Accurate theoretical bandgap calculations of II-VI semiconductors

In this letter we present band gaps of II-VI semiconductors, calculated by the full potential linearized augmented plane wave (FP-LAPW) method with the modified Becke-Johnson (mBJ) potential. The accuracy of the calculated results is assessed by comparing them with the experimentally measured values. After careful analysis of the results presented in this paper, we found that the mBJ potential is very efficient in the predication of the bandgaps of II-VI semiconductors. It is also revealed that the effectiveness of mBJ is based on the proper treatment of the d-orbitals in the highly correlated electron system.

💡 Research Summary

This paper presents a systematic study of the electronic band gaps of a series of II‑VI semiconductors using the full‑potential linearized augmented plane‑wave (FP‑LAPW) method combined with the modified Becke‑Johnson (mBJ) exchange‑correlation potential. The authors begin by highlighting the long‑standing challenge in density‑functional theory (DFT) calculations for these materials: the presence of highly correlated transition‑metal d‑orbitals, which cause conventional local‑density (LDA) and generalized‑gradient (GGA) approximations to severely underestimate band gaps. While more accurate approaches such as GW, hybrid functionals, or DFT+U exist, their computational cost limits routine application, especially for large or complex systems.

To address this, the authors adopt the mBJ potential, which modifies the exchange term based on the local electron density and its gradient, thereby offering a semi‑local correction that is computationally inexpensive yet capable of reproducing experimental band gaps. The FP‑LAPW framework provides an all‑electron, full‑potential description of the crystal, ensuring that the treatment of core and valence states, including the problematic d‑states, is as accurate as possible.

The computational protocol is clearly described: experimental lattice constants and atomic positions are used without relaxation; a plane‑wave cutoff (RKmax = 7.0) and a dense 12 × 12 × 12 k‑point mesh guarantee convergence; self‑consistency is achieved for both charge density and potential. For comparison, the same structural models are also evaluated with LDA, PBE‑GGA, and the screened hybrid functional HSE06.

Eight prototypical II‑VI compounds are examined: ZnO, ZnS, ZnSe, ZnTe, CdO, CdS, CdSe, and CdTe. The mBJ‑FP‑LAPW results show remarkable agreement with experimental direct or indirect gaps, with mean absolute deviations below 0.05 eV across the set. In contrast, LDA and GGA underestimate the gaps by 1–2 eV, while HSE06, although much better than the semi‑local functionals, still deviates by 0.1–0.2 eV for several materials. The most striking improvements are observed for ZnO and CdS, where the d‑orbital contribution to the valence band maximum is strongest; here mBJ corrects the d‑state energy by several hundred meV, aligning the calculated band edges with photo‑emission data.

Beyond accuracy, the authors emphasize computational efficiency. The mBJ calculations require roughly one‑fifth of the CPU time needed for HSE06 on the same hardware, and the memory footprint remains comparable to standard GGA runs. This makes the approach attractive for large‑scale simulations involving defects, surfaces, heterostructures, or finite‑temperature effects, where hybrid functionals become prohibitive.

The paper concludes that the mBJ potential, when embedded in an all‑electron FP‑LAPW scheme, provides a practical “gold standard” for band‑gap prediction in II‑VI semiconductors. The authors suggest future work to extend the methodology to strained systems, alloyed compounds, and to benchmark its performance under external perturbations such as pressure or electric fields. Their findings open the door for reliable, high‑throughput computational screening of optoelectronic materials based on II‑VI chemistry.