Large Scale Raman Spectrum Calculations in Defective 2D Materials using Deep Learning

TL;DR

This paper presents a machine learning workflow for large-scale Raman spectrum calculations in defective 2D materials, enabling simulations with tens of thousands of atoms and good agreement with experiments.

Contribution

The study introduces a novel combination of machine-learned potentials and density of states methods for scalable Raman response simulations in 2D materials.

Findings

Successful simulation of Raman spectra in large defective 2D systems

Good agreement between predicted and experimental Raman responses

Demonstration of scalability to tens of thousands of atoms

Abstract

We introduce a machine learning prediction workflow to study the impact of defects on the Raman response of 2D materials. By combining the use of machine-learned interatomic potentials, the Raman-active $\Gamma$-weighted density of states method and splitting configurations in independant patches, we are able to reach simulation sizes in the tens of thousands of atoms, with diagonalization now being the main bottleneck of the simulation. We apply the method to two systems, isotopic graphene and defective hexagonal boron nitride, and compare our predicted Raman response to experimental results, with good agreement. Our method opens up many possibilities for future studies of Raman response in solid-state physics.