Search for singly produced new heavy quarks at CDF
 
Florencia Canelli1,2, Martina Hurwitz1

1University of Chicago
2Fermilab

Public note

  


  • Theory
  • Baseline event selection
  • Analysis method
  • Modeling validation
  • Systematic uncertainties
  • Results


    • Download plots in EPS format by clicking on the plot.
    To view GIF format with full resolution, right-click and select "View Image."


    Theory

    We search for single production of new heavy quarks via a charged current interaction. We assume B.R.(Q → Wq)=100%, where q is a first generation SM quark. We follow the framework of Heavy Quarks Above the Top at the Tevatron (A. Atre, M. Carena, T. Han, J. Santiago), which gives a model-independent parameterization of the interaction of new quarks with the SM.

    We focus on the production of D or U-type heavy quarks produced via charged current interactions and decaying 100% of the time to a W boson and a light flavor SM quark. The leading-order production / decay of these quarks is shown below.

    s-channel (left) and t-channel (right) production of D-type heavy quark and subsequent decay to W and SM quark.

    The production cross section of the new quarks can be factorized: σ(qQ)=\tilde{κ}qQσSM, where \tilde{κ}qQ is a model-dependent coupling between the new quark and an SM quark. σSM is the production cross section of the new heavy quark assuming SM couplings adjusted by a factor of v/mQ, where mQ is the new heavy quark mass. We will derive limits on \tilde{κ} and σ(qQ).

    We generate signal events using MadGraph with a Pythia parton shower and use the CDF detector simulation to evaluate the expected kinematics. We use a baseline event selection (described below) and compare the expected signal kinematics to those of the backgrounds. Some relevant examples are shown below.

    Expected kinematics of signal and background processes.

    Due to the decay of the heavy quark, the leading jet and the W boson are boosted with respect to the background. The Δφ between the lepton and the missing transverse energy is smaller for the signal than for the W plus jets background. The second jet is produced at a higher η in the signal than in the background.

    Event Selection

    We define a baseline event selection designed to select events consistent with a W plus two jet topology. In the baseline event selection, we require a single, isolated electron or muon with transverse momentum greater than 20 GeV/c, missing transverse energy greater than 20 GeV (25 GeV in events with an electron), and exactly two jets. The leading jet in the event is required to have transverse energy larger than 25 GeV, while the second jet must have transverse energy larger than 20 GeV. Both jets must be central with |η| less than 2.0.

    We veto events with additional leptons or jets or with a second loose lepton consistent with Z decay. The background due to QCD multi-jet events (non-W) is suppressed by requiring the transverse mass of the W candidate to be larger than 10 GeV (20 GeV in events with an electron) and by placing additional cuts on the angles between the missing transverse energy and the leptons and jets.

    Events from various standard model processes will pass this baseline event selection. We calculate the number of expected events from diboson, Z plus jets, top pair, and single top processes using known cross sections, acceptances from the Monte Carlo simulation, and the known integrated luminosity. The contribution from QCD non-W and W plus jets events is derived by fitting the missing transverse energy distribution observed in data with predicted shapes. The fit to the missing transverse energy is performed separately in the four different lepton categories we use. The fits to the missing transverse energy and the expected composition of the sample are shown in the figures below.

    Fits to the missing transverse energy for central electrons (top left), central and extended muons (top right and bottom left), and extended muons from a MET+jets trigger (bottom right).

    The Z and W plus jets backgrounds are modeled using the Alpgen event generator and Pythia parton shower; the top pair and diboson processes are modeled with Pythia; and the single top background is modeled with MadGraph plus Pythia.

    Analysis Method

    Some additional cuts are imposed to isolate a region with enhanced signal contribution. In particular, we require the transverse energy of the leading jet to be larger than 60 GeV, the missing transverse energy to be larger than 30 GeV, and the Δφ between the missing transverse energy and the lepton to be less than 2.2. The expected numbers of events are shown below.

    CDF Run II Preliminary, L=5.7 fb-1
    Expected contributions of Standard Model processes
    Process Expected number of events
    W+jets 10648 ± 319
    Diboson 599 ± 60
    non-W 581 ± 232
    Z+jets 497 ± 75
    Top pair 355 ± 43
    Single top 149 ± 18
    Total 12829 ± 409
    Data 13243
    Expected contributions of Standard Model processes to the signal region and number of events observed in data

    CDF Run II Preliminary, L=5.7-1
    D-type quark signal with \tilde{κ}uD=1
    Heavy quark mass [GeV/c2] Expected number of events
    300 1275 ± 115
    400 293 ± 26
    500 81 ± 7
    600 20 ± 2
    Expected contributions of heavy D-type quark when coupling to SM u quark is 1.

