Search for direct pair production of top squarks in $pp$ collisions at $\sqrt{s}= 13$ TeV and $13.6$ TeV in events with two oppositely charged leptons using the ATLAS detector

Abstract

This paper presents the search for direct pair production of top squarks decaying into two on-shell top quarks and two neutralinos in final states with two oppositely charged leptons (electrons or muons), $b$-jets and large missing transverse momentum. The search uses the full Run 2 dataset, corresponding to an integrated luminosity of 140 fb$^{-1}$ of proton-proton collisions at a centre-of-mass energy of $\sqrt{s} = 13$ TeV collected by the ATLAS detector from 2015 to 2018, and the early Run 3 dataset, corresponding to an integrated luminosity of 53 fb$^{-1}$ of proton-proton collisions at $\sqrt{s} = 13.6$ TeV collected in 2022 and 2023. Machine-learning-based classifiers are used to distinguish the signal from Standard Model backgrounds, to enhance signal discrimination across a wide kinematic phase space. No significant excess is observed with respect to the Standard Model prediction and 95% confidence level limits are set on top squark and neutralino mass combinations reaching up to 1060 GeV for the former and 560 GeV for the latter, improving mass limits by about 10% compared to previous ATLAS analyses in the same channel.

Figures (8)

• Figure 1: Diagram of top-squark pair production, representing the two-body $\tilde{t}$ decay into an on-shell top quark and the lightest neutralino ($\tilde{t}\xspace \rightarrow t \hbox{$\tilde{\chi}^0_1$}\xspace$).
• Figure 2: Expected event yield as a function of the NN signal score for the (a) low, (b) medium, and (c) high $\Delta m(\tilde{t}_1\xspace,\hbox{$\tilde{\chi}^0_1$}\xspace)\xspace$ regions. The scores are shown for Run 3 simulated events with two $b$-tagged jets. The distribution in grey represents the expected number of SM background events, the dashed lines show the distributions of three representative signal models: $m(\tilde{t}_1\xspace,\hbox{$\tilde{\chi}^0_1$}\xspace) = (800, 500)~\text{Ge V}\xspace$ for the low-$\Delta m(\tilde{t}_1\xspace,\hbox{$\tilde{\chi}^0_1$}\xspace)\xspace$ category, $m(\tilde{t}_1\xspace,\hbox{$\tilde{\chi}^0_1$}\xspace) = (900, 500)~\text{Ge V}\xspace$ for the medium, and $m(\tilde{t}_1\xspace,\hbox{$\tilde{\chi}^0_1$}\xspace) = (1000, 1)~\text{Ge V}\xspace$ for the high. The expected signal yields are multiplied by 50. CRs (Control Regions), and VRs (Validation Regions) are defined in Section \ref{['sec:background']}.
• Figure 3: Distributions of (a) the $E_{\text{T}}^{\text{miss}}$ in $\mathrm{CR^{Medium,2b}_{\Pqt{}\Paqt}}$ and (b) the $E_{\mathrm {T,\mathrm{corr}}}^{\mathrm{miss}}$ in $\mathrm{CR}_{\Pqt{}\Paqt Z}$, each after the background fit. The contributions from all SM backgrounds are shown as a histogram stack. "Others" includes the contributions from $VVV$, $\Pqt{}\Paqt t$, $\Pqt{}\Paqt\Pqt{}\Paqt$, $\Pqt{}\Paqt W$, $\Pqt{}\Paqt WW$, $\Pqt{}\Paqt WZ$, $\Pqt{}\Paqt H$, $tZ$ and $tWZ$ processes. The hatched bands represent the total combined statistical and systematic uncertainty. The rightmost bin includes overflow events. The bottom panels show the ratio of the observed data to the total SM background prediction, with hatched bands representing the total uncertainty of the background prediction.
• Figure 4: Expected and observed yields in the VRs for the (a) Run 2 and (b) Run 3 datasets. The upper panel shows the observed number of events together with the expected SM backgrounds obtained after the fit. "Others" includes the contributions from $VVV$, $\Pqt{}\Paqt t$, $\Pqt{}\Paqt\Pqt{}\Paqt$, $\Pqt{}\Paqt W$, $\Pqt{}\Paqt WW$, $\Pqt{}\Paqt WZ$, $\Pqt{}\Paqt H$, $tZ$ and $tWZ$ processes. The hatched bands represent the total combined statistical and systematic uncertainty on the SM background. The lower panel shows the significance as defined in Ref. sig-atlas.
• Figure 5: Comparison of the relative uncertainty for the total background yield in each SR for the (a) Run 2 and (b) Run 3 datasets, including the contribution from the different sources of uncertainty, calculated as results of the profile likelihood fit. The category "Other experimental" includes minor detector uncertainties, such as the ones related to leptons, flavour tagging and pile-up modelling. The total uncertainty is calculated considering the correlation across the various systematic components, and therefore it does not necessarily correspond to the sum of the single components.