KEK-TH-2452

Flavor Hierarchy from Smooth Confinement

arXiv:2209.15244v2 [hep-ph] 20 Oct 2022

Yuta Hamada∗ and Juven Wang†

?
Theory Center, IPNS, High Energy Accelerator Research Organization (KEK),
1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan
?
Graduate University for Advanced Studies (Sokendai), 1-1 Oho, Tsukuba,
Ibaraki 305-0801, Japan
?
Department of Physics, Harvard University, Cambridge, MA 02138 USA
†
Center of Mathematical Sciences and Applications, Harvard University, MA 02138, USA

Abstract

We present a model to explain the Standard Model flavor hierarchy. Our model is based on explicit
smooth confinement (namely confinement without chiral symmetry breaking in a supersymmetric
gauge theory) at an intermediate energy scale, before the electroweak symmetry breaking by the
Higgs condensation at lower energy. In our context, the smooth confinement preserves the SU(3)
and the chiral SU(2)L × U(1)Y symmetry in a supersymmetric Standard Model, while this internal
symmetry becomes dynamically gauged in the end. In contrast to Razamat-Tong’s symmetric mass
generation model also preserving the GSM ≡ SU(3)C × SU(2)L × U(1)Y internal symmetry, our model
introduces different matter contents with a different kind of superpotential deformation irrelevant at
UV, which further induces Yukawa-Higgs terms marginal at IR, breaking the GSM down to SU(3)C
and the electromagnetic U(1)EM only when Higgs condenses. In our model, the IR fermions in the
first and second families are composite of UV fields, while the third family elementary fermions match
between UV and IR. The smallness of the first and second family fermion masses is explained by
the exponential hierarchy between the cutoff scale and the smooth confinement scale via dimensional
transmutation. As a result, our UV Lagrangian only contains the natural parameters close to the
order one.

∗
E-mail: yhamada@post.kek.jp
†
E-mail: jw@cmsa.fas.harvard.edu

1
1 Introduction

One of the mysteries of the Standard Model (SM) is the flavor hierarchical structure. The observed
lightest charged fermion in our world is the electron, and its mass is 511 keV [1]. On the other hand,
the heaviest one is the top quark, and its mass is about 172 GeV [2]. Both fermions acquire masses
from the Yukawa coupling with the condensation of the Higgs field. Therefore, the huge difference in
the fermion masses indicates that the Yukawa coupling has a hierarchical structure. More concretely,
the ratio between the electron and top Yukawa coupling is 3 × 105 . Why is there such a large flavor
hierarchy?

One of the known solutions was provided by Froggatt and Nielsen in Ref. [3]. They introduced a
flavor-dependent U(1) symmetry and a scalar field called flavon charged under it. The small parameter
is introduced if the vev of the Higgs is smaller than the cutoff scale.

In this work, we present a new solution, inspired by the dynamics of the nonperturbative strong
coupling of gauge theory, known as the smooth confinement (or the s-confinement [4, 5], which means
confinement without chiral symmetry breaking in the supersymmetry context). The idea is that we
hypothesize some matter fields of the (supersymmetric) SM at IR can be obtained as the composite fields
of UV under s-confinement. Moreover, the SM gauge group GSM ≡ SU(3)C × SU(2)L × U(1)Y (including
the chiral SU(2)L × U(1)Y ) is preserved not spontaneously broken by the s-confinement dynamics.

Recently, the s-confinement is also applied to another problem: the symmetric mass generation (see
[6] for an overview) of the SM in four dimensions [7, 8]. It has been previously stressed that there is
no obstruction to give masses and energy gaps to all of the SM fermions without breaking the GSM
or Grand Unified Theory’s internal symmetry as long as there are no anomalies among these internal
symmetries [9] (see the systematic checks on the local and global anomaly cancellations for the SM in
[10, 11, 12] via cobordism). Ref. [7, 8] provides an explicit model realizing this idea, in the context of
the supersymmetric or non-supersymmetric SM. The key idea in [7, 8] is that the dangerously irrelevant
interaction term in the UV becomes the GSM -preserving relevant mass term to achieve the symmetric
mass generation in the IR.

Although both Razamat-Tong [7] and our models use the s-confinement dynamics at an intermediate
energy scale, the two models (including UV Lagrangian, the deformations, and the IR dynamics) are
rather different. In comparison, Ref. [7] considered enlarging from the SM’s 15Nf or 16Nf Weyl fermions
to the 27Nf Weyl fermions that can be symmetrically gapped out by preserving GSM via a symmetric
mass generation deformation, with the family number Nf . Instead, we consider a different UV model
with 63 Weyl fermions in total (see Table 1, including the 15Nf Weyl fermions in Nf = 3 families)
with a different deformation. In our scenario, the standard Yukawa coupling at IR corresponds to the
higher-scaling-dimensional operator at UV. The GSM -preserving dangerously irrelevant higher-scaling-
dimensional operator at UV becomes the marginal Yukawa coupling at IR. Only when the Higgs field
condenses to get a vev at low energy, then the IR dynamics further spontaneously breaks GSM down
to the electromagnetic U(1)EM . This naturally explains the smallness of the first and second family
Yukawa couplings, where the smallness is a consequence by the ratio between the s-confinement and
cutoff scales (see Fig. 1 (a)). On the other hand, the third family fields are treated as elementary, which
is consistent with O(1) top Yukawa coupling. We present an explicit model to demonstrate the idea. We
show that the flavor hierarchy gets milder in our UV model, the coupling of the first and second family
(i.e., superpotential coefficients), and the third family Yukawa coupling become closer to O(1) at high
energy. Namely, our UV Lagrangian only contains the natural parameters close to the order one. The

