Top Quark Width in Lepton+Jets Channel Using Template Method

Summer 2007: 955 pb^-1

S. Shiraishi, J. Adelman, E. Brubaker, Y.K. Kim

University of Chicago

CDF Collaboration

We present the first direct experimental bound on the total decay width of the top quark, Γ_t, using 955 pb^-1 of the Tevatron's ppbar collisions recorded by the Collider Detector at Fermilab. We identify 253 top-antitop pair candidate events. The distribution of reconstructed top quark mass from these events is fitted to templates representing different values of the top quark width. Using a confidence interval based on likelihood-ratio ordering, we extract an upper limit at 95% C.L. of Γ_t <13.1 GeV for an assumed top quark mass of 175 GeV/c².

Description	Figures
Feldman-Cousins confidence bands with stastical and systematic uncertainties incorporated.
Combined likelihood for 1-tag and 2-tag samples in the data.
The reconstructed top mass distributions from the data and the results of the likelihood fits for 1-tag and 2-tag samples.
The parameterized signal m_t^reco distributions and goodness of the parameterization are shown for the 2-tag subsample at three different values of Γ_t.
Top mass distributions at the generator-level for the samples of Γ_top^input = 1.5, 30, and 50 GeV.
Top mass distributions at the generator-level with events that passed the event selection cuts after the full detector simulation. The samples are for Γ_top^input =1.5, 30, and 50 GeV.
Reconstructed top mass distributions for 1-tag samples for Γ_top^input =1.5, 30, and 50 GeV.
Reconstructed top mass distributions for 2-tag samples for Γ_top^input =1.5, 30, and 50 GeV.
Parameterization of 1-tag reconstructed top mass for the likelihood fit. The fit function is two Gaussian and a Gamma PDF whose parameters are quadratic functions of Γ_top^input.
Parameterization of 2-tag reconstructed top mass for the likelihood fit. The fit function is two Gaussian and a Gamma PDF whose parameters are quadratic functions of Γ_top^input.
Systematic uncertainties on the likelihood.
Top width distributions from 3000 pseudoexperiments for Γ_top^input =1.5, 5, 15, 30, and 50 GeV.
Coverage for all input top width samples.
Reconstructed top mass distribution for 1-tag sample in the data.
Reconstructed top mass distribution for 2-tag sample in the data.
Likelihood fit to 1-tag sample in the data.
Likelihood fit to 2-tag sample in the data.
Results of pseudoexperiments assuming the Standard Model (Γ_top^input = 1.5 GeV). Roughly 25% of the time, we set an upper limit lower than 12.7 GeV.

Description

Figures

Feldman-Cousins confidence bands with stastical and systematic uncertainties incorporated.

Feldman-Cousins confidence bands including systematics

Combined likelihood for 1-tag and 2-tag samples in the data.

The reconstructed top mass distributions from the data and the results of the likelihood fits for 1-tag and 2-tag samples.

The parameterized signal m_t^reco distributions and goodness of the parameterization are shown for the 2-tag subsample at three different values of Γ_t.

Top mass distributions at the generator-level for the samples of Γ_top^input = 1.5, 30, and 50 GeV.

Top mass distributions at the generator-level with events that passed the event selection cuts after the full detector simulation. The samples are for Γ_top^input =1.5, 30, and 50 GeV.

Generator-level distributions after selection

Reconstructed top mass distributions for 1-tag samples for Γ_top^input =1.5, 30, and 50 GeV.

Reconstructed top mass distributions (1-tag)

Reconstructed top mass distributions for 2-tag samples for Γ_top^input =1.5, 30, and 50 GeV.

Reconstructed top mass distributions (2-tag)

Parameterization of 1-tag reconstructed top mass for the likelihood fit. The fit function is two Gaussian and a Gamma PDF whose parameters are quadratic functions of Γ_top^input.

Parameterization of 2-tag reconstructed top mass for the likelihood fit. The fit function is two Gaussian and a Gamma PDF whose parameters are quadratic functions of Γ_top^input.