Some words of caution from Buchwald & Bolm this week in the Angewandte (DOI) on the use of catalysts. They observe that in certain cross-coupling reactions the reaction yield very much depends on the quality of the metal catalyst. In several reactions with ferric chloride, the yields range from the very low (10%-30%) with pure FeCl3 material (commercial grade > 99.99%) to high (70%-90%) with regular material (commercial > 98%) which is odd.
The regular iron salt is found to contain a small amount of copper and in fact by adding copper oxide at the ppm level to high-grade FeCl3, the yields are restored to their proper values.
So the question is, whether iron is the true catalyst or copper. The results also require a rethink on other reactions such as the copper-free Sonogashira reaction and the so-called metal-free reactions as already discussed here and here in this blog.
Update 30 Sept: Bedford et al. (DOI) also find that metal contamination , this time Pd contamination in (again) an iron catalyst can interfere with an organic reaction , this time a Suzuki reaction. Palladium is found to be an effective catalyst at the ppb level!
Buchwald, S., & Bolm, C. (2009). On the Role of Metal Contaminants in Catalyses with FeCl Angewandte Chemie International Edition DOI: 10.1002/anie.200902237