L. V. Usova

Some features of X-ray registration using a JXA-8100 electron-probe microanalyzer

Specific sources of measurement errors have been studied. In the presence of perturbing radiation, the automatic setting of the spectrometer discrimination mode is not reasonable. When choosing the transmittance window, the dependence of the shift of the amplitude distribution on the counting rate should be taken into account. Losses of intensity can be significant. The value of effective dead time is 1.3 μs for spectrometers with flow methane-argon counters and 1.2–1.5 μs for spectrometric channels with sealed xenon counters. The mode of constant dead time is retained till the intensities of 100–150 kcps. The instrumental error of routine analysis is insignificant as compared with the statistical error of count collection. In the spectra from LiF crystals, a dip in intensities is adjacent to the shortwave wing of the AuLα radiation line. Measuring intensities on it results in overestimates of the analyte concentration, and line weakening, which occurs because of the spectral hole, in contrast, leads to its underestimation. For this reason, in determining small concentrations of gold, using the Lα line as an analytical one seems unreasonable.

Accuracy in the determination of the compositions of main rockforming silicates and oxides on a JXA-8100 microanalyzer

The individual sources of errors in procedures used for considering matrix effects were studied. The ZAF and PRZ analytical programs provided good correction in the predominance of one of the matrix effects, namely, absorption or the atomic number. Their simultaneous occurrence increased the total error to 2–4 rel %. In the ZAF method, normalization to oxygen in terms of stoichiometry introduced errors if the sample simultaneously contained the test element in various oxide forms. The greatest error of up to 1–2 wt % appeared in simple and double oxides. In the analysis of the main rockforming silicates with the determination of total iron as FeO, the presence of Fe2O3 can change the concentration of components to ∼0.2%.

Synthetic TRPO4 crystals as reference samples in the quantitative X-ray electron probe microanalysis of rare-earth elements

Crystals of 15 rare-earth element phosphates, which provide a set of standard reference samples in X-ray electron probe microanalysis, were synthesized and tested for homogeneity and stoichiometry. The phosphates were synthesized in two stages, the preparation of aqueous phosphates and crystal growth in a solution of potassium polymolybdate K2Mo3O10. The tests of the synthesized crystals confirmed their reliability as reference samples.

X-ray electron probe microanalysis in the vicinity of M absorption edges

Nine correction methods with five methods of calculating absorption coefficients are tested for the cases when the analytical line is located in the vicinity of the M adsorption edges of elements with atomic numbers from 76–83. Calculations were performed using the CARAT program and experiments, on the Camebacs Micro microanalyzer. It was found that none of the correction methods can maintain the appropriate accuracy within all absorption ranges. It was proposed to change the μ/ρ ratio in the corresponding inner intervals, which will ensure an accuracy of the analysis at a level of 1 rel %.

Problem of the Account of Matrix Effect in Electron Probe Microanalysis of Rock-Forming Minerals

Four methods of correction and three methods of calculation of absorption coefficients are tested in electron probe microanalysis of rock-forming minerals. Experimental data by Pouchou and Pichoir, Sewell–Love–Scott, and Armstrong are attracted in tests. It is shown that the correction factor can be calculated with the error no more than 1 rel. % if a short-wave line (matrix effect is determined mainly by the effect of atomic number) is used as an analytical line or the analytical line belongs to the absorption K-edge of elements present in the sample. In the presence of binary matrix effect, when the analytical line absorbs in the K–L1 region, the situation is more complex and additional studies are required.