Cluster-Based Catalysts for Asymmetric Synthesis

Sammanfattning: In this work, the synthesis and characterisation of new low-valence transition metal carbonyl clusters, and an investigation into their viability to act as catalysts/catalyst precursors in asymmetric synthesis, are described. Carbonyl clusters based on ruthenium and osmium have been tested as (pre)catalysts for asymmetric hydrogenation of alpha-unsaturated carboxylic acids, and cobalt carbonyl clusters have been used as (pre)catalysts in (asymmetric) Pauson-Khand synthesis. The clusters [Os3(µ-H)2(CO)8(µ-1,2-P-P)] (x=1,2; P-P=chiral diphosphine) did not show any tendency to catalyse hydrogenation of the prochiral substrate tiglic acid. The new triosmium clusters were found to be generally inert and did not react with carbon monoxide or acetylene, or undergo hydride exchange with styrene-d8, all of which are reactions characteristic for the parent cluster [Os3(µ-H)2(CO)10]. Clusters of the type [H4Ru4(CO)12-2x(P-P)x] (x=1,2; P-P=chiral diphosphine) are viable as catalysts for (asymmetric) hydrogenation. Their reactivity strongly depends on the phosphine ligand coordinated. The reactivity of the ruthenium clusters derivatised with ferrocene-based chiral diphosphines showed a significant increase in reactivity compared to other clusters investigated. In the case of clusters coordinated with Walphos ligands, enantiomeric excess of >90 % was obtained in hydrogenation of tiglic acid. The Pauson-Khand reaction was catalysed by cobalt carbonyl clusters of the general formulae [Co3(µ3-CH)(CO)9-x(PR3)x] (x=1-3; PR3=achiral/chiral monophosphines) and Co3(µ3-CH)(CO)7(P-P) (P-P=chiral diphosphine). The clusters were tested both in the intermolecular and the asymmetric intramolecular versions of the Pauson-Khand reaction. In the asymmetric Pauson-Khand reaction, it was found that the clusters that gave the highest product yields also gave racemic mixtures. In contrast, the cluster that gave the lowest cyclopentenone yield (6%), [Co3(µ3-CH)(CO)7(CHIRAPHOS)] was able to generate the highest enantiomeric excess of 13 %. The absence of the original cobalt cluster after the catalytic tests indicates that the cobalt clusters act as catalyst precursors. Regardless of the nature of the active catalyst, the results obtained with these cluster-based catalytic systems are too poor to make these cobalt clusters of practical use in cyclopentenone synthesis.