Catalog · Top / Higgs / EW · T018

T018 $\mathcal{B}(h\to\mu\tau)$

Lepton-flavor-violating Higgs decay h $ \to \mu \tau$

Status REVIEWED VERIFIED High Code: NO Priority Medium

PDG / equivalent values

Observable	Value	Year	Experiment / source	Provenance
B(h $ \to \mu \tau$), CMS full Run-2 direct search	< 0.15% branching fraction	2025	PDG2025:T018:cms_run2	source ↑
B(h $ \to \mu \tau$), ATLAS full Run-2 direct search	< 0.18% branching fraction	2025	PDG2025:T018:atlas_run2	source ↑
B(h $ \to \mu \tau$), CMS full Run-2 direct search	< 0.15% branching fraction	2021	CMS2021:T018:mutau_limit	source ↑
CMS dataset for h $ \to \mu \tau$ direct search	137 fb^-1 at sqrt(s) = 13 TeV integrated luminosity and collision energy	2021	CMS2021:T018:dataset	source ↑
B(h $ \to \mu \tau$), ATLAS full Run-2 direct search	< 0.18% branching fraction	2023	ATLAS2023:T018:mutau_limit	source ↑
ATLAS dataset for h $ \to \mu \tau$ direct search	138 fb^-1 at sqrt(s) = 13 TeV integrated luminosity and collision energy	2023	ATLAS2023:T018:dataset	source ↑
ATLAS branching-fraction difference B(H $ \to \mu \tau$) - B(H $ \to $ e $\tau$)	0.25% $branching-fraction difference B(H \to \mu \tau) - B(H \to e \tau)$	2023	ATLAS2023:T018:symmetry_difference	source ↑
CMS 8 TeV local excess and best-fit B(H $ \to \mu \tau$)	2.4 sigma local excess significance and best-fit branching fraction	2015	CMS2015:T018:hint	source ↑
B(H $ \to \mu \tau$), CMS 8 TeV direct-search limit	< 1.51% branching fraction	2015	CMS2015:T018:limit	source ↑
CMS 8 TeV dataset for Higgs LFV search	19.7 fb^-1 at sqrt(s) = 8 TeV integrated luminosity and collision energy	2015	CMS2015:T018:dataset	source ↑
CMS 2015 $\mu^-\tau$ Yukawa-coupling norm interpretation	< 3.6e-3 $\mu^-\tau Yukawa-coupling norm from CMS 2015 interpretation$	2015	CMS2015:T018:yukawa_limit	source ↑
Generic LFV-Higgs EFT allowance for h $ \to \tau \mu$ or h $ \to \tau$ e	order 10% $possible h \to \tau \mu or h \to \tau e branching ratios allowed by low-energy constraints \,\text{in}\, generic LFV-Higgs EFT$	2012	HarnikKoppZupan2012:order10	source ↑
Generic RS KK-gluon lower scale from 2008 warped-flavor abstract	about 21 TeV generic RS KK-gluon lower scale from 2008 warped-flavor abstract	2008	CsakiFalkowskiWeiler2008:kk_gluon_rs	source ↑

Why this constrains the RS scan

In warped or partially composite lepton sectors, fermion localization explains charged-lepton masses, but misalignment between the mass matrix and physical Higgs couplings can generate off-diagonal Yukawa interactions. The $h\to\mu\tau$ limit directly tests the $\mu$-$\tau$ charged-lepton Yukawa entries and is complementary to dipole LFV constraints such as $\mu\to e\gamma$. It is therefore relevant only to a lepton-sector extension of the current quark scan, not to the existing $\Delta F=2$ quark lane.

