Modeling the g-factors, hyperfine interaction and optical properties of semiconductor QDs: the atomistic and eight-band $k \cdot p$ approaches

TL;DR

This study compares atomistic tight-binding and continuum eight-band $k \, p$ methods for modeling the spin and optical properties of semiconductor quantum dots, proposing corrections to improve accuracy and establishing criteria for method selection.

Contribution

It provides a detailed comparison of two modeling approaches, introduces targeted corrections to the $k \, p$ method, and benchmarks hyperfine interactions for better accuracy in quantum dot simulations.

Findings

Both methods show qualitative agreement but differ quantitatively.

Targeted corrections improve $k \, p$ accuracy for g-factors and energies.

Hyperfine interaction implementation converges between models.

Abstract

We present a detailed comparative study of two important theoretical approaches: atomistic sp$^3$d$^5$s$^*$ tight-binding and continuum eight-band $k \cdot p$ methods, for modeling the spin and optical properties of quantum dots (QDs). Our investigation spans key physical observables, including single-particle energy levels, g-factors, exciton radiative lifetimes, and hyperfine-induced Overhauser field fluctuations. We perform our calculations for self-assembled InGaAs/GaAs QD systems as representative case studies. While both methods yield qualitatively consistent trends, quantitative discrepancies arise due to different treatment of atomistic details, strain effects, and confinement. We introduce targeted corrections to the eight-band $k \cdot p$ framework, including a modified deformation potential scheme and adjusted remote-band contributions, to improve agreement with atomistic results, especially for electron g-factors and single-particle energies. Furthermore, we validate the eight-band implementation of hyperfine interactions by benchmarking it against the tight-binding model, showing reasonable convergence for both electrons and holes. Our results establish criteria for selecting the appropriate modeling framework based on the desired physical accuracy and computational efficiency in spin-optical studies of semiconductor QDs.