Ye Qin, Meng Xianghe
Institute of Food Sciences, Zhejiang Academy of Agricultural Sciences, Hangzhou 310014, China.
College of Food Science and Technology, Zhejiang University of Technology, Deqing 313200, China.
Food Chem. 2022 Aug 15;385:132661. doi: 10.1016/j.foodchem.2022.132661. Epub 2022 Mar 8.
A novel improved method for the authentication of edible oil samples based on Fourier-transform infrared (FTIR) spectroscopy coupled with chemometrics has been developed. A discrimination analysis model has been developed. On this basis, 100% correct classification of 135 samples from eleven species has been achieved. Recognition rates with respect to external validation for 91 pure oil samples and 231 blend samples were 100% and 92.6%, respectively. A general quantitative model for detecting edible oil adulteration (taking Camellia oil as an example) has also been built. An optimal backward interval partial least-squares model, based on the spectral regions ν = 3100-2900, 1800-1700, 1500-1400, and 1200-1100 cm, has been determined, giving good performances. A specific sub-model using a single adulterant oil has also been constructed, which showed higher prediction accuracy. Based on the developed qualitative and quantitative FTIR methods, adulterant oils in Camellia blends could be rapidly detected, effectively differentiated, and accurately quantified.
基于傅里叶变换红外（FTIR）光谱结合化学计量学，开发了一种新型的食用油样品鉴别改进方法。已建立了判别分析模型。在此基础上，实现了对来自11个品种的135个样品的100%正确分类。91个纯油样品和231个混合油样品的外部验证识别率分别为100%和92.6%。还建立了一个检测食用油掺假的通用定量模型（以山茶油为例）。基于ν = 3100 - 2900、1800 - 1700、1500 - 1400和1200 - 1100 cm光谱区域，确定了一个最优的向后区间偏最小二乘模型，表现良好。还构建了一个使用单一掺假油的特定子模型，其预测准确率更高。基于所开发的定性和定量FTIR方法，可以快速检测、有效区分和准确量化山茶混合油中的掺假油。
