RESUMEN
Neutron double scatter imaging exploits the kinematics of neutron elastic scattering to enable emission imaging of neutron sources. Due to the relatively low coincidence detection efficiency of fast neutrons in organic scintillator arrays, imaging efficiency for double scatter cameras can also be low. One method to realize significant gains in neutron coincidence detection efficiency is to develop neutron double scatter detectors which employ monolithic blocks of organic scintillator, instrumented with photosensor arrays on multiple faces to enable 3D position and multi-interaction time pickoff. Silicon photomultipliers (SiPMs) have several advantageous characteristics for this approach, including high photon detection efficiency (PDE), good single photon time resolution (SPTR), high gain that translates to single photon counting capabilities, and ability to be tiled into large arrays with high packing fraction and photosensitive area fill factor. However, they also have a tradeoff in high uncorrelated and correlated noise rates (dark counts from thermionic emissions and optical photon crosstalk generated during avalanche) which may complicate event positioning algorithms. We have evaluated the noise characteristics and SPTR of Hamamatsu S13360-6075 SiPMs with low noise, fast electronic readout for integration into a monolithic neutron scatter camera prototype. The sensors and electronic readout were implemented in a small-scale prototype detector in order to estimate expected noise performance for a monolithic neutron scatter camera and perform proof-of-concept measurements for scintillation photon counting and three-dimensional event positioning.
RESUMEN
The position-sensitive alpha-particle detector used to provide the starting time and initial direction of D-T neutrons in a fast-neutron imaging system was simulated with a Geant4-based Monte Carlo program. The whole detector system, which consists of a YAP:Ce scintillator, a fiber-optic faceplate, a light guide, and a position-sensitive photo-multiplier tube (PSPMT), was modeled, starting with incident D-T alphas. The scintillation photons, whose starting time follows the distribution of a scintillation decay curve, were produced and emitted uniformly into a solid angle of 4π along the track segments of the alpha and its secondaries. Through tracking all photons and taking into account the quantum efficiency of the photocathode, the number of photoelectrons and their time and position distributions were obtained. Using a four-corner data reconstruction formula, the flood images of the alpha detector with and without optical grease between the YAP scintillator and the fiber-optic faceplate were obtained, which show agreement with the experimental results. The reconstructed position uncertainties of incident alpha particles for both cases are 1.198 mm and 0.998 mm respectively across the sensitive area of the detector. Simulation results also show that comparing with other faceplates composed of 500 µm, 300 µm, and 100 µm fibers, the 10-µm-fiber faceplate is the best choice to build the detector for better position performance. In addition, the study of the background originating inside the D-T generator suggests that for 500-µm-thick YAP:Ce coated with 1-µm-thick aluminum, and very good signal-to-noise ratio can be expected through application of a simple threshold.