Research Topic
Computational Chemistry and Reaction Engineering
The goal of our research is to develop algorithms to rapidly generate kinetic models that can accurately simulate reaction systems. Because a complex reaction system could potentially involve thousands of kinetically significant chemical species, it is difficult to experimentally derive a detailed kinetic model that incorporates all the important reaction channels. Therefore, models derived by experiments are often heavily simplified and may fail to predict detailed chemistry of a large reaction network. Computer-generation of kinetic models should allow us to overcome this problem since it can keep track of all the important species, and thus can help us to move beyond the resolution limitations of experimental techniques. By combining quantum mechanical calculations and machine learning algorithms, we are dedicated to making it practical to accurately predict the reaction activity and selectivity of complex systems, based on fundamental understanding of the detailed chemistry. The long-term goal of our research is to allow one to use simulations to quickly test many hypotheses in a high-throughput manner, and facilitates the development of reaction technology by replacing some of the many required experiments.