⚛️ phenomenology

First Search for Ultraheavy Dark Matter Using a Magnetically Levitated Particle

The POLONAISE experiment conducted the first search for ultraheavy dark matter using a magnetically levitated milligram-scale ferromagnet, setting leading constraints on dark matter-neutron interactions for masses between $10^6$ and $10^{15}\,\mathrm{GeV}/c^2$ and extending the reach of levitated sensing by seven orders of magnitude.

Original authors: Dennis G. Uitenbroek, Dorian W. P. Amaral, Juehang Qin, Jurriaan Langendorff, Andrew Gingerich, Tjerk H. Oosterkamp, Christopher D. Tunnell

Published 2026-08-24

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Dennis G. Uitenbroek, Dorian W. P. Amaral, Juehang Qin, Jurriaan Langendorff, Andrew Gingerich, Tjerk H. Oosterkamp, Christopher D. Tunnell

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The universe is filled with a mysterious substance called dark matter. Astronomers can see its gravity pulling on stars and galaxies, yet no one has ever directly caught a single particle of it. For decades, scientists focused their search on a specific type of heavy particle, hoping to find it in deep underground labs. But when those searches came up empty, the scientific community began to look elsewhere. They started asking if dark matter could be much heavier than anyone expected, or if it could be made of complex clumps rather than simple particles. This shift in thinking opened the door to a new kind of hunt, one that required sensors capable of feeling the faintest possible push from a passing ghost.

A team of researchers has now taken a significant step in this new direction using a device that floats in mid-air. In a laboratory in the Netherlands, they suspended a tiny, milligram-scale piece of magnet inside a superconducting trap. This magnet, which is roughly the size of a grain of sand, was held in place by magnetic fields and cooled to temperatures near absolute zero. The goal was to watch this floating object for any sudden, tiny jolt that could not be explained by normal vibrations or known forces. If a heavy dark matter particle were to fly past the magnet, it would exert a tiny force, giving the magnet a gentle kick. Because the magnet is so light and the environment is so quiet, even the smallest kick would be visible.

The researchers monitored this floating magnet for over nine days, listening for those kicks with extreme precision. Their instrument was sensitive enough to detect a force as small as 0.07 femtonewtons, a level of sensitivity that allows it to feel impulses far smaller than what previous experiments could catch. During the observation period, the team saw eight distinct events where the magnet seemed to receive a push. However, these pushes were not a clear signal of new physics. Instead, the pattern of these events matched the behavior of random, non-Gaussian noise—essentially, the kind of electronic glitches and environmental vibrations that happen in any sensitive experiment. Because they could not prove these were dark matter, and because they could not perfectly model the background noise to subtract it, the team treated every single one of these eight events as a potential dark matter signal.

By assuming the worst-case scenario—that every one of those eight jolts could have been a dark matter particle—the researchers calculated the limits of what their experiment could rule out. They found that if dark matter exists in the form of ultraheavy particles interacting through a long-range force, it cannot be interacting with neutrons as strongly as their experiment would have detected. Specifically, they excluded a range of interaction strengths for dark matter particles weighing between 400,000 and 800 trillion times the mass of a proton. This result is a major expansion of the search territory. While earlier experiments using light to levitate tiny objects could only look for dark matter up to a certain weight, this new magnetic approach extends the search window by seven orders of magnitude into the realm of the ultraheavy.

The study also looked at a specific model where dark matter is not a single particle but a composite object, like a tiny nugget made of many smaller dark constituents. For these composite candidates, the team set the most stringent limits to date on how often they might collide with ordinary matter. They determined that such particles cannot interact with neutrons more frequently than a rate corresponding to a cross-section of 4.5 times 10 to the negative 28th power square centimeters. This finding is particularly important because it covers a mass range that traditional underground detectors cannot reach, as the heavy, composite nature of these particles would cause them to lose their signal coherence in those other experiments.

Ultimately, the experiment did not find dark matter, but it successfully mapped a new frontier. By proving that a magnetically levitated sensor can operate with the necessary sensitivity and stability for months at a time, the team has demonstrated a viable path forward for hunting the heaviest forms of dark matter. They have shown that the universe might be hiding its most massive particles in a mass range that was previously invisible to us, and they have provided the first concrete boundaries for where those particles might still be found. The search continues, but the tools to find the heaviest dark matter candidates are now firmly in hand.

Technical Summary: First Search for Ultraheavy Dark Matter Using a Magnetically Levitated Particle

Problem and Motivation
Despite overwhelming astrophysical evidence for dark matter (DM), its particle nature remains unknown. While Weakly Interacting Massive Particles (WIMPs) in the GeV–TeV mass range have been the primary focus of direct detection, the absence of confirmed signals has driven interest in Ultraheavy Dark Matter (UHDM), defined here as particles with masses exceeding $\sim 10^4$ GeV/ $c^2$ . Candidates include asymmetric dark matter, quark nuggets, Q-balls, and WIMPzillas.

A specific class of UHDM models posits that these particles interact with Standard Model matter via a long-range, Yukawa-like force mediated by a new light boson ( $\phi$ ). If the mediator is sufficiently light, this interaction can remain coherent across a macroscopic target, producing a force proportional to the number of coupled constituents. This coherence offers a potential detection advantage over conventional nuclear recoil experiments, which suffer from signal suppression due to the loss of coherence for composite or spatially extended DM candidates.

