Successes and failures of DFT functionals in acid/base and redox reactions of organic and biochemical interest

Abstract

Abstract:The performance of 18 different DFT functionals in the prediction of absolute and relative energies of organic and biochemical acid/base and redox reactions was evaluated, using MP2 extrapolated to the complete basis set limit and CCSD(T)/aug-cc-pVTZ energies as benchmark.Absolute reduction energies were predicted with relatively large average errors (2-4 kcal.mol -1 ) except for the best functional, PBE0 (1.3±1.2 kcal.mol -1 ). The DFT predictions of relative reduction energies afforded mean unsigned errors (2.1 kcal.mol -1 for the best functional, PBE0) which, although relatively large, are smaller than those obtained with MP2 with comparable basis sets (3.2 kcal.mol -1 ). This relatively poor performance was mostly due to significant underprediction of the reduction energy of a model disulfide bridge, and overprediction of that of a model quinone, as eliminating these reaction from the test set enables most of the tested hybrid-GGA functionals (B3LYP, B3PW91, B97-1, BHHLYP, PBE0, PBE1PW91, X3LYP), though not meta-GGA or meta-hybrid-GGA functionals, to achieve small mean unsigned errors (<1.5 kcal.mol -1 ). In the acid/base reaction test set, several hybrid-

SEI-forming electrolyte additives for lithium-ion batteries: development and benchmarking of computational approaches

“…Overall, the M06-2X functional was found to be the best performer, although the other functionals of the Minnesota family (M06L, M11, and MN12L) also provided good results. This is consistent with other reported studies that have benchmarked the DFT prediction of redox properties against highly accurate wavefunction methods [42–45]. …”

“…However, since then there has been further progress in this area of research, including a new family of Minnesota functionals [38–41]. Several redox-process benchmarking papers have already been published which show that some functionals, especially M06-2X, provide better results than previously thought [42–45]. However, while they generally show improved performance as compared to wavefunction-based methods, it is essential to compare the results they yield to concrete experimental data for additives used in battery electrolytes in order to identify the best methodology.…”

Evaluation of density functional methods on the geometric and energetic descriptions of species involved in Cu+-promoted catalysis

“…1, structures 4-7). As expected from previous studies of deprotonation of organic molecules [41,55], most density functionals afforded optimized geometries very similar to the MP2 geometry, although the precise energetics of the deprotonation had considerable errors at the BS1 level, which lacks diffuse basis functions ( Table 5). The larger BS2 basis set, which does include diffuse basis functions on the heavy atoms, yields a much better description of the anionic, deprotonated species 5 , thereby strongly reducing the absolute error in the energies.…”

“…DFT reaction energies are known to be quite stable after increasing the basis set triple-ζ-quality with diffuse functions on the heavy atoms and two sets of polarization functions. Inclusion of many additional polarization functions in these studies does not usually afford much better agreement with experiment or benchmarks [39][40][41]. MP2 is however much more sensitive to the completeness of the basis set, and for this theory level we computed the energies using aug-cc-pVDZ [38,42,43] and aug-cc-pVTZ [38,[42][43][44] basis sets combined with the SBKJ pseudo-potential for the core electrons of Cu and I.…”

On constitutional isomers and tautomers of oxadiazolones and their mono- and disulfur analogues (C2H2N2XY; X,Y=S,O)

“…However, for selected type of reactions and small variation of functional groups, the error can be reduced to ca. 1 kcal/ mol [39,40]. Nevertheless, a careful verification of the error of energy differences was beyond this study.…”