Détermination de l’acidité et de l’indice de peroxyde dans les huiles d’olive vierges et dans les huiles raffinées par spectrométrie proche infrarouge à transformée de Fourier

Abstract

A rapid Fourier transform near-infrared (FT-NIR) spectroscopic method was developed for quantitative determination of peroxide value (PV) and acidity, expressed in oleic acid % (w/w), in virgin olive oils and for quantitative determination of PV in sunflower oils. Calibration models were constructed using Partial Least Squares (PLS) regression with different wavenumber ranges. The best results were obtained between 5100 and 4520 cm-1 for the evaluation of acidity and between 7500 and 6300 cm-1 for the determination of PV. Models obtained allowed the prediction of PV with an uncertainty of 1,0 meqO2/kg within a 3 to 32 meqO2/kg analytical range and acidity with an uncertainty of 0,06% (w/w) within a 0,15 to 1,3% analytical range. In the case of refined sunflower oils the indirect measurement of PV is obtained with an uncertainty of 1,1 meqO2/kg between 1,5 to 24 meqO2/kg. However, prediction obtained for PV determination of refined oils from different origins was not so efficient. It seems better to carry out a calibration model for each type of oil. Due to its simplicity and rapidity, FT-NIR method provides an alternative means of measuring acidity and PV. Also, FT-NIR methods can be automated to allow routine and on-line quality control analysis in fats and oils industry.