In 2013 we were awarded a $100k Doctoral New Investigator (DNI) Grant by the American Chemical Society Petroleum Research Fund.

In January 2014 we began work on the project, titled **“Transition-State Prediction for High-Throughput Calculation of Accurate Chemical Reaction Rates”**.

Even conventional engine designs require hundreds of hours of costly tuning to optimize, for each fuel; novel engine designs depend even more critically on the combustion behavior of the fuels. More accurate kinetic models will allow proposed fuels, and engine designs, to be screened more reliably in silico, enabling significant savings in time and money.

Modern quantum chemistry methods allow reaction rate coefficients to be calculated ab inito with high accuracy, but the current pace of these high-accuracy rate calculations is at least an order of magnitude too slow. With the advance of High Performance Computing the bottle-neck is no longer the computation, but the human interaction required to set up the calculation. Only once these calculations are automated will predictive kinetics realize the full potential of ever-increasing computational power.

This project aims to determine how to locate transition state geometries automatically, so that quantum mechanical electronic structure calculations can calculate accurate reaction rate expressions without human input. The hypothesis is that the transition state geometry is mostly determined by the reaction family, the chemical structure near the reacting site, and the geometries of the reactants; a hierarchical group-contribution scheme, similar to that used to estimate reaction rates, should therefore, with modification, be able to estimate transition state geometries. Incorporated into a software platform that automatically generates detailed reaction mechanisms, and coupled with increasingly high-performance computers, this will enable the high-throughput calculation of thousands of accurate reaction rates for petroleum combustion and other fields.