Effective electronic band structure of monoclinic β−(AlxGa1−x)2O3 alloy semiconductor

Abstract

In this article, the electronic band structure of a [Formula: see text] alloy system is calculated, with β−Ga2O3 as the bulk crystal. The technique of band unfolding is implemented to obtain an effective band structure for aluminum fractions varying between 12.5% and 62.5% with respect to gallium atoms. A 160-atom supercell is used to model the disordered system that is generated using the technique of special quasi-random structures, which mimics the site correlation of a truly random alloy by reducing the number of candidate structures that arise due to the large number of permutations possible for alloy occupation sites. The impact of the disorder is then evaluated on the electron effective mass and bandgap, which is calculated under the generalized gradient approximation.

Effective Band Structure and Crack Formation Analysis in Pseudomorphic Epitaxial Growth of (InxGa1–x)2O3 Alloys: A First-Principles Study

“…As previously discussed for similar materials, the PBE band structure is expected to be reasonably accurate, apart from the band gap underestimation. For instance, the band curvature given by PBE is expected to closely agree with that of HSE06 . The spectral function of the EBS, which provides a qualitative measure of Bloch’s character conserved after the band unfolding, is computed using eq .…”

“…This is likely due to the S size (80 atoms) used in previous studies, which is not sufficient to accurately model the substitutional disorder. As shown by Sharma et al 27 for (Al Ga ) O…”

“…Band unfolding is based on the expression of SC eigenvectors as a linear combination of PC eigenvectors. The spectral weight, defined as P boldK⃗ m ( boldk⃗ i ) = prefix∑ n false| ⟨ ψ boldK⃗ m false| ψ k⃗ i n ⟩ false| 2 determines the probability of an SC eigenstate to have the same Bloch character as a PC eigenstate . From the knowledge of the spectral weight, the spectral function A ( boldk⃗ i , E ) = ∑ m P boldK⃗ m ( boldk⃗ i ) δ ( E − E K⃗ m ) is obtained, where E boldK⃗ m is the energy of the SC eigenvector ψ boldK⃗ m .…”

“…The majority of the experimental and theoretical studies of alloys between Ga 2 O 3 and group-III element oxides focuses on false( normalA normall x normalG normala 1 − normalx false) 2 O 3 , with several works focusing on the engineering of their band gap and growth mechanisms. − Less attention has been paid to false( normalI normaln x normalG normala 1 − normalx false) 2 O 3 alloys. ,,,− Although indium, gallium, and aluminum are isovalent, their ambient stable oxides are not isomorphous due to different cation sizes. For instance, there are significant differences between the monoclinic β-Ga 2 O 3 and body-centered cubic (BCC) In 2 O 3 (bixbyite) structures, including lattice parameters and cation coordination.…”

“…The electronic structure of false( normalI normaln x normalG normala 1 − normalx false) 2 O 3 alloys has been already studied from first-principles by Liu and Tan . In that study, semilocal functionals were employed, which are known to underestimate the band gap, and a supercell of 80 atoms was used, which is rather too small for modeling accurately the disordered structures . At variance with Liu and Tan, we use the Heyd, Scuseria, and Ernzerhof (HSE06) hybrid functional, , which returns accurate band gaps without the need for empirical adjustments (e.g., by means of “scissor operators”).…”