SU (3) U (1) Model For Electroweak Interactions: IFT-P.008/92
SU (3) U (1) Model For Electroweak Interactions: IFT-P.008/92
008/92
hep-ph/9206242 SU (3) U (1) Model for Electroweak Interactions
F. Pisano and V. Pleitez
Instituto de Fsica Te
orica
Universidade Estadual Paulista
Rua Pamplona, 145
CEP 01405S
ao Paulo, SP
Brazil
PACS numbers:
I. INTRODUCTION
Some years ago, it was pointed out that processes like e e W V in Fig. 1(a), if
induced by right-handed currents coupled to the vector V , imply violation of unitarity at
high energies. Then, if the right-handed currents are part of a gauge theory, it has been
argued that at least some neutrinos must have non-zero mass [1].
The argument to justify this follows exactly the same way as in the usual electroweak
theory for the process W + W . The graph induced by an electron exchange has bad
high-energy behavior; when the energy goes to infinity, the respective amplitude violates
unitarity [2].
In Fig. 1(a) the lower vertex indicates a right-handed current which absorbs the righthanded antineutrino coming from the upper vertex, which represents the left-handed current
of the electroweak standard model. The part of the amplitude, corresponding to Fig. 1(a),
in which we are interested is
X
L
Uem
q2
q/
UR ,
M2m em
(1)
where U L (U R ) is the mixing matrix in the left(right)-handed current, and q is the 4momentum transfer [1]. The space-time structure of Eq. (1) is the same as the charged
lepton exchange amplitude in the process W + W [2]. Then, we must have the same
bad high energy behavior of the last process. One way to avoid this is to have cancellation
among the contributions from the various m exchanges when we add them up; at high
energy and large q 2 , the latter dominates the denominator in Eq. (1), and if we require that
X
L
R
Uem
Uem
= 0,
(2)
the amplitude in Eq. (1) vanishes even at low energies, unless at least one of the masses
Mm is non-zero. On the other hand, the diagram in Fig. 1(a) or its time reversed one,
W W e e , appearing in Fig. 1(b), when both vertices are left-handed, proceeds via
Majorana massive neutrinos.
2
Here we are concerned with a gauge model based on a SUL (3) UN (1) symmetry. The
original motivation leading to the study of this model stemmed from the observation that a
gauge theory must be consistent, that is, unitary and renormalizable, independently of the
values of some parameters, like mixing angles. Then, from this point of view, instead of the
condition in Eq. (2) in order to solve the problem placed by the graph in Fig. 1(a), we prefer
the introduction of a doubly charged gauge boson which, like the Z 0 in the standard model,
will restore the good high energy behavior.
Although there exist in the literature several models based on a SU(3) U(1) gauge
symmetry [37], our model has a different representation content and a quite different new
physics at an, in principle, arbitrary mass scale. The main new features of our model occur
in processes in which the initial electric charge is not zero. Even from the theoretical point
of view, that sort of processes have not been well studied, for instance, general results exist
only for zero initial charge [8].
The plan of this paper is as follows: Sec. II is devoted to present the model. Some
phenomenological consequences are given in Sec. IV. In this way we can estimate the allowed
value for the mass scale characterizing the new physics. In Sec. III we study briefly the scalar
potential and show that there is not mixing between the lepton-number-conserving and
lepton-number-violating scalar fields which could induce decays like the neutrinoless double
beta decay. The last section is devoted to our conclusions and some comments and, in the
Appendix we give more details about the definition of the charge conjugation operation we
have used in this work.
As we said before, the gauge model that we shall consider is one in which the gauge
group is SUL (3) UN (1). This is possibly the simplest way to enlarge the gauge group
SUL (2) UY (1) in order to have doubly charged gauge bosons, without losing the natural
features of the standard electroweak model. The price we must pay is the introduction of
3
A. Yukawa Interactions
We start by choosing the following triplet representations for the left-handed fields of
the first family,
EL =
ec
and
u
d
J1
(3, + 32 ),
(3)
(4)
for the respective right-handed fields. Notice that we have not introduced right-handed
neutrinos. The numbers 0, 2/3 in Eq. (3) and 2/3, 1/3 and 5/3 in Eq. (4) are UN (1)
charges. The electric charge operator has been defined as
1
Q
3 38 + N,
=
e
2
(5)
where 3 and 8 are the usual Gell-Mann matrices; N is proportional to the unit matrix.
