【作者】 孙剑奇；

【导师】 吴海龙；

【作者基本信息】 湖南大学 ， 分析化学， 2003， 硕士

【摘要】 本论文概述了化学计量学多元分辩与校正这一分支的发展情况以及研究现状，重点描述了采用交替最小二乘原理的三维数据即立方阵数据三线性分解算法近年来的发展和应用，在此基础上，将立方阵三线性分解算法与高效液相色谱-二极管阵列检测(HPLC-DAD)相结合，应用于多样品复杂多组份体系的二阶校正研究，以解决环境污染物、药物及药物代谢物等样品分析中由于混合物色谱、光谱严重重叠所引起的组份分离和分析极为困难的问题，实现了化学计量学以“数学分离”代替或增强“化学分离”的设想，展示了二阶校正问题中，立方阵三线性分解方法所具有的“二阶优势”的迷人魅力，此外，还提出了一个采用交替最小二乘原理的新的立方阵数据三线性分解算法，也用以解决传统平行因子分析方法(PARAFAC)的收敛缓慢和对组份数的估计敏感等问题。 第一章，绪论，概述了化学计量学多元分辨与校正这一分支的发展情况以及研究现状，重点描述了采用交替最小二乘原理的三维数据即立方阵数据交替三线性分解算法近年来的发展和应用，以及本论文研究工作开展的意义。 第二章，酚类化合物是重要的化工原料，同时也是一类广泛存在的环境污染物，因此对酚类物质进行定量分析具有较为重要而现实的意义。邻、间及对苯二酚的化学和物理性质十分相似，难以用常规的化学或物理方法分离及同时定量测定，本文利用交替三线性分解算法与反相高效液相色谱-二极管阵列检测相结合，对苯二酚位置异构体的重叠色谱及光谱体系进行了分辨研究，同时测定了水溶液中共存的邻苯二酚、间苯二酚和对苯二酚的含量，其回收率分别为(101.3±2.0)％，(98.3±1.3)％，(100.6±1.4)％，研究结果表明：该方法定量快速准确，实验操作步骤简单，解决了背景干扰存在条件下三者同时分辨难的问题，同时也说明了化学计量学在分析科学中具有广泛的应用前景。 第三章，水杨酸类化合物的研究一直是一个比较活跃的领域。阿司匹林能够止痛，同时也是一种抗菌消炎试剂并具有退烧作用。水杨酸和龙胆酸是阿司匹林的主要代谢产物，而2，3-二羟基苯甲酸作为另一种活性医药成份有可能与水杨酸和龙胆酸在人体内共存。本文利用交替三线性分解算法结合高效液相色谱-二极管阵列检测在与待测物性质相似的未知干扰物2，3-二羟基苯甲酸共存条件下同时测定了水溶液中的水杨酸和龙胆酸，其回收率分别为(102.2±6.7)％和(102.1±4.1)％，分辨结果和实际结果相一致。实验结果表明，本实验操作简单，定量快速准确，实现了未知干扰物存在条件下，色普及光谱严重重叠的复杂分沂体系 中感兴趣组份的直接同时测定，实验也表明交替三线性分解、ATLD)算法是有 效解决高效液相色谱一二极管阵列检测中二阶校正问题的途径之一。 第四章，对食用动物过量使用磺胺类药物会导致其在动物产品中的残留，这些残留物最终会对人体造成不良影响，因此，对磺胺类药物的医药剂量进行质量控制，对其在体内的变化进行监测便成了一项非常重要的分析任务。本文利用交替三线性分解算法与高效液相色谱一二极管阵列检测相结合，在与待测物性质相似的未知干扰物磺胺共存条件下同时测定了人工合成样及增效联磺片实际样中的磺胺嗜咤及磺胺甲嘻哇，并同时给出了分析物的光谱及色谱分辨轮廓，实验表明，本实验在不增加经济和时间耗费的前提下，可以“数学分离”增强传统的 “化学分离”，在未知干扰物共存条件下，对感兴趣的多组份进行了直接的同时定量测定。表明了交替三线性分解(ATLD)算法具有收敛速度快、对体系因子数估计不敏感的特点，能有效解决高效液相色谱一二级管阵列检测中的二阶校正问题 第五章，原先的R气RAFAC算法收敛速度较慢且对因子数的选择十分敏感，本文通过交替优化三个具有很强内在联系的目标函数，新提出了用于立方阵数据分析的交替同时最小化对角矩阵误差及协变矩阵误差(A SMDCME)算法，并通过解析模拟的HPLC一DAD数据及真实的HPLC一D人D数据验证了ASMDCME算法的特点。结果表明ASMDCME算法与传统的RA凡廿AC算法相比具有收敛速度快，_且对体系过高因子数估计不敏感的特性，为二阶校正及研究立方阵数据表征的化学体系或过程提供了有价值的工具。更多还原

