【作者】 杨丽丽；

【导师】 王玉田；

【作者基本信息】 燕山大学 ， 仪器科学与技术， 2013， 博士

【摘要】 石油作为重要的能源和化工原料，在造福人类给社会带来巨大经济价值的同时，也对生态环境产生了极大的危害，海洋溢油事故频发，水中石油类物质污染严重，最终威胁到人类健康。石油类污染物质的识别和测量成为鉴别污染来源、控制污染状况，保护生态环境的重要手段。荧光光谱分析技术以其灵敏度高、选择性好、分析速度快、受风化程度影响小等特点成为石油类污染物质检测的重要手段之一。论文主要针对荧光光谱鉴别能力有限，多组分石油类物质混合体系光谱重叠严重难以区分的问题，应用三维荧光光谱法结合化学计量学中的二阶校正方法，充分利用荧光光谱信息，以对复杂体系的石油类物质进行有效识别和定量分析。分析荧光产生的机理，研究石油类污染物质的化学组成及其荧光特性，确定荧光法识别和测量原油或各种石油产品等石油类污染物的可行性。研究测量仪器的结构原理、工作性能、各种参数的设置以及散射对光谱的影响，以具有代表性的多种石油产品为测量对象，论述了从表观光谱到真实光谱的校正方法和实现手段，消除了各种因素使光谱的歪曲，并获得了样品的三维荧光光谱图。研究石油产品的荧光光谱随溶液浓度变化的现象，针对水中油类的溶解性、稳定性差的问题给出解决的办法。测量了不同的石油产品乙醇溶液及其多组乙醇稀释溶液的荧光光谱，通过对光谱特征及其差异性的分析，发现了光谱随浓度增大产生的连续红移现象，并从每种石油产品荧光成分化学组成上解释了产生这种现象的原因，确定每种石油产品的荧光强度随浓度线性变化的范围。研究了有序介质在荧光测量分析中的原理和方法，测量了每种石油产品在特定的有序介质中的荧光光谱，验证其用于增敏水中石油类物质荧光分析的可行性和定量分析的准确性。应用化学计量学中的二阶校正方法对混合石油类物质荧光光谱进行分析，以解决复杂多组分混合石油类物质的识别及定量的问题。分别以两种不同石油产品不同浓度的混合和三种不同石油产品不同浓度的混合为校正样品集和预测样品集，测量了每个样品集中各样品的荧光光谱，通过平行因子算法，应用核一致诊断法并综合考虑选择不同因子数时的解释方差，确定最佳的平行因子分析的因子数，对混合样品中特定组分的光谱进行正确的分辨。测量了混合样品加入特定的干扰成分时的荧光光谱并对其应用平行因子及自加权交替三线性分解算法进行分析，验证二阶校正法区别于其它方法的“二阶优势”。更多还原

【Abstract】 Petroleum as an important energy and chemical raw materials bring benefit to humanbeing and create tremendous economic value for society, but also does harm to theecological environment. Marine oil spill accidents take place frequently, petroleumsubstances make water polluted seriously, and human health has been threatenedultimately. Identification and measurement of petroleum pollutants have became animportant tool in identifying pollution sources, controlling pollution and protecting theenvironment.Fluorescence Spectroscopy technology is an important means of detecting petroleumpollutants with the advantages of high sensitivity, good selectivity, analysis rapidly, smallweathering effects. Thesis focuses on limited identification capacity of fluorescencespectroscopy and serious spectral overlap problem of multi-component petroleumsubstances mixed system. Three-dimensional fluorescence spectroscopy combined withthe second-order calibration methods in chemometrics is applied to identify and analyzecomplex systems of petroleum substances effectively.Making good full use of the fluorescence spectrum information, chemicalcomposition and fluorescence characteristics of petroleum pollutant are researched, andthe feasibility of using fluorescence spectroscopy to identify and measure the variety ofpetroleum pollutants such as crude oil or petroleum products is determined. Structuralprinciple, performance, various parameter settings and scattering affects to measurementof the measuring instrument are researched. Correction methods from apparent spectrumto the true spectrum to eliminate various factors making the spectral distortion arediscussed in detail and three-dimensional fluorescence spectra of samples are obtained.The varying on the fluorescence spectra of petroleum products with concentration isresearched, and the problem on poor solubility, stability of oils in water is solved. Throughmeasuring fluorescence spectra of the same kinds of petroleum products in differentsolvents and different petroleum products in the same solvent, the differences betweenfluorescence spectra are discussed. The fluorescence spectra of various different concentrations petroleum products in ethanol diluted solution are measured, and spectralcharacteristics and differences on the composition of the chemical nature from each of thepetroleum diluted solution are verified. The concentration-dependent wavelength shifts inthree-dimensional fluorescence of petroleum samples and red-shift cascade areinvestigated. Linear relationship between concentration and fluorescent intensity ofpetroleum solution are ensured. The measurement principle and method of each petroleumproducts in surfactant micelle solution as an ordered medium are researched and thefluorescence spectra of them are measured. The feasibility and accuracy of quantitativeanalysis using it to enhance the fluorescent intensity are verified.Second-order calibration methods in chemometrics are used to analyze fluorescencespectra of mixed petroleum substances and solve identification and quantification problemof complex multi-component mixed petroleum pollutants. By mixing two or three type ofpetroleum products of different concentrations as calibration sample set and predictionsample set, the three-dimensional fluorescence spectra of samples in calibration set andprediction set are measured. Combining core consistency diagnosis and explained varianceof each factor to ensure optimal number of factor, PARAFAC model is built to analyze thethree-dimensional data and fluorescence loading spectra is obtained correctly.Fluorescence spectra of mixed samples added interference components in are measured,and the second-order advantage of second-order calibration methods is verified byPARAFAC and SWATLD algorithm.更多还原