Available online at www.sciencedirect.

com

ScienceDirect
Nuclear Physics B 896 (2015) 281–310
www.elsevier.com/locate/nuclphysb

Higgs → μτ in Abelian and non-Abelian flavor

symmetry models
Julian Heeck a,b,∗ , Martin Holthausen a , Werner Rodejohann a ,
Yusuke Shimizu a
a Max-Planck-Institut für Kernphysik, Saupfercheckweg 1, 69117 Heidelberg, Germany
b Service de Physique Théorique, Université Libre de Bruxelles, Boulevard du Triomphe, CP225,
1050 Brussels, Belgium
Received 16 December 2014; received in revised form 27 April 2015; accepted 28 April 2015
Available online 30 April 2015
Editor: Tommy Ohlsson

Abstract
We study lepton flavor violating Higgs decays in two models, with the recently found hint for Higgs →
μτ at CMS as a benchmark value for the branching ratio. The first model uses the discrete flavor symmetry
group A4 , broken at the electroweak scale, while the second is renormalizable and based on the Abelian
gauge group Lμ − Lτ . Within the models we find characteristic predictions for other non-standard Higgs
decay modes, charged lepton flavor violating decays and correlations of the branching ratios with neutrino
oscillation parameters.
© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/). Funded by SCOAP3 .

1. Introduction

After the discovery of the Higgs boson in 2012 [1,2], the obvious next step is to check whether
the new particle behaves exactly as predicted by the Standard Model (SM). Expectations for
departure from SM behavior are based on the fact that a variety of new physics scenarios can

* Corresponding author at: Service de Physique Théorique, Université Libre de Bruxelles, Boulevard du Triomphe,
CP225, 1050 Brussels, Belgium.
E-mail addresses: [email protected] (J. Heeck), [email protected] (M. Holthausen),
[email protected] (W. Rodejohann), [email protected] (Y. Shimizu).

http://dx.doi.org/10.1016/j.nuclphysb.2015.04.025
0550-3213/© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/). Funded by SCOAP3 .
282 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

cause deviations. In particular in light of flavor symmetries, which seem necessary to explain the
peculiar structure of lepton mixing, one expects non-trivial Higgs decays, be it unusual decays in
SM particles or in new particles, see e.g. Refs. [3–6]. A particularly interesting possible departure
from the Higgs standard properties is flavor violation in its decays [7,8].
Indeed, in the first direct search for lepton flavor violating (LFV) Higgs decays, the CMS
Collaboration has recently reported an interesting hint for a non-zero branching ratio [9], namely
 
BR(h → μτ ) = 0.89+0.40−0.37 % . (1.1)
Translated into Yukawa couplings defined by the Lagrangian
−LY = yμτ μL τR h + yτ μ τ L μR h + h.c., (1.2)
with decay rate (h → μτ ) = (|yμτ |2 + |yτ μ |2 )mh /8π , one needs to explain values around

|yμτ |2 + |yτ μ |2  0.0027 ± 0.0006 . (1.3)
Though (1.1) represents only a 2.5σ effect, the measurement has caused some attention [10–14].
While the signal in Eq. (1.1) is not unlikely a statistical fluctuation, it is surely tempting to apply
flavor symmetry models to the branching ratio given above, to study the necessary structure of
models that can generate it, and to investigate other testable consequences of such models. At
least it demonstrates again that some flavor symmetry models have testable consequences outside
the purely leptonic sector, and that precision studies of the Higgs particle can put constraints on
such models. In this paper we show that the signal in Eq. (1.1) can be generated in two different
approaches based on quite different flavor symmetries: a continuous Abelian approach and a
more often studied non-Abelian discrete Ansatz.
It is clear that in order to enforce non-standard Higgs phenomenology one needs to intro-
duce new physics around the electroweak scale. The Higgs could also be the member of a larger
multiplet of states. These aspects occur frequently in flavor symmetry or other models, and in
particular in one of the approaches that we follow. Our first model applies the frequently used
non-Abelian discrete flavor symmetry group A4 , broken at the electroweak scale,1 and features
the Higgs particle as a member of a scalar A4 triplet. The second approach gauges the difference
between muon and tau flavor, Lμ − Lτ , and is therefore an anomaly-free Abelian gauge symme-
try. Both models have in common that there are additional Higgs doublets with non-trivial and
specific Yukawa coupling structure. They are distinguishable and falsifiable. We demonstrate that
charged lepton flavor violation bounds are fulfilled: the model based on gauged Lμ − Lτ is bro-
ken in such a way that only the μτ sector is affected, where constraints are in general weaker than
in decays involving electrons. The A4 model benefits essentially from a residual Z3 symmetry
that survives the A4 breaking, sometimes known as triality [15]. However, its breaking causes in
particular the decay μ → eγ , inducing constraints on the model. Anomalous Higgs decays other
than h → μτ are predicted, most noteworthy h → eτ , whose testable correlations with h → μτ
are governed by the model parameters. As the breaking of the respective flavor symmetry also
generates lepton mixing, we investigate the impact of the Higgs branching ratios on observables
in the neutrino sector. For example, the Abelian model links the chiral nature of the leptons in
the h → μτ decays with the octant of θ23 and the neutrino mass ordering.

1 As usual, the discrete symmetry group is broken in different directions at different scales. The “visible” breaking
takes place at the electroweak scale. For colliders, the neutrino masses are irrelevant and the other breaking is therefore
“invisible”.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 283

In what follows we first deal with the non-Abelian model based on A4 (Section 2), before
turning to the Abelian model in Section 3. We summarize our results in Section 4.

2. Non-Abelian case: an A4 example

Non-Abelian discrete flavor symmetries have been used to account for the large mixing an-
gles measured in the lepton sector [16,17]. The symmetry A4 is the smallest discrete group with
a 3-dimensional representation [18–23] and is therefore an economic and popular choice given
the three generations of leptons in the SM. In typical models the discrete symmetry is broken
to non-commuting subgroups, which form remnant symmetries of the charged lepton and neu-
trino mass matrices [24–26]. In the vast majority of models the breaking of the flavor symmetry
happens at very high and untestable scales.
Here we aim to employ non-Abelian discrete symmetries with a slightly different point of
view, namely we want to emphasize the possibility of additional phenomenology of non-Abelian
flavor symmetries at the electroweak scale [3,5,15,18,27–36]. Thus, instead of only concentrating
on predicting mixing angles, we have additional tests of models at our disposal, e.g. lepton flavor
violation in the Higgs sector.
Related to this topic there are two aspects of non-Abelian discrete symmetries that are worth
pointing out: first, embedding the SM Higgs in a multiplet of Higgs fields allows one to predict
the Yukawa couplings of the additional Higgs fields. We will put electroweak scalar doublets
into an A4 triplet, which then automatically induces LFV Higgs phenomenology. Second, the
often occurring possibility that breaking of A4 results in a remaining Z3 subgroup – which helps
obeying charged lepton flavor violating bounds – is also of use to us.
To make the presentation self-contained, we first remind the reader about ‘lepton triality’ [15]
and then discuss our model and the resulting phenomenology.

2.1. Lepton triality in A4 models

We here describe lepton triality [15], i.e. the Z3 subgroup typically conserved in the charged
lepton sector of A4 models where the Higgs transforms as a triplet 3 under A4 . The discrete
symmetry group A4 is the smallest group containing an irreducible 3-dimensional representation;
we use the basis
⎛ ⎞ ⎛ ⎞
1 0 0 0 1 0
ρ(S) = ⎝ 0 −1 0 ⎠ , ρ(T ) = ⎝ 0 0 1 ⎠ (2.1)
0 0 −1 1 0 0
and implement a model describing the lepton sector at the electroweak scale, following Refs. [3,
5,15,18,27–36], only caring about the charged lepton sector for now. The particle content is given
in Table 1. The necessary vacuum configuration for χ ≡ (χ1 , χ2 , χ3 )T ∼ 3,

0
χi  = √v , i = 1, 2, 3, (2.2)
6

can be naturally obtained from the most general scalar potential following the discussion in
Ref. [36]. Obviously these fields break the discrete symmetry group A4 down to the subgroup
T |T 3 = E ∼
= Z3 , while simultaneously breaking the electroweak gauge group SU(2)L × U(1)Y
down to the electromagnetic U(1)em . The normalization in Eq. (2.2) is chosen such that v cor-
284 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Table 1
Particle content of the minimal model that realizes flavor symmetry breaking at the electroweak scale, which may be
UV completed in the fashion of Ref. [35]. The flavon χ contains the Higgs field and ties the electroweak to the flavor
breaking scale.
eR μR τR χ ξ
A4 3 11 13 12 3 3 11
Z4 i i i i 1 −1 −1
SU(2)L 2 1 1 1 2 1 1
U(1)Y −1/2 −1 −1 −1 1/2 0 0

√ 2 √
2
responds to the SM value, i.e. v 2 ≡ i χi0 = 3 2 √v = ( 2GF )−1  (246 GeV)2 . The
6
charged lepton sector is described by the couplings2
−Le = ye ¯χeR + yμ ¯χμR + yτ ¯χτR + h.c. (2.3)
Because of the unbroken Z3 symmetry in the charged lepton sector it is useful to change to the
basis where this symmetry is represented diagonally:
 T  T
ϕ, ϕ , ϕ ≡ †T χ ∼ (1, ω2 , ω) , L ≡ Le , Lμ , Lτ ≡ †T ∼ (1, ω2 , ω) , (2.4)
with a unitary matrix T
⎛ ⎞
1 1 1
1 ⎝
T ≡ √ 1 ω2 ω ⎠ and ω ≡ e2πi/3 . (2.5)
3 1 ω ω2
In (2.4) we have indicated the transformation properties under the unbroken subgroup T  ∼
= Z3 ,
under which (eR , μR , τR ) transform as (1, ω2 , ω). This has been denoted flavor triality in
Ref. [15] and naturally suppresses flavor changing effects, which usually severely constrain
multi-Higgs doublet models. To see this, note that in this basis the vacuum configuration (2.2)
 √ T
implies that only the field ϕ acquires a vacuum expectation value (VEV) ϕ = 0, v/ 2 ,
while ϕ and ϕ are inert (VEV-less) doublets. In the basis of Eq. (2.4) the Yukawa terms read
   
−Le = ϕ ye L̄e eR + yμ L̄μ μR + yτ L̄τ τR + ϕ ye L̄τ eR + yμ L̄e μR + yτ L̄μ τR
 
+ ϕ ye L̄μ eR + yμ L̄τ μR + yτ L̄e τR + h.c. (2.6)
and we thus see that ϕ couples diagonally to leptons while ϕ and ϕ do not. The mass matrix,
defined by Le  = ēL Me eR with eL = − , is thus given by
v
Me = √ T diag(ye , yμ , yτ ) . (2.7)
2
Me is diagonal in the Z3 basis of Eq. (2.4), which therefore
√ corresponds to the charged-lepton
mass basis for the case of unbroken triality with y = 2m /v. As it stands, the model (which
was originally motivated from neutrino considerations) does not exhibit tree-level LFV Higgs
decays, as can be read-off of Eq. (2.6). The scalars ϕ, ϕ , and ϕ do not mix because they carry
different charges under the unbroken Z3 symmetry. Corrections to the VEV alignment (2.2) are

2 As there is only one A invariant that can be formed out of these fields, we do not specify the contraction here.
4
In ambiguous cases, we always specify the contraction.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 285

thus needed for LFV, as will be discussed in the next section.3 We will show later that lepton
mixing can successfully be reproduced in this model as well.
This model seems to be an excellent starting point when discussing Higgs LFV decays: first of
all, we have introduced multiple Higgses (which are a necessity for LFV, according to Paschos–
Glashow–Weinberg [37,38]) without introducing additional free Yukawa couplings; the Yukawa
couplings of the additional Higgses are not free, but rather dictated by lepton masses. Further-
more, there is a well-defined SM limit, which is the ‘lepton triality’ case, giving an ‘explanation’
for why we have not seen LFV processes yet. Finally, the tau Yukawa is the only large Yukawa
coupling and the model therefore predicts large LFV processes predominately in processes in-
volving taus.