2
 Field SU(3)C SU(2)L U(1)Y SU(2)01 SU(2)02 U(1)B−L Z4
L0I=1 1 2 1/2 2 1 −3 1
L0I=2 1 2 1/2 1 2 −3 1
LI=1,2,3 1 2 −1/2 1 1 6 2
0
DI=1 3 1 −1/3 2 1 1 1
0
DI=2 3 1 −1/3 1 2 1 1
DI=1,2,3 3̄ 1 1/3 1 1 2 2
0
SI=1 1 1 0 2 1 3 1
0
SI=2 1 1 0 1 2 3 1
E3 1 1 1 1 1 −6 2
U3 3̄ 1 −2/3 1 1 2 2
Q3 3 2 1/6 1 1 −2 2
Hu 1 2 1/2 1 1 0 0
Hd 1 2 −1/2 1 1 0 0
D̃I=1,2 3̄ 1 1/3 1 1 −4 0
L̃I=1,2 1 2 −1/2 1 1 0 0

Table 1: The UV matter contents of superfields of our N = 1 minimal supersymmetric model in the
left-handed basis. U(1)B−L is the anomaly free global symmetry, and the last column is Z4 subgroup
of U(1)B−L symmetry. There are 22 Weyl fermions from the chiral multiplets for each of the first and
second family (I = 1, 2). There are 15 Weyl fermions from the chiral multiplets for the third family
(I = 3). Each of the two Higgs multiplets introduces 2 Weyl fermions. So the model contains 63 Weyl
fermions in total. If we include additional 3 families of right-handed neutrinos neutral to GSM , there will
be 66 Weyl fermions in total.

physics at different energy scales from UV to intermediate to IR are shown in Fig. 1 (b).

Energy scale
Cutoff: Planck or GUT scales
M
<latexit sha1_base64="1IRxmdNxymXmXq8IFxNa9aLT/vc=">AAAB6nicbVA9SwNBEJ2LXzF+RS1tFhPBKtwFUcugjY0Q0ZhAcoS9zV6yZG/v2J0TQshPsLFQxNZfZOe/cZNcoYkPBh7vzTAzL0ikMOi6305uZXVtfSO/Wdja3tndK+4fPJo41Yw3WCxj3Qqo4VIo3kCBkrcSzWkUSN4MhtdTv/nEtRGxesBRwv2I9pUIBaNopfvybblbLLkVdwayTLyMlCBDvVv86vRilkZcIZPUmLbnJuiPqUbBJJ8UOqnhCWVD2udtSxWNuPHHs1Mn5MQqPRLG2pZCMlN/T4xpZMwoCmxnRHFgFr2p+J/XTjG89MdCJSlyxeaLwlQSjMn0b9ITmjOUI0so08LeStiAasrQplOwIXiLLy+Tx2rFO6+c3VVLtassjjwcwTGcggcXUIMbqEMDGPThGV7hzZHOi/PufMxbc042cwh/4Hz+AF7pjTQ=</latexit>

UV
1
⇤2
<latexit sha1_base64="5c7qhZmBcLzwmUe1xzOa93XGbOE=">AAAB8nicbVDLSgMxFM3UV62vqks3wVZwVWaKqMuiGxcuKtgHTIeSyWTa0EwyJHeEUvoZblwo4tavceffmLaz0NYDgcM555J7T5gKbsB1v53C2vrG5lZxu7Szu7d/UD48ahuVacpaVAmluyExTHDJWsBBsG6qGUlCwTrh6Hbmd56YNlzJRxinLEjIQPKYUwJW8qu9e5uNSL9e7Zcrbs2dA68SLycVlKPZL3/1IkWzhEmgghjje24KwYRo4FSwaamXGZYSOiID5lsqScJMMJmvPMVnVolwrLR9EvBc/T0xIYkx4yS0yYTA0Cx7M/E/z88gvg4mXKYZMEkXH8WZwKDw7H4ccc0oiLElhGpud8V0SDShYFsq2RK85ZNXSbte8y5rFw/1SuMmr6OITtApOkceukINdIeaqIUoUugZvaI3B5wX5935WEQLTj5zjP7A+fwB9OGQZQ==</latexit>
⇡ s-confinement for 2nd family <latexit sha1_base64="+XpfpThBLrgbbfHqddVu2aTMA/s=">AAAB7nicbVBNSwMxEJ2tX7V+VT16CRbBU9mVoh6LXjxWsB/QLiWbZtvQbBKSrFiW/ggvHhTx6u/x5r8xbfegrQ8GHu/NMDMvUpwZ6/vfXmFtfWNzq7hd2tnd2z8oHx61jEw1oU0iudSdCBvKmaBNyyynHaUpTiJO29H4dua3H6k2TIoHO1E0TPBQsJgRbJ3U7mGltHzqlyt+1Z8DrZIgJxXI0eiXv3oDSdKECks4NqYb+MqGGdaWEU6npV5qqMJkjIe066jACTVhNj93is6cMkCx1K6ERXP190SGE2MmSeQ6E2xHZtmbif953dTG12HGhEotFWSxKE45shLNfkcDpimxfOIIJpq5WxEZYY2JdQmVXAjB8surpHVRDS6rtftapX6Tx1GEEziFcwjgCupwBw1oAoExPMMrvHnKe/HevY9Fa8HLZ47hD7zPH5ZQj78=</latexit>

