Probing For Non-Gravitational Long-Range Dark Matter Interactions

TL;DR

The study addresses whether non-gravitational, long-range interactions between dark matter and ordinary matter exist by using a precision rotating torsion balance to search for differential accelerations toward the Galactic center. A Be-Al dipole pendulum interrogated via a turntable-demodulated signal at frequencies $\omega_{TT}$ and $\omega_0$ yields a bound on $\eta_{DM,Be-Al}$ of $2.4\times10^{-5}$ (95% CL), improving prior limits by about a factor of four. No evidence for non-gravitational DM interactions is found, implying dark matter interacts gravitationally over long distances within current sensitivity and constraining models with differential Aluminum-Beryllium couplings. These results constrain B-L charged DM scenarios and provide a laboratory-scale test of the Equivalence Principle for dark matter at galactic scales.

Abstract

Dark matter remains a mystery in fundamental physics. The only evidence for dark matter's existence is from gravitational interactions. We constructed a precision torsion balance experiment to search for non-gravitational, long-range interactions between ordinary matter in our lab and the Milky Way's dark matter. We find no evidence of such interaction and set strict upper bounds on its strength. These results suggest that dark matter only interacts gravitationally over long distances and constrains a variety of dark matter theories.

Figures (8)

• Figure 1: A schematic of the experimental setup in the galactic frame. The experiment searches for differential accelerations of two different materials towards the dark matter in the Milky Way galaxy. Adapted from ESA/Gaia/DPAC, Stefan Payne-Wardenaar. milkyway
• Figure 2: Schematic of the rotating torsion balance apparatus with embedded picture of the pendulum held before installation. The thermal shields, vacuum chamber, and magnetic shields have been cut away to show the torsion balance held inside. The torsion pendulum has four test bodies made of aluminum and four of beryllium in a dipole orientation. The pendulum is suspended by a quartz torsion fiber attached to a magnetic swing damper whose position can be actuated by the positioning stage. The vacuum chamber, torsion balance, autocollimator, and magnetic shield are suspended from the air-bearing turntable. The rotating tilt sensors are used to level the turntable by actuating the thermal expansion feet. The angle of the pendulum relative to the vacuum chamber is measured by the autocollimator.
• Figure 3: Once per turntable revolution differential acceleration amplitudes along with the corresponding fits to the galactic basis functions. The mean of each data set has been removed. Each two-day long segment was independently fit. Short gaps in the data were due to local construction activity. Data for sidereal days 189 to 285 was taken in 2024, while the remaining data was taken in 2025. The right panel shows the histogram of the differential acceleration amplitudes, the bottom panel shows the orientations of the pendulum with respect to the vacuum chamber over the data run. The zoom shows a two-day long segment and it's corresponding fit to the galactic basis functions.
• Figure 4: Measured accelerations from each two-day long segment towards the galactic center (in-phase) and the orthogonal direction (out-of-phase). The blue indicates the points taken at a pendulum orientation of $0^\circ$ and yellow indicates $180^\circ$. The red cross and circle, respectively, indicate the mean and corresponding uncertainty. The top and right panels show the histograms for the in-phase and out-of-phase accelerations, respectively.
• Figure 5: Misfit-squared, fit to a $\chi^2$ distribution, and 99th-percentile cutoff.