betaFeedback

WorksPhysics and Astronomy

Complete set of bounded-from-below conditions for the A4A_4-symmetric three-Higgs-doublet model

Abstract

We derive bounded-from-below necessary conditions for the scalar potential of an A4A_4-invariant three-Higgs-doublet model. The field Ansätze that we have used were devised with the help of Claude. Testing our set of conditions on 10610^6 random potentials, that in parallel we minimize by brute force using the Mathematica package StableWein, we find that the set seems to be necessary and sufficient for boundedness-from-below.

Introduction

We consider the SU(2)×U(1)SU(2) \times U(1) gauge model with scalar sector consisting of three SU(2)SU(2) doublets with equal hypercharge; this is called the three-Higgs-doublet model (3HDM). Let χk\chi_k (k=1,2,3k = 1, 2, 3) denote the doublets. We study the version of the 3HDM where, besides the SU(2)×U(1)SU(2) \times U(1) gauge symmetry, there is also invariance under the discrete non-Abelian group A4A_4 generated by the transformations

T1:χ2→−χ2,χ3→−χ3;(1)T_1:\quad\chi_2 \to-\chi_2,\quad\chi_3 \to-\chi_3; \tag*{(1)}
T2:χ1→χ2→χ3→χ1.(2)T_2:\quad\chi_1 \to\chi_2 \to\chi_3 \to\chi_1. \tag*{(2)}

Such a model has been used to address the flavour puzzle [1, 2, 3, 4, 5, 6, 7, 8], to generate neutrino masses [9, 10, 11, 12, 13, 14, 15, 16], as an explanation for dark matter [17, 18, 19, 20, 21, 22, 23, 24], amongst other applications [25, 26, 27, 28, 29, 30, 31].

Using

Akl:=χk†χl,(3)A_{kl} := \chi_k^\dagger\chi_l, \tag*{(3)}

we write the most general renormalizable SU(2)×U(1)×A4SU(2) \times U(1) \times A_4-symmetric scalar potential (SP) as V=V2+V4V = V_2 + V_4, where V2=μ(A11+A22+A33)V_2 = \mu(A_{11} + A_{22} + A_{33}) and [32]

V4=Λ0+Λ33(A112+A222+A332)+2Λ0−Λ33(A11A22+A22A33+A33A11)(4)V_4 = \frac{\Lambda_0 + \Lambda_3}{3}\left(A_{11}^2 + A_{22}^2 + A_{33}^2\right) + \frac{2\Lambda_0 - \Lambda_3}{3}\left(A_{11}A_{22} + A_{22}A_{33} + A_{33}A_{11}\right) \tag*{(4)}
+Λ1+Λ22(∣A12∣2+∣A23∣2+∣A31∣2)+Λ1−Λ22Re⁡(A122+A232+A312)(5)+ \frac{\Lambda_1 + \Lambda_2}{2}\left(\lvert A_{12}\rvert^2 + \lvert A_{23}\rvert^2 + \lvert A_{31}\rvert^2\right) + \frac{\Lambda_1 - \Lambda_2}{2}\operatorname{Re}\left(A_{12}^2 + A_{23}^2 + A_{31}^2\right) \tag*{(5)}
+Λ42Im⁡(A122+A232+A312).(6)+ \frac{\Lambda_4}{2}\operatorname{Im}\left(A_{12}^2 + A_{23}^2 + A_{31}^2\right). \tag*{(6)}

Both μ\mu and Λ0,1,2,3,4\Lambda_{0,1,2,3,4} are real. In the special case where Λ4=0\Lambda_4 = 0, the potential is S4S_4-symmetric [32, 33].

In this work, we are interested in the conditions that guarantee that an SP with V4V_4 given by Eq. (3) is bounded from below (BFB).1 That is, we look for conditions that ensure that V4≥0V_4 \ge0 for all possible field configurations.2 To that end, it is convenient to rewrite V4V_4 as

V4=N2(Λ0+Λ33−Λ3x+Λ1+Λ22x′+Λ1−Λ22y+Λ42z),(7)V_4 = N^2\left(\frac{\Lambda_0 + \Lambda_3}{3} - \Lambda_3 x + \frac{\Lambda_1 + \Lambda_2}{2}x' + \frac{\Lambda_1 - \Lambda_2}{2}y + \frac{\Lambda_4}{2}z\right), \tag*{(7)}

where

N≡A11+A22+A33,(8)N \equiv A_{11} + A_{22} + A_{33}, \tag*{(8)}
x≡A11A22+A22A33+A33A11N2,(9)x \equiv\frac{A_{11}A_{22} + A_{22}A_{33} + A_{33}A_{11}}{N^2}, \tag*{(9)}
x′≡∣A12∣2+∣A23∣2+∣A31∣2N2,(10)x' \equiv\frac{\lvert A_{12}\rvert^2 + \lvert A_{23}\rvert^2 + \lvert A_{31}\rvert^2}{N^2}, \tag*{(10)}
y+iz≡A122+A232+A312N2.(11)y + iz \equiv\frac{A_{12}^2 + A_{23}^2 + A_{31}^2}{N^2}. \tag*{(11)}

The set of values that (x,x′,y,z)(x,x',y,z) may take is called the orbit space.

Notice that Λ0\Lambda_0 appears only as an additive constant in Eq. (7), i.e. the Λ0\Lambda_0 term is independent from the orbit-space point. Therefore, requiring V4≥0V_4 \ge0 at every point of orbit space implies BFB conditions of the form [34]

Λ0+Σ(Λ1,Λ2,Λ3,Λ4)≥0.(12)\Lambda_0 + \Sigma(\Lambda_1,\Lambda_2,\Lambda_3,\Lambda_4) \ge0. \tag*{(12)}

In sections 2 and 3 we derive various Σ(Λ1,Λ2,Λ3,Λ4)\Sigma(\Lambda_1,\Lambda_2,\Lambda_3,\Lambda_4) that provide a set of necessary BFB conditions. In section 4 we suggest that our set of necessary conditions actually is necessary and sufficient (n&s) for BnessFB. Section 5 is devoted to the S4S_4-symmetric case.

Vertices

We say that a configuration of the scalar doublets is ‘neutral’ if one may use an SU(2)SU(2) transformation to nullify all the upper components of the doublets. If a configuration of the doublets is not neutral then we call it ‘charge-breaking’. In neutral configurations the doublets behave like singlets, meaning that ∣Aij∣2=AiiAjj ∀i,j\lvert A_{ij}\rvert^2 = A_{ii}A_{jj}\ \forall i,j. Thus, in the notation of (4), the formal definition of a neutral configuration is x=x′x = x'.

