UC Berkeley

The barium fluoride (BaF2) core-valence (CV) luminescence mechanism (the ”fast” emissioncomponent) offers an unrivaled balance between decay time constant and light yield, with the potential to enable a level of timing performance precluded by the intrinsic properties of other scintillators. Now more than ever, the technology to realize this potential is readily available, including photodetectors with sensitivity and timing amenable to the fast component, and various techniques to selectively suppress the much slower self-trapped exciton (STE) luminescence (the ”slow” component) for faster total response. In particular, STE luminescence suppression reduces pulse pile-up by eliminating the long-lived pulse tail.

Inspired by the unique properties of BaF2 crystal and the advance and integration of relevanttechnologies, a prototype BaF2-based radiation detection system has been developed from commercially available components and characterized. Comprising multiple pixelated detector modules with a compact, practical form factor, the system provides a versatile test bed for timing applications, with commensurately variable geometry. Each module includes an 8x8 array of BaF2 crystals, coupled on one or both ends to a segmented microchannel plate photomultiplier tube (MCP-PMT), and optical transmission interference filters (242 nm long wavelength transmission cutoff) as a simple means of STE luminescence suppression (if desired). The original data acquisition control software rosetta was developed by the author to operate the system and was used to collect data characterizing the unfiltered system coincident time resolution (CTR), the performance of the filters for STE luminescence suppression, and the impact of the filters on the system CTR.

The unfiltered system CTR is calculated to be 1.2 ns—roughly six times worse than the 210 nsCTR of Siemens’ Biograph Vision digital PET/CT system. The poor relative performance of the prototype system is predominantly attributed to 1. increased scintillation transit time spread due to larger crystals (three times the size of the lutetium oxyorthosilicate (LSO) crystals used in the Biograph Vision), and 2. degraded timing signal quality due to flawed readout hardware. The filter performance is assessed relative to unfiltered BaF2 in terms of the ”fast fraction,” i.e., how much of the scintillation signal consists of fast component light, and ”fast transmission,” that is, how much of the fast component light is preserved by the filter. The filters increase the fast fraction from 12% (unfiltered) to 50%, but they transmit only 17% of the fast light and degrade the system CTR by at least a factor of two. This degradation in timing performance is attributed to the severe loss of photons represented by the poor filter fast transmission.

The full potential of BaF2 for high time resolution has not yet been tapped, and photodetectordevelopment will act as the primary driver for future progress. Some applications would benefit from the excellent time resolution of BaF2-based systems, but cannot tolerate pulse pile-up. Thus, the refinement of slow component suppression techniques will broaden the application space of BaF2 by eliminating the long-lived tail without compromising time resolution. Finally, the development of new materials with superior ratios of decay time constant to light yield would push the limits of timing further than ever before. Each component in the prototype test bed can be swapped out to accommodate such progress, providing a system-level performance assessment capability that is sufficiently flexible to keep pace with an ever-evolving state of the art.