The hydroformylation of olefins represents one of the largest homogeneously catalyzed processes in industry and n-butyraldehyde is the largest volume hydroformylation product produced worldwide, the majority of which is hydrogenated to n-butanol. Butyraldehyde is produced by the hydroformylation of propylene in the presence of a rhodium catalyst, where the rhodium atom is bound by one or more phosphorus-containing ligands (Scheme 1). Notably, two aldehyde isomers are produced from the propylene hydroformylation reaction, n-butyraldehyde (linear product) and iso-butyraldehyde (branched product), and the activity and regioselectivity of the rhodium-catalyzed reaction are largely governed by the nature of the phosphorus ligand(s) bound to the metal center. In commodity chemicals manufacturing, the linear isomer is generally more valuable than the branched isomer, and as such, catalysts that yield high linear to branched product ratios (N:I ratios) are sought.

In constructing a detailed model of the propylene hydroformylation energy surface, we discovered that the computed carbonyl stretching frequency of the catalyst resting state is a predictor of catalyst rate and selectivity. DFT calculations were used to survey hundreds of bidentate phosphorus ligands on the rhodium center and the model predicts that more electron-withdrawing phosphorus ligands will lead to both higher catalyst rates and higher N:I selectivity. We targeted a new class of highly electron-withdrawing bisphosphoramidite ligands for synthesis and evaluation in the rhodium-catalyzed hydroformylation of propylene and the results are discussed herein. High pressure NMR spectroscopy was used to characterize the catalyst resting states as well as explore mechanisms of catalyst deactivation. Ultimately, high reaction rates and high regioselectivity (N:I >100) were realized with the new ligand set, but the catalysts also demonstrated higher rates of deactivation during continuous processes.

Robert Froese

Appearances