Effective Passivant Pseudopotentials for semiconductors: beyond the spherical approximation

TL;DR

This paper introduces a novel complex pseudopotential approach for semiconductor surface passivation, surpassing the spherical approximation, and demonstrates its high transferability and accuracy across multiple materials and surface configurations.

Contribution

It develops a new complex pseudopotential method for passivants that improves accuracy and transferability over traditional spherical models, with a detailed implementation and extraction procedure.

Findings

Accurately models passivants for six semiconductor compounds.

Demonstrates excellent agreement with density functional theory.

Effective on various surfaces, including quantum wires.

Abstract

A new type of effective atomic pseudopotential for passivation of semiconductor surfaces is presented. It is shown that the spherical approximation used in the effective and empirical pseudopotential methods is not suitable for describing passivants and that, instead, they have to be regarded as complex quantities. Since the new seudopotentials cannot be handled as usual bulk ones, the way of implementing them in atomistic methods is described here, together with a methodology for extracting them trough an analytic connection to density functional theory, meaning that the new type of surface potential keeps the {\it ab inito} identity. The effectiveness and high transferability of this new approach is demonstrated by generating passivant potentials for six different semiconductor compounds (GaAs, AlAs, Si, Ge, CdSe and ZnO) and testing them on different kinds of surfaces, including a passivated 68 atoms CdSe quantum wire in wurtzite structure, obtaining always an excellent agreement to density functional theory calculations done on the same systems.