Published March 20, 2023 | Version v1
Dataset Open

Fermi energy determination for advanced smearing techniques

  • 1. Theory and Simulation of Materials (THEOS), and National Centre for Computational Design and Discovery of Novel Materials (MARVEL), École Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland
  • 2. Laboratory for Materials Simulations (LMS), Paul Scherrer Institut, 5232 Villigen PSI, Switzerland

* Contact person

Description

Smearing techniques are widely used in first-principles calculations of metallic and magnetic materials, where they improve the accuracy of Brillouin zone sampling and lessen the impact of level-crossing instabilities. Smearing introduces a fictitious electronic temperature that smooths the discontinuities of the integrands; consequently, a corresponding fictitious entropic term arises and needs to be considered in the total free energy functional. Advanced smearing techniques – such as Methfessel-Paxton and cold smearing – have been introduced to guarantee that the system's total free energy remains independent of the smearing temperature at least up to the second order. In doing so, they give rise to non-monotonic occupation functions (and, for Methfessel-Paxton, non-positive definite), which can result in the chemical potential not being uniquely defined. We explore this shortcoming in detail and introduce a numerical protocol utilizing Newton's minimization method that is able to identify the desired Fermi energy. We validate the method by calculating the Fermi energy of ~20,000 materials and comparing it with the results of standard bisection approaches. In passing, we also highlight how traditional approaches, based on Fermi-Dirac or Gaussian smearing, are actually equivalent for all practical purposes, provided the smearing width is appropriately renormalized by a factor ~2.565. This data set contains the AiiDA databases, scripts, and other data necessary for the reproduction of our results.

Files

File preview

files_description.md

All files

Files (10.8 GiB)

Name Apps Size
md5:a277075889a58179c538b47b045f46d4
1.0 KiB Preview Download
md5:d39e60cd472ede30680f3577bacedce0
6.3 GiB Preview Download
md5:5c0ccbb4b2271bff6fa43d04a2e3f578
195.0 KiB Preview Download
md5:a6675ae792ba1195cfbf1d8a198328cf
567.7 KiB Preview Download
md5:9513647274239cb9915e241a1b9250ff
16.8 MiB Download
md5:f073cd93c0d1f9882b991557caa514fc
18.1 MiB Download
md5:acd6ae3ccc1d47597f3b1b3c5462f55d
2.0 GiB Download
md5:c83f9633fea9c5c0bd66d53b052bae38
19.1 MiB Download
md5:16f32a0b0871b43362aa759a9c973754
13.4 MiB Preview Download
md5:3a80efaaffd23d5754f75c9731ebaa5e
321.5 KiB Preview Download
md5:ad36adc44c855f36d903a4f2afd5437e
155.3 KiB Preview Download
md5:2328486f0a8970b47ecbdbafc4eb42fe
703.5 KiB Preview Download
md5:55c1a832e8109598c8bdc50435c884e3
2.4 GiB Download
md5:ee93af0eaa06367768bd4a2516ec21e4
35.0 KiB Preview Download
md5:22755e794ba14d43b341edad3515a1f1
2.9 KiB Download

References

Journal reference (Paper in which the data is used and the method is explained.)
F. J. dos Santos and N. Marzari, Phys. Rev. B 107, 195122 (2023), doi: 10.1103/PhysRevB.107.195122

Preprint (Open version of the paper in which the data is discussed.)
F. J. dos Santos and N. Marzari, arXiv:2212.07988 (2023), doi: 10.48550/arXiv.2212.07988