CDF Logo

A Limit on the Branching Ratio of the Flavor-Changing Top Quark Decay t→Z0c



  • Authors: AlexanderParamonov and Henry J. Frisch
  • Blessed: April 03, 2008
  • Data: Run II, 1.52  fb-1
  • Documentation:  CDF Note 9101 (internal only)
  • Public Note: CDF Public Note 9285
  • Large-format Poster: PDF

  • Abstract.

    We measure an upper limit of 8.3% (95% C.L.) on the branching ratio of the flavor-changing  top quark decay t→Z0c for 100% longitudinally polarized Z-bosons using 1.52 fb−1 of data. We parametrize the upper limit as a function of Z-boson’s helicity to cover the full range of possible decay structures. The analysis is based on the comparison of two processes: pp→tt→WbWb→ l+mET+bbjj and pp→tt→Z0cWb→l+lcjjb. The use of these two decay modes together allows cancellation of major systematic uncertainties of acceptance, efficiency, and luminosity. We validate the MC modeling of acceptance and efficiency for lepton identification over the multi-year dataset by a precision measurement of the ratio of the inclusive production of W- and Z-bosons. To improve the discrimination, we calculate the top mass for each event with two leptons and four jets assuming it is a tt event with one of the top quarks decaying to Z0c. The calculation of the top mass is performed with a fitter on an event-by-event basis. The upper limit on the Br(t→Z0c) is estimated from a 2-dimensional likelihood of the l+lcjjb top mass distribution and the number of “l + mET + 4jets” events. The results are limited by statistics at present.



    1. Analysis overview.

    The analysis is based on comparison of two final states:
    Simultaneous consideration of the two final states allows cancelation of some systematic uncertainties. All the possible structures of the unknown FCNC decay t→Z0c  are parametrized using the fraction of longitudinally polarized Z-bosons in t→Z0c decays.

    The analysis of the two final states results in measurement of top-pair production cross-section and a limit on Br(t→Z0c). We focus on the measurement of the limit on Br(t→Z0c) from now on. The statistical machinery is done using the Bayesian approach.


    2. Event Selection

    We select events using high-Pt lepton trigger (electrons or muons), for the tight lepton we regure Pt > 20 GeV. Each selected event is required to have a loose B-tag (a displaced secondary vertex consisten with a decay of heavy flavour quark). The Z-bosons are identified via their delopton decays by asking for an additional loose (Pt > 12 GeV) lepton of the same flavor and opposite charge. The invariant mass of the dilepton pair should be between 66 and 116 GeV. The W-bosons are reconstructed  in events with large missing Energy (mET) where the transverse mass of the lepton and the missing energy is greater that 20 GeV.


    2.1 Validation of the Monte Carlo simulations with Inclusive W and Z bosons

    We used inclusuve W- and Z- bosons to test presicion of Monte Carlo simulations in modeling of electrons and muons. The comparison of the invariant and transverse masses of the inclusive Z and W bosons, respectively, are shown below. In addition we estimate the R-ration between the production cross-sections of the W- to the Z- bosons. The ratio agrees with the NNLO predictions within 2%.

    M_W,inclusive
    [png][eps][pdf]

    M_W,inclusive
    [png][eps][pdf]

    M_Z,inclusive
    [png][eps][pdf]
    M_Z,inclusive
    [png][eps][pdf]


    2.2 Final state with  a W boson and four jets with at least one B-tag

    We select event with a  final state containing a leptonic decay of a W-boson and at least one b-tagged jet. The limit measurement relies on event with four jets (at least one of the jets should have a b-tag) and a W. The four-jet events have high contribution from decays of the tt-bar pairs.


    We, Njets
    [png][eps][pdf]

    Wm, Njets
    [png][eps][pdf]

    We, Ht (4j)
    [png][eps][pdf]


    Wm, Ht (4j)
    [png][eps][pdf]



    2.3 Final state with a Z boson and four jets with at least one B-tag

    Events with a leptonic decay of Z bosons and at least one b-tagged jets are of our interest. Consequesntly we select events with four jets in the final state. The jet multiplicity distributions are normalized to the 2nd jet bin.

    Ze, Njets
    [png][eps][pdf]

    Zm, Njets
    [png][eps][pdf]


    2.3.1 Top mas fitting

    Additional discrimination against standard mdel backgrounds can be achived by reconstructing invariant mass of the top quark in events with a Z-boson and four jets. The reconstructed topmass information is used in the limit-setting machinery.

    M_top
    [png][eps][pdf]

    M_top
    [png][eps][pdf]

    Chi^2
    [png][eps][pdf]
    Chi^2
    [png][eps][pdf]




    3. Statistical procedure of calculation of the limit on Br(t→Z0c)

    The Bayesian approach is used to analyze events in two final states: "Z + 4jets" and "W=+4 jets". Top-production cross-section and Br(t→Z0c) are treated as independent variables.

    2d likelihood
    [png][eps][pdf]

    Post Distr.
    [png][eps][pdf]


    4. Results and conclusions

    The analysis is performed in two stages:

    First we validate reconstruction of inclusive W- and Z-bosons that play a crucial role in the study. This test lepton ID, acceptances, and analysis code for both data and Monte Carlo simulations over multi-year set of runs.

    The second stage if the actual measurement of the upper limit on Br(t→Z0c). To comput the limit on Br(t→Z0c) we apply Bayesian approach for a 2-diminsional likelihood distribution of Br(t→Z0c) and number ot top-antitop pairs produced (cross-section times luminosity). To be assumption-independent of the preffered FCNC coupling we parametrize the limit as a function of longitudinal polarization of the Z-bosons. This parametrization allows us to cover the full range of possible helicity structures of the FCNC coupling. The results obtained with 1.52 fb−1 of data are presented in the table below.

    Fraction of Longitudinally Polarized Z-bosons
    0.00
    0.25
    0.50
    0.75
    1.0
    95% C.L. limit on Br(t→Z0c) using theoretical pp→tt cross-sectionas a prior
    9.0%
    8.8%
    8.6%
    8.5%
    8.3%
    95% C.L. limit on Br(t→Z0c) using flat prior
    10.2%
    10.0%
    9.7%
    9.5%
    9.2%


    Page was last updated July 02, 2008 by Alexander Paramonov (paramon _at_ hep.uchicago.edu).
    Contact, Legal Notices, and Fermilab Policy on Computing