拉曼光谱技术在管花肉蓉多糖纯化过程中含量快速测定研究  

Rapid determination of polysaccharides in Cistanche tubulosa during purification by Raman spectroscopy technique

作　　者：王宇 张思程 张辉 沈静 姚军[1] WANG Yu;ZHANG Si-cheng;ZHANG Hui;SHEN Jing;YAO Jun(School of Pharmacy,Xinjiang Medical University,Urumqi 830011,China;Department of Pharmacy,The First Affiliated Hospital of Xinjiang Medical University,Urumqi 830054,China)

机构地区：[1]新疆医科大学药学院,新疆乌鲁木齐830011 [2]新疆医科大学第一附属医院药学部,新疆乌鲁木齐830054

出　　处：《化学研究与应用》2024年第6期1230-1235,共6页Chemical Research and Application

基　　金：新疆维吾尔自治区自然科学基金项目(2022D01C327)资助;中央引导地方科技发展专项项目(ZYYD2023B11)资助;新疆天然药物活性组分与释药技术重点实验室项目(XJDX1713)资助。

摘　　要：采用拉曼光谱技术结合化学计量学建立管花肉蓉中多糖大孔吸附树脂纯化过程的分析模型。收集管花肉蓉中多糖的大孔吸附树脂洗脱液,采用苯酚-硫酸显色法测定其多糖含量,采集洗脱液的拉曼光谱信息。通过PCA主成分分析分检测异常光谱,利用Kennard-Stone算法将样本集划分为训练集和预测集,通过卷积平滑法(SG)、标准正态变量(SNV)、多重散射校正(MSC)、一阶导数(FD)、二阶导数(SD)以及它们的组合进行预处理优化模型,采用连续投影算法(SPA)和稳定性竞争性自适应重加权采样(sCARS)进行变量选择,采用随机森林(RF)、偏最小二乘法(PLS)和支持向量机(SVM)建立模型,采用决定系数(R^(2))均方根误差(RMSE)和剩余预测偏差(RPD)对模型进行评价。结果表明,三种模型的R^(2)值均>0.95,RPD均>4,RMSE均<0.12。拉曼光谱技术结合化学计量学法可用于管花肉蓉中多糖的纯化过程质量评价。Raman spectroscopy combined with chemometrics was used to establish analytical models for the purification process of polysaccharides in Cistanche tubulosa with macroporous adsorbent resin.The eluent of the macroporous adsorbent resin of polysac-charides in Cistanche tubulosa was collected,and the polysaccharide content was determined by phenol-sulfuric acid colorimetric method,and the Raman spectral information of the eluent was collected.Anomalous spectra were detected by PCA principal compo-nent analysis score.The samples were divided into training and prediction sets using the Kennard-Stone algorithm.The model is op-timized by preprocessing through convolutional smoothing(SG),standard normal variables(SNV),multiple scattering correction(MSC),first-order derivatives(FD),second-order derivatives(SD),and combinations of them.The successive projection algorithm(SPA)and Stability Competitive Adaptive Re-weighted Sampling(sCARS)were used for variable selection.Random Forest(RF),Partial Least Squares(PLS)and Support Vector Machine(SVM)are used to build the model.Finally,the model was evaluated using the coefficient of determination(R^(2)),root mean square error(RMSE)and residual prediction deviation(RPD).The results showed that the R?values of all three models were>0.95,the RPD were>4,and the RMSE were<0.12.Raman spectroscopy combined with chemometrics can be used for the quality evaluation of the purification process of polysaccharides in Cistanche tubulosa.

关 键 词：管花肉蓉 多糖 拉曼光谱 纯化过程 化学计量学 

分 类 号：O657.3[理学—分析化学]