Then, the exotic quark J1 , has electric charge +5/3.
The other two lepton generations also belong to triplet representations,
ML =
(3, 0); TL =
L
(3, 0).
L
(6)
The model is anomaly free if we have equal number of triplets and antitriplets, counting
the color of SU(3)c , and furthermore requiring the sum of all fermion charges to vanish. As
in the model of Ref. [3], the anomaly cancellation occurs for the three generations together
and not generation by generation.
Then, we must introduce the antitriplets:
Q2L =
J2
c
(3
, 13 );
Q3L =
J3
(3 , 13 ),
(7)
also with the respective right-handed fields in singlets. The quarks J2 and J3 have both
charge 4/3.
In order to generate fermion masses, we introduce the following Higgs triplets, , and
:
1
2+
(3, 0);
0
++
(3, 1);
(3, 1),
(8)
These Higgs triplets will produce the following hierarchical symmetry breaking
<>
<,>
(9)
The Yukawa Lagrangian, without considering the mixed terms between quarks, is
LY =
1X
1L (Gu uR + Gd dR + GJ1 J1R )
Gl ijk lic lj k + Q
2 l
2L cR + Gt Q
3L tR + Gs Q
2L sR + Gb Q
3L bR
+ Gc Q
2L J2R + GJ3 Q
3L J3R + h.c.
+ GJ2 Q
(10)
X
l
c c
c c
c
c
Gl (lR
lLlR lL ) 0 (
lR
lLlR lL )1 + (
lR
lLlR
lL )2+ + h.c.,
(11)
and using the definition of charge conjugation, c = 5 C T , that we shall discuss in the
Appendix, we can write Eq. (11) as
5
LlY =
Gl (lR lL 0 + lR L 1 + Rc lL 2+ + h.c.).
(12)
In Eq. (12) there is lepton number violation through the coupling with the 2+ Higgs scalar.
For the first and second quark generations we have the following Yukawa interactions
LQY = Gu (
uL uR 0 + dLuR 1 + J1L uR 2+ )
+ Gd (
uL dR + + dLdR 0 + J1L dR ++ )
+ Gc (J2L cR + cL cR 0 + sL cR )
+ Gs (J2L sR 2 + cL sR 1+ + sL sR 0 )
+ GJ1 (
uL J1R + dL J1R + J1L J1R 0 )
+ GJ2 (J2L J2R 0 + cL J2R ++ + sL J2R + ) + h.c.
(13)
The Yukawa interactions for the third quark generation is obtained from those of the second
generation making c t, s b and J2 J3 . In Eq. (10), since the neutrinos are
massless there is not mixing between leptons, it is not necessary at all to consider terms like
1
2
n,m
c
hlm niL
mjL H [ij] + h.c. where n, m = e, , , the coupling constants hnm = hmn
< 0 >=
1
2
< 0 >=
1
2
< 0 >=
1
2
0
0
(14)
So, the masses of the fermions are ml = Gl v2 , for the charged leptons and
mu = Gu v2 , mc =
Gc v2 , mt = Gt v2 ,
md = Gd v2 , ms =
Gs v2 , mb = Gb v2 ,
(15)
v
v
mJ1 = GJ1 v2 , mJ2 = GJ2
, mJ3 = GJ3
,
2
2
for the quarks. The exotic quarks obtain their masses from the -triplet. Notice that, if we
had had introduced right-handed neutrinos, we would have massive Dirac neutrinos through
their couplings with the Higgs triplet.
