Proof of the potential for using near-infrared spectroscopy for monitoring and optimization of phospholipase C assisted vegetable oil degumming

Abstract: The Phospholipase C (PLC) assisted degumming of vegetable oils is a well-known process with main benefit being increased oil yields. Phospholipases C hydrolyze selected phospholipids into diglycerides (DG) that are released to oil phase, and phosphate groups that are removed with heavy gums phase. The DG content is therefore considered a good indicator of the enzymatic reaction progress. In this study we investigated the use of Near-Infrared Spectroscopy (NIRS) and chemometrics to monitor the degumming process by quantifying diglycerides in three soybean oils in both at-line and on-line settings. Fifteen enzymatic degumming lab scale batches originating from a definitive screening design (with varying water, acid, and enzymes dosages) was investigated with the aim to develop a NIRS prediction method. We have found out that the diglyceride concentrations in process samples can be reliably and accurately predicted using the NIR spectra and Partial Least Squares (PLS) regression. The prediction errors were in the range from 0.06% to 0.1% (w/w) for both at-line and on-line data, which is sufficient for monitoring the degumming reaction. Despite the challenges posed by the heterogeneous nature of the degummed oil matrix, on-line analysis with NIRS proved to be successful in identifying the effect of different enzyme dosages upon hydrolysis kinetics. The Process Analytical Technology tool developed (NIRS+ PLS) demonstrate great potential for being transferred to a full-scale process by installation a side stream in oil refining plants. Continuous NIRS monitoring of enzymatic degumming would result in a more tailored use of reagents and the enzyme itself, improving the process sustainability and possibly increasing yield due to a better control of the process.