Efficient Method for Modeling Polarons Using Electronic Structure Methods

doi:10.1021/acs.jctc.0c00374

Return

Efficient Method for Modeling Polarons Using Electronic Structure Methods

2020-06-30

PRE

Thang Duc Pham

N. Aaron Deskins *

DOI：10.1021/acs.jctc.0c00374

Original

Original request for help

Save

Abstract

En 中文

Polarons are localized electronic states that occur in many semiconductors. Modeling polarons at the quantum or atomic scale is often performed using electronic structure methods such as density functional theory (DFT). A problem using DFT to model polarons is that self-interaction errors (SIEs) often result in delocalized electronic states rather than localized states. Methods such as DFT + U or hybrid functionals can be used to overcome SIE, but these methods may still not form stable polarons. The initial geometries and wavefunctions strongly influence and determine how and if polarons may arise during electronic structure calculations. In this paper, we have assessed different strategies to efficiently obtain low-energy localized polarons in several semiconductors (TiO2, m-HfO2, and m-BiVO4). These strategies involve distorting the initial geometry to create polaron-like geometries or generating initial wavefunctions that mimic polaronic states. We show that perturbing the crystal's structure to induce polaron formation (which we call the bond distortion method) is a very efficient approach to form stable polarons, requiring less computational time than other methods. In contrast, other methods that we assessed may not lead to stable polaron states or may require much greater time (up to four times more computational time). Having a reliable, efficient method to ensure polaron formation is crucial to modeling polarons. The results described herein will save wasted computational efforts and also enable efforts such as high-throughput simulation of polarons.

Keywords:

INITIO MOLECULAR-DYNAMICS

DENSITY-FUNCTIONAL THEORY

TOTAL-ENERGY CALCULATIONS

EXCESS ELECTRONS

REDUCED RUTILE

PLUS U

ANATASE

TIO2

SURFACE

BIVO4

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.