Development of analytical procedures with industrial samples for atomic emission and atomic absorption spectrometry

Abstract

In this thesis a new approach is described for the validation of an analytical method. Emphasis in this study has been placed on the calibration and specificity of the method. Specificity is ensured by building an interference matrix, which is used along with multiple linear regression to find potential interfering elements. In addition to successfully find the potential interferents, it is essential that the practical impact of these factors are known; the impact is controlled using multiple linear regression equations. One key element in this study was the elimination of critical multiple linear regression assumption violations at various stages. It has been found that the utilizitation of these procedures is beneficial when extreme accuracy and reliability is required in industrial analysis.