6
The gauge bosons of this theory consist of an octet Wa associated with SUL (3) and a
singlet B associated with UN (1). The covariant derivatives are:
~
~ )ji j + ig N i B ,
D i = i + ig(W
2
(16)
where N denotes the N charge for the Higgs multiplet, = , , . Using Eqs. (14) in
Eq. (16) we obtain the symmetry breaking pattern appearing in Eq. (9).
Notice that even if v = v v/ 2, being v the usual vacuum expectation value of the
Higgs in the standard model, the v must be large enough in order to keep the new gauge
bosons, V + and U ++ , sufficiently heavy in order to have consistence with low energy phenomenology. On the other hand, the neutral gauge bosons have the following mass matrix
in the (W 3 , W 8 , B) basis
1
M 2 = g2
4
v2
1 (v 2
3
v2
v2 )
2 gg v2
1 (v 2
3
1 2
(v
3
v2 )
+ v2 + 4v2 )
2 g (v 2
3 g
+ 2v2 )
2 gg v2
2 g (v 2
3 g
2
+ 2v2 )
4 gg2 (v2 + v2 )
(18)
and, since det M 2 = 0 we must have a photon after the symmetry breaking. If we had had
introduced a 6 , the matrix M 2 in Eq. (18) would be such that det M 2 6= 0. In fact, the
eingenvalues of the matrix in Eq. (18) are:
MA2
= 0,
MZ2
g 2 g 2 + 4g 2 2
(v + v2 ),
2
2
4 g + 3g
1
MZ2 (g 2 + 3g 2 )v2 ,
3
(19)
where we have used v v, for the case of MZ and MZ . Notice that the Z 0 boson has
a mass proportional to v and, like the charged bosons V + , U ++ , must be very massive. In
the present model we have
7
MZ2
1 + 4t2
=
2
MW
1 + 3t2
(20)
2
where t = g /g tan , and in order to obtain the usual relation cos2 W MZ2 = MW
, with
cos2 W 0.78, we must have 54o i.e., tan2 11/6. Then, we can identified Z 0 as the
neutral gauge boson of the standard model.
The neutral physical states are:
1
(W3
3W8 )t + B ,
1
2
2
(1 + 4t )
2
"
#
3t
t
1
2 12
8
3
0
(1 + 3t ) W +
Z
1
1 W
1 B ,
(1 + 4t2 ) 2
(1 + 3t2 ) 2
(1 + 3t2 ) 2
1
8
Z0
3tB
W
+
.
1
(1 + 3t2 ) 2
A =
(21)
The interactions among the gauge bosons and fermions are read off from
~ )L,
( + ig B N)R + Li
( + ig B N + ig ~ W
LF = Ri
2
(22)
g sin
(1 + 3 sin2 )
1
2
g cos
1
(1 + 3 sin2 ) 2
(23)
+
c
+
c
++
l
W
+
l
V
+
l
l
U
+
h.c.
.
LCC
=
lL
L
lL
L
L
L
l
2 l
(24)
Notice that as we have not assigned to the gauge bosons a lepton number, we have explicit
breakdown of this quantum number induced by the V + , U ++ gauge bosons. A similar
8
mechanism for lepton number violation was proposed in Ref. [9] but in that reference the
lepton-number-violating currents are coupled to the standard gauge bosons and they are
proportional to a small parameter appearing in this model.
For the first generation of quarks we have the following charged current interactions:
g
LCC
uL dL W+ + J1L uL V+ + dL J1L U + h.c. ,
Q1 W =
2
(25)
LCC
=
d
W
J
V
+
c
J
U
+
h.c.
.
L
L
L
2L
L
2L
Q2 W
2
(26)
The charge changing interactions for the third generation of quarks are obtained from those
of the second generation, making c t, s b and J2 J3 . We have mixing only in
the Q= 31 and Q= 43 sectors, then in Eqs. (25) and (26) d , s and J2 mean CabibboKobayashi-Maskawa states in the three and two-dimensional flavor space d, s, b and J2 , J3
respectively.
