Difficulty replicating experimental CZT response function

Hello - I’m trying to learn GEANT4 in order to simulate the response functions of my CZT detectors for incident photons of energies between 10 and 300 keV.
By adapting some starter code, I was able to fire photons at a cube of CZT, which mirrors the crystal geometry of my detector. (I have also tried a slightly more detailed setup that matches experiment better, but that does not solve the problem detailed below)

However, the resulting simulated response function does not match the experimental one well. Attached are the simulated and experimental response functions, and one can see that the photopeak is greatly over estimated compared to experiment.

I am told that this may be due to GEANT4 not modelling charge carrier trapping, but I am unsure of how to include this. Any advice would be greatly appreciated. Thanks!

responseFunctionComparison.pdf (59.6 KB)