What's changed since the original paper

The requested baseline is arXiv:0804.1954, before the LHC Higgs LFV program (CsakiFalkowskiWeiler2008WarpedFlavor). Harnik, Kopp, and Zupan later emphasized that flavor-violating Higgs decays could be a direct discovery channel and that $h\to\tau\mu$ searches could beat rare-tau bounds (HarnikKoppZupan2012:order10). CMS then reported the historical $8~\mathrm{TeV}$ hint: a $2.4\sigma$ excess with $p=0.010$, best fit $\mathcal{B}(H\to\mu\tau)=(0.84^{+0.39}_{-0.37})\%$, and a 95\% CL limit of $1.51\%$ using $19.7~\mathrm{fb}^{-1}$ (CMS2015:T018:hint; CMS2015:T018:dataset; CMS2015:T018:limit). The full Run-2 CMS and ATLAS analyses did not confirm a significant signal and pushed the direct limits to the $10^{-3}$-branching-fraction level (CMS2021:T018:mutau\_limit; ATLAS2023:T018:mutau\_limit).

Validity and model dependence

Experimentally this is a direct, clean upper limit on a forbidden SM decay. The interpretation assumes the signal production mixture and total Higgs width used by each collaboration; a model that changes Higgs production or the total width needs a recast rather than a literal branching-ratio cut. Low-energy LFV bounds constrain related Yukawa products, but the direct $h\to\mu\tau$ search is the least model-dependent way to constrain this specific final state.

Code coverage in this repo

Coverage is NO. The required greps and a targeted search for $h \to \mu \tau$ , $H \to \mu \tau$ , mutau, and Y\_mu\_tau across quarkConstraints/, qcd/, flavorConstraints/, neutrinos/, yukawa/, warpConfig/, solvers/, scanParams/, and tests/ found no observable implementation. The existing LFV backend is specifically $\mu\to e\gamma$ (flavorConstraints/muToEGamma.py:1; flavorConstraints/muToEGamma.py:75), and the scanner applies that filter at scanParams/scan.py:523. Charged-lepton Yukawa helpers are diagonal mass-relation utilities, not an $h\to\mu\tau$ prediction.

Linked evidence (opens GitHub blob at flavor-catalog-website/2026q2):

Implementation difficulty

Production integration is HIGH. The external limit is easy to store, but a faithful RS constraint needs a new charged-lepton Higgs-Yukawa misalignment calculation, a convention for $\mu\tau$ versus $\tau\mu$ couplings, and a Higgs-rate/width interpretation before scan points can be accepted or rejected.

Reason: The direct limit is simple to catalog, but production-grade RS integration needs a new charged-lepton Higgs-Yukawa misalignment calculation, a convention for Y_mu_tau/Y_tau_mu, and a Higgs production/total-width interpretation. The current code only has diagonal charged-lepton Yukawas and a $\mu \to e \gamma$ dipole filter.

Key references

PDG2025HiggsLFVReview; CMS2021HiggsLFVMuTau; ATLAS2023HiggsLFVMuTau; CMS2015HiggsLFVMuTauHint; HarnikKoppZupan2012FlavorViolatingHiggs; CsakiFalkowskiWeiler2008WarpedFlavor; PerezRandall2008WarpedLeptonFlavor.

value < 0.15%

RESOLVED

Match snapshot line 21

L18: limits on LFV Higgs branching ratios. For h -> mu tau, the observed
L19: (expected) limits are:
L20:   B(h -> mu tau) < 0.18% (0.09%) from ATLAS [439].
L21:   B(h -> mu tau) < 0.15% (0.15%) from CMS [440].
L22: 
L23: The same paragraph states that direct ATLAS/CMS limits on Br(h -> mu tau) and
L24: Br(h -> e tau) are more than an order of magnitude stronger than limits

value < 0.18%

RESOLVED

Match snapshot line 20

L17: The full 13 TeV Run-2 datasets collected by ATLAS and CMS set 95% CL upper
L18: limits on LFV Higgs branching ratios. For h -> mu tau, the observed
L19: (expected) limits are:
L20:   B(h -> mu tau) < 0.18% (0.09%) from ATLAS [439].
L21:   B(h -> mu tau) < 0.15% (0.15%) from CMS [440].
L22: 
L23: The same paragraph states that direct ATLAS/CMS limits on Br(h -> mu tau) and

value < 0.15%

RESOLVED

Match snapshot line 26

L23: violating branching fractions of the Higgs boson.
L24: 
L25: The observed (expected) upper limits on the branching fractions are:
L26:   B(H -> mu tau) < 0.15% (0.15%) at 95% confidence level.
L27:   B(H -> e tau) < 0.22% (0.16%) at 95% confidence level.

