Search for Charged Higgs Bosons with the ATLAS Detector at the LHC

The discovery of a charged Higgs boson, $H^+$, would be an unambiguous evidence for physics beyond the Standard Model. In this thesis a search for the $H^+$, with the ATLAS experiment at the Large Hadron Collider, LHC, at CERN based on data taken in 2011, are described. A re-analysis of the charged Higgs boson search, utilising the ratio-method, was performed, which greatly enhanced the sensitivity compared to the traditional direct search approach. Light charged Higgs bosons, with a mass lower than the top quark mass, can be produced in top quark decays. Due to the large production cross-section of top quark pairs the light charged Higgs bosons are accessible with early LHC data, in contrast to charged Higgs bosons heavier than the top quark mass. For light charged Higgs bosons the decay via $H^\pm \to \tau^\pm \nu$ is predominant in most theories and scenarios, in the most parts of their phase space. Therefore the $\tau$ identification and the $\tau$ mis-identification probabilities play an important role in the charged Higgs boson search. A tag-and-probe selection of $Z\to ee$ events was developed in order to asses the mis-identification probability of electrons as hadronically decaying $\tau$ leptons, utilising the very first data taken in 2010. The results of this analysis on the one hand provided scale factors crucial for all analyses utilising the electron veto algorithms of the $\tau$ identification. On the other hand a data-driven estimation technique for backgrounds stemming from electrons mis-identified as hadronically decaying $\tau$ leptons, dedicated for the charged Higgs boson search, was developed and successfully implemented, based upon the results of the tag-and-probe mis-identification results. Trigger studies for the charged Higgs boson search aiming at the highest feasible signal efficiencies with the utilised combination of $\tau$ trigger and missing transverse energy trigger, during the 2012 data taking at a centre of mass energy of 8 TeV, were performed. Novel trigger items were designed, thoroughly tested and finally implemented into the trigger menu of the ATLAS experiment for data taking in 2012. A direct charged Higgs boson search in three channels with $\tau$ leptons in the final state, was performed on the full dataset, with an integrated luminosity of 4.6 fb$^{-1}$, taken in 2011 at a centre of mass energy of 7 TeV. No significant excess over the Standard Model expectation was observed in data, thus limits on ${\cal B}( t \to bH^+)$ were set. Finally the channel with a hadronically decaying $\tau$ lepton and an additional electron or muon in the final state from the $t\bar{t}$ decay was re-analysed with the so-called ratio-method. This method measures ratios of event yields instead of evaluating distributions of discriminating variables, thus most of the dominating systematic uncertainties intrinsically cancel to first order. The observed data are found to be in agreement with the Standard Model predictions. Utilising the ratio-method the limits obtained on ${\cal B}( t \to bH^+)$ were significantly improved, compared to the direct search in this channel. When the results of the ratio-method analysis are combined with the results of the direct search for charged Higgs bosons in $t\bar{t}$ decays, with a hadronically decaying $\tau$ lepton and jets in the final state, upper limits on ${\cal B}(t\to bH^+)$ are set in the range of 0.8 % $-$ 3.4 %, for $m_{H^+}$ between 90 GeV and 160 GeV. The limits set on ${\cal B}(t\to bH^+)$ have direct implications on the identity of the Higgs-like particle, with a mass of approximately 126 GeV discovered in 2012 at the LHC, if the Minimal Supersymmetric extension of the Standard Model (MSSM) is realised in nature.