Free Access
Microsc. Microanal. Microstruct.
Volume 7, Number 4, August 1996
Page(s) 265 - 277
Microsc. Microanal. Microstruct. 7, 265-277 (1996)
DOI: 10.1051/mmm:1996121

Quantitative Microanalysis Using Electron Energy-Loss Spectrometry: II. Compounds with Heavier Elements

Ferdinand Hofer et Gerald Kothleitner

Forschungsinstitut für Elektronenmikroskopie, Technische Universität Graz, Graz, 8010, Austria

Electron energy-loss spectrometry (EELS) in the TEM can be used for the quantitative analysis of compounds including both light and heavy elements at a submicrometre scale. However, EELS-quantification can become complicated due to low edge-to-background ratios, problems with multiple scattering in case of thicker samples, inaccuracies due to background extrapolation or edge-overlapping. All these problems can be overcome by careful use of well known procedures. A further problem of quantitative EELS is the need for precise partial ionization cross-sections that are sometimes not well known. In this work, we have quantified EELS-spectra of compounds with known composition: $\rm LaCoO_3$, $\rm YBa_2Cu_3O_7$ and $\rm
LaB_6$. To demonstrate the achievable accuracy with different models, the quantifications have been calculated with theoretical cross-sections (hydrogenic model and Hartree-Slater model) and experimentally determined k-factors. In every single case more accurate quantification results could be obtained when experimentally derived cross-sections (or k-factors) were used, giving concentration values that lie within 5 rel% of the nominal composition.

7920K - Other electron surface impact phenomena.
0780 - Electron and ion microscopes and techniques.
8280P - Electron spectroscopy for chemical analysis photoelectron, Auger spectroscopy, etc..

Key words
electron energy loss spectra -- electron probes -- electron energy loss spectrometry -- TEM -- quantitative analysis -- low edge to background ratios -- background extrapolation -- edge overlapping -- precise partial ionization cross sections -- hydrogenic model -- Hartree Slater model -- k factors

© EDP Sciences 1996