Neutral vertices

The possible neutral vacua of the A4A_4-symmetric 3HDM have been found in Refs. [32, 34]. They are

  • point AA: (x,x′,y,z)=(0,0,0,0)(x,x',y,z)=(0,0,0,0);

  • point BB: (x,x′,y,z)=13(1,1,1,0)(x,x',y,z)=\frac{1}{3}(1,1,1,0);

  • point CC: (x,x′,y,z)=16(2,2,−1,3)(x,x',y,z)=\frac{1}{6}(2,2,-1,\sqrt{3});

  • point C′C': (x,x′,y,z)=16(2,2,−1,−3)(x,x',y,z)=\frac{1}{6}(2,2,-1,-\sqrt{3});

  • points in circumference DD: (x,x′,y,z)=14(1,1,cos⁡2α,sin⁡2α)(x,x',y,z)=\frac{1}{4}(1,1,\cos2\alpha,\sin2\alpha), with α∈[0,π]\alpha\in[0,\pi].

Figure 1 depicts the orbit space sliced along the hyperplane x=x′x=x'. The points AA, BB, CC, and C′C' are marked, and the circumference DD is drawn in blue.

The points of orbit space with x = x′

Figure 1. The points of orbit space with x=x′x=x'.

The points and circumference may be obtained, for example, by using the configurations of doublets4

A:χ1=(0ϱ2),χ2=χ3=(00);(13)\mathrm{A}: \quad\chi_1 = \begin{pmatrix} 0 \\ \varrho^2 \end{pmatrix}, \quad\chi_2 = \chi_3 = \begin{pmatrix} 0 \\ 0 \end{pmatrix}; \tag*{(13)}
B:χ1=χ2=χ3=(0ϱ);(14)\mathrm{B}: \quad\chi_1 = \chi_2 = \chi_3 = \begin{pmatrix} 0 \\ \varrho\end{pmatrix}; \tag*{(14)}
C:χ1=(ω2ϱ0),χ2=(ωϱ0),χ3=(ϱ0);(15)\mathrm{C}: \quad\chi_1 = \begin{pmatrix} \omega^2\varrho\\ 0 \end{pmatrix}, \quad\chi_2 = \begin{pmatrix} \omega\varrho\\ 0 \end{pmatrix}, \quad\chi_3 = \begin{pmatrix} \varrho\\ 0 \end{pmatrix}; \tag*{(15)}
C′:χ1=(ωϱ0),χ2=(ω2ϱ0),χ3=(ϱ0);(16)\mathrm{C'}: \quad\chi_1 = \begin{pmatrix} \omega\varrho\\ 0 \end{pmatrix}, \quad\chi_2 = \begin{pmatrix} \omega^2\varrho\\ 0 \end{pmatrix}, \quad\chi_3 = \begin{pmatrix} \varrho\\ 0 \end{pmatrix}; \tag*{(16)}
D:χ1=(00),χ2=(ϱ0),χ3=(eiαϱ0),(17)\mathrm{D}: \quad\chi_1 = \begin{pmatrix} 0 \\ 0 \end{pmatrix}, \quad\chi_2 = \begin{pmatrix} \varrho\\ 0 \end{pmatrix}, \quad\chi_3 = \begin{pmatrix} e^{i\alpha}\varrho\\ 0 \end{pmatrix}, \tag*{(17)}

where ϱ=2\varrho= \sqrt{2} and ω=e2iπ/3\omega= e^{2i\pi/3}.

As stated, for BFBness, V4V_4 cannot be negative at any of these points. The resulting necessary BFB conditions happen to be n&s if one restricts oneself to neutral configurations of doublets [33, 35]. Written in the form of inequality (6), they have Σ(Λ1,Λ2,Λ3,Λ4)\Sigma(\Lambda_1,\Lambda_2,\Lambda_3,\Lambda_4) equal to

Λ3≡ΣA,(18)\Lambda_3 \equiv\Sigma_A, \tag*{(18)}
Λ1≡ΣB,(19)\Lambda_1 \equiv\Sigma_B, \tag*{(19)}
Λ1+3Λ2+3Λ44≡ΣC,(20)\frac{\Lambda_1 + 3\Lambda_2 + \sqrt{3}\Lambda_4}{4} \equiv\Sigma_C, \tag*{(20)}
Λ1+3Λ2−3Λ44≡ΣC′,(21)\frac{\Lambda_1 + 3\Lambda_2 - \sqrt{3}\Lambda_4}{4} \equiv\Sigma_{C'}, \tag*{(21)}
Λ34+38[Λ1+Λ2+(Λ1−Λ2)cos⁡2α+Λ4sin⁡2α]≡ΣD(α),(22)\frac{\Lambda_3}{4} + \frac{3}{8}\left[\Lambda_1 + \Lambda_2 + (\Lambda_1 - \Lambda_2)\cos2\alpha+ \Lambda_4\sin2\alpha\right] \equiv\Sigma_D(\alpha), \tag*{(22)}

respectively. Since Λ0+ΣD(α)≥0 ∀α\Lambda_0 + \Sigma_D(\alpha) \ge0\ \forall\alpha, one may minimize ΣD(α)\Sigma_D(\alpha) with respect to α\alpha and obtain [33]

Λ34+38(Λ1+Λ2−R)≡ΣD,(23)\frac{\Lambda_3}{4} + \frac{3}{8}(\Lambda_1 + \Lambda_2 - R) \equiv\Sigma_D, \tag*{(23)}

where

R:=(Λ1−Λ2)2+Λ42.(24)R := \sqrt{(\Lambda_1 - \Lambda_2)^2 + \Lambda_4^2}. \tag*{(24)}

Charge-breaking vertices

In Figure 2 we show the orbit space projected onto the (x,x′)(x,x') plane. The neutral configurations lie on the black line where x′=xx' = x. One sees that there are additional vertices on the green line x′=x−1/4x'=x-1/4. Their coordinates are

Orbit space projected on the $(x,x')$ plane

Figure 2. Orbit space projected on the (x,x′)(x,x') plane. The black line contains neutral configurations. The green line contains the charge-breaking vertices. Red line: the three tt-paths (they all coincide in this projection). Blue line: the qq-paths (they all coincide in this projection).

  • point A‾\overline{A}: (x,x′,y,z)=14(1,0,0,0)(x,x',y,z)=\frac{1}{4}(1,0,0,0);

  • point B‾\overline{B}: (x,x′,y,z)=112(4,1,1,0)(x,x',y,z)=\frac{1}{12}(4,1,1,0);

  • point C‾\overline{C}: (x,x′,y,z)=124(8,2,−1,3)(x,x',y,z)=\frac{1}{24}(8,2,-1,\sqrt{3})

  • point C′‾\overline{C'}: (x,x′,y,z)=124(8,2,−1,−3)(x,x',y,z)=\frac{1}{24}(8,2,-1,-\sqrt{3})

  • circumference D‾\overline{D}: (x,x′,y,z)=116(5,1,cos⁡2α,sin⁡2α)(x,x',y,z)=\frac{1}{16}(5,1,\cos2\alpha,\sin2\alpha).

