Studies of ZZγ production in pp collisions at √s = 13 TeV with the ATLAS detector

This thesis presents a study of the feasibility of observing ZZγ triboson production in proton-proton collisions at a centre of mass energy root s = 13 TeV using data collected by the ATLAS detector during Run 2 of the LHC. This process has not yet been studied experimentally because of the small cross-section expected. The full Run 2 data set offers an integrated luminosity of 139/fb. Using a cut-based method to select ZZγ candidate events, an expected signal significance of 1.8σ is achieved. Machine learning processes such as Boosted Decision Trees are used to raise the expected significance to 2.5σ which gives hope of finding evidence of ZZγ triboson production. Background contributions considered include those due to tt ̄γ and W Zγ production, as well as photons from pile-up interactions and from jets misidentified as photons (fake photons). In order to estimate the latter background, usual methods suffer from the low number of events and a novel technique is thus proposed. The contribution from final state radiation (FSR) is also included as part of the estimated background. The FSR process corresponds to the production of a pair of Z bosons where one of the four decay leptons radiates a photon. By treating the FSR process as signal rather than background, it is also possible to study inclusive llllγ production. In this case, the predicted signal significance obtained using the cut-based analysis is 8.0σ . A longer-term goal of the ZZγ study is to search for the presence of anomalous gauge couplings, and this is also discussed here.