Stages of arc atomic emission spectrometry development as applied to the solid geological samples’ analysis

The atomic emission spectrometry (AES) with arc discharge method evolution is inextricably linked with the fundamental scientific discoveries made in the 19th and 20th centuries, and it also reflects the change of scientific paradigms in a specific field of natural science – analytical chemistry. Theoretical comprehension and generalization of experimental data, along with the improving spectral equipment and methodological techniques for determining the elemental and material composition of solid geological samples, increased the accuracy of the analysis results i.e. the results were translated from qualitative to semi-quantitative and quantitative. Modern computerized equipment for direct AES with arc discharge provides minimal errors in measuring the spectral intensity due to the high stability of the excitation source of the spectra of atoms and molecules, the use of high-power polychromators and express digital recording of spectra by multi-channel detectors. However, in the commercial software of spectrometers, only the methods of manual spectra processing proposed in the 30s of the last century are programmed. That limits the possibilities of improving the analysis quality. The time has come to use the developed concept of computer processing of big spectral data, which is based on the information models of chemical analysis and the back propagation of error, in order to select the best models. Current article shows that the information models of computer spectra interpretation obtained from direct arc AES multi-element techniques of geological samples’ analysis using the injection-spillage method provide better quantitative results (category III of accuracy) due to a more complete account of spectral and matrix influences compared to the routine processing techniques.