Quantitative Detection of Virgin Olive Oil Adulteration with Oil Obtained from Ground-Picked Olives Using Mid- and Near-Infrared Spectroscopy

Abstract This study investigated the quantification of virgin olive oil (VOO) adulteration with oil obtained from ground-picked olives (GPOO) using mid- and near-infrared spectroscopy combined with partial least squares regression (PLSR). Nine GPOO and nine VOO source oils (18 in total) were characterized by reference analyses, including free fatty acidity, peroxide value, specific absorbance, viscosity, refractive index, and color parameters. A total of 101 binary mixtures were prepared gravimetrically at randomly varied proportions ranging from 0 to 100% GPOO (w/w), comprising 68 calibration and 33 external validation samples. Spectral data were acquired using benchtop FT-MIR, benchtop FT-NIR, and handheld NIR instruments. All models showed strong predictive performance (rPred = 0.98). Benchtop FT-NIR provided the best results (RPD = 4.84; SEP = 6.14%), followed by benchtop FT-MIR (RPD = 4.77; SEP = 6.22%) and handheld NIR (RPD = 3.92; SEP = 7.58%). These findings demonstrate that IR spectroscopy coupled with chemometrics can successfully quantify GPOO adulteration in VOO, with handheld NIR showing promise for rapid on-site authenticity screening.