As first noticed for the S4S_4-symmetric model in [33], these vertices can only be reached through charge-breaking field configurations. Examples of such configurations are

A‾:χ1=(01),χ2=(10),χ3=(00);(25)\overline{A}: \quad\chi_1=\begin{pmatrix}0\\1\end{pmatrix},\quad\chi_2=\begin{pmatrix}1\\0\end{pmatrix},\quad\chi_3=\begin{pmatrix}0\\0\end{pmatrix}; \tag*{(25)}
B‾:χ1=(ω2ω),χ2=(ωω2),χ3=(11);(26)\overline{B}: \quad\chi_1=\begin{pmatrix}\omega^2\\\omega\end{pmatrix},\quad\chi_2=\begin{pmatrix}\omega\\\omega^2\end{pmatrix},\quad\chi_3=\begin{pmatrix}1\\1\end{pmatrix}; \tag*{(26)}
C‾:χ1=(ω1),χ2=(ω21),χ3=(11);(27)\overline{C}: \quad\chi_1=\begin{pmatrix}\omega\\1\end{pmatrix},\quad\chi_2=\begin{pmatrix}\omega^2\\1\end{pmatrix},\quad\chi_3=\begin{pmatrix}1\\1\end{pmatrix}; \tag*{(27)}
C′‾:χ1=(ω21),χ2=(ω1),χ3=(11);(28)\overline{C'}: \quad\chi_1=\begin{pmatrix}\omega^2\\1\end{pmatrix},\quad\chi_2=\begin{pmatrix}\omega\\1\end{pmatrix},\quad\chi_3=\begin{pmatrix}1\\1\end{pmatrix}; \tag*{(28)}
D‾:χ1=(0ρ),χ2=(10),χ3=(eiα0).(29)\overline{D}: \quad\chi_1=\begin{pmatrix}0\\\rho\end{pmatrix},\quad\chi_2=\begin{pmatrix}1\\0\end{pmatrix},\quad\chi_3=\begin{pmatrix}e^{i\alpha}\\0\end{pmatrix}. \tag*{(29)}

As the naming suggests, each charge-breaking configuration has an associated neutral configuration: at the charge-breaking configuration, Σ(Λ1,Λ2,Λ3,Λ4)\Sigma(\Lambda_1,\Lambda_2,\Lambda_3,\Lambda_4) is one quarter of its value at the corresponding neutral configuration. Thus,

Σ‾A=ΣA4,Σ‾B=ΣB4,Σ‾C=ΣC4,Σ‾C′=ΣC′4,Σ‾D=ΣD4.(30)\overline{\Sigma}_{A}=\frac{\Sigma_{A}}{4},\qquad\overline{\Sigma}_{B}=\frac{\Sigma_{B}}{4},\qquad\overline{\Sigma}_{C}=\frac{\Sigma_{C}}{4},\qquad\overline{\Sigma}_{C'}=\frac{\Sigma_{C'}}{4},\qquad\overline{\Sigma}_{D}=\frac{\Sigma_{D}}{4}. \tag*{(30)}

Connecting the neutral and charge-breaking vertices

Mathematical preliminary

Consider the function f:R→Rf:\mathbb{R}\to\mathbb{R} defined as

f(u)=au2+2ku+c,(31)f(u)=au^2+2ku+c, \tag*{(31)}

where the coefficients aa, kk, and cc are real. Suppose one wants to guarantee that f(u)≥0f(u)\geq0 in the whole domain u∈[−1,1]u\in[-1,1]. This, of course, implies that both f(−1)≥0f(-1)\geq0 and f(1)≥0f(1)\geq0. Thus, one must enforce

a−2k+c≥0,a+2k+c≥0.(32)a-2k+c\geq0,\qquad a+2k+c\geq0. \tag*{(32)}

Additionally, if f(u)f(u) has a minimum at the point u=u0u=u_0, and if u0u_0 happens to be inside the interval ]−1,1[]-1,1[, then one must require f(u0)≥0f(u_0)\geq0. This means that

if a>0,a−k>0,anda+k>0,thenc−k2a≥0.(33)\text{if }a>0,\quad a-k>0,\quad\text{and}\quad a+k>0,\quad\text{then}\quad c-\frac{k^2}{a}\geq0. \tag*{(33)}

The tt-paths

tt-path 1

One may connect vertices BB, C‾\overline{C}, and C′C' at Eqs. (7b), (27), and (16), respectively, through the Ansatz

χ1=(ω1+u1−u),χ2=(ω21+u1−u),χ3=(1+u1−u),(34)\chi_1=\left(\frac{\omega\sqrt{1+u}}{\sqrt{1-u}}\right),\qquad\chi_2=\left(\frac{\omega^2\sqrt{1+u}}{\sqrt{1-u}}\right),\qquad\chi_3=\left(\frac{\sqrt{1+u}}{\sqrt{1-u}}\right), \tag*{(34)}

where u∈[−1,1]u\in[-1,1]. Indeed, when u=−1u=-1, Eq. (34) reproduces Eq. (7b); when u=0u=0, it reproduces Eq. (11c); and when u=1u=1, it reproduces Eq. (7d). Now, configuration (34) produces A11=A22=A33=2A_{11}=A_{22}=A_{33}=2 and A12=A23=A31=−ω2(1+iu3)A_{12}=A_{23}=A_{31}=-\omega^2(1+iu\sqrt{3}), hence

N=6,(35)N=6, \tag*{(35)}
x=13,(36)x=\frac{1}{3}, \tag*{(36)}
x′=3u2+112,(37)x' = \frac{3u^{2}+1}{12}, \tag*{(37)}
y=3u2−6u−124,(38)y = \frac{3u^{2}-6u-1}{24}, \tag*{(38)}
z=3(−3u2−2u+1)24.(39)z = \frac{\sqrt{3}\left(-3u^{2}-2u+1\right)}{24}. \tag*{(39)}

We call the orbit-space line given (in parametric form) by Eqs. (17b)–(17e) the tt-path 1.

