THE HIGGS SECTOR IN THE MINIMAL 3-3-1 MODEL WITH THE

MOST GENERAL LEPTON-NUMBER CONSERVING POTENTIAL

arXiv:hep-ph/0011201v1 16 Nov 2000

Nguyen Tuan Anh a , Nguyen Anh Ky a,b

and Hoang Ngoc Long a,c

a
Institute of Physics, NCNST, P. O. Box 429, Bo Ho, Hanoi 10000, Vietnam

b
Department of Physics, Chuo University, Bunkyo-ku, Tokyo 112-8551, Japan

c
LAPTH, B.P. 110, F–74941, Annecy–le–Vieux Cedex, France

Abstract

The Higgs sector of the minimal 3 - 3 - 1 model with three triplets and

one sextet is investigated in detail under the most general lepton–number

conserving potential. The mass spectra and multiplet decompostion structure

are explicitly given in a systematic order and a transparent way allowing they

to be easily checked and used in further investigations. A previously arising

problem of inconsistent signs of f2 is also automatically solved.

PACS numbers: 11.15.Ex, 12.60.Fr, 14.80.Cp.

1
1 Introduction

The standard model (SM) combining the Glashow–Weinberg–Salam (GWS) model

with the QCD under the gauge group SUC (3)⊗SUL (2)⊗UN (1) is one of the greatest

achievements of physics in the 20th century. Many predictions of the SM have been

confirmed by various experiments. However, this model works well only at the

energy range below 200 GeV and gradually losses its prediction power at higher

energies. Therefore, any extension of the SM to fit the theory with the higher

energy phenomenology is needed. In addition, the observation of the Higgs bosons

which play a central role in symmetry breaking is still an open problem. Hence

the mechanism for electroweak symmetry breaking is, in some way, still a mystery.

The scalar sector has been throughly studied not only in the framework of the SM

but also in its various extensions including the so–called 3 - 3 - 1 models based

on the SUC (3) ⊗ SUL (3) ⊗ UN (1) gauge group [1–6]. The later models have the

following intriguing features: firstly, the models are anomaly free only if the number

of families N is a multiple of three. Further, the condition of QCD asymptotic

freedom valid only for the number of quark families less than five, leads to N equal

to 3. Secondly, the Peccei–Quinn (PQ) [7] symmetry – a solution of the strong

CP problem naturally occurs in the 3 - 3 - 1 models [8]. It is worth mentioning

that the implementation of the PQ symmetry is usually possible only at a classical

level (broken by quantum corrections through instanton effects) and there has been

a number of attempts to find models solving the strong CP question. In these 3 - 3 -

2
1 models the PQ symmetry following from the gauge invariant Lagrangian does not

have to be imposed. The third interesting feature is that one of the quark families

is treated differently from the other two. This could lead to a natural explanation

for the unbalancing heavy top quarks, deviations of Ab from the SM prediction, etc.

Additionally, the models predict not very high new mass scales, at the order of a

few TeV only [9].

Recently, the scalar sector of the minimal 3 - 3 - 1 model was in detail studied in

[10] and [11]. There, three Higgs triplets were firstly analysed and then the sextet

was added in a further consideration. It was also shown in [11] that the potential

used in [10] leaded to inconsistent results and therefore it should be further modified

or replaced by some more relevant potential. Another precise investigation on the

model with a new potential is, in our opinion, really interesting and necessary. Fol-

lowing the previous paper [11] the present paper is devoted to such an investigation.

Here, instead of the potential in [10, 11] the most general gauge–invariant postential

conserving lepton numbers [12] is used. With the latter potential the scalar sector

of the 3 - 3 - 1 model is investigated again at tree–level. The multiplet decomposi-

tion structure remains the same as in [11] but the masses of most of the scalars get

corrections and the problem with inconsistent signs of f2 is automatically solved.

We emphasize that the above mentioned potential was also considered in [12, 13],

but only mass matrices [12] and some their eigenvalues [13] were presented.

The paper is organized as follows. The Higgs potentials, constraint equations

3
and main notations are presented in section 2, while the characteristic equations

are solved in section 3 where the multiplet decompositions and mass spectra of the

Higgs sector are given. Our results are summarized in the last section, section 4.

