On 13.02.2020 07:33, PyMca general purpose mailing list. wrote:
> I´m refering to Bragg diffraction peaks that appear in an XRF
> spectrum when analyzing materials with a crystalline structure. While
> these peaks can generally be avoided by rotating the sample to change
> the crystal orientation, this can be difficult when irradiating some
> materials with multiple crystal orientations.
> These peaks have been a problem for us when analyzing mineral samples.
>
If the peaks are relatively separated you can try to put an anchor at
their maximum height, but I do not expect any miracle there. Other
workaround could be to add those energies as part of the excitation
beam, provide a ridiculous weight to them so that they do not influence
the fit and to set them as scatter peaks. For the later you have to save
the BEAM profile to a file, to edit the file and to load it again.
Again, I do not expect it to be a generic solution.
Your problem falls in the generic issue of allowing the user to provide
arbitrary peaks and peak shapes (diffraction peaks, Raman peaks,...).
PyMca does not offer that possibility (yet?).
A solution that it would not be too difficult to implement is to ask the
fitting routine to simply ignore the regions where the peaks/artifacts
are. Unless all the peaks of an element (including their escape peaks)
fall on those regions, the XRF results should be correct. This solution
is incompatible with the pile-up algorithm implemented in PyMca. Given
my workload it is the only approach I can imagine to implement within
relatively short time.
Best regards,
Armando
|
