Johnny Alejandro Mora Grimaldo ; Hiroaki Aihara

2020
University of Tokyo Tokyo

Abstract: Flavor Changing Neutral Currents (FCNC) constitute excellent probes for physics beyond the Standard Model, since their branching fractions can be affected by the presence of new, heavy particles as mediators. Of special interest are the B → Kl+l− decays—whereBiseitheraB0 oraB+ meson,KiseitheraK+ oraK0,andliseither a μ or an e— given their relatively high branching fraction and smaller theoretical uncertainties (in comparison with other FCNC), and the multiple new couplings they can test. Great efforts have been put into measuring the value of RK, defined as RK = B(B→Kμ+μ−) / B(B → Ke+e−) since it is theoretically very clean, and recent measurements show tension with the Standard Model predicted value of 1. hese decays are extremely rare, and as such, their uncertainties are dominated by the size of the data sample used to study them. New particle accelerators, focused on extreme luminosities, have the potential to narrow these uncertainties and to provide more precise measurements. The Belle II Experiment, which started recording physics events from the summer of 2019, aims for an integrated luminosity of 50 ab−1, by achieving a record of instantaneous luminosity (∼ 8 × 1035cm−2s−1). Since it is operating at the SuperKEK-B, a B meson factory, its reconstruction efficiency for B decays is higher than more general experiments such as the LHC; thus, it represents an ideal setup for a more precise study of the B → Kl+l−. The current work is the first analysis of the B → Kl+l− decays using the Belle II experiment data accumulated in its first physics runs up to March of 2020. It consists of 11.5 fb−1 of collisions recorded at a Center of Mass Energy equal to the Υ(4S) mass; besides the novelty of being the first study of these processes at the Belle II Experiment, this analysis also introduces a new, improved algorithm for Bremsstrahlung recovery, uses Boosted Decision Trees as multivariate classifiers in order to reduce the different background components in the data samples, and applies the technique of boosting to flatness in order to avoid introducing bias through the multivariate classifiers’ outputs. In the study, the ratio RK is measured for the dilepton invariant mass regions corresponding to the J/ψ and ψ(2S) resonances to be B(B→KJ/ψ[e+e−])/B(B → KJ/ψ[μ+μ−]) =0.99±0.09(stat)±0.01(sys), B(B→Kψ(2S)[e+e−])/B(B → Kψ(2S)[μ+μ−]) =1.03±0.41(stat)±0.01(sys). The decay mode B+ → K+e+e− is observed with a significance of 1.35 and a 90% C.L. upper bound on its branching fraction is measured to be B(B+ → K+e+e−) < 3.5 × 10−6. The other three decays are not observed; the 90% C.L. upper bound on their branching fractions are: B(B0 → K0μ+μ−)< 3.2 × 10−6, B(B+ → K+μ+μ−)< 9.9 × 10−7, B(B0 → K0e+e−)< 2.3 × 10−6. All of these are consistent with measurements by previous experiments, and with the predictions made by the Standard Model. We also perform a sensitivity projection for both the branching fractions and the RK ratio at higher luminosities, and show that the improvements in the detector performance are required to meet previous expectations for the RK ratio at Belle II— derived from extrapolations of the Belle measurements—; based on these projections, and pending on the improvements mentioned, we predict that the Belle II experiment will be able to settle the RK anomaly with a 5σ significance at 35 ab−1.

Note: Presented on 27 07 2020
Note: PhD