Inline monitoring of chemical reactions using spectroscopy can lead to great improvements in process efficiency and economics. In this work, we performed inline attenuated total reflectance infrared spectroscopy on two reactions involving bioderived lactones. Using partial least squares regression, we were able to gain information on the kinetics of the reactions, during batch reactor operation at high temperature and pressure.

Inline monitoring of chemical processes using spectroscopy is a powerful approach to replace conventional sampling. Spectroscopic data can be used with regression models to determine the endpoint or the onset temperature of reactions and can thus improve the efficiency and economics of processes. Spectroscopic data recorded during reactor operation is often complex, but partial least squares (PLS) is well-suited for dealing with this complexity. Here, we employed in situ attenuated total reflectance infrared (ATR-IR) spectroscopy in the hydrogenation of levulinic acid (LA) into γ-valerolactone (GVL) and operando ATR-IR spectroscopy in the ring-opening of γ-stearolactone (GSL), at high temperatures and pressures. Using PLS, we constructed a model to predict the GVL concentrations from spectra recorded during LA hydrogenation and determined the onset temperature of GVL formation. During the hydrogenation of branched unsaturated fatty acids, we determined the concentrations of the dilute byproduct GSL. The onset temperature of GSL ring-opening could not be determined, due to excessive variance caused by temperature differences during reactor heating, compared to limited variance caused by GSL conversion. We believe that in the oleochemical industry, this methodology can be used as a quality control tool to ensure that the produced branched saturated fatty acid fall within the GSL concentration specification limits.