Search for top squarks in multijet events with large missing momentum in proton-proton collisions at 8 TeV

Link to PAS-SUS-13-015

Abstract:

A search for pairs of top squarks decaying to a top quark and a stable, weakly interacting, massive particle is presented. The search selects events containing multiple jets, with at least one identified as originating from a b-quark, and large missing transverse momentum. The search also employs a novel top quark tagging algorithm for identifying a top quark candidate decaying hadronically. It is based on 19.4 fb^-1 of proton-proton collision data collected at a center-of-mass energy of 8 TeV by the CMS detector at the Large Hadron Collider. The observed event yields are consistent with the SM expectations. Assuming both top squarks decay always into a top quark and a weakly interacting, neutral particle, the production of top squarks with mass less than 535 GeV is excluded at 95% confidence-level for small LSP masses less than 10 GeV.

Results:

Plots in the PAS

Figure/Table	Links	Description
	pdf, png	Figure 1: Top squark pair production with the top squark decaying into a top quark and neutralino.
	(a) pdf, png, (b) pdf, png	Figure 2: The 2D distributions of m23/m123 versus arctan(m13/m12) for ttbar (left) and QCD (right). The three horizontal solid red lines which form a band demonstrate the criteria in Eq. (1.a) with the central line taken as the nominal ratio of mW/mtop and the other two lines showing the boundaries defined by the Rmin and Rmax. Similarly the criteria in Eqs. (1.b) and (1.c) are in groups of solid lines. The dashed line indicates the requirement of m23/m123 > 0.35 as in Eqs. (1.b) and (1.c). All distributions are normalized by the total number of events and the label "a.u." means arbitrary units.
	(a) pdf, png, (b) pdf, png, (c) pdf, png, (d) pdf, png	Figure 3: The pTmiss, MT2, MTRsys, and MT3jet distributions for ttbar and three signal points after the pre-selection and a pTmiss ≥ 200 GeV requirement. All distributions are normalized by the total number of events and the label "a.u." means arbitrary units. The signal points are labeled as "(X, Y)" where X is the top squark mass and Y is the neutralino mass.
	(a) pdf, png, (b) pdf, png, (c) pdf, png, (d) pdf, png	Figure 4: The 2D distributions of MT3jet versus MTRsys for ttbar and three signal points after the pre-selection and a pTmiss >200 GeV requirement. All distributions are normalized by the total number of events and the label "a.u." means arbitrary units. The red line indicates the cut threshold. The signal points are labeled as "(X, Y)" where X is the top squark mass and Y is the neutralino mass.
	(a) pdf, png, (b) pdf, png, (c) pdf, png	Figure 5: The pTmiss, MT2, and 0.5MT3jet+MTRsys distributions from observed data (black dots), predicted backgrounds (solid filled areas) and two selective signal points after applying the baseline selection cuts. The signal yields are scaled to 19.4 fb-1. The QCD prediction is not included in the plots since its contribution in the baseline search region is negligible. The dashed area indicates the total uncertainties including both statistical and systematic uncertaintes for each bin. The last bin contains all the events above the maximum values of pTmiss, MT2, and 0.5MT3jet+MTRsys in the figures, respectively.
	pdf, png	Table 1: Event yields and cut flow from MC simulated samples after various requirements in the pre-selection and kinematic cuts from the top quark reconstruction. All numbers are scaled to 19.4 fb-1. Only statistical uncertainties are shown in the table. The signal points are labeled as "(X, Y)" where X is the top squark mass and Y is the neutralino mass. Data yields are omitted as the trigger is inefficient before the offline pTmiss cut at 200 GeV. The number of QCD multijet events is not included in the sum total because QCD multijet background overwhelms all the other backgrounds which renders the sum total uninformative.
	pdf, png	Table 2: Event yields and cut flow continued from Table 1 for the four signal regions. All numbers are scaled to 19.4 fb-1. Only statistical uncertainties are shown in the table. The signal points are labeled as "(X, Y)" where X is the top squark mass and Y is the neutralino mass. The number of QCD multijet events is not included in the sum total because QCD is negligible in all four signal regions.
	pdf, png	Table 3: Predicted hadronic tau background corresponding to 19.4 fb-1 from the methods presented here for all four search regions.
	(a) pdf, png, (b) pdf, png	Figure 6: Comparison of predicted distributions of pTmiss (left) and MT2 (right) with the truth in MC simulated ttbar and W events containing a hadronically decaying tau lepton. Only the statistical uncertainties are shown on the predicted distributions.
	pdf, png	Table 4: Lost lepton background prediction results for 19.4 fb-1.
	(a) pdf, png, (b) pdf, png	Figure 7: Predicted distributions of pTmiss (left) and MT2 (right) compared to the truth in the ttbar and W MC simulated events containing an undetected lepton (e or μ).
	pdf, png	Table 5: Prediction for the Z to invisible background for the various signal regions.
	pdf, png	Table 6: QCD multijet background prediction for each of the four search regions.
	pdf, png	Table 7: Predicted event yields from the different background estimation methods for the search regions. The uncertainties on various backgrounds are added in quadrature for the uncertainty on the total background estimation.
	pdf, png, C	Figure 8: Combined 95% C.L. exclusion limits for the top squark pair production (signal topology). The plot shows the expected limit in red dashed line. The observed limit is shown in black solid line. The thinner dashed lines around the expected limit shows the variation in the expected limit, while the thinner solid curves show the variation in the observed limit when the signal cross section is varied by its theoretical uncertainties.

Additional materials

Figure/Table	Links	Description
	pdf, png	Figure 9: 95% C.L. exclusion observed limits with different branching ratio assumptions of top squark to top+neutralino decay. The results are based on the assumption that the search has no acceptance for top-squark pair events if one of the top squarks decays in a different mode.
	pdf, png	Figure 10: The signal region with the best limit for each signal point.
	(a) pdf, png, C (b) pdf, png, C (c) pdf, png, C (d) pdf, png, C	Figure 11: Signal efficiencies for the four search regions.