Search for heavy metastable particles decaying to quark pairs in proton anti-proton collisions at s**1/2 =1.96 TeV.


Shawn Kwang
Mel Shochet
Enrico Fermi Institute
CDF II Logo University of Chicago




Abtract

We report on an analysis searching for heavy particles that are metastable and decay into quark pairs with a macroscopic lifetime (c*tau ~ 1 cm) at a displaced vertex, which in term decay into quark pairs. We use a data driven background approach, where we build probability density functions to model Standard Model secondary vertices from known processes in order to estimate the background contribution from the Standard Model. No statistically significant excess is observed above the background. Limits on the production cross section in a Hidden Valley benchmark phenomenology are set for various Higgs boson masses as well as metastable particle masses and lifetimes.


Supporting Documents

CDF Public Note 10356
Description of the analysis
Web page of figures


Hidden Valley Model

Feynman Diagram Hidden Valley
This analysis serches for metastable particles decaying at a displaced vertex. However, a phenomenological theory is useful for a benchmark. The Hidden Valley (HV) phenomenology provides a framework in which we can generate signal MC, search for discriminants, optimize our search, and compare results.
See the Description of the analysis for diagrams of Hidden Valley decay


Signal Search Results

A counting experiment is performed using the variables developed for this search. The resulting number of expected signal MC, background estimate, and number of observed events is presented below. Two searches were performed, a low HV mass search and a high HV mass search.
The background estimate is the mean of a distribution of the number of events that pass the analysis cuts, performed manycal_dis/ (10,000) times. The uncertainties on the background estimate, as well as systematic uncertianties on the signal MC are shown the table below.
Results of the Search
Systematic Uncertainties


P-values and Limit Calcuation

We calculate the p-values for this search. No statistically significant excess is observed. Since we do not observe a statistically significant excess we proceed to set a limit on the production cross section times branching ratio of the Hidden Valley model for the particular masses and lifetimes we studied. The counting experiment was performed with a small discrete number of events, where the background estimate is less than one. Thus the expected limit can only fluctuate up (from 0). The result is that the negative sigma expected limits will be identical to the median limit.

Graphs of the limit calculations are shown below.
p-values for search
Table of calculated limits
Limit Graph 1-Low HV Mass Search
Observed and Expected limit with +1 & +2 sigma bands for signal
MC with HV mass 20 GeV/c^2.

Download: eps png svg
Limit Graph 2-High HV Mass Search
Observed and Expected limit with +1 & +2 sigma bands for signal
MC with HV mass 40 GeV/c^2.

Download: eps png svg
Limit Graph 3-High HV Mass Search
Observed and Expected limit with +1 & +2 sigma bands for signal
MC with HV mass 40 GeV/c^2.

Download: eps png svg
Limit Graph 4-High HV Mass Search
Observed and Expected limit with +1 & 2 sigma bands for signal
MC for differing HV particle lifetimes.

Download: eps png svg