Predicting the relative stability of simple versus ansa-sandwich systems across groups: structure, bonding, and...

We examine theoretically the bonding and thermodynamic stability of a proposed class of tris(sandwich)benzene complexes with the general formula C(6)(RMR')(3). In these systems, a single central (benzene) ring is flanked by three distorted 18-electron sandwich fragments with M = Groups 6, 7, or 8 metals, and R and R' = Ph or cyclopentadienyl (Cp) substituents. Remarkably, the computed free energy changes for the binding of the metal atoms to the organic fragments in 1) the ansa-(C(6)(RMR')(3)) complexes and 2) the simple (RMR') sandwich complexes--for all the Groups 6, 7, and 8 metals--conform to the linear relationship: [ΔG(bind)(C(6)(RMR')(3))]/3 = β(1)(ΔG)[ΔG(bind)(RMR')]+β(2)(ΔG) (R=0.989). The bonding and relative stabilities of these unusual tris(sandwich) and simple sandwich complexes are assessed at the B3PW91 level of theory employing small-core relativistic MDF pseudopotentials and the corresponding basis sets for all the metals. The possible existence of strain-induced bond localization (the so-called Mills-Nixon effect) in the C(6)(RMR')(3) complexes and the aromaticity at the central benzene ring in C(6)(RMR')(3) are investigated. Despite the strain on the central ring, no bond fixation is observed. The tris(sandwich) complexes of Groups 6, 7, and 8 are all thermodynamically stable, relative to the free M atoms and organic fragments, with increasing stability as M gets heavier (going down the groups). The equation above also enables us to predict, by extrapolation, the (in)stabilities of the C(6)(RMR')(3) complexes of several other transition metals and helps us to better understand thermodynamic aspects of the ansa effect. Similar functions likely apply for other ansa variants of sandwich complexes.