# Explaining hypervalency in the framework of molecular orbital theory

In preparing for teaching duties in an introductory course in inorganic chemistry, I seem to have made a digression into how we understand the hypervalent molecules.

Just to re-iterate what this concept is (most likely a bad definition, correct me if this is bad): We say that a molecule is hypervalent if an atom has more than its maximum number of valence electrons participating bonding, as drawn in Lewis structures. Three examples are $$\ce{PCl5}$$, $$\ce{SF6}$$, and $$\ce{XeF2}$$.

When reading up on this, it seems to me that mostly the valence bond framework is used to explain hypervalent molecules. There is talk of ionic configurations accounting for the majority of the energy, that of single electrons localized on the ligands preventing the "octet rule" from breaking.

But how do we explain this with concepts from molecular orbital theory? If I run a high level calculation of $$\ce{SF6}$$, my computer does not explode (presumably, I have not tried it).

Assuming that d-electrons do not participate in bonding to any significant extent, then we only have the valence electrons. If I add up the partial occupancies of the orbitals surrounding the central atom, will I then get a number very close to 8? So only 8 electrons will participate in bonding? The "remaining" electrons could then occupy nonbonding orbtils?

Is "hypervalent molecules" a non-issue in molecular orbital theory?

• chemistry.stackexchange.com/a/49844/16683 there’s an MO scheme here. Not exactly backed up by computational work, but I don’t think it will be too different – orthocresol Jun 4 '19 at 12:09
• A reasonably good overview can be found here Wikipedia - Hypervalent Molecule – Michael Lautman Jun 4 '19 at 13:17
• That MO diagram is definitely interesting. It is still unclear to me how exactly the two remaining F atoms are bonded to the S, since the two electron pairs are localized on the F atoms and do not participate in bonding. Are they supposed to be "ionic bonds"? – Yoda Jun 5 '19 at 10:00
• In MO theory there are no hypervalent molecules. Hypervalency is a figment of Valence Bond Theory, and the apparent misrepresentation only arises from localising electrons, and the inherent inability to describe delocalisation with one structure in the VBT framework (ref. also resonance). Since MOT uses delocalised descriptions, there cannot be such limitations. It might sound strange, but when you break it down in SF6, you basically have a 2-electron-7-centre bond (completely symmetric overlap) and three degenerate 2-electron-7-centre bonds (from the p orbitals). – Martin - マーチン Jun 7 '19 at 17:17