Search for direct top squark pair production in events with a single isolated lepton, jets and missing transverse energy at √s = 8 TeV (SUS-12-023)

Further information

This analysis is documented in SUS-12-023.

Abstract

This note presents a search for top squark pair production in events with a single isolated electron or muon, jets, and large missing transverse energy. At least one of the jets is required to be identified as originating from a b-quark and events are required to have large transverse mass. The data sample used corresponds to an integrated luminosity of 9.7 fb^-1 of pp collisions, collected in 2012 by the CMS experiment at the LHC, at a centre-of-mass energy of 8 TeV. No significant excess in data is observed above the expected standard model backgrounds. The results are interpreted in the context of models of top squark pair production where the top squarks decay either to a top quark and a neutralino or to a bottom quark and a chargino. Depending on the decay mode, the results probe top squarks with masses in the range of 160–430 GeV.

Analysis Summary

The search presented focuses on two decay modes of the top squark (t ̃): t ̃ → tχ ̃0 and t ̃ →bχ ̃± → bWχ ̃0, which are expected to have large branching fractions if kinematically accessible. The signature of the signal process includes high transverse momentum jets, including two b-jets, and missing ET. The selection applied requires exactly one isolated, high pT electron or muon, at least 4 jets, out of which at least one is b-tagged, and missing ET. We search for an excess of events with large transvere mass (MT) since the largest backgrounds in the single lepton topology contain a single leptonically-decaying W boson, and the MT of the lepton-neutrino system has a kinematic endpoint requiring MT < MW. For the signal, the presence of additional LSP’s in the final state allows MT to exceed MW. The dominant background in this kinematic region is dilepton tt where one of the leptons is not identified, since the presence of the additional neutrino from the second leptonically-decaying W boson allows MT to exceed MW .Seven signal regions are defined, based on missing ET and MT requirements, in order to be sensitive to a range of signal kinematics, dependent on the masses of the top squark and LSP. Backgrounds are estimated from Monte Carlo, with scale factors (where necessary) derived in control regions. The control regions are used to validate the background modeling in the MC and derive the systematic uncertainties on the backgrounds. Agreement is observed between the data and the predicted backgrounds for all signal regions. The results are interpreted in the context of simplified SUSY models where the stops decay to top-neutralino or b-chargino and are used to place constraints on the stop mass.

Approved Tables and Plots ( click on plot to get larger version )

Introduction

Figure	Caption
	Figure 1a : Diagram for top squark pair production for the t ̃ → tχ ̃0 decay mode.
	Figure 1b : Diagram for top squark pair production for the t ̃ → bχ ̃± → bWχ ̃0 decay mode.

Signal and Control Regions

Table	Caption
	Table 1 : Signal region definitions based on MT and missing ET requirements. These requirements are applied in addition to the baseline single lepton selection (1 high pT isolated electron or muon, ≥ 4 jets out of which ≥ 1 is b-tagged).
	Table 2 : Summary of signal and control regions defined in terms of the lepton and b-tagged jet multiplicities. The control regions (CRs) are used to test the background estimation procedure and derive the systematic uncertainties on the dilepton ttbar (CR4 and CR5) and single lepton (CR1 and CR2) background contributions.

Control Region Studies

Figure	Caption
	Figure 2 : Comparison of the jet multiplicity distribution in data and MC for a sample composed primarily of dilepton ttbar events. This distribution is used to validate the modeling of additional jets from initial/final state radiation that are required for the dilepton ttbar background to satisfy the analysis selection requirement of at least four jets.
	Figure 3a : Comparison of the missing ET distributions in data and MC for events satisfying the requirements of CR4 (two reconstructed leptons and at least 1 b-tagged jet), which is a sample predominantly composed of dilepton ttbar.
	Figure 3b : Comparison of the MT distributions in data and MC for events satisfying the requirements of CR4 with missing ET > 150 GeV. This control sample is used to validate the MC modeling of the dominant dilepton ttbar background and derive the systematic uncertainties on this background.
	Figure 4a : Comparison of the missing ET distributions in data and MC for events satisfying the requirements of CR5 (a reconstructed lepton, an isolated track and at least 1 b-tagged jet). The sample is predominantly composed of dilepton ttbar and serves to validate the MC modeling of this dominant background.
	Figure 4b : Comparison of the MT distributions in data and MC for events satisfying the requirements of CR5 with missing ET > 150 GeV. As with CR4, this control sample is used to validate the MC modeling of the dominant dilepton ttbar background and derive the systematic uncertainties on this background.
	Figure 5a : Comparison of the missing ET distributions in data and MC for events satisfying the requirements of CR1 (the full analysis selection is applied but no b-tagged jets are required). The resulting sample is dominated by W+jets events and is used to validate the MC modeling of this background.
	Figure 5b : Comparison of the MT distributions after a missing ET > 150 GeV requirement in data and MC for events satisfying the requirements of CR1 (the full analysis selection is applied but no b-tagged jets are required). The resulting sample is dominated by W+jets events and is used to extract data/MC corrections for the modeling of this background in the MT tail, described in terms of the tail-to-peak ratio, and estimate the corresponding uncertainty. In the tail of the MT distribution, the data/MC scale factor on the tail-to-peak ratio is applied to the MC. The band corresponds to the uncertainty on the data/MC scale factor.
	Figure 6a : Comparison of the pseudo-missing ET distributions in data and MC for CR2, a control region dominated by Z+≥ 4 jet events.The pseudo-missing ET is constructed by adding one of the leptons to the missing ET.
	Figure 6b : Comparison of the pseudo-MT distributions in data and MC for CR2, a control region dominated by Z+≥ 4 jet events. The pseudo-missing ET is constructed by adding one of the leptons to the missing ET and the pseudo-MT includes the pseudo-missing ET and the other lepton. This sample is used to validate the MC modeling of the MT tail associated with jet resolution effects, since the Z is required to be on-shell (we require 81 < Mll < 101 GeV). In single lepton top background, jet resolution effects are the dominant contributors to the MT tail, since the W boson from the top decay cannot be far off-shell. The tail of the pseudo-MT distribution is used to extract data/MC corrections for the modeling of this background in the MT tail, described in terms of the tail-to-peak ratio, and estimate the corresponding uncertainty. In the tail of the pseudo-MT distribution, the data/MC scale factor on the tail-to-peak ratio is applied to the MC. The band corresponds to the uncertainty on the data/MC scale factor.

