Eric A. Walker

In this talk, examples demonstrate how data science tools can elucidate physical insight to chemical reactions.  The type of data science tool utilized depends upon the chemical application.  In the first part of the talk, the appropriate solvent for organic reactions is selected using a graph-based algorithm and a large database of experimental reaction data.  The graph algorithm is based upon the similarity in functional groups of the molecules involved in the reaction.  Upon visualizing the graph, the reaction data form clusters of consistency in terms of the reaction solvent, without the solvent being provided beforehand.  This indicates that the graph model has chemical-based predictive power.  The predictions of the graph algorithm for a test set of reactions are compared against the predictions of the same test set from 3 human chemists.  The second part of the talk examines the active site of the water-gas shift reaction on Pt supported by ceria oxide catalyst.  Three active site models are generated and populated with density functional theory calculations.  Two of the active site models are of the interface of the Pt and ceria.  The third active site model is of Pt only.  Using Bayesian statistical tools to determine which active site best explains experimental data, the interface is found to play an important physical role in the reaction.  The third part of the talk lays out ongoing and future work.  Data science tools will be applied to adsorption predictions of zeolite materials, mechanistic modeling of the dry-reforming of methane reaction, and parameter estimation of the non-reactive temperature programmed desorption of methanol from ceria. 

Dr. Eric A. Walker obtained his bachelor of science from the Georgia Institute of Technology and his doctor of philosophy degrees at the University of South Carolina, both in chemical engineering.  His doctoral dissertation focused on Uncertainty Quantification in Computational Catalysis and he was awarded an Eastman fellowship and presidential fellowship.  After graduate school, he completed postdoctoral research at the University of Michigan where he applied machine-learning tools to chemical challenges in catalyst and materials optimization.  In January 2019, he moved to the State University of New York at Buffalo where he is a research assistant professor in the Institute for Computational and Data Sciences with an adjunct appointment in the Chemical and Biological Engineering Department (since January 2020).  He serves as session chair in the catalysis and reaction engineering division of the American Institute of Chemical Engineers, and his research interests span catalysis, alternative energy, environment catalysis, data science, and machine learning.

• Time: 11:00 AM
• Location: 206 Furnas Hall
• Seminar Flyer