Am. Porte et al., DESIGN AND OPTIMIZATION OF NEW PHOSPHINE OXAZOLINE LIGANDS VIA HIGH-THROUGHPUT CATALYST SCREENING, Journal of the American Chemical Society, 120(36), 1998, pp. 9180-9187
This paper uses the phosphine oxazoline ligands 1 and an allylation tr
ansformation (reaction 1) to illustrate the value of divergent ligand
syntheses and high-throughput screening in catalyst discovery and opti
mization. Thus, a diverse set of Ligands 1 (Table 1) was prepared via
a divergent synthesis involving the pivotal intermediate, phosphine-su
bstituted amino alcohol 7 (Scheme 1). Single-crystal X-ray crystallogr
aphic data was obtained for a nickel complex 8 (Figure 3) of the pheny
l-substituted ligand ii. This analysis illustrated some structural fea
tures of the ligand systems 1 that may be conducive to asymmetric cata
lysis. High-throughput screens were then used to correlate the ligand
1 R-substituents with asymmetric induction in the allylation reaction
1, and it emerged that the pseudo-spherical adamantyl substituent was
superior to other R-substituents. Other parameters in the catalyst sys
tems were also varied, sometimes in ''two-dimensional'' screens. No pr
onounced solvent effects were identified. Abstraction of chloride was
shown to be detrimental, whereas addition of chloride provided no adva
ntages. One of the most critical of all the variables probed was, rath
er surprisingly, the effect of ligand-to-metal ratio; enantioselectivi
ties dropped sharply and eventually reversed when this ratio was incre
ased above 1:1. These observations were rationalized in terms of a che
lated complex A and a nonchelated one B (Scheme 3). The implications o
f these results for high-throughput screening of catalyst systems in g
eneral, and for Ligands 1 in particular, are discussed.