Reductive Hydroformylation of Linseed Oil: from Rhodium Catalyst Stability to Machine Learning Data Analysis

Abstract

The reductive hydroformylation of linseed oil catalyzed by a rhodium/triethylamine system was optimized for alcohols production. Catalyst stability toward dioxygen, water, and primary and secondary oxidation products from linseed oil (such as peroxides or carboxylic acids) was critically assessed, revealing their impact on catalytic performance. The catalytic system can be applied (i) without special precautions during the preparation of the catalyst and reactants, (ii) in the presence of oxidised substrates and (iii) in the presence of water. The robustness of this catalytic system is particularly interesting, as it could be applied to inedible waste oils. The reaction parameters such as rhodium/triethylamine ratio, temperature, pressure, reaction time, and selected combinations of these factors were varied to maximise alcohol yield and to identify the relationships between the experimental reaction conditions and products yields. Systematic parameter screening combined with machine learning analysis identified the key variables controlling alcohol yield.