{getToc} $title={Table of Contents}
Introduction
Delocalization in chemical bonds refers to cases where electrons cannot be considered as part of a single bond, and are instead spread over a few atoms. In typical nonpolar pi bonds, for instance, we would expect the pi electron density to be spread equally between the two atoms forming the pi bond; however, in delocalized pi bonds, it is observed that the electron density is spread over more than just the two nuclei of the atoms forming the pi bond. So, how do we explain why the electron density for delocalized bonds is not simply concentrated between the two atoms forming the bonds? The answer lies in canonical forms, where a molecule (on paper) can assume more than one bond structure, with the true molecule being a weighted average of all the canonical forms of this molecule. Note that we say weighted average; this is especially important, as we will see later, because evidently some canonical forms are ‘worth’ more than others and have a higher weightage in the structure of the true molecule.
For a clear explanation of delocalization, nothing is better than a picture showing the possible resonance forms of a delocalized bond. A great example of why certain resonance forms have a greater weightage, as well as to demonstrate canonical forms, is benzene. The two canonical forms of benzene (Fig. 1) both involve six pi electrons as well as three conjugated bonds. The three bonds constantly ‘shift’ around the ring (a concerted, one-step movement of pi electrons), creating two different structures. This is known as resonance, while the two different structures created by the movement of electrons are known as resonance (or sometimes canonical) forms. Recall that we mentioned that the true structure would be a hybrid (known as a resonance hybrid) of the resonance forms; in each of the two resonance forms, every pair of adjacent carbon atoms would form a pi bond. Thus the resonance hybrid would have the electron density of the pi electrons equally distributed between all the carbon atoms; this is represented by a circle showing the delocalization.
Fig. 1: Resonance forms of benzene.
Now, what about the bond order of benzene? The bond order refers to the number of bonding pairs of electrons between two atoms. Consider two adjacent atoms in benzene. In the first resonance form, we see that a double bond is formed between the two adjacent atoms, giving the carbon-carbon bond a bond order of 2. However, in the second resonance form, the same two atoms only form a carbon-carbon single bond, giving it a bond order of 1. Since the two structures should have similar stabilities (although this is not always the case), they have an equal weightage in the resonance hybrid, and the true structure of benzene. This means that the true bond order of each bond in benzene (every bond in benzene is equivalent) would be the mean of 1 and 2, which is 1.5. Note, however, that this is not the correct bond order for benzene, with the actual bond order being slightly less, due to the presence of ‘Dewar structures’. A Dewar structure refers to a resonance form of benzene where the three bonds are instead (1,4), (2,3) and (5,6) bonds, and it contributes 7.3% to the overall resonance hybrid while the contribution for each of the conventional structures we have discussed is 39%. However, we will not discuss Dewar structures in detail; more information about these structures can be found here.
The above treatment, known as the valence bond method, of the bonding in benzene is slightly problematic; in the valence bond theory, bonds are considered two-electron two-center, making them more suitable to explain localized bonding rather than delocalized bonding. A further problem of this treatment is revealed when we consider that bond orders refer to the number of bonding pairs of electrons; since we calculated the bond order to be 1.5, it would mean that there are 1.5 bonding pairs of electrons, or 3 electrons, in one of the bonds in benzene, which is clearly untrue. For delocalized structures, a better treatments comes from the molecular orbital method. By this method, each of the carbons in benzene are sp2 hybridized, resulting in trigonal planar geometry, which prefers bond angles of 120 degrees. Since benzene already has the correct 120 degree bond angles, this would mean that it is more stable for benzene to stay planar (a concept we have already explored in the previous article). Since one s and two p orbitals are used in bonding, one p orbital remains for each of the six carbon atoms. Interaction between the carbon atoms occurs equally, resulting in equivalent bonds in benzene.
Fig. 2: Structure of the molecular orbitals in benzene.
Examples
Benzene is only one example of many, many compounds which have delocalized bonds. In general, these compounds can be separated into four classes. The first class of molecules have doubly or triply conjugated bonds (double or triple bonds that are adjacent to each other). Benzene is an example that we have already discussed. However, while the molecule of benzene is completely conjugated, some molecules are only partly conjugated. In such cases, resonance exists as well. Conjugated species should not be confused with allenes such as propadiene, where two double bonds are present directly next to each other. For conjugated molecules, there can be a concerted movement of pi electrons from one carbon to an adjacent carbon, as we have seen in benzene. Because there is resonance in conjugated species, they are more stable than non-conjugated but structurally similar compounds, and this is especially so if the whole molecule is conjugated.
The second type of molecules also involves molecules with multiple bonds, but this time there is only a single double or triple bond, and it must be adjacent to a p orbital. While this may sound confusing at first, it can simply be regarded as a generalization of the first class of molecules. In the first class of molecules, one double bond would be adjacent to a carbon which is sp2 hybridized (it participates in a double bond as well since there is conjugation). The carbon that is sp2 hybridized will have an extra p orbital for bonding, thus it would fit into this second type of molecules (as well as the first type). The extra p orbital is so that resonance can occur to move the double bond’s pi electrons to interact with the vacant p orbital of the carbon atom. However, this second class can also include molecules such as vinyl chloride (Fig. 3), where the chlorine atom has a p orbital available for bonding as well.
Fig. 3: Resonance structures of vinyl chloride.
The third class of molecules, notably, is rarely discussed in detail, because the delocalization is relatively unique compared to the other examples; here, the delocalization occurs to a dative covalent bond. A dative covalent or coordinate bond refers to a unique type of bond occurring commonly for metal compelexes. The bond forms between the metal complex and the donor atom of the substituent (the substituent is known as a ligand), and this donor atom donates a lone pair of electrons to the metal ion (which has vacant orbitals to facilitate this bonding). Typically, the electrons from the bond are unevenly distributed throughout, but delocalization may occur to distribute electrons equally. We will not dwell too much on this topic; it appears there is very little chemical literature online that discusses this type of delocalization in detail. We will also not explain in specific terms the last class of molecules, which involve hyperconjugated molecules. Hyperconjugation (Fig. 4) refers to cases where sigma electrons interacts with the p orbitals of an adjacent atom, and the delocalization acts in a way similarly to that of pi electrons’ interaction with p orbitals.
Fig. 4: Example of hyperconjugation.