Electronic Structure Theory of Weakly Interacting Bilayers

TL;DR

This paper develops transferable electronic structure models for weakly interacting bilayers like graphene, enabling analysis of their electronic properties under various configurations, including twists and translations, with applications in heterostructure design.

Contribution

The authors derive a new, transferable interlayer coupling model based on density functional theory and Wannier functions, capturing angular dependence and applicable to arbitrary bilayer configurations.

Findings

Successfully applied the model to rotated graphene bilayers, reproducing Fermi velocity renormalization.

Predicted van Hove singularities and Moiré patterns in electronic localization.

Demonstrated the model's utility for designing van der Waals heterostructures.

Abstract

We derive electronic structure models for weakly interacting bilayers such as graphene-graphene and graphene-hexagonal boron nitride, based on density functional theory calculations followed by Wannier transformation of electronic states. These transferable interlayer coupling models can be applied to investigate the physics of bilayers with arbitrary translations and twists. The functional form, in addition to the dependence on the distance, includes the angular dependence that results from higher angular momentum components in the Wannier $p_z$ orbitals. We demonstrate the capabilities of the method by applying it to a rotated graphene bilayer, which produces the analytically predicted renormalization of the Fermi velocity, van Hove singularities in the density of states, and Moir\'{e} pattern of the electronic localization at small twist angles. We further extend the theory to obtain the effective couplings by integrating out neighboring layers. This approach is instrumental for the design of van der Walls heterostructures with desirable electronic features and transport properties and for the derivation of low-energy theories for graphene stacks, including proximity effects from other layers.