2.2. Perturbation to the vacuum alignment

The potential for the electroweak doublets χ ∼ 3 is given by4

  
Vχ (χ) = μ2χ χ † χ + λχr (χ † χ)r (χ † χ)r ∗ + λχA Im (χ † χ)3S (χ † χ)3A , (2.8)
r=11,2 ,3S,A

which leads to the VEV of Eq. (2.2) for a certain choice of parameters (see for example Ref. [36]
and references therein). In the following, we will always present results in the limit λχA = 0,
which simplifies the mixing in the scalar sector. We do not expect qualitative changes for small
non-zero λχA , merely additional small mixing among the scalars.
The choice λχA = 0 lets the potential gain another symmetry, namely the exchange of χ2 and
χ3 , generated by the Z2 generator
⎛ ⎞
1 0 0
ρ(U ) = ⎝ 0 0 1 ⎠ . (2.9)
0 1 0
Together with A4 , this leads to an S4 symmetry of the potential, which protects λχA = 0 from
corrections of the other scalar couplings. However, as the Yukawa couplings do not respect this
symmetry, the (technically) natural size of λχA is of the order yτ4 /(16π 2 ).
To discuss symmetry breaking, we should also discuss how the symmetry is implemented in
the neutrino sector. Following standard literature, we assume the existence of a scalar singlet field
∼ 3 (see Table 1) to break the A4 symmetry in the (1, 0, 0) direction, as well as an A4 singlet
ξ which breaks the Z4 . Since we are interested in a phenomenological analysis, we assume the
following VEV hierarchy:
  v. (2.10)
The alignment then proceeds as follows:

• The potential for is decoupled from the other scalars and obtains a VEV   ∼ (1, 0, 0).
This is a natural outcome for a large range of potential parameters (see e.g. [39, p. 34] or
[40] and references therein).

3 The only LFV lepton decays allowed by the Z are τ ± → μ± μ± e∓ and τ ± → e± e± μ∓ , others being induced
3
exclusively by breaking of triality [15].
4 See Ref. [35] for a definition of the various Clebsch–Gordon coefficients and the notation. r ∗ is the complex conjugate
representation, i.e. r ∗ = r except for 1∗2 = 13 .
286 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

• The interaction λm ( )13 (χ † χ)12 is the only term communicating the A4 breaking to χ .
Effectively, this results in the soft-A4 -breaking term
 
λm ( )13 (χ † χ)12 + h.c. → MS2 (χ † χ)12 + h.c. (2.11)

in the scalar potential of χ , that has to be added to Vχ (χ) in Eq. (2.8). Let us remind the
reader that the VEV of points in the (1, 0, 0) direction, so the VEV of is only non-zero
when coupled to a singlet, i.e. ( )3  = 0. A trivial singlet ( )11 just redefines μ2χ in
Vχ , so the above is the only relevant coupling.
• The inclusion of Eq. (2.11) then leads to a VEV shift in χ (without back-reaction on ) with
the following structure:
 T  v
χ ∼ (1 + 2ε, 1 − ε, 1 − ε) ⇔ ϕ, ϕ , ϕ = √ (1, ε, ε) , (2.12)
2
where ε ∝ MS2 /v 2 , defined properly below in Eq. (2.15). We thus need the soft A4 breaking
below the electroweak scale, which can be achieved with small λm despite the hierarchy of
Eq. (2.10). Note that these VEVs are in the CP-even neutral direction.

This triality-breaking VEV correction (2.12) with identical entries in χ2 and χ3 is a consequence
of the symmetry U of the potential, which is left invariant by this VEV. Its form has been ob-
served before in alignment models with driving fields [22] and non-trivial group extensions [36].
Contrary to the philosophy employed in those references, we do not assume ε  1, and therefore
rather use the parametrization5
 T  v 1 1
ϕ, ϕ , ϕ = √ cβ , √ sβ , √ sβ . (2.13)
2 2 2
A non-zero β will give rise to lepton flavor violating Higgs decays as well as rare leptonic
decay modes, e.g. i → j γ , otherwise forbidden by triality (see also footnote 3 on p. 285).
Since
 the VEV structure (2.13) leaves invariant the generator U it makes sense to define ψ1,2 =
√1 ϕ ± ϕ . Of these additional two Higgs doublets, only ψ1 develops a non-vanishing VEV:
2
ψ1  ∼ ε. Effectively, we therefore have a 2HDM-like model with an additional VEV-less doublet
ψ2 .6
To see precisely how MS of Eq. (2.11) leads to the quoted VEV configuration of Eq. (2.13),
we consider the minimization conditions ∂V ∂η = 0, where η is any of the scalar fields including
their neutral components ϕ 0 and ψ1,20 . Assuming the form Eq. (2.13), they all vanish except for

∂V 1 √ 
0= ⇒ μ2χ = − v 2 3λχ 11 + λχ 31,S , (2.14)
∂ϕ 0 3
∂V 1  √  √ 
0= ⇒ MS2 = − v 2 sβ sβ + 2 2cβ 3λχ 12 − λχ 31,S . (2.15)
∂ψ10 12
This shows that the VEVs can be obtained from the potential once one adds a soft-breaking
term (which may originate from the coupling to the neutrino-flavon as in Eq. (2.11)). Note

5 We use the standard abbreviations c = cos β, s = sin β and t = tan β.

β β β
6 Care has to be taken when comparing our tan β to other two-Higgs-doublet models (2HDMs), as the replacement
tan β → 1/ tan β can easily be more appropriate depending on the fermion couplings.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 287

the simplicity of the minimization conditions as a result of the non-Abelian symmetry of the
model. The scalar mass spectrum will lead to the conditions λχ 31,S < 0 (see Eq. (2.17)) and
√
λχ 12 > λχ 31,S / 3 (see Eq. (2.19)), while MS2 can take on any sign. The sign difference between
ϕ 0  and ψ10  – the sign of β – is physical and cannot be rotated away, as the Higgs fields
originate from the same multiplet. For small β  1, we find from Eq. (2.15)
√ √
−3 2 MS2 2 MS2
ε = sβ  β  √  − √ , (2.16)
3λχ 12 − λχ 31,S v 2 3 M2
as expected from the observation that MS → 0 reinstates triality. For the last equation we already
inserted the scalar mass M, to be introduced in the next section (see Eq. (2.19)). Since values
of interest to explain the CMS excess in h → μτ lie around |β| ∼ 0.2, we will actually only
occasionally make use of the small-β limit to gain analytic insights but otherwise use the full
expression for β.

2.3. Scalar masses

After symmetry breaking, the nine physical scalars contained in χ arrange themselves in the
following multiplets under the remnant U(1)em × Z3T × Z2U symmetry of the χ ’s: The first four
degrees of freedom are in the charged scalars H + = cβ ψ1+ − sβ ϕ + and ψ2+ , which both have the
mass
λχ 31,S 2
m2H + = − √ v . (2.17)
2 3
The quartic coupling λχ 31,S is thus required to be negative for an electrically neutral vacuum,
which leads to consistency conditions on the parameters of Eq. (2.8)
√ by demanding bounded-
ness of the potential. The next two degrees of freedom are A = 2(cβ Imψ10 − sβ Imϕ 0 ) and
√
2Reψ20 , which are degenerate with mass
λχ 31,A 2
m2A = m2H + − √ v . (2.18)
2 3
√
We also have the neutral state 2Imψ20 with mass
√ 1  1 √
m2 ( 2Imψ20 ) = λχ 12 v 2 + 2m2H + (3 + c2β − 2 2s2β )
3 4
1 √
≡ M2 (3 + c2β − 2 2s2β ) . (2.19)
4
In the last
√line we defined a new mass parameter M for convenience, which corresponds to the
mass of 2Imψ20 in the triality limit β → 0. The final two real scalars sit in the two complex
neutral scalars ψ10 and ϕ 0 that acquire VEVs. The mass eigenstates are given by the neutral
scalars
√ 0
H
=
cα sα √ 2Reϕ 0 , (2.20)
h −sα cα 2Reψ1
with masses m2h = (m0h )2 −  and m2H = (m0H )2 + . We can express the last remaining potential
parameter in terms of physical quantities:
288 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Fig. 1. cos(α − β) as a function of mH and β. In the parameter regions where the contour is white, some masses are
negative/imaginary, and the VEV is not a minimum.

2  1  √ 
(m0h )2 = λχ 11 v
2
− 2m2H + , (m0H )2 = M 2 2 2s2β − c2β + 5 (2.21)
3 4
and
 √ 
2
(m0h ) sβ2 4 2s2β + 7c2β + 9  4

= √ + O (m0h ) /(m0H )2 . (2.22)
2 2s2β − c2β + 5
Positivity of masses restricts the values of β, see Fig. 1. Note that the mass splitting is predicted
in terms of the other scalar masses; this non-trivial relation is due to the fact that there is a smaller
number of parameters in the scalar sector than in the general case, courtesy of the non-Abelian
flavor symmetry. In the same vein, the mixing angle α is predicted in terms of scalar masses:
   √ 
2
4sβ 2 M 2 + (m0h ) cβ + 2M 2 sβ )
tan 2α = −    √ . (2.23)
3M 2 − 4(m0h )2 c2β + M 2 2 2s2β + 1

Note that the CP-even Reψ20 does not mix with H and h because it is odd under the Z2U we
obtained by setting λχA = 0. Since the Z2U is broken by the Yukawa interactions, Reψ20 is not
stable and will mix with h and H at loop level. The same comment applies to the mixing of the
charged scalars and pseudoscalars. We will neglect this complication, which is anyways expected
to give only small modifications to our results.
The state h will play the role of the SM-like Higgs particle that has been produced at the LHC.
The limit of cos(α − β) = 0 is the SM limit, as in other 2HDMs [41]. We can eliminate m0h by
using (125 GeV)2  m2h = (m0h )2 −  and therefore end up with the free parameters mH + , mA ,
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 289

M and β. Note that we have cos(α − β)  −2β in the limit of small β and mH mh (see Fig. 1).
The parameters mH + and mA are not particularly important for the following discussion and can
be made large to evade experimental constraints (see also the discussion for the Abelian model
in Section 3). Lower limits on mH + typically range from 90 GeV (LEP) up to O(300) GeV
(B physics) [42], but depend strongly on the H + couplings to quarks, which are not specified in
our model (see Section 2.5). Similar comments apply to mA . √
As a numerical example, we consider β = 0.2 and M = 400 GeV, which leads to  
51 GeV, m0h  135 GeV – in order to obtain the Higgs mass mh = 125 GeV – mH  460 GeV
and the scalar mixing angle sin α  −0.98 (and hence cos(α − β)  −0.4). Keep in mind that
our
√ notation for α and β is somewhat different from the standard 2HDM notation. The state
2Imψ20 has mass 336 GeV, whilst the other four scalars have masses that depend on an addi-
tional coupling (Eqs. (2.17) and (2.18)). The soft-breaking parameter from Eq. (2.15) is given by
MS2  −(200 GeV)2 .

