Search for the Higgs Boson produced in association with a Vector Boson at ATLAS

On the 4th of July, 2012, the ATLAS and CMS experiments at CERN announced the discovery of a particle consistent with the Standard Model Higgs boson through its decay to bosons. Subsequent measurements have established that the particle’s spin, parity and coupling strengths are in line with Standard Model expectation, making this the strongest evidence of the Standard Model electroweak symmetry breaking mechanism. This thesis concerns the search for Higgs boson decays to two tau leptons, one of the leading Standard Model Higgs decay modes, with particular focus on events in which the Higgs is produced in association with a weak vector boson. Evidence for the existence of such events constitutes a crucial test of the Higgs’ coupling to leptons. The search is conducted using the 20.3 fb−1 of data collected by the ATLAS detector in 2012 from proton-proton collisions in the Large Hadron Collider at a centre-of-mass √ energy of s = 8 TeV. The search is optimised for events in which the associated vector boson decays to an electron or muon and one or both tau leptons from the Higgs decay subsequently decay to hadrons. The dominant background comes from events in which a jet is misidentified as the hadronic decay products of a tau decay, which is estimated from data samples in a signal-free control region using the rate of misidentification. The trigger and identification efficiencies are measured using dedicated calibration data samples with a tag-and-probe technique. No significant excess above the expected background yield is observed in any signal region. The signal strength of best fit, for a Higgs boson mass of 125 GeV, is 2.3 ± 1.6. Instead, limits are placed on the rate at which the event rate might exceed the Standard Model expectation. The event rate is found to be no greater than 5.6 times that expected by the Standard Model at a 95% confidence interval. This observed limit is less than the expected limit of 3.5 times the SM event rate, under the no signal hypothesis, or 3.7 times the SM event rate with signal injected, but is consistent within uncertainties.