Computational material property and discovery from density functional theory and machine learning 

Dr. Mengen Wang

Department of Electrical and Computer Engineering

The State University of New York at Binghamton 

Abstract: First-principles density functional theory (DFT) calculations have contributed to assessing material properties and revolutionized materials science and condensed matter physics. Examples of material properties include mechanical properties, electrical and thermal conductivity, catalytical properties, and optical properties. The predictive power of first-principles calculations in assessing material properties provides opportunities for high-throughput calculations and accelerated screening of new materials, which help the experimental search of chemical spaces and synthesizing conditions.

Defects and surfaces are ubiquitous in materials and can alter their functionalities. Functional defects can be shallow donors or acceptors contributing to the n-type or p-type conductivity of semiconductors. Surface and grain boundary properties are relevant to material degradation mechanisms and defect incorporation. Using halide perovskite as an example, I will present how we use DFT calculations combined with machine learning to understand the defect and surface properties of semiconductor materials. In the first part of my talk, I will introduce the DFT computation of defect formation energy and charge transition levels of dopants in a perovskite material CsSnI₃ and the training and prediction of machine learning regression models. In the second part of my talk, I will introduce how machine learning interatomic potentials are trained and applied to constructing surface phase diagrams of perovskite materials, including CsSnI₃ and CsPbI₃. To conclude, I will discuss the remaining challenges and outlooks in computational defects and surface studies combining high-throughput DFT calculations with machine learning methods.

Bio: Mengen Wang is an Assistant Professor from the Department of Electrical and Computer Engineering at Binghamton University. Mengen received her Ph.D. in 2019 in Materials Science and Chemical Engineering from Stony Brook University, where she worked at Brookhaven National Laboratory, focusing on applying first-principles computation for heterogeneous catalysis. From 2019 to 2022, she worked as a postdoc in the Materials Department at the University of California, Santa Barbara, focusing on first-principles computation of the growth and defects in wide-bandgap semiconductors for power electronics and quantum information science applications. Her current research is focused on combining first-principles approaches with machine learning to design semiconductor materials for optoelectronic and quantum information applications.