Feeling a sense of déjà vu? Don't worry, it's not the solvents from your column. This is indeed our second examination of these targets, both dimers of another natural product: the ever-fascinating resveratrol. The latter has been the subject of much scientific discussion, as its naturally occurring oligomers possess a wide array of biological properties.¹ However, such fame has had unintended consequences in the form of many resveratrol supplements with dubious medical benefits.

Happily, the target duo have well characterised biological behaviour, including potent activity in antitumour and acetylcholinesterase inhibition assays. Interest in their synthesis has been plentiful, with both asymmetric and racemic syntheses from KC Nicolaou's team at the Scripps Research Institute in La Jolla, US (see Chemistry World, April 2009, p32), but also including a previous effort by Scott Snyder of Columbia University, New York, US. Snyder's interest in these targets has not waned, and this latest synthesis showcases an innovative use of rearrangement chemistry.²

The synthesis began with an advanced intermediate, the product of six synthetic operations, which was produced in a reasonable yield of 48%. The carbonyl group in the material was targeted with a Corey-Chaykovsky epoxidation, which delivered the required epoxide with great stereocontrol. The epoxide was protonated with a little acetic acid and opened to generate a stabilised cation, which was trapped by the remaining acetate.

Oxidising the resulting alcohol with Dess-Martin periodinane formed the corresponding aldehyde, which was then attacked with an aryl Grignard reagent. However, with little to restrict the stereochemistry of this reaction, the result was a poor diastereomeric ratio of 1:1.3.

Fortunately, this was inconsequential, as the team had planned to impart asymmetry in this region through a reagent-controlled rearrangement reaction (I hope you all read last month's spotlight on rearrangement reactions!). The weapon of choice was a pinacol rearrangement, named not after the chemist responsible for its discovery, but after the archetypal substrate.³ Like many rearrangements, it begins with acid-promoted cation formation: protonating the tertiary alcohol leads to loss of a molecule of water, leaving a relatively stable cation. This is quenched by migration of the nearby aryl group to leave an aldehyde.

The key insight was to use a chiral Brønsted acid. Using a phosphoric acid derivative of (R)-2,2'-diphenyl-(4-biphenanthrol) engendered an exceptional diastereomeric excess, while maintaining high overall reaction efficiency.

A simple oxidation converted the freshly installed aldehyde to a carboxylic acid (leaving the cap off the flask would probably have done it). Treating the per-methylated substrate with boron tribromide revealed all six phenol groups, whereupon the acid group snapped closed onto the nearest phenol to form a gamma-lactone in a single neat step. 

At this point, the team was extremely close to the targets, requiring only a change of oxidation state to complete hopeahainol A. Their hope of directly oxidising the final methylene group on the seven-membered ring was unfortunately not realised, so they were obliged to reprotect all five remaining phenol groups. Per-benzylation produced a substrate that was more amenable to oxidation with ceric ammonium nitrate (CAN). Simply removing the protecting groups with boron trichloride then delivered hopeahainol A, and addition of some sodium methoxide (as in Nicolaou's synthesis) provided hopeanol too.

Paul Docherty is a science writer and blogger based in Reading, UK