中红外光谱技术对临泽小枣中多糖的定量分析

Establishment of Predictive Model for Quantitative Analysis of Polysaccharides of Linze Jujube by Middle Infrared Spectroscopy

探索基于中红外漫反射光谱技术对临泽小枣中多糖快速定量检测的可行性。对36份临泽小枣样品按传统方法测定其多糖含量,同时采集样品的中红外光谱,采用TQ-Analyst软件,对比多种建模方法与光谱预处理技术,找出最佳建模方法,并进行验证。结果表明,在进行对比之后,最终确定以偏最小二乘法(PLS)来进行模型构建,通过一阶导数+多元散射校正的方法预处理光谱,从而全面地构建多糖含量分析模型。所建立的模型相关系数R2为0.9655,校正均方差(RMSEC)为2.71,验证均方差(RESEP)为2.93,验证集相对误差最大为5.05%,建立的模型效果较好,可用于临泽小枣中多糖快速定量。基于中红外漫反射光谱技术对临泽小枣中多糖快速定量具有可行性,有具有操作简单、检测效率高、不耗费试剂等优点,可为临泽小枣的质量控制和进一步开发潜在价值奠定一定基础。

In order to determine polysaccharides content quickly and nondestructively,totally 36 Linze jujube samples were measured in laboratory based on mainstream survey method.The middle infrared spectrum of the samples was collected and analyzed with TQ-Analyst software.The results showed that the quantitative model had good prediction effect,and the prediction model had strong stability and high prediction precision with the decision coefficient(R2)as 0.9655 and standard deviation as 2.93.The mid-infrared spectroscopy technique can quickly quantify the content of polysaccharides of Linze jujube.The model which had good prediction results,was successful and could rapidly analyze the content of polysaccharides of Linze jujube.