2 The Higgs contents and scalar potentials

Presently, the minimal 3 - 3 - 1 models are considered with three Higgs triplets [1]
     

 ηo 


 ρ+ 


 χ− 

     
     
η= 
 η1− 
 ∼ (1, 3, 0), ρ = 
 ρo 
 ∼ (1, 3, 1), χ = 
 χ−− 
 ∼ (1, 3, −1),
     
     
     
η2+ ρ++ χo
(2.1)

and one Higgs sextet [2]

 
√ √

 σ1o s+
2
/ 2 s−
1
/ 2 

√ √ 
 
∗

S= 
 +
s2 / 2 s1 ++ o
σ2 / 2  ∼ (1, 6 , 0). (2.2)
 
√ √
 
 
s−
1
/ 2 σ2o / 2 s−−
2

The latter is needed in order to give masses to all leptons. In [10, 11], the scalar

sector of the minimal 3 - 3 - 1 models is investigated by using the potential

VS (η, ρ, χ, S) = VT (η, ρ, χ) + µ24 T r(S † S) + λ10 T r 2 (S † S) + λ11 T r[(S † S)2 ]

 
+[λ12 η † η + λ13 ρ† ρ + λ14 χ† χ]T r(S † S) + 2f2 ρT Sχ + h.c. ,

with

VT (η, ρ, χ) = µ21 η † η + µ22 ρ† ρ + µ23 χ† χ + λ1 (η † η)2 + λ2 (ρ† ρ)2 + λ3 (χ† χ)2

4
 +λ4 (η † η)(ρ† ρ) + λ5 (χ† χ)(η † η) + λ6 (ρ† ρ)(χ† χ) + λ7 (ρ† η)(η † ρ)
√  
+λ8 (χ† η)(η † χ) + λ9 (ρ† χ)(χ† ρ) + 2f1 εijk ηi ρj χk + h.c. , (2.3)

where µ’s are mass parameters and the coefficients f1 and f2 have dimensions of

mass, while λ’s are dimensionless. Unfortunately, as shown in [11], this potential

(which is not most general) leads to nonlogical results as f2 cannot take a consistent

sign. Analysing the results obtained, we conclude that we need a wider potential

in order to give necessary corrections to those masses showing contradict signs of

f2 . On the other hand, we suggest that the potential needed should guarantee the

conservation of the lepton numbers and the continuous symmetry not higher than

SU(3) × U(1). Fortunately, there exist such potentials. The most general potential

satisfying our requirements is [12]

VE (η, ρ, χ, S) = VS (η, ρ, χ, S) + λ15 η † S † Sη + 4λ16 ρ† S † Sρ + 4λ17 χ† S † Sχ

√ √
+2 2λ18 ρ† S † ρη + 2 2λ19 χ† S † χη + λ20 S † S † ηη + h.c. (2.4)

where
ρ† S † ρη = εijk ρ†l S li† ρj η k ,

χ† S † χη = εijk χ†l S li† χj η k ,

S † S † ηη = εijk εlmn η k η n S il† S jm† .

The triplet Higgs fields η o , ρo ,and χo develop VEVs v, u, and w, respectively, as

follows
     
√

 v/ 2 


 0 


 0 

√
     
     
hηi = 
 0 ,
 hρi = 
 u/ 2 ,
 hχi = 
 0 ,
 (2.5)
     
√
     
     
0 0 w/ 2

5
while for the sextet, only the σ2o field develops a VEV (a nonzero VEV of σ1o should

provide a nonzero neutrino mass which, however, we do not consider here)

 

 0 0 0 

 
 
hSi = 
 0 0 vσ /2 .
 (2.6)
 
 
 
0 vσ /2 0

Here, all VEVs are taken to be real. (We are restrained ourselves the possibility

of CP–violation arising from complex VEVs which has already been investigated in

detail by D. G. Dumm [12]). The expansion of the scalar fields reads

v u w
η o = √ + ξη + iζη , ρo = √ + ξρ + iζρ , χo = √ + ξχ + iζχ ,
2 2 2
σ2o vσ
√ = + ξσ + iζσ , (2.7)
2 2

and

σ1o = ξσ′ + iζσ′ . (2.8)

Below we call a real part ξ scalar and an imaginary one ζ pseudoscalar. In this case

the symmetry breaking ladder is

SUC (3) ⊗ SUL (3) ⊗ UN (1)

↓ hχi

SUC (3) ⊗ SUL (2) ⊗ UY (1)

↓ hρi, hηi, hSi

SUC (3) ⊗ UQ (1)

where the VEVs satisfy the relation

v 2 + u2 + vσ2 ≡ vW
2
≈ (246 GeV )2 , (2.9)

6
with vW being the standard model VEV. At the first step of the symmetry breaking,

hχi generates masses for exotic quarks and new heavy gauge bosons. At the subse-

quent breaking, nonzero values of hρi, hηi and hSi generates masses of the familiar

quarks and leptons. To keep the 3 - 3 - 1 models consistent with the low-energy

phenomenology, the VEV hχi must be large enough in comparison with other VEV’s

w ≫ v, u, vσ .

