Errors in standardless analysis arise from inaccuracies in atomic data, detection efficiency of the detector, the normalization and the sensitivity to the errors of the spectrum processing, and ZAF correction (which converts the ratio of the measured intensities into concentration). Other sources of error encountered in fully quantitative analysis (using standards) also remain in the standardless version. Calculation of the Lα standard intensities also requires incorporation of non-radiative transitions. Both the experimentally found dependence of the net Lα intensities on the atomic number and the corresponding peak-to-background ratios (P/B) indicate that an empirical increase in the calculated non-radiative transitions is necessary. Either the subshell ionization cross-sections are higher for the L1 and L2 subshells than they are for the L3 subshell or the Coster-Kronig transition rates are larger than reported in the literature. Independent experimental data would be needed especially for the ionization cross-sections of the individual L subshells. Experimental data pertinent to the conventional energy dispersive (ED) detectors with a Be window are discussed in this paper. The average error of the standardless analysis is 3-10% which should be compared with the average error of the fully quantitative ED analysis, which is 2-6%, depending on the ZAF procedure used. An extensive comparative assessment of the standardless procedures is urged.
Lábár, János L.
"Standardless Quantitative Analysis by X-Ray Spectrometry,"
Scanning Microscopy: Vol. 1993
, Article 10.
Available at: https://digitalcommons.usu.edu/microscopy/vol1993/iss7/10