Physics list for e-/hole tracking in semiconductor

Physics Lists
1

Hi all,

I’m trying to simulate the charge spread in a semiconductor detector (CdTe) due to a photon interacting in the detector (100keV). For this electrons should be generated with a minimum energy of 4eV and therefore I used the EmLivermorePhysicsList. I expected around 25000 electrons produced inside the semi-conductor but I’am getting around 2000 electrons, even though the production cut energy is set to 4eV. Is this an error in the choice of the physics list and should I be using an ultra low energy like DNA?

Geant4 10.06
Ubuntu 18.04

Thanks for your help,
Juan

2

Hi Juan,

I am not sure of my comment on this but I was faced with the same question at some point. I understood looking online that it is a known limit of Geant where you cannot generally track e-/hole pairs as the energy level is too low for the model used. See https://geant4.web.cern.ch/node/1619
The lower is close to 10eV for specific models in Livermore physics list.
The method I found was to use Geant energy deposition output and then process it into other sofwares focusing on e-/holes tracking.

3

I thought I had replied to this much earlier; apologies. As Florian says, Geant4 does not deal with anything at that low energy. Materials are treated as continuua of atoms, not as crystals, and the low energy properties of crystals (bandgaps, Brillouin zones, etc.) are not treated at all.

Using G4 with higher energy projectiles to figure out where the energy deposits are is fine. Then you can take those energy deposits and “convert” them to e/h pairs which would be modelled in a dedicated semiconductor simulation.

The G4CMP package, available on GitHub, is an ongoing attempt to do this. It is still in development, and there are known limitations with the physics there, especially at high fields. But it is in use for cryogenic (~50 mK) silicon and germanium detectors.

4

Hello Mike,

your explanation is correct. The standard (and Livermore) EM physics provide secondary production only above mean ionisation potential of atoms. Energy of electron/hole is much lower. In the $G4INSTALL/examples/extended/electromagnetic/TestEm8 it is shown how to sample electron/hole pairs. Extra user software software is required to propagate these pairs and produce electronic signal.

VI