What is simulated by the FritiofCaptureAtRest for antiprotons?

Physics Processes, Models and Cross Sections
1

Geant4 Version: 11.2
Operating System: WSL Ubuntu 22.04.5 LTS
Compiler/Version:
CMake Version:

Dear Geant4 Community,

I have a simulation where I shoot slow (1.005 keV) antiprotons into different materials, and then I look for the “CaptureAtRest” process from which I store information about the secondaries produced. I compared different physics lists (FTFP_BERT_HP, QGSP_BIC_HP, QGSP_BERT_HP, FTFP_INCLXX_HP, QGSP_INCLXX_HP) to see the difference between implemented models. To my knowledge, the antiproton stopping was handled solely by hFritiofCaptureAtRest in Geant4 v.10.0-v11.1. In Geant4 v11.2, the INCL++ model was also updated to include hINCLXXCaptureAtRest for antiprotons. When I check the “CaptureAtRest” process name, I indeed see only two: hFritiofCaptureAtRest and hINCLXXCaptureAtRest.

According to the “processes//hadronic/stopping/src/G4HadronicAbsorptionFritiof.cc”, the hFritiofCaptureAtRest doesn’t include any intranuclear cascade, and it goes directly to Preco.

G4HadronicAbsorptionFritiof::
G4HadronicAbsorptionFritiof( G4ParticleDefinition* pdef )
  : G4HadronStoppingProcess( "hFritiofCaptureAtRest" ), 
    pdefApplicable( pdef ) {
  
  G4TheoFSGenerator * theModel = new G4TheoFSGenerator( "FTFP" );
  G4FTFModel * theStringModel = new G4FTFModel;
  theStringModel->SetFragmentationModel(new G4ExcitedStringDecay());

  // Not a cascade - goes straight to Preco
  G4HadronicInteraction* p =
    G4HadronicInteractionRegistry::Instance()->FindModel("PRECO");

Therefore, I would expect the distributions to be identical across all physics lists in versions before v11.2, and to be split into two types (one for FTFP and one for INCLXX-based physics lists) in v11.2. Below are the distributions of atomic mass numbers of nuclei produced in the “CaptureAtRest” process for different physics lists in Geant4 versions v11.1 and v11.2:

geant4_v11.1_vs_v11.2_A-disribution
geant4_v11.1_vs_v11.2_A-disribution4400×1767 878 KB

I ran the simulation for each physics list and Geant4 version with the same initial random number seed. From the distributions I showed, I would conclude that the cascade model has some effect on the hFritiofCaptureAtRest process, even if it isn’t explicitly called.

I would like to understand how the nucleus is modelled in the hFritiofCaptureAtRest process. Is the model based on the included INUCL model? In the documentation for The Geant4 Bertini Intranuclear Cascade Model there is:

The target nucleus is modelled by up to six concentric shells of constant density as an approximation to the continuously changing density distribution of nuclear matter within nuclei.

Is the same approach applied in the hFritiofCaptureAtRest process?

2

Hello,

yes, you are right in most of your sentences.

There are essentially 2 models in Geant4 to deal with antiproton at rest:
the hadronic string model Fritiof (FTF), coupled directly to Precompound/de-excitation (i.e. without any intranuclear cascade model), and the intranuclear cascasde model INCLXX (which is coupled directly to de-excitation, without precompound).

Therefore, as you said, for anti-proton annihilation at rest in thin-target, there should be only 2 family of results, with no effect of intranuclear cascade models.

However, if the target is not thin, i.e. the secondaries produced by the antiproton annihilation at rest can then have hadronic inelastic interactions in other target nuclei, then differences among the various physics lists can arise, due to different intranuclear cascades - e.g. between FTFP_BERT
and QGSP_BIC.

The plots are too busy for me, and I am not able to see statistically significant differences where they should not be.

An exception of what said above is FTF_BIC and QGS_BIC (please notice the absence of “P” is the names of these physics lists): in these two cases only, the antiproton annihilation at rest is handled by Fritiof (FTF), but coupled with the Binary Cascade intranuclear cascade (BIC) - which in turn makes use of Precompound and de-excitation.

In principle, the INCLXX treatment of antiproton at rest should be the most precise available in Geant4, followed by FTF_BIC / QGS_BIC, and then, the “default one”, i.e. FTF + Precompound/de-excitation. In practice, more physics validation is needed in order to draw firm conclusions…

Please note that the Bertini intranuclear cascade model (BERT) has nothing to do with the treatment of antiproton annihilation - in any physics list.