HPGe detector efficiency

Hello, I want to simulate the HPGe detector efficiency in Geant4 but I have no idea how to do it since I am a new in Geant4 and MC simulations.

Which physics list should I use? If you recommend me G4RadioactiveDecay, why? Since I can directly simulate some discrete gamma ray energies and to plot the determined efficiency as a function of energy. Also, what is meant by “detecting a particle” in Geant4? How can I find this efficiency? I mean, what do I have to divide because the formula for intrinsic efficiency is:
(# of gammas detected) / (# of incident gammas)

Thank you in advance, I really need to have a very clever picture about this kind of simulation!

1 Like

The meaning of “detected” varies depending on one’s point of view. If a particle enters a detector volume, one could say it is detected. However, if it leaves no energy in the detector, it is not really detected in any practical sense. Since your simulation is intended to approximate reality, detection should include a measure of energy deposited. So I would define intrinsic efficiency as (# of gammas that deposit some energy in the detector)/ (# gammas that are incident on the detector surface).

If you just want to estimate the efficiency through simulation, it’s best to keep it as simple as possible. Modify one or the basic examples, such as B1 or B4, for the geometry you want. You can use a GPS or ParticleGun source for the primary source of gamma rays or neutrons. For gamma rays, choose an energy that corresponds to one of the lines of one of the common check sources, say 0.662 MeV for Cs137. I prefer to use a GPS source and configure it using macro commands for source type, position, energy, number of source particles, etc.,

For a simple simulation, you do not need RadioactiveDecay. I would suggest you use QGSP_BERT_HP (in case you want to measure efficiency from neutron interactions too) with the G4EmStandardPhysics_option4.

There are many ways to determine the intrinsic efficiency. The simplest is to emit n primary source particles with a point source isotropic emitter sitting on one surface of the detector. Record the total energy deposited in the detector for each event. Count the number of events with non zero energy and divide the sum by n/2. (The source emits into 4pi, so 1/2 are incident on the detector volume.)

If you use a more realistic geometry and/or source, you have to figure out how many source particles are incident on the detector. You can also save the deposited energy event by event to a file. After a run of n primary particles, you will have a file with n entries: each one the deposited energy for its event. This is essentially the spectrum that the ideal HPGe would measure and can be processed outside Geant4.

To be more realistic, one can set energy thresholds above which a gamma is considered detected. One can also convolve the energy spectrum with a response function to account for electronic noise and other effects such as energy broadening.

1 Like

Ok, I like the idea of a point source on one of the surfaces of the detector. Thank you for it :).

My second question is how can I record and count the number of events with non zero energy? What classes can I use and how? Can I find any example of such simulation?

I think example B1 can be simply modified to do what you want.

1 Like