We measure the cross section for producing a photon and a b-jet at the Tevatron in the
Assuming that the photon and the jet (required to have a separation of at least 0.7 in η-Φ) are independent, we can determine the trigger efficiency as the ratio of events passing the SVT track requirements with respect to all those passing the unbiased inclusive photon trigger.
This efficiency is shown in the following plot as a function of jet Et:
Photons in the barrel region are identified as electromagnetic clusters with no nearby associated track compatible in Pt, with a proper shower shape and small hadronic energy deposition. The photon identification efficiency is extracted from Monte Carlo, and is the product of the efficiency of finding an electromagnetic cluster close to the position of the generated photon (Electromagnetic cluster reconstruction in the plot), times the efficiency of that cluster to pass the photon requirements (photon identification in the plot). The latter has two steps at 18 and 35 GeV since the cuts change at these energies. The two efficiencies are shown in the following figure:
In a hadron collider environment numerous neutral pions are present and can possibly fake photons. CDF has installed a preshower detector before the calorimeter, and a "shower maximum" detector at about one third of the electromagnetic calorimeter depth. For each shower, the measured distribution is compared to test-beam shapes for photons and pions, and each event is given a weight corresponding to the probability of actually being a photon. Due to the large background, the photon purity is about 30% for low photon energies, and grow to around 60% at higher energies:
Jets with minimal Et of 20 GeV are tagged using the standard CDF b-tagging algorithm. The efficiency of this algorithm for b-jets is obtained from the Monte Carlo and is corrected for a scale factor derived from the comparison between data and simulation for jets which contain leptons. The corrected efficiency is shown in the following figure as a function of jet Et:
The b-purity of the tagged sample is extracted directly from data, fitting the invariant mass of the tracks composing the secondary vertex. In an ideal world where all decay products of the decaying quarks (including neutrals) could be reconstructed, the distributions of the secondary vertex masses would peak at the values of the quark masses themselves. Even with real-world reconstruction, the mass of the secondary vertex is a powerful discriminator. The next figure shows the distribution of reconstructed secondary vertex masses in data (black histogram with error bars) as well as the fitted contributions from light, charm and b quark templates.
The main systematic uncertainties for this measurement come from the luminosity determination (6%), the determination of the trigger efficiency (10%), and from the effect of the limited knowledge of the tracking reconstruction efficiency on the fitted b-purity (15-20%).
We measure a total cross section of σ(γ + b-jet) = 90.5 ± 6.0 (stat.) +21.7 -15.4 (syst.) pb for
The differential cross section as a function of photon transverse energy is shown in the next figure. The error bars represent the statistical errors, while the systematic uncertainties are represented by a light blue band.
The differential cross section as a function of jet transverse energy is shown in the following figure:
