Simple approach to the analysis of oxides, silicates and carbonates using x‐ray fluorescence spectrometry

Abstract

AbstractX‐ray fluorescence spectrometry has always been considered to be a comparative method of analysis requiring chemically analysed samples or reference materials for calibration. The fusion technique, which is widely used as the method of sample preparation to eliminate particle size and mineralogical effects, allows the use of synthetic ‘standards’ as the means of calibration, but most laboratories use large numbers to establish the primary calibration line and calculate empirical interelement correction coefficients. In the 1970s, de Jongh proposed the use of theoretically based correction coefficients, and his concept of ‘apparent concentration,’ which is related linearly to the measured intensity, means that in principle only two samples are necessary to define the primary calibration line. This paper describes the use of this principle to calibrate a spectrometer for the 13 most commonly occurring oxides, plus six others, over wide composition ranges. Each calibration line requires one sample, which is made from a commercially available source of the element, a blank or zero sample, which is used for all calibration lines, and a set of theoretical α coefficients calculated using de Jongh's program. Sample preparation is by fusion in Li2B4O7. The calibration was tested using reference materials and was found to be accurate for every element except sulphur.