First Sub-MeV Dark Matter Search with the QROCODILE Experiment Using Superconducting Nanowire Single-Photon Detectors

TL;DR

This work introduces QROCODILE, a sub-MeV dark matter search using a microwire-based superconducting nanowire single-photon detector that doubles as both target and sensor, achieving a 0.11 eV energy threshold and sensitivity to DM masses down to ~30 keV. The authors develop a comprehensive rate framework incorporating DM-electron scattering, absorption, and nuclear scattering, leveraging thin-layer geometry to induce anisotropic responses and potential directional sensitivity. They report world-leading constraints from 415 hours of data with 15 events, carefully framing the results within conservative background assumptions and outlining future expansions (NILE) to push thresholds lower and exposures higher. The work demonstrates the viability of quantum-sensor technologies for probing light DM, including electron- and nucleus-coupled channels, and highlights the substantial gains achievable with larger, lower-threshold detector arrays and underground operation.

Abstract

We present the first results from the Quantum Resolution-Optimized Cryogenic Observatory for Dark matter Incident at Low Energy (QROCODILE). The QROCODILE experiment uses a microwire-based superconducting nanowire single-photon detector (SNSPD) as a target and sensor for dark matter scattering and absorption, and is sensitive to energy deposits as low as 0.11 eV. We introduce the experimental configuration and report new world-leading constraints on the interactions of sub-MeV dark matter particles with masses as low as 30 keV. The thin-layer geometry of the system provides anisotropy in the interaction rate, enabling directional sensitivity. In addition, we leverage the coupling between phonons and quasiparticles in the detector to simultaneously constrain interactions with both electrons and nucleons. We discuss the potential for improvements to both the energy threshold and effective volume of the experiment in the coming years.

Figures (4)

• Figure 1: The QROCODILE experiment.Left: Experimental setup: The detector is mounted into an oxygen-free copper sample holder aligned with the light beam of the quantum-cascade laser (QCL). The initial power of the QCL is significantly reduced by the neutral density filter (NDF). The detector is biased via a high-precision and low-noise DC source. A bias T splits the DC and RF signals from the detector. The RF signal is then amplified with a low-noise amplifier (LNA) to readout via the pulse counter. Inset: Schematic description of the detector stack. The SEM image represents the meander-shaped SNSPD, with a scale bar of $\qty{18}{\micro\meter}$. Right: Normalized count rate as a function of the bias current applied to our SNSPD detector under irradiation with 5 and 11 wavelength photons.
• Figure 2: Dark matter results.Top: new constraints on DM scattering with electrons via a light (left) or heavy (center) mediator. Bottom: new constraints on DM scattering with nucleons via a light (left) or heavy (center) mediator. Right: constraints on absorption on electrons of kinetically-mixed dark photon DM. In all panels, green shaded regions indicate the new limits we place using our QROCODILE SNSPD. Blue shaded regions indicate previous SNSPD limits Hochberg:2021yudHadas:2024. Dotted orange curves indicate the projected reach of a 10 megapixel SNSPD array with an exposure of one year and a threshold of 29 in wavelength, corresponding to 43. Dot-dashed curves indicate cross sections in which the experiment is sensitive to the direction of a DM stream, as a proxy for directional sensitivity, assuming that all events originate from DM. Other existing terrestrial limits from Refs. EDELWEISS:2019vjvEDELWEISS:2022kttDarkSide-50:2022qzhDarkSide_2023SuperCDMS:2020ausSuperCDMS:2023sqlCRESST_III_2019SENSEI:2023zdfPandaX_4T_2023LUX:2018akbXENON_2020Barak:2020fqlAmaral:2020rynAguilar-Arevalo:2019wdiEssig:2017kqsAgnes:2018oejXENON:2019gfnAn:2014twaAgnese:2018colAguilar-Arevalo:2019wdiArnaud:2020svbFUNKExperiment:2020ofvBarak:2020fql are shown in shaded gray, with complementary stellar constraints on absorption An:2013yuaAn:2014twaAn:2020bxd appearing in shaded yellow. Upper limits of the scattering panels are determined by estimated atmospheric overburden Emken:2019tni.
• Figure S1: Anisotropic response function.Top left: the linear response function $\chi(\bm{\mathrm{q}}, \omega)$ as a function of energy for three different magnitudes and directions of momentum transfer. The angle $\theta_{\bm{\mathrm{q}}}$ is measured from the normal direction of the detector layer to the direction of $\bm{\mathrm{q}}$. Top center: angular dependence of $\chi$ at fixed energy and momentum transfer. The response function is plotted in ratio to its value at $\theta_{\bm{\mathrm{q}}}=\pi/2$ for ease of comparison across different curves. Top right: the ratio of the response function at its maximum ($\theta_{\bm{\mathrm{q}}}=\pi/2$) to its minimum ($\theta_{\bm{\mathrm{q}}}=0$) at fixed momentum transfer, as a function of energy. Bottom: modulation in the scattering rate for DM at fixed speed as a function of the angle $\theta_{\bm{\mathrm{v}}_{\mathrm{DM}}}$ between the DM velocity and the normal direction of the detector layer.
• Figure S2: Left:Waveforms. The 15 measured waveforms recorded over the DM science run drawn superimposed. Inset: arrival time of each event throughout the run. Right:Energy spectrum. Energy spectrum of the silicon substrate wafers (orange), measured with the Gator low-background germanium spectrometer, compared to the background spectrum (blue). Statistical uncertainties are indicated as transparent bands. The background-subtracted spectrum, normalized to the combined uncertainty in units of standard deviations, is given in red.