Plugging the values (17) into Eq. (7) one obtains V4=12(au2+2ku+c)V_{4}=12\left(au^{2}+2ku+c\right), with

a=9Λ1+3Λ2−33Λ416,(40)a = \frac{9\Lambda_{1}+3\Lambda_{2}-3\sqrt{3}\Lambda_{4}}{16}, \tag*{(40)}
k=−3Λ1+3Λ2−3Λ416,(41)k = \frac{-3\Lambda_{1}+3\Lambda_{2}-\sqrt{3}\Lambda_{4}}{16}, \tag*{(41)}
c=Λ0+Λ1+3Λ2+3Λ416.(42)c = \Lambda_{0}+\frac{\Lambda_{1}+3\Lambda_{2}+\sqrt{3}\Lambda_{4}}{16}. \tag*{(42)}

So, according to the previous subsection, one must demand

Λ0+ΣB≥0,Λ0+ΣC‾′≥0,(43)\Lambda_{0}+\Sigma_{B}\geq0,\qquad\Lambda_{0}+\Sigma_{\overline{C}'}\geq0, \tag*{(43)}

and, besides,

if ∣−3Λ1+3Λ2−3Λ4∣≤9Λ1+3Λ2−33Λ4, then Λ0+Σt1≥0,(44)\text{if }\left|-3\Lambda_{1}+3\Lambda_{2}-\sqrt{3}\Lambda_{4}\right|\leq9\Lambda_{1}+3\Lambda_{2}-3\sqrt{3}\Lambda_{4},\text{ then }\Lambda_{0}+\Sigma_{t_{1}}\geq0, \tag*{(44)}

where

Σt1=4Λ1Λ2−Λ424(3Λ1+Λ2−3Λ4).(45)\Sigma_{t_{1}}=\frac{4\Lambda_{1}\Lambda_{2}-\Lambda_{4}^{2}}{4\left(3\Lambda_{1}+\Lambda_{2}-\sqrt{3}\Lambda_{4}\right)}. \tag*{(45)}

Since point C‾\overline{C} lies in tt-path 1, the BFB condition Λ0+ΣC‾≥0\Lambda_{0}+\Sigma_{\overline{C}}\geq0 is made redundant.

t-path 2

In a similar fashion to what has been done in the previous subsection, one may connect vertices BB, C‾′\overline{C}', and CC at Eqs. (7b), (11d), and (7c), respectively, through the Ansatz

χ1=(ω21+u1−u),χ2=(ω1+u1−u),χ3=(1+u1−u),(46)\chi_{1}=\left(\frac{\omega^{2}\sqrt{1+u}}{\sqrt{1-u}}\right),\qquad \chi_{2}=\left(\frac{\omega\sqrt{1+u}}{\sqrt{1-u}}\right),\qquad \chi_{3}=\left(\frac{\sqrt{1+u}}{\sqrt{1-u}}\right), \tag*{(46)}

which is just Ansatz (16) with χ1↔χ2\chi_{1}\leftrightarrow\chi_{2}. Indeed, when u=−1u=-1, Eq. (46) reproduces Eq. (7b); when u=0u=0, it reproduces Eq. (11d); and when u=1u=1, it reproduces Eq. (7c).

As is obvious from Eqs. (5), making χ1↔χ2\chi_1 \leftrightarrow\chi_2 is equivalent to making z→−zz \to-z. In Eq. (7) this is equivalent to making Λ4→−Λ4\Lambda_4 \to-\Lambda_4. So, one must demand

Λ0+ΣB≥0,Λ0+ΣC≥0,(47)\Lambda_0+\Sigma_B\geq0,\qquad\Lambda_0+\Sigma_C\geq0, \tag*{(47)}

and, besides,

if ∣−3Λ1+3Λ2+3Λ4∣<9Λ1+3Λ2+33Λ4, then Λ0+Σt2≥0,(48)\text{if }\left|-3\Lambda_1+3\Lambda_2+\sqrt{3}\Lambda_4\right|<9\Lambda_1+3\Lambda_2+3\sqrt{3}\Lambda_4,\text{ then }\Lambda_0+\Sigma_{t2}\geq0, \tag*{(48)}

where

Σt2=4Λ1Λ2−Λ424(3Λ1+Λ2+3Λ4).(49)\Sigma_{t2}=\frac{4\Lambda_1\Lambda_2-\Lambda_4^2}{4\left(3\Lambda_1+\Lambda_2+\sqrt{3}\Lambda_4\right)}. \tag*{(49)}

We call tt-path 2 the orbit-space line given by Eqs. (17b)–(17e) with z→−zz\to-z. Since point C′C' lies in tt-path 2, the BFB condition Λ0+ΣC′≥0\Lambda_0+\Sigma_{C'}\geq0 is redundant.

tt-path 3

One may connect vertices C′C', B‾\overline{B}, and CC through the Ansatz

χ1=(ω21+uω1−u),χ2=(ω1+uω21−u),χ3=(1+u1−u).(50)\chi_1=\begin{pmatrix}\omega^2\sqrt{1+u}\\ \omega\sqrt{1-u}\end{pmatrix},\qquad \chi_2=\begin{pmatrix}\omega\sqrt{1+u}\\ \omega^2\sqrt{1-u}\end{pmatrix},\qquad \chi_3=\begin{pmatrix}\sqrt{1+u}\\ \sqrt{1-u}\end{pmatrix}. \tag*{(50)}

When u=−1u=-1, Eq. (50) reproduces Eq. (7d) with the upper and lower components of all three doublets interchanged; when u=0u=0, it reproduces Eq. (11b); and when u=1u=1, it reproduces Eq. (7c). Configuration (26) produces A11=A22=A33=2A_{11}=A_{22}=A_{33}=2 and A12=A23=A31=−1−iu3A_{12}=A_{23}=A_{31}=-1-iu\sqrt{3}, hence

N=6,(51)N=6, \tag*{(51)}
x=13,(52)x=\frac{1}{3}, \tag*{(52)}
x′=3u2+112,(53)x'=\frac{3u^2+1}{12}, \tag*{(53)}
y=−3u2+112,(54)y=\frac{-3u^2+1}{12}, \tag*{(54)}
z=36u.(55)z=\frac{\sqrt{3}}{6}u. \tag*{(55)}

We call the orbit-space line parametrically given by Eqs. (27b)–(27e) the tt-path 3. Plugging the values (27) into Eq. (7) one obtains V4=12(aˉu2+2kˉu+cˉ)V_4=12\left(\bar{a}u^2+2\bar{k}u+\bar{c}\right), with

aˉ=3Λ24,(56)\bar{a}=\frac{3\Lambda_2}{4}, \tag*{(56)}
kˉ=3Λ48,(57)\bar{k} = \frac{\sqrt{3}\Lambda_{4}}{8}, \tag*{(57)}
cˉ=Λ0+Λ14.(58)\bar{c} = \Lambda_{0} + \frac{\Lambda_{1}}{4}. \tag*{(58)}

Accordingly, one must demand

Λ0+ΣC′≥0,Λ0+ΣC≥0,(59)\Lambda_{0} + \Sigma_{C'} \ge0,\qquad\Lambda_{0} + \Sigma_{C} \ge0, \tag*{(59)}

and, besides,

if ∣Λ4∣<23Λ2, then Λ0+Σt3≥0,(60)\text{if } \lvert\Lambda_{4}\rvert< 2\sqrt{3}\Lambda_{2},\text{ then }\Lambda_{0} + \Sigma_{t3} \ge0, \tag*{(60)}

where

Σt3=4Λ1Λ2−Λ4216Λ2.(61)\Sigma_{t3} = \frac{4\Lambda_{1}\Lambda_{2} - \Lambda_{4}^{2}}{16\Lambda_{2}}. \tag*{(61)}

Since point Bˉ\bar{B} lies in tt-path 3, the BFB condition Λ0+ΣBˉ≥0\Lambda_{0} + \Sigma_{\bar{B}} \ge0 becomes redundant.

