Angular correlations in beauty production at the Tevatron at √s = 1.96 TeV

Abstract

Measurements of the b quark production cross section at the Tevatron and at Hera in the final decades of the 20th century have consistently yielded higher values than predicted by Next-to-Leading Order (NLO) QCD. This discrepancy has led to a large efforts by theorists to improve theoretical calculations of the cross sections and simulations of b quark production. As a result, the difference between theory and experiment has been much reduced. New measurements are needed to test the developments in the calculations and in event simulation. In this thesis, a measurement of angular correlations between b jets produced in the same event is presented. The angular separation between two b jets is directly sensitive to higher order contributions. In addition, the measurement does not depend strongly on fragmentation models or on the experimental luminosity and efficiency, which lead to a large uncertainty in measurements of the inclusive cross section. At the Tevatron, b$\bar{b}$ quark pairs are predominantly produced through the strong interaction. In leading order QCD, the b quarks are produced back to back in phase space. Next-to-leading order contributions involving a third particle in the final state allow production of b pairs that are very close together in phase space. The Leading Order and NLO contributions can be separated into three different processes: flavour creation, gluon splitting and flavour excitation. While the separation based on Feynman diagrams is ambiguous and the three processes are not each separately gauge invariant in NLO QCD, the distinction can be made explicitly in terms of event generators using LO matrix elements. Direct production of a b{bar b} quark pair in the hard scatter interaction is known as flavour creation. The quarks emerge nearly back to back in azimuth. In gluon splitting processes, a gluon is produced in the hard scatter interaction. The gluon subsequently splits into a b$\bar{b}$ quark pair. The quarks are very close in phase space. The flavour excitation process can be interpreted as production of a b{bar b} quark pair before the hard scatter interaction. One of the b quarks interacts with a particle from the other beam hadron and emerges with high pT. The other quark stays close to the beam axis but may still be recorded by the detector. The azimuthal correlation between the b quarks is weak. In leading order event generators, the gluon splitting and flavour excitation processes are simulated by final- and initial state showering. The b quarks produced in the proton-antiproton collisions at the Tevatron are detected through the signature of their decay products in the D0 detector. The particles associated with the production and decay of a b hadron are reconstructed as a jet in the calorimeter. These b jets are distinguished from light flavour background using two methods. The first method is based on the association of a muon with the jet. In about 20% of b hadron decays, a muon is created. Due to the large mass of the b hadron, this muon will have large transverse momentum with respect to the flight axis of the b hadron. This relative transverse momentum or P$Rel\atop{T}$ is approximated by the P$Rel\atop{T}$ of the muon with respect to the jet axis. The fraction of b jets in a muon plus jet sample can be determined by fitting the P$Rel\atop{T}$ distributions for b jets and background jets determined from Monte Carlo to the data distribution. The second method uses the relatively long lifetime of b hadrons. The tracks of the decay products of the b hadron do not point back to the production point but to the decay point of the hadron, which is displaced from the primary vertex by an average of cτγ ~ 0.5γ mm. Combined with the large mass of the hadron, this means the tracks are also displaced from the production point. By comparing the distance of shortest approach of each track to the distribution for background tracks, the probability that each track comes from a background process is determined. The probabilities of all tracks associated with a jet are combined to compute the lifetime probability for the jet to come from a background process. In this thesis, the angle between a pair of b jets is determined as the angle ΔΦ between a jet with an associated muon and a jet with a very low background lifetime probability. After selection, 1062 events remain. About 67% of all selected jet pairs are b jet pairs.