    CDF Run II Preliminary, L=5.7-1
    U-type quark signal with \tilde{κ}dU=1
    Heavy quark mass [GeV/c2] Expected number of events
    300 302 ± 27
    400 54 ± 5
    500 12 ± 1
    600 2 ± 0.2
    Expected contributions of heavy U-type quark when coupling to SM d quark is 1.

    The presence (or lack) of signal is evaluated by fitting the invariant mass of the leading jet, the lepton, and the neutrino to templates in this variable derived for signal and background processes.. The transverse momentum components of the neutrino are taken from the measured missing transverse energy, while the z component of its momentum is taken assuming the invariant mass of the lepton and neutrino must be the W mass. The templates of Mlνj used in the fit are shown below.
    Mlνj templates for backgrounds (left) and signals of different quark masses (right).

    Validation

    We reverse the selection criteria described above to create several control regions:

    • Low-ET control region: ET(leading jet) < 60 GeV
    • High-Δφ control region: Δφ(MET, lepton) > 2.2 and ET(leading jet) > 60 GeV
    • One-jet control region: All of the same cuts as in the signal region, but no second jet

    We check the modeling of the invariant mass of the lepton, neutrino, and leading jet in each of these three control regions. The plots below show the sum of the backgrounds in a stack plot with the data superimposed.

    Modeling of the invariant mass of the lepton, neutrino, and jet in the three control regions

    In addition, we check the modeling of variables that are relevant for Mlνj in the signal region.

    Modeling of jet ET, jet η, lepton pT, lepton η, missing transverse energy, and the Δφ between the objects in the signal region.

    Systematic uncertainties

    We take into account systematic uncertainties from several sources when performing the signal extraction fit. We evaluate the effect of uncertainties on the rates of signal and background processes as well as the shape of the Mlνj template.

    The rate uncertainties are listed in the table below.

    CDF Run II Preliminary
    Rate uncertainties
    Source of uncertainty Process affected Size of uncertainty
    Jet Energy Scale Signals 2%
    Initial / final state radiation Signals 2%
    Parton distribution functions Signals 2%
    NLO cross section contribution Signals 5%
    Background normalizations W+jets 20%
    Diboson 10%
    QCD non-W 40%
    Top pair and single top 12%
    Z+jets 15%
    Integrated luminosity All except W+jets and non-W 6%
    Trigger and ID efficiencies All except W+jets and non-W 2%
    Uncertainties on normalizations of signal and background processes

    Uncertainties on the shape of the Mlνj templates are defined due to the JES (evaluated for the signal and the W+jets background) and the factorizaton and renormalization (Q2) scale in Alpgen (evaluated for the W+jets background). The change in the templates shapes are shown below.

    Systematic shape uncertainties on the templates: uncertainty in signal template for a 400 GeV heavy quark due to the JES (left), uncertainty in the template for the W+jets background due to the JES (centerl), and uncertainty in the template for the W+jets background due to the Q2 scale (right).

    Finally, we define two uncertainties associated with mismodeling observed in control regions. The pT of the W boson is observed to be mismodeled in the 1-jet control region, and the ΔR between the two jets is mismodeled in all control regions, with the largest mismodeling observed in the low-ET control region. The W+jets model is reweighted to agree with the data in these variables, and the W+jets template is rederived with these weights. The change in the template is defined as a systematic shape uncertainty on the W+jets background. The mismodeling and resulting shape systematics are shown below.

    Mismodeling observed in control regions and corresponding shape uncertainties on the W+jets background template.

    Results

    A maximum likelihood fit of the templates to the Mlνj distribution observed in the signal region in data is carried out. No significant excess is observed, so we set limits on the cross section of single production of the heavy quarks. We translate the limit on the production cross section into limits on the couplings between the SM and heavy quarks (limits on \tilde{κ}qQ2). The Mlνj observed in data and the resulting limits are shown below.

    Mlνj observed in data, and the resulting limits on the cross section of heavy quark production and the coupling between heavy quarks and SM quarks as a function of mass.

    CDF Run II Preliminary, L=5.7 fb-1
    Limits on new heavy quark production
    MQ [GeV/c2] Limit on σ(qQ) [pb] Limit on \tilde{κ}uD2 Limit on \tilde{κ}dU2
    300 2.4 0.34 1.5
    400 0.67 0.45 2.4
    500 0.27 0.74 5.4
    600 0.18 1.9 15