Depiction of the tt-paths

The tt-paths 1, 2, and 3 all have

x=13,x′=3u2+112(62)x = \frac{1}{3},\qquad x' = \frac{3u^{2}+1}{12} \tag*{(62)}

for u∈[−1,1]u \in[-1,1], and therefore they are all represented by the red vertical line in Figure 2.

Figure 3 shows the orbit space projected onto the x=1/3x = 1/3 hyperplane, cf. Eq. (62). The three tt-paths are marked in that figure.

The qq-paths

Both at point AA and at circumferences Dˉ\bar{D} and DD one has A12=A31=0A_{12} = A_{31} = 0. We connect those three configurations through the Ansatz

χ1=(0ρ1−u),χ2=(1+u0),χ3=(eiα1+u0),(63)\chi_{1} = \begin{pmatrix} 0 \\ \rho\sqrt{1-u} \end{pmatrix},\qquad \chi_{2} = \begin{pmatrix} \sqrt{1+u} \\ 0 \end{pmatrix},\qquad \chi_{3} = \begin{pmatrix} e^{i\alpha}\sqrt{1+u} \\ 0 \end{pmatrix}, \tag*{(63)}

with u∈[−1,1]u \in[-1,1]. In orbit space, we call these the qq-paths (there is one path for each value of α\alpha).

Configuration (33) has, besides A12=A31=0A_{12} = A_{31} = 0, A23=eiα(1+u)A_{23} = e^{i\alpha}(1+u), A11=2(1−u)A_{11} = 2(1-u), and A22=A33=1+uA_{22} = A_{33} = 1+u. Hence,

N=4,(64)N = 4, \tag*{(64)}
x=−3u2+2u+516,(65)x = \frac{-3u^{2}+2u+5}{16}, \tag*{(65)}
Orbit space projected on the $x = 1/3$ hyperplane

Figure 3. Orbit space projected on the x=1/3x = 1/3 hyperplane. Top (in yellow): the neutral configurations and the vertices BB, CC, and C′C'. Bottom (in yellow): the vertices Bˉ\bar{B}, Cˉ\bar{C}, and Cˉ′\bar{C}'. In grey: all the points of orbit space. The three tt-paths, each of them joining two neutral vertices through a charge-breaking vertex, are depicted in red (tt-path 1), brown (tt-path 2), and pink (tt-path 3$).

x′=u2+2u+116,(66)x' = \frac{u^2 + 2u + 1}{16}, \tag*{(66)}
y=x′cos⁡2α,(67)y = x' \cos2\alpha, \tag*{(67)}
z=x′sin⁡2α.(68)z = x' \sin2\alpha. \tag*{(68)}

So, defining

s(α):=Λ1(1+cos⁡2α)+Λ2(1−cos⁡2α)+Λ4sin⁡2α2,(69)s(\alpha) := \frac{\Lambda_1(1 + \cos2\alpha) + \Lambda_2(1 - \cos2\alpha) + \Lambda_4 \sin2\alpha}{2}, \tag*{(69)}

it has

V4=163[9Λ3+3s(α)16u2+−6Λ3+6s(α)16u+Λ0+Λ3+3s(α)16].(70)V_4 = \frac{16}{3}\left[\frac{9\Lambda_3 + 3s(\alpha)}{16}u^2 + \frac{-6\Lambda_3 + 6s(\alpha)}{16}u + \Lambda_0 + \frac{\Lambda_3 + 3s(\alpha)}{16}\right]. \tag*{(70)}

Therefore, one must demand

Λ0+ΣA≥0,Λ0+ΣD≥0,(71)\Lambda_0 + \Sigma_A \ge0,\qquad\Lambda_0 + \Sigma_D \ge0, \tag*{(71)}

and, besides,

if Λ3>0 and Λ3+s(α)>0, then Λ0+Λ3s(α)3Λ3+s(α)≥0.(72)\text{if } \Lambda_3 > 0 \text{ and } \Lambda_3 + s(\alpha) > 0,\text{ then } \Lambda_0 + \frac{\Lambda_3s(\alpha)}{3\Lambda_3 + s(\alpha)} \ge0. \tag*{(72)}

Getting rid of the α\alpha-dependence of conditions (38) takes some work. Firstly, we note that, from definition (35),

Λ1+Λ2−R2≤s(α)≤Λ1+Λ2+R2.(73)\frac{\Lambda_1 + \Lambda_2 - R}{2} \le s(\alpha) \le\frac{\Lambda_1 + \Lambda_2 + R}{2}. \tag*{(73)}

Secondly, we note that the second inequality (72) together with the negation of the third inequality (72) are equivalent to

−Λ3<s(α)<−3Λ0Λ3Λ0+Λ3.(74)-\Lambda_{3}<s(\alpha)<\frac{-3\Lambda_{0}\Lambda_{3}}{\Lambda_{0}+\Lambda_{3}}. \tag*{(74)}

(Remember that the first condition (71) is precisely Λ0+Λ3≥0\Lambda_{0}+\Lambda_{3}\geq0, which we therefore take for granted.) So, when Λ3>0\Lambda_{3}>0 we need to impede that any s(α)s(\alpha) in interval (73) also is in interval (74). One possibility of doing this is by setting

−3Λ0Λ3Λ0+Λ3≤−Λ3⇔Λ0≥Λ32,(75)\frac{-3\Lambda_{0}\Lambda_{3}}{\Lambda_{0}+\Lambda_{3}}\leq-\Lambda_{3}\quad\Leftrightarrow\quad\Lambda_{0}\geq\frac{\Lambda_{3}}{2}, \tag*{(75)}

which means that interval (74) simply does not exist. We concentrate on the case where inequality (75) does not hold. Then, one must impede that the two intervals intersect. Hence,

if Λ3>max⁡(0,2Λ0), then either Λ1+Λ2+R2≤−Λ3 or Λ1+Λ2−R2≥−3Λ0Λ3Λ0+Λ3.(76)\text{if }\Lambda_{3}>\max(0,2\Lambda_{0}),\text{ then either }\frac{\Lambda_{1}+\Lambda_{2}+R}{2}\leq-\Lambda_{3}\text{ or }\frac{\Lambda_{1}+\Lambda_{2}-R}{2}\geq\frac{-3\Lambda_{0}\Lambda_{3}}{\Lambda_{0}+\Lambda_{3}}. \tag*{(76)}

Since circumference D‾\overline{D} is in all qq-paths—it corresponds to u=0u=0 in Ansatz (63)—the BFB condition Λ0+ΣD‾≥0\Lambda_{0}+\Sigma_{\overline{D}}\geq0 is now superfluous.

