ACS PRF | ACS
All e-Annual Reports
43137-G1
[2,3]-Sigmatropic Rearrangements of Vinyl Oxiranes and Vinyl Oxetanes: New Construction of Medium Sized Rings and Macrocycles
We have initiated a research program focused on the chemoselective activation of cyclic and acyclic allylic ethers using electrophilic copper catalysts. We proposed that six and seven membered cyclic ethers could be assembled in a single operation by reacting a vinyl oxirane or a vinyl oxetane with a metallocarbenoid. An electrophilic copper carbene would react with the lewis basic oxygen of the strained oxacycle (oxirane or oxetane), thus generating and oxonium ylide ready to participate in a [2,3]-sigmatropic rearrangement affording the corresponding cyclic ethers. Our studies have shown that these proposed ring expansions are feasible. Suppressing competing C-H insertions and cyclopropantions turned out not to be challenging using copper carbenoids and all observed products resulted from the desired oxonium ylide. In the vinyl oxiranes case we have been unable to fully suppress deoxygenation. Vinyl oxetane are far better substrates for this reaction, although some competing Stevens rearrangement is observed. Vinyl-tetrahydrofurans ring expand in high yield. We have designed a new oxonium ylide based approach towards these same six and seven membered cyclic ethers, using acyclic allyl ethers substrates. We propose to chemoselectively activated these substrates by an electrophilic copper carbenoid, forming an oxonium ylide which would then undergo a [2,3]-rearrangement followed by ring closing metathesis (RCM). Although an additional synthetic step is added, the starting materials are much simpler and the potential for one pot rearrangement/RCM was considered an achievable goal. This approach represents a variant of our general program, one in which the protecting group has been replaced by a group that will be incorporated into the final structure. These two examples (R = allyl and homoallyl) represent only the tip of the iceburg in terms of possible variations.
Our new oxonium ylide based approach, using acyclic allyl ethers as substrates, has shown that competing cyclopropanation is easily suppressed by having the olefin substituted at the terminus. In our previous ring expansion route using vinyl oxirane our efforts were hampered by other competing oxonium ylide pathways, the Stevens rearrangement and deoxygenation. No Stevens rearrangement takes place when these acyclic allyl ethers are used as substrates, and products are formed in good isolated yields and diastereoselectivity. The resulting diene products can be readily cyclized to dihyropyrans by adding Grubbs second generation catalyst to the crude reaction mixture. Simply by using a homoallyl “protecting” group instead of an allyl group the corresponding seven membered oxepine rings can be accessed. Rearrangement and ring closing metathesis takes place with similar efficiency and selectivity as before, furnishing the oxepine ring in good yield. Our new route is clearly superior to our previous oxonium ylide mediated route to form six and seven membered rings. We are in the process of revisiting these substrates with our new catalyst. These results are examples of the incredible diversity of products one could access by varying the “protecting” group. We are in the process of writing an Organic Letters paper detailing our work in this area.
Future studies are focused on designing better catalysts for selectively activating and rearranging allylic ethereal oxygen atoms. The “Ideal” catalyst should be able to: 1) Efficiently decompose the diazo precursors, 2) Suppress carbenoid dimerization, 3) Chemoselectively activating the ethereal oxygen atom, 4) Form a metal associated oxonium ylide which rearranges readily, 5) Induce asymmetry. Our research program has made great headway towards accomplishing four of these five criteria. Our catalysts readily decompose the diazo compounds. Carbenoid dimerization is still a challenge, but our most successful catalyst (Cu(ptfmdbm)2) is breaking important ground in this area. Chemoselective activation of the ethereal oxygen atom has not been as challenging as we expected. We assume, based primarily on the difference in diastereoselectivity when different catalysts are used, that a metal associated oxonium ylide is involved in the rearrangement. We have not started addressing the fifth and final criteria, the development of a chiral copper catalyst.
Several students have been part of this important project in our group. Lindsay Batory is the graduate student in charge of this project. Undergraduates are an essential part of our research group, and several excellent undergraduates, Christopher Daeffler, Christine A. McGinnis and Eleanor Bucholz have worked with Lindsay on this project. Chris and Christine started their graduate studies in chemistry last fall at Cal. Tech (working for Grubbs) and Wisconsin Madison respectively. What started as a small project in our group, focusing only on oxiranes and oxetanes has grown into a much larger and more important research program focused on heteroatom activation. We are extremely grateful for the support from the ACS Petroleum Research Fund.
