DFT Calculations on the Spin-Crossover Complex Fe(salen)(NO):  A Quest for the Best Functional

Abstract

DFT calculations on the spin-crossover complex Fe(salen)(NO) provide a striking illustration of the comparative performance of different exchange-correlation functionals vis-à-vis the issue of transition metal spin state energetics. Thus, although the "classic" pure functionals PW91 and BLYP favor the S = 1/2 state by about 10 kcal/mol, relative to the S = 3/2 state, the hybrid functional B3LYP favors the latter state by nearly the same margin. In contrast, the newer pure functionals OLYP and OPBE, based on the OPTX exchange functional, as well as the B3LYP* hybrid functional (which has 15% Hartree-Fock exchange, compared with 20% for B3LYP) predict nearly isoenergetic S = 1/2 and 3/2 states, as required for a spin-crossover complex. Intriguingly, the OLYP and B3LYP* spin density profiles for the S = 1/2 state of Fe(salen)(NO) are substantially dissimilar.

Spin in density‐functional theory

“…q e v nuc ðr i Þ ¼T þV ee þV nuc (37) According to this structure of the Hamiltonian, the energy expectation value in the above minimization is usually split up as…”

“…[29][30][31][32][33][34][35][36] Moreover, the spin density-which serves as an additional fundamental quantity in the spin-DFT formalism commonly employed for open-shell systems-is qualitatively incorrect in some cases. [15,[37][38][39] To make things even worse, the treatment of low-spin states usually requires the use of a broken-symmetry description, [40][41][42][43] which provides an unphysical spin density by construction (see, e.g., Refs. [24,44] for a discussion).…”

Iron Pincer Complexes as Catalysts and Intermediates in Alkyl–Aryl Kumada Coupling Reactions

“…33,51 The later functional feature OPTZ exchange, which assists in the accurate reproduction of energies of inorganic complexes with different spin states. [52][53][54] The M06/def2-SVP geometries of relevant compounds were then recomputed as single point energies using a density-dependent dispersion correction [35][36][37][38] appended to the PBE0 33,34 functional (PBE0-dDsC) with the triple- slater-type orbital TZ2P basis set in ADF. 55,56 Solvation corrections (in THF) employed the continuum solvent model for realistic solvents 41 (COSMO-RS), as implemented in ADF.…”

Spin‐State Energetics of FeII Complexes – The Continuing Voyage Through the Density Functional Minefield

“…[22] A second study by the latter authors added to these three systems CASPT2 results for the [Fe II (N H S 4 )L] series, [23] which indicated that OLYP performed better than PBE0 (25 % HF exchange), B3LYP, and B3LYP*, although all four DFT methods give identical trends (vide infra). Ghosh and coworker also favour OLYP, albeit they tested it on somewhat different systems including Fe III porphyrins, [24,25] as do Tewary et al [26] for the interesting case of computing magnetic exchange interactions and their possible effects on SCO, which involves "non-innocent" oxazolidine N-oxide iron nitroxide complexes.…”