A Search for Ultra-Light Vector Dark Matter with a Rotating Torsion Balance
Ross, Shaw, Gettings et al.
We search for ultra-light vector dark matter interacting with a rotating torsion balance with a baryon minus lepton number composition dipole. Our search spans candidate masses in the ultra-low mass range from 1.3~$\times10^{-22}$ to 1.9~$\times10^{-18}$ eV. We set limits on the coupling strength to baryon minus lepton number for each dark matter candidate reaching a peak sensitivity of $g_{B-L} \leq 9 \times 10^{-26}$.
academic
A Search for Ultra-Light Vector Dark Matter with a Rotating Torsion Balance
This study employs a rotating torsion balance apparatus equipped with baryon number minus lepton number (B-L) component dipoles to search for ultra-light vector dark matter. The research covers candidate particle masses in the ultra-low mass range of 1.3×10⁻²² to 1.9×10⁻¹⁸ eV. The experiment establishes constraints on the coupling strength between each dark matter candidate particle and B-L number, achieving peak sensitivity of gB-L ≤ 9×10⁻²⁶, representing a threefold improvement over previous limits.
Despite widespread evidence for dark matter's existence, all direct detection experiments to date have been unsuccessful. This leaves a vast parameter space for non-conventional dark matter candidates, among which ultra-light dark matter (ULDM) represents an underexplored frontier.
Mass Range: ULDM candidate particles can have masses ranging from 10⁻²² eV to 1 eV, far below traditional WIMP candidates
Classical Field Behavior: If dark matter consists entirely of such particles, the dark matter density observed in the Milky Way would be sufficient for ULDM to behave as a classical field
Characteristic Frequency: ULDM fields possess a characteristic frequency fDM ≃ mDMc²/h, linking particle mass to observable frequencies
First Application of Rotating Torsion Balance to Search for B-L Coupled Vector ULDM: Applies rotational modulation techniques to ULDM searches, covering a mass range of 1.3×10⁻²² to 1.9×10⁻¹⁸ eV
Significant Sensitivity Enhancement: Achieves coupling strength constraints of gB-L ≤ 9×10⁻²⁶ at peak sensitivity, representing approximately 3-fold improvement over previous limits
Extended Detection Range: Through rotational modulation techniques, enables detection of lower-mass dark matter candidates compared to static torsion balance experiments
Long-Term Data Accumulation: Utilizes one full year of data from July 2024 to July 2025, obtaining valid data despite harsh conditions including construction noise
Three-Dimensional Field Direction Analysis: Independently fits ULDM fields in three directions (X, Y, Z) of the Earth-centered coordinate system, providing comprehensive constraints
Input: Angular displacement measurements from the rotating torsion balance, recording the angular displacement of the pendulum relative to the vacuum chamber during turntable rotation
Output: 95% confidence level upper limits on coupling strength gB-L for different mass ULDM candidates
Constraints:
Mass range limited by dataset length (low-mass end) and turntable frequency (high-mass end)
Must exclude signals near the diurnal frequency (4.8×10⁻²⁰ eV) caused by Earth's rotation
Overall Characteristic: Constraints do not exhibit strong frequency dependence
Dominant Factors:
Torsion balance thermal noise at turntable frequency
Data acquisition duration
Diurnal Frequency Peak: Peak appears at 4.8×10⁻²⁰ eV, corresponding to Earth's diurnal frequency; the experiment cannot distinguish dark matter signals at this mass from Earth's rotation modulation
No Signal Evidence: No evidence for ULDM signals found across the entire search mass range
System Stability: Despite severe environmental interference, the experiment maintains good systematic control through data quality control and periodic pendulum flips
No Signal Detection: No evidence for B-L coupled ULDM found in the mass range 1.3×10⁻²² to 1.9×10⁻¹⁸ eV
Constraint Setting: Establishes 95% confidence level upper limits on coupling strength between ULDM and B-L number, with peak sensitivity gB-L < 9×10⁻²⁶
Performance Enhancement: Achieves approximately 3-fold improvement over previous limits, extending the achievable range for torsion balance dark matter searches
Parameter Space Constraint: Constrains candidate particles at the ultra-low end of the possible dark matter mass spectrum
Rotational Modulation Application: Innovatively applies rotational modulation to ULDM searches, breaking through static torsion balance mass lower limits
Three-Dimensional Field Analysis: Independently analyzes ULDM fields in three Earth-centered coordinate directions, providing comprehensive field direction constraints
Graham et al. (2016), Phys. Rev. D 93:075029 - Proposes theoretical framework for direct dark matter detection using accelerometers; provides theoretical foundation for this study
Shaw et al. (2022), Phys. Rev. D 105:042007 - Previous torsion balance ULDM search; direct predecessor to this work
Ross et al. (2025), arXiv:2509.10701 - Study using same apparatus for long-range dark matter interaction search; provides detailed apparatus description
Hui (2021), Ann. Rev. Astron. Astrophys. 59:247 - Review of wave dark matter; provides extensive theoretical background
Frerick et al. (2024), Phys. Lett. B 848:138328 - LISA Pathfinder constraints on general dark photons; provides space experiment comparison
Zimmermann et al. (2025), Phys. Rev. Lett. 134:151001 - Dwarf galaxy observation constraints on dark matter mass lower limit
Overall Assessment: This is a high-quality experimental physics paper making substantive contributions to ultra-light dark matter searches. Despite severe environmental interference, the research team successfully completed one year of data acquisition and obtained meaningful results. The application of rotational modulation technique represents an important innovation, significantly extending torsion balance experimental detection capabilities. The paper's primary value lies in filling a gap in the dark matter parameter space and establishing new benchmarks for future experiments. Recommended future work should focus on improving experimental environment, extending observation time, and exploring ULDM-optimized pendulum designs.