I am currently involved in a project which we are investigating if a type of gas detector will actually work in real life. It involves scintillation, and right now the goal with the simulation is to determine approx. how many optical photons will be produced from the process. Pretty basic stuff. However, I have gotten stuck in a catch 22, since to produce scintillation simulation (from my understanding) we have to specify things such as the scintillation yield. Is there anyway to produce scintillation WITHOUT specifying the yield since we want to use the simulations to compare against calculations.
I have some idea which maybe worth and it is an alternative way i.e. e-dep simulations from geant4 and postAnalysis inframe of G4 o/p.
Step 1
In gas detector the energy loss by the particle (say alpha) will lose its energy along the track. /tracking/verbose 1. This information controlled by Geant4.
Step2:
The generation of optical photons depends on many factors however, the approximation can be done based on the energy loss by the particle (alpha). The energy loss by the particle means generation of the scintillation based on the work function of the gas. For more details any standard general Physics text book.
Step2 continue:
I am unable to understand catch 22? However, the way I said is the conversion of energy loss by the particle into number of photons is in general is = sintillation eff. x E_dep where E_dep is the energy loss by the particle along the track and stored in steps. The responsible function is G4Step.hh
Once a particle has ionized the medium, scintillation photon production is determined by chemical processes, not nuclear processes. It is a function of molecular composition and the processes are complex. (See J. B. Birks, The Theory and Practice of Scintillation Counting, Pergamon Press, 1964, for example.) It is very specific to the molecular species and I’m not sure how one would make a general simulation module that would cover a wide range of molecules and still be able to calculate photon yield, time profile, etc. from scratch.
It might be possible to write your own ab initio scintillation yield module for a specific simplistic molecule. There is also general purpose ab initio quantum chemistry software (see Quantum Chemistry Software) that might possibly do that to some degree, but that’s outside my area of expertise.
In any case, Geant4 does not have that modelling capability.
@John_McFee I truly agree that one can not develop such complex simulations in only Geant4, however, the most important information like e-dep, global time, phase space, etc along the track [only geant4 can do with more precision] is worth in developing own model via post processing methods.
