Application of real-time refrigeration turbidimetry for rapid discrimination and purity assessment of refined peanut oils adulterated with other seed oils

Abstract

Real-time refrigeration turbidimetry and common chemometric methods, including principal component analysis (PCA) and partial least squares (PLS), were applied to the discrimination and purity assessment of some refined peanut oils adulterated with other seed oils. A special turbidimeter was developed, and the working temperature fluctuation could be lower than ±0.25 °C. The curves of turbidance versus refrigeration time of RPO samples were constructed and analyzed by PCA and PLS-1. Discrimination of pure RPO and adulterated samples was successfully achieved by calculating and comparing the Euclidean or Manhattan distances of their first two principal components obtained by PCA. The PLS-1 was applied to the purity assessment of RPO samples. The correlation coefficient and root-mean-square error of prediction were 0.94 and 2.75%, respectively. Forty-six RPO samples were analyzed by the proposed method and gas chromatography (GC). The analytical results obtained show that only one adulterated sample confirmed by PCA was misjudged. The absolute errors of the purity assessment results obtained by PLS-1 relative to those obtained by GC were less than 4%.