Spin-States of Diastereomeric Iron(II) Complexes of 2,6-Bis(thiazolin-2-yl)pyridine (ThioPyBox) Ligands and a Comparison with the Corresponding PyBox Derivatives

Inorg Chem. 2021 Sep 20;60(18):14336-14348. doi: 10.1021/acs.inorgchem.1c01988. Epub 2021 Sep 2.

Abstract

This report investigates homoleptic iron(II) complexes of thiazolinyl analogues of chiral PyBox tridentate ligands: 2,6-bis(4-phenyl-4,5-dihydrothiazol-2-yl)pyridine (L1Ph), 2,6-bis(4-isopropyl-4,5-dihydrothiazol-2-yl)pyridine (L1iPr), and 2,6-bis(4-tert-butyl-4,5-dihydrothiazol-2-yl)pyridine (L1t-Bu). Crystallographic data imply the larger and more flexible thiazolinyl rings reduce steric clashes between the R substituents in homochiral [Fe((R)-L1R)2]2+ or [Fe((S)-L1R)2]2+ (R = Ph, iPr, or t-Bu), compared to their PyBox (L2R) analogues. Conversely, the larger heterocyclic S atoms are in close contact with the R substituents in heterochiral [Fe((R)-L1Ph)((S)-L1Ph)]2+, giving it a more sterically hindered ligand environment than that in [Fe((R)-L2Ph)((S)-L2Ph)]2+ (L2Ph = 2,6-bis(4-phenyl-4,5-dihydrooxazol-2-yl)pyridine). Preformed [Fe((R)-L1Ph)((S)-L1Ph)]2+ and [Fe((R)-L1iPr)((S)-L1iPr)]2+ do not racemize by ligand redistribution in CD3CN solution, but homochiral [Fe(L1iPr)2]2+ and [Fe(L1t-Bu)2]2+ both undergo partial ligand displacement in that solvent. Homochiral [Fe(L1Ph)2]2+ and [Fe(L1iPr)2]2+ exhibit spin-crossover equilibria in CD3CN, centered at 344 ± 6 K and 277 ± 1 K respectively, while their heterochiral congeners are essentially low-spin within the liquid range of the solvent. These data imply that the diastereomers of [Fe(L1Ph)2]2+ and [Fe(L1iPr)2]2+ show a greater difference in their spin-state behaviors than was previous found for [Fe(L2Ph)2]2+. Gas-phase DFT calculations (B86PW91/def2-SVP) of the [Fe(L1R)2]2+ and [Fe(L2R)2]2+ complexes reproduce most of the observed trends, but they overstabilize the high-spin state of SCO-active [Fe(L1iPr)2]2+ by ca. 1.5 kcal mol-1. This might reflect the influence of intramolecular dispersion interactions on the spin states of these compounds. Attempts to model this with the dispersion-corrected functionals B97-D2 or PBE-D3 were less successful than our original protocol, confirming that the spin states of sterically hindered molecules are a challenging computational problem.