A comparison of recently developed correction procedures for electron probe microanalysis

Abstract

Three recently developed absorption and atomic number correction procedures have been investigated with regard to possible improvements and to the performance of these improved models, which has been studied depending on different values of the absorption correction factor. The test file includes 655 measurements of elements with Z > 10 and 258 B-, C-, and Nmeasurements. The results show that the models of Bastin and Heijligers and of Sewell, Love, and Scott generally are substantially better than the one of Tanuma and Nagashima or the established ZAF approach.

A characteristic fluorescence correction factor for use in electron probe microanalysis. II. Evaluation of experimental results and comparisons

“…Karttunen et al [16] found [19] and further developed by Schiebl [20] (2) According to August [24] this expression, which is similar to that presented by Ruste [25], seems to reflect the physical situation better than other quantifications found in literature. This finding is also proved by the investigations of Sevov et al [4].…”

“…It quantifies the ratio of the absorption corrections for the secondary fluorescence radiation to that of the primarily generated radiation of a given line of a given element, viz., the excited element. 4.1 THE DEPTH DISTRIBUTION FUNCTION 03A6(03C1z). -As model for the depth distribution function 03A6(03C1z) we chose to take one that is commonly used in electron probe microanalysis, viz.…”

“…the atomic number and the absorption corrections, the latter are likely to be responsible for the systematic deviations. Of course also the errors introduced by the fundamental parameters must be considered; Bastin and Heijligers [51] and Sevov et al [4] showed remarkable changes of the k'/k-values resulting from even small uncertainties of the mass absorption coeh'icients. [1].…”

“…Sevov et al [4], on the other hand, could show that in connection with electron probe microanalysis clear differences appear. In connection with the depth distribution functions proposed by Bastin and Heijligers [5] Sevov et al [4] recommended the use of the values presented by Henke et al [6] for radiation energies below 1 keV and of those of Henrich [7] for higher radiation energies.…”

A Method for Determining Detector Efficiencies In-Situ by Variation of the Radiation Incidence Angle

“…The acceleration energy, however, should not be smaller than x5 keV, because even the best of the commonly used absorption correction models (e.g., [33 to 351) generally are not applicable without restrictions to lower overvoltage ratios or acceleration energies, respectively [36].…”