Qualitative and quantitative analysis of mineral oil pollution in peanut oil by Fourier transform near-infrared spectroscopy

Emerging contaminants pose a potential threat to the safety of edible oils. This study combined Fourier Transform Near-Infrared (FT-NIR) spectroscopy with chemometrics for the qualitative and quantitative analysis of five contaminants in peanut oil. The results show that the Partial Least Squares Discriminant Analysis (PLS-DA) classifier effectively differentiates between normal and contaminated samples with a classification accuracy of 100 %. In specific contaminant identification, PLS-DA achieved 100 % accuracy for diesel, white mineral oil, and lubricating oil, and 97.04 % for kerosene and engine oil. Quantitative results revealed that Support Vector Regression (SVR) exhibited high precision in predicting diesel (R_P = 0.9852), white mineral oil (R_P = 0.9908), and lubricating oil (R_P = 0.9929), while Partial Least Squares Regression (PLSR) demonstrated good predictive ability for kerosene (R_P = 0.9335) and engine oil (R_P = 0.9270). Therefore, NIR spectroscopy can be an effective tool for monitoring the safety of edible oils.