Instrumental Factors and Figure of Merit in the Detection of Low Concentrations by X-Ray Spectrochemical Analysis

Abstract

Detection of low concentrations in x-ray spectrochemical analysis depends upon the degree of correlation between concentration and spectral line intensity. Various factors causing deviations of points from a calibration curve are mentioned with particular emphasis on deviations due to inadequate spectral resolution and to x-ray counting statistics, both of which are affected by various instrumental parameters. The effects of various choices of instrumental parameters on the counting statistical deviations may be measured in terms of a figure of merit z=A/√B, where A is the slope of the calibration curve (intensity vs. concentration) and B is the background or intensity at zero concentration. The statistical deviations are inversely proportional to z. Values of z obtained in the analysis of some of the constituents in the NBS low alloy steel standards for several different combinations of instrumental parameters and experimental conditions are presented. It is shown that the use of higher resolution optics or greater dispersion crystals does not always reduce z drastically and in some cases actually increases z The effects on B and hence on z, of spectral purity of the exciting x-ray tube, of pulse height analysis techniques, and of monochromators having no second order reflection, are also examined and are shown to depend upon the spectral nature of B, which in turn depends upon the type of specimen under analysis.