“Standardless”, quantitative electron probe x-ray microanalysis with energy-dispersive spectrometry: What is the distribution of errors?

Abstract

Conferences on quantitative electron probe X-ray microanalysis are typically highlighted by presentations which seek to advance the state of analytical science in this field through improvements in knowledge of the fundamentals of electron/X-ray-specimen interactions, inter-element (matrix) corrections, corrections for irregular specimen geometry, etc. However, in light of recent developments in computer-assisted, energy dispersive X-ray spectrometry analytical systems, we need to re-evaluate what is actually taking place in the everyday application of this analytical technique. The central paradigm in quantitative electron probe X-ray microanalysis, as established in the earliest work of Castaing and built upon by generations of contributors, is the measurement of the ratio of the intensity of an element in the unknown to the intensity for the same element in a standard under constant conditions of beam energy, known dose, and spectrometer performance. The set of such “k-values” determined for all constituents in the unknown is then converted into a set of concentrations through the use of matrix correction factors calculated through one of several schemes, e.g., “ZAF”, ϕ(ρz), etc. The great power of electron-excited X-ray microanalysis derives from the simplicity of the standards required.