Due to the requirement the potential to reach a minimum at the chosen VEV’s

we obtain the following constraint equations in the tree–level approximation

!
2 λ λ λ12 λ u2 vσ λ w 2 vσ f uw
µ1 = 2
−λ1 v − 4 u2 − 5 w 2 − − λ20 vσ2 + 18 − 19 − 1 ,
2 2 2 v v v
!
λ λ λ13 f vw f2 vσ w
µ22 = −λ2 u2 − 4 v 2 − 6 w 2 − + λ16 vσ2 + 2λ18 vvσ − 1 − ,
2 2 2 u u
!
λ5 2 λ6 2 λ14 f vu f v u
µ23 = 2
−λ3 w − v − u − + λ17 vσ2 − 2λ19 vvσ − 1 − 2 σ ,(2.10)
2 2 2 w w
! ! ! !
λ11 λ12 λ13 λ14
µ24 = − λ10 + 2
vσ − 2
− λ20 v − 2
+ λ16 u − + λ17 w 2
2 2 2 2
λ u2 v λ w 2 v f2 wu
+2 18 − 2 19 −
vσ vσ vσ

which, in fact, exlude the linear terms in fields from the potential. The mass matri-

ces, thus, can be calculated, using

∂ 2 VE
Mij2 =
∂φi ∂φj

evaluated at the chosen minimum, where φi ’s are fields (ξ, ζ, η, ρ, χ, s).

7
3 Higgs mass spectra and physical particles

Since the σ1o field does not develop a VEV, the associated scalar ξσ′ and pseudoscalar

ζσ′ do not mix with other fields and we have the physical field Hσ′ ≃ ξσ′ with mass

f2 uw λ11 2 λ15 λ18 u2 v λ19 w 2v

m2H ′ = + v −( 2 2 2
+ λ20 )v + λ16 u + λ17 w − + . (3.1)
σ vσ 2 σ 2 vσ vσ

In the basis of ξη , ξρ , ξσ and ξχ the square mass matrix, after imposing the con-

straints (2.10), reads

 
 −m2ξ −m2ξ ξ
−m2ξ ξ
−m2ξ ξ

η η ρ η σ η χ
 
 
 
−m2ξ −m2ξ −m2ξ
 
 
2 ρ ρ ξσ ρ ξχ
M4ξ = (3.2)
 
 
 
2 2
 −mξ −mξ 
σ ξχ
 
 σ 
 
−m2ξ
 
χ

where

f1 uw λ u2 vσ λ w 2 vσ
m2ξ = − 2λ1 v 2 − 18 + 19 , m2ξ ξ = −f1 w − λ4 uv + 2λ18 uvσ ,
η v v v η ρ

w
m2ξ = (f v + f2 vσ ) − 2λ2 u2 , m2ξ ξ = −(λ12 − 2λ20 )vvσ + λ18 u2 − λ19 w 2 ,
ρ u 1 η σ

m2ξ = −f2 w − (λ13 + 2λ16 )uvσ + 2λ18 uv , m2ξ = −f1 u − λ5 vw − 2λ19 wvσ ,
ρ ξσ η ξχ

m2ξ = −f1 v + f2 vσ − λ6 uw , m2ξ = −f2 u − (λ14 + 2λ17 )wvσ − 2λ19 wv,

ρ ξχ σ ξχ

and
f2 uw λ u2 v λ19 w 2v
m2ξ ≡ − (2λ10 + λ11 )vσ2 − 18 + ,
σ vσ vσ vσ
u
m2ξ ≡ (f v + f2 vσ ) − 2λ3 w 2 .
χ w 1

8
As in [10] we use here the approximation |f1 | , |f2 | ∼ w and maintain only terms of

the second order in w in (3.2) (using w ≫ v, u, vσ ). This immediately gives us one

physical field

Hχ ≃ ξ χ (3.3)

with a mass

m2H ≃ −2λ3 w 2 , (3.4)

