Quantitative or semi‐quantitative?–laboratory‐based WD‐XRF versus portable ED‐XRF spectrometer: results obtained from measurements on nickel‐base alloys

Abstract

Semi-quantitative' analytical procedures are becoming more and more popular. Using such procedures, the question of the accuracy of results arises. The accuracy of an analytical procedure depends to a great extent on spectral resolution, counting statistics and matrix correction. Two 'semi-quantitative' procedures are compared with a quantitative analytical program. Using a laboratory-based wavelength-dispersive x-ray fluorescence (WD-XRF) spectrometer and a portable energy-dispersive x-ray fluorescence (ED-XRF) spectrometer, 28 different nickel-base alloy Certified Reference Materials (CRMs) were analyzed. Line interferences and inaccurate matrix correction are reasons for deviations from the reference value. As the comparison shows, 'semi-quantitative' analyses on the WD-XRF spectrometer can be accepted as quantitative determinations. The investigations show that the results obtained with the portable ED-XRF spectrometer do not meet the quality requirements of laboratory analysis, but they are good enough for field investigations.

Atomic and Nuclear Analytical Methods

“…Several hundred alloy compositions exist and the most important alloying elements include Ti, V, Cr, Mn, Fe, Ni, Cu, and Zn which are virtually adjacent in atomic number and several of these elements may occur together in a single alloy. The comparison of results obtained from measurements on nickel-base alloys using the laboratory-based WD-XRF vs. portable ED-XRF spectrometer has been made by Zwicky and Lienemann (2004). The analyses of steels (Mn, Cr, and V in carbon and low alloy steels), brasses (Mn, Ni, and Fe in brasses) and aluminium alloys (Si, Cr, Mn, Fe, Ni, Cu, Zn) have been successfully carried out and reported by Clayton et al (1973).…”

Application of Kβ/Kα in selecting calibration standards for X‐ray fluorescence analysis

“…For most XRF analysis software, however, these parameters are assumed to be the same for different types of samples . Evidently, the difference of these parameters between reference samples and unknown samples can result in large measurement errors …”

“…[15][16][17] Evidently, the difference of these parameters between reference samples and unknown samples can result in large measurement errors. [18] This paper has proposed a method to correct the mineralogical effects for measurement of geological samples by selecting suitable calibration samples from commonly used certified reference materials (CRMs). In traditional mode, calibration curves used for quantitative analysis are established by regression with multiple reference samples, and all of the analytes are measured using the same group of reference samples.…”