Missing n,n+a for 14.1 MeV neutrons in Silicon

I have a beam of 5E9 14.1 MeV neutrons on 25 um thick silicon, with an interest in the high-energy tail of the inelastic energy deposition. As a result, I’ve been combing through secondary particle creation and energy deposition.

I have found most of what I would expect, but do not see any n(Si28,Mg24) n + alpha reactions. I’ve double-checked Jeff 3.3 ENDF data to confirm the reaction should be happening at this energy, and poking around the GNDL1.4 data it looks like that reaction channel is available.

I have copied the physics from Hadr04.

I’m trying to determine:

  1. If there’s some reason I’ve missed why I shouldn’t see energy deposition from the resulting Mg24 like the knock off ions from other reactions, or see Mg24 in the list of secondaries?
  2. Is there a way to force/bias specific inelastic channels, or check probability of specific inelastic channels during a run?

If there’s any other info on my sim that would help, I’m happy to provide it. Thanks in advance!

Geant4-v11.2.0
Ubuntu 22.04
g++ 11.4.0
CMake 3.22.1


How many (n,na) reactions do you expect for 5e9 incident 14.1MeV neutrons? I would guess the number is small given that you have a very thin 25um foil.

From the EXFOR data base, the (n,na) reaction cross section is ~6mb. Other energetically allowed reactions at that energy have much higher cross sections: (n,p) 290mb, (n,a) 900mb, (n,np) 27mb, (n,el) 600mb. This means that only ~1/300 reactions will be an (n,na). So, for example, how many (n,a) reactions are you seeing? That number should be signifcantly greater than 300, if you expect to see one or more (n,na) products.

To somewhat answer my own question, I did a back of the envelope (literally) calculation for 5e9 14.1MeV neutrons incident on the face of a 25um Si wafer assuming that all of the neutron beam hits all of the wafer. Assuming the 6mb cross section, I estimate there should be ~372 (n,na) reactions. Do you see anywhere near that many 24Mg atoms?

I am seeing no Mg24 created at all. I’m looking at the list of secondaries created at each step in my stepping action; if there’s a better place to monitor, let me know. In particular, if there’s a way to check for specific inelastic channels explicitly that would be helpful.

I am, for comparison, seeing >90k Mg25 created. I’ll change my seed and run 20E9 overnight to be sure I’m not just getting particularly unlucky and report back, but not seeing any Mg24 seems suspect to me when I should be averaging a few hundred.

I ran Hadr03 from the 11.2 distribution with 100,000,000 14.1 MeV neutrons incident on a 1000m block of 28Si. It only scores first interactions and hadron elastic scattering is turned off. I chose QGSP_BIC_HP as the registered hadron physics. I see lots of reactions, but there are none involving 24Mg as a product. From the output:

List of nuclear reactions: 

                 neutron + Si28 --> N gamma or e- + He3 + Mg26:     698   Q = -2.9565 MeV
                       neutron + Si28 --> N gamma or e- + Si29:   60765   Q =  8.4736 MeV
               neutron + Si28 --> N gamma or e- + alpha + Mg25: 17749095   Q = -2.9226 MeV
            neutron + Si28 --> N gamma or e- + deuteron + Al27: 1301922   Q = -5.3188 MeV
             neutron + Si28 --> N gamma or e- + neutron + Si28: 44613104   Q = -1.2316 MeV
              neutron + Si28 --> N gamma or e- + proton + Al28: 24632479   Q = -3.9417 MeV
              neutron + Si28 --> N gamma or e- + triton + Al26:    2390   Q =  -7.108 MeV
                               neutron + Si28 --> alpha + Mg25: 1262392   Q = -2.6449 MeV
                            neutron + Si28 --> neutron + alpha: 1324412   Q = -10.776 MeV
                              neutron + Si28 --> proton + Al28:  738409   Q = -3.8549 MeV
                           neutron + Si28 --> proton + neutron: 8314334   Q = -12.205 MeV

                          number of gamma or e- (ic): N = 1 --> 7 

So it would suggest that the 28Si(n,n’a)24Mg reaction is not included when using the HP neutron physics.
I’ll try it with the FTFP_BERT_HP (I suspect result will be the same) and then without HP to see if there’s any change.

In the mean time, I did the same exercice, with FTFP_BERT_HP ! Of course, similar results.
In fact, the reaction (n, n’a) is simulated, but wrongly : Mg24 is not created. Third line from the end of the list.
I do not know the origin of this behavior. Better to open a bug report or to ask directly the maintainers of ParticleHP package : Alberto Ribon, Dennis Wright, Vladimir Ivantchenko.

joel.mac.txt (313 Bytes)
joel.out.txt (3.2 KB)

Thank you Michel. I noticed the third line from the bottom, but could not figure out what reaction it was supposed to be because there was no residual nucleus like most of the other reactions show. The last reaction shown (–> proton +neutron) is similarly confusing.

And there are no 28Si(n,n’a)24Mg reactions using Hadr03 with QGSP_BIC.

hadronic conservation : Al 27

nor FTFP_BERT, QBBC, INCLXX …

hadronic conservation : Al 27

I figured out the reaction without problem. My confusion is that they include the residual nucleus in most of the reactions, but not in some (3rd from last, last).

Unrelated to that, I can volunteer to file a bug report. But should I contact the maintainers first?

Thank you for this, and for pointing me to Hadr03 as a way to look at this! I’d certainly appreciate you filing the bug report.

thanks you in advance for opening a bug report

I have created Bugzilla problem 2591 ‘Hadronic physics does not include 28Si(n,n alpha)24Mg reaction for 14.1 meV neutrons’. I will keep you posted as to the outcome.

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.