Compton Edge and Gamma Spectroscopy

[zeebo17]

Hi,

I have measured some spectrums from various gamma sources and I was wondering what kind of fitting I should do to determine the Compton edge. I have access to Logger Pro, Excel, and mathematica. I have fit the photopeak with a Gaussian so I know the uncertainty in the Compton edge, I just need to determine its value.

Thanks!

 

[Bob S]

Hi Zeebo-
This is a very good question, and there are two answers. One is called Compton edge, and the other is called the backscatter peak.

When a photon backscatters off a free electron, and leaves the detector (e..g, NaI(Tl)), only the energy of the Compton recoil electron remains in the detector (Compton edge). When a photon backscatters in the shielding around the detector, and the backscattered photon is absorbed by the detector, it creates a backscatter peak.

Let's solve this situation for a cesium-137 photon backscattering in or near the detector.

The basic equation for energy balance in Compton scattering is

E'/E₀ = [1+α(1-cos(θ)]^-1

where E' is scattered photon energy, E₀ = 661 keV, α= E₀/mc² = 661/511 = 1.29, and θ = scattering energy.

So for this case, θ = 180 degrees,

E'/E₀= [1+ 2.58]^-1 = 0.28

So the backscattered photon is 0.28*661 = 185 KeV, (backscatter peak) and

E_electron = 661 - 185 = 476 KeV (Compton edge)

Bob S

 

[zeebo17]

Sorry, I should have been clearer. I know the theoretical values for the various spectral features, but I was wondering how to determine my experimental result from the spectrum I have taken (Something more quantitative then eye-balling it).

I had seen some things that said the edge was half way between the maximum value (near the edge) and half of the height of this maximum. But I'm not sure how trustworthy this source is.

I'm also not sure how to find the Compton edges when there are two right next to each other.

Thanks for your help!

 

[Bob S]

You know the detector resolution from the width of the cesium gamma line (661 KeV). So you can do a convolution of the theoretical Compton edge shape (need to integrate over scattering angle) and the detector resolution.
Bob S