2.4. Lepton masses

With the Lagrangian of Eq. (2.3) and VEV structure of Eq. (2.13) we find the charged-lepton
mass matrix
    
v ye sβ y μ yτ
Me = √  T cβ yμ +√ ye yτ , (2.24)
2 yτ 2 y y
e μ

which reduces to the matrix of Eq. (2.7) in the triality limit β → 0. The off-diagonal mass-matrix
elements all scale with sβ and their relative magnitude
√ is fixed by the charged lepton masses
(for small β we have the SM-like relations y  2m /v). In particular, the eτ and μτ entries
dominate and have the same magnitude, which will ultimately lead to large rates for h → μτ , eτ
of similar magnitude, discussed below. We go to the charged-lepton mass basis eL 0 , e0 ,
R

eL = VeL eL
0
, eR = VeR eR
0
, (2.25)
where the unitary matrices satisfy

Ve†L Me VeR = diag(me , mμ , mτ ) . (2.26)

Both mixing matrices will be functions of β and the lepton masses. A good approximation to the
left-handed rotation matrix can be parametrized as follows
⎛ 1 ⎞
−√ √1 0
2 2
⎜ ⎟
VeL ≡ T WL  T R O23 (β)R T O12 (αL ), R ≡ ⎝ √1 √1 0 ⎠, (2.27)
2 2
0 0 1
where T is defined in Eq. (2.5). WL describes the deviation from the triality case β = 0, which
just has VeL = T . Here we have expanded in small Yukawa couplings (ye  yμ  yτ ), but not
in small values of β. The Oij denote rotations in the ij plane, and we have
 √ √ 
sβ −3sβ − 7s3β + 12 2cβ + 4 2c3β
tan 2αL  √ , (2.28)
8 2sβ3 + 6cβ + 10c3β
290 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Fig. 2. Relevant couplings in the charged-lepton sector as functions of the triality-violating angle β. Left: Off-diagonal
charged-lepton mixing matrix elements |(WL )ij |2 , with (WL )31  (WL )21 and (WL )13  (WL )23 . Right: Yukawa cou-
√
plings of the charged leptons relative to the SM values yαSM = 2mα /v (see Eq. (2.24)), as well as the angle αL of WL
(see Eq. (2.28)).

2  
or approximately αL = √β − 3β4 + O β 3 , which is true to relative order in small Yukawas and
2
to leading order only depends on β. For small β, this simply yields
 √   √ √ 
1 αL β/√2 1√ β/ 2 β/√2
WL  −α√L 1√ β/ 2  −β/√2 1√ β/ 2 , (2.29)
−β/ 2 −β/ 2 1 −β/ 2 −β/ 2 1
which gives non-negligible contributions to the Pontecorvo–Maki–Nakagawa–Sakata (PMNS)
mixing matrix for the values required to explain the CMS excess (as we will see, values of
interest are around |β| ∼ 0.2). The approximation of Eq. (2.29) is pretty good for the 13 and 23
elements of WL , but quickly breaks down for all others, see Fig. 2. This is where our definition
of αL kicks in. Note that our parametrization of WL from Eq. (2.27) obeys (WL )23 = (WL )13 ,
which is valid to order m2μ /m2τ (see Fig. 2) and (WL )31 = (WL )21 , valid to order m2e /m2μ . These
are dictated by the flavor structure in Me with its equal 23 and 13 elements, etc. (see Eq. (2.24)).
The right-handed mixing angles are all suppressed by small Yukawas and it therefore suffices
to expand in first order:
⎛ √ √ ⎞
1 − 2 yyμe sin β − 2 yyτe sin β
⎜√ y √ y ⎟
VeR  ⎝ 2 yμe sin β 1 − 2 yμτ sin β ⎠ . (2.30)
√ ye √ yμ
2 yτ sin β 2 yτ sin β 1
The Yukawa couplings y deviate from their SM values for β = 0; the relative corrections are
larger for the first and second generation Yukawa couplings, with a behavior at small β reading
me   mμ  
ye  √ 1 + 2β 2 , yμ  √ 1 + β 2 ,
v/ 2 v/ 2
 
mμ2
mτ
yτ  √ 1 − 2 β 2 . (2.31)
v/ 2 mτ
√
The relations between the Yukawa couplings yα and their SM values of yαSM = 2mα /v are
shown in Fig. 2.
For the neutrino sector, we have introduced a scalar field ∼ 3 that breaks the group A4 to
the subgroup generated by S of Eq. (2.1), and therefore has a VEV in the (1, 0, 0) direction [36].
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 291

Its Z4 charge is −1 in order to couple only to neutrinos, similar to the A4 singlet scalar ξ . Using
the particle content of Table 1 we then obtain the leading order effective operators
       
L ⊃ xa T σ 2 σ χ T σ2 σ χ ξ + xd T σ 2 σ χ T σ2 σ χ
11 11 3 11
         
+ xe T σ2 σ χ T σ2 σ χ ξ + xbi T σ2 σ χ T σ2 σ χ
3 3
i=2,3
1i 3 1∗i
    
+ xci T
σ2 σ χ T σ2 σ χ ξ + h.c., (2.32)
1i 1∗i
i=2,3
where   ∼ (1, 0, 0) and the xj have mass dimension −2. The Majorana neutrino mass matrix
is then given by
⎛ ⎞
a + b 2 + b3 e e
Mν = ⎝ e b 3 ω 2 + b2 ω + a d +e ⎠, (2.33)
e d +e b 2 ω + b3 ω + a
2

with
1 1  √  √ 
a = v 2 xa |ξ | , d = v 2 4 3xd | | + 6xe |ξ | sβ sβ − 2cβ ,
3 24
1 2 √ 
e = v xe |ξ | 2s2β + 4c2β ,
24
1  √  1  √ 
bi = v 2 xbi | | −2 2s2β + c2β + 3 + v 2 xci |ξ | sβ sβ + 2 2cβ . (2.34)
36 12
The matrix is diagonalized by going to the mass basis νL = Vν νL0 with
VνT Mν Vν = diag(mν1 , mν2 , mν3 ) , (2.35)
leading to the unitary PMNS matrix U ≡ Ve†L Vν = WL† †T Vν
relevant for charged-current inter-
actions.
In the limit b2 = b3 the matrix Mν becomes μ–τ symmetric and hence gives a Vν with θ13 ν =0

and θ23 = π/4 (setting further b2 = b3 = d/3 gives tri-bimaximal mixing (TBM) values in Vν ,
ν

i.e. additionally sin2 θ12ν = 1/3). Neglecting the triality-breaking W would then result in U 
L
†
T Vν with θ13 = 0 and θ23 = π/4, incompatible with current data [43]. Triality breaking WL = I
√
contributes corrections of order β/ 2 (see Fig. 2), and thus roughly of order θ13 when the CMS
excess is to be explained (β ∼ 0.2, see below). One could thus hope to take the μ–τ -symmetric
(or TBM) limit in Mν as a starting point and use the WL corrections to generate a non-zero θ13 .
Unfortunately this does not work; the reason for this is the relation (WL)31 = (WL )21 (see Fig. 2),
ultimately√due to the mass matrix structure in Me (Eq. (2.24)). This gives U13 = ((WL )31 −
(WL )21 )/ 2 ∼ m2e /m2μ , so θ13 is highly suppressed (θ13  4 × 10−6 for β = 0.2). WL does
√
hence lead only to β/ 2 corrections to θ12 and θ23 .
We thus need a μ–τ -asymmetric (non-TBM) structure in Vν , easily accomplished for b2 = b3
(= d/3). If all the xj are of similar order, this means in particular that the VEVs of and ξ should
be non-hierarchical,   ∼ ξ , to get a large enough θ13 . The mass matrix Mν in Eq. (2.33) has
sufficient parameters to fit the present global data, so we omit a detailed discussion. The flavor
symmetry can then no longer predict specific values for mixing angles (and/or sum-rules for
neutrino masses [44]), but rather just motivate the mixing angle hierarchy. Definite predictions
arise, however, in the LFV observables, as discussed below.
292 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

2.5. Quark couplings

Having discussed the lepton sector of the model, which serves as a major motivation for
the discrete flavor group Ansatz, we turn to the other fermions. To extract experimental limits
on the scalars, in particular the SM-like h, one has to take the quark sector into account. So
far, all introduced scalars carried charges under the flavor group A4 × Z4 in order to generate
viable lepton mixing patterns. Having treated h as the potential candidate for the 125 GeV scalar
discovered at the LHC, we have to specify its couplings to quarks and how quark masses/mixing
arises in our model. This is important, because the very same scalar particle that we study below
via its h → μτ decay has been observed to decay/couple to third-generation quarks, forcing
us to include quarks in our discussion. While the coupling of h to bottom quarks is not yet
established at a statistically significant level (around 1–2σ [45,46]) and the top-quark couplings
are so far only inferred indirectly (e.g. via the loop-induced gluon production rate of h), we will
not entertain the ludicrous idea of h not coupling to quarks. Two qualitatively different scenarios
emerge [35]:

1. Including the quarks in the flavor group and generating their masses by the VEV of χ . One
possibility is to generate quark masses analogously to lepton masses, by putting QL ∼ 3 and
uiR , dRi ∼ 1i , which gives the couplings

−LQ = yd Q̄L χdR + ys Q̄L χsR + yb Q̄L χbR + yu Q̄L χ̃ uR + yc Q̄L χ̃ cR

+ yt Q̄L χ̃ tR + h.c.,
in complete analogy to the charged leptons. For simplicity we insert the SM Yukawa cou-
plings yq in the above formula. For vanishing β (triality limit) one finds a trivial Cabibbo–
Kobayashi–Maskawa (CKM) matrix VCKM = I. A non-zero β  0.2 introduces non-trivial
mixing, but too small to accommodate the rather large Cabibbo angle. This can already
be observed from the diagonalization in Eq. (2.27): the left-handed rotations VuL and VdL
will depend to high accuracy only on β and not on the Yukawa couplings/masses, so even
though both rotations have off-diagonal entries O(β), the overlap VuL Vd†L remains very close
to I (for β = 0.2 the Cabibbo angle is  6 × 10−4 ). Consequently,  one has to introduce a
higher-dimensional operator to generate viable CKM mixing, e.g. QL χ 3 ξ dR /2 . This
introduces an additional parameter, which we can adjust to reproduce the Cabibbo angle.
As an example, we give the CKM matrix for β = 0.2 and operator strength | ξ | /2 =
7 × 10−4 ,
⎛ ⎞
0.972356 0.233472 0.003771
|VCKM | = ⎝ 0.233458 0.972356 0.004559 ⎠ , (2.36)
0.004534 0.003801 0.999982
which does not appear to be completely unrealistic. Here, we have only used SM Yukawa
couplings and have not fitted all parameters of the theory. Obviously, including other opera-
tors will allow us to fit the CKM matrix to even better precision, and also to include the quark
masses. Flavor-violating Higgs decays will also be induced, suppressed by small Yukawas,
and more importantly heavily depending on the various possible higher-dimensional oper-
ators and on details of charge assignments. We are confident that such an analysis can be
performed and the point of this discussion is to outline ways of how this can be achieved.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 293

2. Introduction of an additional scalar doublet H , uncharged under the flavor group, which
couples to quarks in the usual manner and acquires a VEV H  = 0. The VEV of χ (split
among its components with angle β as in Eq. (2.13)) is then no longer fixed to yield 246 GeV,
but we rather have H 2 + i χi0 2 = (246 GeV)2 , so another angle β has to be introduced
in order to describe the ratio H 2 / i χi0 2 . New scalar mixing angles α arise as well,
giving rise to a rather large parameter space.

