Chiral Amines via Enantioselective π-Allyliridium-C,O-Benzoate-Catalyzed Allylic Alkylation: Student Training via Industrial-Academic Collaboration.

Abstract

Cyclometalated π-allyliridium-C,O-benzoate complexes discovered in the Krische laboratory display unique amphiphilic properties, catalyzing both nucleophilic carbonyl allylation and electrophilic allylation of diverse amines as well as nitronates. Given the importance of chiral amines in FDA-approved small-molecule drugs, a collaboration with medicinal chemists at Genentech that included on-site graduate student internships was undertaken to explore and expand the scope of π-allyliridium-C,O-benzoate-catalyzed allylic amination and related processes. As described in this Account, our collective experimental studies have unlocked asymmetric allylic aminations of exceptionally broad utility and scope. Specifically, using racemic branched alkyl-substituted allylic acetate proelectrophiles, primary and secondary aliphatic or aromatic amines, including indoles, engage in highly regio- and enantioselective allylic amination. Additionally, unactivated nitronates were found to be competent nucleophilic partners for regio- and enantioselective allylic alkylation, enabling entry to β-stereogenic α-quaternary primary amines. Notably, these π-allyliridium-C,O-benzoate-catalyzed allylic substitutions, which display complete branched regioselectivity in reactions of alkyl-substituted allyl electrophiles, complement the scope of corresponding iridium phosphoramidite-catalyzed allylic aminations, which require aryl-substituted allyl electrophiles to promote high levels of branched regioselectivity. Computational, kinetic, ESI-CID-MS, and isotopic labeling studies were undertaken to understand the mechanism of these processes, including the origins of regio- and enantioselectivity. Isotopic labeling studies suggest that C-N bond formation occurs through outer-sphere addition to the π-allyl. DFT calculations corroborate C-N bond formation via outer-sphere addition and suggest that early transition states and distinct trans effects of diastereomeric chiral-at-iridium π-allyl complexes render the reaction less sensitive to steric effects, accounting for complete levels of branched regioselectivity in reactions of hindered amine and nitronate nucleophiles. Reaction progress kinetic analysis (RPKA) reveals a zero-order dependence on allyl acetate, a first-order dependence on the catalyst, and a fractional-order dependence on the amine. As corroborated by ESI-CID-MS analysis, the 0.4 kinetic order dependence on the amine may reflect the intervention of cesium-bridged amine dimers, which dissociate to form monomeric cesium amide nucleophiles. Hence, the requirement of cesium carbonate (vs lower alkali metal carbonates) in these processes may reside in cesium's capacity for Lewis acid-enhanced Brønsted acidification of the amine pronucleophile. Beyond the development of catalytic processes for the synthesis of novel chiral amines, the present research was conducted by graduate students who benefited from career development experiences associated with training in both academic and industrial laboratories.