One of the main goals of automated scanning electron microscopy analysis (ASEM) of particles is to provide macroscopic phase and compositional information on particle populations based on the energy dispersive x-ray spectrometry, EDS, analysis of several hundred to several thousand individual particles from the sample. The selection and identification of the various elemental groupings is often accomplished by applying techniques for grouping data, such as multivariate or cluster analysis methods, to characteristic x-ray intensities or normalized elemental concentrations. Particle groupings are often based on major and minor elements with concentrations greater than about 1-2 wt. percent. At these concentration levels, the peak-to-background ratios for the characteristic x-ray peaks are sufficiently large that in the absence of severe peak overlaps the elements can be easily identified even in spectra with poor counting statistics. Additional refinement in particle groupings may be possible if reliable information can be obtained on the trace elements present in particles at less than about 1-2 wt.%.