χ

and a square mass matrix of ξη , ξρ , ξσ mixing

 
f u λ19 wvσ

 − 1v − v
f1 λ19 w 

 
M3ξ2 ≃
 
w
 f1 − u1 (f1 v + f2 vσ ) f2 .
 (3.5)
 
 
 f u λ19 wv 
λ19 w f2 − v2 − vσ
σ

Solving the charateristic equation for the matrix (3.5) we get one massless field H1

and two physical ones (H2 , H3 ) with masses

" #
w f1 2 f2 2 λ w
x2,3 = − (u + v 2 ) + (vσ + u2 ) + 19 (v 2 + vσ2 )
2 uv uvσ vvσ

" #2
w  f1 2 2 f2 2 2 λ19 w 2 2
± (v + u ) + (v + u ) + (v + vσ )
2  vu uvσ σ vvσ
" !#)1/2
2 f1 f2 λ19 w f1 f
−4vw + + 2
vvσ u vσ v
≡ m2H2,3 (3.6)

The characteristic equation corresponding to x2,3 can be given in the following com-

pact form

v [F2 (n) + G1 ] + u [F1 (n)F2 (n) − G1 G2 ] + vσ [F1 (n) + G2 ] = 0, n, = 2, 3, (3.7)

9
where

u v xi
F1 (i) = + G1 σ + ,
v v f1 w
u v xi
F2 (i) = + G2 + ,
vσ vσ f2 w
λ19 w
G1 = ,
f1
λ19 w
G2 = . (3.8)
f2

To construct physical fields we now consider the equation

 
M 2 − xi Hi = 0, i = 1, 2, 3, (3.9)

where M 2 is a square mass matrix and Hi ≡ (Hi1 , Hi2 , Hi3 )T . For M3ξ2 we obtain a

system of three equations

!
f uw λ19 w 2 vσ
− 1 + + xi Hi1 + f1 wHi2 + λ19 w 2 Hi3 = 0,
v v
w
 
f1 wHi1 − (f v + f2 vσ ) + xi Hi2 + f2 wHi3 = 0, (3.10)
u 1 !
2
f uw λ w v
λ19 w 2 Hi1 + f2 wHi2 − 2 + 19 + xi Hi3 = 0.
vσ vσ

It is clear that this system of equations is over defined and can be reduced to two

equations, say, the first and the last ones. Thus we have a freedom to suppose

Hi1 = k(i), (3.11)

where k(i) will be defined by the normalization of the states. Hence,

F1 (i)F2 (i) − G1 G2
Hi2 = k(i) ≡ Γ2 (i) k(i) (3.12)
F2 (i) + G1

10
and

F1 (i) + G2
Hi3 = k(i) ≡ Γ3 (i) k(i). (3.13)
F2 (i) + G1

Then k(i) can be found

h i−1/2
k(i) = 1 + Γ22 (i) + Γ23 (i) , (3.14)

by normalizing the states Hi written now in the form

   

 1 


 Hi1 

   
   
Hi = k(i) 
 Γ2 (i) 
 ≡ 
 Hi2 .
 (3.15)
   
   
   
Γ3 (i) Hi3

In the massless (x1 = 0) approximation i = 1 we immediately find

 

 v 

1
 
 
H1 = 
u . (3.16)
vW 



 
 
vσ

In the next approximation (when the λ’s are taken into account) the field H1 acquires

a mass. Solving the characteristic equation for the exact 3 × 3 mass matrix M3ξ2

and the H1 , namely

 
M3ξ2 − x1 H1 = 0, (3.17)

we obtain the following formulas for the H1 mass

u2 vσ
m2H1 = x1 ≈ 2λ1 v 2 + λ4 u2 + (λ12 − 2λ20 )vσ2 − 2λ18
v
≈ λ4 v 2 + 2λ2 u2 + (λ13 + 2λ16 )vσ2 − 4λ18 vvσ

≈ (λ12 − 2λ20 )v 2 + (λ13 + 2λ16 )u2 + (2λ10 + λ11 )vσ2

u2 v
−2λ18 . (3.18)
vσ

11
 The coupling constants λ’s must be chosen such that the Eqs. (3.18) to be

compatible with the Eq. (2.9) or, in the geometric language, we need to find on the

sphere (2.9) all point(s) (v, u, vσ ) where all the surfaces (3.18) get together. The

simplest solution of this system of equations (2.9) and (3.18) could be found if we

accept the following relation among coupling constants

1 1 1
λ ≈ λ1 ≈ λ12 − λ20 ≈ λ4 /2 ≈ λ2 ≈ ( λ13 + λ16 ) ≈ (λ10 + λ11 ). (3.19)
2 2 2

