Application Of Chemometric Approaches To The Differential Kinetic Spectrophotometric Determination Of Some Substances In Complex Systems

Differential kinetic approaches involve similar species reacting with a common reagent or undergoing a common process. Differences in the reaction or process kinetics are used to distinguish among the components without any physical separation. The major limitation of many conventional techniques for processing kinetic data is their reliance on an accurate model of the kinetics of the system under study. Such techniques require a knowledge of the reaction order and rate constants for each of the reactions in the chemical system. It is becoming increasingly clear that perhaps the most useful chemometric techniques for handling kinetic data are those that do not assume a kinetic model. In the thesis, various chemometric approaches were applied to differential kinetic spectrophotometric determination of some substances in complex systems. Among these chemometric approaches, there were those that were used to handle two-way kinetic data, such as kalman filter (KF), iterative target transformation factor analysis (ITTFA), classical least squares (CLS), ridge regression (RR), principle component regression (PCR), partial least squares regressionl (PLS 1), partial least squares regression2 (PLS2), uninformation variable elimination-partial least squares (UVE-PLS), locally weighted regression (LWR), global-radial basis function-artificial neural network (G-RBF-ANN), single-radial basis function-artificial neural network (S-RBF-ANN), global-principle component-radial basis function-artificial neural network (G-PC-RBF-ANN), single-principle component-radial basis function-artificial neural network (S-PC-RBF-ANN), back propagation-artificial neural network (BP-ANN) and so on. In addition, some other chemometric pretreatment techniques such as taking derivative and wavelet packet transform (WPT), were also applied to handle the original data in order to improve the accuracy and precision of analysis.This thesis is composed of five chapters.ChapterⅠIn chapter one, it was reviewed the application of chemometric approaches to differential kinetic analysis. Two-point and multi-point data pretreatment approaches, Kalman filter approach, H-point standard addition approach, mean centering of ratio kinetic profiles approach, multivariate linear regression approaches, multivariate calibration approaches based on factor analysis, multivariate calibration approaches based on synergism effects, artificial neural networks approaches are summarized and discussed in this section. In addition, the prospects of chemometrics in differential kinetic analysis were also given.ChapterⅡIn chapter two, a kinetic spectrophotometric approach for the simultaneous determination of iodate and periodate in mixtures was proposed. The approach is established on the different kinetic behaviours of the analytes, which react with starch-iodide in the presence of sodium chloride in sulfuric acid medium. The kinetic data were collected from 260 to 900 nm every 10 nm, within a time range of 0-180 s at 1 s interval, and the absorbance collected at 291, 354 and 585 nm, respectively, increased linearly with the concentration between 0.1-1.2 mg L^-1 for both iodate and periodate. The mechanism investigation revealed that the iodate/periodate-iodide-starch system is a consecutive reaction. Subsequently, the mathematical model for the quantitative kinetic determination based on the consecutive reactions by utilizing chemometric approaches was deduced, and it was then applied the simultaneous determination of synthetic mixtures of iodate and periodate. Kinetic data collected at 291,354 and 585 nm, and kinetic-spectra data collected in the wavelength of 260-550 nm were processed by chemometric approaches, such as CLS, PCR, PLS, BP-ANN, RBF-ANN and PC-RBF-ANN. The results showed that calibration model with the data collected at 354 nm had some advantages for the prediction of the analytes as compared with the ones of other two wavelengths, and the PLS and PC-RBF-ANN gave the lower prediction errors than other chemometric approaches. The proposed approach was applied to the simultaneous determination of iodate and periodate in several real samples; and the standard addition approach yielded satisfactory recoveries in all instances.ChapterⅢIn chapter three, a simple and sensitive kinetic spectrophotometric approach for the simultaneous determination of amaranth, proceau 4R, sunset yellow, tartrazine and brilliant blue in mixture was described for the first time. The approach is based on the chemical system involved two coupled reactions, viz. the reduction of iron(Ⅲ) by the analytes to the corresponding iron(Ⅱ) in presence of the NaAc/HCl buffer solution (pH=1.71) and the chromogenic reaction between iron(Ⅱ) and hexacyanoferrate(Ⅲ) to yield Prussian blue peaked at 760 nm. The absorbance data for the reaction were recorded at 1 nm intervals over the wavelength range 500-1000 nm every 2 s for 600 s. Under the optimized experimental conditions, the absorbance at 760 nm was utilized for constructing the calibration graph of each colorant, which was found to be linear in the range from 0.2 to 8.0 mg L^-1 for amaranth, from 0.5 to 6.0 mg L^-1 for proceau 4R, from 0.5 to 18.0 mg L^-1 for sunset yellow, from 2.0 to 44.0 mg L^-1 for tartrazine and from 0.1 to 2.0 mg L^-1 for brilliant blue, respectively. Various calibration models, such as ITTFA, PCR, PLS, UVE-PLS, LWR, PC-RBF-ANN with and without wavelet packet transform pre-treatment on the original kinetic data have been investigated to