Knowledge-based computer system for the detection of matrix interferences in atomic absorption spectrometric methods

Abstract

A knowledge-based computer system which assists with the detection of matrix interferences during the validation of atomic absorption spectrometry (AAS) is described. The system compares the slopes of a standard addition line and an aqueous calibration line. An experimental design is proposed which guarantees that the probability of not detecting important interferences (β-error) is acceptably low. To permit a detection of interferences by the comparison of curved calibration lines, a linearization is proposed. The system is developed in a Windows environment and is one module of a more complete hypertext system for the final validation of AAS methods.