MOE Key Laboratory of Laser Life Science & Laboratory of Photonic Chinese Medicine, College of Biophotonics, South China Normal University, Guangzhou 510631, China.
MOE Key Laboratory of Laser Life Science & Laboratory of Photonic Chinese Medicine, College of Biophotonics, South China Normal University, Guangzhou 510631, China; Analysis and Testing Center, South China Normal University, Guangzhou 510631, China.
Spectrochim Acta A Mol Biomol Spectrosc. 2023 Dec 5;302:123014. doi: 10.1016/j.saa.2023.123014. Epub 2023 Jun 15.
The juice drink industry has repeatedly been exposed to adulteration. Unscrupulous producers, for example, use cheap juice for substitution in the pursuit of more significant economic benefits, which presents a tremendous challenge for the control of the quality of drinks. The objective of this study was to apply Raman spectroscopy combined with chemometrics to rapidly quantify the adulteration concentration of apple juice or grape juice in pomegranate juice. Two supervised learning algorithms: partial least squares regression (PLSR) and support vector machine regression (SVR) were used to analyze the Raman spectra of 114 samples. The coefficient of determination (R), root mean square error (RMSE), and residual prediction deviation (RPD) of the prediction set when using PLSR and SVR to predict the adulterated concentration of apple juice in pomegranate juice were 0.9357 and 0.9465, 6.446% and 5.974%, 3.945 and 4.322, respectively. The R, RMSE, and RPD of the prediction set when using PLSR and SVR to predict the adulteration concentration of grape juice in pomegranate juice were 0.9501 and 0.9502, 6.334% and 5.571%, and 4.475 and 4.481, respectively. It was concluded that Raman spectroscopy combined with chemometrics has excellent potential for application as a rapid quantitative method to detect adulterated concentrations of pomegranate juice.
果汁饮料行业屡屡被曝光存在掺假现象。例如，为追求更大的经济效益，不法生产者在生产过程中使用廉价果汁进行替代，这给饮料质量控制带来了巨大挑战。本研究旨在应用拉曼光谱结合化学计量学方法，快速定量测定石榴汁中苹果汁或葡萄汁的掺假浓度。本研究采用两种有监督学习算法：偏最小二乘回归（PLSR）和支持向量机回归（SVR）对 114 个样本的拉曼光谱进行分析。当使用 PLSR 和 SVR 预测石榴汁中苹果汁的掺假浓度时，预测集的决定系数（R）、均方根误差（RMSE）和预测偏差（RPD）分别为 0.9357 和 0.9465、6.446%和 5.974%、3.945 和 4.322。当使用 PLSR 和 SVR 预测石榴汁中葡萄汁的掺假浓度时，预测集的决定系数（R）、均方根误差（RMSE）和预测偏差（RPD）分别为 0.9501 和 0.9502、6.334%和 5.571%、4.475 和 4.481。综上所述，拉曼光谱结合化学计量学具有作为一种快速定量检测石榴汁掺假浓度的方法的巨大潜力。
