The goal of this work is to model the heterogeneous recombination of atomic oxygen on silica surfaces, which is of interest for accurately predicting the heating on vehicles traveling at hypersonic speeds. This is accomplished by creating a finite rate catalytic model, which describes recombination with a set of elementary gas-surface reactions. Fundamental to a description of surface catalytic reactions are the in situ chemical structures on the surface where recombination can occur. Using molecular dynamics simulations with the ReaxFFSiOGSI potential, we find that the chemical sites active in direct gas-phase reactions on silica surfaces consist of a small number of specific structures (or defects). The existence of these defects on real silica surfaces is supported by experimental results and the structure and energetics of these defects have been verified with quantum chemical calculations. The reactions in the finite rate catalytic model are based on the interaction of molecular and atomic oxygen with these defects. Trajectory calculations are used to find the parameters in the forward rate equations, while a combination of detailed balance and transition state theory are used to find the parameters in the reverse rate equations. The rate model predicts that the oxygen recombination coefficient is relatively constant at T (300-1000 K), in agreement with experimental results. At T > 1000 K the rate model predicts a drop off in the oxygen recombination coefficient, in disagreement with experimental results, which predict that the oxygen recombination coefficient increases with temperature. A discussion of the possible reasons for this disagreement, including non-adiabatic collision dynamics, variable surface site concentrations, and additional recombination mechanisms is presented. This thesis also describes atomistic simulations with Classical Trajectory Calculation Direction Simulation Monte Carlo (CTC-DSMC), a particle based method for modeling non-equilibrium rarefied gas flows that employs trajectory calculations to determine the outcome of molecular collisions. We compare CTC-DSMC to direct molecular dynamics calculations for one-dimensional shocks, where exact agreement between the two methods is demonstrated. We also discuss a number of topics important in CTC-DSMC simulations, including GPU enabled acceleration, a preliminary algorithm for modeling three-body collisions, and characterizing high temperature rovibrational effects.