【作者】 彭韵燕；

【导师】 倪永年；

【作者基本信息】 南昌大学 ， 分析化学， 2007， 硕士

【摘要】 本论义阐述了近年来化学计量学在复杂药物体系中的应用情况,分别介绍了化学计量学中的多元曲线分辨和化学模式识别方法的发展和应用,并将多元曲线分辨用于高效液相色谱-二极管阵列检测器(HPLC-DAD)的二维色谱-光谱数据的解析,将化学模式识别用于建立和解析中药指纹图谱,从而对中药进行质量控制提供一定的理论基础。第一章对多元曲线分辨和中药指纹图谱的研究现状及发展前景进行了综述,主要叙述了多元分辨算法近年来的发展和应用,以及中药指纹图谱的概念、特点、研究内容和在中药质量控制中的意义,并对化学模式识别方法在中药指纹图谱中的应用进行了详细评述。第二章采用高效液相色谱建立黄连上清片的色谱指纹图谱,利用基于主成分分析的投影判别法分析实验结果,并用反传人工神经网络对未知样品进行预报。结果表明,不同厂家生产的黄连上清片存在显著差异,主成分分析投影判别法能对样品进行正确分类,从而建立了识别不同厂家黄连上清片的方法,能有效地控制中药黄连上清片的质量。此外,主成分分析还用于优化反传人工神经网络,统计多次预报的结果,表明经过优化的反传人工神经网络能对未知样品的来源进行准确预报。第三章采用高效液相色谱-二极管数组检测器获得46个杜仲药材的色谱指纹图谱。这些杜仲样品收集于中国八个不同的省份,存在地理位置不同、气候条件不同、地形不同以及土壤条件不同等差异。七个共有色谱峰用来构建这种传统中药的指纹图谱,质谱技术用来鉴定杜仲样品中的共有化合物,最终确定了其中六个。从主成分分析的结果来看,四川、湖北、陕西和安徽四省的样品聚为一类。其它来自贵州、江西、甘肃和河南的样品差异较大。结果表明,杜仲样品所含的物质取决于它们的地理位置和环境。主成分投影双图反映出来的结果也得到直观聚类分析的支持,这表明,本文提出的方法可以对杜仲样品进行成功地分析。第四章首次将高效液相色谱资料和原子发射光谱资料结合起来研究中药材,实现了两种不同仪器资料的有效结合,并将其应用于中药的二维指纹图谱的构建。对构建的二维资料进行主成分投影,并将其结果与单独的色谱主成分投影结果和原子发射光谱主成分投影结果进行比较,结果表明,本文提出的方法能成功的对不同的苍术样品按一定的目的进行正确的分类,从而能有效地控制中药苍术的质量,同时为其它中药的化学模式识别提供了参考。本文还采用了三种有监督模式识别方法对资料进行分类,结果与主成分分析结果基本一致。第五章采用高效液相色谱联用二极管数组检测器,同时测定了饲料中氧氟沙星、依诺沙星以及甲磺酸培氟沙星等三种喹诺酮类药物的含量。测定过程中发现各组分色谱峰相互重叠而影响测定结果,但它们的紫外吸收光谱在320-380nm波长范围内存在较大差异,因此本文采用经典最小二乘法(CLS)和演进特征投影法(HELP)进行重叠色谱峰解析。利用本法对饲料中的三种抗生素添加剂进行了测定,可为饲料的质量控制提供参考。更多还原

【Abstract】 The development and application of chemometrics multivariate curve resolution and chemical pattern recognition in recent years are reviewed in this article. Moreover, multivariate curve resolution methods are applied in resolution of the two-dimensional data obtained by high performance liquid chromatography with diode array detection (HPLC-DAD), and pattern recognition methods are employed to establish and resolve the fingerprints of traditional Chinese medicine, which could be used for the quality control of Chinese medicine.In the first chapter, the development and application of multivariate curve resolution and chemical pattern recognition are summarized. A review on research status and development foreground was arisen, which the conception, characteristics, research content and its importance for quality control of TCM were introduced. In addition, the pattern recognition methods applied for fingerprint analysis of TCM were also specially discussed.In the second chapter, high performance liquid chromatography was adopted to develop the fingerprint of Huanglianshangqing Tablets. In this study the technique of projection discriminant analysis based on principal component analysis (PCA) was applied to investigate the experimental data, and back-propagation network (BPN) was applied to predict the attribution of unknown samples. The results based on principal component analysis showed that there were obvious differences among the Huanglianshangqing Tablets produced by different manufacturers, which established a method to distinguish Huanglianshangqing Tablets produced by different manufacturers, and the method developed can be effectively used to control the quality of traditional Chinese medicine Huanglianshangqing Tablets. Besides, the predicted results based on optimized back-propagation network can be correctly used to predict the attribution of unknown samples.In the third chapter, chromatographic fingerprints of 46 Eucommia Bark samples were obtained by HPLC-DAD. These samples were collected from eight provinces in China, with different geographical locations, and climates. Seven common HPLC peaks that could be used for fingerprinting this common popular traditional Chinese medicine were found, and six were identified as substituted resinols (4 compounds), geniposidic acid and chlorogenic acid by HPLC-MS. Principal components analysis (PCA) indicated that samples from the Sichuan, Hubei, Shanxi and Anhui - the SHSA provinces, clustered together. The other objects from the four provinces, Guizhou, Jiangxi, Gansu and Henan, were discriminated and widely scattered on the biplot in four province clusters. Thus, such results suggested that the composition of the Eucommia Bark samples was dependent on their geographic location and environment. These observations from the PCA biplot were supported by hierarchical cluster analysis, and indicated that Eucommia Bark preparations may be successfully compared with the use of the HPLC responses from the seven marker compounds and chemometrics methods such as PCA and the complementary HCA.In the fourth chapter, Chinese medicine is studied by use of the combined data derived from HPLC-DAD and inductively coupled plasma-optical emission spectrometry (ICP-OES) to establish the two-way fingerprint. With the application of principal component analysis (PCA), it was found that, HPLC fingerprints could be used to explain the processed methods of Atractylis chinensis DC, and the ICP-OES metal ions data could be used to explain its origins. Moreover, from the principal component projection plot of the two-way data of 51 Atractylis chinensis DC. samples, not only the four main classes based on processed methods could be classified, but also sub-classes based on origins could also be observed sequentially. The proposed method could be successfully applied to classify the Atractylis chinensis DC. samples into different groups that correlated to its properties, and it could be used for the quality control of Atractylis chinensis DC. effectively. Moreover, three supervised pattern recognition algorithms were applied to process the data, and similar results could be obtained.In the fifth chapter, Ofloxacin, Enoxacin and Pefloxacin Mesylate in feedstuff were determined by HPLC-DAD. During the determination process, it was founded that the results were affected by the over-lapped chromatography peaks, therefore, classic least-squares (CLS) and heuristic evolving latent projection (HELP) techniques were applied to resolve the over-lapped chromatography peaks based on the slight difference of their spectrum in the range 320-380nm of the two-dimensional HPLC-DAD datum. The results obtained show that CLS and HELP can be applied in resolution of seriously overlapped chromatographic peaks of multicomponent system. Furthermore, the determination of three antibiotic feed additives can be regarded as a reference method for quality control.更多还原