The motivation behind the work lies in one of the deepest puzzles in modern physics. Dark matter is overwhelmingly supported by astrophysical and cosmological evidence, from the rotation curves of galaxies to the anisotropies of the cosmic microwave background and the growth of large-scale structure, yet no experiment has ever identified the particle responsible. As direct detection experiments grow more sensitive, they are approaching what physicists call the neutrino fog — the point at which coherent elastic neutrino–nucleus scattering, an irreducible flux of neutrinos from the Sun and other sources, begins to mimic the signals that dark matter hunters are looking for. Once inside the fog, distinguishing a genuine dark matter signal from neutrino backgrounds becomes extraordinarily difficult.

This looming challenge is precisely why the RES–NOVA strategy emphasizes diversity of targets and technologies. Different nuclear targets respond differently to dark matter particles, because interaction rates depend on nuclear properties such as mass number and spin content. Comparing results across materials — xenon, germanium, silicon, and now lead tungstate — offers a powerful way to disentangle a real signal from backgrounds and to cross-check any future discovery. There is also a deep symmetry in the underlying physics: dark matter scattering and coherent neutrino scattering both produce tiny nuclear recoils, so a detector optimized for one is naturally suited to the other. RES–NOVA is primarily designed to catch low-energy neutrinos from astrophysical sources, including the burst of neutrinos that would accompany a supernova in our galaxy, but the same apparatus can search for dark matter.

The secret ingredient is the lead itself. Archaeological lead — recovered from ancient shipwrecks and Roman artifacts that have sat underwater or underground for centuries — has had enormous time for its radioactive contaminants to decay away. Modern lead contains troublesome quantities of lead-210, a long-lived isotope whose decay products can generate signals in exactly the energy region where rare-event searches operate. Archaeological lead, shielded from cosmic rays for two millennia, shows strongly suppressed levels of lead-210 and its progeny. For a field in which the dominant background is often the intrinsic radioactivity of the detector materials themselves, this is a decisive advantage. The RES–NOVA team grew their PbWO4 crystal from this ancient metal, creating a scintillating, high-density target with a mass number high enough to be sensitive to a wide range of dark matter masses.

The prototype detector was operated in the Ieti dilution refrigerator, a custom-built cryostat housed in Hall C of the INFN’s underground laboratory at Gran Sasso, where roughly 3600 meters of water equivalent of rock overburden suppresses the cosmic-ray muon flux by about six orders of magnitude. The cryostat reaches a base temperature below 7 millikelvin and offers about 500 microwatts of cooling power at 100 millikelvin. The crystal itself, a rectangular bar of 0.7 by 0.7 by 4 cubic centimeters, was held by PTFE clamps inside an oxygen-free high-conductivity copper frame and read out with a neutron transmutation doped germanium thermistor glued to its surface. When a particle deposits energy in the crystal, the tiny temperature rise changes the thermistor’s resistance, producing a heat pulse that is carried out of the cryostat via gold bonding wires and superconducting cabling to room-temperature electronics.

One of the most technically demanding aspects of cryogenic rare-event searches is vibration. The pulse tube refrigerator that pre-cools the cryostat generates low-frequency mechanical vibrations that can propagate through the thermal stages, couple to the sensitive thermal sensors, and degrade the baseline energy resolution — fatal for a detector that must resolve signals near its threshold. The Ieti team decoupled the cold head mechanically, mounted it on an independent support structure, and suspended the detector from the mixing chamber plate on a low-thermal-conductivity structure engineered to block vibration transmission. Remarkably, they characterized the vibrational environment using geophones operated at around 7 millikelvin — the first in-situ vibration measurement inside a working cryogenic detector environment. Previous studies in the literature relied on room-temperature piezoelectric sensors, which miss how thermal contraction, capable of shrinking a cryogenic system by centimeters, alters its mechanical response. The measurements showed vibrational amplitudes more than an order of magnitude lower than in comparable setups in the 1–50 hertz band, reaching the sub-nanometer level.