best calibrate the resulting signals from a set of synthetic colorant mixture samples. The linearity relationship between the absorbance and the concentration in the complex system were carefully investigated with the aid of the Mallows augmented partial residual plot (APaRP) approach proposed by Massart et al. Compared with the results obtained by the different calibration models, it showed that UVE-PLS with wavelet packet transform pre-treatment was the preferred models with％RPE_T=8.0. The proposed approach was applied for the simultaneous determination of amaranth, proceau 4R, sunset yellow, tartrazine, brilliant blue in foodstuff samples and compared with the reference (HPLC) approach. The results confirm the successful use of the proposed approach.ChapterⅣIn chapter four, a new spectrophotometric approach for the simultaneous determination of the important pharmaceuticals, pefloxacin and its structurally similar metabolite, norfloxacin, is described. This is the first reported spectrophotometric approach for the simultaneous predictions of the two analytes. The analysis is based on the monitoring of a kinetic spectrophotometric reaction of the two analytes with potassium permanganate as the oxidant. The spectrophotometric analysis is made possible by chemometric data interpretation and the application of multivariate calibrations. This approach overcomes the significant practical difficulties of severe spectral overlap, some consequent non-linearities as well as similarities in reaction kinetics. These problems preclude any approach with the use of conventional spectrophtometry or the direct kinetic methodology. The approach monitored the reaction process by following the absorbance decrease of potassium permanganate at 526 nm, and the increase of the product of potassium manganate at 608 nm. Under the optimized experimental conditions, the calibration curves for the individual analytes showed a linear relationship over the concentration ranges of 1.0-11.5 mg L^-1 at 526 and 608 nm for pefloxacin and those of 0.15-1.8 mg L^-1 at 526 and 608 nm for norfloxacin. Subsequently, various multivariate calibration models were applied for the simultaneous prediction of the two analytes with the aid of CLS, PCR, PLS, RBF-ANN and PC-RBF-ANN, at the two analytical wavelenghths. It was shown that PLS and PC-RBF-ANN calibrations with the data collected at 526 nm, were the preferred approaches with％RPE_T～5. The proposed approach was then applied for the simultaneous determination of pefloxacin and norfloxacin in pharmaceutical and human plasma samples with satisfactory results; these compared well with those from the alternative analysis by HPLC.ChapterⅤIn chapter five, a kinetic spectrophotometric approach is described for the simultaneous determination of isoniazid and rifampicin for the first time, based on their reduction of iron(Ⅲ) in NaAc/HCl buffer solution (pH=3.63) followed by the chromogenic reaction with potassium hexacyanoferrate(Ⅲ) to form the Prussian blue peaked at 760 nm. Different variables affecting the system under study were carefully evaluated. Under the optimum conditions, the analytical curves give a linear range of 0.1-2.6 mg L^-1 and a detection limit of 0.036 mg L^-1 for isoniazid and 0.5-20.0 mg L^-1 and 0.2 mg L^-1 for rifampicin. Subsequently, the mathematical model for the quantitative kinetic determination by utilizing chemometric approaches was derivatived in the case that the reactant of potassium permanganate and the product of potassium manganate showed highly overlapped spectral bands in the wavelength range studied. To avoid tedious separation procedure and improve the accuracy and precision of determination for both analytes, various chemometric approaches, such as CLS, KF, PCR, PLS1 and PLS2, RR, G-RBF-ANN and S-RBF-ANN, G-PC-RBF-ANN and S-PC-RBF-ANN, without any pretreatment and with the pretreatment procedure by taking derivative on the original kinetic data were introduced to model the kinetic data of the synthetic mixtures according to a three-level orthogonal design. Subsequently, the chemometric approaches were vertified with a separate set of mixtures and compared. Principal component analysis (PCA) was used to detect outliers and non-linearities of the system. The multi-criteria decision making approach, PROMETHEE and GAIA was applied for ranking the performance of the calibration models based on several different figures of merit. The results showed that G-PC-RBF-ANN and S-PC-RBF-ANN with the pretreatment procedure by taking derivative on the original kinetic data were the best performance approaches with％RPE_T～6 and％Recovery～100. The proposed approach was successfully applied to the determination of isoniazid and rifampicin in pharmaceutical preparations and human urine. The results showed no significant differences with those found by means of high-performance liquid chromatography.In general, the analytes in complex system cannot be determined simultaneous by thermodynamics equilibrium spectrophotometric techniques if the sample solutions were not pretreated by masking and separation. On the other hand, the spectra overlapped severely and will lead to large relative predition errors obtained by chemometric approaches. But nonequilibrium thermodynamics situations were investigated in differential kinetic analysis from which two-way kinetic or three-way kinetic-spectra data can be obtained and can give more opportunities and challenges to chemometrics, while the application of chemometrics in kinetic analysis can make some further developments in multicomponent kinetic analysis keep well up with modern analytical chemistry.