Roush and Va. JACS, 2006, ASAP. DOI: 10.1021/ja066663j.
Nope, this isn’t dÃ©jÃ vu (or the film), this really is another synthesis of amphidinolide E, the second this month (the first was by Lee in Korea). This time it’s Bill Roush from Scripps, using some their own methodology to construct that THF. So, first the retro:
As you can see, some of the disconnections are more obvious than others, but I assure you, their methodology isn’t “shoe-horned” in… The top-half of the molecule was discussed first, preparing the substrates for that funky [3+2], which is where I found my interest piqued initially. They did a silylallylboration of the aldehyde, referencing their earlier work on that procedure, which returned the product allylsilane in fantastic yield and d.r.
TES protection of this compound left them ready to perform the [3+2]. They used 2.5eq of the alkyne, achieving a reasonable yield of the product, and with great diastereomeric control. They mention that they recover the excess of alkyne, but the remaining aldehyde cyclotrimerises under the reaction conditions.
With this fragment in hand, they were ready to couple the next fragment. This might appear to be a relatively simple esterification, and that Yamaguchi conditions would have this done before breakfast tea, but apparently not. They found that under all conditions attempted, the diene-acid isomerised to the more conjugated material, leaving with a tricky problem. However, their solution is super-smart – protect the diene!! They used the iron complex shown to hold the diene unit in place, performed the modified Yamaguchi in great yield, and then oxidised the iron off.
RCM completed the macrocycle, leaving them with only the side-chain to append. A great synthesis of this rather popular molecule.