Unified ab initio description of Fr\"ohlich electron-phonon interactions in two-dimensional and three-dimensional materials

TL;DR

This paper develops a unified ab initio framework for describing Fr"ohlich electron-phonon interactions in both 2D and 3D materials, eliminating the need for Coulomb truncation in 2D systems.

Contribution

It generalizes the Fr"ohlich electron-phonon matrix element to 2D materials using a supercell approach, providing accurate formulas without Coulomb truncation.

Findings

Validated approach with DFT calculations on monolayer BN and MoS₂.

Derived a simple expression for the 2D Fr"ohlich matrix element.

Unified formalism reduces to known formulas in 3D and 2D limits.

Abstract

\textit{Ab initio} calculations of electron-phonon interactions including the polar Fr\"ohlich coupling have advanced considerably in recent years. The Fr\"ohlich electron-phonon matrix element is by now well understood in the case of bulk three-dimensional (3D) materials. In the case of two-dimensional (2D) materials, the standard procedure to include Fr\"ohlich coupling is to employ Coulomb truncation, so as to eliminate artificial interactions between periodic images of the 2D layer. While these techniques are well established, the transition of the Fr\"ohlich coupling from three to two dimensions has not been investigated. Furthermore, it remains unclear what error one makes when describing 2D systems using the standard bulk formalism in a periodic supercell geometry. Here, we generalize previous work on the \textit{ab initio} Fr\"ohlich electron-phonon matrix element in bulk materials by investigating the electrostatic potential of atomic dipoles in a periodic supercell consisting of a 2D material and a continuum dielectric slab. We obtain a unified expression for the matrix element, which reduces to the existing formulas for 3D and 2D systems when the interlayer separation tends to zero or infinity, respectively. This new expression enables an accurate description of the Fr\"ohlich matrix element in 2D systems without resorting to Coulomb truncation. We validate our approach by direct \textit{ab initio} density-functional perturbation theory calculations for monolayer BN and MoS$_2$, and we provide a simple expression for the 2D Fr\"ohlich matrix element that can be used in model Hamiltonian approaches. The formalism outlined in this work may find applications in calculations of polarons, quasiparticle renormalization, transport coefficients, and superconductivity, in 2D and quasi-2D materials.