Searches for direct pair production of third generation squarks with the ATLAS detector

The most recent results of the searches for direct production of third generation squarks with the ATLAS detector are presented. All the selected analyses use 21 fb−1 of pp collisions at √ s = 8 TeV and follow the assumptions of a Minimal Supersymmetric extension of the Standard Model of Particle Physics where the R-parity is conserved and the lightest supersymmetric particle (LSP) is the lightest neutralino. Naturalness considerations require that the partners of the top and bottom quarks are light (O(1 TeV)) and hence have a relatively high production cross-section at the LHC. Since no deviations from the Standard Model predictions are observed, limits are set for different mass hierarchies and mass relations between the bottom and top squarks and the lightest chargino and neutralino.


Introduction
The results of the searches for direct stop and direct sbottom production with the ATLAS detector [1], all updated to the full 2012 √ s = 8 TeV dataset of ∼ 20 fb −1 , are presented.
The searches are based on a Minimal Supersymmetric (SUSY) extension of the Standard Model (SM) of particle physics, where the R-parity is conserved. SUSY particles are therefore pair-produced and the lightest supersymmetric particle (LSP) is a stable, neutral and only weakly interacting particle and hence a good Dark Matter candidate.
Naturalness requires low fine tuning in the theory and implies a light higgsino sector, gluinos with a mass below 2 TeV and light partners of the 3 rd generation SM quarks, making their production cross-section accessible for searches at the LHC.

Searches for direct sbottom production
The direct sbottom production and decay scenarios have been covered by two analyses, one focusing on final states with two leptons with the same charge and jets (Sect. 2.1) and one that requires two b-tagged jets and high E miss T (Sect. 2.2).

Searches in final states with 2 same-sign leptons and jets
The search for Supersymmetry in final states with two Same Sign (SS) leptons [3] has interpretations in many SUSY production and decay scenarios. Here only the interpretation for direct sbottom production withb 1 → t +χ ± 1 ,χ ± 1 → W ( * ) +χ 0 1 , with branching ratio BR=1, is considered. This analysis exploits the low SM background to get a good sensitivity to SUSY processes. The only sources of SS leptons in the Standard Model are the associated production of tt+vector-boson and the diboson production, that have to be added to the contributions containing misidentified leptons. These consist of processes where the SS pair arises either from charge-misidentification (radiation of a hard photon bremsstrahlung followed by an asymmetric conversion) or from a non-prompt lepton mainly produced in a b-hadron decay.   Figure 1. Expected (dashed lines) and observed (solid lines) limits at 95% CL for direct sbottom pair production. On the left the limit for theb 1 → tχ ± 1 model with m(χ ± 1 ) = 2 · m(χ 0 1 ) is shown [3]; the numbers give the limits on the excluded model cross sections in fb. The right plot shows the limit obtained in theb 1 -χ 0 1 plane assumingb 1 → bχ 0 1 with 100% BR [2].
The signal selection requires, on top of the SS lepton pair, the presence of at least three jets with p T > 40 GeV, of which one identified as b-jet with p T > 20 GeV, > 100 GeV and high effective mass, m e f f = 2 i=1 p i T + n− jets k=1 p jet k T + E miss T , in the event.
To maximise the exclusion reach, a shape fit in m e f f is performed. The limits are computed as 95% Confidence Levels (CL), using the CL s method, with a fit based on the profile likelihood ratio [13]. Exclusion limits are extracted for two different signal assumptions in the m(b 1 ) − m(χ ± 1 ) plane. For a fixed neutralino mass m(χ 0 1 ) = 60 GeV sbottom masses up to 470 GeV are excluded at 95% CL for all the allowed chargino masses. Figure 1 left shows the excluded area for the model m(χ ± 1 ) = 2 · m(χ 0 1 ): sbottom masses up to 450 GeV are excluded at 95% CL for the allowedχ ± 1 masses.

Searches in final states with 2 b-jets and E miss T
The search focusing on final states with 2 b-tagged jets, E miss T and very low additional hadronic activity [2] is developed to investigate direct sbottom and direct stop production, for which the results are reported in section 3.1. The scenario considered for direct sbottom production and decay isb 1 → bχ 0 1 , BR=1, and limits are extracted in the m(b 1 ) − m(χ 0 1 ) plane. The main backgrounds to this analysis come from the semi-leptonic decay of tt pairs, where the lepton is either an hadronic decaying τ or is lost or misidentified, and Z → νν and W production in association with b-jets. Two different signal selections (SR) are defined, one sensitive to large mass splitting between the relevant SUSY particles, in this caseb 1 andχ 0 1 , and one that exploits an initial state radiation (ISR) signature, that is, a hard jet emitted back to back to theb 1b * 1 system giving higher E miss T in the event, and is sensitive to compressed scenarios. For both SRs the selected events must pass the requirements for a E miss T trigger, not contain any lepton, have two b-tagged jets, and have almost no additional hadronic activity on top of the requested jets. The selection for large mass splittings scenarios uses the contransverse mass [14] of the two b-jets, m 2 2 , as main discriminating variable against tt, and the invariant mass of the bb system, m bb > 200 GeV, against the Z background. Five signal enhanced selections are defined in step of 50 GeV for increasing values of m CT starting from m CT > 150 GeV. For the ISR selection, the leading jet is required to be a high p T non b-tagged jet with a big angular separation with the E miss T , and the b-tagged jets are selected as the second and third leading jets. A tight requirement is then applied on E miss T > 250 GeV and on the scalar sum of the momenta carried by the additional jets in the event, GeV. No significant excess over the SM model was observed and 95% CL limits were set using the CL s prescription. The obtained exclusion curve is shown in figure 1 on the right. Sbottom masses up to 620 GeV are excluded for neutralino masses up to 220 GeV.

