The inner detector
The inner detector (ID) is responsible for measuring the tracks of the charged particles. A solenoid magnet system covers the ID and assures a 2 T magnetic field. From the curvature of the tracks the charge and the transverse momenta pT is determined in a range covering || < 2.5. Structurally, the ID is composed from three sub-detectors: the pixel detector, the semiconductor tracker and the transition radiation tracker. Figure 2.6 (left) shows a sketch of the ID structure.
The material budget in terms of radiation lengths is shown by component in figure 2.6 (right). This information is particularly important for electrons and photons which can interact with the material and loose energy through bremsstrahlung photons and e−e+ pair creation respectively. A detailed layout of the ID is given in figure 2.7.
The semiconductor tracker (SCT) is composed of 4 layers (2112 modules) in the barrel and 2 layers in the end-caps of 9 disks each (1976 modules). Detailed view of the SCT structure with sizes and positioning is given in figure 2.7. Each layer has up to 4 sensors separated by an angle of 40 μrad angle to obtain a three dimensional information (in z for the barrel and radial position for the endcap). The achieved resolution in the SCT is 17 μm in the transversal direction and 580 μm in the longitudinal direction.
Transition radiation tracker
The transition radiation tracker (TRT) is the outermost layer of the ID and is composed of 350848 polyimide drift tubes (straws) of 4 mm diameter with a gold-plated tungsten wire (anode) in the center of the tubes (cathode). The detector functions based on the transition radiation emitted when a charged particle passes through an inhomogeneous media (ex. boundary of two different media). The straws are filled with a mixture of gas based mainly on xenon (Xe 76%, CO2 27% and O2 3%) which is ionized by the transition radiation photons. The positioning and the overall size of the TRT is shown in figure 2.7. The overall spatial resolution in the TRT is around 130 μm (in r direction) mainly thanks to the high number of points per track.
The calorimeter system
ATLAS’s calorimeter system is composed of the electromagnetic (EM) calorimeter and the hadronic calorimeter. Figure 2.8 show a schematic view of the ATLAS calorimeter system. The calorimeters allow to measure the energy of the electrons, photons and hadrons. The EM calorimeter is designed to measure the energy of the particles that interact primarily electromagnetically, the electrons and the photons, which develop showers that are fully contained. The hadronic calorimeter relies on the strong nuclear force as the interaction mechanism between the detector and the particles and as the name suggests it targets the hadrons.
The electromagnetic calorimeter is a Liquid Argon (LAr) sampling calorimeter with the liquid argon as the active medium and lead (Pb) is used as absorber. The EM calorimeter is composed of three main parts:
— LAr electromagnetic barrel (EMB), covering ||<1.475. The EMB is formed from 2 half-barrel (-1.475 << 0 and 0 <<1.475) each containing 16 modules (one module covers = 22.5o) and presents no discontinuities in azimuth angle.
— LAr electromagnetic end-caps (EMEC), two on each side of the EMB covering 1.375<||<3.2.
— LAr forward calorimeter, covering the 3.1<||<4.9 region The central part of the detector (||<2.5) allows for precision measurements of electron and photon energies and position using more than 170000 cells. It is composed of three layers with different granularities with an additional layer, the pre-sampler (||<1.8), used to recover for energy loss in front of the calorimeter. layer. Full details on the granularity of the barrel and end-caps are given in table 2.2. Layer 1 features cells with a fine granularity × = 0.025/8 × 0.1 in to allow to distinguish isolated single photons from collimated two closeby photons from 0 meson decays. This factor is crucial for analyses involving photons in the final state like H ! . The second layer is the largest one and encompasses most of the energy of the electromagnetic shower. A coarse granularity of × = 0.1 × 0.1 (4×4 cells in layer 2) is used for the trigger system for fast decision making on storing or not the event at the first level trigger.
Track and vertex reconstruction
The charged-particle trajectories (tracks) are bent by the magnetic field in the ID and get an helical geometry. The tracks are reconstructed using the spatial information from the ID. In figure 2.13 is shown the parametrization of the tracks based on the cylindrical geometry of the detector and the particle’s properties.
The tracks are defined by 5 parameters. The distance of the closest point of the track to the beam axis, d0, is referred to as the transverse impact parameter. The longitudinal impact parameter, z0, defines the distance from the transverse plane (x-y plane) containing the closest point of the track to the beam axis to the primary vertex or origin of the coordinate system (in case no primary vertex
is defined yet). The orientation of the track is given by the azimuth angle and the polar angle . These angles are defined based on the momentum vector at the closest point. The curvature of the track is parametrized by the ratio of the charge and transverse momentum Q/pT .
