Hypervalence and the delocalizing versus localizing propensities of H3-, Li3-, CH5- and SiH5-

Lithium and silicon have the capability to form hypervalent structures, such as Li-3(-) and SiH5-, which is contrasted by the absence of this capability in hydrogen and carbon, as exemplified by H-3(-) and CH5- which, although isoelectronic to the former two species, have a distortive, bond-localizing propensity. This well-known fact is nicely confirmed in our DFT study at BP86/TZ2P. We furthermore show that the hypervalence of Li and Si neither originates from the availability of low-energy 2p and 3d AOs, respectively, nor from differences in the bonding pattern of the valence molecular orbitals; there is, in all cases, a 3-center-4-electron bond in the axial X-A-X unit. Instead, we find that the discriminating factor is the smaller effective size of C compared to the larger Si atom, and the resulting lack of space around the former. Interestingly, a similar steric mechanism is responsible for the difference in bonding capabilities between H and the effectively larger Li atom. This is so, despite the fact that the substituents in the corresponding symmetric and linear dicoordinate H-3(-) and Li-3(-) are on opposite sides of the central atom.