We conclude that while the quark sector of this model is not completely satisfactory, options exist
which can make the framework holistic, and which can render the quark part largely decoupled
from the lepton part. With our main focus on lepton flavor physics phenomenology, we leave the
discussion on the quark sector as it is. Independent of the fermion couplings one can set a limit
of | cos(β − α)| < 0.45 at 95% C.L. using the vector boson couplings of h alone [47]. This is the
minimal bound employed in this paper.

2.6. Higgs interactions

The SM state that is carrying

√ all the VEV,
√ and therefore couples with SM strength to the gauge
bosons, is given by HSM = 2cβ Reϕ 0 + 2sβ Reψ10 . The coupling to gauge bosons of the state
h is therefore suppressed when compared to the SM,
ghW W ghZZ
SM
= SM = sin(β − α) , (2.37)
ghW W ghZZ
just like in other 2HDMs. As discussed above, measurements of the Higgs–vector boson cou-
plings give a limit | cos(β − α)| < 0.45 at 95% C.L., typically strengthened depending on the
underlying 2HDM couplings to fermions. We will not perform a scan of the currently allowed
parameter range of our 2HDM-like model, but rather focus on the LFV aspects. We always dis-
play the employed cos(β − α) (which is closely related to the triality-violating angle β, but easier
to access experimentally in this form) to enable cross checks with LHC results.
The interaction Lagrangian of the lepton mass eigenstates with the Higgs mass eigenstate h is
given by
−Lhf¯f = yαβ ēLα
0 0
heR β + h.c. (2.38)
and the parameters are given to first order in O(yμ /yτ ) as (see (2.31) and Fig. 2)
mτ mτ
yeτ = √ (cαL − sαL )cα−β , yμτ = √ (cαL + sαL )cα−β ,
2v 2v
mτ
yτ τ = sβ−α . (2.39)
v
All other couplings vanish in this approximation. Note again that the couplings become SM-like
for cβ−α → 0 as in other 2HDMs. The third LFV coupling is suppressed by the muon Yukawa
coupling
yμ   √ 
yeμ = cα cβ (cαL − sαL ) + 2sβ (cαL − sαL ) + sαL + cαL
4    
√
+ 2 2 sin α sec β sαL (3 cos β − 2)cβ/2
2
+ cαL sβ/2
2
(3cβ + 2) + O yμ2 , (2.40)
and hence small, but tightly constrained by the LFV decay μ → eγ . While not obvious in any
way, yeμ also vanishes in the limit cβ−α → 0. With the off-diagonal μτ couplings of h at our
294 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Fig. 3. The leptonic branching ratios h → i j as a function of cos(β − α). The shaded horizontal areas denote the 1σ
and 2σ ranges of CMS for h → μτ (red) [9] and h → τ τ (green, only 1σ ) [48]. There is a small dependence on the
heavy Higgs mass mH through the scalar mixing angle αL (Eq. (2.23)): solid (dashed) lines are for mH = 200 GeV
(800 GeV). The shaded vertical areas |cβ−α | ≥ 0.45 are excluded by the conservative limits from Higgs–vector–vector
measurements [47]. The SM-values are recovered for cos(β − α) = 0. (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)

disposal, we can determine the parameter values necessary to explain the CMS excess (Eq. (1.3))
as
mτ
|yμτ | = | √ (cαL + sαL )cα−β |  7 × 10−3 | sin(αL + π/4) cos(α − β)|
2v
!
 3 × 10−3 . (2.41)
Note that the chiral coupling μL τR h dominates the decay h → μτ in this model. The branching
ratio depends only on the parameter β (slightly on mH due to the scalar mixing angle α, see
Fig. 1), but we show it as a function of cos(β − α) in Fig. 3 because this quantity is directly
related to the Higgs couplings to vector bosons. We see that rather large values |cβ−α |  0.4
(|β|  0.2) are required to describe the CMS excess. Because of the few free parameters in our
flavor model, this has direct consequences for other LFV rates. For one thing, the LFV rate
h → eτ is expected to be close to the h → μτ rate,
2
BR(h → μτ ) cαL + sαL β=0.2
 = tan2 (αL + π/4) −−−−→ 1.59 . (2.42)
BR(h → eτ ) cαL − sαL
A sensitivity to h → eτ of similar order as h → μτ seems feasible at the LHC [8], even though
a dedicated analysis has so far only been performed in the μτ channel. The above prediction
will thus serve as the most important discriminator between models once this channel has been
probed. The rate h → eμ is suppressed by yμ2 /yτ2 and hence unobservably small compared to
the other two LFV channels. The flavor conserving rates h → μμ, τ τ are reduced in this model,
but only slightly so (compared to the Abelian explanation of the CMS excess in Section 3). The
h → τ τ rate lies comfortably in the 1σ region of CMS [48]: 0.78 ± 0.27 (relative to the SM), as
shown in Fig. 3.
Not only the Higgs–vector–vector coupling limit | cos(β − α)| < 0.45 constrains the non-
Abelian CMS explanation, the induced LFV rate μ → eγ further impacts our model, as we will
discuss now, and actually excludes the region of interest with positive cβ−α .
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 295

2.7. Bounds from indirect measurements

The most stringent constraints are expected from the rare decays μ → eγ [49] and τ →
μγ [50]. Note that the decay τ → μγ gives typically stronger limits on the scalar sector than
τ → 3μ, even though the experimental limit on the branching ratio is a factor ∼2 weaker. This is
because τ → 3μ is either suppressed by an additional muon Yukawa coupling (tree-level scalar
exchange) or fine-structure coupling (off-shell photon in τ → μγ → 3μ) [8]. The same holds
for μ → eγ vs. μ → 3e. The Wilson coefficients cL and cR , which affect the rate for τ → μγ as
αm5τ  
(τ → μγ ) = 4
|cL |2 + |cR |2 , (2.43)
64π
are given at one-loop as

cL = F (mτ , mα , mμ , ms , 0, Ys ) , (2.44)
α=e,μ,τ
s=h,H,A,Reψ2 ,Imψ2

cR = F (mτ , mα , mμ , ms , 0, Ys† ) , (2.45)
α=e,μ,τ
s=h,H,A,Reψ2 ,Imψ2

with Yukawa coupling matrix Ys of scalar s and the loop function F given in Eq. (A.1) of Ref. [8].
The corresponding equations for μ → eγ can be obtained by obvious replacements. Note that
these complicated expressions only depend on β, mA and M (or equivalently, mH ) as free pa-
rameters.
We find the most constraining bound to come from μ → eγ with recent MEG result [49]
BR(μ → eγ ) < 5.7 × 10−13 at 90% C.L., (2.46)
see Fig. 4. We also plot the relevant branching ratios, BR(h → μτ ) and BR(μ → eγ ), against
each other in Figs. 5 and 6. The MEG bound is so strong that it forbids a resolution of the CMS
excess for positive cos(β − α) (negative β); A cancellation among the scalar contributions to
μ → eγ occurs however for negative cos(β − α) (positive β) for mH  300–400 GeV, opening
up parameter space in CMS’ 1σ region for cβ−α  −0.4.
Note that two-loop contributions to the radiative lepton decays i → j γ can be dominant
in some parts of parameter space because the stronger scalar coupling to top quarks or vector
bosons compared to leptons can compensate the additional loop suppression. Since this requires
a specific model for the quark couplings we do not take it into account here, but this will pose a
challenge for the way quarks are included in the model; Ref. [8] found that an SM-like h with
LFV couplings would have dominating two-loop contributions to μ → eγ , ultimately resulting
in BR(h → eμ)  10−8 , orders of magnitude below our prediction (see Fig. 3). We most likely
need some fine-tuning to suppress μ → eγ in our A4 model once we take quark couplings and
two loops into account.
The constraint from μ → eγ is shown in Fig. 4 on top of the relevant parameter space
for h → μτ . For further visualization of the parameter space and constraints we directly plot
BR(h → μτ ) against BR(μ → eγ ), fixing either β (Fig. 5) or mH (Fig. 6). We observe again
the cancellation that suppresses μ → eγ for certain values of mH and β. Seeing as our model
demands a large cos(β − α) ∼ −0.4 to explain the CMS excess in h → μτ and a rather light
‘heavy’ Higgs mH  280–380 GeV for sufficient one-loop cancellation of μ → eγ , it might be
possible to see H at the LHC. The specifics depend strongly on the employed quark couplings
and will be left for a future publication.
296 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Fig. 4. Relevant parameter space of the model in order to explain the CMS excess in h → μτ (colored regions, the light
lines give steps in 0.001). The dashed red contour denotes the 90% C.L. bound from μ → eγ (MEG [49]); the region
inside is allowed. The mass of A is taken to be mA = 600 GeV. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)

Fig. 5. Branching ratios of h → μτ vs. μ → eγ . Horizontal lines are best-fit value and the 1σ or 2σ ranges for the Higgs
branching ratio, see Eq. (1.1). The vertical line is the MEG bound on μ → eγ [49]. The various lines correspond to the
different values for β indicated in the plot; color coding is in mH . mA is fixed to mA = 600 GeV.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 297

Fig. 6. Branching ratios of h → μτ vs. μ → eγ . Horizontal lines are best-fit value and the 1σ or 2σ ranges for the Higgs
branching ratio, see Eq. (1.1). The vertical line is the MEG bound on μ → eγ [49]. The various lines correspond to the
different values for mH indicated in the plot; color coding is in cos(β − α). mA is fixed to mA = 600 GeV.

