Mott p-n Junctions in layered materials

TL;DR

This paper introduces a new theoretical approach called 'Dynamical Layer Theory' to model charge redistribution in layered Mott insulator p-n junctions, revealing complex charge behaviors beyond traditional semiconductor models.

Contribution

It develops a novel modeling framework combining Cellular Dynamical Mean Field Theory and Hartree-Fock approximation for layered Mott insulators, capturing charge reconstructions at interfaces.

Findings

Charge redistribution can be controlled by Fermi energy differences.

Quasi-two-dimensional charge carriers can form within Mott insulators.

Density of states profiles differ from simple band bending models.

Abstract

The p-n junction has provided the basis for the semiconductor-device industry. Investigations of p-n junctions based on Mott insulators is still in its infancy. Layered Mott insulators, such as the cuprates or other transition metal-oxides, present a special challenge since strong in-plane correlations are important. Here we model the planes carefully using plaquette Cellular Dynamical Mean Field Theory with an exact diagonalization solver. The energy associated with inter-plane hopping is neglected compared with the long-range Coulomb interaction that we treat in the Hartree-Fock approximation. Within this new approach, "Dynamical Layer Theory", the charge redistribution is obtained at the final step from minimization of a function of the layer fillings. A simple analytical description of the solution, in the spirit of Thomas-Fermi theory, reproduces quite accurately the numerical results. Various interesting charge reconstructions can be obtained by varying the Fermi energy differences between both sides of the junction. One can even obtain quasi-two dimensional charge carriers at the interface, in the middle of a Mott insulating layer. The density of states as a function of position does not follow the simple band bending picture of semiconductors.