Theory of band gap bowing of disordered substitutional II-VI and III-V semiconductor alloys

TL;DR

This paper presents a theoretical approach to understanding and predicting the band gap bowing in disordered III-V and II-VI semiconductor alloys using empirical tight binding models, CPA, and supercell calculations.

Contribution

It introduces a method to accurately describe band gap bowing in semiconductor alloys based on empirical tight binding parameters and probabilistic modeling.

Findings

Accurately models band gap bowing in various semiconductor alloys.

Demonstrates the effectiveness of tight binding and CPA methods.

Provides insights into electronic properties of disordered alloys.

Abstract

For a wide class of technologically relevant compound III-V and II-VI semiconductor materials AC and BC mixed crystals (alloys) of the type A(x)B(1-x)C can be realized. As the electronic properties like the bulk band gap vary continuously with x, any band gap in between that of the pure AC and BC systems can be obtained by choosing the appropriate concentration x, granted that the respective ratio is miscible and thermodynamically stable. In most cases the band gap does not vary linearly with x, but a pronounced bowing behavior as a function of the concentration is observed. In this paper we show that the electronic properties of such A(x)B(1-x)C semiconductors and, in particular, the band gap bowing can well be described and understood starting from empirical tight binding models for the pure AC and BC systems. The electronic properties of the A(x)B(1-x)C system can be described by choosing the tight-binding parameters of the AC or BC system with probabilities x and 1-x, respectively. We demonstrate this by exact diagonalization of finite but large supercells and by means of calculations within the established coherent potential approximation (CPA). We apply this treatment to the II-VI system Cd(x)Zn(1-x)Se, to the III-V system In(x)Ga(1-x)As and to the III-nitride system Ga(x)Al(1-x)N.