Study of ATLAS sensitivity to FCNC top quark decays

According to the Standard Model (SM), the top quark Branching Ratios (BR) through Flavour Changing Neutral Currents (FCNC), $t \to qX,X=\gamma, Z$ or q are extremely small (10$^{-14}$ to 10$^{-12}$). Nevertheless there are several extensions to the SM which predict higher values for these FCNC BR (up to ~10$^{-4}$). This work is dedicated to the study of the ATLAS experience sensitivity for the FCNC top quark decays in ttbar events produced at the Large Hadron Collider (LHC) at CERN. Probabilistic analyses were developed for high (L=10 fb$^{-1}$ and L=100 fb$^{-1}$) and low (L=1 fb$^{-1}$) luminosities, using Monte Carlo simulations. Fast and full ATLAS simulations were used. Several sources of systematic uncertainties were studied. With an integrated luminosity of 10 fb$^{-1}$, ATLAS will be able to observe FCNC top quark decays, with a 5sigma significance, if these decays have a BR larger than 9.4x10^${-5}$, 4.4x10$^{-4}$ or 4.3x10$^{-3}$, for the $t \to q\gamma, t \to qZ and t \to qg$ decay channels, respectively. If no signal evidence is found, the following 95% confidence level limits on the BR can be derived: 4.1x10$^{-5}$, 3.1x10$^{-4}$ and 1.3x10$^{-3}$, for the same channels. The expected limits for low luminosity are compatible with these, if the differences in luminosities, in lepton and photon identification efficiencies, in trigger efficiencies and in background cross sections are taken into account. These results were compared with previous analyses from ATLAS, as well from CMS (another general purpose detector at LHC). For the $t \to q\gamma$ channel, the obtained results for the BR are better than the previous ones from ATLAS, and those from CMS. For the $t \to qZ$ channel, the results are similar. The expected limits for these two channels are about two orders of magnitude better than the present experimental limits. For the $t \to qg$ the expected result is about one order of magnitude better then the current limit from Tevatron, although the results from the single top production analyses could be even better. Even if the SM predicts much lower BR than the expected limits, some extensions of the SM can be probed, rejecting them or constraining their parameters space.