Conclusions and hypothesis

We now summarize our findings: in order to guarantee that the SP of the A4A_{4}-invariant 3HDM is BFB, one must

  • Enforce inequalities5

Λ0+Λ3≥0,(77)\Lambda_{0}+\Lambda_{3}\geq0, \tag*{(77)}
Λ0+Λ1≥0,(78)\Lambda_{0}+\Lambda_{1}\geq0, \tag*{(78)}
Λ0+Λ1+3Λ2−3∣Λ4∣4≥0,(79)\Lambda_{0}+\frac{\Lambda_{1}+3\Lambda_{2}-\sqrt{3}\lvert\Lambda_{4}\rvert}{4}\geq0, \tag*{(79)}
Λ0+Λ34+3(Λ1+Λ2−R)8≥0,(80)\Lambda_{0}+\frac{\Lambda_{3}}{4}+\frac{3(\Lambda_{1}+\Lambda_{2}-R)}{8}\geq0, \tag*{(80)}
Λ0+Λ34≥0.(81)\Lambda_{0}+\frac{\Lambda_{3}}{4}\geq0. \tag*{(81)}
  • Impede the situation where

∣Λ4∣−23Λ2<0,(82)\lvert\Lambda_{4}\rvert-2\sqrt{3}\Lambda_{2}<0, \tag*{(82)}
Λ0+4Λ1Λ2−Λ4216Λ2<0.(83)\Lambda_0+\frac{4\Lambda_1\Lambda_2-\Lambda_4^2}{16\Lambda_2}<0. \tag*{(83)}
  • Forbid the situation where

∣3(Λ2−Λ1)−3Λ4∣−3(3Λ1+Λ2−3Λ4)<0,(84)\left|3(\Lambda_2-\Lambda_1)-\sqrt{3}\Lambda_4\right|-3\left(3\Lambda_1+\Lambda_2-\sqrt{3}\Lambda_4\right)<0, \tag*{(84)}
Λ0+4Λ1Λ2−Λ424(3Λ1+Λ2−3Λ4)<0.(85)\Lambda_0+\frac{4\Lambda_1\Lambda_2-\Lambda_4^2}{4(3\Lambda_1+\Lambda_2-\sqrt{3}\Lambda_4)}<0. \tag*{(85)}
  • Forbid the situation where

∣3(Λ2−Λ1)+3Λ4∣−3(3Λ1+Λ2+3Λ4)<0,(86)\left|3(\Lambda_2-\Lambda_1)+\sqrt{3}\Lambda_4\right|-3\left(3\Lambda_1+\Lambda_2+\sqrt{3}\Lambda_4\right)<0, \tag*{(86)}
Λ0+4Λ1Λ2−Λ424(3Λ1+Λ2+3Λ4)<0.(87)\Lambda_0+\frac{4\Lambda_1\Lambda_2-\Lambda_4^2}{4(3\Lambda_1+\Lambda_2+\sqrt{3}\Lambda_4)}<0. \tag*{(87)}
  • Discard cases where

max⁡(0,2Λ0)−Λ3<0,(88)\max(0,2\Lambda_0)-\Lambda_3<0, \tag*{(88)}
max⁡(−Λ1−Λ2−2Λ3,Λ1+Λ2+6Λ0Λ3Λ0+Λ3)−(Λ1−Λ2)2+Λ42<0.(89)\max\left(-\Lambda_1-\Lambda_2-2\Lambda_3,\Lambda_1+\Lambda_2+\frac{6\Lambda_0\Lambda_3}{\Lambda_0+\Lambda_3}\right)-\sqrt{(\Lambda_1-\Lambda_2)^2+\Lambda_4^2}<0. \tag*{(89)}

In our numerical work, we have constructed one million sets of quartic couplings Λ0\Lambda_0, Λ1\Lambda_1, Λ2\Lambda_2, Λ3\Lambda_3, and Λ4\Lambda_4 that satisfy conditions (43).6 We checked how many of our million sets pass the brute-force minimization procedure for V4V_4 implemented in mode 4 of the Mathematica package StableWein [36];7 we found that, out of the one million initial sets, 957 888 sets correspond to BFB potentials, according to the numerical minimization procedure of StableWein. Now, exactly the same 957 888

a1=a2=a3=2(Λ0+Λ3)3,(90)a_1=a_2=a_3=\frac{2(\Lambda_0+\Lambda_3)}{3}, \tag*{(90)}
b1=b2=b3=2Λ0−Λ33,(91)b_1=b_2=b_3=\frac{2\Lambda_0-\Lambda_3}{3}, \tag*{(91)}
c1=c2=c3=Λ1+Λ22,(92)c_1=c_2=c_3=\frac{\Lambda_1+\Lambda_2}{2}, \tag*{(92)}
d1=d2=d3=∣κ∣,(93)d_1=d_2=d_3=|\kappa|, \tag*{(93)}
ϵ=3arg⁡(κ),(94)\epsilon=3\arg(\kappa), \tag*{(94)}

where

κ=12(Λ1−Λ2−iΛ4).(95)\kappa=\frac{1}{2}(\Lambda_1-\Lambda_2-\mathrm{i}\Lambda_4). \tag*{(95)}

set is obtained when one discards, from the original one million sets, those where any of conditions (44)–(47) holds true.8

We therefore conjecture that our necessary conditions for BnessFB are, in fact, necessary and sufficient.

