【摘要】 诃子是一种重要的传统药用植物,它的有效成分没食子酸具有抗菌、抗病毒以及抗癌的作用.本试验旨在利用近红外光谱技术建立诃子中没食子酸含量的预测模型,将89个样品分为校正集(73个样品)和验证集(16个样品).结果表明,选择950 nm~1650 nm光谱,标准正态变量变换法处理光谱时所建的预测模型效果最佳,其校正集和交互验证的相关系数分别为0.986和0.960.利用该模型对16个外部验证集样品进行预测,结果显示预测值与HPLC测定值之间有好的相关性,其相关系数为0.958.T检验的结果表明二者差异不显著(P=0.919>0.05).说明该方法能够用于诃子中没食子酸含量的快速测定中.更多还原

【Abstract】 Terminalia chebula is an important plant species in Chinese traditional medicine. Its sarcocarp is rich in gallic acid, a very important chemical substance that has the effects on anti-bacterial, anti-viral and anti-tumor. The aim of this study was to build an evaluation method rapidly identifying gallic acid content in T. chebula with near infrared diffuse reflectance spectroscopy(NIR). A total of 89 samples were collected and used as calibration(n=7)and prediction(n=16)sets. The applied wavelength range of near infrared reflectance spectroscopy is 950 nm-1650 nm. Partial least squares regression was used to develop model to forecast the content of gallic acid. The best model was built by comprehensive analysis. The correlation coefficients of calibration set and cross validation set were 0.986 and 0.960 respectively. The model was further tested with the external validation set including 16 samples. It was found that the correlation coefficient between the predicted and the measured values was 0.958. For the given significance level 0.05, the result showed that they had no significance difference by paired samples T test(P= 0.919>0.05). The equations developed in this study demonstrate that NIR can effectively predict gallic acid content in T. Chebula.更多还原