便携拉曼光谱快速无损定量检测酱料中山梨酸钾

Rapid nondestructive detection of potassium sorbate in sauces by portable Raman spectroscopy

中文关键词:  拉曼光谱技术  酱料调味品  分子光谱  化学计量学  质量评价

英文关键词:Raman spectroscopy  sauces  molecular spectroscopy  chemometrics  quality evaluation

中文摘要:

      目的 基于拉曼光谱技术(Raman spectroscopy)并结合化学计量学方法建立定量分析模型, 以实现对酱料中山梨酸钾进行快速定量检测。方法 采用便携式拉曼光谱检测酱料中的山梨酸钾, 并对比不同的预处理方法进行模型性能评估, 建立偏最小二乘法(partial least squares regression, PLSR)回归模型。结果 甜辣酱的光谱信息经S-G(5,2)卷积平滑处理后建立PLSR模型的校正相关系数为0.9670, 预测相关系数为0.9616, 预测集的残差预测偏差(residual prediction deviation, RPD)达到5.0335; 香辣酱的光谱信息经S-G(3,1)卷积平滑加一阶导数处理, 建立PLSR模型的校正相关系数为0.9766, 预测相关系数为0.9432, 模型的RPDc值和RPDp值分别可以达到6.6727和4.1143, 均大于3, 定量效果良好。结论 本研究采用的便携式拉曼光谱技术联用化学计量学方法所建立的光谱定量模型获得酱料中关键成分的含量信息是可行的, 该方法经过进一步优化有望用于酱料质量的快速检测。

英文摘要:

      Objective To establish a quantitative analysis model based on Raman spectroscopy combined with chemometric methods, aiming to achieve rapid quantitative detection of potassium sorbate in sauces. Methods A portable Raman spectrometer was used to detect potassium sorbate in sauces, and various preprocessing methods were compared to assess the performance of the model. A partial least squares regression (PLSR) model was constructed. Results For sweet and spicy sauce, the spectral information was preprocessed with S-G convolution smoothing (5,2), leading to a calibration correlation coefficient of 0.9670 and a prediction correlation coefficient of 0.9616. The residual prediction deviation (RPD) of the prediction set reached 5.0335. For spicy sauce, spectral information preprocessed by S-G convolution smoothing (3,1) plus first derivative, obtained a calibration correlation coefficient of 0.9766 and a prediction correlation coefficient of 0.9432. The RPDc and RPDp values of the model were 6.6727 and 4.1143, respectively, both greater than 3, indicating a good quantitative effect. Conclusion Overall, the results indicate that the spectral quantitative model established using portable Raman combined with chemometrics in this study is feasible for obtaining key component content in sauces. With further optimization, this method is expected to be applied for the rapid detection of sauce quality.