Aggarwal, Robinson. ACIEE, 2010, EarlyView. DOI: 10.1002/anie.201003236
A question – how long does one spin-out natural products based on one particular method? I ask this as this is the second synthesis I’ve written about this year by Aggarwal, both using his lithiation–borylation–allylation sequence (the first one was in Chemistry World in June). Aggarwal’s work is still very fresh – that’s why it’s on it’s second appearance here – but other groups have been stuck in a methodological rut for sometime (not wanting to name names here, but this is certainly true of some macrolide obsessives…).
Putting that question aside for now, Aggarwal’s work with this chemistry continues to impress me, along with the targets he chooses to make with it. His heritage (working for Stuart Warren at Cambridge) betrays it’s self early in the synthesis, with a really neat enantioselective cyclopropane formation. Working from an allylic alcohol derivative, an SAE gave them an epoxide, which after treatment with base and acetonitrile, gave them the cyclopropane. Really neat work, leaving some interesting functionality in the product.
A second fragment, required for the lithiation–borylation–allylboration sequence, also required installation of a stereodefined stannyl group – something I’d have thought would be quite tricky, but was actually done with ease. Using a commercially available starting material, the group formed a carbamate and then did a substrate directed lithiation and capture of the stannyl group. The yield might not be stunning, but the diastereoselectivity is impressive, so I’m sure they were very happy with this – even though I’m sure the lab stank…
With this fragment set to go, they were ready to try out their new methodology. However, the chemistry didn’t go quite to plan – as they only achieved a 15% yield. Analysis of the reaction mixture showed that the lithiated carbamate was attacking the aldehyde directly, and that the reaction with the vinyl borane was incomplete. However, by reordering the addition of TMEDA, they were able to get past this problem and boost the yield to 73%, and in a very respectable diastereoselectivity.
To complete the molecule, they still had a few steps to go; firstly they did a selective epoxidation of the allyl silane using Ti(OiPr)4/t-BuOOH. Treatment of this fragile intermediate with a bit of acid promoted a rearrangement and delivery of the target. Top work – and I’m looking forward to seeing this methodology again!