Process spectroscopy in microemulsions—Raman spectroscopy for online monitoring of a homogeneous hydroformylation process

Abstract

A major industrial reaction based on homogeneous catalysis is hydroformylation for the production of aldehydes from alkenes and syngas. Hydroformylation in microemulsions, which is currently under investigation at Technische Universität Berlin on a mini-plant scale, was identified as a cost efficient approach which also enhances product selectivity. Herein, we present the application of online Raman spectroscopy on the reaction of 1-dodecene to 1-tridecanal within a microemulsion. To achieve a good representation of the operation range in the mini-plant with regard to concentrations of the reactants a design of experiments was used. Based on initial Raman spectra partial least squares regression (PLSR) models were calibrated for the prediction of 1-dodecene and 1-tridecanal. Limits of predictions arise from nonlinear correlations between Raman intensity and mass fractions of compounds in the microemulsion system. Furthermore, the prediction power of PLSR models becomes limited due to unexpected by-product formation. Application of the lab-scale derived calibration spectra and PLSR models on online spectra from a mini-plant operation yielded promising estimations of 1-tridecanal and acceptable predictions of 1-dodecene mass fractions suggesting Raman spectroscopy as a suitable technique for process analytics in microemulsions.