AMC is offering a Seminar Series on Atomistic Modeling this summer semester!

Kjell Jorner is Assistant Professor of Digital Chemistry at the Department of Chemistry and Applied Biosciences ETH Zürich. His group develops digital solutions to chemical problems in catalysis and molecular design.

The talk will cover computational tools to enable fast predictions of structure and reaction mechanisms of catalysts and ligands, including expert descriptors, reaction dataset generation, transition state sampling, and diffusion models, all aimed at advancing generative models for catalysis.

Date: June 05, 2025, 10:15 am

Location: MIBE Lecture Hall

Abstract:
Machine learning models based on high-quality simulated or experimental data can potentially provide both fast and accurate predictions to support the development of new chemical reactions and routes to compounds of practical interest. Developing such tools presents challenges such as developing the necessary datasets and making the models generalize well outside of the immediate domain of their training data. 

In my group, we develop computational tools to enable fast predictions based on atomistic structure and an understanding of the reaction mechanisms. In the area of traditional tools, we develop and maintain software to calculate expert descriptors based on the steric and electronic properties of catalysts and ligands. I will also cover our ongoing work on fast tools for reaction dataset generation and transition state sampling, and diffusion models for molecular structure generation. Based on such tools and the corresponding datasets, we are building towards generative models for catalysis.