【作者】 方冬梅；

【导师】 吴海龙；

【作者基本信息】 湖南大学 ， 分析化学， 2005， 硕士

【摘要】 本论文首先在绪论部分概述了化学计量学的产生背景、发展历史及研究现状,然后重点阐述了化学计量学中二阶校正这一极具特色的核心研究领域,包括三维数据阵三线性模型与算法的发展,以及在分析化学领域(药物、环境、医学和食品等)中的应用。尤其近几年来,三维荧光光谱分析在复杂基质(血浆或尿液)中对药物进行无干扰直接定量测定获得了极大发展和应用。第二章是通过激发发射(Ex-Em)矩阵荧光法获取三维数据,结合化学计量学中交替三线性分解(ATLD)二阶校正方法,对干扰物依诺沙星(ENO)共存下的氧氟沙星(OFL)和诺氟沙星(NOR)进行了同时定量测定,并用于实际药物分析; 同时对血浆液中的依诺沙星(ENO)、氧氟沙星(OFL)和诺氟沙星(NOR)进行了同时定量测定。体现了化学计量学的二阶校正法能解决在未校正干扰物存在下对被分析物因光谱严重重叠难定量的问题。第三章中对血浆液中的盐酸普鲁卡因(Procaine hydrochloride)进行了测定。盐酸普鲁卡因是一种目前常用的局部麻醉药,能暂时阻断神经纤维的传导而具有麻醉作用,它的用量必须适当控制。因此,测定盐酸普鲁卡因在血浆液中的含量具有重要的意义。通过激发发射矩阵荧光法获取三维数据,结合化学计量学中二阶校正方法,对复杂基质血浆中的盐酸普鲁卡因进行了直接定量测定。实现了复杂基质血浆中的被分析物的直接定量测定,具有“二阶校正优势”,获得了满意的结果。第四章中对血浆液和尿液中的抗肿瘤药物盐酸柔红霉素( Daunomycin hydrochloride)进行了测定。盐酸柔红霉素能嵌入DNA双螺旋中与DNA结合,使其模板发生改变,抑制DNA和RNA聚合酶,阻止DNA和RNA的合成。测定盐酸柔红霉素在血浆液和尿液中的含量具有重要的意义。本章利用三维激发发射荧光光谱与化学计量学二阶校正法相结合的优势,即能够解决复杂基质血浆和尿液中的盐酸柔红霉素药物因血浆和尿液内源物质与分析物光谱重叠所引起的难分辨的问题,在复杂基质共存条件下对待测药物组分进行直接快速定量分析,获得较满意的结果。第五章应用一个新的化学计量学算法—交替归一加权残差法(ANWE)进行激发-发射矩阵荧光数据分析应用。ANWE的一个明显的优点是只要选择的组分数不少于实际组分数时,分辨的结果也很稳定。而且,能避免所谓的二阶退化问题,以及解决随机初始化或高度共线性带来的慢收敛问题。并将此算法与激发-发射三维荧光相结合对萘普生(Naproxen)为干扰物下的水杨酰胺(Salicylamide)和双更多还原

【Abstract】 In the introductory part of this thesis, the background and history of chemometrics and its research advances are illustrated. The second-order calibration in chemometrics was a core field with mighty advantages. It included the development of the trilinear decomposition algorithms as well as the applications of these algorithms to different fields of analytical chemistry, such as in pharmaceutical chemistry, environment, medicine, food and so on. In particular, three-way fluorescence (excitation-emission matrix fluorescence) coupled with second-order calibration has gained important development and applications in pharmaceutical analysis in complex matrix, e.g. plasma or urine. In Chapter 2, a method was proposed for simutaneous determination of ofloxacin (OFL), norfloxacin (NOR) in the presence of enoxacin (ENO) by combining excitation –emission fluorescence with second-order calibration based on alternating trilinear decomposition algorithm. This method was satisfactorily applied to the quantification of OFL and NOR in pharmaceutical preparations. Moreover, this method has also been used for simultaneous direct determination of OFL, NOR and ENO in plasma with satisfactory results. The results showed that this algorithm could solve the problem of serious fluorescence spectral overlapping of the sought-for analytes and determine simultaneously them even in the presence of uncalibrated interferents. An excitation-emission matrix fluorescence (EEMs) method combined with chemometrics methods was also utilized to detect directly procaine hydrochloride in plasma in Chapter 3. Procaine hydrochloride is a local anesthetic drug, which can block the conduction of nerve fibers. So it has narcotism and its doseage must be controlled. The determination of procaine hydrochloride in plasma is of an important meaning. It was showed that these methods are able to handle the occurrence of interference not modeled in the calibration set, which presents “second-order advantage”. In Chapter 4, a combination of EEMs with second-order calibration was applied for determination of daunomycin hydrochloride, an antineoplastic drug, in plasma and urine. It has characters of binding of daunomycin to DNA and the inhibition of RNA and DNA synthesis. The determination of daunomycin hydrochloride in plasma or urine has a vitally important meaning. The method was used for direct determination of daunomycin hydrochloride in plasma or urine by combining excitation-emission fluorescence with second-order calibration based on alternating trilinear decomposition algorithm. The satisfactory results can be gained in complex matrix (plasma or urine), which contains some uncalibrated interferences. In Chapter 5, a newly developed method, an alternating normalization-weighted error (ANWE), was utilized to three-way excitation-emission matrix fluorescence data analysis for second-order calibration. A prominent advantage of the method is that the resolved results are very stable with respect to the model dimensionality when the dimensionality chosen is not less than the actual number of components, escaping the dilemma in selection of an actual component number for the model. Moreover, the proposed algorithm can overcome the slow-convergence brought about by random initialization or high multicollinearity to some extent. By treating one real excitation-emission spectral data sets, the results demonstrated that the proposed method performs well as long as the model dimensionality chosen is not less than the actual number of components. This method has been satisfactorily applied to determinate salicylamide and salsalate in presence of naproxen.更多还原