Automatization of theoretical kinetic data generation for tabulated TS models building - Part 1: Application to 1,3-H-shift reactions

Estimation of kinetic parameters is a key aspect of chemical combustion modeling and several approaches were developed to approximate unknown data. In this work, a code in Python was developed to build tables of transition state (TS) models automatically for intramolecular H-shift reactions in alkyl radicals. The code generates the kinetic rules for all the possible combinations of methyl-substituted reactions based on the structures of the minimal, non-substituted, reactant, TS, and product. The code is able to create and differentiate multiple transition state configurations, considering the axial and equatorial positions for the cyclic substituents and including all the possible pathways for the reactions, which is shown to be an important feature in performing accurate automatic kinetic calculations. Each structure is automatically submitted to geometry optimization and electronic energy calculations, as well as the relaxed scans of the torsional modes identified by the code. From the results of electronic calculations, the rate constants for each pathway are obtained automatically by the application of the transition state theory with tunneling corrections, in a defined temperature range. The kinetic coefficients, as well as the modified Arrhenius parameters, are then assembled and organized to create a final table that connects the kinetic data with TS structure characteristics. These tables can be directly applied as a kinetic data source for reaction mechanism development. The ability of the code to generate reliable rate constants was tested for 1,3-H-shift reactions and the results were compared with theoretical data manually produced, and showed a good agreement. In particular, the code was able to create all the transition state configurations, with an exhaustive description of all possible reaction pathways, using a rigorous and systematic counting based on symmetry, stereocenters, and diastereomers. The proposed method leads to more accurate results on these aspects, compared to repetitive hand calculations of dozens of rate constants.