Research - Ligand Diversity

Which reaction and which metal/ligand combination will provide a required product efficiently and in high enantioselectivity? Our solution to this problem starts with targeted libraries in which a common precursor is used to generate a divergent array of ligand structures. These can then be assayed in a given process and further optimised to achieve maximum enantioselectivity. In our first demonstration of this approach we employed (S)-serine for the synthesis of a series of phosphinite-oxazolines 1, from which 2 was quickly identified as an optimum structure for controlling palladium catalysed allylic alkylation of 1,3-diphenylpropenyl acetate (96% e.e.). Similarly 3 was found to give an encouraging 70% e.e. with the more demanding dimethyl substituted precursor. [1, 2]
The planar chiral ferrocene building block 4 has also proved to be good starting point for generating ligand diversity. After oxidation to aldehyde 5, Grignard addition proceeds cleanly and with high diastereoselectivity. Amines can be introduced stereospecifically as can phosphine substituents via bromine-lithium exchange to give ligands 6 displaying three points of diversity. [3]
To help identify which ligands are likely to be successful in a given process we are also mapping correlations between know ligands and the selectivity and absolute configuration of the products arising from their application in various metal catalysed reactions.
 
[1] G. Jones and C. J. Richards, Tetrahedron Lett., 2001, 42, 5553. [2] .G. Jones and C. J. Richards, Tetrahedron: Asymmetry, 2004, 15, 653. [3] Abstracts of Papers, 224th National Meeting of the American Chemical Society, Boston, MA, 2002; ORGN 442.