Materials Science Research Lecture

Abstract:

The ab-initio prediction of properties has made tremendous progress in the last three decades, making it now possible to predict many functional properties of compounds with high reliability. This has led to many examples of computer-designed materials.

As computational property prediction continues to extend its reach, we will become increasingly limited by the time required needed to synthesize new materials in a predictive and controlled manner. In particular, an understanding of which metastable compounds can be formed and under which condition they can be made would enhance the focus on computational materials design towards feasible, synthesizable materials

I will discuss two approaches towards a predictive approach to synthesis: A classic, more deductive approach can be formulated by combining the ideas of nucleation with first principles computations of the quantities that enter nucleation theory. I will show how this can be used to predict the synthesis path for several metastable compounds. A second, radically different, approach is to use artificial intelligence and machine learning methods to attempt to learn materials synthesis. I will show results from machine learning about 3 Million research papers.

More about the Speaker:

Gerbrand Ceder is the Daniel M. Tellep Distinguished Professor in Engineering, Professor of Materials Science and Engineering at UC Berkeley and Faculty Staff Scientist at LBNL. He received an engineering degree from the University of Leuven, Belgium, and a Ph.D. in Materials Science from the University of California at Berkeley in 1991. Before moving to California he was a Professor in Materials Science at the Massachusetts Institute of Technology between 1991 and 2015. His research interests lie in materials design through first-principles computations and selected experiments. He has developed methods for high-throughput computation and integrates data mining and statistical learning in materials science. Current interests include nucleation and formation of materials and metastability. He has published over 400 scientific papers, and holds several U.S. patents. He has served on MIT's Energy Council as well as on several DOE committees, including the workgroup preparing the Basic Needs for Electrical Energy Storage report, and has advised the government's Office of Science and Technology Policy on the role of computation in materials development, leading to the Materials Genome Initiative and is currently on the International Scientific Advisory Board of the Fritz Haber Institute. He is a Fellow of the Materials Research Society and a member of the National Academy of Engineeringand the Royal Flemish Academy of Belgium for Sciences and the Arts. He is a TMS Fellow and has received the TMS Cohen Award, the MRS Gold Medal, the Battery Research Award from the Electrochemical Society, the Career Award from the National Science Foundation, and the Robert Lansing Hardy Award from The Metals, Minerals and Materials Society, as well as several teaching awards. He is a co-founder of Computational Modeling Consultants, Pellion Technologies, and The Materials Project.