Embedding Dynamical Mean-Field Theory for Superconductivity in Layered Materials and Heterostructures

TL;DR

This paper extends embedding dynamical mean-field theory to layered superconducting systems, enabling efficient large-scale simulations and revealing how interfaces influence superconductivity penetration and proximity effects.

Contribution

It introduces a feedback-enhanced embedding scheme for superconducting layered systems within dynamical mean-field theory, improving computational efficiency and accuracy.

Findings

Superconductivity can penetrate about ten layers across interfaces.

Proximity effects can induce superconductivity in layers with repulsive interactions.

The method allows for large system simulations with reduced computational cost.

Abstract

We study layered systems and heterostructures of s-wave superconductors by means of a suitable generalization of Dynamical Mean-Field Theory. In order to reduce the computational effort, we consider an embedding scheme in which a relatively small number of active layers is embedded in an effective potential accounting for the effect of the rest of the system. We introduce a feedback of the active layers on the embedding potential that improves on previous approaches and essentially eliminates the effects of the finiteness of the active slab allowing for cheap computation of very large systems. We extend the method to the superconducting state, and we benchmark the approach by means of simple paradigmatic examples showing some examples on how an interface affects the superconducting properties. As examples, we show that superconductivity can penetrate from an intermediate coupling superconductor into a weaker coupling one for around ten layers, and that the first two layers of a system with repulsive interaction can turn superconducting by proximity effects even when charge redistribution is inhibited.