3. Abelian case: an Lμ − Lτ example

In the second part of this paper, we study the realization of h → μτ in the framework of
Abelian flavor symmetries, specifically U (1)Lμ −Lτ . Not only is this an anomaly-free global sym-
metry within the SM [51–53], it is also a good zeroth-order symmetry for neutrino mixing with a
quasi-degenerate mass spectrum, predicting maximal atmospheric and vanishing reactor mixing
angles [54–56]. Breaking of Lμ − Lτ is, of course, necessary for a viable neutrino sector, and
can also induce the (Lμ − Lτ ) = 2 process h → μτ , as we will show below. This will also
lead to the lepton-flavor-violating decays τ → 3μ and τ → μγ [57,58]. Since the Z of a gauged
U (1)Lμ −Lτ does not couple to first generation fermions, the experimental limits are not as strin-
gent as for other U (1) models, and it might even be possible to use (a light) Z to resolve the
longstanding 3–4σ anomaly surrounding the muon’s magnetic moment [58–66]. An even lighter
Z may induce long-range forces modifying neutrino oscillations [67], although this is not the
limit of interest here.
We work within gauged U (1)Lμ −Lτ with three right-handed neutrinos Ne,μ,τ , qualitatively
similar to Ref. [58]. For symmetry breaking, we introduce two scalar doublets 1,2 , with Lμ −Lτ
charge −2 and 0, respectively, as well as an SM-singlet scalar S with Lμ − Lτ charge +1 (see
Table 2). A small VEV of 1 – induced by the larger VEV of S that generates right-handed
neutrino masses – will break Lμ − Lτ by two units in the charged-lepton sector and subsequently
lead to the LFV decay mode h → μτ . A particular feature of this model is LFV only in the μτ
sector, evading strong constraints from, e.g., μ → eγ . This is opposite to the model Ref. [58],
where 1 was given the Lμ − Lτ charge +1, leading to charged-lepton processes with (Lμ −
Lτ ) = ±1, with (Lμ − Lτ ) = ±2 being highly suppressed. We will comment on variations of
our model in Section 3.7, which have a similar structure but different phenomenology.
298 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Table 2
Particle content of the Lμ − Lτ model; quarks are uncharged under the new U (1). j and S denote the scalar bosons of
the model, uncharged under the color group SU(3)C .
Le Lμ Lτ eR μR τR Ne Nμ Nτ 1 2 S
U(1)Lμ −Lτ 0 1 −1 0 1 −1 0 1 −1 −2 0 1
SU(2)L 2 2 2 1 1 1 1 1 1 2 2 1
U(1)Y −1/2 −1/2 −1/2 −1 −1 −1 0 0 0 1/2 1/2 0

3.1. Scalar potential and Yukawa couplings

With the particle content from Table 2, the scalar potential takes the form
λ1 λ2 †
2 , S) = m1 | 1| +
2 | 1 | − m2 | 2 | + 2 | 2 | + λ3 | 1 | | 2 | + λ4 | 2|
2 2 4 2 2 4 2 2 2
V( 1, 1
− μ2S |S|2 + λ2S |S|4 + λ 1 S | 1 |2 |S|2 + λ 2 S | 2 |2 |S|2
− δ S 2 †2 1 + h.c. (3.1)
The scalar S acquires a high-scale VEV, and for simplicity we assume it also to be heavy and
have negligible mixing with the other scalars (similar to the flavon field in the A4 model, see
Eq. (2.11)). In this limit, we can simply consider the effective 2HDM potential (after renaming
coefficients)
λ1 λ2 †
2 )  m1 | 1| + 2 | 1| − m22 | 2| + 2 | 2| + λ3 | 1| | 2| + λ4 | 2|
2 2 4 2 4 2 2 2
V( 1, 1
(3.2)
†
− m23 2 1 + h.c., (3.3)
which is just a U (1)-invariant 2HDM [41], softly broken by the mass-mixing term in the last
line, m23 ≡ δS2 , again similar to the soft-breaking term MS2 in the A4 potential.7 Our choice
U (1)Lμ −Lτ acts here as a very simple anomaly-free horizontal symmetry in the scalar potential
(see Ref. [68] for other U (1)H choices). In Section 3.2 we will see that S contributes to the
right-handed neutrino masses, and is therefore expected to be close to the seesaw scale, at least
S v. We work with a low-scale seesaw in mind in order to have more interesting Z phe-
nomenology (MZ  g S), but a high-scale seesaw is of course possible. In this case, δ might
have to be chosen very small if we still want the new scalars to be at the electroweak scale. In
this regard we note that δ → 0 would lead to an additional global U (1) symmetry in the scalar
potential, but not in the full Lagrangian, so a small δ is not technically natural; loop contributions
to this operator arise at one loop, see for example Fig. 7.
With positive m21,2 , 2 acquires a VEV from its Mexican-hat potential, and the m23 term
subsequently induces a small VEV for 1 :  1    2 m23 /m21 , where we neglected the portal
couplings λ. We will assume the hierarchy tan β ≡  2  /  1  = v2 /v1 1 in the following,
as this suffices for our purposes. Again neglecting the portal terms, the new scalars contained
in 1 , namely the heavy CP-even H , the CP-odd A, and the charged H + , are then degenerate
with mass m2A = m23 /sβ cβ  m21 .
More accurately, the charged scalar has mass m2+ = m2A − λ4 v 2 , whereas the neutral CP-even
√
scalars hj inside j = (φj+ , (vj + hj − izj )/ 2)T mix according to the symmetric mass matrix

7 The model can also be identified with a CP-conserving 2HDM with softly broken Z symmetry and λ = 0 [41].
2 5
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 299

Fig. 7. Example of a loop contribution to the scalar coupling S 2 †2 1 that generates  1  =

 0 and ultimately h → μτ .

λ1 v12 + m23 tβ (λ3 + λ4 )v1 v2 − m23

· λ2 v22 + m23 /tβ
cα −sα m2h cα sα
≡ , (3.4)
sα cα m2H −sα cα
which leads to a light SM-like scalar h = cα h2 − sα h1 and a heavy H = cα h1 + sα h2 . Note that
the seven real parameters of the potential can be expressed in terms of the physical quantities mh ,
mH , mA , m+ , v, tan β, and α (see, e.g., Ref. [69] for the interchangeable parameter sets). The
couplings of h (H ) to ZZ and W W are given by their respective SM values times sin(β − α)
(cos(β − α)), as usual in 2HDMs, so the value β − α = π/2 makes h SM-like and decouples H
in the gauge boson sector. The limit β − α = π/2 also reduces all fermion couplings of h to their
SM values, but does not decouple H from them.
Yukawa couplings and fermion mass terms are dictated by the quantum numbers of Table 2
and take the form
−LY = LL Y 1 1 R + LL YN1 ˜ 1 NR
+ LL Y 2 2 R + LL YN2 ˜ 2 NR + QL Yu ˜ 2 uR + QL Yd 2 dR
+ 2 N R M N NR
1 c
+ 1 c
2 N R YS1 SNR + 1 c
2 N R YS2 SNR + h.c. (3.5)
The quark Yukawa matrices Yu,d are arbitrary and quarks will receive their masses (and CKM
mixing) just from 2 . Consequently, there are no tree-level flavor-changing neutral currents in
the quark sector, as enforced by our U (1) gauge symmetry. In the lepton sector, on the other
hand, the U (1)Lμ −Lτ symmetry enforces diagonal 2 couplings
Y 2 = diag(ye , yμ , yτ ) , YN2 = diag(y1 , y2 , y3 ) , (3.6)
and even more selective 1 couplings
   
0 0 0 0 0 0
Y 1 = 0 0 0 , YN1 = 0 0 ξ23 . (3.7)
0 ξτ μ 0 0 0 0
It is the off-diagonal τ –μ entry in Y 1 that will ultimately lead to h → μτ . The right-handed
neutrino Majorana mass matrix will be built from the pieces
     
M1 a13 a12
MN = M2 , YS1 = , YS2 = a12 . (3.8)
M2 a13
Overall we recognize our model as being basically a 2HDM of type I (albeit slightly restricted in
the scalar potential through λ5 = 0 [41]), plus 1 interactions (3.7) that exclusively modify the
μτ lepton sector. Since only the off-diagonal charged-lepton coupling τ L μR to scalars exists, as
enforced by the gauge symmetry U (1)Lμ −Lτ , our model provides a very minimal explanation
300 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

for the fact that flavor violation has only been (potentially) observed in the μτ sector. No flavor-
changing neutral currents arise in the quark sector, nor will (Lμ − Lτ ) = ±1 processes such as
μ → eγ be generated at an observable rate.

3.2. Neutrino masses and mixing

Having defined our setup, let us first take a look at neutrino masses and mixing, which serve
as a major motivation for Lμ − Lτ , independent of any charged-lepton flavor violation. The
symmetric right-handed neutrino mass matrix has contributions from Lμ − Lτ symmetric parts
(M1,2 ) and (Lμ − Lτ ) = ±1 pieces induced by the VEV S:
 
M1 a12 S a13 S
MN ≡ MN + SYS1 + SYS2 = · 0 M2 . (3.9)
· · 0
The Dirac mass matrix, coupling NR to the active left-handed neutrinos, similarly contains a
diagonal Lμ − Lτ symmetric part and one off-diagonal term generated by  1 :
 
v y1 sβ
mD =  2 YN2 +  1 YN1 = √ y2 sβ ξ23 cβ . (3.10)
2 y3 sβ
In the seesaw limit, we obtain the active-neutrino Majorana mass matrix νcL Mν νL

Mν  −m∗D (MN∗ )−1 m†D . (3.11)

Diagonalization VνT Mν Vν = diag(mν1 , mν2 , mν3 ) then gives the neutrino mixing matrix Vν ,
which forms the PMNS mixing matrix together with the charged-lepton contribution VeL via
U = Ve†L Vν . As we will see in Section 3.3, VeL consists only of a 23 rotation, so it influences
exclusively the atmospheric mixing angle θ23 (in the standard parametrization for U ).
The Abelian gauge group U (1)Lμ −Lτ obviously introduces a structure in Mν , making it a
flavor symmetry. Without the breaking terms from S and  1 , Mν would have the same
Lμ − Lτ symmetric structure as MN in Eq. (3.8), predicting maximal atmospheric mixing and
vanishing solar and reactor angles, as well as m232 ≡ m2ν3 − m2ν2 = 0 [56]. The U (1)Lμ −Lτ
breaking terms are thus needed in order to accommodate the observed neutrino parameters, and
it has been shown that small perturbations suffice to obtain viable mass splittings and angles for
quasi-degenerate neutrinos [56,58].
Let us briefly comment on our specific setup: MN has two texture zeroes due to the
(Lμ − Lτ ) = ±1 breaking structure of S. With ξ23 = 0 (or  1  = 0), these two texture zeroes
propagate to Mν as two vanishing minors, i.e. (M−1 −1
ν )33 = 0 = (Mν )22 [70], as is obvious from
the seesaw formula. These two vanishing minors ultimately imply correlations among the mixing
angles and phases [71,72], and our texture is among the seven patterns that are compatible with
current data [70]. One of the vanishing minors survives even for ξ23 = 0, namely (M−1 ν )22 = 0,
which leads to a weaker relation among the parameters [73]. The (Lμ − Lτ ) = 2 perturba-
tion from the Dirac sector thus helps to reduce the required fine-tuning otherwise necessary for
viable two-texture-zero/two-vanishing-minor neutrino mass matrices. The corrections from the
charged-lepton sector, U = Ve†L Vν , only influence the atmospheric mixing angle, leaving θ12 and
θ13 to be determined solely from Mν . Since the left-handed charged-lepton mixing angle will
turn out to be small, see Eq. (3.14), even the atmospheric mixing angle is essentially determined
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 301

Fig. 8. Scatter plot of neutrino mixing angles θ23,13 of Mν (Eq. (3.11)). The red (blue) points have ε = 0.1 and inverted
(normal) ordering, the green (black) points have ε = 0.01 and inverted (normal) ordering. The global-fit 3σ ranges are
sin2 θ23 = 0.385–0.644 and sin2 θ13 = 0.0188–0.0251 [43]. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)

by the neutrino mass matrix only, and hence expected to be close-to-maximal due to the approx-
imate Lμ − Lτ structure.
We visualize the dependence of θ23 on the breaking scale with a scatter plot in Fig. 8. Here,
the Lμ − Lτ symmetric entries in mD and MN are generated with absolute values ∈ [1, 3] and
random phases in order to generate the desired quasi-degenerate neutrino mass spectrum. Entries
breaking Lμ − Lτ , i.e. (MN )12,13 and (mD )23 in (3.9) and (3.10), are taken random ∈ [0, ε]
with random phases. We impose the 3σ constraints on sin θ12 and m221 /m231 from Ref. [43].
The units/scale of mD and MN , and hence Mν , are not fixed, but we expect a quasi-degenerate
spectrum, i.e. mνj  0.1–1 eV, with 0ν2β rates testable in the near future [74]. As can be seen,
normal ordering (NO) and inverted ordering (IO) correspond to s23 2 > 1/2 and s 2 < 1/2, re-
23
spectively, with perturbations |s23 − 1/2|  ε/2. Note that the global fit of Ref. [43] prefers
2 8
2 < 1/2 (> 1/2) for normal (inverted) ordering, opposite to our prediction; while this is not
s23
statistically relevant at the moment, it will become an important constraint on the breaking struc-
ture of Lμ − Lτ in the future (note that other global fits have different preferences [75,76]).
2 ≶ 1/2 (NO/IO) is fixed by the chiral structure in h → μP
Ultimately, the relation s23 L,R τ in
our Lμ − Lτ model, at least if the charged-lepton contribution to θ23 is small. As we will see in
Section 3.7, a flip in the Lμ − Lτ charge of 1 will modify the chiral structure to h → μPR τ
and give s232 < 1/2 for NO, opposite to the case discussed above. Determination of the θ octant
23
as well as the mass ordering are hence important discriminators for our model.