D. Neutral Currents
Similarly, we have the neutral currents coupled to both Z 0 and Z 0 massive vector bosons,
according to the Lagrangian
C
LN
=
1
g MZ
1
Z ],
lL lL [Z q
2 MW
3 h(t)
(27)
g MZ
[l (vl + al 5 )lZ + l (vl + al 5 )lZ ],
4 MW
(28)
for the charged leptons, where we have used lLc lLc = lR lR and defined
vl = 1/h(t),
al =
1,
(a)
(29)
As it was said before, the quark representations in Eqs. (3) and (7) are symmetry eigenstates, that is, they are related to the mass eigenstates by Cabibbo-like angles. As we have
one triplet and two antitriplets, it should be expected to exist flavor changing neutral currents. Notwithstanding, as we shall show below, when we calculate the neutral currents
explicitly, we find that all of them, for the same charge sector, have equal factors and the
GIM [2] cancellation is automatic in neutral currents coupled to Z 0 . Remind that in the
standard electroweak model, the GIM mechanism is a consequence of having each charge
sector the same coupling with Z 0 , for example for the charge +2/3 sector,
U
vSM
=1
8 2
sin W ,
3
aUSM = 1.
(30)
g MZ X i
[i (v + ai 5 )i ]Z ,
4 MW i
(31)
aU = 1, (a)
v D = (3 + 8t2 )/3h(t),
aD =
J1
J1
= 20t /3h(t),
v J2 = v J3 = 16t2 /3h(t),
1, (b)
0, (c)
aJ2 = aJ3 =
0, (d)
(32)
U, and D mean the charge +2/3 and 1/3 respectively, the same for J1,2,3 . Notice that as it
was said above, there is not flavor changing neutral current coupled to the Z 0 field and that
the exotic quarks couple to Z 0 only through vector currents. It is easy to verify that for the
Q = 23 , 13 sectors the respective coefficients v and a also coincide with those of the standard
electroweak model if t2 = 11/6, as required to maintain the relation cos W MZ = MW .
The same cancellation does not happen with the corresponding currents coupled to the
Z 0 boson, each quark having its respective coefficients. Explicitly, we have
LZ Q =
g MZ X i
[i (v + ai 5 )i ]Z ,
4 MW i
where
10
(33)
v u = (1 + 8t2 )/ 3h(t),
q
v c = v t = (1 2t2 )/ 3h(t),
2
au = 1/ 3h(t),
v = (1 + 2t )/ 3h(t),
q
v s = v b =
h(t)/3,
(a)
q
a = a ,
(34)
(c)
as = ab = au ,
(d)
17t2
2
,
3
h(t)
1+3t
aJ1 = 23
, (a)
(b)
h(t)
15t
, aJ2 = aJ3 = aJ1 ,
v J2 = v J3 = 23
h(t)
(35)
The most general gauge invariant potential involving the three Higgs triplets is
V (, , ) = 21 + 22 + 23 + 1 ( )2 + 2 ( )2 + 3 ( )2
+ ( )[4 + 5 ] + 6 ( )( )
+
ijk (f i j k + h.c.).
(36)
ijk
the coupling f has dimesion of mass. We can analyse the scalar spectrum defining
0 = v1 + H1 + ih1 ,
0 = v2 + H2 + ih2 ,
0 = v3 + H3 + ih3 ,
(37)
23 v32
5 v12
6 v22
Re f v1 v2 v31
= 0,
Im f = 0.
11
(38)
Then, it is possible to verify that there is a doubly charged Goldstone boson and a doubly
charged physical scalar. There are also two singly charged Goldstone bosons
1
2
2
G
1 = (v1 2 + v3 )/(v1 + v3 ) 2 ,
1
(39)
2
2
G
2 = (v1 1 + v2 )/(v1 + v2 ) 2 ,
(40)
In this model, the lepton number is violated in the heavy charged vector bosons exchange
but it is not in the neutral exchange ones, because neutral interactions are diagonal in the
lepton sector. However, we have flavor changing neutral currents in the quark sector coupled
to the heavy neutral vector boson Z 0 . All these heavy bosons have a mass which depends
on v and this vacuum expectation value is, in principle, arbitrary.