Systematic Uncertainties and Results

Table	Caption
	Table 3 : Summary of the contributions to the relative uncertainties in the background predictions and the total relative background uncertainty. All values are quoted in %. In SRA and SRB the dominant uncertainty is from the single lepton top backgrounds. For all other signal regions the uncertainty in the dilepton ttbar background, assessed based on the data to MC comparisons in CR4 and CR5, dominates.
	Table 4 : The result of the search. For each of the 7 signal regions the individual background contributions, total background, and observed yields are indicated, separately for the electron and muon channels and for the combination of the two. The uncertainty includes both the statistical and systematic components. The observed yields in the signal regions are in agreement with the predicted background. No evidence for an excess is observed.

Figure	Caption
	Figure 7 : Comparison of the missing ET distributions in data and for the predicted background, after the requirement MT > 120 GeV. Expected missing ET distributions from two sample signal points are indicated: t ̃ → tχ ̃0 where m(t ̃) = 450 GeV and m(χ ̃0) = 50 GeV and t ̃ → bχ ̃± → bWχ ̃0 where m(t ̃) = 450, m(χ ̃±) = x m(t ̃) +(1−x) m(χ ̃0) with x = 0.75, and m(χ ̃0) = 50 GeV.
	Figure 8a : Comparison of the MT distributions in data and MC for events with missing ET > 100 GeV. The region with MT > 150 GeV corresponds to SRA. Expected MT distribution from two sample signal points are indicated: t ̃ → tχ ̃0 where m(t ̃) = 450 GeV and m(χ ̃0) = 50 GeV and t ̃ → bχ ̃± → bWχ ̃0 where m(t ̃) = 450, m(χ ̃±) = x m(t ̃) +(1−x) m(χ ̃0) with x = 0.75, and m(χ ̃0) = 50 GeV. The ratio of data to expected background is indicated in the top plot. The error bars indicate the statistical uncertainty in the data.
	Figure 8b : Comparison of the MT distributions in data and MC for events with missing ET > 150 GeV. The region with MT > 120 GeV corresponds to SRB. Expected MT distribution from two sample signal points are indicated: t ̃ → tχ ̃0 where m(t ̃) = 450 GeV and m(χ ̃0) = 50 GeV and t ̃ → bχ ̃± → bWχ ̃0 where m(t ̃) = 450, m(χ ̃±) = x m(t ̃) +(1−x) m(χ ̃0) with x = 0.75, and m(χ ̃0) = 50 GeV. The ratio of data to expected background is indicated in the top plot. The error bars indicate the statistical uncertainty in the data.