0.1
⇤1
<latexit sha1_base64="SxnYLVTCYD6+QQza18B3ZmaUCuE=">AAAB8nicbVDLSgMxFL1TX7W+qi7dBFvBVZkpoi6Lbly4qGAfMB1KJpNpQzPJkGSEUvoZblwo4tavceffmLaz0NYDgcM555J7T5hypo3rfjuFtfWNza3idmlnd2//oHx41NYyU4S2iORSdUOsKWeCtgwznHZTRXESctoJR7czv/NElWZSPJpxSoMEDwSLGcHGSn61d2+zEe571X654tbcOdAq8XJSgRzNfvmrF0mSJVQYwrHWvuemJphgZRjhdFrqZZqmmIzwgPqWCpxQHUzmK0/RmVUiFEtlnzBorv6emOBE63ES2mSCzVAvezPxP8/PTHwdTJhIM0MFWXwUZxwZiWb3o4gpSgwfW4KJYnZXRIZYYWJsSyVbgrd88ipp12veZe3ioV5p3OR1FOEETuEcPLiCBtxBE1pAQMIzvMKbY5wX5935WEQLTj5zDH/gfP4A81yQZA==</latexit>

s-confinement for 1st family

10-4 Intermediate
Λ2 /M2

energy
⇤2 ⇡
<latexit sha1_base64="V3LN8hZcc56muEZ7ZwJLsVgW2o8=">AAACAHicbVC7TsMwFL0pr1JeAQYGlogWialKKgSMFSwMIBWJPqQmVI7rtFYdJ7IdpCrKwq+wMIAQK5/Bxt/gthmg5Ui2js65R/Y9fsyoVLb9bRSWlldW14rrpY3Nre0dc3evJaNEYNLEEYtEx0eSMMpJU1HFSCcWBIU+I21/dDXx249ESBrxezWOiReiAacBxUhpqWceVNxAIJy6NzrTRw+1LL3VV6Vnlu2qPYW1SJyclCFHo2d+uf0IJyHhCjMkZdexY+WlSCiKGclKbiJJjPAIDUhXU45CIr10ukBmHWulbwWR0Icra6r+TqQolHIc+noyRGoo572J+J/XTVRw4aWUx4kiHM8eChJmqciatGH1qSBYsbEmCAuq/2rhIdKFKN1ZSZfgzK+8SFq1qnNWPb2rleuXeR1FOIQjOAEHzqEO19CAJmDI4Ble4c14Ml6Md+NjNlow8sw+/IHx+QOtp5XS</latexit>

<latexit sha1_base64="+XpfpThBLrgbbfHqddVu2aTMA/s=">AAAB7nicbVBNSwMxEJ2tX7V+VT16CRbBU9mVoh6LXjxWsB/QLiWbZtvQbBKSrFiW/ggvHhTx6u/x5r8xbfegrQ8GHu/NMDMvUpwZ6/vfXmFtfWNzq7hd2tnd2z8oHx61jEw1oU0iudSdCBvKmaBNyyynHaUpTiJO29H4dua3H6k2TIoHO1E0TPBQsJgRbJ3U7mGltHzqlyt+1Z8DrZIgJxXI0eiXv3oDSdKECks4NqYb+MqGGdaWEU6npV5qqMJkjIe066jACTVhNj93is6cMkCx1K6ERXP190SGE2MmSeQ6E2xHZtmbif953dTG12HGhEotFWSxKE45shLNfkcDpimxfOIIJpq5WxEZYY2JdQmVXAjB8surpHVRDS6rtftapX6Tx1GEEziFcwjgCupwBw1oAoExPMMrvHnKe/HevY9Fa8HLZ47hD7zPH5ZQj78=</latexit>

M2 SUSY breaking scale ranging

from TeV to s-confinement scale
10-7

IR ⇤
<latexit sha1_base64="00CJiTFLgVmPd2m6N7qtVxhYmI8=">AAAB+XicbVDLSgMxFM34rPU1KrhxEyyCqzIjoi5LRXDhogX7gM4wZDKZNjTJDEmmUIb+iRsXigiu/AW/wJ0bv8X0sdDWA4HDOedyb06YMqq043xZS8srq2vrhY3i5tb2zq69t99USSYxaeCEJbIdIkUYFaShqWaknUqCeMhIK+xfj/3WgEhFE3GvhynxOeoKGlOMtJEC2/buTDhCQe5JDm9ao8AuOWVnArhI3BkpVQ7r3/St+lEL7E8vSnDGidCYIaU6rpNqP0dSU8zIqOhliqQI91GXdAwViBPl55PLR/DEKBGME2me0HCi/p7IEVdqyEOT5Ej31Lw3Fv/zOpmOr/ycijTTRODpojhjUCdwXAOMqCRYs6EhCEtqboW4hyTC2pRVNCW4819eJM2zsntRPq+bNqpgigI4AsfgFLjgElTALaiBBsBgAB7AE3i2cuvRerFep9ElazZzAP7Aev8B2YqW3A==</latexit>
⇡ 246 GeV SM Higgs h <latexit sha1_base64="+XpfpThBLrgbbfHqddVu2aTMA/s=">AAAB7nicbVBNSwMxEJ2tX7V+VT16CRbBU9mVoh6LXjxWsB/QLiWbZtvQbBKSrFiW/ggvHhTx6u/x5r8xbfegrQ8GHu/NMDMvUpwZ6/vfXmFtfWNzq7hd2tnd2z8oHx61jEw1oU0iudSdCBvKmaBNyyynHaUpTiJO29H4dua3H6k2TIoHO1E0TPBQsJgRbJ3U7mGltHzqlyt+1Z8DrZIgJxXI0eiXv3oDSdKECks4NqYb+MqGGdaWEU6npV5qqMJkjIe066jACTVhNj93is6cMkCx1K6ERXP190SGE2MmSeQ6E2xHZtmbif953dTG12HGhEotFWSxKE45shLNfkcDpimxfOIIJpq5WxEZYY2JdQmVXAjB8surpHVRDS6rtftapX6Tx1GEEziFcwjgCupwBw1oAoExPMMrvHnKe/HevY9Fa8HLZ47hD7zPH5ZQj78=</latexit>

