目的:利用红外光谱定量分析模型快速测定进口血竭中血竭素的含量.方法:根据进口血竭在1135.87~1025.94/cm、779.10~719.32/cm、715.46~649.89/cm的中红外吸收光谱以及4485.61~4420.05/cm的近红外吸收光谱,采用偏最小二乘法(PLS)建立了校正模型.采用二阶导数预处理方法时能最有效的提取光谱的信息.结果:中红外定量校正模型中血竭素的校正集标准偏差(RMSEC)为0.034,校正集相关系数为0.9995,预测集标准偏差(RMSEP)为0.536,预测集相关系数为0.9061;近红外定量校正模型中血竭素的校正集标准偏差为0.169,校正集相关系数为0.9939,预测集标准偏差为0.351,预测集相关系数为0.9843.结论:红外光谱技术快速简便,适合中药指标成分的快速分析.%Objective:To determine the contents of dracorhodin in foreign XueJie(Daemonorops draco Bl.) rapidly by using infrared spectrum quantitative analysis model. Methods: Based on Mid-infrared absorption spectrum of foreign XueJie ranged from 1135.87 to 1025.94 cm-1, 779.10 to 719.32 cm-1 and 715.46 to 649.89 cm-1 and its near infrared absorption spectrum at 4485.61~4420.05 cm-1, correction model was built by partial-least-square (PLS). Spectral information could be extracted most effectively through adopting second derivative pretreatment method. Results: In mid-infrared quantitative calibration model, the RMSEC was 0.034, the calibration set correlation coeffi-cient was 0.9995, the RMSEP was 0.536, the prediction set correlation coefficient was 0.9061; and in near-infrared quantitative calibration model, the RMSEC was 0.169, the calibration set correlation coefficient was 0.9939, the RMSEP was 0.351, the prediction set correlation coefficient was 0.9843. Conclusion: Infrared spectroscopy is rapid and simple, suitable for rapid analysis of the index and ingredients of Chinese herbs.
展开▼