Searches in final states with 2 b-jets and E miss T
The structure of the 2 b-jets and E miss T analysis has already been described in detail in section 2.2 in relationship to the interpretation in a direct sbottom production model. This final state is powerful also to detect direct stop production in the scenariost 1 → bχ ± 1 ,χ ± 1 → W ( * )χ0 1 , with BR=1, where the mass difference between the chargino and the neutralino becomes small and therefore the decay products of the W get lost [2]. The exclusion contour at 95% CL in thet 1 -χ 0 1 mass plane for a fixed m(χ ± 1 ) − m(χ 0 1 ) = 5 GeV covers stop masses up to 580 GeV and neutralino masses up to 250 GeV for a mass of the stop of 500 GeV, see figure 2 teal-coloured contour. When considering a m(χ ± 1 ) − m(χ 0 1 ) = 20 GeV the excluded area shrinks and stop masses up to 420 GeV are excluded at 95% CL for neutralino masses below 200 GeV. The simulated signal grids respect the LEP limit on the mass of theχ ± 1 just above 100 GeV.

Searches in final states with jets and E miss T
The search for direct stop production in all hadronic final states with high missing transverse momentum [4] has an interpretation fort 1 → tχ 0 1 with BR=1, orange curve in figure 2.
The main background sources to the analysis come from the tt semi-leptonic decay, the associated production of Z → νν with b-jets, the associated production of tt+vectorboson, which is the only irreducible background, and fake sources such as multi-jet production and all-hadronic tt.
The signal signature is obtained requiring high E miss T and six jets of which at least two b-tagged, the two leading ones with a 80 GeV transverse momentum threshold, the others with p T > 35 GeV. The rejection of the semileptonic decaying top and the electroweak background is achieved by vetoing the events with a τ lepton, by requiring that the transverse mass between the nearest bjet to the E miss T and the E miss T itself is above the top mass (m T (b near , E miss T ) > 175 GeV) and by reconstructing the two hadronic decaying tops. The final discriminating variable is E miss with right polarisation. It has been verified that the analysis results are independent on the top polarisation choice. A study has then been done to determine the maximum branching ratio excluded at 95% CL assuming that the analysis is completely insensitive to the decay products not coming from thet 1 → tχ 0 1 decay. For a BR of 75% stop masses between 340 GeV and 600 GeV are excluded for an almost massless neutralino.

Searches in final states with 1 lepton, jets and
The search for direct stop production in final states with one lepton, jets and E miss T [5] is the analysis able to cover better the bulk of thet 1 ,χ ± 1 ,χ 0 1 parameter space. It has interpretations fort 1 → tχ 0 1 and fort 1 → bχ ± 1 scenarios. The baseline signal selection requires exactly one well identified and isolated lepton, e or µ, with p T > 25 GeV, 4 jets or more, the four leading with p T > 80, 60, 40, 25 GeV; among them at least one or two, depending on the signal model under consideration, has to be identified as bjet. The main background processes to the analysis arise, therefore, from tt and W + jets production. To suppress the SM contributions, a series of selections are further applied. The most powerful variables used to reject the tt and W backgrounds are the E miss T and the transverse mass between the lepton and the E miss T . The analysis makes also use of a τ veto and of two asymmetric version of the m T2 variable [15,16]. The m T2 variable is powerful to discriminate against the semi-leptonic tt background since the variable presents a sharp endpoint for the top and not for the signal. The endpoint depends on the mass of the invisible particles in the tt decay, while for the signal the dependancy comes from thet 1 -χ 0 1 orχ ± 1 -χ 0 1 mass difference. It is defined as and the two variations considered are am T2 and m τ T2 . The former is an asymmetric definition where the missing particles in the dileptonic top decay are defined as the W and the neutrinos and has the end point at m(t); the latter is defined for dileptonic top decays where one of the leptons is an hadronic decaying τ and in this case the end point of the top distribution is at m(W). Different thresholds and combinations of the discriminating variables are chosen depending on the targeted scenariot 1 → tχ 0 1 ort 1 → bχ ± 1 and three signal enhanced regions are defined for each of the two signal models. Fort 1 → tχ 0 1 a SR is defined in bins of m T and E miss T and a combined shape fit is performed to extend the analysis sensitivity as much as possible towards the region m(t 1 ) ∼ m(t) + m(χ 0 1 ). In thet 1 → tχ 0 1 scenario, for an almost massless neutralino, stop masses are excluded at 95% CL between 200 and 620 GeV and for a stop mass of 500 GeV a neutralino is excluded up to 250 GeV, as shown in figure 2 with a dark yellow contour. As for the stop search in all hadronic final states, described in section 3.2, the signal grid has been simulated with a stop decaying mainly in t R . This analysis is found to be sensitive to the top polarisation; when considering a stop decaying only in left polarised top, a loss in acceptance is observed that translates in a weaker upper limit on the stop mass of about 50 GeV. The exclusion contour at 95% CL for thet 1 figure 2 blue contour, greatly extends the area excluded with the 7 TeV ATLAS dataset, dark blue contour. Now stop masses between 200 and almost 500 GeV are excluded at 95% CL for neutralino masses ranging between 70 and 140 GeV. The light green contour of figure 2 shows the excluded area at 95% CL for a fixed m(χ ± 1 ) = 150 GeV, here the analysis is able to exclude stop decaying in bχ ± 1 up to a mass of 450 GeV for a 40 GeV neutralino.