<latexit sha1_base64="5jzQ9TCiKN71RrIP8wZUUdMGyv4=">AAACB3icbVDLSgMxFM3UV62vUZeCBFvBVZkpoi6LbrqsYB/QGYZMetuGZjJjkhFK6c6Nv+LGhSJu/QV3/o1pOwttPXDh5Jx7yb0nTDhT2nG+rdzK6tr6Rn6zsLW9s7tn7x80VZxKCg0a81i2Q6KAMwENzTSHdiKBRCGHVji8mfqtB5CKxeJOjxLwI9IXrMco0UYK7OOSx4noc8BeMmBBzZPZS8A9dkqBXXTKzgx4mbgZKaIM9cD+8roxTSMQmnKiVMd1Eu2PidSMcpgUvFRBQuiQ9KFjqCARKH88u2OCT43Sxb1YmhIaz9TfE2MSKTWKQtMZET1Qi95U/M/rpLp35Y+ZSFINgs4/6qUc6xhPQ8FdJoFqPjKEUMnMrpgOiCRUm+gKJgR38eRl0qyU3Yvy+W2lWL3O4sijI3SCzpCLLlEV1VAdNRBFj+gZvaI368l6sd6tj3lrzspmDtEfWJ8/NDqYPg==</latexit>

Hi 6 0
=
EW
10-10
0.5 1.0 1.5 2.0 2.5 3.0
(a) g(M ) (b) 0
<latexit sha1_base64="VT6yIiYJgXD69S6n9ye3yHwX9bk=">AAAB7XicbVBNSwMxEJ2tX7V+VT16CbZCvZTdIuqx6MWLUMF+QLuUbJptY7PJkmSFsvQ/ePGgiFf/jzf/jWm7B60+GHi8N8PMvCDmTBvX/XJyK6tr6xv5zcLW9s7uXnH/oKVloghtEsml6gRYU84EbRpmOO3EiuIo4LQdjK9nfvuRKs2kuDeTmPoRHgoWMoKNlVrlYeX2tNwvltyqOwf6S7yMlCBDo1/87A0kSSIqDOFY667nxsZPsTKMcDot9BJNY0zGeEi7lgocUe2n82un6MQqAxRKZUsYNFd/TqQ40noSBbYzwmakl72Z+J/XTUx46adMxImhgiwWhQlHRqLZ62jAFCWGTyzBRDF7KyIjrDAxNqCCDcFbfvkvadWq3nn17K5Wql9lceThCI6hAh5cQB1uoAFNIPAAT/ACr450np03533RmnOymUP4BefjG+eMjgo=</latexit>

g(M)

Figure 1: (a) Ratio between the confinement and cutoff scales as a function of the gauge coupling g(M )
at the cutoff scale M . This figure reveals that the tiny change of g(M ) coupling can correspond to a
Λ
huge exponential hierarchy between the confinement and the cutoff scales M . (b) Physics at different
energy scales: The UV Lagrangian is valid somewhere below the cutoff M . The s-confinement happens
at intermediate energy Λ1 , Λ2 while the GSM is still preserved, e.g., ΛM1 , ΛM2 ranges from 10−3 to 10−1 for
phenomenological fitting. Below the supersymmetry breaking scale Λ SUSY
  , the SM Higgs φH (given by
a linear combination of Hu and Hd ) condenses breaking GSM to U(1)EM below ΛEW .

Let us also compare Froggatt-Nielsen model [3] with our model. Froggatt-Nielsen introduced the

3
flavon field, which suffers from the hierarchy problem associated with the quadratic divergence to the
mass squared. Moreover, it is not clear why the vacuum expectation value of the flavon is smaller than
the cutoff scale. One of the advantages of our model is that the new hierarchy problem does not occur
since we use s-confinement rather than the Higgs mechanism. The confinement scale appears as the
result of the dimensional transmutation, and is naturally smaller than the cutoff scale.

The idea of applying the composite fermions [13] and smooth confinement [4, 5] to fermion mass hier-
archy dates back to Strassler [14], Nielsen and others [15]. There were also other earlier supersymmetric
composite models based on technicolor [16] with similar goals. However, there are significant differences
between our model and other previous models: 1. We only require the SU(3) × SU(2) × U(1) SM instead
of the Grand Unified Theory (GUT) group. 2. We choose the third family fermions elementary, while
previous works pursue different routes: such as top-Yukawa-Higgs coupling is dynamically generated, top
quark and bottom quark are composite [15]. 3. We choose the large smooth confinement Λ ' 10−3 M to
10−1 M close to the cutoff scale M , while others choose the much smaller TeV confinement scale [15]. 4.
The smooth confinement scale is also the symmetric mass generation scale in our theory, which naturally
demands that the coupling strengths of the original SM Lagrangian gSM and the deformation interaction
terms gint have an order 1 ratio at our UV theory. This gint /gSM ∼ O(1) (or more precisely the energy
ratio of the action Eint /ESM ∼ O(1)) is due to the fact that the symmetric energy gap is generated
nonperturbatively, so we should not rely on a perturbative renormalization group analysis around fixed
points at either limit gint /gSM  1 or gint /gSM  1.

