Abstract:A preliminary study of used frying oil demonstrates that a fluorescence detection method based on fluorescent sensor has been developed for oil quality determination. Copper tetraphenylporphyrin was selected according to its special spectrum differentiating from the endogenous fluorophores in oil and its sensitivity to total polar compounds typically producing during the oxidation process. The fluorescence landscapes of oil samples were analyzed by parallel factor analysis to decompose high-dimensional variables and extract feature data. Then, support vector regression (PLSR) and partial least squares regression (SVR) were used to construct regression model based on the extracted fluorescence data and chemical reference measurements. In contrast to the PLSR model, the SVR model obtained the better regression results as follow: R = 0.9645 and 0.9118 in the training set and test set, respectively. The overall results demonstrate that a fluorescence detection system based on fluorescent sensor is useful for the quality evaluation of frying oil.

Qualitative and Quantitative Analysis of Oxidative Degradation Products in Frying Oil by Three-Dimensional Fluorescence Spectroscopy with Metalloporphyrin-Based Sensor