Catalysts are almost invariably metal compounds and for a very good reason: they can attract molecules to themselves and thereby open them up to specific reactions which depend on the manner in which the molecules interact with the metal. There are hundreds of metal catalysts performing this service in innumerable reactions ranging from enzymes in living cells to inorganic compounds in large-scale industrial processes.

Organic molecules, on the other hand, have fewer opportunities for interacting with other compounds in a catalytic fashion, but the amino acid proline has this ability and can be used to catalyze the long-known chemical reaction between aldehyde molecules, the so-called aldol condensation. (Proline consists of a five-membered ring which incorporates the amino group with the carboxylic acid group on an adjacent carbon atom.)

Paper #6 comes from Prof. David MacMillan, and research student Alan Northrup, who are based at Caltech, and it reports not only the catalytic coupling together of two aldehydes, but also that the product is almost exclusively one enantiomer.

There are other surprises in paper #6. For example, the coupling of the aldehyde propionaldehyde can be carried out in a variety of solvents ranging from non-polar benzene to highly polar dimethylformamide, with conversion rates ranging from 32% with the former to 91% with the latter. Among other solvents that were tried were chloroform (29% yield), ethyl acetate (41%), dioxane (41%), and dimethylsulfoxide (38%). The reactions used a10% molar ratio of proline, and were carried out at 4 degrees Celsius for 11 hours. The product was essentially the ee enantiomer and in none of the solvents was the ratio of this less than 96%. Other direct aldehyde cross-aldol reactions were performed in dimethylformamide as the solvent with yields of product in excess of 75% and enantioselectivity of around 99%.

In a more recent paper, published in Science, (see A.B. Northrup, D.W.C. MacMillan, Science, 305[5691]: 1752-5, 17 September 2004), Macmillan and Northrup go a step further. By reacting the aldols of one cross-aldol reaction with a third oxyaldehyde, they observed not only aldol addition but cyclization of the product to form a carbohydrate. They have managed to achieve in two steps a process that in the past has required as many as 44. Not only that, but they can now easily make carbohydrates labelled with carbon-13 isotopes, thereby making it possible to identify products and centers within carbohydrate molecules by means of NMR analysis.

The new process should enable chemists to study all kinds of carbohydrates and not just those that Nature makes—and to make them almost as effortlessly. It has always been known that carbohydrates have a critical role in plants and microbes, but they are now recognized as having roles in human biology, as part of the immune system and within the brain itself.

"One of the central goals of chemical synthesis is to design new ways to build molecules that will greatly benefit other scientific fields and ultimately society as a whole,"says MacMillan. "We think that this new chemical sequence will help towards this goal, and there is a bounty of chemical reactions waiting to be discovered that will greatly impact the biological and physical sciences."