All data collected by the ATLAS detector undergo careful scrutiny to ensure the quality of the recorded information; data used for the studies in this chapter and in Chapter 5 are filtered by requiring that all detector subsystems needed in the analysis (calorimeters and tracking detectors) are operating nominally. After all data-quality requirements, 37.1 fb−1 ofppcollision data from the 2015–2016 dataset are available for analysis. Some results shown are based on only a fraction of this total dataset (e.g. only 2015 data, only 2016 data, or only a portion of one of these datasets), and thus the integrated luminosity quoted for some results is less than 37.1 fb−1.
Samples of simulatedZ →eeand J/ψ→eedecays as well as single electron samples are used to benchmark the expected prompt electron efficiencies and to define the electron identification criteria. The Z → ee Monte Carlo (MC) samples were generated with the Powheg-Box v2 MC
4. Electron Reconstruction, Identification, and Isolation 32
program [84–88] interfaced to the PYTHIA v8.186 [89] parton shower model. The CT10 parton distribution function (PDF) set [90] was used in the event generation with the matrix element, and the AZNLO [91] set of generator-parameter values (tune) with the CTEQ6L1 [92] PDF set were used for the modeling of non-perturbative effects. The J/ψ → ee samples were generated with
PYTHIAv8.186; the A14 set of tuned parameters [93] was used together with the CTEQ6L1 PDF set for both event generation and for the parton shower. The simulated single electron samples were produced with a flat distribution inη as well as inpTin the region 3.5 GeV to 100 GeV, followed by a linear ramp down to 300 GeV, and then a flat distribution again to 3 TeV. For studies of electrons in simulated event samples, the reconstructed electron track is required to have hits in the Inner Detector which originate from the true electron during simulation.
Backgrounds that may mimic the signature of prompt electrons were simulated with two-to-two processes in the PYTHIA v8.186 event generator using the A14 set of tuned parameters and the NNPDF2.3LO PDF set [94]. These processes include multijet production, qg →qγ, qq¯→gγ, W
andZ production (as well as other electroweak processes), and top quark production. This sample of simulated background events is sometimes referred to as a “dijet” sample, as that is the process with the largest cross-section (though it clearly is not the only contributor). A filter was applied to the simulation to enrich the final sample in electron backgrounds. This filter retains events in which particles produced in the hard scatter (excluding muons and neutrinos) have a summed energy that exceeds 17 GeV in an area of ∆η×∆φ= 0.1×0.1, which mimics the highly localized energy deposits that are characteristic of electrons. When using this background sample, prompt electrons fromW
andZ decays are excluded using generator-level simulation information.
Multiple overlaidppcollisions were simulated with the soft QCD processes ofPYTHIAv8.186 using the MSTW2008LO PDF set [95]. The Monte Carlo events were reweighted so that the distribution of the average number of interactions per crossing hµimatches the one observed in the data. All simulated samples were processed with theGeant-based simulation [96,97] of the ATLAS detector.
4.2.1
Selection of Prompt Electrons and Background Objects in Data
To obtain a relatively pure sample of prompt electrons in data, the “tag-and-probe” method is used. This method relies on the decay products of known resonances such as theZ, W, or J/ψ, and is used to select unbiased electron candidates (“probes”) by applying strict requirements on the other object in the decay (“tags”). Z →eeevents are used for the electron identification optimization for electronpT>15 GeV, whileJ/ψ →eeevents are used for electron pT<15 GeV. More details on
4. Electron Reconstruction, Identification, and Isolation 33
the triggers and selection criteria used to collect these events can be found in Sections 5.2 and 5.3. To obtain a sample of objects which serve as backgrounds to electrons, prescaled single electron triggers are used. These triggers require a reconstructed electron candidate to pass a given pT threshold (ranging from 5 GeV to 120 GeV in steps of 5–20 GeV), but do not apply any identification requirement at the HLT. Note that hadronic jets are frequently reconstructed as electrons due to the requirements imposed on the tracking and calorimetric quantities (as will be discussed in Section 4.3). Consequently, this sample primarily contains dijet events, as this is the process with the largest production cross-section at the LHC. However, this is a very inclusive selection, and includes other processes as well. To reduce potential contamination from electroweak processes containing prompt electrons, several requirements are imposed:
• IfETmiss>25 GeV, veto the event (to reduce contamination fromW →eν).
• IfmT>40 GeV, veto the event (to reduce contamination fromW →eν). mTis the transverse mass computed between the reconstructed electron and the pmiss
T , and serves as a proxy for theW boson mass; see Equation 7.1 for more detail.
• If a second electron (which satisfies the Medium identification operating point and has pT> 4 GeV) is present in the event and forms an invariant mass within 70 GeV< mee<110 GeV,
veto the event (to reduce contamination fromZ →ee). Note that no electric charge require- ments are imposed on these electrons, to suppress the (admittedly small) contamination from
Z →eeelectrons with an incorrect reconstructed charge.
This selection has been used to collect electron backgrounds for the optimization of the electron identification since the start of the 2017 data-taking period; before that, simulation was used to model these backgrounds.