X ray Fluorescence angular distribution

[Nolane_Delille]

Hi,

I’m trying to simulate the electron cloud generated by x rays in HgCdTe as part of my PHD. While analyzing the angular distribution for the fluorescence X rays, I noticed that :

their angular distributions are the same than the one for the primary photoelectric electron distribution.

However, thoses x rays are supposed to be isotropic (at least we can suppose that they are , accoding to https://www.oecd-nea.org/upload/docs/application/pdf/2020-10/penelope-2018__a_code_system_for_monte_carlo_simulation_of_electron_and_photon_transport.pdf). I simulated everything with the Peneloppe model and the fluorescence generation is working well (i can see some fluo x rays inside my logs). This is the code that logs this data (part of it) : ```

// — Pre-calculate kinematics —

const auto momentumDir = track->GetVertexMomentumDirection();

double theta = momentumDir.theta() * 180.0 / M_PI;

double phi = momentumDir.phi() * 180.0 / M_PI;

/*// Correction for negative X momentum (specific legacy logic)

if (track->GetMomentum().x() < 0) {

theta = 360.0 - theta;

}

*/

const auto pos = (post->GetPosition() + pre->GetPosition()) / 2.0;

const auto corrected_pos = pos - this->fOriginPrimary; // Now guaranteed to be initialized

if (!fanalysisManager)

{

fanalysisManager = G4AnalysisManager::Instance();

}

// — Ntuple 0: Primary Electron Stopping —

// Specific check: Track 2 (Primary e-), Step 2

if (trackId == 2 && track->GetCurrentStepNumber() == 2 && particleName == “e-” && isHgCdTe)

{

fanalysisManager->FillNtupleIColumn(kNtuple_ElectronPos, 0, globalStepId);

fanalysisManager->FillNtupleIColumn(kNtuple_ElectronPos, 1, eventID);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 2, edep / keV);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 3, track->GetVertexKineticEnergy() / keV);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 4, corrected_pos.x() / um);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 5, corrected_pos.y() / um);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 6, corrected_pos.z() / um);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 7, track->GetMomentum().x());

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 8, track->GetMomentum().y());

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 9, track->GetMomentum().z());

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 10, theta);

fanalysisManager->FillNtupleDColumn(kNtuple_ElectronPos, 11, phi);

fanalysisManager->AddNtupleRow(kNtuple_ElectronPos);

}
```. This code logs a bunch of data inside ROOT files, and I process them with a python script in order to plot them. Do you know what causes this problem ? I looked a little bit at the code and it is supposed to work (i only looked at the photoelectric process for peneloppe).

Geant4 Version: 11.40
Operating System: WSL2 (Ubuntu 24.04 on Win11)
Compiler/Version: GCC 13.3.0
CMake Version: 3.28.3

---

PS : another problem is that thoses deexcitation particules (fluorescence gammas and auger electrons) are not created by their own process,so they are more difficult to track.!

[jrellin]

I may be misreading what you have but isn’t this just a case of not normalizing by the spherical jacobian? In other words:

dΩ/dθ ∝ sin(θ)

Isotropic emission means the number of particles per unit area so you just need to divide your theta distribution by sin(θ). You can add a very small offset to theta to avoid blowing up at the poles. Phi does not have this problem, of course. Lattitude lines have different lengths but longitude lines do not, for example.

[Nolane_Delille]

That make senses now ! I tested it and it works great !

Thank you very much !