- This event has passed.
PICS Colloquium: “Designing energy conversion materials with ab-initio and active machine learning computations of electron-phonon and ion dynamics”
November 20 at 2:00 PM - 3:00 PM
Abstract: Accurate atomistic computations of transport and reaction dynamics are an important challenge and an opportunity for designing materials for energy conversion and storage. In the context of thermoelectric materials, we develop new automatable computational methods for describing electron-phonon scattering dynamics. By predicting electrical transport properties, we computationally discovered several new low-cost thermoelectric alloys with record device performance. In the context of solid-state batteries, computations of ionic transport reveal how strong ionic interactions lead to disorder and surprising collective phenomena in amorphous polymer electrolyte materials and enable us to design new electrolyte chemistries.
High-fidelity ab-initio simulations of atomistic dynamics are limited to small systems and short times, and development of surrogate machine learning models for force fields is an emerging promising direction to access long-time large-scale dynamics of complex materials systems. However, the main challenges are high accuracy, reliability, and computational efficiency of these models, which critically depend on the training data sets. We develop ML interatomic potential models that are interpretable and uncertainty-aware, and orders of magnitude faster than reference quantum methods. Principled uncertainty quantification built into these models enables the construction of autonomous data acquisition schemes using active learning. We demonstrate on-the-fly learning of machine learning force fields and use them to gain insights into previously inaccessible physical and chemical phenomena in ion conductors, catalytic surface reactions, 2D materials phase transformations, and shape memory alloys.
Associate Professor of Computational Materials Science at the Harvard School of Engineering and Applied Science
Boris Kozinsky is an Associate Prof. at the Harvard School of Engineering and Applied Sciences. He studied at MIT for his B.S. degrees in Physics, Mathematics, and Electrical Engineering and Computer Science, and received his PhD degree in Physics also from MIT. He then established and led the atomistic computational materials science team at Bosch Research in Cambridge MA, before moving to Harvard in 2018. He works at the intersection of fundamental materials physics, efficient computational algorithms, and data science approaches. His group develops and uses atomistic and electronic structure computations and machine learning algorithms for understanding design rules governing quantum-level microscopic effects, particularly ionic, electronic and thermal transport in materials for energy storage and conversion. His work on development and application of computational methods led to computation-driven discovery of new materials advances and over 50 patents applications in a wide range of materials systems, including 1D and 2D materials, piezoelectrics, thermoelectrics, batteries, super-ionic conductors, catalysts, and functional polymers. Website: http://bkoz.seas.harvard.edu