Hi!
I’m trying to simulate the gamma production due to neutron activation in copper, but I’m a bit confused by the results; in the attached files you can see the energy spectrum of the gamma produced in the target via nCapture and neutronInelastic processes.
I am using the Shielding Physics List, in particular the HP cross section for low energy neutrons
-
As expected, there are some peaks and then a continuous spectrum (because neutrons are not only thermal, they could have more energy, so there are gamma’s produced via inelastic collisions as well). The two peaks are around 7 and 8 MeV, which correspond to the neutron capture of Cu65 and Cu63, respectively. However, from here, it seems that there should not be a single peak for each isotope, but several gamma lines. In particular, it seems that it is more likely for Cu65 to emit several low-energy gamma’s than a single one at ~7 MeV, which should lower significantly that peak: indeed, the cross sections reported for the ~7 MeV emission is only 0.0132 barns, while the one for the ~8 MeV peak for Cu63 is 0.869 barns.
Are those features not implemented in Geant4, or am I doing something wrong? Am I missing something?
-
What causes the step-like behavior of the continuous part of the spectrum (it is much more evident in the log plot, however it is possible to see it in the linear plot as well). Is it a physical effect, or just an artifact of the simulations, for example due to the binning of the neutron cross section (which, from what I understand, at those energies is taken from a database, so I’m guessing it will have some kind of binning…)
Thank you very much for your help!
CopperLin.pdf (50.2 KB)
CopperLog.pdf (48.5 KB)
Hi, I had a problem with blocky-looking spectra, much like your graphs, doing photoneutron simulations about four years ago Photoneutron process is inaccurate at low gamma ray energy . I was using the ShieldingLEND physics list on Geant4 10.5.1. It was suggested that it might be a bug or peculiarity in LEND. I do not know if it is still a problem under 11.1. I usually use QGSP_BERT_HP or QGSP_BIC_HP for (n,gamma) simulations, so you might want to try one of those.
As for gamma ray line energies and relative intensities, it depends on your geometry (target size and composition, thickness, incident neutron energy, etc.) More information is required to see if it makes sense.
Thanks for the reply!
Just out of curiosity, which materials you are considering in your (n,gamma) simulations?
I’m using Shielding (not Shielding_LEND), which if I remember correctly uses QGSP_BERT_HP or QGSP_BIC_HP (the relevant part for the simulations is that it uses the low-energy neutron cross sections from database). I also tried using directly QGSP_BERT_HP (with a bunch of other stuff), or the Physic List from the Activation example (in extended/radioactivedecay), but I had the same problems…
I run some more simulations and studied a bit more the issue.
First, I tried considering thermal neutrons (0.025 eV) on a solid target (cylinder, 45 cm thick), in this way there should not be any gamma’s created via inelastic scattering. I considered different materials (Cu, Pb and Co) and recorded all the gamma’s created via neutron capture, you can see the plots in the attached file, but the results do not make any sense to me.
- There is still a continuous spectrum in the gamma created via neutron capture, and I cannot understand where it comes from, nor how it could be physical (there should be only gamma lines, not a continuous spectrum). For the record, the energy I recorded is the one when the gamma’s was created, using the function
G4double gammaKinEn=aTrack->GetVertexKineticEnergy();
which gives the kinetic energy when the gamma was created, so it cannot be due to gamma’s losing energy in the target.
- It seems that all the energy released in the process is carried out by a single gamma, but that’s in contradiction with the data reported here.
For example, if a thermal neutron is captured by a Cu65 atom, there are 7.06572 MeV of energy released. From the results of my simulations, it seems that there is a single 7.066 gamma emitted (there is another peak at ~7.9 MeV, which is related to the neutron capture on Cu63), then there is the step-like continuum spectrum, then a bunch of peaks at low energies.
However according to the database there are a lot of other peaks at intermediate energies, corresponding to two or more gamma’s emitted. For example, it is 7-8 times more likely that, instead of the single 7.066 gamma, there are two gamma’s, one at ~6.6 MeV and one at ~0.4 MeV, these peaks are not present at all.
I tried the same simulation with Pb, here the results are much better (you can see the various peaks), however with Co I had the same problems (only one peak at maximum energy, and then a continuo spectrum. And it is one of the material used in the Activation example! These results were confirmed also by running that code, by the way…)
In your simulations, were you able to see multiple peaks? Which materials did you used?
Activation-Thermal.pdf (131.5 KB)
I have used many target materials in my neutron capture gamma ray detection simulations. Some (Cl-35 Ar-36, Ar-40, Li-6, Li-7, Be, C, N-14, Al-27, B-10, Ni-60) accurately agree with the NNDS thermal capture data base in both relative intensity and energy of the gamma ray lines. Others match the energy well but the intensity not so well (Ni-58, Gd-155, Gd-157). Others are poor matches for both (Ni-58, Li-6, Li-7). The level of agreement depends on the setting of the HPManager parameter settings DoNotAdjustFinalState, UseOnlyPhotoEvaporation.
I noticed that the Geant4 process generates non physical gamma rays to conserve energy in individual reactions (it is also mentioned in the Application Developer’s Manual). This yields lines in the spectra that do not correspond to those in the neutron capture data bases (such as NNDC). There is also a continuum component to the spectra, I presume due to how the process software generates the gamma rays in the exit channel.