Neutrino Mixing based on Mass Matrices with a $2 \leftrightarrow 3$ Symmetry
Yoshio Koide, Eiichi Takasugi

TL;DR
This paper systematically investigates lepton mass matrices with a $2 ightarrow 3$ symmetry broken only by phases, exploring different symmetry scenarios and their implications for neutrino phenomenology without relying on explicit parameter values.
Contribution
It introduces a comprehensive analysis of lepton mass matrices with $2 ightarrow 3$ symmetry broken by phases, considering different symmetry implementations and their effects on neutrino masses.
Findings
Neutrino seesaw mass matrix can obey $2 ightarrow 3$ symmetry even if $M_R$ does not.
Different symmetry scenarios lead to distinct phenomenological implications.
The approach does not depend on explicit parameter values, allowing broad applicability.
Abstract
Under the assumption that the symmetry is broken only through phases, we give a systematical investigation of possible lepton mass matrix forms without referring to the explicit parameter values. The two types of the symmetry are investigated: one is that the left- and right-handed fields obey the symmetry, and another one is that only obeys the symmetry. In latter case, in spite of no symmetry in the Majorana mass matrix for , the neutrino seesaw mass matrix still obey the symmetry. Possible phenomenologies are discussed.
| Type | Type II for | Type I for |
|---|---|---|
| Type II for | Model A: non-GUT type | Model D: unrealistic |
| & | ||
| Type I for | Model B: SU(5)-GUT type | Model C: SO(10)-GUT type |
| & | & |
| Hierarchy | Normal | Inverse | Normal | Inverse |
|---|---|---|---|---|
| Limit of | ||||
| & | ||||
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Abstract
Under the assumption that the symmetry is broken only through phases, we give a systematical investigation of possible lepton mass matrix forms without referring to the explicit parameter values. The two types of the symmetry are investigated: one is that the left- and right-handed fields obey the symmetry, and another one is that only obeys the symmetry. In latter case, in spite of no symmetry in the Majorana mass matrix for , the neutrino seesaw mass matrix still obey the symmetry. Possible phenomenologies are discussed.
**Neutrino Mixing based on Mass Matrices **
with a Symmetry
Yoshio Koide and Eiichi Takasugi
*Institute for Higher Education Research and Practice, Osaka University,
1-16 Machikaneyama, Toyonaka, Osaka 560-0043, Japan
1 Introduction
We usually consider that the quarks and leptons should be understood by a unification theory. Then, the concept of “symmetry” will become important in the understanding of “flavor”. It is well-known that the requirement of the symmetry [1] for the neutrino mass matrix leads to the maximal mixing between the and components. The idea of the the symmetry is very promising for understanding the observed neutrino mixing.
When a matrix satisfies the relation
[TABLE]
where is defined as
[TABLE]
the matrix is called “ symmetric”. The mass matrix form of is explicitly expressed as
[TABLE]
Firstly, we would like to notice that the mass matrix which satisfies Eq.(1.1) is considered a consequence of the invariance of the mass matrix under the field transformation. Explicitly, for the Dirac mass matrix , Eq.(1.1) is derived by requiring the invariance under the transformation, and . This is true if the neutrino mass matrix is derived by the seesaw mechanism, because this matrix is invariant under with any . Next, we focus on the transformation of and . Since they forms an doublet of the electroweak symmetry, the transformation for them should be the same. That makes a big trouble to realize the reasonable neutrino mixing as we see in the next section.
Now we extend the symmetry according to multiplets under the electroweak symmetry. In general, the transformation between and are different. This is true even we consider the SU(5) GUT. On the other hand, in the SO(10) GUT, and will be transformed under the same operator . According to this classification, two types of symmetry arises. The one (we call it Type I) is that both and obey the symmetry. Eq.(1.3) is obtained for charged leptons and also for neutrinos. Consider that and and we find the Majorana mass matrix in Eq.(1.3).
The other one (we call it Type II) is the case where only obeys the symmetry. Then, we find for the Dirac mass matrix (we define a Dirac mass matrix as )
[TABLE]
and the explicit form of the mass matrix is given by
[TABLE]
The neutrino mass matrix is given as a special case of Eq.(1.3) by taking as we shall see later.
