Disorder by design: A data-driven approach to amorphous semiconductors without total-energy functionals

TL;DR

This paper presents a data-driven, multi-objective optimization method to reconstruct realistic amorphous semiconductor models from diffraction data without relying on total-energy functionals, achieving high accuracy in structural and electronic properties.

Contribution

It introduces a novel inverse modeling approach that produces highly realistic amorphous silicon structures without using total-energy functionals, resolving longstanding debates on model uniqueness.

Findings

Models have ≤1% coordination defects

Bond-angle distribution width of 9-11.5 degrees

Electronic gap of 0.8-1.4 eV matching experimental data

Abstract

This paper addresses a difficult inverse problem that involves the reconstruction of a three-dimensional model of tetrahedral amorphous semiconductors via inversion of diffraction data. By posing the material-structure determination as a multi-objective optimization program, it has been shown that the problem can be solved accurately using a few structural constraints, but no total-energy functionals/forces, which describe the local chemistry of amorphous networks. The approach yields highly realistic models of amorphous silicon, with no or only a few coordination defects ($\le$ 1%), a narrow bond-angle distribution of width 9-11.5 degree, and an electronic gap of 0.8-1.4 eV. These data-driven information-based models have been found to produce electronic and vibrational properties of amorphous silicon that match accurately with experimental data and rival that of the Wooten-Winer-Weaire (W3) models. The study confirms the effectiveness of a multi-objective optimization approach to the structural determination of complex materials, and resolves a long-standing dispute concerning the uniqueness of a model of tetrahedral amorphous semiconductors obtained via inversion of diffraction data.