Publisher © Czech Geological Survey, ISSN: 2336-5757 (online), 0514-8057 (print)
The evaluation of precision and relative error of the main components of silicate analyses in Central Laboratory of the Czech Geological Survey
Abstract
The scheme of silicate analysis is based on using two solutions. Solution A is used for SiO2 analysis, from solution B Al2O3, MnO, MgO, CaO, Fe2O3, TiO2, Na2O, K2O, P2O5 are determined. Analytical methods of the scheme involve sample decomposition, atomic absorption spectrometry, photometry, titration with complexon III. Analyses of Fe(II), CO2, C, S, H2O-, H2O+ are needed for completness. Reliability of individual analytical methods is given using seven reference materials: calcite KH, basalt NIST 688, basalt BM, black shale TB, granodiorite JG-3, granite JAM 01, feldspar Na-K JAM 08. Relative standard deviations and relative errors obtained on seven given samples were calculated for the set of oxides from SiO2 to P2O5 and CO2. The methods for SiO2, Al2O3, CO2 reached very acceptable RSD < 1 %, 8 <5 %. The methods for MnO, Fe2O3, TiO2, P2O5 reached worse precision than 5 %, but relative errors are still not bad. The methods for MgO, CaO, FeO, Na2O, K2O reached also satisfying values on reference materials, i.e. RSD <5%, 8 <10%. Having in mind, the types of studied sample matrixes were limited to six and some published values were established on seven values only, the high standard of routine analytical methods was confirmed.