It would be
u v
v ≈ vσ ≈ √ ≈ W ≈ 123 GeV , (3.20)
2 2

following from
u2 vσ u2 v
≈ 2vvσ ≈ ≡ δ2. (3.21)
v vσ

The assumption (3.19) is justified by examining the latest VEV’s (3.20). It is easily

to see here that δ 2 = 21 vW

2
. Then the mass of H1 would take the value

m2H1 ≈ 2λvW
2
− 2δ 2 λ18 = vW
2
(2λ − λ18 ) , (3.22)

while the eigenstates can be expressed, according to (3.3), (3.15) and (3.22), as

follows
    
v u vσ

 H1 


 vW vW vW

 ξη 

    
    

 H2 
 ≈ 
 H21 H22 H23 
 ξρ ,
 (3.23)
    
    
    
H3 H31 H32 H33 ξσ

Hχ ≈ ξ χ . (3.24)

12
Since the matrix
 
v u vσ
vW vW vW
 
 
   T
≡ A−1
 
AHξ = 
 H21 H22 H23 
 Hξ
, det AHξ = 1 (3.25)
 
 
 
H31 H32 H33

in (3.23) is an orthonormal matrix SO(3) the relation inverse to (3.23) and (3.24)

can easily be found

    
v

 ξη 


 vW   H1 
H21 H31   
    
    

ξρ  ≈  u
H22 H32 
H2 , (3.26)





 vW 



    
vσ
    
ξσ vW
H23 H33 H3

ξ χ ≈ Hχ . (3.27)

Similarly, in the pseudoscalar sector we obtain one physical field ζσ ≡ ζσ′ with a

mass equal to the mass of Hσ′ , and the square mass matrix of the ζη , ζρ, ζσ , ζχ mixing
 
 −m2ζ −m2ζ ζ
−m2ζζ
−m2ζ ζ

η η ρ η σ η χ
 
 
 
−m2ζ −m2ζ −m2ζ
 
 
2 ρ ρ ζσ ρ ζχ
M4ζ = . (3.28)
 

 
2 2
 −mζ −mζ 
σ ζχ
 
 σ 
 
−m2ζ
 
χ

where
f1 uw λ18 u2 vσ λ w 2vσ
m2ζ = − + 19 − 2λ20 vσ2 , m2ζ = f1 w ,
η v v v ζ
η ρ

w √
m2ζ = (f1 v + f2 vσ ) , m2ζ ζ = λ18 u2 − λ19 w 2 + 4 2λ19 vvσ ,
ρ u η σ

m2ζ = f2 w , m2ζ = f1 u , m2ζ = f1 v + f2 vσ , m2ζ = f2 u ,

ρ ζσ η ζχ ρ ζχ σ ζχ

13
and

f2 uw λ18 u2 v λ19 w 2 v u
m2ζ = − + − 2λ20 v 2 , m2ζ = (f1 v + f2 vσ ).
σ vσ vσ vσ χ w

In the approximation |f1 |, |f2 | ∼ w ≫ v, u, vσ we obtain one Goldstone boson G1 ≈

ζχ and the ζη , ζρ, ζσ mixing

 
f u λ19 wvσ

 − 1v − v
−f1 λ19 w 

 
M3ζ2 =
 
w
 −f1 − u1 (f1 v + f2 vσ ) −f2 .
 (3.29)
 
 
 f u λ19 wv 
λ19 w −f2 − v2 − vσ
σ

It is clear that the characteristic equation in this case gives the same roots as in

the scalar sector, but a different set of the eigenstates (simply, make replacements

Hi2 → −Hi2 )     
v vσ
 A1  
vW
− vu vW
 ζη 
   W  
    
    

 A2 
 ≈ 
 H21 −H22 H23 
 ζρ 
 (3.30)
    
    
    
A3 H31 −H32 H33 ζσ
or equivalently
    
v

 ζη 


 vW
H21   A1 
H31   
    
    

 ζρ 
 ≈ 
 − vu −H22 −H32 
 A2 .
 (3.31)
   W  
    
vσ
    
ζσ vW
H23 H33 A3

In the singly charged sector the mixing occurs in the set of η1+ , ρ+ , s+
1 and in the

