Band alignment of semiconductors and insulators using dielectric-dependent hybrid functionals: Toward high-throughput evaluation

Abstract

The band alignment of prototypical semiconductors and insulators is investigated using first-principles calculations. A dielectric-dependent hybrid functional, where the nonlocal Fock exchange mixing is set at the reciprocal of the static electronic dielectric constant and the exchange correlation is otherwise treated as in the Perdew-Burke-Ernzerhof (PBE0) hybrid functional, is used as well as the Heyd-Scuseria-Ernzerhof (HSE06) hybrid and PBE semilocal functionals. In addition, these hybrid functionals are applied non-self-consistently to accelerate calculations. The systems considered include C and Si in the diamond structure, BN, AlP, AlAs, AlSb, GaP, GaAs, InP, ZnS, ZnSe, ZnTe, CdS, CdSe, and CdTe in the zinc-blende structure, MgO in the rocksalt structure, and GaN and ZnO in the wurtzite structure. Surface band positions with respect to the vacuum level, i.e., ionization potentials and electron affinities, and band offsets at selected zinc-blende heterointerfaces are evaluated as well as band gaps. The non-self-consistent approach speeds up hybrid functional calculations by an order of magnitude, while it is shown using HSE06 that the resultant band gaps and surface band positions are similar to the self-consistent results. The dielectric-dependent hybrid functional improves the band gaps and surface band positions of wide-gap systems over HSE06. The interfacial band offsets are predicted with a similar degree of precision. Overall, the performance of the dielectric-dependent hybrid functional is comparable to the $G{W}_{0}$ approximation based on many-body perturbation theory in the prediction of band gaps and alignments for most systems. The present results demonstrate that the dielectric-dependent hybrid functional, particularly when applied non-self-consistently, is promising for applications to systematic calculations or high-throughput screening that demand both computational efficiency and sufficient accuracy.