value 137 fb^-1 at sqrt(s) = 13 TeV

RESOLVED

Match snapshot line 20

L17: 
L18: A search is presented for lepton-flavor violating decays of the Higgs boson
L19: to mu tau and e tau. The data set corresponds to an integrated luminosity of
L20: 137 fb^-1 collected at the LHC in proton-proton collisions at a
L21: center-of-mass energy of 13 TeV. No significant excess has been found, and
L22: the results are interpreted in terms of upper limits on lepton-flavor
L23: violating branching fractions of the Higgs boson.

value < 0.18%

RESOLVED

Match snapshot line 27

L24: The observed (expected) 95% CL upper limits obtained with the MC-template
L25: method from a simultaneous measurement are:
L26:   B(H -> e tau) < 0.20% (0.12%).
L27:   B(H -> mu tau) < 0.18% (0.09%).
L28: 
L29: The best-fit branching-ratio difference measured with the Symmetry method in
L30: the channel where the tau lepton decays to leptons is:

value 138 fb^-1 at sqrt(s) = 13 TeV

RESOLVED

Match snapshot line 20

L17: 
L18: The searches use proton-proton collision data collected with the ATLAS
L19: detector at sqrt(s) = 13 TeV, corresponding to an integrated luminosity of
L20: 138 fb^-1. Leptonic and hadronic tau decays are considered. No significant
L21: excess of events is observed, and the results are interpreted as upper limits
L22: on lepton-flavour-violating branching ratios of the Higgs boson.
L23:

value < 1.51%

RESOLVED

Match snapshot line 29

L26:   B(H -> mu tau) = (0.84 +0.39 -0.37)%.
L27: 
L28: The 95% confidence-level upper limit is:
L29:   B(H -> mu tau) < 1.51%.
L30: 
L31: The same result constrains the mu-tau Yukawa couplings to be less than
L32: 3.6e-3.

value 19.7 fb^-1 at sqrt(s) = 8 TeV

RESOLVED

Match snapshot line 20

L17: The first direct search for lepton-flavour-violating decays of the recently
L18: discovered Higgs boson is performed in the H -> mu tau_e and H -> mu tau_h
L19: channels. The data sample was collected in pp collisions at sqrt(s) = 8 TeV
L20: with CMS and corresponds to 19.7 fb^-1.
L21: 
L22: The sensitivity of the search is an order of magnitude better than the
L23: existing indirect limits. A slight excess of signal events with a significance

value order 10%

RESOLVED

Match snapshot line 21

L18: The authors rederive constraints from rare decays, electric and magnetic
L19: dipole moments, and meson oscillations. They find that flavor-violating Higgs
L20: decays to leptons can be sizeable; h -> tau mu and h -> tau e branching
L21: ratios of order 10% can be allowed by low-energy constraints. They also find
L22: that LHC limits from tau-tau recasts were already stronger than rare-tau
L23: decay bounds and motivate dedicated h -> tau mu and h -> tau e searches.

value about 21 TeV

RESOLVED

Match snapshot line 19

L16: dual description of a composite pseudo-Goldstone Higgs. They re-examine
L17: flavor constraints on the mass scale of new physics in standard
L18: Randall-Sundrum-type scenarios and find that the KK gluon mass should
L19: generically be heavier than about 21 TeV. In the composite-Higgs models
L20: considered, new flavor-violation contributions and gauge-sized Yukawa
L21: couplings imply a KK gluon mass of at least about 33 TeV, with the conclusion
L22: that fully anarchic warped flavor is implausible without additional flavor