The data analysis was equally careful. Over a total exposure of 32.4 gram-days, the team processed roughly 780,000 candidate pulse windows through a trigger-less pipeline that never imposes a threshold at acquisition time, allowing the region of interest to be defined afterward without selection bias. Pulses were reconstructed with two independent methods: an optimum filter that maximizes the signal-to-noise ratio assuming a known pulse shape and stationary noise, and a maximum-likelihood fit of each pulse to a detector-response model. Comparing the two amplitudes provided a natural quality cut — events whose two estimates disagreed were rejected as glitches, pile-up, or noise-dominated traces. Varying the consistency tolerance between 5 and 30 percent allowed the team to quantify the systematic uncertainty of this selection. The energy scale was anchored to the 2615 kiloelectronvolt gamma line from thallium-208 and cross-checked against the characteristic 46 kiloelectronvolt signature of lead-210, yielding a sensitivity of 112 microvolts per kiloelectronvolt per milligram-scale detector and a baseline resolution of 234 electronvolts.

The measured energy spectrum between 2.5 and 10 kiloelectronvolts — the dark matter region of interest — was dominated not by the crystal’s own radioactivity but by external backgrounds: bremsstrahlung from bismuth-210 in the non-archaeological lead shielding, ambient gamma radiation from the radon decay chain, and a smaller contribution from environmental neutrons. Monte Carlo simulations based on high-statistics screening of the same crystal confirmed that the intrinsic contamination of the archaeological PbWO4 contributes far less. The analysis threshold of 2.5 kiloelectronvolts was set primarily by the germanium thermistor sensors, which are known to be less sensitive at the lowest energies than the transition edge sensors planned for the final RES–NOVA detectors.

Using the Yellin optimum-interval method — a conservative statistical approach that requires no background model — the team derived 90 percent confidence level upper limits on the dark matter–nucleon scattering cross section as a function of mass, incorporating the nuclear structure of lead-207 and oxygen-17 for spin-dependent interactions and deliberately neglecting tungsten-183, whose lower isotopic abundance makes it a less sensitive target. These are the first constraints ever set with PbWO4 as a target material. The sensitivity, limited by the tiny exposure and the environmental background of this research-and-development facility, does not rival that of xenon-based time-projection chambers. But the projections are striking: the full RES–NOVA demonstrator, with roughly 200 kilograms of target mass and an exposure of 170 kilogram-years, is expected to probe previously unexplored regions of dark matter parameter space, with complementary reach to established technologies.

The broader implications extend well beyond dark matter. The validated analysis pipeline processes data in real time, enabling continuous monitoring of detector performance and — once transition edge sensors are implemented — the ability to identify a collective low-energy excess from a galactic supernova and issue an alert within seconds, allowing telescopes around the world to swing toward the event for multi-messenger follow-up. Future detectors will combine larger target masses, optimized low-background shielding, and superior thermal sensors to push thresholds lower and discrimination sharper. For now, a 13-gram crystal of lead smelted before the fall of Rome, cooled to within seven thousandths of a degree above absolute zero beneath the Apennines, has quietly demonstrated that the oldest metal on Earth may be among the best tools for answering one of physics’ newest questions.

Subject of Research: Direct detection of dark matter using an archaeological-lead PbWO4 cryogenic calorimeter prototype

Article Title: Probing dark matter interactions with a RES–NOVA prototype cryogenic detector

Article References: RES–NOVA Collaboration, Alloni, D., Benato, G., Carniti, P., Cataldo, M., Chen, L., Clemenza, M., Consonni, M., Croci, G., Dafinei, I., Danevich, F. A., de Vecchi, C., Di Martino, D., Di Stefano, E., Ferreiro Iachellini, N., Ferroni, F., Filippini, F., Ghislandi, S., Giachero, A., … Yuan, H. (2026). Probing dark matter interactions with a RES–NOVA prototype cryogenic detector. The European Physical Journal C, 86(9), Article 1117. https://doi.org/10.1140/epjc/s10052-026-16215-9

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16215-9

Keywords: dark matter, cryogenic detector, PbWO4, archaeological lead, RES–NOVA, Gran Sasso, coherent neutrino scattering, radiopurity, germanium thermistor, spin-dependent interactions, spin-independent interactions, neutrino fog