CDF Logo WZ Anomalous Triple Gauge Couplings in 7.1 fb-1 of ppbar Collisions at √s=1.96 TeV CDF Logo



The goal of this study is to form limits for anomalous triple gauge couplings on the WWZ vertex. This analysis is an extension of the WZ cross section measurement using 7.1 fb^-1 of integrated luminosity presented here. The WZ final state used will be 3 e,&mu leptons and missing tranverse energy (MET) from the unobserved neutrino in lepton decay of the W. At tree level, the WZ final state has the s-channel (left) and t-channel (right) contributions shown in these diagrams:

The s-channel diagram provides sensitivity to the WWZ vertex coupling. The WZ final state is unique among the triple gauge coupling (TGC) final states, providing access to this coupling separately from the WW&gamma coupling. The analysis of triple gauge couplings in WZ production is based on analysing the Z pT distribution in the observed data events in 7.1fb-1 in the cross section times branching fraction measurement of WZ->lllnu described here. We consider the anomalous triple gauge couplings &lambdaZ, &Delta g1Z, &Delta &kappaZ as defined in PRL D60,113006 and Nucl. Phys. B282, 253 and implemented in MCFM. In the Standard Model all three of these couplings are zero. Additionally the values &Delta &kappaZ = &Delta g1Z = -1 turns off the WWZ vertex leaving only the t-channel production and a cross section that violates unitarity. To avoid this unitarity violation the WWZ vertex is modified by a form factor that implements a cut-off by multiplying the coupling with 1/(1+s/&Lambda^2)^2. We calculate the limits for two values of &Lambda = 1.5TeV and 2.0TeV.

Summary of Results

Z pT Distribution (eps)

Z pT Distribution with WWZ off (eps)

Analysis Methods

The Z pT distribution measured for the observed events in data is fitted for each of the paramaters: &lambdaZ, &Delta g1Z, &Delta &kappaZ. This is done individually with the other two couplings set to 0. The Z pT distribution is used since it is sensitive to these couplings and it can be measured experimentally.


In order to avoid CDF full simulation for every possible coupling it was shown that the efficiency at a given Z pT is the same for any given coupling. This was investigated using leading order MCFM interfaced to Pythia followed by full realistic Monte Carlo detector simulation using cdfSim and standard reconstruction. The resulting efficiency curve is then applied to MCFM next to leading order matrix element simulations of a given coupling combination to arrive at an expected observed Z pT distribution for each combination of coupling values.


A -2log(Likelihood) is then formed for a binned distribution in data to come from an expected Z pT distribution given any coupling value. Shown below are the various -2log(likelihood) distributions for the 3 couplings (&lambdaZ, &Delta g1Z, &Delta &kappaZ).

-2log(likelihood) distribution for &lambdaZ with &Lambda=1.5TeV eps

-2log(likelihood) distribution for &Delta g1Z with &Lambda=1.5TeV eps

-2log(likelihood) distribution for &Delta &kappaZ with &Lambda=1.5TeV eps

-2log(likelihood) distribution for &lambdaZ with &Lambda=2.0TeV eps

-2log(likelihood) distribution for &Delta g1Z with &Lambda=2.0TeV eps

-2log(likelihood) distribution for &Delta &kappaZ with &Lambda=2.0TeV eps

Rami Vanguri