Can a Butadiene-Based Architecture Compete with its Biaryl Counterpart in Asymmetric Catalysis? Enantiopure Me-CATPHOS, a Remarkably Efficient Ligand for Asymmetric Hydrogenation

The double Diels-Alder cycloaddition between 9-methylanthracene and 1,4-bis(diphenylphosphinoyl)buta-1,3-diyne affords the oxide of the atropos diphosphine, Me-CATPHOS, which has an unusual bicyclic buta-1,3-diene-based architecture. Quantum chemical methods using DTF reveal that the barrier to atropinterconversion in Me-CATPHOS is 130 kJ mol(-1), while the corresponding barrier for its unsubstituted counterpart is only 23 kJ mol-1, entirely consistent with the former being an atropos diphosphine while the latter belongs to the tropos class of ligand. rac-Me-CATPHOS can be resolved by fractional crystallization of the diasteresomeric complexes formed with (2R,3R)-(-)-2,3-O-dibenzoyltartaric acid and reduction of the resulting enantiopure oxide, accomplished by silane reduction in xylene at 130 degrees C, to afford an operationally straightforward, three-step synthesis of an entirely new class of atropos buta-1,3-diene-based diphosphine. Rhodium complexes of enantiopure Me-CATPHOS catalyze the asymmetric hydrogenation of a range of dehydroamino acid derivatives, in some cases giving ee's in excess of 99% and in all cases showing a significant enhancement compared with its BINAP counterpart. Gratifyingly, Rh/(S)-Me-CATPHOS outperforms all existing catalysts for the asymmetric hydrogenation of (E)-beta-dehydroamino phosphonates, many of which are based on markedly more expensive biaryl- and ferrocenyl-based diphosphines. Surprisingly, in the case of the dehydroamino acid substrates, (S)-Me-CATPHOS provides product of opposite absolute stereochemistry to that obtained with (S)-BINAP, despite both ligands having an S-ax configuration, whereas (S)-Me-CATPHOS exerts (S)-BINAP-like stereoinduction for the hydrogenation of beta-enamidophosphonates; both ligands afford product with the same absolute configuration.