I did XRD of steel slag powder with copper source. The sample contains iron phase. I was getting a very elevated XRD peak data with very high noise such that the peaks are not distinguishable. The step size is 0.01 degree, and the sample run-time was 30min. The sample size was less than 180 microns as I had sieved out using 180micron mesh. I am skeptical about the fluorescence effect but how can I still manage out to collect XRD data using copper anode only as I do not have accessibility of XRD device with cobalt anode.
Hi Rajeev,
You can reduce fluorescence by trying the following, if you have them:
Use Ni foil beta filter.
Use Monochromators.
Re-define the energy window for the detected pulses on the detector (pulse height distribution, PHD)
You may find the attached file useful.
Hope this helps.
Dear Rajeev Rajepandhare ,
in addition to the answer of Ameer K Ibraheem I would like to add, that a single Ni filter, which is mainly used to reduce diffraction peaks caused by the Cu K-beta line, will only reduce the Fe K-fluorescence background radiation a bit.
But two filters will do the job, which your are looking for...
The key here is to use a set of balanced filters, which are often called 'Ross-filters' (for details please 'google' for).
You have to perform two XRD scans;
a) one of them, for example called XRDNi, is taken by additing a Ni filter of special thickness in front of the detector, and
b) a second XRD pattern (called XRDCo) by putting a Co filter of special thickness in front of the detector.
In case of the Ni filter (a) you will mainly acquire the Fe K-fluorescence including the Cu K-alpha based diffraction peaks, but the Cu K-beta peaks are strongly reduced, as well as the Compton scatter background due to the bremsspectrum of the tube.
In case of the Co filter (b) you will mainly acquire the Fe K-fluorescence, but the Cu K-alpha and Cu K-beta based diffraction peaks will be strongly reduced, as well as the Compton scatter background due to the bremsspectrum of the tube.
The only thing now you have to do is to perform a subtraction:
XRDresult = XRDNi -XRDCo to get your XRD pattern with strongly reduced Fe K-fluorescence background.
The thicknesses of both filters have to be 'balanced' in such a way, that the energy dependencies of the transmissions of both filters are almost equal in the energy range below the Co K-edge energy and above the Ni K-edge energy.
A suggestion for 'balanced' filter thicknesses (commercially available) as well as their transmissions is given in the attachment.
Here you see that the difference of your filtered XRD pattern will result in a quite 'monochromatic' XRD pattern around Cu K-alpha...
However, I will not hide a drawback: unfortunately the noise of the XRDresult pattern will be increased compared to the noise of the unfiltered pattern.
Good luck and
best regards
G.M.
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films