Qin, Shen and Zhang, ACIEE, 2008, EarlyView. DOI: 10.1002/anie.200800566.
Second showing in as many months for this popular alkaloid, this route from Yong Qin at the “Department of Chemistry of Medicinal Natural Products and, Key Laboratory of Drug Targeting, West China School of Pharmacy” (can’t get the website to work…) shows a few similarities to the recently covered route by Larry Overman, along with some interesting alternatives. Eschewing Overman’s ambitious palladium cascade designs, they went a bit old-school, planning a carbene insertion into an indole. However, like Overman, they hoped that the imine formed in this reaction would be attacked by a pendant amine to form the desired aminal center. Then, a rather familiar series of reactions would lead to product.
Let’s start by looking at the carbene insertion chemistry. Building the core of the indole from some chemistry developed by a former tutor of mine (Pat Bailey), they made a variety of substituted and protected indoles, and threw a few copper salts (and one rhodium catalyst) at it. Copper (I) triflate did the trick, and led them to product with a cracking yield.
Interestingly, with some substrates they obtained the product as a mixture of ketone and enol-ether isomers, where the proportion of one to the other was apparently a function of the protecting group on the indole nitrogen. The mixture of isomers, though, was dealt with swiftly by decarboxylating under Krapcho conditions to leave the unmolested ketone. I must say I’m amazed that 7h at 130 Â°C didn’t nuke the Boc group, though…
They then a little bother removing the tosyl group with the ketone present. Reduction of the carbonyl, removal of the Ts group and reoxidation was required, but soon they had a substrate bearing a remarkable similarity to that of Overman’s synthesis. The same palladium chemistry (initially developed by Cook) with a slightly different catalyst lead to the same result – formation of the five-membered ring. Then (surprise), Commins’ reagent gave them the corresponding enol-triflate, allowing appendage of the hydroxymethyl group by a Stille coupling – directly in this case using tri-n-butylstannylmethanol. Hmm.
And there we are; de-Boc and they’re done.