set of η2+ , χ+ , s+
2 with the following square mass matrices
 
2 2 2
 −mη −m −m + 
 1 ρ+ η − s η− 
 1 1 1 
 
2
M+1 = −m2ρ+ ρ− −m2s+ ρ− , (3.32)
 

 
 1 
 
−m2+
 
s s−
1 1

14
where
f1 uw λ7 u2 λ λ u2vσ λ w 2vσ
m2η = − − ( 15 + λ20 )vσ2 − 18 + 19 ,
1 v 2 4 v v
2 λ7 uv 2 w λ7 v 2
m + − = f1 w − − λ18 uvσ , mρ+ ρ− = (f1 v + f2 vσ ) − + λ16 vσ2 − 2λ18 vvσ ,
ρ η
1 2 u 2
λ vv
m2+ − = 15 σ + λ19 w 2 − λ20 vvσ , m2s+ ρ− = −f2 w − λ16 uvσ + λ18 uv ,
s η
1 1
4 1

f2 uw λ11 vσ2 λ λ u2 v λ19 w 2 v

m2+ = + − ( 15 + λ20 )v 2 + λ16 u2 − 18 + ,
s s−
1 1 vσ 4 4 vσ vσ

and
 
λ8 wv
 −m2η −f1 u + 2
− λ19 wvσ ( 14 λ15 + λ20 )vvσ + λ18 u2 
 2 
 
2  
M+2 = 

2
−mχ+ χ− f2 u + λ17 wvσ + λ19 wv .
 (3.33)
 
 
−m2+
 
s s−
2 2

where
f1 uw λ8 w 2 λ λ u2 vσ λ w 2 vσ
m2η = − − ( 15 + λ20 )vσ2 − 18 + 19 ,
2 v 2 4 v v
u λ v2
m2χ+ χ− = (f1 v + f2 vσ ) − 8 + λ17 vσ2 + 2λ19 vvσ ,
w 2
2
f uw λ11 vσ λ λ u2 v λ19 w 2 v
m2+ = 2 + − ( 15 + λ20 )v 2 + λ17 w 2 − 18 + .
s s
−
2 2 vσ 4 4 vσ vσ

2
Applying the approximation for M+2 we obtain one Goldstone boson G+
2
≈ χ+

and two physical fields associated with η2+ and s+

2
with masses

f1 uw λ8 w 2 λ19 w 2 vσ f2 uw λ w2v
m2η = − + , m2+ = + λ17 w 2 + 19 . (3.34)
2 v 2 v s s
−
2 2 vσ vσ

For the η1+ , ρ+ , s+

1
mixing, we have
 
f u λ19 wvσ

 − 1v − v
−f1 −λ19 w 

 
2  
M+1 = w
 −f1 − u1 (f1 v + f2 vσ ) f2 .
 (3.35)
 
 
 f u λ19 wv 
−λ19 w f2 − v2 − vσ
σ

15
As before, the characteristic equation of (2.41) has the same roots, but the eigen-

states are different and are given by (necessary replacements: Hi2 → −Hi2 , Hi3 →

−Hi3 )     
 h1 + 
vW
v
− vu − vvσ  η1 +
   W W  
    
    

 h2 + 
 ≈ 
 H21 −H22 −H23 
 +
ρ 
 (3.36)
    
    
    
h+
3
H31 −H32 −H33 s+
1

or equivalently
    
v

 η1+ 


 vW
H21  + 
  h1 
H31 
    
    

 ρ+ 
 ≈ 
 − vu   h+
−H22 −H32   .
 (3.37)
2
   W  
    
− vvσ
    
s+
1
−H23 −H33 h+
3
W

In the doubly charged sector the mixing occurs up all states ρ++ , s++
2
, χ++ , s++
1
,

and the square mass matrix is given

 


−m2ρ++ ρ−− −m2s++ ρ−− −m2χ++ ρ−− −m2s++ ρ−− 

 2 1 
 
2 2 2
−m ++ −m −m ++
 
 
2  s s−− χ++ s−− s s−− 
M4++ = 

2 2 2 1 2 ,
 (3.38)
2 2
 

 −mχ++ χ−− −ms++ χ−− 

 1 
 
−m2s++ s−−
 
1 1

where
2 w λ9 w 2
mρ++ ρ−− = (f1 v + f2 vσ ) − − 4λ18 vvσ ,
u 2
√ √
m2s++ ρ−− = − 2(f2 w + λ16 uvσ − λ18 uv) , m2χ++ ρ−− = f1 v + f2 vσ − 2λ18 uw ,
2

