Total Width of the top quark using the template method in the lepton+jets channel

J. Adelman, E. Brubaker, Y.K. Kim, S. Shiraishi (U. Chicago)

    About our analysis
  1. Abstract
  2. Introduction
  3. Event Selection
  4. Monte Carlo Samples
    Monte Carlo Studies
  1. Phase I analysis: Studying the kinematic fitter with and without Mtop=Manti-top constraint
  2. Phase II analysis: Further analysis using kinematic fitter using more MC samples.
    1. Top Mass Distributions
    2. Lorentzian+Gamma Fit Method
    3. 1-D Likelihood Method
    4. Feldman-Cousins
    5. Systematic Uncertainties


We present an upper limit of the total width of the top quark, thus the lower limit on the top quark lifetime in the lepton plus jet channel. The data sample corresponds to an integrated luminosity of 955pb-1 and a total of 253 event is observed. In each t-tbar event, the reconstructed top quark invariant mass is determined by minimizing a &chi2 for the overconstrained kinematic system, and the reconstructed mass distribution is produced for 2-tag and 1-tag Tight samples. The top quark width is extracted using a likelihood fit to the reconstructed mass distribution of the data and template distributions which incorporates top quark events and expected backgrounds, and varies with the top quark width. We set an upper limit on the total top quark width of &Gammatop < 12.4 GeV at 95% confidence level. This corresponds to a lower limit on the top quark lifetime of &tautop > 5.3 x 10-26s at 95% confidence level.


The Standard Model predicts the lifetime of the top quark to be around 10^{-24} s. However, precise measurements of such a small time scale are difficult, and current upper limit on the lifetime is 1.75X10-15 s at 95% CL. Instead of measuring the lifetime directly, we set a limit on the width and inversely relate it to the lifetime. In the future, we will use a sample of ttbar decays corresponding to 1 fb^{-1} of proton-antiproton collisions at Sqrt{s}=1.96 TeV and collected using the CDF-II detector for our analysis.
In our analysis, we study Monte Carlo samples (MC samples) generated by Pythia for 14 different input top widths between 0-150 GeV. We have also included MC samples corresponding to $W$ boson production in association with bb-bar and two additional partons (Wbb2P) to represent the dominant background events. After the Monte Carlo decay simulations, the top mass are reconstructed based on a kinematic calculation using the template method. This template methd is a standard kinematic chi2 minimalization method used in the top mass analysis. Since the reconstructed top mass distributions are sensitive to the top width, we have studied sensitivities of a number of variables such as mean and RMS of the distributions as well as variables from the fit to the distributions. So far, our study suggests one dimensional likelihood fit to the reconstructed top mass distributions is the most sensitive to the width of the top quark.

Event Selection

In our analysis, we select events that are consistent with the production of ttbar paris that is followed by the decay of each top quark to a W boson and a b quark, the hadronic decay of one W boson, and the leptonic decay of the other. These lepton plus jets events are identified by requiring an electron or muon candidate, large missing transverse energy of greater than 20 GeV corresponding to the neutrino from the W decay, and at least four jets in the final state. These events are then separated based on the number of jets that are b-tagged using the tight SECVTX algorithm based on the identification of secondary vertices inside jets. We select events with 2 or more b-tagged jets (2-tag events) and one b-tagged jet. At the level-4 correction, we require the 2-tag events to have 3 jets with E_{T} > 15 GeV and a jet with E_{T} > 8 GeV, and we further require the 1-tag events to have at least four jets with E_{T} > 15 GeV (1-tag tight).

Monte Carlo Studies

The data sample collected in the CDF detector from March 2002 to February 2006 will be used in this analysis. In this data sample, we expect 37 background events and observe a total of 255, after the event selection cuts. A majority of the background events come from W boson production associated with heavy quark jets or a light quark jet misidentified as a b quark jet. Each sample contains roughly 1M ttbar signal events, resulting in about 400K lepton plus jets events.