【Abstract】 In this thesis, the developments of muitivariate resolution and calibration in chemometrics. especially the iterative trilinear decomposition algorithms for the three-way data analysis based on alternating least squares principles and their applications have been reviewed. The methods for three-way (cubic) trilinear decomposition have been applied to the second-calibration using high performance liquid chromatography-diode array detection (HPLC-DAD). The problems arising from the serious overlapping of spectra and chromatograms are successfully resolved during the analysis of environment contamination, remedies and their metabolites, and so on. The assume replacing or improving the "chemical separation" with ’’mathematical separation" and the charming of three-way data analysis from "second-order advantage" is exhibited in second-order calibration. In addition, an improved iterative trilinear decomposition making use of the alternating least squares principles has been proposed to overcome the problems of slow convergence and being sensitive to die estimation of factor number of the traditional PARAFAC algorithm.In Chapter 1, die developments of muitivariate resolution and calibration in chemometrics, especially the trilinear decomposition algorithms for the three-way data analysis and their applications have been reviewed.It is realistic and important to determinate quantitatively phenols due to their pollution widely existing in nature. Dihydroxybenzenes are hard to be determined simultaneously by routine chemical or physical methods because of their similarities of physical and chemical properties. In Chapter 2, the overlapping chromatograms and spectra of dihydroxybenzenes were resolved by alternating trilinear decomposition (ATLD) algorithm combined with HPLC-DAD, determining catechol, resorcinol and hydroquinone in aqueous solution simultaneously with their corresponding recoveries being (100.6±1.4)%, (98.3±1.3)% and (101.3±2.0)%, respectively. The results showed the experiment could be easily performed without resorting to time-consuming and complicated procedures and quantified accurately, resolving the problem arising from the overlapping chromatograms and spectra of analytes of interest. It also showed the promising application of chemometrics in analytical sciences.Salicylic and gentisic acids are major metabolites of aspinn. which may coexist with pyrocatechuic acid in biological fluids that serves as another active pharmaceutical ingredient. In chapter 3, ATLD method combined with HPLC-DAD is used to simultaneously determinate salicylic and gentisic acids in the presence of unknown and uncalibrated pyrocatechuic with their chromatograms and spectra seriously overlapping. A satisfying result has been obtained. The method successfully accomplished the direct simultaneous determination of analytes of interest in the presence of interfering components without expensive or time-consuming procedures with their recoveries being (102.2 ± 6.7)% and (102.1 ± 4.1)%, respectively. It is indicated ATLD can be efficiently applied to the second-order calibration in HPLC-DAD.Sulfonamides exceedingly used in the treatment of edible animals will lead to their residue in products, which will do harm to human ultimately. Quality control of sulfonamide formulations and their monitoring in body fluids are important analytical tasks. In chapter 4, sulfadiazine and sulfamethoxazole in synthetic samples and pharmaceutical tablets were simultaneously determined in the presence of unknown sulfanilamide by HPLC-DAD with easy and fast performance and accurate quantity. The study showed the fast convergence and insensitive to the estimation of component number of the ATLD algorithm and "second-order advantage" of the second-order calibration in HPLC-DAD.In chapter 5, a novel algorithm, alternating simultaneously minimized diagonal and covariant matrices’ errors (ASMDCME) algorithm has been proposed for three-way data analysis. Simulated and real HPLC-DAD data have been carried to investigate the performance o更多还原