1- Does Geant4 have the ability to calculate the intrinsic energy resolution of scintillating crystals?
Optical/LXe example does not do this, in that example we should enter a multipiler manually, AddConstProperty(“RESOLUTIONSCALE”, 1.0);
.
2- What about total energy resolution including the effect of pmt and etc?
3- if Geant4 can’t do this, is there any other software suite for this purpose(FLUKA, or any other)
4-can you introduce any reference which have the reliable experimental energy resolution (total energy resolution) for common scintillation crystals (15-20 case)
I don’t think Geant4 is the right approach to this. The intrinsic energy resolution of a scintillator is determined by chemical, optical, molecular, etc. properties which is not the point of a physics package such as this. Even the optical is something Geant handles with empirical values that are from the same papers you could find online. I suppose you could account for crystal size and light transport in the optics libraries but the actual rise time, fall time, index of refraction, radiation length, etc. will all be empirical.
As independent noise sources, the PMT noise and electronic noise you have would just add in quadrature with the intrinsic energy resolution. It also will depend on which PMT and your electronics. The PMT would likely be from the manufacturer since a proper accounting for the fields would require sophisticated EM modeling software. The electronics is just easiest to measure with a pulse generator.
You can do parts of these with different Monte Carlo codes. Perhaps the easiest from an out of the box perspectrive would be GATE, which uses the reported resolution of electronics, PMT, timing, crystal material, etc. for system level. GATE is a wrapper for Geant4 intended for medical imaging so might be challenging to use in high energy physics, for example.
This is something you can find in most textbooks now including Knoll “Radiation Detection and Measurement”. Or any scintillator overview paper such as this one. Intrinsic energy resolution does not change with technological developments.
Geant4 models the energy resolution due to the scintillation process (essentially Poisson or Gaussian statistics for the number of photons generated) and resolution due to photon transport to the readout device. However, it does not model the rest of the readout chain (e.g., PMT, voltage dicider, preamp, etc. The best way to calculate the statistics due to the latter, in my opinion, is to model it outside Geant4 using Root, Matlab, Octave, or other third party software.
To add to what the others have said. While you can’t use Geant4 to simulate the intrinsic energy resolution, you can simulate the lowest values that the intrinsic energy could be. Note, you do need empirical data to do this though.
This does require that you have measured the electron response curve for your materials of interest, this curve does somewhat depend on the manufacturing technique of the scintillator so you would expect it to vary depending on your manufacturer, Measurements of NaI(Tl) Electron Response: Comparison of Different Samples | IEEE Journals & Magazine | IEEE Xplore. Unfortunately, manufacturer’s do not provide the electron response curves in spec. sheets so you may be out of luck depending on what scintillator you want to model.
To reiterate, the papers linked simulate the non-proportionality term, NOT the intrinsic energy resolution, however you can somewhat capture the shape of intrinsic energy resolution as a function of energy by calculating the non-proportionality term.
The above image is taken from Intrinsic energy resolution of NaI(Tl) - ScienceDirect. Do be aware that you can’t actually directly measure intrinsic energy resolution, Moszynski had to make various assumptions to get the measured points in the above image.
I’m far from an expert here, but “two terms, the non-unique manner that x/gamma rays deposit their energy and the non-unique manner that the secondary electrons deposit their energy”. This I don’t understand. x/gamma rays do not in themselves contribute to signals - only the charged secondaries they produce cause scintillation or electron-hole pairs that are detectable.
I may end up butchering the explanation but I will do my best, for an indepth explanation I would suggest looking at the first paper I linked.
In inorganic scintillators like NaI(Tl) or CsI(Tl) the number of scintillation photons produced is not proportional to the deposited energy. The electron response curve represents the relative number of photons produced for an electron of a given energy, and is measured using the Compton Coincidence technique, this curve represents the nonproportional response to energy.
The manner that x/gamma rays deposit their energy determines the number and energies of the secondary electrons produced. For example we expect a different number and energy of electrons from events where only photoelectric absorption occurred vs a case where the x/gamma ray Compton scattered then underwent photoelectric absorption. If the electron response was constant then the secondary electron distribution would not matter, however, since this is not the case the number of scintillation photons produced is somewhat dependent on how the x/gamma deposited it’s energy.
The intrinsic energy resolution accounts for this. Intrinsic energy resolution specifically refers to converting the energy of charged secondaries (mostly electrons) into light. The origin of the non-proportionality (or this) is the difficulty in accounting for the competing non-radiative transitions. The Bethe-Bloch formula leads to charged particles having very high LET near the end of their range. This ionization density is significantly higher than at any other point in the track(s) that cause distortions in the phase space for transitions. For organic scinatillors this would be breaks in carbon rings and in inorganic scintillators in the creation of highly localized traps. In both cases, quenching.
Put another way, there are not enough radiative centers near the end of the track and/or you are effectively destroying them. This effect is always present but diminishes at higher energies because the loss (on average) becomes a smaller fraction of the total deposited energy. This is what your plots show as well. Stephen Derenzo has decades of work modelling this but there are still required empirical inputs.
I have been reading through the first paper you linked, and am happy to have done so! I won’t be needing to fit arbitrary functions to non-linearity curves anymore and can use a physics based one instead! Please correct me if I am wrong but it appears to me the first paper had a terminology change, it appears that what Stephen defined as the non-proportionality term is equivalent to the intrinsic energy resolution as defined in other literature.
In the first paper I linked, Valentine defined two intrinsic energy resolutions, the photon and electron intrinsic energy resolution. Perhaps this is the source of confusion?
In the equation for detector energy resolution listed in the 1995 Dorenbos paper I linked, the intrinsic energy resolution term represents the photon intrinsic energy resolution. Hence, I always thought that when folks discussed the intrinsic energy resolution they were talking about the photon intrinsic energy resolution. So maybe I was working with different definitions?
You can break up the intrinsic energy resolution into linear and nonlinear components but it is still tied to electron energy and should have nothing to do with the gamma’s interaction “history”.
This is non standard terminology. Near as I can tell its being used as a means to distinguish the linear portion from the non-linear portion but its still coming from the ionization density at the end of electron tracks.
Do we disagree that the gamma interaction “history” influences the energy distribution of the secondary electrons produced from gamma ray interaction? Quoting the first reference you linked
We are now ready to explore nonproportionality’s
contribution to the resolution of a scintillator. It is recognized
to consist of two independent phenomena relating to: (1)
variations in the distribution of primary electrons created by
the gamma photon
Here Stephen states that the energy distribution of the secondary electrons impacts the resolution of a scintillator spectrometer. When Stephen says primary electrons I understand that to mean the secondary electrons produced from direct gamma ray interactions, and does not include secondary electrons produced from electrons produced from direct gamma ray interactions. For example delta rays would not be considered primary electrons in this case.
Sure. I guess that definition works but it will cause confusion to name it “photonic” such as @allison. Photonic typically refers to atomic states and gamma to nuclear states. There are thousands or tens of thousands of photonic states being accessed with every incident single gamma ray (from one source nuclear state transition). The number of primary electrons pushes the electrons to “start” further down the curve you have above. And since it is nonlinear, 2 photons will lose more energy than a single photon with the energy of both.
I’ll need to think on x-rays, I think PE absorption is so strong for most materials that the effect is even less pronounced.
