Electronic band structure ofCuInSe2:Bulk and (112) surface

Abstract

The nonlinear optical properties and potential technological applications of chalcopyrites have attracted attention in the last few years. The electronic band structure of their surfaces, and of interfaces, quantum wells, and superlattices with chalcopyrites as one of the components, contributes to a deeper understanding of the details underlying their interesting properties. In this paper, we use the Slater-Koster formalism to set a tight-binding Hamiltonian for Cu-based chalcopyrites. We use an ${\mathrm{sp}}^3$ basis for both an atom belonging to the IIIA group (In) and one belonging to the VIA one (Se). For the Cu atom, we use a full ${\mathrm{sp}}^3{d}^5$ basis. With tight-binding parameters (TBP’s) obtained from Harrison’s formulas, we can reproduce the characteristics of the known ab initio calculated bulk electronic band structures. The gaps are, nevertheless, very badly reproduced. We found a set of more suitable on-site TBP’s for Cu by minimizing the deviation from the experimental gap value for the whole series of chalcopyrites considered. Small adjustments with the other on-site TBP’s allowed us to fit the experimental optical gap values accurately. Here we present our result for the bulk band structure of ${\mathrm{CuInSe}}_2,$ and compare it with the existing ab initio calculations. We also consider the effect of tetragonal deformation. We then proceed to use this Hamiltonian to calculate the (112) surface local density of states making use of the surface Green’s-function matching method. This particular direction is the one in which this chalcopyrite is grown to produce an interface with some zinc-blende II-VI semiconductors (CdS, for example) in highly efficient solar cells.