ABSTRACT
Sciences, University of Amsterdam, Nieuwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands. 3 Institut de Recherches sur la Catalyse, 2 Avenue Albert Einstein,
69626 Villeurbanne Cedex, France
gadi@science.uva.nl
Abstract We present in this communication an alternative concept for optimizing homogeneous catalysts. With this method, one extracts and screens virtual catalyst libraries, indicating regions in the catalyst space where good catalysts are likely to be found. This automated screening is done in two stages: A ‘rough screening’ using 2D descriptors, followed by a “fine screening” using 3D descriptors. The model is demonstrated for a set of Pd-catalyzed Heck reactions using bidentate ligands. Introduction
One of the biggest challenges in chemistry is discovering new chemical reactions that will enable society to function in a sustainable manner. Atom economy and waste minimization are at the heart of industrial policy, driven both by governmental incentives and by market considerations. Homogeneous catalysis is a most promising area in this respect (1). The Nobel prizes (2, 3) awarded in 2001 (to Sharpless, Noyori and Knowles for their discovery of chirally-catalyzed oxidation and hydrogenation reactions), and in 2005 (to Chauvin, Grubbs, and Schrock for their discovery of metathesis catalysts) exemplify how a new catalyst can cause a paradigm shift in the chemical industry (4).
