Ley, Francais, Leyva, Etxebarria-Jardi. Org. Lett., 2009, ASAP. DOI: 10.1021/ol902676t.
For those UKian readers out there, there’s a fittingness to my posting a Cambridge paper following an Oxford last week. Having spent a bit of time at both, it’s hard for me to pick a particular allegience, but I guess I have to go light blue, and congratulate the boys for their dug-in performance at Twickenham a few weeks back. But I am particularly lucky to have been at both, as their approach to organic chemistry was so markedly different. Both universities might be perpetually mentioned in the same breath (or word, the horriffic OxBridge, cause of many a confused tourist in Uxbridge), but the departments couldn’t be more different. Simply consider the staffing rosters: Ox Cam. And that difference will only get more dramatic in the near future, as Steve Ley heads for the retiral office (something I’m still uncertain of…). However, he’s a long way from done, as this new Org. Lett. points out.
Bongkrekic acid is a scene of much synthetic action, with an early triumph by Corey, and a less satisfying effort by Ley last year. In this paper, Ley admits that their previous effort was somewhat crippled by an ill-fitting methodology, and that this most recent synthesis is an attempt to right this. And quite success it is too, focusing on using the best methods available for each coupling. The retro outlines that this paper is indeed coupling-tastic, using quite a few pots from Strem. But what was often more interesting is how the group made the coupling precursors, with a particular emphasis on installing vinyl iodides. First up is actually a vinyl-stannane, installed using a stereoselective Piers hydrostannylation, nabbed from this JOC paper. But what was of more interest to me is a simple piece of convergence I haven’t read before – concommitant deprotection of a TBS ether and oxidation directly to the acid using Jones reagent. Neat.
And we’re into our first vinyl iodide synthesis, and it’s a prep I’ve used before. You take your acetylene in hexane, bang in some DiBAL-H, and vac-off the solvent. The DiBAL-H, meanwhile, has added across your triple-bond in a hydro-alumination, and you isolate this vinyl-alane – not a particularly nice species. This is then quenched by addition of iodine in THF (better hope it’s dry), giving a pretty decent yield of vinyl iodide, stereoselectively. This was all well-and-good when I was doing this on a small scale in the lab at Cambridge, but when I moved for my industrial placement at AstraZeneca, the lab-saftey people went nuts, so I ended-out getting a custom synthesis done at their expense…
One iodine isn’t enough here, though – a gem-diiodide (looks real-stable, right?) was formed by oxidising the alcohol to the aldehyde, and then adding hydrazine and iodine. The details are apparently in this 1970 J. Aust. Chem. paper, but I don’t have access – anyone like to elaborate?
Okay, time to couple something – and using what Ley refers to as the Stille-Migita coupling. Using CuTC in Stille type couplings is something I’ve written quite a bit about, but I’d always attributed the reaction name to Liebeskind. A difference is that in most of the Liebeskind work, copper is the sole metal, whereas Ley uses a spot of palladum too. He also uses a phosphonate base, and explains it’s role as that of tin scavenger. I wonder if it a) makes the reaction smell less evil and b) allows for a less-streaky column. One can only hope…
Time to make another vinyl iodide, and this time the chemistry is old school. Using diethyl methyl malonate (which smells so nice it’s really difficult not to eat it…), a little base and some iodoform, a displacement of one iodine seems to be order of the day. Following this with a bit more base, and we get a decarboxylation, saponification (of the remaining ester) and elimination, binning the penultimate iodine and generating the vinyl iodide.
This little fragment was coupled up in a Suzuki coupling, quickly generating a diene and finally getting the group ready for a bit of non-flatland chemistry. Their intention was to do an asymmetry propargylation, and they opted for an Indium mediate reaction using Singaram’s chemistry, which uses an ephedrine-like ligand to provide asymmetry. The resulted in a cracking yield, but also a disappointing enantiomeric ratio. The group managed to boost this with some of Greg Fu’s work, using a planar-chiral DMAP-ferrocene catalyst (commercially available) to do a non-enzymatic resolution. More about that in this Chem. Comm.
And now we get to the final coupling, a Sonogashira using a vinyl-iodide. Typical conditions using a bog-standard palladium source and similarly common copper salt resulted in a rather neat yield of the enyne. However, what they desired was of course the E,Z-diene, so a bit of reduction was in order. The most obvious choice, Lindlar catalyst, failed, giving the group a bit of a headache. I’ll presume that a few sets of conditions were examined for this reduction, as their success with copper/silver activated zinc seems a little obsure. I’d have tried diimide, but they didn’t ask me…
It did work, though, giving them a very respectable yield of the desired product (the tri-methyl ester of the natural product), along with some of the undesired Bongkrekic Acid. Separation at this stage, followed by saponification (without any isomerisation) gave the group a portion of each natural product, rounding-off a cracking read and an interesting synthesis.