When I started to test XRF, I noticed the echo (or phantom) image of background Am241 peaks (59.5KeV and 29KeV), appearing at lower energy and often around 10KeV below.
Following the suggestion of people who replied to my post in Yahoo group, I studied about Compton scattering and the shift.
- Some links
http://www.science.mcmaster.ca/medphys/images/files/courses/4R06/4R6Notes5_Photon_Spectrometry.pdf
http://www.iup.uni-bremen.de/PEP_master_thesis/thesis_2007/thesis_chehade.pdf
http://www.qsl.net/k/k0ff/Compton%20Scattering/Compton%20Scattering%20Experiment%20by%20Prutchi.pdf
When gamma ray hits something and change the course, it looses some energy depending on the angle of the course change.
http://hyperphysics.phy-astr.gsu.edu/Hbase/quantum/compdat.html
When I put 59.5KeV and 180 degree, the page gives 11.2389KeV as the resulting shift.
http://hyperphysics.phy-astr.gsu.edu/Hbase/quantum/compton.html
And resulting ghost peak (which corresponds to back scatter peak) would be around 48.26KeV.
So, it seems to match the result I'm getting.
However, for 29KeV peak, ghost peak should be around 26KeV...
And what I'm seeing is the same amount of shift (about 10KeV) as the 59.5KeV peak instead of mere 3KeV.
(3 spectrums are that of small water bottle, big water bottle and Maple Sugar)
Now, I have to study and think again for the lower ghost peak that appears around 17 to 18KeV.
If it's not the peak shifted from 29KeV, it can be the peak of slightly higher energy (1 to 2 KeV for 180 degree Compton scattering).
I listed up possible ingredients for the 17KeV peak in Am Spectrums.
17KeV peak
- U XRF 16.4KeV
- Pa XRF 16.7KeV
- U XRF 17.2KeV
- Am XRF 17.6KeV
- Np237 17.74KeV
- Am XRF 18.8KeV
If it appears at 17KeV, maybe the original peak is a bit higher (again, due to Compton scattering), around 18.6KeV or even slightly higher.
This peak (unlike main 59.5KeV and 29KeV peaks) is easily blocked by thin metal like the cover around smoke detector button.
Even thin Aluminum foil would absorb this wave, but thicker and heavier the higher peak we get.
- Compton edge and back scatter calculation <== I don't think this is exact...
Back scatter = E x 0.28 , Compton edge = E x ( 1 - 0.28) = E x 0.72