天然产物研究与开发 ›› 2022, Vol. 34 ›› Issue (11): 1857-1864.doi: 10.16333/j.1001-6880.2022.11.006

• 研究论文 • 上一篇    下一篇

傅里叶变换中红外光谱结合偏最小二乘法快速测定蕨菜总多糖含量

徐惠敏1,2,郝经文1,2,3,陈乃富1,2,3*,陈乃东1,2,3*   

  1. 1安徽中医药大学,合肥 230031;2皖西学院 生物与制药工程学院;3安徽省中药资源保护与持续利用工程实验室,六安 237012
  • 出版日期:2022-11-25 发布日期:2022-11-25
  • 基金资助:
    国家自然科学基金(81573536,81274021);皖西学院大学生创新创业项目(S202010376121,wxxy2021168)

Rapid quantification of polysaccharide in Pteridium aquilinum by Fourier transform mid infrared spectroscopy and partial least square method

XU Hui-min1,2,HAO Jing-wen1,2,3,CHEN Nai-fu1,2,3*,CHEN Nai-dong1,2,3*   

  1. 1School of Pharmacy,Anhui University of Traditional Chinese Medicine,Hefei 230012,China;2College of Biotechnology and Pharmaceutical Engineering,West Anhui University;3Anhui Engineering Laboratory for Conservation and Sustainable Utilization of Traditional Chinese Medicine Resource,Lu’an 237012,China
  • Online:2022-11-25 Published:2022-11-25

摘要:

本文建立了一种快速预测蕨菜总多糖含量的方法。采用蒽酮-硫酸比色法测定140份蕨菜总多糖含量作为参比值,采集140份蕨菜样品的傅里叶变换中红外光谱图,结合偏最小二乘法,建立蕨菜总多糖定量分析模型。通过比较多元散射校正法、标准正态变换法、一阶导数法、二阶导数法、多元散射校正+一阶导数法、标准正态变换+一阶导数法、多元散射校正+二阶导数法以及标准正态变换+二阶导数法共8种不同的光谱预处理方法,运用多糖类化合物的中红外光谱学特征吸收筛选建模波段,对蕨菜多糖含量预测模型进行优化。 结果表明:采用标准正态变换法作为光谱数据预处理方法,以1 750~1 600、1 500~1 400、1 350~1 290、1 160~1 070、1 060~970、930~800 cm-1作为建模波段,获得校正集相关性系数R2为0.930 8,校正均方差(RMSEC)为0.374%,检验集相关性系数R2为0.914 5,预测均方差(RMSEP)为0.418%,20组样品进行完全外部验证误差为:-0.35%~0.31%,相对误差值为:-0.83%~5.24%,所构建定量模型可用于蕨菜总多糖含量预测。

关键词: 蕨菜, 傅里叶变换中红外光谱, 总多糖, 偏最小二乘法

Abstract:

In this paper,a method was established for rapid determination of total polysaccharides in Pteridium aquilinum.The real contents of total polysaccharides in 140 P. aquilinum samples were determined by the anthrone-sulphuric acid method.A rapid quantitative model for the contents of total polysaccharides in the P. aquilinum samples was established by fourier transform mid infrared spectroscopy (FTIR) combined with partial least square method (PLS).Eight different spectral preprocessing methods (multiplicative scattering correction,standard normal variate,first derivative,second derivative,first derivative combined with standard normal variate method,first derivative combined with multiplicative scattering correction method,second derivative combined with standard normal variate method and second derivative combined with multiplicative scattering correction method) were compared and the characteristic FTIR signals of polysaccharides were utilized as selection modeling bands,which were used to optimize the predicting model.The results showed that when the selected ranges for the model were 1 750-1 600,1 500-1 400,1 350-1 290,1 160-1 070,1 060-970,930-800 cm-1 and the standard normal transformation method was used as the preprocessing method,the correlation coefficient R2 of the calibration set was 0.930 8,the calibration mean square error (RMSEC) value was 0.374%,the correlation coefficient R2 of the test set was 0.914 5 and the mean square error of prediction (RMSEP) value was 0.418%.The error of complete external validation of 20 samples was -0.35%-0.31% and the relative error of complete external validation was -0.83%-5.24%.The established quantitative model could be used to predict the contents of total polysaccharides in P. aquilinum.

Key words: Pteridium aquilinum, FTIR, total polysaccharides, partial least square method

中图分类号:  R282.5