The Green research group focuses on the central problem of reactive chemical engineering: quantitatively predicting the time evolution of chemical mixtures. Accurate chemical kinetic models are extremely powerful and valuable since they allow predictions about the impact of modifying a system. We combine ab initio quantum chemical calculations, fundamental equations, and data-driven approaches to develop predictive models for kinetic and thermochemical parameters and construct detailed kinetic models based on predictions.

In this video, Kevin Spiekermann, a PhD student in Chemical Engineering at MIT, presents his work to automate reaction screening by predicting barrier heights using message-passing neural networks.