Synthesis of O-Heterocycles : the 2,3-sigmatropic rearrangement of transition metal carbenoid-generated allylic oxonium ylides

摘要

In many natural products, heterocycles, such as cyclic ethers, are important features. While there are many different approaches to the formation of these ethers, there are only a few methods available for the synthesis of substituted cyclic ethers with good diastereoselectivity and even less so with good enantioselectivity. The rearrangement of oxonium ylides has proved to be a versatile method for the stereoselective synthesis of cyclic ethers. However, to date, there is no general efficient enantioselective method for the rearrangement of oxonium ylides. This project aimed to develop an enantioselective synthesis of O-heterocycles from chiral copper carbenoids. Screening of various catalysts generated in situ from [Cu(MeCN)4]PF6 and chiral ligands led us to identify a class of ligands, specifically chiral bisoxazoline ligands, that generally resulted in asymmetric induction during the [2,3]-sigmatropic rearrangement of oxonium ylides. Unfortunately, while asymmetric induction was obtained, generally the rearrangement reaction resulted in quite low enantiomeric excess. During the course of this project, iridium-mediated reactions was also investigated. It was found that the catalyst [Ir(COD)Cl]2 could be used for the same transformation as the copper catalysts, which to the best of our knowledge was the first example of the use of an iridium catalyst for the tandem oxonium ylide generation and subsequent [2,3]-sigmatropic rearrangement of diazoketones. As only rather low asymmetric induction was obtained during this transformation, our attention turned towards achieving a greater mechanistic understanding of this reaction, as good mechanistic understanding is essential for enantioselective development. Isotopic labelling of the diazoketone starting materials provided information on the rearrangement products, from which conclusions could be drawn as to the rearrangement mechanism. It was concluded that the rearrangement reactions in question, that take place via copper or iridium carbenoid-mediated reactions, either do not proceed through a free oxonium ylide, but rather through the metal-associated oxonium ylide derivative, or follow a major competing non-ylide route that delivers apparent [2,3]-sigmatropic rearrangement products of oxonium ylides. With regard to rhodium-catalysed reactions, firm conclusions could not be drawn, although there is some suggestion that this reaction also does not proceed solely through the free oxonium ylide pathway. Further investigations of the iridium-catalysed reaction through crossover experiments suggest that the metal-associated oxonium ylide derivative dissociates during the reaction to give an allylic cation and an iridium enolate, which then recombines to give the apparent [2,3]-rearrangement product.