Search for a supersymmetric partner to the top quark in final states with jets and missing transverse momentum at sqrt(s) = 7 TeV with the ATLAS detector

A search for direct pair production of supersymmetric top squarks (stop_1) is presented, assuming the stop_1 decays into a top quark and the lightest supersymmetric particle, neutralino_1, and that both top quarks decay to purely hadronic final states. A total of 16 (4) events are observed compared to a predicted Standard Model background of 13.5+3.7-3.6 (4.4+1.7-1.3) events in two signal regions based on int(Ldt) = 4.7 fb^-1 of pp collision data taken at sqrt(s) = 7 TeV with the ATLAS detector at the LHC. An exclusion region in the stop_1 versus neutralino_1 mass plane is evaluated: 370<m(stop_1)<465 GeV is excluded for m(neutralino_1)~0 GeV while m(stop_1)=445 GeV is excluded for m(neutralino_1)<=50 GeV.

Search for a supersymmetric partner to the top quark in final states with jets and missing transverse momentum at √ s = 7 TeV with the ATLAS detector The ATLAS Collaboration (Dated: August 7, 2012) A search for direct pair production of supersymmetric top squarks (t1) is presented, assuming thet1 decays into a top quark and the lightest supersymmetric particle,χ 0 1, and that both top quarks decay to purely hadronic final states. A total of 16 (4) events are observed compared to a predicted Standard Model background of 13.5 + 3.7 The Standard Model (SM) is a successful but incomplete model of particle interactions. Supersymmetry (SUSY) [1][2][3][4][5][6][7][8][9] provides an elegant solution to cancel the quadratic mass divergences that would accompany a SM Higgs boson by introducing supersymmetric partners of all SM particles, such as a scalar partner of the top quark (t). Like tt, directtt ⋆ is produced primarily through gluon fusion at the Large Hadron Collider (LHC). The production cross section depends mostly on the mass of the top partner and has minimal dependence on other SUSY parameters [10][11][12]. The LHC enables searches for directtt ⋆ production at higher mass scales than previous accelerators [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27]. The viability of SUSY as a scenario to stabilize the Higgs potential and to be consistent with electroweak naturalness [28,29] is tested by the search fort below the TeV scale.
In this Letter, we present a search for directtt ⋆ production assumingt 1 → tχ 0 1 → bWχ 0 1 wheret 1 is the lightestt eigenstate andχ 0 1 represents the lightest supersymmetric particle (LSP) in R-parity conserving models [30][31][32][33][34]. We consider events where both W bosons decay hadronically, yielding a final state with six high transverse momentum (p T ) jets from the all-hadronic tt final state and large missing transverse momentum (E miss T ) from the LSPs. The kinematics of both top quarks are therefore fully specified by the visible decay products. Additionally, SM backgrounds from all-hadronic tt are suppressed as there is no intrinsic E miss T except from semi-leptonic c-and bquark decays. The dominant background consists of tt events that contain a W → ℓν decay where the lepton (ℓ), often of τ flavor, is either lost or mis-identified as a jet and which have large E miss T from the neutrino.
The data sample was acquired during 2011 in LHC pp collisions at a center-of-mass energy of 7 TeV with the ATLAS detector [35], which consists of tracking detectors surrounded by a 2 T superconducting solenoid, calorimeters, and a muon spectrometer in a toroidal magnetic field. The high-granularity calorimeter system, with acceptance covering |η| < 4.9 [36], is composed of liquid argon with lead, copper, or tungsten absorbers and scin-tillator tiles with steel absorbers. This data set, composed of events with a high-p T jet and large E miss T as selected by the trigger system, corresponds to an integrated luminosity of 4.7 fb −1 with a relative uncertainty of 3.9% [37,38].
Jets are constructed from three-dimensional clusters of calorimeter cells using the anti-k t algorithm with a distance parameter of 0.4 [39,40]. Jet energies are corrected [41] for losses in material in front of the active calorimeter layers, detector inhomogeneities, the noncompensating nature of the calorimeter, and the impact of multiple overlapping pp interactions. These corrections are derived from test beam, cosmic ray, and pp collision data, and from a detailed Geant4 [42] detector simulation [43]. Jets containing a b-hadron are identified with an algorithm exploiting both the impact parameter and secondary vertex information [44,45]. A factor correcting for the slight differences in the b-tagging efficiency between data and the Geant4 simulation is applied to each jet in the simulation. The b-jets are restricted to the fiducial region of the tracker, |η| < 2.5. Non-tt backgrounds are minimized by requiring either ≥ 1 b-jets with a selection corresponding to a 60% efficiency with a low < 0.2% mis-identification rate (tight), or ≥ 2 bjets each with 75% efficiency but a higher ≈ 1.7% misidentification rate per b-jet (loose).