Processes like e e are the typical ones, involving leptons, which are induced by
lepton-number-violating charged currents in the present model. It is well known that the
ratio
R=
( e e )
( all)
12
(41)
tests the nature of the lepton family number conservation, i.e., additive vs. multiplicative.
Roughly we have
A(3.a)
R
A(3.b)
MW
MV
4
where A(3.a) and A(3.b) are the amplitudes for the processes in Fig. 3(a) and (b) respectively.
Experimentally R < 5 102 [11], then we have that the occurrence of the decay
e e implies that MV > 2MW .
events impose constraints on the masses of the vector bosons which are weaker than those
coming from the decays. Notice that the incoming negative charged lepton is right-handed
because the lepton-number-violating interactions with the V + vector boson in Eq. (24) is a
right-handed current for the electron.
The doubly charged vector boson U will produce deviations from the pure QED Moller
scattering which could be detected at high energies.
Stronger bounds on the masses of the exotic vector bosons come from flavor changing
neutral currents induced by Z 0 . The contribution to the KL0 KS0 mass difference due to the
exchange of a heavy neutral boson Z 0 appears in Fig. 4. From Eq. (33) we have explicitly
g MZ
(v d + ad 5 )s + d
(v s + as 5 )s]Z 0 ,
cos c sin c [d
4 MW
(42)
with v d,s and ad,s given in Eq. (34c,d) respectively, and for simplicity we have assumed only
two-family mixing. Then, Eq. (42) produces at low energies the effective interaction,
Lef f
g 2 MZ
=
16 MW
2
i2
cos2 c sin2 c h
5
d
(c
+
c
)s
,
v
a
MZ2 0
(43)
cv v d v s = 23 (1 + 3t2 )/ h(t),
d
ca a a = cv .
13
(44)
2
with g 2 /8MW
m2c
GF
1 (1 5 )s]2 ,
(45)
LSM
=
cos2 c sin2 c [d
ef f
2
2
2
2 4 MW sin W
as usual, that any additional contribution to the KS0 KL0 mass difference from the Z 0 boson
cannot be much bigger than the contribution of the charmed quark [13]. Then, from Eqs. (43)
and (45) we get
MZ2 0
>
2
1 4 2 MW
2
,
ca 2 tan2 W MW
2
mc
(46)
3
2
v2 >
(40 T eV )2 ,
2
8GF MW (1 + 3t2 )
that is, v > 12 T eV . As the vacuum expectation value of the Higgs is <0 >= v / 2
then we have that <0> > 8.4 T eV . This also implies, from Eq. (17), that the masses of the
charged vector bosons V , U are larger than 4 T eV .
V. CONCLUSIONS
If we admit lepton number violation, SU(3) could be a good symmetry at high energies,
at least for the lightest leptons (, e , e+ ). Assuming that this is a local gauge symmetry,
the rest of the model follows naturally, including the exotic quarks, Js. To the best of our
knowledge, there is not laboratory or cosmological/astrophysical constraints to the masses
of the exotic quarks (Josions) J1 and J2,3 with charge + 53 and 43 respectively but, they
must be too massive to be detected by present accelerators. For the case of the heavy vector
bosons, charged U, V (Wanios) and the neutral Z 0 (Zezeons), rare decays restraint their
masses as we have shown before. It is interesting to note that no extremely high mass scale
emerges in this model making possible its experimental test in future accelerators.
14
(47)
and also with , , when these two new triplets are added to the model. Then we
have mixing in the scalar sector which imply 1-loop contributions to the ()0 involving
the vector bosons V , U but these are less than contributions at tree level through scalar
exchange [10]. On the other hand, this model cannot produce processes like K + e
and l+ with l = e, .