This article is organized as follows. The UV model is provided in Sec. 2. The model exhibits the
s-confinement at the intermediate energy scale, and the minimal supersymmetric SM is realized in the
IR, as we show in Sec. 3. We show that the flavor hierarchy is solved by comparing with the experimental
data in Sec. 4. The summary and discussions are in Sec. 5.

2 Model in the UV

Inspired by the symmetric mass generation in four dimensions [7], we propose a new model which reduces
to the supersymmetric version of the SM in the IR.

The UV matter contents of our model is listed in Table 1. Regarding the first and second families,
we introduce the superfields (L0I , DI0 , SI0 ) charged under SU(2)01 × SU(2)02 (I = 1, 2 is the index for the
family). There are also superfields (DI , D̃I , LI , L̃I ), which are neutral under SU(2)01 × SU(2)02 . As for
the third family, the matter contents are the same as those of the N = 1 Minimal Supersymmetric SM,
(Q3 , U3 , D3 , L3 , E3 ). On top of that, we introduce the Higgs superfields (Hu , Hd ). These are necessary
to cancel the dynamical gauge anomaly of GSM ≡ SU(3)C × SU(2)L × U(1)Y . In total, there are 63 Weyl
fermions in Table 1.

4
 The superpotential of the model is1

u (L0I DI0 )(DJ0 DJ0 )Hu 0 0

d (LI DI )DJ Hd
0 0
l LI (LJ LJ )Hd
WUV =yIJ + y IJ + y IJ
M2 M M
Q (D 0 D 0 )H (L 0 D 0 )U H 0 0
u 3 I I u u I I 3 u d d (LI DI )D3 Hd
+ y3I + yI3 + y3I Q3 DI Hd + yI3
M M M
0 0
l L3 (LI LI )Hd l u d
+ y3I + yI3 LI E3 Hd + y33 Q3 U3 Hu + y33 Q3 D3 Hd
M
l 0 0 0 0
+ y33 L3 E3 Hd + λD L
I D̃I (DI SI ) + µHu Hd + λIJ L̃I (LJ SJ ) (1)

where M is the cutoff scale of the model. This could be the GUT or Planck scale. This superpoten-
tial possess U(1)B−L defined in Table 1. As we will see, the Z4 subgroup of U(1)B−L is identified as
matter parity in the Minimal Supersymmetric SM. If we do not impose U(1)B−L or Z4 , the following
superpotential is allowed:
LI LJ (L0K L0K ) LI (L0J DJ0 )DK
W∆L=1,UV = λIJK + λ0IJK + µ0I LI Hu + µ00IJ LI (L0J SJ0 ),
M M
(D0 D0 )DJ DK
W∆B=1,UV = λ00IJK I I . (2)
M
These terms must be small in order to avoid the fast proton decay.

3 Effective theory in the IR

The beta function of SU(2)01 × SU(2)02 gauge coupling is negative, and therefore the system is strongly
coupled at IR, and asymptotic free at UV. The one-loop beta functions of the model is
 3
13gI3

dgI 11 1 2 gI
µ =− C2 (G) − ns T (Rs ) − 2nf T (Rf ) = − , I = 1, 2. (3)
dµ 3 3 3 16π 2 48π 2

where gI is the gauge coupling of SU(2)0I at the energy scale µ, the quadratic Casimir is C2 (G) = 2 of
G = SU(2)0I . The number of complex scalars ns = 6 and the number of Weyl fermions 2nf = 6 that
couple to G are all from the 6 chiral multiplets. T (Rs ) = T (Rf ) = 21 for the fundamental representation.
The strongly coupled s-confinement scale ΛI is

48π 2
 
2 2
ΛI = M exp − (4)
13gI2 (M )

where gI (M ) is the coupling at the cutoff scale µ = M .

In the IR, the system exhibits the s-confinement [4, 5], which magic requires nc + 1 = nf where
nc = 2 is from SU(2)0I and nf = 3 implies the 6 Weyl fermions that couple to the SU(2)0I . The IR fields
are the composites of UV fields:

ab L0a 0b
I LI ijk DI0i DI0j ab L0a 0i
I DI D0i S 0 L0 S 0
EI ≡ , UkI ≡ , QibI ≡ , D̃I0i ≡ I I , L̃0I ≡ I I , (5)
ΛI ΛI ΛI ΛI ΛI
To be precise, more terms are allowed since Hd and L̃I have the same quantum number. Similarly, Hu and L0I SI0 has
1

the same quantum number. In generic situation, the last two terms in Eq. (1) will be 3 by 3 mass matrix in the IR. We
assume that one of the linear combination is relatively light, which we call Hu and Hd . The electroweak Higgs is obtained
from a linear-combination of Hu and Hd .

5
where a, b = 1, 2 is SU(2)L index, i, j, k = 1, 2, 3 is SU(3)C index, and ΛI is the confinement scale of
SU(2)0I .

