Highly diastereoselective and enantioselective synthesis of α-hydroxy β-amino acid derivatives: Lewis base catalyzed hydrosilylation of α-acetoxy β-enamino esters.

Chiral a-hydroxy b-amino acid moieties are important structural components in a wide variety of biologically active compounds as well as natural products, of which the side chains of Taxol and its analogues are the most famous examples. Consequently, the synthesis of chiral a-hydroxy b-amino acid derivatives has attracted considerable attention. Some syntheses include the Sharpless asymmetric aminohydroxylation, asymmetric dihydroxylation, ring opening of chiral epoxides, asymmetric nitroaldol reactions, asymmetric Mannich reaction, asymmetric 1,3dipolar cycloaddition, and other transformations. Among the successful strategies developed for obtaining optically active a-hydroxy b-amino acid derivatives, those that lead to substrates containing certain functional groups are of great significance. Accordingly, it can be reasoned that direct asymmetric reduction of the corresponding C=N double bond in substrates would be the most straightforward way to construct chiral a-hydroxy b-amino acid derivatives. However, to the best of our knowledge, the research on the abovementioned reaction has not yet been reported. Recently, asymmetric reactions involving the Lewis base activation of Lewis acids has attracted much attention. Among these reactions, chiral Lewis base catalyzed asymmetric hydrosilylation of C=N double bonds has become an important approach to chiral nitrogen-containing compounds because of the mild reaction conditions, cheap reagents, and the environmentally benign nature of this transformation. Several groups have achieved impressive progress in this field. During our ongoing studies on chiral Lewis base catalyzed asymmetric hydrosilylation of C=N double bond compounds such as b-enamino esters, we envisioned that the design and synthesis of b-enamino esters bearing various functional groups on the a position would provide a wide range of precursors to a-substituted b-amino acid derivatives. Therefore, we first tried to introduce an acetoxy group to the a position of b-enamino esters so as to generate a-acetoxy benamino ester 1 in which every functional group was finely assembled as depicted in Figure 1. (1S,2S)-2-Amino-1-(4-nitrophenyl)propane-1,3-diol (2) is the intermediate of chloramphenicol. It is very cheap and easily accessible. The two hydroxy groups can undergo condensation with a ketone or an aldehyde to generate a six-membered ring. Thus it occurred to us that we could make a novel chiral, rigid picolinamide Lewis base catalyst through the same transformation. We reasoned that this rigid catalyst might be highly selective in promoting asymmetric hydrosilylation of C=N double bonds. Hence we synthesized catalysts through two facile steps. As can be seen in Scheme 1, 2 was condensed with picolinic acid to give amide 3. The two hydroxy groups of amide 3 were then condensed with formaldehyde or a ketone to generate the cyclic catalysts 4a–4d.