Amidine synthesis by amine addition to nitriles normally requires high temperatures or harsh catalysts. Here, we report that boronate esters can facilitate amidination of proximal amines with moderate heating. With amidines present in a number of drugs and the synthetic handle provided by the boron, this chemistry should find useful applications.
Amidines are important pharmacophores in many drugs (e.g., pentamidine and quetiapine) and drug-like molecules.1 These can be constructed from the reaction of nitriles and amines with strong Lewis acid activation of the cyano group by catalysts such as AlCl3. 2 Reaction of nitriles with aniline in the presence of AlCl3 and BCl3 can result in electrophilic aromatic substitution by the nitrile ortho to the amine as originally reported by Sugasawa.3 The iminium intermediates can then be hydrolysed to the corresponding ketones. When this chemistry is carried out by addition of AlCl3 and BCl3 sequentially, amidination occurs first followed by boron substitution at the ortho position as outlined in Scheme 1.4 The boron is likely delivered in an intramolecular manner following complexation with the amidine. We have discovered an amidination reaction that creates the same amidine-boronate system through a different pathway and developed it into a simple synthetic transformation. Boron acids are useful catalysts through two different manifolds acting either as Lewis acids or as reversible covalent catalysts through exchange of the boron’s hydroxyl group with other ligands, a mechanism particularly useful in amidations.5 We have discovered a variation of the latter pathway to allow for chemoselective esterification of hydroxy-carboxylic acids and di-acids with boric acid.6 Ishihara has developed amidations of hydroxy-carboxylic acids using methylboronic acid as a catalyst with additional carboxylic acid necessary to promote the reaction.7 During our studies exploring possible intramolecular amide formation between 2-aminophenylboronic acid (1) and salicylic acid we noted formation of unexpected side products (Scheme 2). The first of these was an amide derived from reaction with ethyl acetate during attempted recrystallisation.8 The other was an amidine that formed in significant amounts when the reaction was run in acetonitrile. The structures of these compounds were each confirmed by X-ray crystallography and shown to be tetrahedral in nature at the boron through coordination with either the CvO (4) or CvN bond (5),‡ respectively. An amidine product was not observed when 3-aminophenylboronic acid and salicylic acid were heated in acetonitrile suggesting a cooperative interaction between the boron and the amine in these reactions.