【作者】 柯博克；

【导师】 瞿海斌； 刘雪松；

【作者基本信息】 浙江大学 ， 药物分析学， 2006， 硕士

【摘要】 目前中药生产过程中缺乏快速、有效的质量分析技术及控制手段,严重制约了中药产业的发展。建立中药生产关键环节中的快速质量分析方法,有助于解决上述难题、提升产业质量控制水平,进而推进中药现代化进程。近红外光谱分析技术是目前发展迅速、应用前景广阔的一种快速、无损的分析技术,已逐步应用于中药产品的生产过程分析,有望解决中药生产过程中的快速检测难题。然而,对近红外光谱分析技术在中药生产关键环节中的应用尚缺乏深入研究。本文就近红外光谱分析技术在中药硅胶柱层析过程分析中的应用展开研究,以期建立硅胶柱层析过程的快速分析方法。 本文针对中药硅胶柱层析生产过程中活性成分洗脱时间的判断难题,以白芍提取液硅胶柱层析过程为例,研究建立了基于近红外光谱分析技术的中药柱层析过程中活性成分洗脱时间的快速判断方法。通过采集白芍提取液硅胶柱层析洗脱液的近红外光谱,以硅胶薄层色谱检测为参照,用马氏距离判别法对近红外光谱进行快速分类,判断洗脱液中是否含有芍药苷或芍药内酯苷,进而确定芍药苷或芍药内酯苷的收集时间。光谱分类正确率为96.4%,可望指导实际生产。 本文建立了基于近红外光谱分析技术的烟叶提取液硅胶柱层析洗脱液中茄尼醇的快速分析方法,解决了柱层析洗脱液中活性成分洗脱时间和含量的快速检测难题。先用DPLS(PLS Discriminant)对所得光谱进行定性分类,判断茄尼醇的洗脱时间,再用偏最小二乘算法回归(PLSR)建立定量校正模型,并描述硅胶柱层析过程中茄尼醇的洗脱曲线。结果表明所建模型的预测准确性好,符合HPLC分析值。 本文针对烟叶提取液硅胶柱层析过程中茄尼醇含量的在线监测难题,利用近红外光谱分析技术建立定量校正模型,解决了硅胶柱层析洗脱液中活性成分含量的在线分析问题。此方法具有实时在线、无损和准确的优点,可直接对柱层析过程中的洗脱液进行在线监测。烟叶提取液柱层析洗脱液中茄尼醇含量的在线监测结果表明,所建模型的预测结果与HPLC分析值比较接近,可用于在线监测中药硅胶柱层析生产过程中活性成分的含量。更多还原

【Abstract】 The progress of TCM (Traditional Chinese Medicine) is blocked by the lack of fast and effective quality analytical techniques in manufacturing process. Researches on key manufacturing processes of TCM can help to solve the problem mentioned above, improve the industry’s quality control level and promote the modernization of TCM. As a fast, nondestructive analytical method, NIRS (Near Infrared Spectroscopy) have been gradually applied in manufacturing process analysis of TCM in the past few years, and provided a possible solution for the problem. However, the researches and applications of the technology were not in-depth. In this thesis, analysis of silica-gel column chromatography process in TCM was studied, expecting to establish a fast quality analytical method using NIRS.This thesis aimed at the problem of the timing for collection of active ingredients in silica-gel column chromatography process.. Radix Paeonia Alba extract was studied as an example for establishment of a fast qualitative analysis method using NIRS. TLC (Thin layer chromatography) was used as a reference method, determining whether the elutes contained paeoniflorin or albiflorin. MD (Mahalanobis Distance) method was used to classify near infrared spectrum collected from column chromatography process. The accuracy of the established model was 96.4 %. This method was expected to supervise industrial manufacturing process.This thesis also established a fast analytical method for determination of the concentration of solanesol in Tobacco leave extract in silica-gel column chromatography process using NIRS. DPLS (Partial Least Squares Discriminant) was utilized to determine the time when solanesol was eluted, then PLSR (Partial Least Squares Regression) was utilized to construct the quantitive calibration model, which was applied to describe the elution curve of solanesol in Silica-gel column chromatography process. The results showed high accuracy close to the value measured by HPLC (High Performance Liquid Chromatography) .On-line NIRS was used for monitoring and measurement of the concentration of solanesol for Tobacco leave extract’s silica-gel column chromatography process in this thesis. The proposed on-line method was nondestructive and accurate, could analyze silica-gel column chromatography directly and fast. The results were close to HPLC determined value, and competent to bedeveloped as an on-line monitoring and feedback-control method for TCM extracts’ silica-gel column chromatography.更多还原