Detailed Monte Carlo simulation of energy integrating and photon counting semiconductor detectors

X-ray imaging techniques widely employ semiconductor detectors. Energy integrating (EI) detectors are used in digital radiography and photon counting (PC) in CT. This work aims to implement a detailed Monte Carlo modeling of these sensors. The model was divided into radiation interaction and electron-hole pairs (EHP) creation and dispersion. The PENELOPE code simulated the radiation transport. In each electron interaction, the absorbed energy was converted into EHP considering the pair creation energy and the Fano factor. The detection position was sampled using a Gaussian distribution, where the standard deviation was from the Einstein diffusion equation. The Hetch equation models the charge trapping. In the PC mode, the photon was counted if the energy deposited was higher than a threshold (ethr). Monoenergetic pencil beams between 10 and 100 keV were simulated, with 107 histories. The detector material was cadmium tellurite, with 50 mu m pixel size, whose thicknesses, applied electric field, and ethr vary, respectively from 250 to 1000 mu m, 0.01 to 1 V/mu m, and from 1 to 50 keV. The results show a wider detector response as the beam energy increases. For energies above 32 keV the fluorescence is greatly responsible for this spread. The detector's efficiency increases with the sensor thickness and decreases with the photon energy. Charge trapping decreases the efficiency up to 43,53%. For the PC mode, an ethr increase yields a narrower detector response and increases the image noise. This study provides a detailed detector modeling and, consequently, insight into the imaging system's fundamental limitations.