The E miss T is the magnitude of p miss T , the negative vector sum of the p T of the clusters of calorimeter cells, calibrated according to their associated reconstructed object (e.g., jets and electrons), and of the p T of muons above 10 GeV within |η| < 2.4. Events containing fake E miss T induced by jets associated with calorimeter noise or non-collision backgrounds [46], or by cosmicray muons or poorly reconstructed muons [47,48], are removed from consideration. Large p miss T colinear with a high-p T jet could indicate a significant fluctuation in the reconstructed jet energy or the presence of a semileptonic c-or b-quark decay. Therefore, the difference in azimuthal angle (∆φ) between the p miss T and any of the three highest-p T jets in the event, ∆φ(p miss T , jet), is re-quired to be > π/5 radians. Fake E miss T is also suppressed by requiring that the ∆φ between the above computed p miss T and one calculated with the tracking system, using tracks having p T > 0.5 GeV, is < π/3 radians.
Events are required to have at least one jet with p T > 130 GeV in |η| < 2.8 and E miss T > 150 GeV to ensure full efficiency of the trigger. At least five other jets having p T > 30 GeV and |η| < 2.8 must be present. In addition to the jet and E miss T requirements, events containing "loose" electrons [49,50] with p T > 20 GeV and |η| < 2.47 that do not overlap with any jet within ∆R < 0.4, where ∆R = (∆η) 2 + (∆φ) 2 , are rejected. Similarly, events with muons [47, 51] having p T > 10 GeV and |η| < 2.4 that are separated by ∆R > 0.4 from the nearest jet are rejected. A jet with 1-4 tracks and ∆φ(p miss T , jet) < π/5 indicates a likely W → τ ν decay. Events with such τ -like jets that have transverse mass The presence of high-p T top quarks that decay through t → bW → bjj in thet 1t ⋆ 1 final state is exploited to further reduce SM backgrounds by only considering events with reconstructed three-jet invariant masses consistent with the top-quark mass (m t ). A clustering technique resolves the combinatorics associated with highmultiplicity jet events. The three closest jets in the η − φ plane are combined to form one triplet; a second triplet is formed from the remaining jets by repeating the procedure. The resulting three-jet mass (m jjj ) spectrum is shown in Fig. 1 for the control region constructed from ℓ+jets events (defined below). There is a clear peak associated with the hadronically-decaying top quarks above a small non-tt background. A requirement of 80 < m jjj < 270 GeV is placed on each reconstructed triplet in the event. The kinematics of the t → bW → bℓν decay is also exploited to further reduce the dominant ℓ+jets tt background, as the m T distribution of the p miss T and b-jet (m jet T ) has an endpoint at m t . When there are ≥ 2 loose b-jets, the m jet T for the b-jet closest to the p miss T is required to be > 175 GeV. The largest m jet T , calculated for each of the four highest-p T jets, is required to be > 175 GeV in the case of only one tight b-jet.
Two signal regions (SR) are defined including the above kinematic and mass requirements. The first, which requires E miss T > 150 GeV (SRA), is optimized for low mt 1 , while the second, requiring E miss T > 260 GeV (SRB), is used for higher mt 1 . Using these signal regions, the search is most sensitive tot 1t  cross sections are calculated to next-to-leading order in the strong coupling constant, including the resummation of soft gluon emission at next-to-leading-logarithmic accuracy (NLO+NLL) [10][11][12]. The nominal production cross section and associated uncertainty are taken from an envelope of cross section predictions using different PDF sets and factorization and renormalization scales, as described in Ref. [55]. Thet 1t ⋆ 1 cross section for mt 1 = 400 GeV is σt 1t ⋆ 1 = 0.21 ± 0.03 pb. In the signal region, the dominant source of SM background is tt → τ +jets events where the τ lepton is reconstructed as a jet. Additional, smaller, backgrounds include other tt → ℓ+jets final states, tt + V where V represents a W or Z boson, single top quark production, V +jets, and V V +jets. The tt events are produced with ALPGEN [56] using the CTEQ6L1 PDF [57] and interfaced to HERWIG [58,59] for particle production and JIMMY [60] for the underlying event model. Additional tt samples generated with MC@NLO [61,62] and AcerMC [63], interfaced to HERWIG and JIMMY, are used to estimate event generator systematic uncertainties. Samples of tt+ V are produced with MadGraph [64] interfaced to PYTHIA [59,65,66]. Single top events are generated with MC@NLO [67,68] and AcerMC. The associated production of W and Z bosons and light and heavy-flavor jets is simulated using ALPGEN; diboson production is simulated with SHERPA [69].
