6.3.1 Refinements and difficulties
In Section 6.2, we said that inter-axis repulsions vary in the order:
non-bonded pair-non-bonded pair > non-bonded pair-bond pair > bond pair-bond pair
There is evidence for this in the inter-bond angles in molecules. For example, in water and ammonia (Structures 7.4 and 7.5), the bond angles are about 5° and 2° less than the tetrahedral angle of 109.5°.
Does this support the quoted order of inter-axis repulsions?
Yes; non-bonded pair-bond pair repulsions tend to reduce the inter-bond angle; bond pair-bond pair repulsions tend to increase it. The observed reduction shows that non-bonded pair-bond pair repulsions are dominant. The reduction is greater in H2O than NH3 because the water molecule has two non-bonded pairs.
Similar effects suggest that there are differences in the repulsive effects of single, double and triple bonds. As Lewis theory implies that these consist of one, two and three pairs of electrons, we might expect that their repulsive effects would vary in the order:
triple bond > double bond > single bond.
The geometry of the ethene molecule can be seen in Structure 7.9. The four outer electrons on each carbon atom are distributed between three repulsion axes: one double bond to carbon and two single bonds to hydrogen.
Do the inter-bond angles support our assumed difference in the repulsive effects of single and double bonds?
Yes; the stronger repulsion exerted by the C=C bond forces the two C—H bonds together. The inter-bond angle falls below 120°, the value for regular trigonal-planar coordination.
Recognition of the different repulsive effects of single and double bonds can therefore be useful in choosing a molecular shape or predicting bond angles.
Nevertheless, when step 5 of the procedure of Section 6.2 leaves us with two or more structures to choose from, it is sometimes hard to make an informed choice. Minimising the strongest repulsions is usually effective, but not always. A particular problem can arise when there are five repulsion axes - the trigonal-bipyramidal disposition. We can illustrate it with ClF3, a liquid that boils at 12 °C, and reacts with water with a sound like the crack of a whip. The central chlorine atom has seven outer electrons, three of which are used in forming the three Cl\F bonds. The other four electrons become two non-bonded pairs, which, with the three bonds, give us five repulsion axes disposed in the trigonal-bipyramidal arrangement. There are three possibilities (Structures 7.10-7.12):
In all three, the strongest repulsive interaction is of the non-bonded pair-non-bonded pair type. If we minimise this, we would decisively reject 7.10, where the non-bonded pair axes are at right-angles, and choose 7.12, where they both occupy the axial positions and so are at 180° to one another. This predicts a planar ClF3 molecule. But experiment shows that the correct structure is 7.11, where the non-bonded pairs occupy equatorial positions at 120° to each other. Indeed, it seems that in molecules based on the trigonal-bipyramidal disposition of repulsion axes, the non-bonded pairs avoid the axial, and occupy the equatorial sites. In this case, our recommended procedure must be modified.
Note that throughout this Section we have confined ourselves to typical element molecules containing an even number of valence electrons. The valence electrons can then always be divided into pairs, and each repulsion axis consists of a pair or pairs of electrons. But a few typical element molecules contain an odd number of electrons, and the application of VSEPR theory then forces us to deal with repulsion axes consisting of a single electron. An example of this sort is considered in Question 21 below.
Finally, you should recognise that the restriction of VSEPR theory to typical elements is important. It is very much less successful in predicting the molecular shape of transition-metal compounds.