Methodology
The POLONAISE collaboration presents the first experimental search for UHDM using a magnetically levitated particle. The experiment utilizes a milligram-scale ferromagnet (0.356 mg) composed of Nd $_2$ Fe $_{14}$ B cubes and a glass sphere, levitated within a Type-I superconducting tantalum trap.

Detection Principle: The experiment monitors the translational motion of the levitated magnet for momentum kicks ( $q$ ) delivered by passing DM particles. The interaction is modeled as a neutron-philic force, where the DM couples to neutrons via a mediator $\phi$ with mass $m_\phi$ and range $\lambda$ .
Sensitivity: The system operates at cryogenic temperatures with ultra-low mechanical dissipation, achieving a force sensitivity of $0.07 $fN Hz$ ^{-1/2}$. The effective radius of the sensor is $R_{\text{eff}} \approx 260$ $\mu$ m, with a neutron charge number $N_n \approx 1.07 \times 10^{20}$ .
Data Acquisition: Data were recorded between December 2025 and January 2026. To minimize environmental noise, the analysis focused on night-time data (19:00–07:00 local time) and excluded periods of maintenance or calibration. The total livetime was 219.66 hours (9.15 days).
Signal Processing: The magnet's motion is read out via a superconducting pick-up coil coupled to a DC SQUID. The data are demodulated at three translational eigenmode frequencies. The analysis focuses on Mode 1 (51.2 Hz, $Q = 2.25 \times 10^6$ ). Impulses are identified as persistent steps in the complex amplitude of the demodulated signal. A test statistic $T$ , based on a profile-likelihood ratio, is used to distinguish candidates from noise. Candidates are selected with $T > 100$ , corresponding to a signal-to-noise ratio of 10.
Atmospheric Attenuation: The analysis accounts for the attenuation of DM particles as they traverse the atmosphere. For strong couplings, DM particles may scatter with air molecules and lose momentum, potentially decelerating below the detector's threshold. The velocity distribution at the detector ( $f_{\text{det}}$ ) is computed by evolving the standard halo model velocities through the momentum-transfer equation.

Key Contributions and Results
The experiment observed eight candidate impulses in Mode 1 with momenta ranging from 1.52 TeV/ $c$ to 12.8 TeV/ $c$ . These candidates are consistent with instrumental transients and correlate with laboratory activity. Consequently, no discovery is claimed. Instead, the authors treat every candidate as a potential DM event to set conservative upper limits using the optimum-interval method.

Exclusion Limits: The study sets 95% confidence level (CL) upper limits on the DM-neutron coupling $\alpha_n$ $α_{n}$ for DM masses ranging from $4 \times 10^5$ $4 \times 1 0^{5}$ GeV/ $c^2$ $c^{2}$ to $8 \times 10^{14}$ $8 \times 1 0^{14}$ GeV/ $c^2$ $c^{2}$ .
- For a massless mediator, the best constraint excludes couplings as low as $\alpha_n = 3.2 \times 10^{-9}$ at a DM mass of $\approx 5 \times 10^6$ GeV/ $c^2$ .
- Sensitivity extends to mediator masses up to $\sim 30$ meV/ $c^2$ (force range $\lambda \approx 7$ $\mu$ m). For heavier mediators, the signal is suppressed due to the loss of coherence and the exponential Yukawa factor.
Composite Dark Matter Constraints: The results are recast for a concrete composite DM model (spatially extended bound states of fundamental constituents). For a benchmark model with constituent mass $\bar{m}_d = 1$ keV/ $c^2$ and a mediator mass of 10 meV/ $c^2$ , the experiment sets leading constraints on the DM-neutron cross section $\sigma_{\chi n}$ in the mass range $2 \times 10^6$ GeV/ $c^2$ to $4 \times 10^7$ GeV/ $c^2$ , constraining cross sections to $\sigma_{\chi n} \lesssim 4.5 \times 10^{-28}$ cm $^2$ . This improves upon existing bounds by an order of magnitude at specific masses.
Comparison to Previous Work: The results extend the mass reach of levitated sensing experiments by approximately seven orders of magnitude compared to optical levitation strategies (which are limited to $\sim 10^8$ GeV/ $c^2$ due to smaller target masses). This is achieved by utilizing milligram-scale masses, which provide a coherent enhancement factor of $\sim 10^2$ over nanogram-scale targets, despite a higher momentum threshold.

Significance
The paper establishes magnetically levitated sensors as a sensitive new probe for ultraheavy dark matter. By combining a large target mass (milligram scale) with sub-femtonewton force sensitivity and continuous monitoring, the POLONAISE experiment successfully probes a DM mass window ( $10^6$ – $10^{15}$ GeV/ $c^2$ ) that was previously inaccessible to direct detection methods. The results provide the first experimental constraints on neutron-philic UHDM interacting via long-range forces in this regime and demonstrate the viability of macroscopic force sensors for detecting composite dark matter candidates that would otherwise be invisible to nuclear recoil experiments due to coherence loss.

Technical Summary: First Search for Ultraheavy Dark Matter Using a Magnetically Levitated Particle

More like this