Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  • Any charging present on the surface of the sample will quickly degrade the quality of EDS spectra. In order to avoid charging: 
    • Use the lowest possible accelerating voltage that is appropriate for the material and elements of interest
    • Use an even smaller beam current
    • Ensure good electrical grounding using carbon tape,
    • Perhaps remove the sample to apply a thicker C coat to improve electron transport
    One proxy for quality of spectrum is to right click on the spectra and then select Logarithmic graph display, to make visible the background curve of the spectra. The background should taper smoothly to the highest values at the accelerating voltage used, e.g. if the microscope is being operated at 10 kV, the
    • Check to make sure that the sample isn't charging. You can often tell charging because there will be visual streaking on the image, however also you can look at a few things on the quantification and spectra. If you are seeing charging, let me know and I can give more specific results for how to try to remedy that.
      1. Check the mass% total, which should be around 100% for non-hydrous minerals (circled on attached image)
      2. Check the "Dwayne-Hunt limit" which is where the background on the spectra goes to 0. This should be around the energy of the accelerating voltage you are operating the microscope on, e.g. for 10 kV microscope setting, you should have a Dwayne-Hunt limit of 10 keV (circled on image)
      3. If there is charging, under Additional Settings, click A (circled on image) to make the software adjust the quantification down to the "effective high voltage"
  • One proxy for quality of spectrum is to right click on the spectra and then select Logarithmic graph display, to make visible the background curve of the spectra. The background should taper smoothly to the highest values at the accelerating voltage used, e.g. if the microscope is being operated at 10 kV, the spectra should have smooth values up to 10 kV.
  • The carbon peak is caused by sputtering: It can be disregarded in the Esprit software Sample Configurator menu. It will ignore the carbon peak in the quantification (aka it will know that the carbon is from the coating, not the sample). See page 27 of the Bruker Esprit Compact user manual
  • There are two quantification methods loaded. They can be changed by clicking the arrow on the Quantify button, and click Load spectra should have smooth values up to 10 kV.
  • Check the non-normalized mass % totals, which ideally should be around 100% for non-hydrous minerals.
  • In order to interpret mass percent oxides, it is helpful to have an idea of the anticipated mineralogy of the sample, then compare the compositions to the ideal stochiometric mineral compositions from an online mineral database such as Webminerals of Mindat.
  • Technicians can share access to a number of technical books on electron microanalysis if desired.

...