Global Lepton Flavour Violating Constraints on New Physics
Enrique Fernández-Martínez, Xabier Marcano, Daniel Naredo-Tuero
TL;DR
This study performs a comprehensive global analysis of charged lepton flavour violation (cLFV) within effective field theory, contrasting low-energy LEFT and high-energy SMEFT frameworks to constrain the full operator space beyond one-operator-at-a-time approaches. It shows that fully leptonic four-fermion operators have no flat directions in LEFT, allowing robust global bounds, while semileptonic operators exhibit flat directions that are only partially lifted by SMEFT matching and are highly sensitive to nuclear uncertainties in μ–e conversion. The authors map operator constraints across τ–ℓ and μ–e sectors, quantify flat directions, and provide correlation matrices to facilitate UV completion studies, including a scenario with only first-generation quarks where near-complete global constraints emerge. The results highlight the pivotal role of nuclear overlap uncertainties and suggest additional observables (e.g., τ decays to KK) to close remaining degeneracies, offering practical tools for interpreting cLFV bounds in UV models.
Abstract
We perform a global analysis of the bounds from charged lepton flavour violating observables to new physics. We parametrize generic new physics through the Effective Field Theory formalism and perform global fits beyond the common one-operator-at-a-time analyses to investigate how much present data is able to constrain the full parameter space. We particularly focus on leptonic and semileptonic operators with light quarks, identifying unbounded flat directions, detailing how many are present and which operators are involved. The analysis is performed in the general LEFT formalism, which contains all possible low-energy effective operators relevant for lepton flavour violation, as well as in more restricted scenarios, when operators come from a SMEFT completion. We find that flat directions play no role in the fully leptonic four-fermion operators. Conversely, they significantly hinder the ability to derive global bounds on semileptonic operators, with several flat or at least very poorly constrained directions preventing to fully constrain the parameter space. These results are particularly affected by the proper inclusion of uncertainties in the parameters describing $μ-e$ conversion, which decrease the number of well-constrained directions in operator space when treated as nuisance parameters in the fit. While present data is able to provide global constraints on all operators only in the more restricted scenarios we investigated, very strong correlations among the parameters must exist to avoid conflict with the different observables. We provide correlation matrices approximating our full results as a useful tool to compare present data with particular UV completions.