8 The relation of the octant of θ

23 and the mass ordering is dictated by our breaking structure in mD , leading to
(M−1 )
ν 22 = 0. Had we chosen an Lμ − Lτ charge +2 instead of −2 for 1 , effectively replacing mD by its trans-
2 > 1/2 (< 1/2) for inverted (normal) ordering, the vanishing minor being
pose, the correlation would be flipped, i.e. s23
−1
(Mν )33 = 0. We will comment on this scenario in Section 3.7.
302 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Note that the generic perturbations required for successful neutrino phenomenology are in
the range S/M1,2  10−1 –10−2 (Yukawa couplings of order one), so a lower bound on S 
MZ /g also implicitly bounds the seesaw scale M1,2 .
As a final remark, the solar mixing angle θ12 vanishes in the Lμ − Lτ -symmetric limit, but
can easily be large for a quasi-degenerate neutrino spectrum. In the μ–τ -symmetric case with
θ13 = 0 and θ23 = π/4, it is given in the form
S/M1,2
tan 2θ12 ∝ , (3.12)
1 − (Mν )11 /(Mν )23
(Mν )11,23 being the Lμ − Lτ -symmetric entries of Mν , naturally of similar magnitude. θ12
is hence given by the ratio of two small numbers [56]: U (1)Lμ −Lτ -breaking entries and devia-
tions from degeneracy. The expressions for θ12 become more intricate for the realistic cases with
θ13 = 0, but remain qualitatively similar.

3.3. Charged lepton masses

After spontaneous symmetry breaking of SU (2)L × U (1)Y × U (1)Lμ −Lτ , the charged-lepton
mass matrix takes the form
 
v ye sβ
Me = √ yμ sβ
2 ξτ μ cβ yτ sβ
   
1 me 1
≡ cL sL mμ cR −sR
−sL cL mτ sR cR
≡ VeL diag(me , mμ , mτ )Ve†R , (3.13)

which defines the charged-lepton mass basis via 0L = Ve†L L , 0R = Ve†R R . We assumed all
Yukawa couplings to be real here, so the two new mixing angles θL,R are given by
tan θL mμ v ξτ μ
=  1 and sin θR  √ cos β , (3.14)
tan θR mτ mτ 2
and replace ξτ μ as a parameter. Note that θL is automatically small and does not play a significant
role for neutrino mixing.

3.4. Lepton flavor violating interactions

The couplings of the light scalar h to the lepton mass eigenstates, −LY ⊃ 0 y 0 h,
L R can be
calculated in a straightforward manner to be
 
cα mτ cos(α − β) 0
y  diag(me , mμ , mτ ) − sR −cR sL −sL sR . (3.15)
vsβ v cβ sβ cL cR cL sR
The first (diagonal) term corresponds to the standard type-I 2HDM couplings, proportional to
cα /sβ , just like in the quark sector [41]; the second matrix is proportional to ξτ μ (or, equivalently,
sin θR ) and obviously induces the desired lepton flavor violation in the μτ sector. The τ μ entry
is dominant due to the mixing-angle hierarchy sL  sR , as expected from the structure of the
Yukawa couplings in Y 1 . Writing cos(α − β) = cos α cos β + sin α sin β allows us to picture the
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 303

Fig. 9. Pictorial depiction of the U (1)Lμ −Lτ -breaking structure behind h → μτ .

induced h → μτ as a combination of lepton mixing (proportional to the type-I cα /sβ ) and scalar
mixing (sα /cβ ), see Fig. 9.
Note that the choice cos(β − α) = 0 reduces all couplings of h (to vector bosons and fermions)
to their SM values [41], and eliminates the LFV couplings. In order to obtain a measurable
h → μτ rate, we thus need to allow for cos(α − β) = 0 and hence generically predict (slightly)
reduced h rates compared to the SM. This is easily allowed in our type-I-like 2HDM even in the
limit of large tan β, as discussed below in Section 3.5.
The coupling of the heavy scalar to leptons (−LY ⊃ 0L y H 0R H ) follows from the h couplings
with α → α − π/2:
 
sα mτ sin(α − β) 0
y  diag(me , mμ , mτ )
H
− sR −cR sL −sL sR . (3.16)
vsβ v cβ sβ c c c s L R L R

In particular, the τ μ entry is given by yτHμ = tα−β yτhμ .

Also note the chiral nature of these cou-
plings, due to our non-Hermitian coupling matrix.
The charged-scalar couplings to active neutrinos take the form
√
2 0 † 1 0 †
−LY ⊃ ν U diag(me , mμ , mτ ) 0R H + − ν V Y Ve 0 H + + h.c. (3.17)
vtβ L sβ L ν 1 R R
with νμR H + dominating in the large tan β limit, as expected. The first term is again the type-I
coupling, while the second one has the by now familiar matrix structure
 
ξτ μ 0 † 0
+
− ν U −cR sL −sL sR 0
RH . (3.18)
sβ L cL cR cL sR
√
Pseudoscalar interactions, following our parametrization j = (φj+ , (vj + hj − izj )/ 2)T and
A ≡ cβ z2 − sβ z1 , are given by
   
−i iξτ μ 0
−LY ⊃ 0 0
L diag(me , mμ , mτ ) R A+ √ 0
L −cR sL −sL sR 0
R A
vtβ 2sβ cL cR cL sR
+ h.c. (3.19)
√
Again, the dominant term in the large tan β limit is iξτ μ / 2 τ PR μA. Note the chiral nature even
of the pseudoscalar couplings (only in the off-diagonal couplings since we assume ξτ μ to be real
for simplicity).
304 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

3.5. Hints and constraints

Since we have basically a type-I 2HDM (slightly restricted via λ5 = 0, which however barely
changes the phenomenology [77]), we inherit the bounds on masses and mixing angles from
Ref. [78]. We only have to worry about the new interactions we introduced in the μτ sector.
Without going into any details, let alone a scan of the huge 2HDM parameter space, we
simply take | cos(β − α)|  0.4 for tan β  3, following the recent scan of the type-I 2HDM
with LHC bounds from Ref. [78]. This means in particular A, H , and H + masses below roughly
800 GeV [78], otherwise cβ−α is highly suppressed (decoupling regime) and makes our job in
explaining h → μτ slightly more difficult. The main difference of our 2HDM to type-I is the
absence of the λ5 term in the potential, fixing the two otherwise free mass parameters mA and
m3 via m2A = m23 /sβ cβ , which has little impact on the valid cβ−α –tβ values [77]. Furthermore,
our model features additional fermion couplings of H , A, and H + , predominantly in an LFV
manner to μ and τ fermions. This should in principle strengthen the bound on H + compared to
the type-I 2HDM, seeing as H + → μν is potentially enhanced; since the fermionic decay modes
under investigation at colliders are however only H + → τ ν or quarks, there are no additional
constraints.
This leaves the additional non-collider constraints from e.g. τ → μγ and (g − 2)μ to impose
on our model, which we will discuss below, as well as the decay h → μτ that motivates our
study.
With the τ L μR h coupling at our disposal, we can explain the CMS excess [9] in h → μτ with
Yukawa couplings (see Eq. (1.3))
   
mτ  cos(α − β) 


−3  cos(α − β)
 !
|yτ μ | =  cR cL sR   7 × 10  cR sR   3 × 10−3 . (3.20)
v cβ sβ sβ cβ
Working in the limit cβ ∼ sR  1, this simply fixes the parameter combination |ξτ μ cα−β | 
4 ×10−3 ; deviations of h’s couplings with respect to the SM, parametrized by sin(α −β) = 1, can
hence be easily made unobservably small, even for perturbatively small Yukawa coupling ξτ μ .
The scalar h is then very much SM-like, and will not lead to additional (LFV) processes in
conflict with observation, e.g. τ → μγ or (g − 2)μ , following the work of Ref. [8]. For example,
the current bound from τ → μγ translates into |yτ μ | < 0.016 at 90% C.L. for a sufficiently
SM-like h. Since all non-h rates can be suppressed by choosing mH , mA , and mH + large enough,
we clearly have a large allowed parameter space at our disposal.
Let us however briefly discuss possible effects of our model away from the decoupling limit
mentioned above. From the matrix structure of the h couplings in Eq. (3.15) we see that a large
θR will induce changes in the μμ and τ τ couplings of h, and hence to potentially observable
modified rates for h → μμ and h → τ τ . We plot the three branching ratios of interest in Fig. 10
for some sample values of α and β. The h → μμ branching ratio, even when enhanced in our
model, is currently not experimentally accessible [79]. The di-tauon rate on the other hand has
been observed by CMS [48] and ATLAS [80], with rates (relative to the SM) of 0.78 ± 0.27 and
1.42+0.44
−0.38 , respectively. In our case, the modified τ rate is
2
BR(h → τ τ ) cα v
 + yτhμ tR  (1 ± 0.4 |tR |)2 , (3.21)
BR(h → τ τ )|SM sβ mτ
inserting |yτhμ |  3 × 10−3 in the last step and assuming cα /sβ  1. The rate is enhanced (re-
duced) for cα−β < 0 (> 0). As expected, a large θR can strongly modify the rate h → τ τ ;
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 305

