Dear researchers,
I would appreciate if you could guide me:
I have filtered some river water samples to characterize the trapped particles on the filters by EDS. The EDS analysis shows Ca, Mg, K etc peaks together with C, N and O ones. As the particles can be composed of some organic and inorganic materials, then
should I remove C and N peaks from our EDS software? When I keep the C peak, the Ca wt% is about 10% but when I remove the C peak then the Ca wt% increases to 25 wt%.
or I keep C and N but not report them?
FYI: (1) I know EDS is not a good technique to characterize the light elements such as C and N. (2) some particles are small and therefore the beam hits both particles and the membrane but some are not.
Regards,
I would use element ratios, Ca/Mg etc related to the count rate. The absolute valueas are anyway more speculative because of exinsting/non-existing C and N.
Thank you for your reply. But the question is should I remove C and N from EDS software before extracting the data? because removing them increases the wt% of other elements.
Yes, but the ratios are not affected whether you keep both inside or not... in other words, I would keep everything inside since you do not know what the software actually does.
What probably happens is is related to the spectral lines of the elements. You want to only see the main spectral lines, but due to fore example power of the electron gun you get coincidental lines also. If you remove C, N and O from the results and the machine interprets the signals to be coming only from Ca you will get higher results than it actually is. Unless you can get to the raw data and see what happens in different matrices I think Gert Nolze's suggestion to keep everything in is much more safer bet than than excluding them.
Yes, you need to ensure good spectral fits for which you need to identify and quantify all x-ray lines; after quantification by your software you can then exclude those elements you know are either not in your sample, due to surface contamination or are detection artefacts. If e.g. C and N make up 50at% of an original quantitative analysis result form your software but you omit them, then the relative contributions from all other elements will double.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science