Column: Totally Synthetic
Barekoxide and barekol
Davies' entry point was a critical domino cyclisation reaction in Sarpong's original work. The chemistry works by firstly building a cis -divinylcyclopropane intermediate, which then undergoes a Cope rearrangement to create a seven-membered ring (overall, a formal [4+3] cycloaddition).3 Using a chiral dirhodium tetracarboxylate catalyst (Rh2[DOSP]4), Sarpong's group performed this reaction diastereoselectively, but with only modest yield and stereoselectivity (figure 1). Davies and his small team have quite some expertise in this area, and believed that their catalyst systems might work better.
Balancing yin and yang
Key to Davies' success was understanding two conflicting reaction modes. The opposing principles - the yin and yang of stereoselectivity - are substrate control and reagent control. Substrate control exploits a molecule's shape and conformation to favour one product, whereas reagent control is provided by an external source - a reagent or catalyst. When the two reinforce, driving the same outcome, fantastic selectivity is possible. However, in Sarpong's case the two were opposed, muddying the waters and leading to mediocre results.
Davies overcame the substrate bias by using a catalyst that exerts stronger control over the reaction (Rh2[PTAD]4). This improved both the yield and diastereoselectivity, and expanded its scope to include other ring systems - including the decalin system present in barekoxide and barekol. The trick to improving the catalyst is one that is quite familiar: make it bigger, so it has a greater steric influence. However, this has to be done with care - the position of the huge adamantyl group close to the metal centre is critical.
Working with a substrate derived from sclareolide, which like many terpenes is used as a fragrance, the team created a suitable diene for the reaction in three steps. Applying their improved catalyst, they achieved excellent yield and selectivity with a particularly challenging system (figure 2).
The group then moved on with the total synthesis, starting with an entirely substrate controlled hydrogenation. This selectively reduced the simple alkene and generated an additional stereocentre at the junction with the new ring. Further reduction with diisobutylaluminium hydride (DIBAL-H) targeted the ester, which when coupled with acidic cleavage of the silyl enol ether neatly formed an enone. Another dose of DIBAL-H reduced the enone to an allylic alcohol, but substrate control dictated that the new stereocentre was oppositely configured to that required in barekol.
Initiate the Gevorgyan protocol
Using a process that is entirely new to me, the Gevorgyan protocol4 - which sounds like the title of a book - the group converted the allylic alcohol into an epoxide using tri-perfluorophenyl borane and triethylsilane, followed by meta -chloroperbenzoic acid (m CPBA). This presumably works by firstly forming the borane-alcohol complex, reduction by the silane to an isomerised alkene, and then substrate controlled epoxidation (figure 3). This gave barekoxide in good yield, and a simple acid-mediated isomerisation of the epoxide regenerated the alcohol, this time correctly configured to complete the synthesis of barekol.
Paul Docherty is a science writer and blogger based in London, UK
1 Y Lian et al, J. Am. Chem. Soc., 2010, 132, 12 422 (DOI: 10.1021/ja103916t)
2 L C Miller, J M Ndungu and R Sarpong, Angew. Chem. Int. Ed., 2009, 48, 2398
3 H M L Davies and J R Denton, Chem. Soc. Rev., 2009, 38, 3061
4 V Gevorgyan et al, J. Org. Chem., 2000, 65, 6179
Read more from Totally Synthetic
External links will open in a new browser window