NIR spectroscopy for the quality control of Moringa oleifera (Lam.) leaf powders: Prediction of minerals, protein and moisture contents

A rapid methodology was developed to simultaneously predict water content and activity values (a(w)) of Moringa oleifera leaf powders (MOLP) using near infrared (NIR) signatures and experimental sorption isotherms. NIR spectra of MOLP samples (n=181) were recorded. A Partial Least Square Regression model (PLS2) was obtained with low standard errors of prediction (SEP of 1.8% and 0.07 for water content and a(w) respectively). Experimental sorption isotherms obtained at 20, 30 and 40 degrees C showed similar profiles. This result is particularly important to use MOLP in food industry. In fact, a temperature variation of the drying process will not affect their available water content (self-life). Nutrient contents based on protein and selected minerals (Ca, Fe, K) were also predicted from PLS1 models. Protein contents were well predicted (SEP of 2.3%). This methodology allowed for an improvement in MOLP safety, quality control and traceability.