Enhanced Dark Matter Sensitivity using a Hybrid SiPM-SNSPD-Qubit Detector in Liquid Argon

TL;DR

Sub-GeV dark matter remains elusive with conventional detectors. This work proposes a hybrid liquid-argon detector that combines enhanced SiPM optical readout with a qubit-based phonon sensor, leveraging a nuclear dielectric constant correction to boost low-$q$ nuclear recoils and lower detection thresholds to $\gtrsim$30 meV. A two-chip flip-chip architecture enables nearly unit detection efficiency and ~50% energy resolution at 100 meV, yielding orders-of-magnitude improvements in sensitivity for $m_\chi \gtrsim 0.01~\text{MeV}$ and enabling competitive searches for axions and dark photons in the $0.04-0.2$ eV range. The approach integrates detailed modeling of DM scattering/absorption, phonon transport, and quasiparticle dynamics, illustrating a practical pathway to transformative low-mass DM searches with scalable quantum-enhanced detectors.

Abstract

We investigate novel strategies to extend the sensitivity of dark matter direct detection experiments to energy deposits well below the thresholds of conventional detectors. In liquid-argon time-projection chambers equipped with silicon photomultipliers (SiPMs), we show that improved optical readout, combined with a nuclear dielectric constant (NDC) correction to the WIMP nucleus interaction, enhances the response to low-momentum-transfer nuclear recoils. The NDC effectively amplifies the interaction strength at small recoil energies, increasing the expected ionization and scintillation yields without modifying the high-energy behavior constrained by calibration data. When coupled to SiPM based light collection, this mechanism lowers the effective detection threshold to the subKeV regime, significantly improving sensitivity to low-mass WIMPs and other weakly interacting particles. Complementarily, we present the design and projected performance of a qubit-based detector optimized for ultra-low-energy depositions. A novel two-chip architecture is employed to minimize signal dissipation, while quantum parity measurements enable enhanced single-phonon sensitivity. Full simulations of phonon propagation and quasiparticle dynamics demonstrate that energy deposits at the level of $\gtrsim 30 meV$ can be detected with nearly unit efficiency and high energy resolution. This capability is expected to advance sensitivity to dark-matter scattering for masses $m_χ\gtrsim 0.01 MeV$ by several orders of magnitude for both light and heavy mediators, and to enable competitive searches for axion and dark-photon absorption in the $0.04 - 0.2 eV$ mass range.

Figures (12)

• Figure 1: The schematic drawing of the qubit array and its details.
• Figure 2: Recoil spectrum with/without the nuclear dieletric constant and “SiPM threshold” effect.
• Figure 3: the cross section with/without the nuclear dieletric constant and “SiPM threshold” effect.
• Figure 4: Simultaneous fit to the ReD, ARIS, SCENE, and DarkSide-50 datasets assuming the Lenz-Jensen screening function (green solid line). ReD data points are shown with both prior (gray) and posterior (red) uncertainties. The gray line and its corresponding uncertainty band represent the previous ionization model, based on the ZBL screening function and fitted without ReD data DarkSide:2021bnz. For comparison, the global fit including the ReD dataset was also performed using the screening functions of ZBL (orange dashed line) and Molière (purple dashed line).
• Figure 5: Probability density functions of the expected WIMP-induced ionization spectra in DarkSide-50 for WIMP masses of 1.2, 3.5, and 7.0 GeV/$c^2$, shown for three different screening-function models. The $f_q(E_{nr})$ response derived from the ZBL screening function corresponds to the previous fit excluding ReD data DarkSide:2021bnz, while the $f_q(E_{nr})$ curve based on the Molière screening function is obtained in this work. In the upper panel, fluctuations in the nuclear recoil quenching process (NQ) are neglected, whereas in the lower panel, quenching fluctuations are described using a binomial distribution (QF).