Jamison, Tanuwidjaja, Ng. JACS, 2009, ASAP. DOI: 10.1021/ja9052366.
Capitalising upon his pontifications analyses in Science (2007), Tim Jamison has come-up with the goods; a poly-epoxide cascade to build a poly-ether. He’s not the only one – Frank McDonald crafted a nice synthesis last year using the same principles. However, Jamison has upped the epoxide count, controlling formation of three ether rings in one go.
As ever, the initial trial is construction of the linear, poly-epoxy precursor – with few surprises in Jamison’s route. The starting material is farnesol, providing on three alkenes, one of which is allylic, so was a perfect candidate for Sharpless’ tasty catalysts. The homo-terminal olefin was deactivated by oxidation of the terminal methyl group to an allylic alcohol and acetylation, allowing selective Shi epoxidaiton of the internal olefin. It’s role complete, the acetate was removed to restore the isoprene-type terminus, and attention turned to the other end of the chain. An oxidation of the primary hydroxyl and olfination provided the last site for epoxidation, which was again accomplished via an SAE.
With the boring stuff done (…kidding!), it was time to let the molecular zipper rip – using NBS to provide bait for the cascade. Microsieves are essential, as Jamison notes in the Science paper, the presence of water can mess with the stereo- and regio-selectivity of the ring formations, whilst hexafluoroisopropanol is an interesting choice. He mentions that it’s highly polar and non-nucleophilic nature ‘facilitate[s] the presumably cationic cascade and thus maximize the directing influence of the methyl groups‘. However, I’ve also seen it used in many radical reactions where premature termination (!) is caused when the solvent has abstractable hydrogen atoms. Remember, that remaining OH is a homolytically strong bond.
I shouldn’t procrastinate about the solvent choice when such a nice result is evident – all three desired ethereal rings form in excellent yield, with the remaining cyclic carbonate acting as a useful synthetic handle. The only fly in the ointment is that the bromination is unselective, with a pair of equally prevalent epimers in the flask. I’m sure it would be possible to convert the stereocenter to the desired position, but rather than slow-down, the group pushed on to the end with a rather tasty sp2-sp3 Suzuki coupling to stitch the THF moiety to the larger fragment. No loading of catalyst is given, and for some reason my computer doesn’t like the SI, so lets just hope it’s under 10%!
A really nice bit of work; setting up those cascade reactions must be quite nerve-racking!