Direct Sbottom Pair Production Analysis
IFAE-Barcelona
IFAE-Barcelona


Authors

Monica D'Onofrio (donofrio@fnal.gov)

Gianluca De Lorenzo (gdl@fnal.gov)

Mario Martinez (mmp@fnal.gov)

CDF, FNAL,
P.0. Box 500, M.S. 318
Batavia, Illinois 60510
USA


Search for Direct Sbottom Pair Production at CDF

We present preliminary results of the search for sbottom direct pair-production in proton-antiproton collisions with a center-of-mass energy of 1.96 TeV at the Tevatron, based on 2.65 fb-1 of data collected by the CDF detector in Run II. Events with two energetic jets and large missing transverse energy are analyzed within the framework of the minimal supesymmetric standard model (MSSM) and assuming the scalar bottom quark (sbottom) to decay into bottom quark and neutralino with a branching ratio of BR = 100%. No excess with respect to the standard model (SM) prediction is observed and limits on sbottom mass and production cross section are extracted.

  • MSSM Signal
    In MSSM, for high values of tan&beta, the large mass splitting in the third generation sector yields to low masses for the lightest sbottom state. If R-parity is conserved, sbottoms are expected to be produced in pairs via gg fusion or qqbar annihilation. The sbottom is forced to decay into bottom and the lightest neutralino (assumed to be the LSP), leading to final states which include two b-jets and substantial missing transverse energy (MET).
    For this analysis, signal samples have been generated with PYTHIA varying the values of sbottom and neutralino masses (all other MSSM parameters are kept constant). The scan is performed for sbottom masses between 80 and 280 GeV/c2 and neutralino masses between 40 up to 100 GeV/c2. The region of the MSSM phase space where [M(~chi0)+M(b)] > M (~b1) is kinematically forbidden. In the allowed region, kinematics of the signal events vary according to the values of sbottom and neutralino masses. Heavier sbottoms decay into more energetic jets while the amount of missing transverse energy is proportional to the difference between sbottom and neutralino masses.


  • SM Background
    The dominant source of background for this analysis comes from events with light-flavor jets misidentified as b-jets ('mistags'). The amount of mistags in signal regions is estimated directly form data using the standard CDF procedure. Contributions from QCD heavy flavoured multijet processes (QCD-HF) where the observed MET comes from poorly measured jets in the final state are also estimated from data. Other SM processes with a signature of b-jets and MET also include Z/γ*+jets and W+jets decays, top pair and single top events, and WW/WZ/ZZ decays. Monte Carlo simulated events are used to compute these background yields. ALPGEN+PYTHIA is used for Boson+jets, and samples are normalized to the inclusive W/Z cross section; PYTHIA is employed for Diboson and top pair production, and MADEVENT+PYTHIA for single top: samples are normalized to the corresponding NLO theoretical cross section.

    Several control samples have been defined to test the SM predictions estimated with MC-based or data-driven methods. General good agreement is found in all control regions.


  • Event selection
    Events are required to satisfy the following pre-selection criteria:

    • At least one good quality primary vertex with |Vz|<60 cm
    • Tracking activity consistent with the energy measured in the calorimeter (cleanup cuts to reject cosmics and beam halo background)
    • MET > 10 GeV
    • Only two jets per event with ET > 25 GeV and |&eta|<2.0
    • At least one of the two leading jets must be central |&eta| < 1.1
    • EM Fraction of Jet(1,2) < 0.9
    • Δφ (MET-jet1,2)> 0.4
    • No isolated high-PT tracks (lepton veto)

    The standard secondary vertex (SECVTX) b-tagging algorithm with tight-tag requirements is employed. An event is considered b-tagged if at least one of the two leading jets is tight-tag and central (|&eta| < 1.1).
    The final thresholds on jet energies and MET are set in order to maximize the signal over background separation and define the different signal regions. Two different set of cuts are defined to increase sensitivity in regions with different Delta M = M(sbottom) - M(neutralino). The thresholds for the transverse energy (ET) of the jets, the MET and the MET plus the scalar sum of the ET of the two jets (MET+HT) are quoted below for each region:

    • Low ΔM:  ETjet1 > 80 GeV ;  ETjet2 > 25 GeV ;  MET > 60 GeV ;  no cut on (MET+HT);
    • High ΔM:  ETjet1 > 90 GeV ;  ETjet2 > 40 GeV ;  MET > 80 GeV ;  (MET+HT) > 300 GeV;

    Finally, an increased threshold on Δφ (MET-jet2) from 0.4 to 0.7 is found to further improve the signal significance in both signal regions.