Interpretation

Table	Caption
	Figure 9 : Interpretation in the top squark pair production model with t ̃ → tχ ̃0, in the plane of m(χ ̃0) vs. m(t ̃). The t ̃ decay matrix element is treated as flat, which is equivalent to the assumption of 50/50 mixing of left-handed and right-handed top super partners. The upper limits on the signal cross section are calculated separately for each signal region, incorporating the uncertainties in the signal acceptance and efficiency, using the LHC-type CLs criterion. For each point in the signal model parameter space, the observed limit is taken from the signal region with the best expected limit. The shading indicates the resulting upper limit on the signal cross section. The observed, median expected, and ±1σ expected exclusion contours are indicated assuming NLO-NLL cross sections, as well as the observed contours when the theory cross section is varied by ±1σ (the region below the contours is excluded).
	Figure 10a : Interpretation in the top squark pair production model with t ̃ → bχ ̃± → bWχ ̃0, in the plane of m(χ ̃0) vs. m(t ̃), for the mass of the intermediate charginoχ ̃± specified by the parameter x=0.5, where m(χ ̃±) = x m(t ̃) + (1−x) m(χ ̃0) i.e. the χ ̃± mass is half way between the t ̃ and χ ̃0 masses. The t ̃ decay matrix element is treated as flat, which is equivalent to the assumption of 50/50 mixing of left-handed and right-handed top super partners. The upper limits on the signal cross section are calculated separately for each signal region, incorporating the uncertainties in the signal acceptance and efficiency, using the LHC-type CLs criterion. For each point in the signal model parameter space, the observed limit is taken from the signal region with the best expected limit. The shading indicates the upper limit on the signal cross section. The observed,median expected, and ±1σ expected exclusion contours are indicated assuming NLO-NLL cross sections, as well as the observed contours when the theory cross section is varied by ±1σ (the region below the contours is excluded).
	Figure 10b : Interpretation in the top squark pair production model with t ̃ → bχ ̃± → bWχ ̃0, in the plane of m(χ ̃0) vs. m(t ̃), for the mass of the intermediate chargino χ ̃± specified by the parameter x=0.75, where m(χ ̃±) = x m(t ̃) + (1−x) m(χ ̃0) i.e. the χ ̃± mass is close in mass to the t ̃. The t ̃ decay matrix element is treated as flat, which is equivalent to the assumption of 50/50 mixing of left-handed and right-handed top super partners. The upper limits on the signal cross section are calculated separately for each signal region, incorporating the uncertainties in the signal acceptance and efficiency, using the LHC-type CLs criterion. For each point in the signal model parameter space, the observed limit is taken from the signal region with the best expected limit. The shading indicates the upper limit on the signal cross section. The observed,median expected, and ±1σ expected exclusion contours are indicated assuming NLO-NLL cross sections, as well as the observed contours when the theory cross section is varied by ±1σ (the region below the contours is excluded).

Appendix

Figure	Caption
	Figure 11 : The signal region with the best sensitivity for the t ̃ → tχ ̃0 decay, in the plane of m(χ ̃0) vs. m(t ̃).
	Figure 12 : The signal region with the best sensitivity for the t ̃ → bχ ̃± decay for x = 0.5, in the plane of m(χ ̃0) vs. m(t ̃).
	Figure 13 : The signal region with the best sensitivity for the t ̃ → bχ ̃± decay for x = 0.75, in the plane of m(χ ̃0) vs. m(t ̃).

Additional Material

Table	Caption
	Additional Table 1 : The result of the search indicating for each of the 7 signal regions the total background and observed yields, for the combination of the electron and muon channels. The uncertainty includes both the statistical and systematic components. The observed yields in the signal regions are in agreement with the predicted background.

Figure	Caption
	Additional Figure 1 : Comparison of the predicted backgrounds and observed yields for the 7 signal regions. Two expected yields from two sample signal points are also shown: t ̃ → tχ ̃0 where m(t ̃) = 300 GeV and m(χ ̃0) = 50 GeV and t ̃ → bχ ̃± → bWχ ̃0 where m(t ̃) = 350, m(χ ̃±) = x m(t ̃) +(1−x) m(χ ̃0) with x = 0.75, and m(χ ̃0) = 50 GeV. The observed yields in the signal regions are in agreement with the predicted background. The ratio of data to expected background is indicated in the top plot. The band indicates the total background uncertainty and the error bars indicate the statistical uncertainty in the data.
	Additional Figure 2 : Comparison of the dilepton tt central prediction with those using alternative MC samples for the requirements of signal region C (SRC). These include alternative generators: Powheg (default) vs. Madgraph, alternative values for the top mass, variations in the renormalization and factorization scale up and down by a factor of 2, and variations in the matching scale at which the Matrix Element partons from Madgraph are matched to Parton Shower partons from Pythia. The blue band corresponds to the total statistical error for all data and MC samples. The alternative sample predictions are indicated by the datapoints. The uncertainties on the alternative predictions correspond to the uncorrelated statistical uncertainty from the size of the alternative sample only.
	Additional Figure 3 : Same as Fig. 9 but with the cross section upper limit range extended.
	Additional Figure 4 : Same as Fig. 10a but with the cross section upper limit range extended.
	Additional Figure 5 : Same as Fig. 10b but with the cross section upper limit range extended.

Additional Material for conference speakers

When correcting for luminosity and center-of-mass energy, the ATLAS results (see this link, Figure 2) cover more of the stop -> t chi0 space than our result. The explanation for this is provided in slide format (PDF format and ppt format). Speakers presenting these results should put this slide into the backup of their talk.