Timing-performance evaluation of Cherenkov-based radiation detectors

Abstract

With the upgradation of detector components, such as scintillators and photodetectors, the PET-image signal-to-noise ratio of time-of-flight positron emission tomography (TOF-PET) systems has been improved, compared to those of ordinary nonTOF-PET systems. A TOF-PET with an ultrahigh time resolution, for example a coincidence time resolution (CTR) better than few tens of picoseconds, can not only improve the image quality, but also remove the image reconstruction process, significantly impacting medical imaging. Therefore, it is crucial to develop a high-time resolution PET detector. We focus on the prompt emission of Cherenkov radiation, owing to the instantaneousness of which, a high time resolution can be expected. One of the candidates for the Cherenkov radiator is lead fluoride (PbF2) due it has excellent properties, including transparency toward the ultraviolet region, high refractive index (n = 1.82), and high density (7.77 g/cm3). Moreover, it does not contain radioisotopes, unlike lutetium-based scintillators, which are commonly used in the currently available TOF-PET detectors. In this work, we experimentally investigate the timing performance of PbF2-based Cherenkov detectors, breaking down the timing performance into physical components. 3×3×5 mm3 and 9.6×9.6×5 mm3 PbF2 crystals are used as Cherenkov radiators; both are attached to a microchannel plate photomultiplier tube (MCP-PMT) because the single channel MCP-PMT is one of the best photodetectors in terms of the SPTR, which is 25 ps full width at half maximum (FWHM). All the surfaces, except the end surface where the MCP-PMT is connected, are wrapped in black tape to suppress the reflections of the Cherenkov photons in the crystal. The CTR is measured by placing a detector pair face-to-face, using an 22Na point source, and an oscilloscope at 20 GS/s with a set bandwidth of 4.2 GHz. A CTR of 46.9 ps FWHM, corresponding to a position resolution of 7.0 mm, is obtained, consistent with our simulation results.