Analysis of inclusive semileptonic <em>B</em> meson decays with τ lepton final states at the Belle experiment

Abstract

In this thesis, the measurement of the inclusive branching fraction of B→Xτν decays at the Belle experiment is presented. Previous measurements of the inclusive semileptonic b-hadron branching fraction BR(b→Xτν) in Z⁰→bb decays at LEP agree well with the prediction of the Standard Model of particle physics. Conversely, the B-factory experiments BaBar, Belle and LHCb consistently measured the exclusive branching fractions of the decays B→Dτν and B→D^*τν larger than the theoretical prediction. The combined exclusive branching fraction measurements together exceed the Standard Model by 4σ and are even larger than the average of inclusive LEP measurements.<br /> This work is the first measurement of the branching fraction of B→Xτν decays at a B-factory. It is therefore a valuable cross check to the exclusive analyses and also probes the inclusive LEP measurements in a different experimental environment.<br /> At Belle, a large sample of 770 million BB pairs was recorded in e⁺e^- collisions at the Υ(4S) resonance. One of the two B mesons in the BB event is fully reconstructed in a hadronic decay mode (B_tag) and the other one (B_sig) is analysed regarding the signal decay. This hadronic B tagging allows one to constrain the kinematics and the flavour of the signal B meson and enables the study of observables in the B_sig rest frame. The signal is selected in the leptonic decay modes of the τ and thus signal-like events are required to have a single electron or muon in the final state.<br /> This analysis relies on Monte Carlo simulations of backgrounds, particularly on those of semileptonic B decays. The latter produce mainly charmed mesons of which ~25% are the poorly known D^** mesons. To improve the modelling of the data, a new model of these excited charmed mesons and the related semileptonic B decays is developed and implemented in dedicated Monte Carlo samples.<br /> Observables for the signal extraction are the lepton momentum and the squared sum of four-momenta of undetected particles (squared missing mass) in the event. The signal branching fraction is measured, normalised to the branching fraction of B→Xlν (l = e, μ) decays, to be R(X) = BR(B→Xτν)/BR(B→Xlν) = 0.298 ± 0.012(stat) ± 0.018(sys). Assuming the isospin-average branching fraction BR(B→Xlν) = (10.86 ± 0.16)%, this translates to a branching fraction of BR(B→Xτν) = (3.23 ± 0.13(stat) ± 0.21(sys))% and is the most precise single measurement of B→Xτν decays so far.