Simple linear response model for predicting energy band alignment of two-dimensional vertical heterostructure

TL;DR

This paper introduces a linear response model that accurately predicts energy band alignment in 2D van der Waals heterostructures, outperforming traditional models like Anderson and midgap.

Contribution

The authors propose a new linear response model for band alignment prediction in 2D heterostructures, validated against DFT calculations, with a specific coefficient for group-IV monochalcogenides.

Findings

Linear response model predicts DFT band alignment effectively.

The model's parameter α = 0.34 fits group-IV monochalcogenide heterostructures.

The model accounts for interface dipoles and can incorporate effects like strain.

Abstract

The Anderson and midgap models are often used in the study of semiconductor heterojunctions, but for van der Waals (vdW) vertical heterostructures they have shown only very limited success. Using the group-IV monochalcogenide vertical heterostructures as a prototypical system, we propose a linear response model and compare the effectiveness of these models in predicting density functional theory (DFT) band alignments, band types and bandgaps. We show that the DFT band alignment is best predicted by the linear response model, which falls in between the Anderson and midgap models. Our proposed model can be characterized by an interface dipole $\alpha\times(E_{m2}-E_{m1})$, where the linear response coefficient $\alpha$ = 0 and 1 corresponds to the Anderson and midgap model respectively, and $E_{m}$ is the midgap energy of the monolayer, which can be viewed as an effective electronegativity. For group-IV monochalcogenides, we show that $\alpha$ = 0.34 best captures the DFT band alignment of the vdW heterostructure, and we discuss the viability of the linear response model considering other effects such as strains and band hybridization, and conclude with an application of the model to predict experimental band alignments.