Notice that the definition of the charge conjugation transformation we have used in this
work, see the Appendix, has physical consequences only in the Yukawa interactions and in
the currents coupled to the heavy charged gauge bosons where an opposite sign appears
with respect to the usual definition of that transformation.
ACKNOWLEDGMENTS
We would like to thank the Conselho Nacional de Desenvolvimento Cientfico e Tecnologico (CNPq) for full (F.P.) and partial (V.P.) financial support, M.C. Tijero for reading
the manuscript and, finally C.O. Escobar, M. Guzzo and A.A. Natale for useful discussion.
In this appendix we shall treat in more detail how it is possible to get a Yukawa interactions from Eq. (11).
In the present model we have put together in the same multiplet the charged leptons
and their respective charge conjugated field. That is, both of them are considered as the
two independent fermion degrees of freedom. If we use the usual definition of the charge
conjugation transformation c = C T , c = T C 1 the Yukawa couplings in Eq. (11)
vanish, including the mass terms. This is a consequence of the degrees of freedom we have
chosen. Notwithstanding, it is possible to define the charge conjugation operation as follows:
15
c = 5 C T , c = T C 1 5 .
This definition is consistent with quantum electrodynamics since its only effect is to change
the sign of the mass term in the Dirac equation for the charge conjugated spinor c with
respect to the mass term of the spinor , and it is well known that the sign of the mass term in
the Dirac equation has no physical meaning. With the negative sign, the upper components
of the spinor are the large ones, and with the positive sign, the large components are the
lower ones [14].
c c
c c
lL = lR lL instead of lR
lL = +lR lL , which
Using this definition it is easy to verify that lR
(lL i6 lL + c i6 lLc ),
X
l
where the right-handed electron has been interpreted as the absence of a left-handed positron
with (E, ~p).
For charged leptons we have the electromagnetic interaction
e(lL lL lLc lLc )A ,
and using lLc lLc = lR lR we obtain the usual vector interaction el lA , but on the
c
other hand, in the charged currents we have lR
lL = lR lL .
16
REFERENCES
[1] B. Kayser, F. Gibrat-Debu and F. Perrier, The Physics of Massive Neutrinos, World
Scientific, 1989.
[2] C. Quigg, Gauge Theories of the Strong, Weak, and Electromagnetic Interactions, The
Benjamin/Cummings, 1983.
[3] M. Singer, J. W. F. Valle and J. Schechter, Phys. Rev.D22, 738(1980).
[4] J. Schechter and Y. Ueda, Phys. Rev. D8, 484(1973).
[5] P. Langacker and G. Segr`e, Phys. Rev. Lett. 39, 259(1977).
[6] H. Fritzsch and P. Minkowski, Phys. Lett. B63, 99(1976).
[7] B. W. Lee and S. Weinberg, Phys. Rev. Lett. 38, 1237(1977).
[8] J. M. Cornwall, D. N. Levin and G. Tiktopoulos, Phys. Rev. D10, 1145 (1974).
[9] J.W.F. Valle and M. Singer, Phys. Rev. D28, 540(1983).
[10] F. Pisano and V. Pleitez, Neutrinoless Double Beta Decay With Massless Neutrinos,
Preprint IFT-P.07/92, submited for publication.
[11] Particle Data Group, Phys. Lett. B239, 1(1990).
[12] M.K. Gaillard and B.W. Lee, Phys. Rev. D10, 897(1974); R. Shrock and S.B. Treiman,
Phy. Rev. D19, 2148(1979).
[13] R.N. Cahn and H. Harari, Nucl. Phys. B176, 135(1980).
[14] J. Tiomno, Nuovo Cimento 1, 226(1955).
17
FIGURES
FIG. 1.
the handedness of the current at the vertex, and q is the momentum transfer. (b) Diagram for
W W e e with massive Majorana neutrinos, both vertices are left-handed.
FIG. 2. Diagram for W V e e due to the existence of right-handed current (a) and
doubly charged gauge boson (b).
FIG. 3. (a) Lepton number conserving process. (b) Lepton number violating process.
18