The case of symmetry S4S_{4}

When Λ4=0\Lambda_{4}=0 this model has an exact S4S_{4} symmetry. In that limit, our method for assessing BnessFB reads:

  • Enforce inequalities

Λ0+Λ3≥0,(96)\Lambda_{0}+\Lambda_{3}\ge0, \tag*{(96)}
Λ0+Λ1≥0,(97)\Lambda_{0}+\Lambda_{1}\ge0, \tag*{(97)}
Λ0+Λ1+3Λ24≥0,(98)\Lambda_{0}+\frac{\Lambda_{1}+3\Lambda_{2}}{4}\ge0, \tag*{(98)}
Λ0+Λ34+3min⁡(Λ1,Λ2)4≥0,(99)\Lambda_{0}+\frac{\Lambda_{3}}{4}+\frac{3\min(\Lambda_{1},\Lambda_{2})}{4}\ge0, \tag*{(99)}
Λ0+Λ34≥0.(100)\Lambda_{0}+\frac{\Lambda_{3}}{4}\ge0. \tag*{(100)}
  • Impede the situation where

Λ2>0,(101)\Lambda_{2}>0, \tag*{(101)}
Λ0+Λ14<0.(102)\Lambda_{0}+\frac{\Lambda_{1}}{4}<0. \tag*{(102)}
  • Forbid the situation where

∣Λ2−Λ1∣−3Λ1−Λ2<0,(103)\lvert\Lambda_{2}-\Lambda_{1}\rvert-3\Lambda_{1}-\Lambda_{2}<0, \tag*{(103)}
Λ0+Λ1Λ23Λ1+Λ2<0.(104)\Lambda_{0}+\frac{\Lambda_{1}\Lambda_{2}}{3\Lambda_{1}+\Lambda_{2}}<0. \tag*{(104)}
  • Discard cases where

max⁡(0,2Λ0)−Λ3<0,(105)\max(0,2\Lambda_{0})-\Lambda_{3}<0, \tag*{(105)}
max⁡(−Λ1−Λ2−2Λ3,Λ1+Λ2+6Λ0Λ3Λ0+Λ3)−∣Λ1−Λ2∣<0.(106)\max\left(-\Lambda_{1}-\Lambda_{2}-2\Lambda_{3},\Lambda_{1}+\Lambda_{2}+\frac{6\Lambda_{0}\Lambda_{3}}{\Lambda_{0}+\Lambda_{3}}\right)-\lvert\Lambda_{1}-\Lambda_{2}\rvert<0. \tag*{(106)}

On the other hand, Ivanov & Vazão [33] have stated that the n&s conditions for BnessFB are

Λ0+Λ1≥0,(107)\Lambda_{0}+\Lambda_{1}\ge0, \tag*{(107)}
Λ0+Λ3≥0,(108)\Lambda_{0}+\Lambda_{3}\ge0, \tag*{(108)}
Λ0+Λ1+3Λ24≥0,(109)\Lambda_{0}+\frac{\Lambda_{1}+3\Lambda_{2}}{4}\ge0, \tag*{(109)}
Λ0+Λ3+3Λ24≥0,(110)\Lambda_{0}+\frac{\Lambda_{3}+3\Lambda_{2}}{4}\ge0, \tag*{(110)}
Λ0+Λ14≥0,(111)\Lambda_{0}+\frac{\Lambda_{1}}{4}\ge0, \tag*{(111)}
Λ0+Λ34≥0,(112)\Lambda_{0}+\frac{\Lambda_{3}}{4}\ge0, \tag*{(112)}
Λ0+Λ1+3Λ216≥0,(113)\Lambda_{0}+\frac{\Lambda_{1}+3\Lambda_{2}}{16}\ge0, \tag*{(113)}
Λ0+Λ3+3Λ216≥0;(114)\Lambda_{0}+\frac{\Lambda_{3}+3\Lambda_{2}}{16}\ge0; \tag*{(114)}

besides, the situation where

Λ1−∣Λ2∣>0,(115)\Lambda_{1}-|\Lambda_{2}|>0, \tag*{(115)}
Λ0+Λ1Λ23Λ1+Λ2<0(116)\Lambda_{0}+\frac{\Lambda_{1}\Lambda_{2}}{3\Lambda_{1}+\Lambda_{2}}<0 \tag*{(116)}

must be avoided, and the situation where

Λ3−∣Λ2∣>0,(117)\Lambda_{3}-|\Lambda_{2}|>0, \tag*{(117)}
Λ0+Λ2Λ3Λ2+3Λ3<0(118)\Lambda_{0}+\frac{\Lambda_{2}\Lambda_{3}}{\Lambda_{2}+3\Lambda_{3}}<0 \tag*{(118)}

must also be avoided.

We have numerically checked that the two methods for guaranteeing BnessFB are in practice equivalent.

Acknowledgements: We thank Darius Jurčiukonis for independently performing the numerical check mentioned in section 4. Our work was supported by the Portuguese Foundation for Science and Technology (FCT) through projects UIDB/00777/2020 and UIDP/00777/2020, and by the Recovery and Resilience Plan within the scope of investment RE-C06-i06, measure RE-C06-i06.m02, project 2024.01362.CERN. The work of A.M. was supported by FCT - Fundação para a Ciência e Tecnologia, I.P. by project reference 2024.01340.BD and DOI identifier 10.54499/2024.01340.BD.

