High spin d 5 complexes of tris(6-hydroxymethyl-2-pyridylmethyl)amine (H3L): hepta-coordinated [Mn(H3L)]Cl2 and linear trinuclear [Fe3L2](ClO4)3

Reaction of MnCl 2·4H2O with H3L (H3L = tris(6-hydroxymethyl-2-pyridylmethyl)amine) in methanol gives hepta-coordinated [Mn(H3L)]Cl2 involving attachment of Mn(II) to all four nitrogens and three hydroxymethyl arms. Reaction of H3L with Fe(ClO4)2·6H2O in CH3CN in the presence of NaO2CC6H5 in an attempt to make [FeIIIOH(H3L)(O2CC6H5)](ClO4), a putative model for soybean lipoxygenase-1, instead gave rise to the linear triiron(III) complex [Fe3L2](ClO4)3 with all three hydroxymethyl arms deprotonated and forming three alkoxide bridges between each Fe(III) centre. The central Fe(III) is hexa-coordinated to only the alkoxide bridges and flanked by two hepta-coordinated iron(III) centres analogous to the Mn(II) complex. [Fe3L2](ClO4)3 exhibits two reversible 1e− reductions to mixed-valence [Fe3L2]2+ and [Fe3L2]+ forms. Structure data and magnetochemistry on [Fe3L2](ClO4)3 reveals the tightest Fe–O–Fe angle (87.4◦) and shortest Fe · · ·Fe distance (2.834 A˚) yet found for any weakly antiferromagnetically-coupled high spin alkoxide-bridged di- or triiron(III) system and challenges current theories involved in correlating the extent/nature of magnetic interactions in such systems based on Fe–O(bridge) distances and Fe–O–Fe angles. The central hexa-alkoxide coordinated Fe(III) is novel and shows a remarkable resistance towards reduction to Fe(II).