At IR low energy below the s-confinement scale, the superpotential (1) becomes

u ΛI ΛJ d ΛI l ΛJ
WIR = yIJ 2
QI UJ Hu + yIJ QI DJ Hd + yIJ LI EJ Hd
M M M
u ΛI u ΛI d d ΛI
+ y3I Q3 UI Hu + yI3 QI U3 Hu + y3I Q3 DI Hd + yI3 QI D3 Hd
M M M
l ΛI l u d
+ y3I L3 EI Hd + yI3 LI E3 Hd + y33 Q3 U3 Hu + y33 Q3 D3 Hd
M
l 0
+ y33 L3 E3 Hd + µHu Hd + λD L
I ΛI D̃I D̃I + λI ΛI L̃I L̃I
u
≡ ỹαβ d
Qα Uβ Hu + ỹαβ l
Qα Dβ Hd + ỹαβ Lα Eβ Hd + µ̃ Hu Hd + M̃ID D̃I D̃I0 + M̃IL L̃I L̃0I (6)

where α, β = 1, 2, 3, while ỹ u , ỹ d and ỹ l are the low-energy Yukawa couplings.

In the IR, Eq. (2) becomes

W∆L=1,IR = λIJK LI LJ EK + λ0IJK LI QJ DK + µ0I LI Hu + µ00IJ LI L̃0J ,

W∆B=1,IR = λ00IJK UI DJ DK . (7)

Again, this breaks U(1)B−L and its Z4 subgroup.

4 Comparison with experimental data

From [17], the experimental data of the Yukawa coupling is2

u
ỹ11 sin β ∼ 6 × 10−6 , u
ỹ22 sin β ∼ 3 × 10−3 , u
ỹ33 sin β ∼ 0.8,
d
ỹ11 cos β ∼ 1 × 10−5 , d
ỹ22 cos β ∼ 3 × 10−4 , d
ỹ33 cos β ∼ 1 × 10−2 ,
l
ỹ11 cos β ∼ 3 × 10−6 l
ỹ22 cos β ∼ 6 × 10−4 l
ỹ33 cos β ∼ 1 × 10−2 . (8)

The singular value decompositions of Yukawa coupling ỹ u and ỹ d are

ỹ u = U u diag(Yu , Yc , Yt )V u , ỹ d = U d diag(Yd , Ys , Yb )V d , (9)

where U u,d and V u,d are unitary matrices. The CKM matrix is defined as

VCKM = (U u )† U d . (10)

The Wolfenstein parametrization of the CKM matrix is

1 − 12 λ2 Aλ3 (ρ − iη)
 
λ
VCKM =  −λ 1 − 21 λ2 Aλ2  + O(λ4 ) (11)
3
Aλ (1 − ρ − iη) −Aλ 2 1
2
Here tan β is defined as the ratio of the VEV of two Higgs at the electroweak symmetry breaking vacuum, tan β ≡
hHu i/hHd i.

6
 Field SU(3)C SU(2)L U(1)Y U(1)B−L Z4
EI=1,2,3 1 1 1 −6 2
UI=1,2,3 3̄ 1 −2/3 2 2
QI=1,2,3 3 2 1/6 −2 2
LI=1,2,3 1 2 −1/2 6 2
DI=1,2,3 3̄ 1 1/3 2 2
D̃I=1,2 3̄ 1 1/3 −4 0
0
D̃I=1,2 3 1 −1/3 4 0
Hu 1 2 1/2 0 0
Hd 1 2 −1/2 0 0
L̃I 1 2 −1/2 0 0
L̃0I 1 2 1/2 0 0

Table 2: The IR matter contents in the left-handed basis. The IR composite fields are given in Eq. (5).
All fields are singlet under SU(2)01 × SU(2)02 . The original Z4 symmetry is identified as matter-parity [18]
in the IR.

where λ ' 0.23, A ' 0.8, ρ ' 0.14 and η ' 0.35. This means that the observed value of CKM matrix is
 
1 0.2 0.001 − 0.003i
obs
VCKM ∼ 0.2 1 0.04 . (12)
0.01 − 0.003i 0.04 1

Similarly, from [17], the experimental value of Y u,c,t and Y d,s,b is

6 × 10−6 3 × 10−3 0.8

Yuobs ∼ , Ycobs ∼ , Ytobs ∼ ,
sin β sin β sin β
1 × 10−5 3 × 10−4 1 × 10−2
Ydobs ∼ , Ysobs ∼ , Ybobs ∼ ,
cos β cos β cos β
3 × 10−6 6 × 10−4 6 × 10−2
Yeobs ∼ , Yµobs ∼ , Yτobs ∼ . (13)
cos β cos β cos β

Let us estimate the Yukawa couplings in our model. From (6), the order of the up and down quark
Yukawa couplings are
 2
Λ1 Λ1 Λ2 Λ1

Λ Λ1 Λ1 
1
 M2 M2 M M M M
ỹ u ∼  Λ1 Λ2 Λ22

Λ2 

, ỹ d ∼  Λ2 Λ2 Λ2 
, (14)
 
 M2 M2 M M M M
Λ1 Λ2
 
1 1 1 1
M M
while the charged lepton Yukawa coupling is

Λ1 Λ2
 
M 1
M
Λ Λ2
 
ỹ l ∼  1 1 . (15)

M M 
Λ Λ2 
1
1
M M

7
By performing the singular value decomposition, we obtain
Λ21 Λ22
Yu ∼ , Yc ∼ , Yt ∼ 1,
M2 M2
Λ1 Λ2
Yd ∼ 2 , Ys ∼ , Yb ∼ 1,
M M
Λ1 Λ2
Ye ∼ , Yµ ∼ , Yτ ∼ 1, (16)
M M
and
 Λ1 Λ1   Λ1 Λ1 
1 1
 M M  Λ2 M
Λ Λ2  Λ Λ2 
   
Uu ∼  1 1 , Ud ∼  1 1 . (17)
M M  Λ2 M
Λ1 Λ2 Λ1 Λ2
   
1 1
M M M M
Then, the CKM matrix is (here O(N1 , N2 ) chooses the maximal among the N1 and N2 ):
   
Λ1 Λ1 Λ1
1 O ,

  M Λ2 M 

 Λ1 Λ 1 Λ 
2
VCKM = (U u )† U d ∼ O , 1 .