Note that, in the both types I and II, the Hermitian matrix defined by satisfies the constraint
[TABLE]
independently whether the mass matrx has the form in Eq.(1.3) or (1.5).
Now the neutrino mixing matrix is given by
[TABLE]
where are defined by
[TABLE]
From the argument given above, we learned that as far as the mixing matrix is concerned, the structure of the neutrino mixing matrix is independent of the mass matrices of Types I or II. Only difference arises in the mass spectrum.
The purposes of the present paper is to investigate the general properties of the models with the symmetry, paying attention to the difference between types I and II, and taking relations to the grand unification (GUT) scenarios into consideration. Although we investigate the masses and mixings in the lepton sectors, the formulation in this paper is also applicable to the quark sectors. Since, in the quark sectors, there is essentially no case complexity about the mass spectrum such as the inverse hierarchy as in the neutrino sector, the application is more straightforward. Therefore, we will investigate only the lepton sectors in this paper.
2 Extended symmetry and the neutrino mixing
In this section, we will demonstrate that the symmetry in the exact meaning cannot explain the observed neutrino mixing. For the convenience of the discussion in later, let us introduce the so-called extended operator [2]
[TABLE]
instead of the operator (1.2) and consider both types. The operator is unitary and Hermitian. We obtain the constraint
[TABLE]
for the Hermitian matrix irrespective of Type I or II. Note that we can express the operator (2.1) as
[TABLE]
where and
[TABLE]
Therefore, we can express the constraint (2.2) as
[TABLE]
Now we define
[TABLE]
then we find
[TABLE]
where is a Hermitian matrix
In general, the Hermitian matrix which satisfies the constraint (2.6) can be expressed by the form
[TABLE]
where , , and are real, so that can be transformed to a real matrix as
[TABLE]
where
[TABLE]
It is also well-known that the symmetric real matrix is diagonalized by a rotation as
[TABLE]
where
[TABLE]
[TABLE]
[TABLE]
As a result, the Hermitian matrix is diagonalized by
[TABLE]
as
[TABLE]
Since we did not considered the size of masses, the ordering of them is needed. Therefore, the unitary matrix to diagonalize the mass matrix in an proper mass ordering is given by , where is the matrix to exchange the mass ordering. Then, we find the neutrino mixing matrix defined by (1.7) as
[TABLE]
where
[TABLE]
Here, we recall that the operation (2.1) must be the same for and , so that , in the expression, is exactly equal to . Therefore, we obtain
[TABLE]
where . Obviously, the mixing matrix (2.18) cannot give the observed values [3, 4] and simultaneously. (It is a general feature for any flavor symmetry with a transformation that we obtain only a family-mixing between two families. See Ref.[5].)
3 Extended symmetry with the breaking term
We saw in the previous section that the symmetry which arises as a consequence of the transformation for fields cannot reproduce the observed neutrino mixing. However, we consider that the symmetry is still useful from the phenomenological point of view. Therefore, from the phenomenological point of view, we assume [6] that the symmetry is broken only through the phase parameters. Hereafter, we will use the extended symmetry operator (2.1) in the phenomenological meaning, and we will consider the case in the left-handed sectors.
3.1) Charged lepton mass spectrum
First, we investigate the symmetry of Type II. The mass matrix for the charged leptons must also satisfy the relation
[TABLE]
where, for convenience, we have dropped the index “e” from . Then, the explicit form of is also given by
[TABLE]
where the parameters in can be complex. Therefore, we obtain the Hermitian matrix
[TABLE]
where
[TABLE]
Then, we can obtain a real matrix as
[TABLE]
From the formula (2.13), we obtain
[TABLE]
because of in this case. Therefore, Type II transformation in charged lepton sector cannot give a realistic mass spectrum.
Next, we investigate the case of Type I, i.e.
[TABLE]
The case (3.5) may be realized in an SU(5)-GUT model. In this case, instead of the constraint (3.1), we have the constraint
[TABLE]
The explicit form of is given by
[TABLE]
so that we obtain
[TABLE]
where
[TABLE]
where and are defined by and , respectively. Therefore, since
[TABLE]
we can obtain when .
