ASTAROTH: A Novel Detector for Dark Matter Direct Detection Using Cryogenic SiPMs

TL;DR

ASTAROTH investigates replacing PMTs with cryogenic SiPMs for NaI(Tl) dark matter detectors to enhance low-energy sensitivity and reduce background. The authors present a 360 g NaI(Tl) prototype read out by a 64-channel SiPM matrix operated at ~80 K, achieving 4.5 phe/keV after crosstalk correction and validated linear response. The results demonstrate the feasibility of SiPM-based readout for NaI(Tl) and outline concrete steps toward larger, low-background experiments with active vetoes. The work provides a practical path to sub-keV detection thresholds for verifying DAMA-like signals and guiding the design of next-generation dark matter detectors.

Abstract

The DAMA experiment's long-standing claim of dark matter detection remains a key open issue in astroparticle physics. Independent verification requires NaI(Tl)-based detectors with enhanced low-energy sensitivity. Current detectors rely on photomultiplier tubes (PMTs) which features limited detection efficiency, intrinsic radioactivity, and high noise at keV energies. ASTAROTH is an R&D project that developed a proof of concept NaI(Tl) detector where siliconphotomultipliers (SiPMs) have been used instead of PMTs, offering higher photon detection efficiency, negligible radioactivity, and, most of all, a reduction of two orders of magnitude in the dark noise. The setup includes a custom cryostat operating at approximately 80 K. We report the first characterization of an approximately 360 g NaI(Tl) crystal coupled to a 5 x 5 cm SiPM matrix, yielding 4.5 photoelectrons\keV after crosstalk correction. This promising result demonstrates the feasibility of SiPM-based readout for NaI(Tl) and paves the way for future large-scale dark matter experiments.

Figures (5)

• Figure 1: Design of the cryostat. The cryogenic chamber (1) is submerged into a cold fluid (2) exchanging heat through the thermal bridges (3). The detector (5) is hosted inside the chamber, which is connected to the chimney (4).
• Figure 2: Design of the detector. A SiPM array is coupled to the crystal by means of a copper frame, the electronic front-end reads out the signal from the SiPM.
• Figure 3: (a) Baseline subtracted waveforms from the SiPM output, the amplitude of the pulse linearly increases with the number of detected photons. Integration window for the charge computation is indicated between dashed lines. (b) Histogram of charge corresponding to detected photon with superimposed Gaussian fitting for each peak.
• Figure 4: Pictures of the experimental setup, the NaI(Tl) crystal is covered into PTFE tape for increasing the diffuse reflectivity onto the non-instrumented faces and then, thanks to a copper frame, coupled to the SiPM. After positioning a $^{241}$Am source at 25 cm distance, the detector is installed into the cryostat and cooled down to $\approx$ 80 K.
• Figure 5: Energy spectra comparing two parallel run performed with the 241Am source (blue) and without (magenta). The peak stands at 59.5 keV due to gammas from the source. A Gaussian fitting returns the peak position and a gross photoelectron yield of 7.2 phe/keV (4.5 phe/keV after crosstalk correction). The upper x-axis is computed consequently.