Ring-opening C-N bond cleavage reactions provide an effective means to convert widespread, readily accessible chiral N-heterocycles into hard-to-attain stereodefined linear amines. Current strategies either rely on the strain-induced release of small aziridine and azetidine rings or, for larger ring systems, require highly electrophilic reagents, oxidative conditions, or preinstalled reactive functionalities to enable the ring-opening event. Recently, complementary radical strategies that exploit the reactivity of α-amino-ketyl radicals, formed upon single-electron transfer (SET) reduction of common N-carbonyl protecting groups, have emerged. Nevertheless, these methods facilitate the homolytic fragmentation only of up to 5-membered azacycles. In this study, we leveraged electroreductive conditions to switch the nature of the above C-N bond cleavage manifold from radical to ionic and enable the heterolytic ring-opening of a broad array of unstrained cyclic amines (comprising pyrrolidines, piperidines, azepines, azocanes, and N-macrocycles), protected as N-(thio)amides, carbamates, or ureas. Crucially, this electrochemically enabled reactivity switch grants complementary functional group compatibility and a broader ring size and N-carbonyl group scope. Computational and experimental studies indicate that electrochemical settings are crucial for generating the Mg(II)-Lewis acid catalyst, activating the N-carbonyl moiety while prompting the so-formed oxy-iminium ion intermediates to undergo two consecutive cathodic SET reductions, generating "umpoled" α-amino-α-oxy-carbanion species. These, via irreversible E1cB fragmentation of the adjacent C-N bond, lead to the desired ring-opened products. Our electrochemical procedure can be scaled up and miniaturized (enabling its application to high-throughput experimentation screening), and its synthetic utility has been demonstrated by accessing decorated stereodefined linear amides from stereochemically rich pyrrolidine and azepane derivatives.
Buscemi et al. (Sun,) studied this question.