The track reconstruction, i.e. the determination of the 5 track parameters, is based on a staged pattern-recognition approach . The reconstruction starts with an iterative track-finding algorithm based on track seeds that requires a set of three space-points from the pixel detector and first layer of SCT. Selected track candidates are extended to the full SCT and a special procedure is put in place to solve ambiguities by using track scores (based on track quality) and neural networks . Global 2 fit  information, Kalman filter classification  and minimum requirements like pT> 400 MeV, || < 2.5 and other selections based on the number of pixels and SCT clusters are used to reduce fake tracks. Further on, the tracks are extended to the TRT and a high-resolution fit is performed using full information from the ID to complete the track reconstruction.
Once the tracks are reconstructed, primary vertices are obtained by extrapolation of the tracks to the same point on the beam axis. Multiple primary vertices can be reconstructed. The actual primary vertex in an event is considered the one with the highest P i p2 T,i, with pT,i > 400 MeV, while the others are considered pileup vertices. Secondary vertices are displaced from the beam axis and arise from heavy flavor decays (jets from b quarks have a displaced vertex with respect to the primary vertex), photon conversion or interaction with the material of the detector.
Electrons and photons
The electron and photon deposit their energy and are stopped in the EM calorimeter. Unlike the photons, the electrons have an associated track in the ID. Therefore, a reconstructed electron candidate is an energy cluster in the EM calorimeter matched with a track in the ID. Figure 2.14 shows a sketch of the sub-detectors involved in the electron and photon reconstruction, both sharing a similar procedure.
Calorimeter Cell Energy Reconstruction
Energy « seeds » are searched in the EM calorimeter in a × space of 3 x 5 towers using a « sliding window » algorithm . The coverage of one tower is × = 0.025 x 0.0245 and corresponds to the granularity of the second layer of the LAr calorimeter (see figure 2.9). The energy of one tower is obtained by energy summation of all LAr layers for a given tower. A total of 200 x 256 towers cover the × space. Increments of 1 tower in or directions allows to scan the entire × space and localize energy deposits. A seed cluster is found when its energy reaches at least 2.5 GeV. If two seed clusters overlap within a 5×9 towers region, the highest transverse energy seed cluster is retained. A high efficiency of around 95% is obtained at ET = 7 GeV and at least 99% is reached for ET > 15 GeV.
During Run 2 a new algorithm was implemented for energy cluster reconstruction. Unlike the sliding window algorithm, the new method uses topo-clusters of cells which allow for irregular shapes of the reconstructed energy cluster.
Table of contents :
Synthèse en français
1 Elementary Particle Physics
1.1 Standard Model of Particle Physics
1.2 Fundamental forces
1.2.1 Electromagnetic interaction
1.2.2 Strong interaction
1.2.3 Electroweak interaction
1.2.4 Spontaneous symmetry breaking and Higgs mechanisms
1.3 Standard Model limitations and BSM theories
1.4 Higgs boson phenomenology
1.4.1 Production at LHC
1.4.2 Decay modes
1.4.3 Discovery at LHC
1.4.4 Properties measured at LHC
1.4.5 Yukawa couplings
2 The LHC and the ATLAS detector
2.1 The LHC
2.2.1 The inner detector
2.2.2 The calorimeter system
2.2.3 The muon system
2.2.4 The magnets system
2.2.5 The trigger system
2.3 Object reconstruction
2.3.1 Track and vertex reconstruction
2.3.2 Electrons and photons
2.3.6 Missing transverse momentum
3 Electron performance
3.2 Electron efficiencies generalities
3.2.1 Tag and Probe method
3.2.2 Event and object selection
3.2.3 Data and simulation samples
3.3 Electron reconstruction efficiency measurement
3.3.1 Previous measurements
3.3.2 Background estimation
3.3.4 Efficiency and Scale Factors measurement
3.3.5 Results with new reconstruction and more statistics
3.4 Non-prompt electron tagger efficiency
3.4.1 PromptLeptonIso tagger
3.4.2 Charge flip tagger
3.4.3 Efficiency definition
3.4.4 PromptLeptonIso results
3.4.5 PromptLeptonVeto tagger
4 Search for flavor changing neutral currents in top decays
4.2 Previous results
4.3 Analysis strategy
4.4 Object and event selections
4.4.1 2`SS channel selections
4.4.2 3` channel selections
4.5 Signal and Backgrounds
4.5.1 Irreducible backgrounds
4.5.2 Reducible backgrounds
4.6 Event MVA
4.6.1 2`SS category
4.6.2 3` category
4.7.1 Fit model
4.7.3 Fit results
4.7.4 ATLAS combination
5 Search for Higgs boson production in association with a top-antitop quark pair
5.1 Observation of the ttH production mode
5.1.1 ttH Multilepton (36 fb−1)
5.1.2 Combination including t¯tH(H !) (79.8 fb−1) and t¯tH(H ! b¯b) (36 fb−1)
5.2 Two-lepton same-sign analysis with 80 fb−1
5.2.1 μμ(Nb-jets 3) Region
5.2.2 Fake lepton estimation
5.2.3 Event yields in the pre-MVA region
5.2.4 Event MVA
5.2.5 Expected fit results
5.3 Current status of 80 fb−1 analysis