Electronic Structure Calculations of Chiral Matter - From First Principles Methods to Machine Learning Techniques

Amartya Banerjee
University of California, Los Angeles (UCLA)

Chiral materials - particularly, quasi-one-dimensional (1D) nanostructures - are often associated with unique (and sometimes anomalous) transport, optical, electric, and magnetic properties. As such, they offer unparalleled opportunities for impacting the design of novel quantum, photonic and electromagnetic devices. In this talk, I will describe our efforts in formulating and implementing electronic structure calculation techniques for such materials, so as to enable their discovery and computational characterization. First, I will show how the electronic states in chiral 1D nanomaterials can be characterized by means of special solutions to the single electron problem called helical Bloch waves, and how these solutions can be used to reduce the equations of Kohn–Sham Density Functional Theory (KS-DFT) to a suitable fundamental domain. I will then discuss real space and reciprocal space techniques, for discretizing and solving the governing equations. I will present results involving the electromechanical response of various 1D nanostructures simulated using our methods. Finally, I will describe an interpretable machine learning model that can be trained using first principles data, to directly predict the electronic fields of chiral 1D nanostructures, thereby accelerating the electronic structure calculation problem associated with these materials.

Back to Workshop III: Large-Scale Certified Numerical Methods in Quantum Mechanics