Compensation, interstitial defects and ferromagnetism in diluted semiconductors

TL;DR

This paper develops a quantitative theory for ferromagnetism in diluted III-V semiconductors considering common defects, showing Mn interstitials, not As anti-sites, best explain experimental Curie temperatures.

Contribution

It introduces an effective Heisenberg model with ab initio exchange integrals and a semi-analytical solution that accurately predicts Curie temperatures considering Mn interstitials as the main compensating defects.

Findings

Mn interstitials explain measured Curie temperatures well.

As anti-sites are unlikely the dominant compensation source.

The model accurately matches experimental data for various samples.

Abstract

We present a quantitative theory for ferromagnetism in diluted III-V ferromagnetic semi-conductors in the presence of the two types of defects commonly supposed to be responsible for compensation: As anti-sites and Mn interstitials. In each case we reduce the description to that of an effective random Heisenberg model with exchange integrals between active magnetic impurities provided by ab initio calculation. The effective magnetic Hamiltonian is then solved by a semi-analytical method (locally self-consistent RPA), where disorder is treated exactly. Measured Curie temperatures are shown to be inconsistent with the hypothesis that As anti-sites provide the dominant mechanism for compensation. In contrast, if we assume that Mn interstitials are the main source for compensation, we obtain a very good agreement between the calculated Curie temperature and the measured values, in both as-grown and annealed samples.