Fig. 10. The √ branching ratios h → μτ (black), h → τ τ (green), and h → μμ (blue) as a function of sin θR 
v
mτ ξτ μ cos β/ 2. Solid lines are for tan β = 3, cos(α − β) = −0.3, dashed lines for tan β = 10, cos(α − β) = −0.2,
dotted lines for tan β = 20, cos(α − β) = −0.2. The colored regions show the 1σ and 2σ ranges for the CMS hint of
h → μτ [9] (red) and the 1σ range for CMS h → τ τ [48] (green). (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)

we could fit the CMS rate nicely with tR  1/4, which obviously worsens the agreement with
ATLAS, or take tR  1/2 to match ATLAS’ enhanced h → τ τ rate, worsening agreement with
CMS. Future improvement in the accuracy of the di-tauon rate can hence provide important in-
formation for our model, complementary to the h → μτ rate.
We stress again that, as mentioned above, we can in any case work in the limit θR  1, while
still explaining the h → μτ rate, thus rendering even the h → τ τ and h → μμ rates effectively
SM-like. Nevertheless, the generic predictions of our Lμ − Lτ explanation of the LFV excess
h → μτ are modified di-tauon and di-muon rates, together with in general not-too-small cα−β ,
thus suppressing the h couplings to gauge bosons, as in any other 2HDM.
On to other LFV processes: An SM-like h with yτ μ  3 × 10−3 does not lead to τ → μγ rates
in conflict with current constraints, as shown in Ref. [8]. One might still expect additional LFV
processes induced by the other scalars, seeing as they couple more dominantly to μτ the more
SM-like h becomes. This is not necessarily the case, though, because additional suppression
η
factors arise. For τ → μγ , not only the coupling yτ μ is required for η ∈ {H, A} to run in the
η η
loop, but also yτ τ (one loop) or ytt,W W (two loop). Since these couplings are suppressed by
sα /sβ , cot β, or cα−β , the rates are typically small. Using the formulae from Ref. [8] for the
one-loop contribution of h, H , and A to τ → μγ as well as the dominant two-loop diagrams with
top-quark and W -boson loops, we can find a weak correlation between h → μτ and τ → μγ , see
Fig. 11. This is not surprising, as all LFV scales with ξτ μ , the only LFV coupling in our model
(outside the neutrino sector). Since τ → μγ is additionally suppressed by the heavy masses of
A and H , the rates are typically below the current sensitivity. In Fig. 11, we randomly selected
values
mA , mH ∈ [150, 700] GeV, tan β ∈ [3, 60] ,
| cos(α − β)| < 0.4 , sin θR ∈ [0, 0.5] . (3.22)
Note that we did not impose any bounds, but chose values that seem compatible with Ref. [78].
Not surprisingly, we find that the two LFV rates are somewhat correlated, and also that h → τ τ is
generically modified for large LFV. Improvements of τ → μγ searches down to branching ratios
of O(10−9 ) appear feasible [81] and will have a major impact on the allowed parameter space.
306 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

Fig. 11. The LFV branching ratios h → μτ and τ → μγ for the random values of Eq. (3.22). The color coding of
the points refers to the rate h → τ τ relative to the SM prediction. Also shown are the 2σ ranges for the CMS hint of
h → μτ [9] (light red area) and the current 90% C.L. upper limit on BR(τ → μγ ) from BaBar [50] (vertical black
dashed line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version
of this article.)

Note that the decay τ → μγ gives stronger limits on the scalar sector than τ → 3μ, even
though the experimental limit on the branching ratio is a factor ∼2 weaker. This is because
τ → 3μ is either suppressed by an additional muon Yukawa coupling (tree-level scalar exchange)
or fine-structure coupling (off-shell photon in τ → μγ → 3μ) [8]. Finally, the scalar contribu-
tions to the muon’s magnetic moment (g − 2)μ are insignificant for the parameter values chosen
above.

3.6. Gauge boson sector

The gauge boson Z of U (1)Lμ −Lτ couples only to muonic and tauonic leptons (neglecting
kinetic mixing), and is thus subject to quite different constraints than other popular Z models.
It has recently been shown in Ref. [63] that trident production νμ N → νμ N μ+ μ− provides the
strongest limit for a heavy Z , MZ /g  550 GeV at 95% C.L. from CCFR [82], making in par-
ticular a resolution of the muon’s anomalous magnetic moment aμ impossible with a heavy Z .
A lighter Z below 400 MeV might still do the trick [64], but is not considered here.9 Note that
MZ /g  S, so the trident bound effectively also acts as a lower bound on the seesaw scale
(Section 3.2). The trident bound can be improved in future collider measurements, as pointed out
recently in Ref. [84].
The flavor-dependent Z couplings in combination with the charged-lepton diagonalization
lead to the gauge couplings
 
 0
g jLμ −Lτ Zα = g
α 0
i cos 2θi sin 2θi γ α 0i Zα , (3.23)
i=L,R sin 2θi − cos 2θi
which not only contain an axial-vector component due to θL = θR , but more importantly yield
LFV couplings, e.g. g sin 2θR μR γ α τR Zα . These couplings induce a tree-level decay τ → 3μ

9 A resolution of a and h → μτ with such a light Z would allow for the two-body decay τ → μZ . Old limits
μ
from ARGUS [83] then already imply a tiny θR < O(10−5 ) and a correspondingly large tan β ∼ 1/θR to keep yτhμ in
Eq. (3.20) large. Updated searches, e.g. at Belle, could easily improve this limit.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 307

mediated by Z , with a rate proportional to sin2 2θi cos2 2θj (g /MZ )4 for MZ mτ . The
experimental limit is BR(τ → 3μ) < 2.1 × 10−8 at 90 % C.L. [85], resulting in the strong con-
straint [86]

2
sin 4θj 550 GeV
< 1, j = L, R , (3.24)
5 × 10−3 MZ /g

beating out τ → μγ [86]. Obviously lepton flavor violation induced by the Z can be made small
by increasing MZ /g even for large θR , so this bound in no way poses a problem for h → μτ .
Also note that the VEV of 1 automatically induces a Z–Z mixing angle

2g1 g v 2 cos2 β 10−3 TeV 2

10 2
tan 2θZZ  − , (3.25)
MZ − MZ
2 2 g MZ /g tan β

that leads to lepton non-universal Z couplings. Limits on g θZZ are typically around 10−2 –10−3
[58] and can be easily satisfied by increasing MZ ∝ S. There is furthermore no theoretical
reason to forbid a kinetic-mixing angle between our Z and the hypercharge gauge boson, which
will in any way be generated radiatively. We will not consider this any further. Note that the
gauge coupling should satisfy g  0.5 in order to avoid a Landau pole below the Planck scale if
Lμ − Lτ is broken around TeV.

3.7. Other charge assignments

Having focused on the specific Lμ − Lτ charge assignments of Table 2, let us make some
comments about closely related models. First, consider a sign-flip of the 1 scalar doublet’s
Lμ − Lτ charge, i.e. +2 instead of −2. The effective 2HDM potential looks the same, the soft-
breaking parameter m23 being generated by the coupling δ S 2 †1 2 . A difference arises in the
Yukawa couplings of 1 , though, as one effectively replaces Y 1 and YN1 from Eq. (3.7) by their
transposed matrices. Because of this, it is now the μL τR entry that is dominant, not τ L μR , so
the chiral structure behind h → μτ changes. For the neutrino mixing, this changes the vanish-
ing minor from (M−1 −1 2
ν )22 to (Mν )33 and leads to the octant–ordering relation s23 > 1/2 (IO),
< 1/2 (NO) for the 23 mixing angle in Vν . This is the (not yet statistically relevant) preferred
correlation in the global fit of Ref. [43], assuming the charged lepton contribution to the PMNS
matrix is small.10
In the charged-lepton mass matrix diagonalization, the change of quantum numbers switches
θL ↔ θR , so θL is now the dominant charged-lepton mixing angle. Besides this renaming, there
are no changes in the LFV phenomenology. However, the fact that θL can be large has potentially
a huge impact on the neutrino mixing sector, because θL is added to the generically close-to-
maximal mixing angle θ23 ν to form the physical atmospheric mixing angle (which is measured

to be close to maximal: θ23 = 38–53◦ at 3σ [43]). So, in addition to strongly modified h → τ τ

rates (Eq. (3.21)), one expects also large deviations from maximal 23 mixing in the region of
parameter space where θL is large. This is however not a hard prediction, as h → μτ can still be
explained with θL  1 in this scenario.

10 Note that other global fits have different preferences [75,76].

308 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

If we give the doublet 1 an Lμ − Lτ charge ±1 instead of ±2, we recover the model from
Ref. [58]. Here, the (Lμ − Lτ ) = 1 Yukawa couplings will generate LFV processes like μ →
eX or h → eμ, eτ (see Ref. [8] for details), mediated by the scalars and Z . Stronger constraints
apply, of course, due to the couplings to electrons and muons, and there are no indications in the
data that could be taken as a hint.
Analogous considerations for the anomaly-free symmetries Le − Lμ and Le − Lτ – which
suffer from stronger bounds due to the coupling to electrons and cannot be regarded as good
symmetries to describe leptonic mixing – can be found in Ref. [57].

4. Conclusion

Using a recently found hint for non-standard Higgs decays h → μτ as an exemplary bench-
mark value, we have analyzed the possibility to explain lepton flavor violating Higgs decays in
flavor symmetry models. These models usually focus on explaining lepton mixing, but can also
have phenomenology at the electroweak scale, when they are broken at that scale. Two very dif-
ferent approaches were followed here: a continuous Abelian and a discrete non-Abelian case.
Both models enlarge the Higgs sector.
The non-Abelian group A4 is the most often applied flavor group, as it is the smallest discrete
group with a 3-dimensional irreducible representation. We introduce, as almost always in such
models, an A4 -triplet containing the three lepton doublets, but also assume an A4 -triplet of scalar
weak doublets. The model can easily reproduce neutrino mixing parameters in agreement with
current data. It predicts a branching ratio for h → eτ with similar magnitude as the one of h →
μτ , where this decay enjoys a chiral structure mainly of μPR τ . The breaking of the symmetry
introduces also breaking of ‘triality’, which usually suppresses lepton flavor violation in such
models; hence, the decay μ → eγ is induced and can pose important constraints.
The Abelian flavor symmetry U (1)Lμ −Lτ as a particularly attractive symmetry for quasi-
degenerate neutrinos with close-to-maximal atmospheric mixing can also be taken as a horizontal
symmetry in a 2HDM, allowing for flavor violation in selected lepton modes only. These are
specified by the Lμ − Lτ quantum numbers of the non-SM-like scalar doublet, leading to lep-
ton Yukawa couplings that perturb the overall type-I 2HDM structure. We have shown how this
can lead to LFV exclusively in the μτ sector in order to explain the tentative h → μτ signal
at CMS, with natural suppression/absence of other flavor violating modes typically occurring
in other models. The model non-trivially correlates the Higgs branching ratio for h → μτ with
the decay τ → μγ , and generally predicts modified h → μμ and τ τ rates, as well as overall
type-I 2HDM-like suppression of Higgs couplings to gauge bosons, on a smaller scale than the
A4 model. Additionally, the chiral structure of h → μPR,L τ influences the correlation of mass
ordering and the octant of θ23 in the neutrino sector.
In summary, the option to explain non-standard Higgs phenomenology in low scale flavor
symmetry models is an interesting testing ground for those models, and possible in a variety of
different approaches, predicting a large number of other phenomenological consequences.

5. Note added

Recently, CMS has released its final analysis

 of the h → μτ search as a preprint [87], resulting in
slightly changed values – BR(h → μτ ) = 0.84+0.39
−0.37 % – with little impact on our study.
 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310 309

Acknowledgements

This work of WR is supported by the Max Planck Society in the project MANITOP. The work
of JH is funded in part by IISN and by Belgian Science Policy (IAP VII/37). YS is supported by
JSPS Postdoctoral Fellowships for Research Abroad, No. 20130600.

