Pith. sign in

REVIEW 2 major objections 4 minor 1 cited by

A 4 K cavity optomechanical accelerometer searched for ultralight vector dark matter; it sees no signal but reaches quantum-limited sensitivity and demonstrates a scalable resonant-detection approach.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-04 16:40 UTC pith:JRHXKBD2

load-bearing objection First cryogenic optomechanical accelerometer UDM search: honest null result with careful statistics; the weak spot is the 300 K chip transfer function extrapolated to 4 K. the 2 major comments →

arxiv 2509.12175 v3 pith:JRHXKBD2 submitted 2025-09-15 hep-ex astro-ph.IMphysics.ins-detphysics.opticsquant-ph

Optomechanical Accelerometer Search for Ultralight Dark Matter

M. Dey Chowdhury , J. P. Manley , C. A. Condos , A. R. Agrawal , D. J. Wilson This is my paper
classification hep-ex astro-ph.IMphysics.ins-detphysics.opticsquant-ph
keywords ultralight dark mattercavity optomechanicsaccelerometervector dark matterB-L couplingquantum-limited sensingresonant detectioncryogenic mechanical resonator
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper reports a first-generation resonant search for ultralight vector dark matter using a cryogenic cavity optomechanical accelerometer. Its central claim is that no dark-matter signal appears, but the detector operates stably at the quantum limit with roughly 10 n g0/sqrt(Hz) acceleration sensitivity near its 39 kHz trampoline resonance, validating a practical path to tabletop vector-dark-matter searches. The result matters because the same dual-membrane architecture—two silicon nitride membranes of different stiffness on a silicon chip mounted to a copper plate—turns material composition differences into a differential acceleration readout for baryon-minus-lepton coupling. The paper argues that with larger test masses, millikelvin temperatures, and multiplexed arrays, this platform can reach coupling strengths competitive with or beyond current equivalence-principle bounds.

Core claim

The paper's central discovery is an experimental demonstration, not a signal: a 12 ng Si3N4 trampoline in a 4 K optical cavity, read out shot-noise-limited and cold-damped by radiation pressure, reaches an acceleration noise floor near 10 n g0/sqrt(Hz) over a ~30 Hz band. Scanning the resonance by photothermal heating, the authors searched a 1 kHz window around 39 kHz for the narrow spectral line a vector dark-matter field would imprint through differential acceleration between the silicon nitride membranes and the copper mounting plate. The matched-filter analysis yields a Gaussian distribution of power estimates and no detection events; the resulting 95% upper bound on B-L coupling, g_B-L

What carries the argument

The load-bearing element is a dual-membrane Fabry-Perot cavity accelerometer: a low-mass Si3N4 'trampoline' acting as the test mass and a stiffer square membrane acting as a reference, vertically integrated on a silicon chip with the chip glued to a copper plate. Because the two materials carry different baryon-minus-lepton charge per nucleon, a vector dark-matter field differentially accelerates the chip relative to the membranes, producing a relative displacement amplified by the trampoline's quality factor of about 60 million. The machinery combines this differential readout with shot-noise-limited displacement sensing, radiation-pressure feedback cooling to damp the mode, and phototherma

Load-bearing premise

The reported limit relies on a chip transfer function measured at 300 K in air being representative of the chip's response at 4 K in vacuum; if the chip's mechanical response changes with temperature or mounting, the correction applied to the acceleration spectra is wrong and the baryon-minus-lepton coupling bound shifts.