√ √
m2++ = 2(−λ17 wvσ + λ19 wv) , m2s++ ρ−− = − 2(λ16 uvσ + λ18 uv) ,
χ s−−
2 1

√
m2++ = −λ20 v 2 , m2s++ χ−− = − 2(f2 u + λ17 wvσ + λ19 wv) ,
s s−−
1 2 1

16
and
f2 uw λ u2 v λ19 w 2 v
m2++ = + λ16 u2 − λ17 w 2 − 18 + − λ20 v 2 ,
s
2
s−−
2 vσ vσ vσ
u λ u2
m2χ++ χ−− ≡ (f1 v + f2 vσ ) − 9 + 4λ19 vvσ ,
w 2
f2 uw λ18 u2 v λ19 w 2 v
m2s++ s−− ≡ 2 2
− λ16 u + λ17 w − + − λ20 v 2 .
1 1 vσ vσ vσ

By the same way as considered above we obtain one Goldstone boson G++
3
≈ χ++

and one physical field s++

1
with mass

f2 uw λ w2v
m2s++ s−− = + λ17 w 2 + 19 , (3.39)
1 1 vσ vσ

and a matrix of ρ++ , s++

2
mixing

√
 
λ9 w
 − u1 (f1 v + f2 vσ ) + 2
2f2 
2
M2++ = w . (3.40)
 
 √ f u λ19 wv

2f2 − v2 + λ17 w − vσ
σ

Solving the characteristic equation we get two square masses

" ! #
w λ9 w λ wv u v fv
x4,5 = + λ17 w − 19 − f2 + σ − 1
2 2 vσ vσ u u
" ! #2
w  λ9 w λ wv u v fv
± + λ17 w − 19 − f2 + σ − 1
2 2 vσ vσ u u
√ " # ! )1/2
2 2 λ9 uw λ17 wvσ 2
+ (f1 v + f2 vσ ) − f2 u − √ + λ19 wv − f2
uvσ 2 2
≡ m2d . (3.41)
1 ,d2

for two physical fields

    
x4 λ17 w λ19 wv
 
u

 d++
1



 n4 √12 vσ
+ f2 w
− f2
+ f2 vσ
n4   ++ 
 ρ 
    
    



 = 



,
 (3.42)
    
    
x5 λ17 w λ19 wv
 
u
    
d++
2
n5 √12 vσ
+ f2 w
− f2
+ f2 vσ
n5 s++
2

17
corresponding to (3.41) in [11] where
 !2 − 12
1 u x λ w λ wv
ni = 1 + + i − 17 + 19  , (i = 4, 5), (3.43)
2 vσ f2 w f2 f2 vσ

is found by normalizing states. Here, using a shorter notation

 
 a11 a 
2
M2++ = . (3.44)
 
 
a a22

of the matrix (3.40) we, however, can rewrite (3.42) in another way
    
 d++
1
  aN4 X4 N4   ρ++ 
= , (3.45)
    
   
    
++ ++
d2 aN5 X5 N5 s2

where
 − 1
Ni = a2 + Xi2 2
(3.46)

and
1
 q 
Xi = −a11 + a22 ± (a11 − a22 )2 + 4a2 , (i = 4, 5). (3.47)
2

4 Conclusion

We have just considered the Higgs sector of the minimal 3 - 3 - 1 model under the

most general gauge–invariant potential conserving lepton numbers. In comparison

with the pevious paper [11] the content of the particles and their multiplet decom-

position structure remain the same but most of the masses get corrections:

18
 – in the neutral scalar sector, physical fields are: H1 , H2 , H3 , Hσ′ and Hχ

m2H1 ≈ 2λvW
2
− 2δ 2 λ18 = vW
2
(2λ − λ18 ) , m2H2 = x2 , m2H3 = x3 ,
!
f uw λ11 2 λ15 λ u2 v λ19 w 2v
m2Hσ′ = 2 + vσ − + λ20 v 2 + λ16 u2 + λ17 w 2 − 18 +
vσ 2 2 vσ vσ
√
!
1 λ11 λ 2
≈ 2f2 w + − 15 + 2λ16 − 2λ18 − λ20 vW + (λ17 + λ19 ) w 2 ,
4 2 2
m2χ ≈ −2λ3 w 2, (4.1)

where the relation (3.20) is used,

– in the neutral pseudoscalar sector, physical fields are: A2 , A3 , Aσ and two

Goldstone bosons: G1 ≈ ζχ and G2 (corresponding to the massless A1 )

m2A2 = x2 , m2A3 = x3 , m2A′ = m2H ′ , (4.2)

σ σ

– in the singly charged sector, there are two Goldstone bosons G3 = h+

1
, G+
4
≈ χ+

and three physical fields : h+

2
, h+
3
, η2+ , s+
2
with masses:

m2h+ = m2H2 , m2h+ = m2H3 ,

2 3

2 2
f uw λ8 w λ w vσ f2 uw λ19 w 2 v
m2η2 ≈ 1 − + 19 , m+ 2
≈ 2
+ λ17 w + , (4.3)
v 2 2v s s−
2 2 vσ vσ

