Nature 501, 195–199 (2013)

One of the earliest pieces of mechanistic organic chemistry taught to undergraduates the world over is that SN2 reactions proceed with an inversion of the reacting centre. This principle is swiftly followed by the fact that substitutions at a tertiary centre do not occur through an SN2 mechanism and the alternative SN1 mechanism results in the formation of stereochemical mixtures. Such canonical reactivity is problematic when targeting the syntheses of a number of medicinally interesting marine natural products containing tertiary isocyanides. Now, Ryan Shenvi and co-workers from the Scripps Research Institute have developed a Lewis acid-catalysed solvolysis of trifluoroacetate esters (formed from tertiary alcohols) that proceeds with an inversion of the reacting centre.

Textbook descriptions aside, substitution reactions most likely occur by mechanisms along a continuum of pathways in which the SN1 and SN2 are the extremes (pictured). Shenvi and co-workers were convinced that they would be able to perform their desired stereoinversion by using a Lewis acid to activate the starting material towards formation of a carbocation (as in the SN1 mechanism) but then quickly trap the nascent carbocation before the leaving group had fully departed. This would ensure that the nucleophile could approach from only one face (as in the SN2 mechanism). To achieve this they planned to use a large excess of nucleophile while screening for a Lewis acid and activated ester combination that — even in the presence of this large excess of competing Lewis base — would favour activation of the ester starting material. The combination of trimethylsilyl cyanide as nucleophile with scandium triflate as Lewis acid acting on a trifluoroacetate ester was found to be just right.

The methodology is shown to be synthetically useful when starting with a variety of tertiary alcohols being converted to isocyanides — with onward conversion to amines also possible. Chemoselectivity for the reaction of tertiary over secondary and primary esters is observed with more sterically hindered centres reacting fastest — an inversion of the pattern seen for the SN2 reaction. The products, meanwhile, are the same as those that would be expected from a putative SN2 reaction. The high, but imperfect, inversion selectivity militates in favour of a reaction that proceeds through the rapid reaction of a contact ion-pair.