Measurement of the scintillation resolution in liquid xenon and its impact for future segmented calorimeters

TL;DR

This study measures the energy resolution achievable in liquid xenon when using scintillation light only, leveraging high-PDE VUV-SiPMs and a maximized light-collection LXe segmented scintillating block. A Geant4 optical Monte Carlo guides the interpretation and saturation-correction of the data, enabling a close-to-Poissonian assessment of the resolution. At 511 keV, the corrected energy resolution is $R_m = 3.7 \pm 0.4\%\,\sigma$, with a combined intrinsic component $R_c = \sqrt{R_i^2 + R_r^2} = 2.3 \pm 0.8\%\,\sigma$, roughly compatible with Doke's non-proportional expectation ($R_i \sim 1.8\%\,\sigma$) and with NEST predictions ($\sim 1.3\%\,\sigma$). The results demonstrate that LXe scintillation-only segmented calorimeters can be highly competitive for PET and other calorimetric applications, motivating further advances in photodetector technology and optical design.

Abstract

We report on a new measurement of the energy resolution that can be attained in liquid xenon when recording only the scintillation light. Our setup is optimized to maximize light collection, and uses state-of-the-art, high-PDE, VUV-sensitive silicon photomultipliers. We find a value of 3.7 $\pm$ 0.4% $σ$ at 511 keV, once saturation effects are corrected for, a result close to the Poissonian resolution that we expect in our setup (2.8 $\pm$ 0.4% $σ$ at 511 keV). Our results in the intrinsic resolution (2.3 $\pm$ 0.8 % $σ$) are compatible, within errors, at 511 keV, with those found by theoretical estimations which have been standing for the last twenty years, 1.8% $σ$. Our work opens new possibilities for apparatus based on liquid xenon and using scintillation only. In particular it suggests that modular scintillation detectors using liquid xenon can be very competitive as building blocks in segmented calorimeters, with applications to Positron Emission Tomography technology.

Figures (13)

• Figure 1: A schematic illustrating the concept of SSB.
• Figure 2: Intrinsic resolution in LXe. Reproduced from Ref. non-prop. The original data were reported in $\%$ FWHM and have been converted here to $\%$$\sigma$ (FWHM = 2.355 $\sigma$) for consistency with the rest of this work.
• Figure 3: A diagram of our experimental setup (top), with an enlarged view of the SiPM arrays (bottom). The dimensions of the relevant parts of the system are specified.
• Figure 4: The aluminum cube holding the two SSBs used for the measurement. Notice the thermal links attached to the cold head.
• Figure 5: Top: instrumented plane of four arrays of 4$\times$4 SiPMs. Bottom: teflon piece defining the SSB.