References

[1] G. Aad, et al., ATLAS Collaboration, Phys. Lett. B 716 (2012) 1, arXiv:1207.7214.
[2] S. Chatrchyan, et al., CMS Collaboration, Phys. Lett. B 716 (2012) 30, arXiv:1207.7235.
[3] G. Bhattacharyya, P. Leser, H. Päs, Phys. Rev. D 83 (2011) 011701, arXiv:1006.5597.
[4] P.H. Frampton, C.M. Ho, T.W. Kephart, S. Matsuzaki, Phys. Rev. D 82 (2010) 113007, arXiv:1009.0307.
[5] Q.-H. Cao, A. Damanik, E. Ma, D. Wegman, Phys. Rev. D 83 (2011) 093012, arXiv:1103.0008.
[6] S.-Y. Ho, J. Tandean, Phys. Rev. D 87 (2013) 095015, arXiv:1303.5700.
[7] G. Blankenburg, J. Ellis, G. Isidori, Phys. Lett. B 712 (2012) 386, arXiv:1202.5704.
[8] R. Harnik, J. Kopp, J. Zupan, J. High Energy Phys. 1303 (2013) 026, arXiv:1209.1397.
[9] CMS, CMS Collaboration, 2014, CMS-PAS-HIG-14-005.
[10] A. Dery, A. Efrati, Y. Nir, Y. Soreq, V. Susič, Phys. Rev. D 90 (2014) 115022, arXiv:1408.1371.
[11] M.D. Campos, A.E.C. Hernández, H. Päs, E. Schumacher, arXiv:1408.1652, 2014.
[12] A. Celis, V. Cirigliano, E. Passemar, arXiv:1409.4439, 2014.
[13] D.A. Sierra, A. Vicente, Phys. Rev. D 90 (2014) 115004, arXiv:1409.7690.
[14] C.-J. Lee, J. Tandean, arXiv:1410.6803, 2014.
[15] E. Ma, Phys. Rev. D 82 (2010) 037301, arXiv:1006.3524.
[16] H. Ishimori, T. Kobayashi, H. Ohki, Y. Shimizu, H. Okada, et al., Prog. Theor. Phys. Suppl. 183 (2010) 1,
arXiv:1003.3552.
[17] S.F. King, C. Luhn, Rep. Prog. Phys. 76 (2013) 056201, arXiv:1301.1340.
[18] E. Ma, G. Rajasekaran, Phys. Rev. D 64 (2001) 113012, arXiv:hep-ph/0106291.
[19] K. Babu, E. Ma, J. Valle, Phys. Lett. B 552 (2003) 207, arXiv:hep-ph/0206292;
E. Ma, Phys. Rev. D 70 (2004) 031901, arXiv:hep-ph/0404199.
[20] G. Altarelli, F. Feruglio, Nucl. Phys. B 720 (2005) 64, arXiv:hep-ph/0504165.
[21] K. Babu, X.-G. He, arXiv:hep-ph/0507217, 2005.
[22] G. Altarelli, F. Feruglio, Nucl. Phys. B 741 (2006) 215, arXiv:hep-ph/0512103.
[23] X.-G. He, Y.-Y. Keum, R.R. Volkas, J. High Energy Phys. 0604 (2006) 039, arXiv:hep-ph/0601001.
[24] C. Lam, Phys. Lett. B 656 (2007) 193, arXiv:0708.3665.
[25] C. Lam, Phys. Rev. Lett. 101 (2008) 121602, arXiv:0804.2622.
[26] C. Lam, Phys. Rev. D 78 (2008) 073015, arXiv:0809.1185.
[27] E. Ma, Phys. Rev. D 73 (2006) 077301, arXiv:hep-ph/0601225.
[28] J. Kubo, E. Ma, D. Suematsu, Phys. Lett. B 642 (2006) 18, arXiv:hep-ph/0604114.
[29] M. Hirsch, S. Morisi, J. Valle, Phys. Lett. B 679 (2009) 454, arXiv:0905.3056.
[30] D. Ibanez, S. Morisi, J. Valle, Phys. Rev. D 80 (2009) 053015, arXiv:0907.3109.
[31] R. de Adelhart Toorop, F. Bazzocchi, L. Merlo, A. Paris, J. High Energy Phys. 1103 (2011) 035, arXiv:1012.1791.
[32] R. de Adelhart Toorop, F. Bazzocchi, L. Merlo, A. Paris, J. High Energy Phys. 1103 (2011) 040, arXiv:1012.2091.
[33] Q.-H. Cao, S. Khalil, E. Ma, H. Okada, Phys. Rev. Lett. 106 (2011) 131801, arXiv:1009.5415.
[34] G. Bhattacharyya, P. Leser, H. Päs, Phys. Rev. D 86 (2012) 036009, arXiv:1206.4202;
G. Bhattacharyya, I. de Medeiros Varzielas, P. Leser, Phys. Rev. Lett. 109 (2012) 241603, arXiv:1210.0545.
[35] M. Holthausen, M. Lindner, M.A. Schmidt, Phys. Rev. D 87 (2013) 033006, arXiv:1211.5143.
[36] M. Holthausen, M.A. Schmidt, J. High Energy Phys. 1201 (2012) 126, arXiv:1111.1730.
[37] E. Paschos, Phys. Rev. D 15 (1977) 1966.
[38] S.L. Glashow, S. Weinberg, Phys. Rev. D 15 (1977) 1958.
[39] M. Holthausen, Ph.D. thesis, University of Heidelberg, 2012.
[40] I. Ivanov, C. Nishi, J. High Energy Phys. 1501 (2015) 021, arXiv:1410.6139.
[41] G. Branco, P. Ferreira, L. Lavoura, M. Rebelo, M. Sher, et al., Phys. Rep. 516 (2012) 1, arXiv:1106.0034.
[42] K. Olive, et al., Particle Data Group, Chin. Phys. C 38 (2014) 090001.
[43] M. Gonzalez-Garcia, M. Maltoni, T. Schwetz, J. High Energy Phys. 1411 (2014) 052, arXiv:1409.5439.
[44] J. Barry, W. Rodejohann, Nucl. Phys. B 842 (2011) 33, arXiv:1007.5217.
310 J. Heeck et al. / Nuclear Physics B 896 (2015) 281–310

[45] G. Aad, et al., ATLAS Collaboration, J. High Energy Phys. 1501 (2015) 069, arXiv:1409.6212.
[46] S. Chatrchyan, et al., CMS Collaboration, Phys. Rev. D 89 (2014) 012003, arXiv:1310.3687.
[47] V. Barger, L.L. Everett, C.B. Jackson, A.D. Peterson, G. Shaughnessy, Phys. Rev. D 90 (2014) 095006,
arXiv:1408.2525.
[48] S. Chatrchyan, et al., CMS Collaboration, J. High Energy Phys. 1405 (2014) 104, arXiv:1401.5041.
[49] J. Adam, et al., MEG Collaboration, Phys. Rev. Lett. 110 (2013) 201801, arXiv:1303.0754.
[50] B. Aubert, et al., BaBar Collaboration, Phys. Rev. Lett. 104 (2010) 021802, arXiv:0908.2381.
[51] X. He, G.C. Joshi, H. Lew, R. Volkas, Phys. Rev. D 43 (1991) 22.
[52] R. Foot, Mod. Phys. Lett. A 6 (1991) 527.
[53] X.-G. He, G.C. Joshi, H. Lew, R. Volkas, Phys. Rev. D 44 (1991) 2118.
[54] P. Binetruy, S. Lavignac, S.T. Petcov, P. Ramond, Nucl. Phys. B 496 (1997) 3, arXiv:hep-ph/9610481.
[55] N.F. Bell, R.R. Volkas, Phys. Rev. D 63 (2001) 013006, arXiv:hep-ph/0008177.
[56] S. Choubey, W. Rodejohann, Eur. Phys. J. C 40 (2005) 259, arXiv:hep-ph/0411190.
[57] G. Dutta, A.S. Joshipura, K. Vijaykumar, Phys. Rev. D 50 (1994) 2109, arXiv:hep-ph/9405292.
[58] J. Heeck, W. Rodejohann, Phys. Rev. D 84 (2011) 075007, arXiv:1107.5238.
[59] S. Gninenko, N. Krasnikov, Phys. Lett. B 513 (2001) 119, arXiv:hep-ph/0102222.
[60] S. Baek, N. Deshpande, X. He, P. Ko, Phys. Rev. D 64 (2001) 055006, arXiv:hep-ph/0104141.
[61] E. Ma, D. Roy, S. Roy, Phys. Lett. B 525 (2002) 101, arXiv:hep-ph/0110146.
[62] E. Ma, D.P. Roy, Phys. Rev. D 65 (2002) 075021.
[63] W. Altmannshofer, S. Gori, M. Pospelov, I. Yavin, Phys. Rev. D 89 (2014) 095033, arXiv:1403.1269.
[64] W. Altmannshofer, S. Gori, M. Pospelov, I. Yavin, Phys. Rev. Lett. 113 (2014) 091801, arXiv:1406.2332.
[65] T. Araki, F. Kaneko, Y. Konishi, T. Ota, J. Sato, et al., Phys. Rev. D 91 (2015) 037301, arXiv:1409.4180.
[66] S. Gninenko, N. Krasnikov, V. Matveev, arXiv:1412.1400, 2014.
[67] J. Heeck, W. Rodejohann, J. Phys. G 38 (2011) 085005, arXiv:1007.2655.
[68] P. Ko, Y. Omura, C. Yu, Phys. Lett. B 717 (2012) 202, arXiv:1204.4588.
[69] S. Kanemura, Y. Okada, E. Senaha, C.-P. Yuan, Phys. Rev. D 70 (2004) 115002, arXiv:hep-ph/0408364.
[70] T. Araki, J. Heeck, J. Kubo, J. High Energy Phys. 1207 (2012) 083, arXiv:1203.4951.
[71] L. Lavoura, Phys. Lett. B 609 (2005) 317, arXiv:hep-ph/0411232.
[72] E. Lashin, N. Chamoun, Phys. Rev. D 78 (2008) 073002, arXiv:0708.2423.
[73] E. Lashin, N. Chamoun, Phys. Rev. D 80 (2009) 093004, arXiv:0909.2669.
[74] W. Rodejohann, Int. J. Mod. Phys. E 20 (2011) 1833, arXiv:1106.1334.
[75] F. Capozzi, G. Fogli, E. Lisi, A. Marrone, D. Montanino, et al., Phys. Rev. D 89 (2014) 093018, arXiv:1312.2878.
[76] D. Forero, M. Tortola, J. Valle, Phys. Rev. D 90 (2014) 093006, arXiv:1405.7540.
[77] P. Ko, Y. Omura, C. Yu, J. High Energy Phys. 1401 (2014) 016, arXiv:1309.7156.
[78] B. Dumont, J.F. Gunion, Y. Jiang, S. Kraml, Phys. Rev. D 90 (2014) 035021, arXiv:1405.3584;
B. Dumont, J.F. Gunion, Y. Jiang, S. Kraml, arXiv:1409.4088, 2014.
[79] V. Khachatryan, et al., CMS Collaboration, Phys. Lett. B 744 (2015) 184, arXiv:1410.6679.
[80] ATLAS, ATLAS Collaboration, 2014, ATLAS-CONF-2014-061, ATLAS-COM-CONF-2014-080.
[81] T. Aushev, W. Bartel, A. Bondar, J. Brodzicka, T. Browder, et al., arXiv:1002.5012, 2010.
[82] S. Mishra, et al., CCFR Collaboration, Phys. Rev. Lett. 66 (1991) 3117.
[83] H. Albrecht, et al., ARGUS Collaboration, Z. Phys. C 68 (1995) 25.
[84] F. del Aguila, M. Chala, J. Santiago, Y. Yamamoto, J. High Energy Phys. 1503 (2015) 059, arXiv:1411.7394.
[85] K. Hayasaka, K. Inami, Y. Miyazaki, K. Arinstein, V. Aulchenko, et al., Phys. Lett. B 687 (2010) 139,
arXiv:1001.3221.
[86] C.-W. Chiang, T. Nomura, J. Tandean, Phys. Rev. D 87 (2013) 075020, arXiv:1302.2894.
[87] V. Khachatryan, et al., CMS Collaboration, arXiv:1502.07400, 2015.