  • Systematic uncertainties
    • Systematics for expected SM backgrounds:

      • Different sources of systematic uncertainty have been taken into account in the determination of the SM predictions for both ΔM configurations.
      • Uncertainties on the estimation of the mistags: this translates into a 13.4% and a 11.4% uncertainty on the SM prediction for Low and High ΔM selections, respectively. It constitutes the dominant uncertainty.
      • Uncertainties on the Heavy Flavor estimation in Boson+jets samples: it varies the SM prediction of 11% for both selections.
      • Uncertainties on the estimation of tagging efficiency translate into a 3.5% uncertainty for both selections.
      • Uncertainties on the non-QCD predictions (top production, boson+jets and diboson processes) due to the variations in the modeling of the ISR/FSR in MC samples: this translates into a 6% uncertainty in the SM prediction for both selections.
      • Uncertainties on the non-QCD predictions due to PDF and renormalization scale translate into a 2.6% and 2.7% uncertainty in the SM prediction Low and High ΔM selections, respectively.
      • Uncertainties on the estimation of QCD-Heavy Flavor background is about 1.5% in the total SM prediciton.
      • A 3% uncertainty on the Jet Energy Scale translates into a 2% uncertainty on the SM yields.
      • The uncertainty on the JET trigger efficiency accounts for an additional 1% uncertainty on the predictions.

    • Systematics for MSSM signal

      • Different sources of systematic uncertainty are considered in estimation of the MSSM signal efficiency separately for each point in the sbottom-neutralino plane.
      The following sources of systematic uncertainty have been considered for the efficiency:
      • A 3% uncertainty on the Jet Energy Scale.
      • Modeling of the ISR/FSR in the Monte Carlo samples.
      • The uncertainty on the estimation of the tagging efficiency for bottom- and charm-jets.
      • The uncertainty on the trigger efficiency parameterization.

      On the theoretical cross section, the uncertainty due to PDF and to the choice of the renormalization scale in PROSPINO are considered:
      • Renormalization scale: the nominal scale &mu = M(sbottom) is shifted to μ*2 and μ/2. The average effect is about 25% for all MSSM points.
      • PDF: the uncertainty on the theoretical cross section due to the choice of CTEQ6.6 PDFs in PROSPINO have been determined by using the Hessian method. The average effect is about 10% for all MSSM points.


  • Results
    No excess with respect to Standard Model predictions is observed:

    • Low ΔM:   Observed N events = 139; SM Prediction = 133.8 ± 25.2;
    • High ΔM:   Observed N events = 38; SM Prediction = 47.6 ± 8.3;

    Systematic uncertainties on the background do not include 6 % uncertainty on the luminosity.

    The present search excludes at 95% C.L. cross sections of the order of 0.07 pb for the range of sbottom masses considered.
    Sbottom masses up to 230 GeV/c2 are excluded at 95% C.L. for neutralino masses in the range 40-80 GeV/c2 .
    This analysis thus improves previous CDF results by ~ 40 GeV/c2.

  • Public Note In preparation.


  • EXCLUSION PLANE

    • Exclusion Plane (Sbottom Mass vs Neutralino Mass) gif & eps
      Exclusion Plane for sbottom and neutralino masses at 95% C.L. The solid red line shows the area excluded by the current analysis with 2.65 fb-1 of CDF Run II data. The dashed red line shows the expected limit. The PDF and the renormalization scale uncertainties on the theoretical cross section are included in the limit calculation. The dashed black line denotes the kinematically forbidden region, M(b1~) < M(~chi0)+M(b). The grey line denotes the LEP limit. For comparison, expected and observed limits of previously published CDF and D0 results are reported.




  • PROSPINO CROSS SECTIONS VS OBSERVED/EXPECTED 95% C.L. LIMITS

    • Cross section as a function of Sbottom Mass (Neutralino Mass = 70 GeV/c2) gif & eps
      PROSPINO NLO Cross section as a function of the sbottom mass for a neutralino mass of 70 GeV/c2. The yellow band denotes the systematic uncertainty on the theoretical predictions. Solid and dashes lines denote the observed and expected 95 % C.L.




  • MISSING ET, ET(jet1, 2) AND MET+HT DISTRIBUTIONS WITH FINAL CUTS

    • Low ΔM signal region
      ETjet1 (gif & eps), ETjet2 (gif & eps), Missing ET (gif & eps), Missing ET+HT (gif & eps) distributions in the Low ΔM signal region. The data are compared to the SM predictions including the total systematic uncertainties. SM background predictions are also shown separately for QCD (heavy flavor), non-QCD (heavy flavor) and misidentified heavy flavor. For illustrations, the signal for an MSSM point is shown.

    • High ΔM signal region
      ETjet1 (gif & eps), ETjet2 (gif & eps), Missing ET (gif & eps), Missing ET+HT (gif & eps) distributions in the High ΔM signal region. The data are compared to the SM predictions including the total systematic uncertainties. SM background predictions are also shown separately for QCD (heavy flavor), non-QCD (heavy flavor) and misidentified heavy flavor. For illustrations, the signal for an MSSM point is shown.

            Monica D'Onofrio (donofrio@fnal.gov)