Towards reliable hybrid functionals for fundamental and optical gaps of surfaces and bulk materials

ORAL

Abstract

Recently, the Wannier-localized optimally tuned screened range-separated hybrid functional (WOT-SRSH)[1] approach has been proposed as a means to compute reliable fundamental gaps for solids. The WOT-SRSH has been benchmarked on a wide set of semiconductors and insulators [1], lead halide perovskites[2], halide double perovskites, and metal oxides [3]. It has reproduced the fundamental gaps, linear optical absorption spectra, and band structures of such systems with low mean absolute errors (for the systems with available experimental data). However, benchmarking this functional for 2D materials and surfaces has yet to be investigated, which raises new challenges. Here we check the transferability of the tuned bulk parameters for predicting the fundamental and optical gaps of two well-studied reconstructed surfaces - Si (111)-(2x1), and Ge (111)-(2x1). We show that bulk parameters used with this functional can accurately reproduce the fundamental, surface state gap, and optical gaps for these materials. We compare our results with other semilocal and hybrid functionals, and assess the extent to which hybrid functionals can accurately reproduce gaps at surfaces.

[1] D.Wing et al.,PNAS, e2104556118 (2021)

[2] G.Ohad, et al., Phys. Rev. Mater. 6, 104606 (2022)

[3] G.Ohad, et al. arXiv:2309.02117

* The authors acknowledge NSF-BSF DMR-1708892, the Azrieli Foundation for an Azrieli International Postdoctoral fellowship, and ACCESS at TACC for computer allocation under TG-DMR190070.

Presenters

  • Francisca Sagredo

    Lawrence Berkeley National Laboratory

Authors

  • Francisca Sagredo

    Lawrence Berkeley National Laboratory

  • Maria Camarasa Gomez

    Weizmann Institute of Science

  • Francesco Ricci

    Lawrence Berkeley National Laboratory

  • Sijia Ke

    University of California, Berkeley

  • Leeor Kronik

    Weizmann Institute of Science

  • Jeffrey B Neaton

    Lawrence Berkeley National Laboratory and UC-Berkeley