Angew. Chem. Int. Ed.http://doi.org/fz985q (2013)

Credit: © 2013 WILEY

Hydroformylation is an industrially important reaction that converts alkenes into aldehydes by reacting them with a mixture of carbon monoxide and hydrogen. Millions of tonnes of a variety of very useful products are made from relatively cheap alkenes every year through hydroformylation. One particular class of alkene substrate, vinyl arenes, is of great interest. Theoretically, hydroformylation can lead to two possible products — an α-aryl or β-aryl aldehyde — with formation of the branched (α-aryl) product being favoured by interactions of the metal catalyst with the aryl ring. The branched product is chiral and several catalyst systems have been developed to produce this product with both high regio- and stereoselectivity.

Despite being a synthetically useful product, few catalyst systems have been developed that can overcome the natural selectivity of the reaction to produce a linear (β-aryl) aldehyde. Now, Paweł Dydio and Joost Reek from the University of Amsterdam have done just that, by designing a ligand for a rhodium catalyst that dictates the orientation in which a vinyl-2-carboxyarene substrate binds to the catalyst before reaction. The ligand contains two phosphite groups (which coordinate to the rhodium) connected by a diamidodiindolylmethane to form a 'pocket' which binds strongly to the carboxylate group in the substrate. In the key regiochemistry-defining step — migration of a hydride to the alkene (pictured) — it is this strong binding that makes the formation of an α-phenylalkyl rhodium complex (that would ultimately lead to the usual branched product) significantly less favourable. In a control reaction in which the carboxylate group of the substrate is masked as an ester, the natural selectivity (producing 95% of the branched product) is restored.

Dydio and Reek show that the reaction tolerates many other substituents on the aryl ring in addition to the carboxylate directing group, and also demonstrate that the selectivity for the β-aryl aldehyde product is maintained even with additional substituents on the alkene. In this latter case, a product containing a new stereocentre is formed and the ability to control the selectivity of this process offers a new target for catalyst design and development.