In the present work, Monte Carlo simulations using GEANT4 are carried out to estimate the efficiency of semiconductor neutron detectors with depleted UO2 (DUO2) as converter material, in both planar (direct and indirect) and 3D geometry (cylindrical perforation and trenches structure) configurations. The simulations were conducted for neutrons of variable energy viz., thermal (25 meV) and fast (1 to 10 MeV) that were incident on varying thicknesses (0.25 μm to 1000 μm), diameters (1 μm to 9 μm) and widths (1 μm to 9 μm) along with depths (50 μm to 275 μm) of DUO2 for planar, cylindrical perforated and trench structures, respectively. In the case of direct planar detectors, efficiency was found to increase with the thickness of DUO2 and the rate at which efficiency increased was found to follow the macroscopic fission cross section at the corresponding neutron energy. In the case of indirect planar detector, efficiency was lower as compared to direct configuration and was found to saturate beyond a thickness of ∼3 μm. This saturation is explained on the basis of mean free path of neutrons in the DUO2 material. For the 3D perforated silicon detectors of cylindrical (trench) geometry, backfilled with DUO2, the efficiency for detection of thermal neutrons ∼25 meV and fast neutrons ∼ typical energy of 10 MeV was found to be ∼0.0159% (∼0.0177%) and ∼0.0088% (0.0098%), respectively. These efficiency values were two (one) order values higher than planar indirect detector for thermal (fast) neutrons. Histogram plots were also obtained from the GEANT4 simulations to monitor the energy distribution of fission products in planar (direct and indirect) and 3D geometry (cylindrical and trench) configurations. These plots revealed that, for all the detector configurations, the energy deposited by the fission products are higher as compared to the typical gamma ray background. Thus, for detectors with DUO2 as converter material, higher values of low level discriminator (LLD) can be set, so as to achieve good background discrimination.
Read full abstract