【作者】 卿湘东；

【导师】 吴海龙；

【作者基本信息】 湖南大学 ， 分析化学， 2014， 博士

【摘要】 环境化学计量学是一门新兴的专门针对解决复杂环境问题的综合性学科分支,它与数学、统计学和物理方法相结合,借助于先进的计算机工具和现代分析仪器,使以经验、实验为基础的环境化学问题研究趋于理论化和更加科学化,扩展了环境化学研究成果的应用。随着环境研究体系的日益复杂化和环境考察因素的逐渐多样化,海量的环境多维数据很容易获得。这些庞大的数据阵包含大量的信息,如何从这些数据阵中获取有用的环境化学信息,就促使化学计量学多维分辨和多维校正理论和方法在环境分析化学中不断地应用和发展。高维校正方法因具有独特的“二阶优势”(即使有未知干扰共存,也能对感兴趣的目标分析物进行准确快速同时定量分析),为复杂环境问题如单组分和多组分污染物的无扰动、在线实时快速定量分析以及多组分平衡与动态过程的实时解析提供了强有力的分析手段。本文作者在查阅国内外大量文献资料的基础上,通过仔细分析当前化学计量学研究热点和环境分析难点问题,对高维化学计量学方法及其在环境土壤和水体中农残和植物激素定量分析等若干方面进行了研究。本论文主要涉及内容如下：第一部分三维校正方法在土壤和水体等复杂体系中的应用研究(第2章-第4章)克百威是一种高效内吸广谱氨基甲酸酯类杀虫剂,剧毒,广泛应用于水果、蔬菜和谷物类害虫的捕杀,能防治上百种害虫,其残留通过食物链和饮用水进入人体,对人体的健康构成了一定的危害。在第2章中,通过三维荧光与基于交替三线性分解(ATLD)算法的三维校正方法相结合(第2章),提出了直接快速定量分析土壤和污水样中的克百威残留量的方法。结果显示,该方法能够解决分析物与背景基体严重重叠的问题,在存在未知和未校正干扰物的情况下,能够准确、快速地测定克百威的浓度,为监测环境农残等污染物提供了一种很有潜力的分析工具。扑草净(PRO)、敌草胺(NAP)和甲草胺(ALA)是非常重要的除草剂,常被用于高粱、棉花、马铃薯、花生和蔬菜等各种农作物农田的一年生杂草和阔叶杂草的控制。这三种除草剂在农业生产中经常会两两混合使用。因此,同时检测它们在复杂环境体系如河泥和污水样中的残留量就显得十分有意义。在第3章中,我们提出了同时、快速、有效检测河泥和污水样中PRO、NAP和ALA残留量的方法(第3章)。它通过HPLC-DAD联用仪与基于白加权交替三线性分解(SWATLD)算法的三维校正方法相结合,在含有共保留和光谱重叠组分的复杂样本中,可同时获得PRO、NAP和ALA的含量。另外,通过分析不同复杂的基体,对该方法的二阶优势和分解的相对唯一性也进行了探讨。这种基于化学计量学方法的分析策略与传统的方法相比,显示了独特的优势,如样品预处理简单、方法绿色环保以及能够获得可靠的光谱和浓度信息,为环境中农药残留量的监测提供了一种很有潜力的分析手段。2-萘乙酸(2-NOA)和1-萘乙酸甲酯(1-NAAME)是人工合成的重要植物激素。这些植物激素的广泛应用也会使得它们在环境如土壤和污水中长期累积,对人体和动物都存在潜在的毒性。因此,亟需发展一种有效的方法来检测它们在环境如土壤和污水中的残留量。基于此目的,我们首先利用快速扫描三维荧光仪来获得复杂环境样的激发发射矩阵荧光数据,然后采用基于自加权交替三线性分解(SWATLD)算法的三维校正方法从中提取有用的化学信息,从而发展了一种有效、简单、低耗的方法来同时检测土壤和污水样中的2-NOA和1-NAAME的含量(第4章)。另外,为了探索所发展方法的预测能力,采用不同的策略如基于不同环境基体和不同仪器来建立相应校正模型并对其进行比较。这种结合了化学计量学数学分离分析步骤的方法非常有潜力发展成为环境监测中有效的分析手段,并为设计智能化、便携式小型分析仪器提供了理论依据和设计基础。第二部分四维校正新方法及其用于农药水解动力学过程分析研究(第5章-第6章)西维因是一种剧毒农药,被广泛地用于农业上如储粮、景观草坪、水果和蔬菜等各种昆虫的杀灭。已经发展了一些基于荧光和色谱的方法来探索西维因的水解动力学过程。然而,迄今尚无文献报道采用室温磷光(RTP)方法来对其水解动力学过程进行研究。在第5章中,我们发展了四维RTP数据分别结合基于交替四线性分解(AQLD)、交替加权残差约束四线性分解(AWRCQLD)和平行因子分析(PARAFAC)等算法的四维校正方法来探索其降解过程的分析策略(第5章)。在存在未校正磷光背景干扰下,该方法对实际自来水样中的西维因复杂动力学过程进行了准确的定量分析。另外,我们观测到样本皆存在严重的光谱背景漂移。而这种漂移与感兴趣组分一样,已被算法拟合成单个组分而被成功地扣除。在多维数据的量测过程中,由于受到诸如环境因素如温度、压力、时间和人为因素等影响,容易导致获得的多维数据其中一维或多维偏离多线性。常见的就是色谱中保留时间漂移导致色谱一维的偏离,光谱中瑞利散射导致激发发射光谱两维的偏离。如果在实验中引进动力学和温度两个维度,由于这两个实验条件的难以控制性以及不可重复性,也可能造成两维轮廓变动而产生偏离多线性。对于严重偏离多线性加和结构的数据,直接采用基于多线性模型的算法来处理,所获得的结果往往不太准确。在第6章中,我们针对获得的色谱-光谱-动力学-温度形式的非五线性五维数据,采用了一种先拓展后处理的思路,即将该非五线性五维数阵沿着不具有线性的动力学一维方向铺展成四线性四维数阵,然后采用基于四线性成分模型的四维平行因子分析(4-PARAFAC)、交替加权残差约束四线性分解(AWRCQLD)和本章提出的交替加权四线性分解(AWQLD)算法来对其进行解析(第6章)。在处理实际数据之前,我们还通过一组模拟的激发-发射-动力学-温度荧光光谱数据和一组模拟的色谱-光谱-动力学-温度色谱数据对提出的方法进行了验证。通过测试模拟数据和真实实验数据,新发展的AWQLD算法给出了与AWRCQLD和4-PARAFAC相似甚至更好的结果。第三部分五维校正新方法及其用于多种干扰体系中萘草胺水解动力学研究(第7章)萘草胺(NAP)是一种有选择性的芽前和苗后除草剂。在水体中能降解为萘胺(NAA)、邻苯二甲酸(PHT)和N-(1-萘基)邻苯二甲酰亚胺(NPI)。NAP低毒,但NAA却具有细胞毒性和遗传毒性。先前已有报道用磷光、傅立叶转换近红外光谱和色谱的方法来检测各种基体如河水、饮用水、尿液等分析体系中的NAP及其代谢物。在过去的几十年,多维校正方法也被用于分析NAP及其代谢物,如基于PLS的二维校正方法。但二维校正方法需要大量有代表性的校正集,且校正集样本中必须包含预测样中可能含有的所有响应组分,这对于分析存在许多未知干扰物的真实体系是非常困难的。在第7章中,我们使用HPLC-DAD来监测不同pH值下样本的动力学过程,从而获得了HPLC-DAD-动力学-pH形式的五维数阵。为了探索所得到的数据的分析属性,我们发展了一种新的算法——交替五线性分解(AQQLD)算法来对其进行分析(第7章)。新发展的方法用于研究三种复杂干扰体系包括(1)包含有吲哚乙酸和萘乙酰胺作干扰的样本,(2)包含有吲哚乙酸、萘乙酰胺和未知物作干扰的样本,(3)土壤样本中NAP的水解动力学过程,都获得了比较满意的结果,为探索高维数阵的分析性能提供了一个新的很有价值的研究工作,并为五维数阵的生成和后续的五维校正应用提供了一种新的思路和策略。更多还原

