Conduction band spin splitting and negative magnetoresistance in ${\rm A}_3{\rm B}_5$ heterostructures

TL;DR

This paper develops a theory for how spin splitting affects magnetoconductivity in asymmetric quantum wells, revealing non-additive contributions of Dresselhaus and Rashba effects and matching experimental data.

Contribution

It introduces a non-additive model for conduction band spin splitting effects on magnetoconductivity, accounting for Dresselhaus and Rashba interactions in quantum wells.

Findings

Linear spin splitting terms cancel when equal, nullifying their effect on magnetoconductivity.

The theory aligns well with experimental measurements.

Parameters of spin-orbit splitting can be experimentally determined using this model.

Abstract

The quantum interference corrections to the conductivity are calculated for an electron gas in asymmetric quantum wells in a magnetic field. The theory takes into account two different types of the spin splitting of the conduction band: the Dresselhaus terms, both linear and cubic in the wave vector, and the Rashba term, linear in wave vector. It is shown that the contributions of these terms into magnetoconductivity are not additive, as it was traditionally assumed. While the contributions of all terms of the conduction band splitting into the D'yakonov--Perel' spin relaxation rate are additive, in the conductivity the two linear terms cancel each other, and, when they are equal, in the absence of the cubic terms the conduction band spin splitting does not show up in the magnetoconductivity at all. The theory agrees very well with experimental results and enables one to determine experimentally parameters of the spin-orbit splitting of the conduction band.