Reconstruction of and search for charged Higgs bosons decaying to top and bottom quarks

  • Emily Hanson

Student thesis: Phd


A search for heavy charged Higgs bosons produced in association with a top and bottom quark using 36.1 fb$^{-1}$ of proton-proton data collected by the ATLAS experiment is presented. As the mass of the charged Higgs boson is not predicted by any model, a range of masses above the top quark mass are tested. A boosted decision tree (BDT) is used to distinguish between signal and background in both the lepton+jets and the dilepton channels. No significant excesses above the background are observed. As such, exclusion limits are set on the production cross-section times branching fraction for the combination of the lepton+jets and dilepton channels. The limits are interpreted under two benchmark models predicting charged Higgs bosons to exclude masses between 200 and 965 (540) GeV for low (high) values of the model parameter $\tan\beta$. This result is dominated by the lepton+jets channel. A method to improve the sensitivity to search for charged Higgs bosons in the dilepton channel is also presented. The method aims to fully reconstruct events in the $t\bar{t}b\bar{b}$ final state with both top quarks decaying via $t\to b\ell\nu$. The reconstruction is complicated by high b-jet multiplicities and unobserved neutrinos. The first step is a reconstruction BDT to determine the origin of b-jets. The second step uses neutrino weighting, which has not been used previously in a $t\bar{t}b\bar{b}$ final state, to compute neutrino momenta. After event reconstruction, a second BDT is used to classify signal and background. The expected sensitivity is interpreted under the Maximally-Symmetric 2-Higgs-Doublet Model (2HDM) in the Standard Model (SM) alignment limit. At an integrated luminosity of 150 fb$^{-1}$, the dilepton channel could be used to improve the exclusion from ATLAS by up to 19%.
Date of Award1 Aug 2020
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorYvonne Peters (Supervisor) & Alexander Oh (Supervisor)

Cite this