In both cases, Types I and II, the Hermitian matrix is diagonalized by the unitary matrix
[TABLE]
as
[TABLE]
3.2) Neutrino mass spectrum
We consider that the neutrino masses are generated by a seesaw mechanism
[TABLE]
where and are defined by and (), respectively. The Dirac mass matrix is given by the form similar to (3.9) or (3.2) according as Type-I or Type-II. In Type-I, we obtain the neutrino mass matrix form
[TABLE]
where
[TABLE]
[TABLE]
Since the neutrino masses in Type-II are given by
[TABLE]
with , we obtain
[TABLE]
On the other, in Type I, such the constraint (4.6) does not appear.
In both cases, Types I and II, the Hermitian matrix is diagonalized by the unitary matrix
[TABLE]
as
[TABLE]
where is defined by Eq.(2.11) with
[TABLE]
3.3) Neutrino mixing matrix
So far, we have used the notation for the mass eigenstates of the fundamental fermions , whose masses have been defined by Eq.(2.13). Hereafter, in order to distinguish the mass-eigenstates and in the conventional notations from the mass-eigenstates whose masses are defined by Eq.(2.13), we denote the states whose masses are defined by Eq.(2.13) as . The states and , which is the SU(2)L partner of the charged lepton state , are related by
[TABLE]
with the neutrino mixing matrix in the conventional notation. Here, the neutrino mixing matrix in Eq.(3.24) is given by
[TABLE]
On the other hand, as seen in Secs.2 and 3, the mass matrices and are diagonalized by unitary matrices (3.21) and (3.13) (we denote them and ), respectively. When we define the mixing matrix
[TABLE]
where
[TABLE]
and . the mixing matrix does not always denote the observed neutrino mixing matrix . When we define the observed fermions and as
[TABLE]
the observed neutrino mixing matrix is given by
[TABLE]
where denotes the exchange operator . However, as we discuss below, the possible choices of are not so many.
The explicit form of the matrix is given by
[TABLE]
Obviously, the cases (3.29) with are ruled out as we have already discussed in Sec.2.
For convenient, we name Models A, B, C and D for combinations of Types I and II for and as shown in Table 1. In Model A, since only the left-handed fields obey the symmetry, the model cannot be embedded into a GUT scenario. In Model B, the fields and obey the symmetry, but the field is free from the symmetry, so that the model can be embedded into SU(5) GUT. In Model C, all fields , and obey the symmetry, so that the model can be embedded into SO(10) GUT. Model D is unlikely, so that we will not investigate this case.
In Models A and D with Type-II symmetry in the charged lepton sector, we obtain , so that the cases are ruled out.
In Model B (a SU(5)-GUT-type model), we can obtain (but ) because of . (In Model B, although we can, in principle, consider any value of , we have assumed because the case can reasonably be realized in most practical models with symmetry.) Therefore, we may suppose a case in the model. Such the case means the assignment
[TABLE]
Then, from the relation , the observed neutrino mixing matrix is described by
[TABLE]
if we regard the observed neutrino states as with , whose case corresponds to the inverse hierarchy. (Such an inverted assignment between up- and down-sectors was first proposed by Matsuda and Nishiura [7].) The case (3.32) predicts
[TABLE]
[TABLE]
where and are given by Eq.(2.12). In order to give , the condition () is required. In order to (), the relation (i.e. ) is required from Eq.(2.13). Then, the masses (2.13) are given by
[TABLE]
Therefore, a suitable choice of the parameter values of can give keeping . Also, a suitable choice of the parameter values of can give a reasonable value of (3.33). If these conditions are satisfied, the model B is preferable. However, note that the parameter value cannot be realized unless SU(2)L is broken.
By the way, the case does not always mean the inverse hierarchy of neutrino masses. At present, as far as the observed neutrino masses satisfy the relation , we may consider any cases . Therefore, even the case , we can consider a case of the normal hierarchy: . Then, in Model B with , the neutrino mixing matrix is given by
[TABLE]
In order to give and , we have to consider . From the expression (3.23), the limit of requires , which leads and gives the mass spectrum , and . If we choose in the neutrino sector, we can give a reasonable value of because of in the normal mass hierarchy. Therefore, we cannot rule out this case (Model B with and in a normal hierarchy). However, we must accept a phenomenological value () in order to understand .
