Recent progress in theoretical methodologies and computational capabilities make highly-predictive, thermochemical kinetics computations possible. The foundation for these calculations is reliable fundamental properties for the reactants, products, and intermediates in a system, obtained with quantum chemistry methods. This dissertation provides an overview on common approaches to characterize a reaction system and demonstrates a high-level approach to quantum mechanics calculations through a study of the formylperoxy radical and to kinetics computations through addition reactions of stabilized Criegee intermediates. Carrying out these intensive protocols, however, is not manually feasible for a full combustion simulation. We describe Autochem: a large-scale, automated, combustion chemistry modeling software package that we have been developing over the last three years. It accurately predicts the thermochemical properties and temperature and pressure dependence of the thousands of gas phase reactions in a combustion mechanism by providing a unified interface for electronic structure and transition state theories, mechanism generation, classical trajectory simulations, the master equation, and uncertainty quantification methods. This dissertation showcases two studies that employ Autochem. The first uses the software to model branching ratios during the pyrolysis of arbitrary fuels. The second focuses on thermochemistry of core combustion species. It benchmarks common electronic structure methods and several high-level treatments of a partition function.
