TCAD simulation of 3D silicon sensors for thermal neutron imaging

Within the INFN HYDE project, we have developed pixelated hybrid detectors for thermal neutron imaging. Devices are based on microstructured silicon sensors coupled with converter materials. Compared to standard 3D sensors, the proposed device offers a more convenient trade-off between technological complexity and neutron detection performance. Nevertheless, experimental results have shown a lower detection efficiency with respect to the theoretically expected value, which was attributed to an insufficient density of negative fixed charge in the Al2O3 layer used to passivate the cavities. TCAD simulations have been performed aimed at gaining deeper insight into this problem. The results reported in this paper confirm the sensitivity of the sensor performance to the properties of the Si/Al2O3 interface in terms of fixed charge and surface generation/recombination velocity, providing useful hints for future optimization of the device.