Charge redistribution in correlated heterostuctures within nonequilibrium real-space dynamical mean-field theory

TL;DR

This paper investigates charge redistribution in correlated heterostructures under bias using non-equilibrium real-space dynamical mean-field theory, highlighting the effects of interactions and system parameters on charge profiles.

Contribution

It introduces a self-consistent approach combining R-DMFT with Poisson's equation solved via Newton-Raphson, reducing computational costs for studying nonequilibrium charge redistribution.

Findings

Charge redistribution depends on Hubbard U and bias voltage.

Long-range Coulomb interactions significantly influence interface charge profiles.

The method efficiently captures steady-state behavior in correlated heterostructures.

Abstract

We address the steady-state behavior of a system consisting of several correlated monoatomic layers sandwiched between two metallic leads under the influence of a bias voltage. In particular, we investigate the effect of the local Hubbard and of the long-range Coulomb interactions on the charge redistribution at the interface. We provide a detailed study of the importance of the various system parameters, like Hubbard $U$, lead-correlated region coupling strength, and the applied voltage on the charge distribution in the correlated region and in the adjacent parts of the leads. Our results are obtained within non-equilibrium (steady-state) real-space dynamical mean-field theory (R-DMFT), with a self-consistent treatment of the long-range part of the Coulomb interaction by means of the Poisson equation. The latter is solved by the Newton-Raphson method and we find that this significantly reduces the computational cost compared to existing treatment. As impurity solver for R-DMFT we use the auxiliary master equation approach (AMEA), which addresses the impurity problem within a finite auxiliary system coupled to Markovian environments.