
The Baylis-Hillman reaction is a well-known coupling process between aldehydes and activated alkenes, catalysed by tertiary amines or phosphines. Generally the process works with α,β- unsaturated esters and nitriles but is not very successful with acrylamides. The addition of phenol in conjunction with Bu3P “catalyst” has been shown to have a cooperative effect. A recent report (Gong J H et al, Tetrahedron Lett), however, shows that when dimethyl acrylamide is used, the product is not the Baylis-Hillman product but the adduct from a Stetter reaction (usually catalysed by cyanide ion or a thiazolium salt). It is suggested that the tributylphosphine reacts with benzaldehyde to give an anion, which isomerises to an ylide, which adds in Michael fashion to the acrylamide. Deuterium labelling establishes the mechanism shown below. The reaction does not work with triphenyl phosphine, and seems specific for the α,β- unsaturated amides and aromatic aldehydes. Aliphatic aldehydes give complex mixtures apart from pivaldehyde which gives a 54% yield. No benzoin adducts form when Bu3P is used, in contrast to many Stetter reactions with other catalysts.
















