Hi geant4 experts,
I noticed that the simulation of scintillation in geant4 needs experimental data. It use properties of material like SCINTILLATIONCOMPONENT1 to asign the light emission spectrum. This is more like a mathematical simulation insteads of physical simulation. Is it possible to simulate the scintillation from physical models? Which model can describe the amount of photons emited and its wave length distribution of a scintillator?
1 Like
Hi,
Geant4 uses a semi-empirical approach to the scintillation process. I think the reason for this is that scintillation is a molecular and atomic process, not a nuclear one. It is a very intricate and convoluted process involving interactions between molecular orbitals and atomic energy levels, with photon absorption and emission at extremely low energies (from a nuclear physics point of view). Incorporating a computer model of scintillation from physics first principles into the Geant4 framework would be very complicated and time consuming. I am not sure there would be sufficient value added to make it worth the effort, given that the existing scintillation package works very well for the problems for which it was intended.
As stand alone software for studying scintillation itself, rather than nuclear physics problems, such a first principles modeling package might be of interest.
However, I am no expert on modeling the molecular physics of scintillation. Perhaps one of the experts out there might like to offer an opinion.
Hi @John_McFee,
Thanks for your message. I understand that it’s reasonable for geant4 to use a semi-empirical approach to simulate the scintillation. It’s indeed work and effective. But I want to understand the details of the process. I searched online but still didn’t find any physical model which can drive out the photon emission spectrum of scintillation or equations summarize the amount photons emited by different scintillation materials.
1 Like
This feels like a job for QMD, perhaps?
I am not aware of any ab initio modeling codes for scintillation processes, although to be fair I haven’t spent much time looking for them. I am not sure if that is because scintillation is very complicated to model or because there just isn’t a need for it.
1 Like
Hi I do not think QMD is appropriate
Fair enough! I’m just not sure what is out there to model atomic and molecular orbitals and their interactions with light, transition lifetimes, blah blah blah.
no worries. But I see you are very knowledgeable in many, many other aspects of Geant4 ! 
Hi @guatelli,
thanks. I am studying scintillation and is expected to understand it prefoundly. As for the usage of geant4, currectly, I can answer some basic questions for others and I am glad to help.
Hi repan
i am currently trying to make a scintillator in geant4 have you found the values for SCINTILLATIONCOMPONENT1 for NaI or any other model i am not able to get values for this.
Thanks for help in advance.
NaI(Tl) seems to have several time constants, however 250ns is the dominant component at room temperature from literature. You can get SCINTILLATIONCOMPONENT1 by looking up a scintillator manufacturer’s spec sheet for NaI(Tl), e.g., Scionix or St Gobain. There should be a curve of emission intensity vs (optical photon) wavelength. Digitize that curve into two vectors - one of intensity values and one of corresponding wavelength values. Convert those two vectors to a vector of intensity and a vector of corresponding optical photon energy values, with energy increasing as the vector index increases. The intensity vector is SCINTILLATIONCOMPONENT1 (it just has to be in arbitrary units) and the optical photon energy vector is its associated energy vector.
The same procedure applies to any other scintillator, although more extensive investigation of the literature may be required to find the data for particle-dependent scintillation and multiple decay components.
1 Like
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.