So far, we have consider the case with (i.e. ) for the charged lepton masses in Model B. We can also consider the case in Model B. In Model B, the neutrino masses are still given by , so that the cases and correspond to the normal and inverse hierarchies, respectively. The explicit form of for the case has been given in (3.30) because . The explicit form of the case is given by
[TABLE]
In order to see whether those cases cannot be ruled out or not, it is convenient to see whether we can take or not possible parameter values in the limit of , and , without contradicting with the observed neutrino mass hierarchy. The results are listed in Table 2. All cases are acceptable if we neglect the problem whether such a set of the parameter values is natural or not, although we think that the case with is unlikely.
In Model C, since we can take any order of , we cannot say any definite conclusion (predictions) without giving the explicit mass matrix parameters. Therefore, for the case C, we do not give a table such as Table 2.
4 Summary
In conclusion, we have systematically investigated possible lepton mass mass matrix forms and mixings under the expended symmetry. We gave investigated two types of the symmetry: one (Type I) is that the left- and right-handed fields obey the symmetry, and another one (Type II) is that only obeys the symmetry. Note that even in Type II, in spite of no symmetry in the Majorana mass matrix for , the neutrino seesaw mass matrix still obey the symmetry. However, we have concluded that the fermion mass is always zero in Type II. Therefore, the possibility that the charged lepton sector obeys the symmetry of Type II is ruled out. We have been interested in the case B classified in Table 1, where the neutrino sector obeys the symmetry of Type II, because we consider a model with an SU(5)-GUT type scenario [8]. In this case, we have only four cases of the neutrino mixing matrix. The results are summarized in Table 2.
We are also interested in a model with an SO(10)-type scenario. In this case (Model C), the right-handed neutrino is also transformed as , so that we can consider any value of and any mixing matrix form (2.19). However, in the SO(10)-GUT model, a more strict constraint on the neutrino mass matrix appears because the neutrino mass matrix form is strictly related to the quark and charged lepton mass matrices, so that most naive SO(10) models have, at present, not succeeded [9] in giving reasonable fits for all the masses and mixings in the quark and lepton sectors, even without the symmetry.
In the practical point of view, we think that there is a possibility to build a realistic model based on SU(5)-GUT rather than SO(10). In Model B, we are interested in the case of an inverse neutrino mass hierarchy, because the case is likely. The case predicts the effective electron neutrino mass is of the order of eV, which is within the reach of the next generation experiments of the neutrinoless double beta decay.
We hope that the present investigation will be helpful to investigate more explicit model based on a GUT scenario.
Acknowledgment
The one of the authors (YK) is supported by the Grant-in-Aid for Scientific Research, Ministry of Education, Science and Culture, Japan (No.18540284).
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] T. Fukuyama and H. Nishiura, hep-ph/9702253, in Proceedings of the International Workshop on Masses and Mixings of Quarks and Leptons , Shizuoka, Japan, 1997, edited by Y. Koide (World Scientific, Singapore, 1998), p. 252; R. N. Mohapatra and S. Nussinov, Phys. Rev. D 60 , 013002, (1999); E. Ma and M. Raidal, Phys. Rev. Lett. 87 , 011802 (2001); C. S. Lam, Phys. Lett. B 507 , 214 (2001); K. R. S. Balaji, W. Grimus and T. Schwetz, Phys. Lett. B 508 , 301 (2001); W. Grimus and L. Lavoura,
- 2[2] Y. Koide, Phys. Rev. D 69 , 093001 (2004).
- 3[3] D. G. Michael et al. MINOS Collaboration, Phys. Rev. Lett. 97 , 191801 (2006). Also see, J. Hosaka et al. the Super-Kamiokande Collaboration, ar Xiv: hep-ex/0604011.
- 4[4] B. Aharmim et al. SNO Collaboration, Phys. Rev. C 72 , 055502 (2005). T. Araki et al. Kam LAND Collaboration, Phys. Rev. Lett. 94 , 081801 (2005).
- 5[5] Y. Koide, Phys. Rev. D 71 , 016010 (2005).
- 6[6] Y. Koide, H. Nishiura, K. Matsuda, T. Kikuchi and T. Fukuyama, Phys. Rev. D 66 , 093006 (2002).
- 7[7] K. Matsuda and H. Nishiura, Phys. Rev. D 69 , 053005 (2004).
- 8[8] For example, R. N. Mohapatra, S. Nussinov and H. B. Yu, Phys. Lett. B 636 , 114 (2006).
