Observation of the slow racemization of isobornyl chloride in a polar solvent in 1923-24 by Meerwein led to the recognition that mechanistic interpretation is the key to understanding chemical reactivity. The hypothesis of ion pairs in which a chloride anion is partnered by a carbocation long ago entered the standard textbooks (see DOI 10.1021/ed800058c and 10.1021/jo100920e for background reading). But the intimate secrets of such ion-pairs are still perhaps not fully recognised. Here, to tease some of them them out, I use the NCI method, which has been the subject of several recent posts.
To remind, the colour coding of the NCI surface is blue=strongly attractive, red=strongly repulsive, green=weakly attractive, yellow=weakly repulsive. Shown above is the ion-pair transition state for [1,2]methyl migration. Note how the hydrogen bonds between the chloride anion and the water molecules are clearly blue. Only slightly weaker (with a turquoise tint) is a pair of hydrogen bonds between the oxygen atoms and H-C bonds in the isobornyl cation. Such C-H…O bonding in ion-pairs seems to be particularly important. There are other blue regions, between an H…H pair, and a C-H bond and the carbon of the migrating methyl group. Also noteworthy is that many atom pairs have multi-coloured NCI regions, suggesting the interaction is not homogenous, and can be both attractive AND repulsive between any pair of atoms.
The NCI plot below shows the competing 1,6-hydride shift in isobornyl chloride, again involving an ion-pair transition state.Notice in this example how the migrating hydrogen supports an attractive hydrogen bond to the chloride anion (ostensibly between a hydride atom and an anionic chloride?), and again how there are a number of blue regions elsewhere.
Modelling is increasingly focusing on these weaker interactions, that probably mediate much (stereo)selectivity in organic reactions. How long before such approaches themselves enter the text-books?