Searches in final states with 2 opposite
charged leptons and E miss T Final states with two opposite charged (OS) leptons allow to select a clean signature without the requirement of high jet multiplicity and E miss T in the event. The stop search that exploits the two lepton selection [6] is therefore sensitive to all the scenarios were jets are soft or are out of the detector acceptance and hence not correctly reconstructed. In particular, a good sensitivity is reached for thet 1 → bχ ± 1 ,χ ± 1 → W ( * )χ0 1 model when assuming a small ∆m(t 1 ,χ ± 1 ) and for the region m(t 1 ) ∼ m(b) + m(χ ± 1 ) with m(χ ± 1 ) = 2 · m(χ 0 1 ). For the first time a limit under the assumption of a three-body decay of the stop iñ t 1 → b + W +χ 0 1 is extracted. The three-body decay channel opens when the mass difference between the stop and the neutralino becomes smaller than the mass of the top and thet 1 → tχ 0 1 decay with an off-shell top is disfavoured with respect to the stop three-body decay in b + W +χ 0 1 . The main SM backgrounds in this channel are the dileptonic top and the diboson production. The signal selection is based on the requirement of the lepton pair and uses as discriminating variable that has an endpoint at the W mass for the tt background. Four signal enhanced region are defined by increasing the lower threshold of the m T2 variable; the most inclusive SR is defined for m T2 > 90 GeV and has no requirements on jets, the tighter SRs require m T2 > 100, 110, 120 GeV and the presence of two jets in the event.
In thet 1 → bχ ± 1 model for m(χ ± 1 ) = 2 · m(χ 0 1 ), the analysis excludes at 95% CL on the diagonal m(t 1 ) = m b + m(χ ± 1 ) stop masses between 200 and 420 GeV, as shown in figure 2 blue contour; the light blue area of figure 2 represent, instead, the excluded region at 95% CL under the constraint m(t 1 ) − m(χ ± 1 ) = 10 GeV for which stop masses up to 440 GeV are discarded for an almost massless neutralino. The purple contour of figure 2 is the 95% CL limit obtained for the three-body decay scenario where stop masses up to 200 GeV are excluded for almost all allowed neutralino masses.

Searches in decay chains with Z
The search for direct stop production in decay chains with a Z-boson exploits final states where at least two leptons are present [7]. Two different models are considered, the direct pair production and decay oft 2 → Zt 1 ,t 1 → tχ 0 1 , particularly interesting for covering the region m(t 1 ) − m(χ 0 1 ) = m(t), and a natural GMSB (Gauge-Mediated Supersymmetry Breaking) scenario where a pair oft 1 is produced and decays in final states with a Z and a gravitino, in this case the LSP. The signal selection requires two oppositely charged leptons, whose invariant mass is compatible with the Z-boson mass, jets, of which at least one identified as b-jet, and high E miss T ; for thet 2 model an additional lepton is also required. The main SM backgrounds come from tt, also in association with vector boson, and diboson production. The obtained exclusion limits at 95% CL are show in figure 3; on the left, fort 2 → Zt 1 , with m(t 1 ) = m(χ 0 1 ) + 180 GeV,t 2 masses are excluded up to 530 GeV for all the allowed neutralino masses; in the natural GMSB scenario the extracted exclusion is fort 1 masses up to 450 GeV for all the allowed neutralino masses, as shown in the right plot.