Electron microprobe analysis (EMPA) (Stelios Aspiotis)

Stylianos Aspiotis

The principle of EMPA is based on the bombardment of the sample surface with a high-energy beam of accelerated electrons. Upon electron beam irradiation, the inner-shell electrons of a given element can be excited from their ground level to an unoccupied valence level, creating at the same time a vacancy. This vacancy at the core-electron level is energetically preferred for all outer-shell electrons and therefore it is filled by an electron from a higher electron level. As a result of this electron transition, a characteristic set, for each chemical element, of X-ray radiation is emitted, whose analysis can provide qualitative and quantitative compositional information of the analysed sample volume. The major difference between the main forms of EMPA, namely wavelength-dispersive and energy-dispersive (ED) methods, is that in the former one, scans are performed at wavelengths where the specific X-ray radiation is expected for the probed element, whereas in the latter one all wavelengths are scanned at the same time. Thus, WD-EMPA is more time-consuming than the ED counterpart, but has a higher spatial resolution and sensitivity.

In particular, ED-EMP, also referred as energy-dispersive spectrometry (EDS), analyses the emitted energy of the X-ray radiation (inversely proportional to the wavelengths of the electromagnetic radiation; X-rays), which is characteristic of the participating atoms in the studied sample. EDS can preferably be applied to samples whose chemical composition is entirely unknown, for which qualitative and semi-quantitative results are needed. Sample surface conditions affect the detection limit of this technique, as a smooth surface can lower the detection limit of minor elements down to 0.1 weight percent (wt %). Although EDS analysis is faster compared to WDS, overlapping peaks in the resulting spectrum may hinder the separation of light elements (Z < 10), transition metals (groups 3-12 in the periodic table) and rare earth elements (REE).

On the contrary, wavelength-dispersive EMPA (WD-EMPA), also referred as wavelength-dispersive spectrometry (WDS), is the ultimate analytical tool to determine the chemical composition of small volumes (in this case done to 5 μm3) of solid materials, both organic and inorganic, with a concentration of a few hundreds ppm. WD-EMPA reads only the X-rays of a single wavelength, hence producing very sharp peaks of the analysed elements and not a broad spectrum of wavelengths or energies compared to EDS. WD-EMPA can be classified as a destructive analytical method, as thick sections need to be prepared from the investigated sample, which requires cutting, polishing, embedding and carbon coating prior to the analysis. At the same time, for small dimensioned samples (diameter smaller than 3 cm) with an already smooth and even surface, WD-EMPA is relatively non-destructive for the analysed surface and the requirements for sample preparation are minimal.

Back-scattered electron (BSE) images are one of the most important functions of EMPA. The detection of the result between the elastic interaction of accelerated electrons (electron beam) and the atoms of the studied sample is what creates the BSE images. BSE are sensitive to the atomic number (Z) of the analysed elements, as typically heavier elements, i.e. bigger nuclei with a higher probability of an elastic collision, can deflect stronger the incident electron beam than light elements, that are characterised by smaller nuclei and a lower probability of an elastic collision. Thus, heavier elements can produce a higher number of back-scattered electrons that will reach the BSE detector, being as a result proportional to the analysed element. With the usage of BSE images, high-resolution compositional maps of a specimen can be acquired (Element Mapping) assisting in that way in the separation of different phases in a polyphasic material.

Cameca SX100 electron microprobe (EMP) system will be available until the ennd of 2025 (End of UWA I), since from 2026 the Department of Earth System Sciences is not goin to further financially support any further annual maintenance, due to budget cuts.