Fourier transform infrared spectroscopy coupled with chemometrics for the monitoring of virgin olive oil quality during storage up to 18 months

This study examines the feasibility of using mid infrared (MIR) spectroscopy for monitoring the quality of 14 Moroccan virgin olive oil samples belonging to three (3) varieties, originated from three (3) geographical origins, extracted with different systems, irrigated and fertilised with different modes.

The application of principal component analysis (PCA) and factorial discriminant analysis (FDA) on MIR spectra data allowed to discriminate perfectly the samples according to their storage time with 96.87% of correct classification.

The obtained results were confirmed following the application of three different multivariate regression tools namely partial least squares regression (PLSR), principal component regression (PCR) and support vector machine regression (SVMR) applied on MIR spectra data, since excellent prediction of R² = 0.99 and RMSEP = 17.05 days was observed. In addition, the evaluation of the effectiveness of MIR spectroscopy to predict the chemical parameters allowed to obtain excellent validation models with R² ranging between 0.98 and 0.99 for free acidity, peroxide value, chlorophyll level, k₂₃₂ and k₂₇₀.