Density sensitivity of empirical functionals

Suhwan Song, Stefan Vuckovic, Eunji Sim, Kieron Burke

Research output: Contribution to journalArticlepeer-review

Abstract

Empirical fitting of parameters in approximate density functionals is common. Such fits conflate errors in the self-consistent density with errors in the energy functional, but density-corrected DFT (DC-DFT) separates these two. We illustrate with catastrophic failures of a toy functional applied to H2+ at varying bond lengths, where the standard fitting procedure misses the exact functional; Grimme's D3 fit to noncovalent interactions, which can be contaminated by large density errors such as in the WATER27 and B30 data sets; and double-hybrids trained on self-consistent densities, which can perform poorly on systems with density-driven errors. In these cases, more accurate results are found at no additional cost by using Hartree-Fock (HF) densities instead of self-consistent densities. For binding energies of small water clusters, errors are greatly reduced. Range-separated hybrids with 100% HF at large distances suffer much less from this effect.

Original languageEnglish
Pages (from-to)800-807
Number of pages8
JournalJournal of Physical Chemistry Letters
Volume12
Issue number2
DOIs
Publication statusPublished - 2021 Jan 21

Bibliographical note

Funding Information:
The work at Yonsei University was supported by a grant from the Korean Research Foundation (NRF-2020R1A2C2007468 and NRF-2020R1A4A1017737). K.B. acknowledges funding from NSF (CHEM 1856165). S.V. acknowledges funding from the Rubicon project (019.181EN.026), which is financed by The Netherlands Organisation for Scientific Research (NWO). We thank Professor Martin Head-Gordon and Dr. Narbe Mardirossian for stimulating discussions.

Publisher Copyright:
©

All Science Journal Classification (ASJC) codes

  • Materials Science(all)
  • Physical and Theoretical Chemistry

Fingerprint Dive into the research topics of 'Density sensitivity of empirical functionals'. Together they form a unique fingerprint.

Cite this