route to vinyl epoxides and vinyl aziridines via chiral sulfur ylides, we were keen to develop their potential in synthesis further. It had been reported that palladium-catalyzed reactions of vinyl epoxides and aziridines with doubly activated Michael acceptors gave tetrahydrofurans and pyrrolidines in good yield but usually with poor stereocontrol. However, related reactions with singly activated Michael acceptors were not effective. Nevertheless, we recognized that if we could find a way of coercing vinyl aziridines to react with singly activated enones/acrylates, and we were able to control stereochemistry in the annulation process, then we could potentially utilize this methodology in synthesis. Herein, we describe our success in simultaneously meeting these two significant challenges and also describe its application in a formal synthesis of ( )-a-kainic acid 1. Our initial efforts at promoting reaction between vinyl aziridine 2a and methyl vinyl ketone (MVK, A) using [Pd2(dba)3·CHCl3], (p-FC6H5)3P in THF (conditions employed by Yamamoto), however, were fruitless—we only observed decomposition. Using trimethylsilyl-substituted trans vinyl aziridine 2e, we now observed isomerization to a mixture of trans/cis aziridines (1:20). With this substrate, clearly the Pd was performing its role in generating the p-allyl palladium complex as this resulted in isomerization of the vinyl aziridine. However, the amide anion that was generated did not react with the enone. This result reinforced the observations by Yamamoto and Knight that doubly activated Michael acceptors were required to capture the relatively unreactive amide anion. We reasoned that ion pairing of the amide anion with the cationic Pd complex might compromise its reactivity, and that increased nucleophilicity should therefore be possible with an anionic Pd complex instead (Scheme 2).
Read full abstract