Theory of the electronic structure of dilute bismide alloys: Tight-binding and k · p models

Dilute bismide alloys have been the focus of increasing research effort in recent years, due in large part to their novel electronic properties. In particular, they display significant potential for achieving highly efficient photonic devices operating at telecomm wavelengths (1.3–1.5 μm). However, despite substantial progress in the growth and characterisation of dilute bismides, there have been comparatively few theoretical investigations of this novel material system. We summarise here aspects of our theoretical work on the electronic and optical properties of dilute bismide alloys. We present tight-binding and k · p models for the electronic structure of (In)GaBi_xAs_1−x, in which the strong reduction of the band gap (E_g) and increase in the spin-orbit-splitting energy (Δ_SO) are explained in terms of a band-anticrossing interaction between the extended states of the host matrix valence band edge and Bi-related resonant impurity states lying in the valence band. Our results, which are in good agreement with the available experimental data, serve to elucidate the origins of the novel electronic properties of dilute bismide alloys and confirm the crossover to an E_g < Δ_SO regime in GaBi_xAs_1−x for x ≳ 11 %, a condition which should lead to suppressed Auger recombination in long wavelength devices. The dilute bismide k ⋅ p model is applied to calculate the effect of Bi incorporation on the band structure and optical gain of dilute bismide quantum well structures, and some general trends relevant to laser operation are identified. We also extend our models to the quaternary dilute bismide–nitride alloy GaBi_xNyAs_1−x−y and show how co-alloying of Bi and N offers broad scope for band structure engineering which should lead to the realisation of highly efficient GaAs-based long wavelength photonic devices.