Bent Chemical Bonds

See allHide authors and affiliations

Science  14 Jun 1963:
Vol. 140, Issue 3572, pp. 1179-1185
DOI: 10.1126/science.140.3572.1179


The obvious and not original conclusion is that the results obtained from considerations of sp3 hybridization in the substituted methanes breaks down as soon as all of the bonding atoms are not equivalent. Indeed, the concept of sp3 localized carbon hybrids may be invalid even for tetrahedral molecules like CCl4 where the result of tetrahedral angles may be inferred from a naive hard sphere model. In other words the electrons in CCl4 have a more delocalized molecular quality than the apparently more localized atomic quality in CH4. This result implies that perhaps a Hartree-Fock treatment of the outer valence electrons in the substituted methanes (frozen inner atomic shells) might give a fairly good description of the electron distribution. On the other hand, the valence bond method would have to be carried well past the convenient approximation of perfect pairing. McWeeny and Ohno (23) were quite optimistic about the frozen inner atomic shell approximation in their treatment of H2O. As far as the strained small-ring compounds it appears the bending of a bond is a function of where one decides to draw the line between localized and delocalized character. Finally, no more appropriate ending could be contrived than a classic quote by Professor R. S. Mulliken (37), "I believe the chemical bond is not so simple as some people seem to think."