Snider, Zou, Chen and Taylor. Org. Lett., 2007, ASAP. DOI: 10.1021/ol070563g.
Another publication of this popular beastie, though with less drama than the Nicolaou paper, which I covered back in October. This is a formal synthesis, working towards an intermediate in the afore mentioned paper, using methods I feel are quite old-school, though completed in style. To help figure this one out, I’ve shown the target intermediate as drawn in both papers, with a quick retro on the enone.
The synthesis started quickly, completing the dione in good yield, although as a mixture of diastereoisomers. However, the group quickly realised that this decalin could epimerise easily, converting from the desired cis form to the trans. As luck would have it, the cis isomer was the more thermodynamically stable, and even though they only isolated 57% from the reaction, they were able to recycle/equilibriate the trans product. However, punching numbers into the big calculator lead them to realise that the target tricyclic product (also prone to epimerisation) was in a converse ratio. Thus, any thermodynamically-driven reaction would tend to give the wrong product. Fortunately, they were planning a radical reaction, and thus generated the desired product in 84%. A quick experiment with hydroxide showed that the computery thing was correct…
With this dione complete, reduction and closure of the conformationally closer hydroxyl onto the alkene generated the THF, leaving the other hydroxyl to triflate and eliminate. All that was remaining was to perform an allylic oxidation – but this was a bit more traumatic than one might expect. Efforts with chromium trioxide were unsatisfactory, as the allyl cation produced could oxidise on either end. They then moved to selenium dioxide, which either reacted too much, to produce the dienone, or the allylic alcohol. However, a little manganese dioxide allowed the latter to be converted to the desired enone in good overall yield.
I liked the paper a lot, and it certainly seems to be a better route than Nicolaou’s to that intermediate (Nicolaou did it in 10 steps, with considerably smaller yield). However, credit must go to Nicolaou, as his route did look great on paper, and had some top ideas (as well as being a good read). Both papers, however, deserve props for great writing and for explaining the full story. Snider gave several variations for this route, and explained their preferences in detail, which to me, is top :).