【Abstract】 Environmental chemometrics is a burgeoning integrated subdiscipline which aims at studying complex and difficult problems in environmental chemistry. It combines with the methods of mathematics, statistics and physics science, and relys on the tools of advanced computer science and modern analytical instruments. Environmental chemometric methodologies not only make the study of chemistry of the environment, based on the experience and experiment, more theoretical and scientific, but also expand the application of the research in chemistry of the environment. With the increasing complication of environmental systems and the gradual diversification of environmental factors, huge of amounts of environmental multi-dimensional data can be produced in a relatively easy way. The data arrays contain a large amount of information, and then, how to extract useful chemical information from these complex data arrays promotes the continual application and development of the fundamental theories and methods of multi-dimensional resolution and calibration into chemistry of the environment. The benefit of multi-way calibration methods, known as "second-order or higher-order advantage", is capable to determine the content of analytes of interest even in the presence of unknown interferents. Such methods have provided a very potential analytical tool for direct and on-line quantitative analysis of single pollutant or mixture pollutants and their dynamic kinetic process in complex environmental systems. The work in this dissertation focuses on the research on multi-way chemometric methodologies and their applications in environmental soil and water analysis.1. Three-way calibration methods and their applications in soil and water analysis (Chapter2to Chapter4)Carbofuran (CF) is a highly effective, systemic and contact carbamate insecticide. It is deadly poisonous and widely applied to control hundreds of pests on a variety of agricultural crops such as fruits, vegetables and corns. The residue of it can reach the human population via food-chain and drinking water, which poses a threat to the health of human beings. A novel method was proposed for the rapid determination of CF in soil and wastewater using three-dimension fluorescence coupled with second-order calibration based on the alternating trilinear decomposition (ATLD) algorithm (Chapter2). It was shown that the proposed method could solve the problem of serious flourescence spectral overlapping of the sought-for analyte and background interferents and it can be applied to rapid determination of CF even in the presence of unknown, uncalibtation interference(s).Prometryne (PRO), napropamide (NAP) and alachlor (ALA) are important herbicides used on farmland for the control of annual grasses and some broad-leaved weeds in a variety of crops, including maize, cotton, potato, peanuts and some vegetables. The mixed use of the herbicides has become the main means in agricultural production. Therefore, it is very important for developing a method to quantify them in complex environmental system such as river sediment and wastewater samples. A new spectrofluorimetric method for the direct determination of PRO, NAP and ALA contents in river sediment and wastewater samples has been proposed and discussed in the chapter (Chapter3). It combines the HPLC-DAD data with second-order calibration method based on self-weighed alternating trilinear decomposition (SWATLD) algorithm. The method can simultaneously obtain the contents of PRO, NAP and ALA in complex matrices which contain coelution and overlapping spectra of compounds. Additionally, the second-order property of the adopted algorithm together with the uniqueness of the decomposition has been explored by analysis different complex