Search for the Standard Model Higgs boson in the dimuon decay channel with the ATLAS detector

The search for the Standard Model Higgs boson was one of the key motivations to build the world’s largest particle physics experiment to date, the Large Hadron Collider (LHC). This thesis is equally driven by this search, and it investigates the direct muonic decay of the Higgs boson. The decay into muons has several advantages: it provides a very clear final state with two muons of opposite charge, which can easily be detected. In addition, the muonic final state has an excellent mass resolution, such that an observed resonance can be pinned down in one of its key properties: its mass. Unfortunately, the decay of a Standard Model Higgs boson into a pair of muons is very rare, only two out of 10000 Higgs bosons are predicted to exhibit this decay. 2 On top of that, the non-resonant Standard Model background arising from the $Z/\gamma^*\rightarrow\mu\mu$ process has a very similar signature, while possessing a much higher cross-section. For one produced Higgs boson, there are approximately 1.5 million Z bosons produced at the LHC for a centre-of-mass energy of 8 TeV. Two related analyses are presented in this thesis: the investigation of 20.7 $fb^{−1}$ of the proton-proton collision dataset recorded by the ATLAS detector in 2012, referred to as standalone analysis, and the combined analysis as the search in the full run-I dataset consisting of proton-proton collision data recorded in 2011 and 2012, which corresponds to an integrated luminosity of 24.8 $fb^{−1}$ . In each case, the dimuon invariant mass spectrum is examined for a narrow resonance on top of the continuous background distribution. The dimuon phenomenology and ATLAS detector performance serve as the foundations to develop analytical models describing the spectra. Using these analytical parametrisations for the signal and background mass distributions, the sensitivity of the analyses to systematic uncertain- ties due to Monte-Carlo simulation mismodeling are minimised. These residual systematic uncertainties are addressed in a unique way as signal acceptance uncertainties. In addition, a new approach to assess the systematic uncertainty associated with the choice of the background model is designed for the com- bined analysis. For the first time, the spurious signal technique is performed on generator-level simulated background samples, which allows for a precise determination of the background fit bias. No statistically significant excess in the dimuon invariant mass spectrum is observed in either ana- lysis, and upper limits are set on the signal strength $\mu = \sigma/\sigma_{SM}$ as a function of the Higgs boson mass. Signal strengths of $\mu\geq$ 10.13 and $\mu\geq$ 7.05 are excluded for a Higgs boson mass of 125.5 GeV with a confidence level of 95% by the standalone and combined analysis, respectively. In the light of the discovery of a particle consistent with the predictions for a Standard Model Higgs boson with a mass of m H = 125.5 GeV, the search results are reinterpreted for this special case, setting upper limits on the Higgs boson branching ratio of BR($H\rightarrow\mu\mu$) $\geq 1.3 \times 10^{−3}$ , and on the muon Yukawa coupling of $\lambda_{\mu} \leq 1.6 \times 10^{−3}$ , both with a confidence level of 95 %.