J. Am. Chem. Soc. 136, 4109–4112 (2014)

It is well understood that the outcome of substitution reactions on benzene rings depends on the substituents already present. In phenols, for example, additions to the ring system are directed to the ortho- and para- positions, relative to the hydroxyl group. This raises an interesting synthetic challenge: any phenol-derived synthetic target with a substituent in a meta- position must be made by a method that avoids functionalizing a phenol, often requiring lengthy synthetic sequences. Now, Igor Larrosa and co-workers from Queen Mary University of London have developed a one-pot sequence of reactions that allows arylation of unprotected phenols at the meta-position.

In a previous approach to this problem, a remote directing group that directs a palladium catalyst to the meta-position was installed on the phenol.This method, however, introduces additional synthetic steps in installing and subsequently removing the directing group. Larrosa and co-workers' method instead uses the known ortho/para-directing effect of the phenol to install a transient carboxyl group onto the ring. This group subsequently directs a palladium-catalysed cross-coupling reaction at the position that is ortho to the carboxyl group and thus meta to the hydroxyl substituent. Finally, decarboxylation of the arene affords selectively functionalized 1,3-substituted phenol derivatives. The process exhibits good tolerance to various substituents, including electron-donating and electron-withdrawing groups, and substitution patterns on either of the aryl coupling partners.

The elegance of this process comes at a price, however. High-pressure carbon dioxide is required to carry out the carboxylation step, and high temperatures are required throughout the sequence. But the economy of synthesis afforded by this process is shown by the synthesis of a γ-secretase inhibitor, a drug candidate for the treatment of Alzheimer's disease. The previous synthesis of this compound required eight synthetic steps and proceeded with an overall yield of 6%. Starting instead with a meta-arylation of 5-bromophenol, the team prepared the target in just three steps with a yield of 41%.