσ +
v vσ +
v vσ +
vσ1 vΣ − vη vΣ ,
+
Päs, hep-ph/0005045.
2 χ 1
2 ρ 1
2 η 1
2
2
[15] J. C. Montero, C. A. de S. Pires and V. Pleitez, Phys.
λ12 2
λ11 3
2
2
2
v +
v vσ
tσ2 = µ4 vσ2 + λ10 (vσ2 + vσ1 )vσ2 +
Rev. D 60, 115003 (1999); hep-ph/9903251.
2 σ2
2 η 2
[16]
J. C. Montero, C. A. de S. Pires and V. Pleitez, Phy.
λ13 2
λ14 2
λ15
λ16
vχ vσ2 +
vρ vσ2 − √ vχ2 vη + √ vρ2 vη − λ17 vη2 vσ2
+
Rev. D 60, 98701 (1999); hep-ph/9902448.
2
2
2 2
2 2
[17] J. T. Liu and D. Ng, Phys. Rev. D 50, 548 (1994).
λ20 2
λ25
λ18 2
f2
2
[18] M. M. Guzzo et al., hep-ph/9908308.
vχ vσ2 +
vρ vσ2 +
vσ2 vΣ
+
+ vρ vχ ,
4
4
2
4
[19] M. B. Tully and G. C. Joshi, Int. J. Mod. Phys. A13,
λ23 2
λ24 2
λ22 2
5593 (1998).
2
3
v vΣ +
v vΣ +
v vΣ
tΣ = µ5 vΣ + λ21 vΣ +
[20] S. Weinberg, Phys. Rev. Lett. 29, 1698 (1972).
2 η
2 ρ
2 χ
[21] G. B. Gelmini and M. Roncadelli, Phys. Lett. 99B, 411
κ
λ25 2
(vσ1 + vσ22 )vΣ − vη2 vσ .
+
(1981).
2
2
[22] M. C. Gonzalez-Garcia and Y. Nir, Phys. Lett. 232, 383
(1989).
[23] F. Pisano and S. Shelly Sharma, Phys. Rev D 57, 5670
9
(1998).
[24] J. C. Montero, C. A. de Sousa Pires and V. Pleitez, Phys.
Lett. B502, 167 (2001); hep-ph/0011296.
[25] E. Ma, Phys. Rev. Lett. 81, 1171 (1998).
[26] J. Schechter and J. W. F. Valle, Phys. Rev. D 25, 774
(1982).
[27] M. A. Diaz, M. A. Garcia-Jareno, D. A. Restrepo and J.
F. W. Valle, Nucl. Phys. B527, 44 (1998).
[28] D. A. Dicus, E. W. Kolb, V. L. Teplitz and R Wagoner,
Phys. Rev. D 18, 1829 (1978). R. N. Mohapatra and P.
B. Pal, Massive Neutrinos in Physics and Astrophysics,
World Scientific, Singapore, 1991; and references therein.
[29] H. Georgi, S. L. Glashow and S. Nussinov, Nucl. Phys.
B193, 297 (1981).
[30] J. Bockholt et al., Phys. Rev. D 54, 3641 (1996).
[31] H. Päs, M. Hirsch, H. V. Klapdor-Kleingrothaus and S.
G. Kovalenko, Phys. Lett. B453, 194 (1999).
[32] R. N. Mohapatra and E. Takasugi, Phys. Lett. 211B,
192 (1988).
[33] P. Bamert, C. Burguess and R. N. Mohapatra, Nucl.
Phys. B449, 25 (1995).
[34] W. Rodejohann, Phys. Rev. D62, 013011 (2000); H.
V. Klapdor-Kliengrothaus, H. Päs and A. Yu. Smirnov,
Phys. Rev. D 63, 073005 (2001).
[35] M. Czakon, J. Gluza and M. Zralek, hep-ph/0003161.
[36] J. C. Montero, V. Pleitez and M. C. Rodriguez, Phys.
Rev. D 58, 094026 (1998); Phys. Rev. D 58, 097505
(1998); and Int. J. Mod. Phys. A16, 1147 (2001).
10