食用油中のベンゾ[a]ピレンの励起・蛍光同時測定蛍光分光法と並列因子解析を組み合わせた迅速かつ非破壊的検出

Benzo[a]pyrene (BaP), a potent carcinogen commonly present in foods processed at high temperatures, typically exists at trace levels (μg/L) in complex food matrices such as edible vegetable oils. Traditional BaP detection methods based on chromatographic separation techniques are hindered by issues such as complex sample pretreatment, high instrument costs, and time-consuming procedures, which limit their applicability for rapid screening and on-site testing. This study proposes a rapid, non-destructive method that combines excitation-emission matrix fluorescence (EEMF) spectroscopy with parallel factor analysis (PARAFAC), enabling direct detection of BaP content in edible vegetable oils without the need for sample pretreatment. The method integrates external calibration, PARAFAC and ordinary least squares (OLS) regression to construct an analytical framework. By establishing PARAFAC models for standard solutions of BaP at different concentrations, the linear response range compliant with the Lambert-Beer law was determined to be 1-500 μg/L. Dilution experiments with spiked vegetable oils confirmed that peanut oil, soybean oil, and corn oil samples did not exhibit significant inner filter effects without any pretreatment, indicating that this method is suitable for direct quantitative analysis of BaP in complex matrices. Based on these findings, the method was further applied to detect BaP content in three different types of vegetable oils spiked with the compound. The results showed that the proposed method could effectively distinguish BaP from other natural fluorescent components. By combining relative concentration scores and the established concentration prediction model, the method could achieve precise quantification of BaP concentrations above 10 μg/L in vegetable oils and accurately identify BaP levels above 1 μg/L in peanut oil and corn oil, and above 7.5 μg/L in soybean oil. The method is rapid, convenient, and easy to operate, with high sensitivity and accuracy, making it well suited for on-site BaP screening in diverse real-world scenarios such as quality control in industrial production, regulatory monitoring, and import/export inspection of edible oils. Moreover, the proposed framework offers a generalizable strategy that holds significant reference value and potential for extension to the detection of other highly toxic contaminants in complex oil matrices.