– in the doubly charged sector, there are one Goldstone (G++

5
≈ χ++ ) and three

physical fields d++

1
, d++
2
, s++
1
with masses:

f2 uw λ w2v
m2d++ = x4 , m2d++ = x5 , m2s++ s−− = + λ17 w 2 + 19 ≡ m2+ − . (4.4)
1 2
1 1 vσ vσ s s
2 2

Eqs. (4.1 – 4.4) show that f1 and f2 can take a definite consistent sign and there are

three degenerate states H2 , A2 and h+

2
in mass x2 , three degenerate states H3 , A3

and h+
3
in mass x3 and two degenerate states Hσ′ , A′σ in mass m2H ′ .
σ

19
 Combining the assumption (3.19) and the positiveness of the mass squares we

get the following bounds for coupling constants

>
λ ≈ λ1 ≈ 21 λ12 − λ20 ≈ λ4 /2 ≈ λ2 ≈ ( 12 λ13 + λ16 ) ≈ (λ10 + 12 λ11 ) ∼ 0,
(4.5)
<
λ3 ∼ 0.

Note that new coupling constants λ15 , λ17 , λ18 and λ19 remain unconstrained by

(3.19) and (4.5). It is worth mentioning that the system of Eqs. (2.9) and (3.18)

may admit more general solutions with other coupling constants rather than those

constrained by (3.19). This question deserves to be furthermore investigated.

In conclusion, the present paper is an extension of previous investigations [10, 11]

on the Higgs sector of the minimal 3 - 3 - 1 model with three triplets and one sextet.

Under the most general lepton–number conserving potential the mass spectra and

the multiplet decomposition structre of this sector are investigated in detail at tree-

level. Due to the fact that most of the scalar masses get corrections the problem with

inconsistent signs of f2 arising in the previuos case [10, 11] is solved. The results of

this paper are exposed in a systematic order and a transparent way allowing them

to be easily checked and used in further studies.

Acknowledgments

One of the authors (N.A.K.) would like to thank T. Inami and Department of

Physics, Chuo University, Tokyo, Japan, for warm hospitality and Nishina founda-

tion for financial support. H.N.L. would like to thank CNRS for financial support

and Theory Group, LAPP, Annecy, France, for kind hospitality. We are thankful to

20
the referee for useful remarks on the normalization of the sextet and VEV’s.

This work was supported in part by the Vietnam National Research Programme

on Natural Sciences under grant number KT–04.1.2 and the Research Program on

Natural Science of Hanoi National University under grant number QT 98.04.

References

[1] F. Pisano and V. Pleitez, Phys. Rev. D46, 410 (1992);

P. H. Frampton, Phys. Rev. Lett. 69, 2889 (1992).

[2] R. Foot, O. F. Hernandez, F. Pisano, and V. Pleitez, Phys. Rev. D47, 4158

(1993).

[3] M. Singer, J. W. F. Valle, and J. Schechter, Phys. Rev. D22, 738 (1980).

[4] R. Foot, H. N. Long, and Tuan A. Tran, Phys. Rev. D50, R34 (1994).

[5] H. N. Long, Phys. Rev. D53, 437 (1996); Phys. Rev. D54, 4691 (1996).

[6] J. C. Montero, F. Pisano, and V. Pleitez, Phys. Rev. D47, 2918 (1993).

[7] R. D. Peccei and H. R. Quinn, Phys. Rev. Lett. 38, 1440 (1977); Phys. Rev.

D16, 1791 (1977).

[8] P. B. Pal, Phys. Rev. D52, 1659 (1995).

[9] D. Ng, Phys. Rev. D49, 4805 (1994).

21
[10] M. D. Tonasse, Phys. Lett. B381, 191 (1996).

[11] N. T. Anh, N. A. Ky, and H. N. Long, [hep-ph/9810273], Int. J. Mod. Phys.

A15, 283 (2000).

[12] M. B. Tully and G. C. Joshi, Preprint UM–P–98/52, [hep-ph/9810282]; J. T.

Liu and D. Ng, Phys. Rev. D50, 548 (1994); D. G. Dumm, Int. J. Mod. Phys.

A9, 887 (1994).

[13] J. Montero, C. de S. Pires and V. Pleitez, Phys. Rev. D60, 115003 (1999).

22