Search for the rare B0 → K ∗0τ +τ − decay with a hadronic tagged approach at the Belle II experiment

Sumitted to PubDB: 2025-10-24

Category: Phd Thesis

Tags: Run 1

Principal Authors Stefano Moneta, Claudia Cecchi, Elisa Manoni
Date 2024-04-29
Belle II Number BELLE2-PTHESIS-2025-025
Abstract This document describes the search for the rare flavor changing neutral current decay B 0 → K ∗0 τ + τ − at the Belle II experiment. This decay is forbidden at tree level in the Standard Model, occurring only at loop level, and it is expected to be sensible to possible new physics extensions. The analysis exploits the 362.2 fb−1 of data collected in the 2019-2022 period at the center of mass energy of the Υ (4S) by the SuperKEKB electron-positron collider. One of the B mesons, produced in the Υ (4S) → BB decay, is fully reconstructed in hadronic decay modes, while the partner B meson is required to decay in a K ∗ → K + π − and in two τ leptons of opposite signs, reconstructed in their 1-prong modes. A multivariate approach with a binary classifier is adopted to separate signal from background events. The classifier is trained on simulated events with data-driven corrections applied, and used as main tool for the separation between signal and mis-reconstructed events. The signal branching ratio is directly extracted through a binned maximum likelihood fit to the output of the classifier. Based on the reconstructed τ + τ − decay modes (τ → ℓνν, τ → πν, τ → ππ 0 ν, being ℓ an electron or a muon), signal events are categorized into different groups. The background suppression stage is optimized separately for each group, while they are combined to extract the signal branching fraction. Including all the systematic uncertainties, the expected upper limit on the branching ratio is determined to be B(B 0 → K ∗0 τ + τ − ) < 2.7 × 10−3 at 90% confidence level. This expectation is already competitive with the current world-best upper limit B(B 0 → K ∗0 τ + τ − ) < 3.1 × 10−3 , at 90% confidence level, set by the Belle collaboration with about a factor ×2 larger dataset of 711 fb−1 .

Files