[Paper Review] Search for lepton flavor violating decays τ−→ ℓ−π0,ℓ−η,ℓ−η′

Y. Enari, N. Sato+152 more|arXiv (Cornell University)|Sep 1, 2005

Particle physics theoretical and experimental studiesPhysics and Astronomy32 references19 citations

TL;DR

This study searches for lepton flavor violating decays of the tau lepton into a lighter charged lepton and a light pseudoscalar meson (π⁰, η, η′) using 153.8 fb⁻¹ of data from the Belle experiment at KEKB. The analysis sets new, more restrictive upper limits at 90% confidence level, improving previous limits by factors of 10 to 64, with the most stringent bound at B(τ⁻ → μ⁻η) < 1.5×10⁻⁷.

ABSTRACT

We have searched for lepton flavor violating semileptonic τ- decays using a data sample of 153.8 fb-1 accumulated with the Belle detector at the KEKB e+e- collider. For the six decay modes studied, the observed yield is compatible with the estimated background and the following upper limits are set at the 90% confidence level: B(τ-→e-η)&lt;2.3×10-7, B(τ-→μ-η)&lt;1.5×10-7, B(τ-→e-π0)&lt;1.9×10-7, B(τ-→μ-π0)&lt;4.1×10-7, B(τ-→e-η′) &lt;10×10-7, and B(τ-→μ-η′)&lt;4.7×10-7. These results are 10 to 64 times more restrictive than previous limits. © 2005 Elsevier B.V. All rights reserved.

Motivation & Objective

• To search for lepton flavor violating (LFV) semileptonic decays of the tau lepton, which are forbidden in the Standard Model but predicted in some extensions.
• To test the validity of the Standard Model and probe new physics beyond it through rare decays that violate lepton flavor symmetry.
• To improve the sensitivity of existing experimental limits on LFV τ decays using a large, high-precision data sample.
• To constrain models of physics beyond the Standard Model that predict such rare decays, such as those involving flavor-changing neutral currents or heavy sterile neutrinos.

Proposed method

• Utilized 153.8 fb⁻¹ of e⁺e⁻ collision data collected by the Belle detector at the KEKB asymmetric-energy collider.
• Performed a model-independent search for six specific LFV decay modes: τ⁻ → ℓ⁻π⁰, ℓ⁻η, ℓ⁻η′ (ℓ = e, μ).
• Employed kinematic reconstruction and particle identification techniques to isolate signal candidates from background events.
• Used an unbinned maximum-likelihood fit to extract signal yields and set upper limits on branching fractions at 90% confidence level.
• Calibrated background estimates using control samples and Monte Carlo simulations to ensure robust systematic uncertainty control.
• Applied selection criteria optimized for signal sensitivity while minimizing contributions from dominant backgrounds such as τ⁻ → ℓ⁻νν̄ and τ⁻ → ℓ⁻νπ⁰.

Experimental results

Research questions

• RQ1 What are the upper limits on the branching fractions of lepton flavor violating τ⁻ → ℓ⁻π⁰, ℓ⁻η, and ℓ⁻η′ decays?
• RQ2 How do the new experimental limits compare to previous measurements in terms of sensitivity improvement?
• RQ3 Is there any significant signal excess above background in any of the six LFV decay modes studied?
• RQ4 To what extent do the results constrain models of physics beyond the Standard Model that predict LFV decays?

Key findings

• No significant signal was observed in any of the six lepton flavor violating decay modes, consistent with background-only expectations.
• The upper limit on the branching fraction for τ⁻ → e⁻η is B(τ⁻ → e⁻η) < 2.3×10⁻⁷ at 90% confidence level.
• The most stringent limit is set for τ⁻ → μ⁻η, with B(τ⁻ → μ⁻η) < 1.5×10⁻⁷ at 90% confidence level.
• The upper limit for τ⁻ → e⁻π⁰ is B(τ⁻ → e⁻π⁰) < 1.9×10⁻⁷ at 90% confidence level.
• The limit for τ⁻ → μ⁻π⁰ is B(τ⁻ → μ⁻π⁰) < 4.1×10⁻⁷ at 90% confidence level.
• The upper limit for τ⁻ → e⁻η′ is B(τ⁻ → e⁻η′) < 10×10⁻⁷ at 90% confidence level, and for τ⁻ → μ⁻η′ it is B(τ⁻ → μ⁻η′) < 4.7×10⁻⁷ at 90% confidence level.