ORTHOGONAL ARRAY DESIGN AS A CHEMOMETRIC METHOD FOR THE OPTIMIZATION OF ANALYTICAL PROCEDURES AND ITS APPLICATION IN BIOLOGICAL STUDIES

Abstract

The research project began from in vitro study on the effect of mercuric chloride on Na+/K+-ATPase and Mg2+-ATPase activities in the kidney tissue of fancy carp (cyprinus carpio). This was followed by the kinetic studies of mercuric chloride on Na+/K+- ATPase activity in the kidney tissue of fancy carp (cyprinus carpio). Subsequently, a possible kinetic model for elucidating the biochemical mechanism of the inhibition of mercuric chloride on Na+/K+-ATPase in the kidney of fancy carp (cyprinus carpio) was proposed. After that, in order to meet the requirement of investigating the interaction of mercury and selenium in fish metabolism, it is expected to determine the selenium content in fish tissue. Hence, four microwave digestion methods for the determination of selenium in fish tissue by hydride generation atomic absorption spectrometry were compared. The results demonstrated that the decomposition procedures using the digestion media of HNO3/H2SO4/H2O2 enabled adequate digestion of fish tissue and retention of selenium in a state amenable for determination while the dissolution methods employing the digestion media of HNO3/H2O2, HNO3/K2S2O8/H2O2 and HNO3/H3PO4/H2O2 were unreliable. In addition, the decomposition procedures with the digestion media of HNO3/H2SO4/H2O2 were applied in the determination of selenium in fish tissue by differential pulse polarography. The findings showed that the results obtained by differential pulse polarography were in good agreement with those found by hydride generation atomic absorption spectrometry. Considering that the choice of the microwave digestion media and conditions is rather subjective, it is expected to use a chemometric method for the optimization of microwave dissolution procedures. Many chemometric strategies for the optimization of analytical procedures are available, However, the use of orthogonal array design, one of the most important experimental design approaches in chemometrics, is rather scarce. The reason for this might be that a large number of chemists are unfamiliar with the statistical steps and algorithms used in the orthogonal array design method. In view of this, we changed our research topic from "Uptake and Effect of Heavy Metal on Fish" to "Orthogonal Array Design as a Chemometric Method for the Optimization of Analytical Procedures and its Application in Biological Studies". The main purpose of this thesis is to develop a systematic theory and methodology of orthogonal array design for the optimization of analytical procedures. This study is the first to prove theoretically the orthogonality of two-level, three-level, four-level and five-level orthogonal array designs by the mathematical methods using the linear (for two-level), quadratic (for three-level), third-order (for four-level) and fourth-order (for five-level) regression models. By means of the equations obtained from the proof of orthogonality, it is very convenient to establish a polynomial model that can be used to display the represented response surface for an orthogonal array design. Then, based on the proof of orthogonality, the use of two-level, three-level, four-level, five-level and mixed-level orthogonal array matrices and the triangular tables associated with the corresponding orthogonal array matrices arc described in detail. This is followed by the data analysis strategy, in which the significance of the different factor effects is quantitatively evaluated by the analysis of variance (ANOVA) technique and the percentage contribution method. Moreover, for three-level orthogonal array design, the quadratic regression equation with derivative algorithm is used to find the optimum value for each factor that has a significant influence. For live-level orthogonal array design, Duncan's multiple F-test is recommended to decide the best experimental conditions for each factor that has a significant influence. As for four-level orthogonal array design, either polynomial model established with the derivative algorithm or Duncan’s multiple F-test can be used to determine the best/optimum experimental conditions for each factor considered: however, the computational burden of the latter method is much smaller than that or the former one. In practical applications, orthogonal array design has been applied to the optimization of many analytical procedures such as the microwave dissolution techniques for hydride generation atomic absorption spectrometric determination of selenium, the reaction system of polarographic determination of selenium, and the operation conditions of' solid phase extraction and high performance liquid chromatographic analysis of organic compounds, and so forth. The results obtained demonstrate that the orthogonal array design as a chemometric method is reliable, efficient and effective. Orthogonal array design as a type of statistical experimental design method was also successfully employed to investigate the activating and inhibitory effects of the combination of waterborne Cu, Zn Cr(VI) and Se(VI) on alkaline phosphatase (APASE) activity in the liver of red sea bream chrysophyrys major