As machine learning becomes accessible to a broader community, our group is exploring how to use machine learning in designing ligands that selectively bind to particular metal ions. To train the algorithm, scientists at Ames Laboratory are combining experimental data, structural information about ligands, metal ions, and data obtained from ab initio and molecular mechanics calculations. We are expanding the capabilities to include metal ion selectivity rankings obtained from machine learning algorithms of the HostDesigner software that quickly generates theoretical ligands. We anticipate that the new software will lead to the discovery of new, more financially attractive avenues of processing and obtaining critical materials.
Speaker: Dr. Marilu Perez Garcia is an Associate Scientist at Ames Lab and leads a collaborative project with Federico Zahariev and Benjamin Hay in the Critical Materials Institute. She completed her B.S. from Idaho State University as an ACS Scholar and her PhD in chemistry with Teaching and Research Excellence awards at Iowa State University. Her research interests include using traditional electronic structure theory and new machine learning computational tools for predicting properties of metal-ligand complexes, especially as they pertain to critical materials.