Is Scanning Electron Microscopy/Energy Dispersive <scp>X</scp>‐ray Spectrometry (<scp>SEM</scp>/<scp>EDS</scp>) Quantitative?

TLDR

SEM/EDS can identify and quantify all elements except H, He, and Li, but its impact is limited by misidentification of peaks, systematic errors from standardless analysis, and geometric factors that can cause up to ten‑fold errors. By following the k‑ratio measurement protocol developed for electron‑excited WDS, SEM/EDS can achieve WDS‑level accuracy and precision at lower electron dose, even with severe peak overlaps, provided sufficient counts are recorded. High‑throughput silicon drift detector EDS now makes this performance more practical. SCANNING 35:141‑168, 2012 Wiley Periodicals, Inc.

Abstract

Summary Scanning electron microscopy/energy dispersive X ‐ray spectrometry ( SEM / EDS ) is a widely applied elemental microanalysis method capable of identifying and quantifying all elements in the periodic table except H, He, and Li. By following the “ k ‐ratio” (unknown/standard) measurement protocol development for electron‐excited wavelength dispersive spectrometry ( WDS ), SEM / EDS can achieve accuracy and precision equivalent to WDS and at substantially lower electron dose, even when severe X ‐ray peak overlaps occur, provided sufficient counts are recorded. Achieving this level of performance is now much more practical with the advent of the high‐throughput silicon drift detector energy dispersive X ‐ray spectrometer ( SDD ‐ EDS ). However, three measurement issues continue to diminish the impact of SEM / EDS : (1) In the qualitative analysis (i.e., element identification) that must precede quantitative analysis, at least some current and many legacy software systems are vulnerable to occasional misidentification of major constituent peaks, with the frequency of misidentifications rising significantly for minor and trace constituents. (2) The use of standardless analysis, which is subject to much broader systematic errors, leads to quantitative results that, while useful, do not have sufficient accuracy to solve critical problems, e.g. determining the formula of a compound. (3) EDS spectrometers have such a large volume of acceptance that apparently credible spectra can be obtained from specimens with complex topography that introduce uncontrolled geometric factors that modify X ‐ray generation and propagation, resulting in very large systematic errors, often a factor of ten or more. SCANNING 35: 141‐168, 2013. Published 2012 Wiley Periodicals, Inc.