References

  1. [1]L. Lavoura and H. Kuhbock, A(4) model for the quark mass matrices, Eur. Phys. J. C 55 (2008) 303–308, [arXiv:0711.0670].DOI
  2. [2]A. C. B. Machado, J. C. Montero, and V. Pleitez, Three-Higgs-doublet model with A₄ symmetry, Phys. Lett. B 697 (2011) 318–322, [arXiv:1011.5855].DOI
  3. [3]R. de Adelhart Toorop, F. Bazzocchi, L. Merlo, and A. Paris, Constraining Flavour Symmetries At The EW Scale I: The A₄ Higgs Potential, JHEP 03 (2011) 035, [arXiv:1012.1791]. [Erratum: JHEP 01, 098 (2013)].DOI
  4. [4]R. de Adelhart Toorop, F. Bazzocchi, L. Merlo, and A. Paris, Constraining Flavour Symmetries At The EW Scale II: The Fermion Processes, JHEP 03 (2011) 040, [arXiv:1012.2091].DOI
  5. [5]E. Ma, Quark and Lepton Flavor Triality, Phys. Rev. D 82 (2010) 037301, [arXiv:1006.3524].DOI
  6. [6]R. González Felipe, H. Serôdio, and J. P. Silva, Models with three Higgs doublets in the triplet representations of A₄ or S₄, Phys. Rev. D 87 (2013), no. 5 055010, [arXiv:1302.0861].DOI
  7. [7]R. González Felipe, I. P. Ivanov, C. C. Nishi, H. Serôdio, and J. P. Silva, Constraining multi-Higgs flavour models, Eur. Phys. J. C 74 (2014), no. 7 2953, [arXiv:1401.5807].DOI
  8. [8]I. Bree, S. Carrolo, J. C. Romao, and J. P. Silva, A viable A₄ 3HDM theory of quark mass matrices, Eur. Phys. J. C 83 (2023), no. 4 292, [arXiv:2301.04676].DOI
  9. [9]E. Ma and G. Rajasekaran, Softly broken A(4) symmetry for nearly degenerate neutrino masses, Phys. Rev. D 64 (2001) 113012, [hep-ph/0106291].DOI
  10. [10]E. Ma, A(4) symmetry and neutrinos with very different masses, Phys. Rev. D 70 (2004) 031901, [hep-ph/0404199].
  11. [11]S. Morisi and E. Peinado, An A(4) model for lepton masses and mixings, Phys. Rev. D 80 (2009) 113011, [arXiv:0910.4389].DOI
  12. [12]R. Gonzalez Felipe, H. Serodio, and J. P. Silva, Neutrino masses and mixing in A₄ models with three Higgs doublets, Phys. Rev. D 88 (2013), no. 1 015015, [arXiv:1304.3468].DOI
  13. [13]E. Ma, Neutrino mixing: A₄ variations, Phys. Lett. B 752 (2016) 198–200, [arXiv:1510.02501].
  14. [14]Y. Izawa, Y. Shimizu, and H. Takei, Non-SUSY lepton flavor model with the three Higgs doublet model, PTEP 2023 (2023), no. 6 063B04, [arXiv:2209.10201].
  15. [15]B. Dziewit, M. Zralek, and J. Vergeest, Partial A₄ Flavor Symmetry of the Leptonic 3HDM, Acta Phys. Polon. Supp. 19 (2026), no. 2 2–A16, [arXiv:2512.10736].arxiv.org/abs/2512.10736
  16. [16]S. C. Chuliá and R. Kumar, Minimal A₄ type-II seesaw realization of testable neutrino mass sum rules, Phys. Rev. D 113 (2026), no. 5 055023, [arXiv:2512.22343].
  17. [17]M. Hirsch, S. Morisi, E. Peinado, and J. W. F. Valle, Discrete dark matter, Phys. Rev. D 82 (2010) 116003, [arXiv:1007.0871].DOI
  18. [18]J. N. Esteves, F. R. Joaquim, A. S. Joshipura, J. C. Romao, M. A. Tortola, and J. W. F. Valle, A₄-based neutrino masses with Majoron decaying dark matter, Phys. Rev. D 82 (2010) 073008, [arXiv:1007.0898].DOI
  19. [19]D. Meloni, S. Morisi, and E. Peinado, Neutrino phenomenology and stable dark matter with A₄, Phys. Lett. B 697 (2011) 339–342, [arXiv:1011.1371].DOI
  20. [20]M. S. Boucenna, M. Hirsch, S. Morisi, E. Peinado, M. Taoso, and J. W. F. Valle, Phenomenology of Dark Matter from A₄ Flavor Symmetry, JHEP 05 (2011) 037, [arXiv:1101.2874].DOI
  21. [21]R. de Adelhart Toorop, F. Bazzocchi, and S. Morisi, Quark mixing in the discrete dark matter model, Nucl. Phys. B 856 (2012) 670–681, [arXiv:1104.5676].
  22. [22]I. de Medeiros Varzielas and D. Ivo, Softly-broken A₄ or S₄ 3HDMs with stable states, Eur. Phys. J. C 82 (2022), no. 5 415, [arXiv:2202.00681].
  23. [23]C. Bonilla, J. Herms, O. Medina, and E. Peinado, Discrete dark matter mechanism as the source of neutrino mass scales, JHEP 06 (2023) 078, [arXiv:2301.10811].DOI
  24. [24]R. Kumar, N. Nath, and R. Srivastava, Cutting the scotogenic loop: adding flavor to dark matter, JHEP 12 (2024) 036, [arXiv:2406.00188].DOI
  25. [25]J. Heeck, M. Holthausen, W. Rodejohann, and Y. Shimizu, Higgs → μτ in Abelian and non-Abelian flavor symmetry models, Nucl. Phys. B 896 (2015) 281–310, [arXiv:1412.3671].DOI
  26. [26]G. Bhattacharyya and D. Das, Nondecoupling of charged scalars in Higgs decay to two photons and symmetries of the scalar potential, Phys. Rev. D 91 (2015) 015005, [arXiv:1408.6133].DOI
  27. [27]I. de Medeiros Varzielas, O. Fischer, and V. Maurer, A₄ symmetry at colliders and in the universe, JHEP 08 (2015) 080, [arXiv:1504.03955].
  28. [28]G. C. Branco and I. P. Ivanov, Group-theoretic restrictions on generation of CP-violation in multi-Higgs-doublet models, JHEP 01 (2016) 116, [arXiv:1511.02764].DOI
  29. [29]I. de Medeiros Varzielas, S. F. King, C. Luhn, and T. Neder, CP-odd invariants for multi-Higgs models: applications with discrete symmetry, Phys. Rev. D 94 (2016), no. 5 056007, [arXiv:1603.06942].
  30. [30]S. Pramanick and A. Raychaudhuri, Three-Higgs-doublet model under A₄ symmetry implies alignment, JHEP 01 (2018) 011, [arXiv:1710.04433].DOI
  31. [31]N. Chakrabarty and I. Chakraborty, On the Higgs mass fine-tuning problem with multi-Higgs doublet models, Int. J. Mod. Phys. A 34 (2019), no. 05 1950025, [arXiv:1801.05272].DOI
  32. [32]A. Degee, I. P. Ivanov, and V. Keus, Geometric minimization of highly symmetric potentials, JHEP 02 (2013) 125, [arXiv:1211.4989].DOI
  33. [33]I. P. Ivanov and F. Vazão, Yet another lesson on the stability conditions in multi-Higgs potentials, JHEP 11 (2020) 104, [arXiv:2006.00036].DOI
  34. [34]S. Carrol[o], J. C. Romao, and J. P. Silva, Conditions for global minimum in the A₄ symmetric 3HDM, Eur. Phys. J. C 82 (2022), no. 8 749, [arXiv:2207.02928].
  35. [35]N. Buskin and I. P. Ivanov, Bounded-from-below conditions for A₄-symmetric 3HDM, J. Phys. A 54 (2021) 325401, [arXiv:2104.11428].DOI
  36. [36]D. Jurčiukonis, L. Lavoura, and A. Milagre, Assessing boundedness from below in the Z₂ × Z₂-symmetric three-Higgs-doublet model: algorithm and machine learning, SciPost Phys. Core 9 (2026) 041, [arXiv:2603.23590].DOI

Paper details

vlnc:W_rqnsz829
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Claude

View statement
Looking for related papers…
Contents