matrices. Such a chemometrics-based protocol shows several advantages over the traditional methods, such as simple sample pretreatment, green, reliable spectral resolution and concentration prediction, and it may possess great potential to be extended as a promising alternative for more practical applications in environment monitoring.2-Naphthoxyacetic acid (2-NOA) and1-naphthaleneacetic acid methyl ester (1-NAAME) are important plant hormones by synthesis. The widely use of these plant hormones has led to their presence in soils and sewage waters. The potential toxicity of these compounds on humans or animals has raised the need for effective methods for their control. With the purpose of developing an effective and inexpensive method for simultaneous determination of2-NOA and1-NAAME in soil and sewage samples, firstly, fluorescence spectrophotometer was fully utilized to obtain excitation-emission matrix fluorescence (EEMF) data of complex environmental samples. Subsequently, second-order calibration method based on SWATLD was applied to extracted useful chemical information from the obtained data (Chapter4). In order to investigate the prediction quality of the proposed method, different strategies, such as taking spectroscopic measurements in the presence of different matrix interferents and at different fluorescence spectrophotometers, were introduced to build calibration models and comparisons among them were done subsequently. Such a chemometrics-based protocol may possess great potential to be extended as a promising alternative for more practical applications in environment monitoring and for the design of small intelligent and field-portable analytical instruments.2. New four-way calibration algorithms and their applications in analysis of the hydrolysis kinetic of pesticides (Chapter5to Chapter6)Carbaryl (CBL), a toxic pesticide for human systems, is widely used in agriculture to control a number of insect pests on stored grain, grass lawns, fruits and vegetables. Many fluorescence and chromatographic methods have been developed to investigate the hydrolysis kinetic of CBL. However, so far none of the published papers reported on the application of room-temperature phosphorescence (RTP) method to investigate the hydrolysis kinetic of it. A sensitive excitation-emission phosphorescence method with third-order calibration based on four-way parallel factor analysis (4-PARAFAC), alternating weighted residue constraint quadrilinear decomposition (AWRCQLD) and alternating quadrilinear decomposition (AQLD) algorithms, respectively, is proposed to investigate the hydrolysis kinetic of CBL (Chapter5). The latter was developed as an extension of alternating trilinear for decomposition (ATLD) for quadrilinear data and presented for the first time in this chapter. These methodologies are applied to investigate the hydrolysis kinetic of CBL even in the presence of an uncalibrated phosphorescence background in complex practical samples, i.e., tap water samples. Additionally, the serious spectral background drift which is observed in all the samples are overcome by means of regarding the drift as an additional factor as well as the analyte of interest in the mathematical model.In experiments, the multidimension data arrays may deviate from the multilinear structure due to changes in the environmental conditions (e.g. temperature, pressure, time, artificial factors and so