All samples are passed through the Geant4 simulation of the ATLAS detector, and are reconstructed in the same manner as the data. The simulation includes the effect of multiple pp interactions and is weighted to reproduce the observed distribution of the number of interactions per bunch crossing. SM event samples are normalized to the results of higher-order calculations using the cross sections cited in Ref. [70] except for the ℓ+jets tt background. This sample is renormalized by a factor that scales the tt expectation to agree with the observed data in a control region (CR) kinematically close to the signal region but with little expected signal. The CR is constructed from events containing one muon or one "tight" electron [49] with p T > 30 GeV consistent with originating from a W -boson decay (40 < m ℓ T < 120 GeV) and ≥ 5 jets, where m ℓ T is the transverse mass of the electron or muon and p miss T . The lepton must be isolated such that the scalar p T sum of tracks in a cone of ∆R < 0.2 around the lepton, excluding the track of the lepton, is < 1.8 GeV for the muon or is < 10% of the electron p T , respectively. The jet, b-jet, and E miss T requirements remain the same as the standard signal selection; however, some topological constraints are relaxed (∆φ(p miss T , jet) > π/10 radians and m jjj < 600 GeV) and others removed (m jet T ) to gain statistics. The tt purity in the control region is > 80%; the expected signal contamination is < 3%. The lepton is treated as a jet of the same energy and momentum, mimicking the effect of the τ lepton. Effects of the additional E miss T from the τ neutrino are smaller than the statistical uncertainties. The scale factor needed to bring the ≥ 6 jet ℓ+jets ALPGEN tt events into agreement with the data after recalculating all quantities except E miss T is 0.66 ± 0.05; the uncertainty quoted here is statistical only. This scale factor is used in Figs. 1−3. The normalization is validated with an orthogonal tt-dominated sample created from SRA by selecting events with τ -like jets; the requirement on m jet T is removed to increase the sample size. The m T of τ -like jets is shown in Fig. 2, where the tt sample has been normalized as described above. Expectations from the simulation agree with the data within uncertainties. Contributions to the signal region from QCD multi-jet and all-hadronic tt production are estimated with a data-driven technique [71]. Jets are smeared in a low-E miss T data sample using response functions derived from control regions dominated by multi-jet events. Only 0.2 ± 0.2 such events remain in SRA after the full event selection.
The E miss T distribution in SRA is shown in Fig. 3 for data, for the SM backgrounds, and for expectations of t 1t ⋆ 1 production with mt 1 = 400 and mχ0 1 = 1 GeV. Numbers of events and combined statistical and systematic uncertainties, for both SRA and SRB, are tabulated in Table I. Uncertainties in the event generators, including the impact of initial-and final-state radiation, are the dominant source of systematic uncertainty of 28% (23%) in SRA (SRB). Other major sources of uncertainty include 22% (32%) for the jet energy calibration, 6.5% (6.8%) for jet energy resolution, 5.9% (6.2%) for b-jet identification, and 1.4% (1.5%) for E miss T in SRA (SRB). The number of observed events in the data is well matched by the SM background. These results are interpreted as exclusion limits for mt 1 and mχ0 1 using a CL s likelihood ratio combining Poisson probabilities for signal and background [72]. Systematic uncertainties are treated as nuisance parameters assuming Gaussian distributions. Uncertainties associated with jets, b-jets, E miss T , and luminosity are fully correlated between signal and background; the others are assumed to be uncorrelated. The expected limits for the signal regions are evaluated for each (mt 1 , mχ0 1 ) point; the SR with the better expected sensitivity is used for that point. The expected and observed 95% C.L. exclusion limits are displayed in Fig. 4. Top squark masses between 370 and 465 GeV are excluded for mχ0 1 ∼ 0 GeV while mt 1 = 445 GeV is excluded for mχ0 1 ≤ 50 GeV. These values are derived from the −1σ observed limit contour to account for theoretical uncertainties on the SUSY cross sections. The 95% CL s upper limit on the number of events beyond the SM in each signal region, divided by the integrated luminosity, yields limits on the observed (expected) visible cross sections of 2.9 (2.5) fb in SRA and 1.3 (1.3) fb in SRB.
In conclusion, we have presented a search for the direct production oft 1t