Measurement of the top quark pair production cross section in proton-proton collisions at center-of-mass energies of 7 TeV in final states with a $\tau$ lepton with the ATLAS detector

The top quark is the heaviest elementary particle and has the largest coupling to the Standard Model Higgs boson. Top quark decays with a tau lepton in the final state play an important role in the search for SM and BSM Higgs bosons. In particular, if a charged Higgs boson exists as predicted by the MSSM, and its mass is lower than the top quark mass minus the bottom quark mass, the top quark predominantly decays into a charged Higgs boson and a $b$-quark: $t \rightarrow H^{+}b$. A heavy charged Higgs can also be produced associated with a top quark: tH$^{+}$. In some scenarios, the charged Higgs would decay predominantly to a tau lepton and a neutrino, producing then an excess in the $\ell + \tau$ channel over the other dilepton channels which, if observed, would constitute experimental evidence of the the existence of a charged Higgs boson. The $t\bar{t}$ production cross section in the lepton plus tau channel ($pp \rightarrow t\bar{t} \rightarrow\ell + \tau$) measurement presented in this thesis has been obtained analysing 2.05 $fb^{-1}$ of ATLAS data at $\sqrt{s} =$ 7 TeV collected during 2011. The $t\bar{t}(\ell+\tau)$ candidate events are selected based on the event topology of the final state: one high $p_T$ electron or muon, 2 high $E_T$ jets, at least one of them b-tagged, large missing transverse energy and one hadronically decaying $\tau$ lepton. The main background after the selection comes from semileptonic $t\bar{t}$ where a jet has been misidentified as a hadronically decaying tau. The only discriminator between this background and the signal is the tau identification. In this analysis we exploit a discriminant variable, $BDT_j$, obtained from a Boosted Decision Tree multi-variate technique specially trained to separate real taus from other jet types. The signal and background contributions to the signal region are then estimated using a $\chi^{2}$ fit of the $BDT_j$ distribution of the reconstructed taus in the data sample with a signal template derived from MC and background templates derived from data to minimize the systematic uncertainty. Two background templates are used, one for jets produced by the hadronization of light-quarks and another for jets produced by the hadronization of gluon, which constitute the main source of fake taus for the $t\bar{t}(\ell+\tau)$ signal after the selections. The parameters of the fit are the amount of both backgrounds and the amount of signal and the shapes of the templates are fixed. The measured $t\bar{t}(\ell+\tau)$ cross section \begin{equation*} \sigma_{t\bar{t}} = 151 \pm 11 ~{\rm (stat.)} \pm 15 ~{\rm (syst.)} \pm 5 ~{\rm (lumi.)} ~\rm{pb.} \end{equation*} exceeds the precision of the existing measurements in the same $\ell + \tau$ channel performed by CDF, DZero and CMS. Furthermore, it improves the results of previous measurements of the $t\bar{t}(\ell+\tau)$ cross section published at ATLAS, reducing the statistical uncertainty of the cross section measurement and improving the background description, building a background model which properly reflects the jet composition to correctly estimate the fake $\tau$ contamination in the signal region. The systematic uncertainties dominate the measurement and are kept at the same level as previous measurements. The observed cross section value agree well with the Standard Model theoretical expectation predicted with full NNLO precision, within uncertainties, and is consistent with the measured cross-sections using different decay channels by ATLAS. Therefore, we cannot make any claims of new physics. We have, however, developed a novel technique which has a great potential to achieve even higher precision when applied to high statistic samples and it can be used to test the SM and search for new physics in pp collisions in the $t\bar{t}(\ell+\tau)$ channel with the new data expected in 2015 when the LHC resumes the proton-proton collisions at 13-14 TeV.