Electron-excited X-ray microanalysis with energy-dispersive spectrometry (EDS) proceeds through the application of the software that extracts characteristic X-ray intensities and performs corrections for the physics of electron and X-ray interactions with matter to achieve quantitative elemental analysis. NIST DTSA-II is an open-access, fully documented, and freely available comprehensive software platform for EDS quantification, measurement optimization, and spectrum simulation. Spectrum simulation with DTSA-II enables the prediction of the EDS spectrum from any target composition for a specified electron dose and for the solid angle and window parameters of the EDS spectrometer. Comparing the absolute intensities for measured and simulated spectra reveals correspondence within ±25% relative to K-shell and L-shell characteristic X-ray peaks in the range of 1–11 keV. The predicted M-shell intensity exceeds the measured value by a factor of 1.4–2.2 in the range 1–3 keV. The X-ray continuum (bremsstrahlung) generally agrees within ±10% over the range of 1–10 keV. Simulated EDS spectra are useful for developing an analytical strategy for challenging problems such as estimating trace detection levels.

Electron beam X-ray microanalysis with semiconductor energy-dispersive spectrometry (EDS) performed with standards and calculated matrix corrections can yield quantitative results with a distribution such that 95% of analyses fall within ±5% relative for major and minor constituents. Standardless methods substitute calculations for the standard intensities, based either on physical models of X-ray generation and propagation (first principles) or on mathematical fits to remotely measured standards (fitted standards). Error distributions have been measured for three different standardless analysis procedures with a suite of microanalysis standards including metal alloys, glasses, minerals, ceramics, and stoichiometric compounds. For the first-principles standardless procedure, the error distribution placed 95% of analyses within ±50% relative, whereas for two commercial fitted standards procedures, the error distributions placed 95% of analyses within ±25% relative. The implication of these error distributions for the accuracy of analytical results is considered, and recommendations for the use of standardless analysis are given.