When does electronic delocalization become a driving force of chemical bonding?

A curve-crossing diagram model is used to conceptualize the stability trends in delocalized clusters X_n^z (n = 3, 4, 6; z = 0, -1; X = H, Li, Na, K, F, Cl, Br, I). The size of the diagram's gap (G) appears to be the decisive stability factor. Large gaps result in unstable (distortive) clusters with a high percentage of bond stretching and large quantum mechanical resonance energies (QMRE), e.g., H₆. Since G is related to the strength of the two-center X-X bond it follows that only atoms which form very weak bonds will generate stable delocalized clusters. Such clusters will possess slightly stretched bonds (on a percentage basis) and small QMRE's (e.g., Li₃,Li₆). The strength of the π-bond between carbon atoms indicates that π-components, of conjugated π-σ-systems, are distortive. This is confirmed by ab initio σ-π energy partitions for benzene and allyl radical. The results show that the species involve distortive π-transition states which are forced to be delocalized by the σ-frames. In accord with the general trend, the distortive π-components also possess large QMRE. It is shown that the Hückel 4n/(4n + 2) dichotomy is reflected in the corresponding QMRE values but not in the π-distortivity, and that the connection between "aromaticity-antiaromaticity" and geometry is not necessary. These conclusions carry over to other σ-π-systems, X_n (X = CH, SiH, N, P; n = 4, 6). The QMRE and the π-distortivities are two distinct molecular properties which must be probed by different experiments. Measurements of rotational barriers (in allyl) and thermochemical resonance energies are shown to be QMRE-related properties. Measurements of the π-distortivity are rare.