A Trip to the Density Functional Theory Zoo: Warnings and Recommendations for the User

Abstract

This account is written for general users of density functional theory (DFT) methods as well as experimental researchers who are new to the field and would like to conduct such calculations. Its main emphasis lies on how to find a way through the confusing 'zoo' of DFT by addressing common misconceptions and highlighting those modern methods that should ideally be used in calculations of energetic properties and geometries. A particular focus is on highly popular methods and the important fact that popularity does not imply accuracy. In this context, we present a new analysis of the openly available data published in Swart and co-workers' famous annual 'DFT poll' (http://www.marcelswart.eu/dft-poll/) to demonstrate the existing communication gap between the DFT user and developer communities. We show that despite considerable methodological advances in the field, the perception of some parts of the user community regarding their favourite approaches has changed little. It is hoped that this account makes a contribution towards changing this status and that users are inspired to adjust their current computational protocols to accommodate strategies that are based on proven robustness, accuracy, and efficiency rather than popularity. Journal compilation Ó CSIRO 2019 www.publish.csiro.au/journals/ajc Account RESEARCH FRONT BP86-D3-gCP/6-31G * RMSD ϭ 0.105 Å corrected for BSSE and dispersion corrected for BSSE Fig. 6. BP86/6-31G* (light blue) structures of a folded FGG conformer compared with an MP2 (large basis set) geometry (dark blue). The effects of adding the gCP and DFT-D3(BJ) corrections are shown. Root-mean-square deviations (RMSDs) with respect to MP2 (and a large basis set) are also shown. Reprinted with permission from ref. 134.

The Quest for Highly Accurate Excitation Energies: A Computational Perspective

“…28 One closely related issue is the selection of the exchange-correlation functional from an ever-growing zoo of functionals and the variation of the excitation energies that one can observe with different functionals. 29,30 More specifically, despite the development of new, more robust approaches (including the so-called range-separated [31][32][33][34] and double [35][36][37] hybrids), it is still difficult (not to say impossible) to select a functional adequate for all families of transitions. 8 Moreover, the difficulty of making TD-DFT systematically improvable obviously hampers its applicability.…”

Double hybrid DFT calculations with Slater type orbitals

“…For (meta‐)GGAs ( a c , os = a c , ss = a x = 0) and hybrids ( a c , os = a c , ss = 0) our selection is mainly based on their performance in the recent comprehensive benchmarks by Goerigk et al on the GMTKN55 database [ 57,58 ] and Mardirossian et al on the MGCDB84 database. [ 168 ] The only exceptions are the B3LYP and PBE0 hybrid functionals, which we included due to their continuing high popularity despite their now well established [ 9,57,162,168,264,277 ] average performance. As functionals based on the B95 [ 278 ] and B97 correlation functionals were not available to us (except for the hybrids ω B97‐X and B97 which were available to us through libxc [ 279 ] ), we excluded the popular hybrids B97M‐V [ 280 ] and ω B97X‐V [ 281 ] as well as the DHs ω B97X‐2, [ 282 ] ω B97M(2), [ 283 ] PWPB95, [ 162 ] and DSD‐PBEB95 from this study.…”

“…Due to its unrivaled price/performance‐ratio, [ 1 ] Kohn‐Sham (KS) [ 2 ] density functional theory (DFT), [ 3–5 ] is the workhorse of modern computational chemistry. [ 6–9 ] Despite recent tremendous progress in wave function (WF) theory, especially in the field of local Coupled Cluster (CC) [ 10–13 ] methods, [ 14–41 ] little doubt exist that this situation will not change in the foreseeable future.…”

Clam‐like Cyclotricatechylene‐based Capsules: Identifying the Roles of Protonation State and Guests as well as the Drivers for Stability and (Anti‐)Cooperativity

“…[57] It was shown to be the most accurate and robust hybrid functional and a good alternative to double-hybrid functionals for large systems in recently performed benchmark studies that were the largest of their kind. [37,[58][59][60][61] Like other vander-Waals-DFT methods, only two-body contributions to the dispersion energy are included. As we are also interested in analysing potential non-additive three-body effects, we followed the example by Risthaus and Grimme [62] and enhanced the van-der-Waals-DFT type dispersion correction with the three-body term known from Grimme's DFT-D3 correction.…”

“…Advances in Density Functional Theory (DFT) methods have reached a point that allows reliably predicting properties with very good to high accuracy -for a recent account on the "zoo" of DFT methods and how to determine reliable DFT-based strategies for applications, see Ref. 37 -and we will use such state-of-the-art methods to provide answers to the aforementioned open questions to guide future syntheses of these and related compounds.…”