Prediction of a Narrow Direct Bandgap in Tetragonal Zintl-Phase NaMgBi

Abstract

Although the Zintl compound NaMgBi has been synthesized in the tetragonal PbFCl-type phase, its accurate electronic structure is still unknown. This calculation is difficult because the conventional density functional theory (DFT) underestimates the bandgap, and the strong spin–orbit coupling (SOC) decreases the bandgap as well. Consequently, the usual DFT+SOC calculation predicts a false semimetallic band structure. Here, the GW approximation is used to calculate the accurate electronic structure of NaMgBi including the SOC. Our eigenvalue self-consistent ${\mathrm{GW}}_0$ (${\mathrm{EVGW}}_0$) results show that NaMgBi has a direct bandgap of 0.47 eV instead of an indirect one predicted by the DFT calculations neglecting the SOC. Although our calculations based on the ${\mathrm{EVGW}}_0$ results indicate that NaMgBi is trivial topologically, the solution of the Bethe–Salpeter equation shows that its exciton binding energy is 0.14 eV and the excitonic absorption coefficient can reach 10${}^5$–10${}^6$ ${\mathrm{cm}}^{-1}$ in the infrared and visible light range, making it a promising optoelectronic material.