Comment on "Quasi-One-Dimensional Metal-Insulator Transitions in Compound Semiconductor Surfaces"

TL;DR

This paper challenges previous DFT-based claims of Peierls-type metal-insulator transitions in semiconductor surfaces, showing that hybrid DFT and advanced methods predict an antiferromagnetic ground state instead.

Contribution

It demonstrates that the observed charge-density-wave phase is an artifact of GGA, and that hybrid DFT and EX+cRPA methods reveal an antiferromagnetic ground state.

Findings

GGA predicts a CDW phase due to artifacts.

Hybrid DFT and EX+cRPA predict an AFM ground state.

Reassessment of the origin of MI transitions in these surfaces.

Abstract

In a recent Letter, Zhao et al. [1] reported the origin of quasi-one-dimensional metal-insulator (MI) transitions in compound semiconductor surfaces. Based on a density-functional theory (DFT) calculation within the generalized gradient approximation (GGA), they claimed that one-atom-wide metallic structures formed by a selective bonding of H or Li atoms to GaN(10-10) and ZnO(10-10) undergo the Peierls-type MI transitions, leading to a charge-density-wave (CDW) formation with periodic lattice distortion. However, we here demonstrate that such a CDW phase is due to the artifact of the GGA, while the antiferromagnetic (AFM) ground state is predicted by the hybrid DFT calculation and the exact-exchange plus correlation in the random-phase approximation (EX + cRPA). [1] J. Z. Zhao, W. Fan, M. J. Verstraete, Z. Zanolli, J. Fan, X. B. Yang, H. Xu, and S. Y. Tong, Phys. Rev. Lett. 117, 116101 (2016).