on). This phenomenon often is observed in the retention time dimension of chromatography due to retention-time shifts, and the excitation or emission dimension of fluorescence due to Rayleigh scattering. When two of the data dimensions are a temperature and a reaction time, the temperature and temporal profiles may vary from sample to sample. For the data seriously deviating from multilinear structure, the classic algorithms based on multilinear model can not treat with them. In the chapter, a new method based on firstly expanding before processing is proposed to investigate the HPLC-DAD-kinetic-temperature non-quinquelinear five-way data produced in the dissertation for the first time (Chapter6). The five-way data array is firstly rearranged into an expanded four-way data array following the non-linear way, then it is decomposed by four-way calibration based on4-PARAFAC, alternating weighted residue constraint quadrilinear decomposition (AWRCQLD) and alternating weighted quadrilinear decomposition (AWQLD) presented for the first time in this chapter. Before processing real data, a simulated excitation-emission-kinetic-temperature fluorescence data and a simulated HPLC-DAD-kinetic-temperature chromatographic data are used to verify the proposed method. It is found that the results of AWQLD are similar to or ever better than the ones of4-PARAFAC and AWRCQLD.3. New five-way calibration algorithm and its application to the kinetic analysis of naptalam in the presence of interferences (Chapter7)Naptalam (NAP) is a selective pre-and post-emergent herbicide and it may degrade to1-naphthylamine (NAA), O-phthalic acid (PHT) and N-(1-naphthyl) phthalimide (NPI) in aqueous environment. NAP is estimated to be of low toxicity but NAA is well known for its cytotoxic and genotoxic effects. It has been previously reported that phosphorescence, Fourier transform infrared spectrometry and chromatographic methods can be applied to determination of NAP and its metabolites in diverse matrices such as river water, drinking water, urine, etc. Over the last decades, first-order calibration method has been employed for the analysis of NAP and its metabolites. Unfortunately, the first-order calibration methods require a sufficiently large and representative calibration sample set and the calibration samples containing all of the constituents present in the future unknown samples, which is very difficult for analysis of real systems containing many unknown interferents.In the chapter, HPLC-DAD is used to monitor the kinetic evolution of a mixture at different pH values and then the HPLC-DAD-kinetic-pH fourth-order data are produced for the first time. To process the obtained data, a new algorithm, which is called alternating quinquelinear decomposition (AQQLD) and presented for the first time, is developed in the chapter (Chapter7). The proposed methodology is applied to analyze the hydrolysis kinetic of NAP in three sample systems containing (1) Indole-3-acetic acid (IAA) and1-naphthaleneacetamide (NAD),(2) IAA, NAD and unexpected component(s) and (3) extracts from soil samples, respectively, as interferents. Satisfactory results were obtained for the determination of NAP in these samples. The present work was a new and successful example of investigating the analytical properties of fourth-order data, opening a new approach for fourth-order data generation and subsequent fourth-order calibration.更多还原