Why does delocalization (only) occur in molecules represented by resonance hybrids?

Question

Before I start this question, I am aware tha

1. Electrons in all molecules are delocalized to some extent
2. Delocalization is a the phenomenon and resonance is an attempt to explain it.

When I asked this question from my professor, this is the explanation he gave me:

In VBT, it is difficult to approximate the molecular orbital with
localized bonding orbitals, so the electronic configuration is a weighted average of the infinite possible electronic configurations i.e Ψ= Aψ1 +Bψ2+Cψ3......

For most molecules, the most stable lewis structure has the majority of the contribution and can sufficiently approximate the actual electronic by itself. However, for some molecules, there are multiple equally-stable lewis structures with an equal and significant contribution and each of them has to be taken into account. For e.g, for nitrate ion, Ψ=1/3 (ψ1 +ψ2 +ψ3) where ψ1,2,3 represent the 3 equal lewis structures. As a result, the electrons MUST be delocalised between the appropriate bonds

This explanation, in simple words, argues that since the 3 lewis strucutes are identical/indistinguishable, the electron density must be equal and thus delocalised between the bonds by symmetry.

However, this doesn't really explain the physical reasoning behind delocalization. Why is (significant) delocalization more energetically stable for some molecules but not for others? And why dpes delocalization always occur in molecules with some kind of symmetry between the possible atoms for a double/triple bond?

Answer 1

A short answer is that we mostly use resonance structures to represent delocalization when the bond orders or formal charges predicted by VBT are not approximately integers, simply because non-integer bond orders and charges are more cumbersome to draw. A typical example is a carboxylate, which we could draw with 1.5 bonds between both C-O pairs and -0.5 charge on each O, but it is simpler to just draw a resonance structure with one C=O double bond and one C-O single bond and a full -1 charge on one oxygen.

As others have noted, electrons are still delocalized across the entire molecule in molecules for which we do not use resonance structures. The difference is that we can approximate the bond orders in those cases as integers, so only one structure is needed to represent that.

Answer 2

I'll provide a partial answer: resonance and delocalization are invoked to explain the fact that certain molecules do not have specific bonding patterns expected based on Lewis bond theory. In particular, that theory would predict that some molecules exist as mixtures of symmetry-related isoenergetic structures with some symmetry-related bonds differing in bond order, while experiment (and more advanced theories) instead show that there is only one structure and bond order rather than many. One explanation is that a molecule somehow exists as a "mixture" of the various possible structures suggested by Lewis bond theory, exchanging ("resonating") between these forms.

Resonance is a phenomenon in which energy is exchanged between tuned systems (such as radio antennae). In the VB theory case it is simply implied that the energy of the electron(s) involved in resonance is similar in the postulated geometries and that there exists a path allowing the electrons to shift between bonds involved in resonance. This shifting does not happen in practice, since there is really only one structure, but it clarifies what structures contribute to resonance.

Resonance hybridization can be used to explain why molecules do not behave as simple Lewis bond theory predicts. The answer to the question "when can I invoke resonance hybridization" is to generate all possible Lewis structures and then see if any of them are related by symmetry or whether there is a method to interconvert them by shifting bonds. If the answer is "yes" then it makes sense to assume that the real structure is a "resonance hybrid" of the Lewis structures (as opposed to each Lewis structure being an isomer).

Consider for instance the carboxylate ion, which can be represented as a resonance hybrid. It is straightforward to draw two possible Lewis structures for the ion. For the two possible structures not to be isomers a path must be available for exchange of valence electrons from one to the second CO bond center, that is, there has to be a common atom (in this case the carbon) serving as a "transit point" for the electrons.

Electrons are quantum mechanical particle-wave entities, and the resolution of the problem is that more advanced theory is required to properly describe the electron distribution in molecules. Lewis bond theory was developed before the full complexity of QM was understood. A simple QM model suffices however to explain why the energy drops when you increase delocalization of an electron: the particle in a box. If you double the size of the box you reduce the energy of the electrons by a factor of four, increasing their stability.