 M Λ2 M
Λ2 Λ21
   
 Λ1 
O , 1
M M M Λ2

If we choose
Λ1 Λ2
∼ 3 × 10−3 , ∼ 3 × 10−2 , (18)
M M
then, the singular values are

Yu ∼ 9 × 10−6 , Yc ∼ 9 × 10−4 , Yt ∼ 1,
Yd ∼ 3 × 10−3 , Ys ∼ 0.03, Yb ∼ 1,
−3
Ye ∼ 3 × 10 , Yµ ∼ 0.03, Yτ ∼ 1. (19)

and the CKM matrix is

0.1 3 × 10−3
 
1
VCKM ∼  0.1 1 0.03  . (20)
3 × 10−3 0.03 1

By comparing (13) and (19), we observe

Yu Yc Yt
∼ 1.5 sin β, ∼ 0.5 sin β, ∼ 1 sin β,
Yuobs Ycobs Ytobs
Yd Ys Yb
∼ 300 cos β, ∼ 100 cos β, ∼ 100 cos β,
Ydobs Ysobs Ybobs
Ye Yµ Yτ
∼ 103 cos β, ∼ 50 cos β, ∼ 20 cos β. (21)
Yeobs Yµobs Yτobs

8
Similarly, from (12) and (20), we get the O(1) ratio between our theoretical fit and the experimental
data:
 
1 2 1
|VCKM,IJ | 
obs
∼ 2 1 1 . (22)
|VCKM,IJ | 3 1 1

We observe that the CKM matrix is nicely fitted. As for the fermion masses, at least the hierarchy is
milder than the original SM. If we choose smaller cos β, the hierarchy is getting milder. For example,
cos β ∼ 0.1 (so tan β = 10) corresponds to Ye /Yeobs ∼ 102 , and cos β ∼ 0.02 (so tan β = 40) corresponds
to Ye /Yeobs ∼ 20.3

The neutrino sector is the same as the SM. By adding the right-handed neutrino superfield which is
singlet under GSM , we can write down the neutrino Yukawa coupling and Majorana mass:
N
WN = yIJ LI NJ Hu + MIJ NI NJ . (23)

These terms break continuous U(1)B−L symmetry, but preserves Z4 symmetry by assigning charge 2 to
NI . Adding 3 right-handed neutrinos would change the model from 63 to 66 Weyl fermions.

N hH i. By
After the electroweak symmetry breaking, the neutrino gets the Dirac mass (MD )IJ = yIJ u
integrating out right handed neutrino, the light neutrino mass matrix is

(mν )IJ = (mD )IK (M −1 )KL (mD )LJ . (24)

This is diagonalized as

(mν )IJ = U ν diag(mν1 , mν2 , mν3 )(U ν )T . (25)

e l l
ỹ = U diag(Ye , Yµ , Yτ )V , (26)

Then, the PMNS matrix is defined as

s13 e− i δCP
 
c12 c13 s12 c13
UPMNS = U ν U l† = −s12 c23 − c12 s23 s13 e i δCP
 c12 c23 − s12 s23 s13 e i δCP s23 c13 , (27)
s12 s23 − c12 c23 s13 e i δCP −c12 s23 − s12 c23 s13 e i δCP c23 c13

where sIJ = sin θIJ , cIJ = cos θIJ are the mixing angles, and δCP is the CP phase. The matrix V ` is
constrained by the charge assignment in our model, while the matrix U ` is arbitrary. Therefore, the
experimental data is fitted by choosing U ν appropriately.

5 Summary

We have proposed a new mechanism involving the s-confinement to explain the flavor hierarchy in
the Standard Model. The s-confinement dynamics is used to drive the dangerously irrelevant higher-
scaling-dimensional operators at UV to the marginal Yukawa-Higgs deformation at IR in our model. (In
contrast, the s-confinement is used differently in the symmetric mass generation deformation in [7, 8].)
The smallness of the Yukawa coupling in the flavor hierarchy problem is explained as the ratio of the
3
Roughly speaking, the large tan β corresponds to a smaller supersymmetry breaking scale. Given the bound from LHC,
the value of tan β cannot be much bigger than 50.

9
s-confinement and the cutoff scales. We propose an explicit supersymmetric model to resolve the flavor
hierarchy with the O(1) coefficients at UV. The UV Lagrangian is more consistent with Dirac or ’t Hooft
criteria on the naturalness [19].

