Uncertainty of the Results of Electron Probe Microanalysis using a Spectrometer with Wavelength Dispersion in the Study of Geological Samples

Minwu Liu,Xijuan Tan
DOI: https://doi.org/10.1134/S1061934822100094
IF: 1.237
2022-10-08
Journal of Analytical Chemistry
Abstract:In this paper, we presented a detailed study on the effect of electron beam diameter, spectral overlapping, and element assay sequence on quantitative uncertainties of electron probe microanalysis. With a fixed electron beam diameter of 1 μm for external standard materials, the sample signal intensities were found to decline with the increasing electron beam diameter from 1 to 300 μm, resulting in a sample concentration bias higher than 2% and even up to 20%. Only when electron beam diameter difference between external standard materials and analyzed samples was less than 20 μm, the concentration bias was negligible. The spectral overlapping interference investigation revealed that in the case of shoulder-peak overlapping it was possible to remove the concentration bias by appropriately choosing L value and background signal picking-up position. However, quantification calibration was mandatory when there existed on-peak spectral overlapping interference. For sphalerite samples, this effect was prominent, and the artificially produced Na 2 O concentration can be 6.23% (±0.30%, 2σ, n = 12) due to on-peak spectral interference of Zn on Na. Furthermore, because of the on-peak spectral interference of S on Mo, the quantification accuracy of Mo was found to be dependent on the mass fraction of S in the studied samples. We also observed that element assay sequence highly affected the quantification accuracy of geological samples containing alkali. In element composition analysis of albite samples, when using matrix-matched calibration standard, the average Na 2 O concentration decreased substantially from 11.55 to 5.61% when Na was measured the first and last, respectively.
chemistry, analytical
What problem does this paper attempt to address?