Table of Contents
Fetching ...

Light-Based Fast Timing in Bulk CsPbBr3 Crystals for TOF-PET and Proton Range Verification

Nicolaus Kratochwil, Leonor Rebolo, Indra R. Pandey, Joshua W. Cates, Emilie Roncali, Joshua H. Palmer, Gerard Arino-Estrada

TL;DR

This work investigates using the Cherenkov light component in bulk $\mathrm{CsPbBr}_3$ crystals to achieve fast timing in gamma-ray detectors for TOF-PET and proton-range verification. By combining optical simulations with experimental measurements from two CLB crystals coupled to SiPMs, the authors quantify the Cherenkov photon yield and demonstrate a coincidence-time resolution of $\mathrm{FWHM} = 419\ \mathrm{ps}$ (detecting $\mu \approx 1.18$ photons on average per event). A Poisson+crosstalk model plus detailed post-processing reveals that experimental photon yields exceed baseline simulations, suggesting an additional light component possibly from UV transparency or room-temperature scintillation. The results indicate CLB-based Cherenkov timing is feasible for fast timing in practical PET and PRV contexts, while highlighting avenues for optimization and further investigation of the extra light source and detector response at higher energies.

Abstract

Halide perovskite semiconductors such as CsPbBr3 (CLB) are emerging gamma-ray detectors for applications requiring very high energy resolution and potential for fine detector segmentation. Semiconductor detectors typically offer poor time resolution due to the long drift times. Recently, we proposed to use the Cherenkov light component in partially transparent semiconductors to boost the timing capability of such detectors. Cherenkov light produced upon 511 keV gamma-ray interaction with CLB was investigated by means of optical simulations and experimental measurements. The timing capability of a pair of identical CLB crystals ( 3 x 3 x 5 mm3) coupled to NUV- MT silicon photomultipliers was measured. On average, 9.5 Cherenkov photons are produced in CLB between 555 and 900 nm for 511 keV photoelectric interactions based on our simulation framework. Experimentally, we observe 2-to-3 times more photons detected than in the simulation. The two most likely explanations for these additional detected optical photons are either the partial transparency of CLB in the UV, or a mild scintillation light emitted by CLB at room temperature. A coincidence time resolution (CTR) of 419 ps FWHM was obtained by triggering on more than 2 fired SiPM cells and after time walk correction. The measured CTR confirms the feasibility to use the Cherenkov light-component for fast timing applications on top of the charge readout, toward full 3D localization.

Light-Based Fast Timing in Bulk CsPbBr3 Crystals for TOF-PET and Proton Range Verification

TL;DR

This work investigates using the Cherenkov light component in bulk crystals to achieve fast timing in gamma-ray detectors for TOF-PET and proton-range verification. By combining optical simulations with experimental measurements from two CLB crystals coupled to SiPMs, the authors quantify the Cherenkov photon yield and demonstrate a coincidence-time resolution of (detecting photons on average per event). A Poisson+crosstalk model plus detailed post-processing reveals that experimental photon yields exceed baseline simulations, suggesting an additional light component possibly from UV transparency or room-temperature scintillation. The results indicate CLB-based Cherenkov timing is feasible for fast timing in practical PET and PRV contexts, while highlighting avenues for optimization and further investigation of the extra light source and detector response at higher energies.

Abstract

Halide perovskite semiconductors such as CsPbBr3 (CLB) are emerging gamma-ray detectors for applications requiring very high energy resolution and potential for fine detector segmentation. Semiconductor detectors typically offer poor time resolution due to the long drift times. Recently, we proposed to use the Cherenkov light component in partially transparent semiconductors to boost the timing capability of such detectors. Cherenkov light produced upon 511 keV gamma-ray interaction with CLB was investigated by means of optical simulations and experimental measurements. The timing capability of a pair of identical CLB crystals ( 3 x 3 x 5 mm3) coupled to NUV- MT silicon photomultipliers was measured. On average, 9.5 Cherenkov photons are produced in CLB between 555 and 900 nm for 511 keV photoelectric interactions based on our simulation framework. Experimentally, we observe 2-to-3 times more photons detected than in the simulation. The two most likely explanations for these additional detected optical photons are either the partial transparency of CLB in the UV, or a mild scintillation light emitted by CLB at room temperature. A coincidence time resolution (CTR) of 419 ps FWHM was obtained by triggering on more than 2 fired SiPM cells and after time walk correction. The measured CTR confirms the feasibility to use the Cherenkov light-component for fast timing applications on top of the charge readout, toward full 3D localization.
Paper Structure (13 sections, 3 equations, 7 figures)

This paper contains 13 sections, 3 equations, 7 figures.

Figures (7)

  • Figure 1: 3.2$\times$3.4$\times$5.4 mm$^3$ crystals employed in this work.
  • Figure 2: Photon detection efficiency, simulated transparency, and calculated Cherenkov photon emission profile as function of the wavelength. The Cherenkov emission profile was calculated according to Kratochwil_TRPMS_2024.
  • Figure 3: Simulated Cherenkov photons in CLB considering all interactions (top) or 511 keV photoelectric interactions only (bottom).
  • Figure 4: Top: 2D histogram of the signal amplitude. The red and black lines serve as eye-guide to distinguish different number of triggered cells from 3 to 9. Bottom: Histogram of the signal amplitude for one detector. The distribution is fitted with a sum of 7 gaussian distributions (red) with each contribution drawn in blue. The complementary cumulative distribution (normalized) is shown in black with the right y-axis.
  • Figure 5: Frequency of the triggered SPADs for the two detectors (black, red) and average (blue) with fit function considering measured crosstalk (blue solid line). Ignoring crosstalk, the mean number of detected photons is almost 2-fold overestimated.
  • ...and 2 more figures