Exploring the performance of SiPM at cryogenic temperature for the sub-meV threshold detector

TL;DR

The paper addresses achieving sub-meV energy thresholds by exploiting Cooper-pair breaking and coupling quasiparticle signals to a PN-junction avalanche readout. It demonstrates that commercial SiPMs can maintain avalanche gain at $10~\mathrm{mK}$ with a substantial reduction in dark count rate and measurable breakdown behavior, reporting $V_{\mathrm{Gbd}}^{10\mathrm{mK}} = 21.08 \pm 0.20$ V, $V_{\mathrm{Ibd}}^{10\mathrm{mK}} = 21.76 \pm 0.01$ V, and a second-divergence at $V_{sd} = 26.27 \pm 0.01$ V, while achieving $DCR \approx 5$ mHz/mm$^2$ at $V_{ov}=2.5$ V. These results establish the viability of using SiPMs for cryogenic quasiparticle readout and motivate a conceptual detector design that combines a superconducting layer with a tunnel barrier in a PN-junction avalanche readout (S-I-P-N), offering potential advantages in scalability and timing over traditional TES-based approaches. The work lays the groundwork for simulations and prototype development aimed at sub-meV searches for neutrinos, dark matter, and other rare low-energy processes. Overall, the study demonstrates that SiPMs at millikelvin temperatures provide sufficient gain and extremely low noise, enabling a viable path toward a new class of ultra-low-threshold detectors.

Abstract

We propose a novel particle detector concept that exploits the breaking of superconducting Cooper pairs, offering a theoretical energy threshold at the sub-meV level. A major challenge for such detectors is the readout of quasiparticles at mK temperatures, where the Cooper pair binding energy is at the sub-meV level, making conventional electronic devices ineffective. Here we demonstrate that silicon photomultipliers (SiPMs) retain full avalanche multiplication capability at environmental temperatures as low as 10 mK. Compared to room temperature operation, the dark count rate decreases by seven orders of magnitude while the gain is reduced by only a factor of 4.44. we present a characterization of several important performance parameters of the SiPM at 10~mK, including breakdown voltage, second divergence voltage, operating voltage range, output waveform characteristics, gain, single-photoelectron resolution, dark count rate, and the correlated noise probability. These results show that SiPMs operating at 10 mK provide sufficient gain and extremely low noise suitable for quasiparticle detection. Based on this finding, we propose a conceptual readout scheme for sub-meV threshold superconducting detectors using PN junction electron multiplication, which holds promise for advancing new searches for neutrino, dark matter, and other rare low-energy processes.

Figures (15)

• Figure 1: Schematic diagram of the experimental setup and data acquisition system. The SiPM is securely mounted on an MXC flange maintained at 10 mK using screws. A calibrated RuOx temperature sensor is installed on the MXC flange to monitor temperature changes. To enhance thermal conduction, a small amount of thermal grease is applied between the SiPM circuit board and the MXC flange. The plastic scintillator is directly optically coupled to the SiPM through air and is in close contact with the SiPM; it is used solely as a light source. The scintillator is wrapped with ESR reflective film and irradiated by a $^{241}\text{Am}\ \gamma$-ray source to produce light signals. The SiPM circuit board is connected to the external power supply and signal acquisition system outside the refrigerator via two coaxial cables. The entire refrigerator is enclosed in multiple layers of electromagnetic shielding to minimize external thermal radiation and electromagnetic interference.
• Figure 2: Photograph of (a) plastic scintillator, (b) SiPM board, and (c) preamplifier board. The plastic scintillator is cut to the same size as the SiPM, 6 mm × 6 mm, and is used solely as a light source. The SiPM is mounted in the center of the circuit board, and the signal is amplified by a LMH6629 transimpedance preamplifier. A DC power supply (RIGOL DP831A) provides $\pm$2.5 V to the preamplifier.
• Figure 3: (a) The SPE charge spectrum of the SiPM at 10 mK measured at a bias voltage of 23.5 V. The horizontal axis represents the charge, obtained by integrating the SiPM waveform and then dividing by the load resistor, i.e.,$\int V_{\text{out}}\,dt/R_{\text{load}}$. The first peak corresponds to the single-photoelectron (1 P.E.), the second peak corresponds to the double-photoelectron (2 P.E.), and so on. The positions and standard deviations of the first and second peaks were determined by fitting with Crystal Ball functions. $Q_{\text{SPE}}$ is obtained by taking the difference between the positions of the first two peaks and then dividing by $Z_{\text{TIA}}$, i.e.,$Q_{\text{SPE}} = (\mu_2 - \mu_1)/Z_{\text{TIA}}$. (b) The SiPM gain as a function of bias voltage $V_{\text{bias}}$ at 10 mK. The data are fitted with a linear function, and the breakdown voltage is defined as the intersection of the fit line with the x-axis.
• Figure 4: IV curve (top) and its differential (bottom) at 10 mK. After applying cubic spline interpolation to the curve, it is then differentiated. The extremum of $\frac{d(\ln I)}{dV}$ is obtained by fitting with a Landau-Gaussian convolution function, thereby determining $V_{\text{Ibd}}$. The calculation method for $V_{\text{sd}}$ is the same.
• Figure 5: Temperature dependence of the SiPM $V_{\text{Gbd}}$ and $V_{\text{Ibd}}$ from 100 K to 10 mK. The errors on the y-axis originate from the uncertainties in the fitting parameters. The breakdown voltages calculated by the two methods show the same trend: they decrease from 100K to 50K and then increase from 50K to 10mK.