Advances in Computational Analysis, Discovery, and Design of Porous Materials
|Event Date:||April 21, 2014|
|Speaker:||Dr. Richard Luis Martin|
|Speaker Affiliation:||Computational Research Division, Lawrence Berkeley National Laboratory|
|Sponsor:||Purdue School of Materials Engineering|
Porous materials have been exploited in industrial applications for many years, as catalysts for oil refinement, water softeners and membranes for separations. They are also promising candidates for future application to such diverse challenges as vehicular natural gas storage, carbon dioxide capture, drug delivery and sensing. Due to the vast number of possible materials, identifying those with the optimal properties for a particular application remains a great challenge.
In this presentation I will introduce recently-developed computational tools aimed at accelerating and automating the analysis, discovery and design of porous materials. An open source materials analysis suite, Zeo++ , utilizes the Voronoi decomposition to identify and map the guest molecule-accessible pores within a material’s structure. Moreover, the calculated Voronoi network is utilized for rapid comparison of materials and prediction of their properties. I will describe information science techniques, inspired by methods successfully deployed in drug discovery, that enable the (dis)similarity-based searching and sampling of materials and the identification of property-determining features within their structures. Together with a high-performance graphics processing unit (GPU)-based simulation tool , these capabilities have enabled accelerated discovery of energy-efficient materials for carbon dioxide capture . Finally, I will introduce a tool for high-throughput prediction of porous material crystal structures which utilizes a topology-based modeling approach , and demonstrate its utility in mapping the chemical space of metal-organic frameworks, a promising class of material for methane storage applications, via an optimization-based design strategy .
Dr. Richard Luis Martin is a postdoctoral research fellow in the Computational Research Division at Lawrence Berkeley National Laboratory. Richard’s multidisciplinary background comprises a BSc in Computer Science and Mathematics and a PhD in Chemoinformatics, from the University of Sheffield, UK. His research interests concern the development of computational and information science techniques for accelerating scientific discovery and informing experimental design efforts.
Throughout his postdoctoral appointment, Richard has focused on automating the analysis, discovery and design of porous materials. He has actively advanced the new field of Materials Informatics, an endeavor which aims at realizing the data-driven future of materials engineering. His recent work has included the development of network science tools for searching and sampling large databases of porous materials; the application of high-performance parallel GPU/CPU computing to the prediction of material properties; and automating the design of porous materials with specifically tailored properties via mathematical optimization.