| The development of new analytical procedures for heavy metal pollutants, pesticide residues and pharmaceutical components in complex environmental and pharmaceutical matrices are highly demanded. Spectrophotometry is one of the most favorable techniques, due to its simplicity, economy, high accuracy and precision for both organic and inorganic components analysis. However, matrix effects and the interaction between analytes are very often problematic for real sample analysis, including environmental samples and Pharmaceuticals. The separation of analytes with the aid of chemometric methods is an ideal approach for eliminating matrix effects and resolving analytical responses of mixtures by mathematical treatment. At this point, sequential injection spectrophotometry coupled with artificial neural networks (ANN) was used to determine multi-components in a single run.A combination of spectrophotometry and appropriate chemometric methods was widely applied in the multicomponent analysis, due to simple operation, low cost instrument, reasonable accuracy and precision and simultaneous determination of multiple species without carrying out any complex pretreatments. Among these approaches, an excellent alternative is the use of ANN algorithms by their high efficiency as predictors for non-linear systems. After proper training, ANN can accurately model the non-linear behavior between input variables and responses and eliminate or reduce the effects of the interaction between analytes, the synergistic effect (non-additivity of reaction rates), the multi-step process and any other unknown non-linearity. However, manual procedures are tedious and time-consuming, and are sometimes complicated. Such disadvantages in manual procedures can be easily overcome by flow injection analysis (FIA), since FIA technique can precisely and reproducibly control the timing of the measurement of the chemical reaction and the mixing of the solutions of reagent and sample with rapid and low consumption of sample throughput. It is also highly versatile, in terms of the types of data it can provide. Sequential injection analysis (SIA), as a branch of sampling means of FIA, the advantages of SIA over FIA are the simpler flow manifold, the lower consumption of reagents, the easier and more convenient PC-control of the instrumental parameters and the greater potential for fluidic handling. All these superiorities have made SIA technique coupled with spectrophotometry and ANN calibration model a powerful analytical tool in multicomponent analysis. In this thesis, three simple, rapid and reliable procedures for the simultaneous determination of multiple species in complex environmental and pharmaceutical matrices were proposed.Chapter 1 of the thesis briefly reviewed the development of FIA method, and the strategies used in flow systems to achieve multicomponent analysis include solid phase extractions, sequential injection chromatography, multichannel manifolds, kinetics, selective detectors and chemometrics.In Chapter 2 of the thesis, a stopped-flow sequential injection spectrophotometry for the simultaneous determination of Cu(Ⅱ), Zn(Ⅱ), Ni(Ⅱ) and Mn(Ⅱ) using an ANN algorithm for multivariate calibration was developed. The procedure was based on the reaction of these metal ions with 4-(2-pyridilazo) resorcinol in an alkaline medium to form colored complexes. The degree of approximation, a new evaluation criterion of the networks parameters was employed to prove the accuracy of the predicted results. The experimental variables were studied and determination ranges were obtained in the concentration range of 0.1-2.0 mg L’1 for each analyte, at a sampling frequency of 30 h-1, respectively. The proposed method was validated and the determination results for environmental water samples were statistically similar to those from the atomic absorption spectrometry. The results obtained for analysis of reference samples of environmental water were in good agreement with the certified values.In Chapter 3 of the thesis, a new procedure was developed for the simultaneous kinetic spectrophotometric determination of a quaternary carbamate pesticide mixture consisting of carbofuran, propoxur, metolcarb and fenobucarb using SIA. The procedure was based upon the different kinetic properties between the analytes reacted with reagent in flow system in the non-stopped-flow mode, in which their hydrolysis products coupled with diazotized p-nitroaniline in an alkaline medium to form the corresponding colored complexes. The absorbance data from SIA peak time profile were recorded at 510 nm and resolved by the use of ANN for multivariate quantitative analysis. The experimental variables and main networks parameters were optimized and each of the pesticides could be determined in the concentration range of 0.5-10.0μg mL-1, at a sampling frequency of 18 h-1. The proposed method was compared to other spectrophotometric methods for simultaneous determination of mixtures of carbamate pesticides, and it was proved to be adequately reliable and was successfully applied to the simultaneous determination of the four pesticide residues in water and fruit samples, obtaining the satisfactory results based on recovery studies (84.7-116.0%).In Chapter 4 of the thesis, a sequential injection pH gradient spectrophotometry was proposed for the simultaneous determination of vitamin C and paracetamol in pharmaceuticals. The method was based on the significant differences between absorption spectra of vitamin C and paracetamol at different pH, and a pH gradient in the range of 9.2-10.9 was produced by the dispersion of NaOH solution in a flow system using SI technique. ANN algorithms were applied to resolving overlapping spectra of the two pharmaceutical components and to the simultaneous determination of their concentrations. Some experimental variables including NaOH concentration and its volume, sample volume, flow rate were studied, respectively. Under the selected experimental conditions, the determination ranges were obtained in the concentration range of 1.0-10.0 mg L-1 for vitamin C and paracetamol, with the limits of detection of 0.3 and 0.2 mg L-1 at a sampling frequency of 40 h-1, respectively. The proposed method was applied to the simultaneous determination of vitamin C and paracetamol in vitamin C yinqiao tables, and the determination results were not statistically different to the claimed values and the recoveries for all components were in the range of 92.0-109.1%.In conclusion, this thesis demonstrated a multicomponent analytical method by sequential injection spectrophotometry with artificial neural networks multivariate calibration model. These procedures are simple, rapid, accurate and low reagents and samples consumption, and they were successfully applied for the determination several analytes in environmental and pharmaceutical samples in a single run. Thus, it infers that the proposed method provides a very interesting way for the simultaneous quantification of multiple species of the different thermodynamic, kinetic or pH properties in non-linear systems. |
PDF Full Text Request