Some comments and promising future directions follow. 1. It will be interesting to investigate
both the supersymmetric and non-supersymmetric scenarios of the models similar to ours for the flavor
hierarchy resolution. 2. One can also consider the symmetric mass generation deformation of our 63
or 66 Weyl fermion model. One can also enlarge our model to include also the 27Nf Weyl fermion
model of [7], so to introduce both the symmetric mass generation and Yukawa-Higgs deformations under
the umbrella of a parent theory. One can study the phase transitions between different phases of the
parent theory [20]. 3. The symmetric mass generation in [7, 8] still allows a mean-field fermion bilinear
mass term description at IR, although this mean-field mass does not break GSM because that the GSM
is redefined when the matter contents are enlarged by adding more fields from UV. However, a more
intrinsic symmetric mass generation requires a non-mean-field mass deformation, purely driven by the
multi-field interactions or the disorder mass field configurations [6, 21, 22] — this alternative beyond-
mean-field deformation deserves future study [22]. 4. The large tan β mildens the flavor hierarchy,
but the tan β cannot be too large otherwise it lowers supersymmetry breaking scale that violates the
experimental bound. It will be desirable to sharpen the rough constraint [23] tan β < 50 here to find the
future phenomenology evidence. 5. The s-confinement scenario has the asymptotic freedom at the deep
UV  Λ, while the higher-scaling-dimensional operator and multi-field interaction are nonrenormalizable
becoming nonperturbatively strong also at UV approaching to the cutoff scale M . Closer to the cutoff
M challenges the validity of the effective field theory, it is worthwhile to find, in addition to the discrete
lattice formulation, any alternative continuum UV completion of our model. 6. As a condensed matter
application, the (iso)spin and the electrically charged degrees of freedom can be separated by a larger
energy gap. It will be interesting to know if the smooth confinement analogy works or not to explain
the (iso)spin-charge separation hierarchy.

Acknowledgment

The work of YH was supported by Japan Society for the Promotion of Science (JSPS) Overseas Research
Fellowships. At the final stages of the work, YH was supported by MEXT Leading Initiative for Excel-
lent Young Researchers Grant Number JPMXS0320210099. The work of JW is supported by Harvard
University CMSA. JW thanks the participants of Generalized Global Symmetries workshop (September
19-23, 2022) at Simons Center for Geometry and Physics for the inspiring program. After the manuscript
submission, we are grateful to receive helpful comments from Nathan Seiberg and Matthew Strassler.

References

[1] Particle Data Group, P. A. Zyla et al., Review of Particle Physics, PTEP 2020 (2020), no. 8,
083C01.

[2] J. de Blas, M. Pierini, L. Reina, and L. Silvestrini, Impact of the recent measurements of the top-
quark and W-boson masses on electroweak precision fits, (2022), 2204.04204.

[3] C. D. Froggatt and H. B. Nielsen, Hierarchy of Quark Masses, Cabibbo Angles and CP Violation,
Nucl. Phys. B 147 (1979), 277–298.

10
 [4] N. Seiberg, Exact results on the space of vacua of four-dimensional SUSY gauge theories, Phys.
Rev. D 49 (1994), 6857–6863, hep-th/9402044.

[5] N. Seiberg, Electric - magnetic duality in supersymmetric nonAbelian gauge theories, Nucl. Phys. B
435 (1995), 129–146, hep-th/9411149.

[6] J. Wang and Y.-Z. You, Symmetric Mass Generation, Symmetry 14 (2022), no. 7, 1475, 2204.14271.

[7] S. S. Razamat and D. Tong, Gapped Chiral Fermions, Phys. Rev. X 11 (2021), no. 1, 011063,
2009.05037.

[8] D. Tong, Comments on symmetric mass generation in 2d and 4d, JHEP 07 (2022), 001, 2104.03997.

[9] J. Wang and X.-G. Wen, A Non-Perturbative Definition of the Standard Models, Phys. Rev. Res. 2
(2020), no. 2, 023356, 1809.11171.

[10] I. Garcia-Etxebarria and M. Montero, Dai-Freed anomalies in particle physics, JHEP 08 (2019),
003, 1808.00009.

[11] J. Davighi, B. Gripaios, and N. Lohitsiri, Global anomalies in the Standard Model(s) and Beyond,
JHEP 07 (2020), 232, 1910.11277.

[12] Z. Wan and J. Wang, Beyond Standard Models and Grand Unifications: Anomalies, Topological
Terms, and Dynamical Constraints via Cobordisms, JHEP 07 (2020), 062, 1910.14668.

[13] S. Dimopoulos, S. Raby, and L. Susskind, Light Composite Fermions, Nucl. Phys. B 173 (1980),
208–228.

[14] M. J. Strassler, Generating a fermion mass hierarchy in a composite supersymmetric standard model,
Phys. Lett. B 376 (1996), 119–126, hep-ph/9510342.

[15] A. E. Nelson and M. J. Strassler, A Realistic supersymmetric model with composite quarks, Phys.
Rev. D 56 (1997), 4226–4237, hep-ph/9607362.

[16] E. Farhi and L. Susskind, Technicolor, Phys. Rept. 74 (1981), 277.

[17] S. Antusch and V. Maurer, Running quark and lepton parameters at various scales, JHEP 11 (2013),
115, 1306.6879.

[18] S. P. Martin, A Supersymmetry primer, Adv. Ser. Direct. High Energy Phys. 18 (1998), 1–98,
hep-ph/9709356.

[19] G. ’t Hooft, Naturalness, chiral symmetry, and spontaneous chiral symmetry breaking, NATO Sci.
Ser. B 59 (1980), 135–157.

[20] J. Wang, Z. Wan, and Y.-Z. You, Cobordism and deformation class of the standard model, Phys.
Rev. D 106 (2022), no. 4, L041701, 2112.14765.

[21] J. Wang, CT or P Problem and Symmetric Gapped Fermion Solution, Phys. Rev. D (accepted, in
press) (2022), 2207.14813.

[22] Strong CP problem and Symmetric Mass Solution, to appear.

[23] G. Degrassi, S. Di Vita, J. Elias-Miro, J. R. Espinosa, G. F. Giudice, G. Isidori, and A. Strumia,

Higgs mass and vacuum stability in the Standard Model at NNLO, JHEP 08 (2012), 098, 1205.6497.

11