Hello,

Looking at the decay of the meta-stable state Ag109[88.034] I would expect the total deposited energy in my detector to be 88.034 keV. While I get a small full-energy peak at 88.034 keV (corresponding to the 3.7% of X-rays which should be emitted), I do get peaks at 62.5 keV and 84.5 keV corresponding to 88 keV minus 25.5 (K-edge) or minus L-edges, etc.

Looking into the source code of G4PhotonEvaporation, G4NulcearDecayChannel and G4DiscreteGammaDeexcitation and G4VGammaDeexitation, in these cases internal conversion happens (excitation energy is transfered to electron). The code handles correctly the binding energy and thus transfers less energy to the electron and generated a vacancy in the shell -- that's all as expected. However, the information about the generated shell vacancy in G4VGammaDeexcitation::GenerateGamma is not passed on to the classes upstream, thus G4NulcearDecayChannel::DecayIt cannot perform the atomic deexcitation, since it simply doesn't know it happened. As consequence, we do not have energy conservation since we are missing some fluoresence X-rays/Auger electrons.

The problem can be easily reproduced using Ag109[88.034] with the rdecay01 example.

This problem has been found with Geant4 10.01.p01 and also in 10.00.p04 the spectrum does not show the correct peaks.

If you have further questions, feel free to contact me.

Cheers,
Andreas