What would settle it

Mount a piezoelectric transducer on the 4 K sample stage and measure the chip transfer function under cryogenic vacuum conditions across the 38.2-39.2 kHz search window; if the ratio of effective to base acceleration differs from the room-temperature curve used in the analysis by more than its quoted uncertainty, the reported upper limit on g_B-L must be revised.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • With a larger optimized test mass such as a ~1.5 microgram 'sail', the projected acceleration sensitivity improves roughly thirtyfold at 4 K, and by another factor of about six at 10 mK.
  • Operating at millikelvin temperatures with year-long integration and multiplexed arrays of N sensors could push the baryon-minus-lepton coupling bound toward 1e-25, beyond current torsion-balance limits.
  • Photothermal tuning turns a single 39 kHz resonance into a 25 Hz scan window, roughly 330 independent dark-matter bins, establishing a practical frequency-scanning scheme for resonant detectors.
  • The demonstrated quantum-limited, shot-noise-limited readout around 10 n g0/sqrt(Hz) provides a benchmark for tabletop optomechanical accelerometers in the 1 Hz to 1 MHz ultralight-dark-matter band.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The room-temperature chip-transfer-function correction is the main vulnerability; a cryogenic shaker measurement is the obvious next step and would either confirm or move the reported bound.
  • The same differential-acceleration readout could be repurposed as a compact gravity or inertial sensor in materials science, since it measures composition-dependent forces between dissimilar materials at the nanog level.
  • If the anomalously large photothermal heating rate of about 0.4 K per microwatt is a material property of the Si3N4 membranes, it could be exploited for wider frequency scans; if it is a stray absorption effect, reducing it would lower the thermal noise floor and improve sensitivity.
  • The observed Gaussian null statistics suggest the platform is already limited by thermal and shot noise rather than technical backgrounds, which is a precondition for scaling to arrays with squeezed light.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. The manuscript reports a first-generation resonant search for ultralight vector dark matter using a cryogenic cavity optomechanical accelerometer. A 12 ng Si3N4 trampoline (fm ≈ 39 kHz, Qm ≈ 6×10^7) forms a dual-membrane optical cavity on a Cu base; displacement readout is shot-noise-limited and radiation-pressure feedback cooling is used. Photothermal tuning scans the mechanical resonance over about 25 Hz, and acceleration PSDs are analyzed with a matched-filter estimator. After a look-elsewhere correction, no statistically significant excess is found. A 95% upper limit on g_{B-L} of order 10^-17 is quoted, with and without a correction for the measured chip transfer function. The paper’s stated central claim is the demonstration of stable, quantum-limited operation and the validation of a scalable platform for future searches, not a competitive exclusion limit.

Significance. If the result holds, this is a useful first demonstration of a resonant optomechanical accelerometer UDM search in the ~39 kHz band. The statistical work is careful: noise variances are measured rather than assumed, a 30% excess over ideal averaging is incorporated, Gaussianity of the estimator is checked, the look-elsewhere effect is applied, and a chip transfer function is measured and used to correct the limit. The thermalization evidence (Boltzmann histogram, energy autocorrelation, ringdown Q) is strong, and the null result itself is robust. The main weakness is the environmental extrapolation of the chip transfer function, which directly affects the quantitative g_{B-L} bound but does not threaten the sensitivity demonstration or the null result.

major comments (2)
  1. [SI Sec. III.F (Eq. S38, Fig. S6); main text Fig. 4(c)] The quantitative upper limit in Fig. 4(c) is obtained by correcting the measured acceleration PSD as S_a(ω) → S_a(ω)/r_chip^2(ω), where r_chip is the base-to-chip transfer function measured at 300 K in atmospheric pressure on an auxiliary chip without suspended membranes. This is explicitly an assumption that the chip response is similar at 4 K in vacuum. At cryogenic temperature, the silicon chip’s elastic modes, the epoxy mounting stiffness, and the chip–PZT interface can shift in frequency and quality factor. Since the analysis restricts the search to a 1 kHz window where r_chip ∈ [0.5,1] at 300 K, a cryogenic mode shift could move r_chip outside this range and change the inferred bound by a factor comparable to 1/r_chip (the bound scales as σ_D^{1/2} ∝ 1/r_chip). The paper itself identifies an in-situ cryogenic PZT measurement as a future upgrade, confirming this is an admitted limit
  2. [Main text Eq. (6), Fig. 4(a,b); SI Eq. (S20)] Please reconcile the look-elsewhere threshold calculation. The text states N_b = τ_p Δf = 10^4 with τ_p = 20 s and Δf = 1 kHz, but 20 s × 1000 Hz = 20,000. More importantly, Eq. (S20) with CL = 0.95 and N_b = 10^4 gives a detection threshold of about 4.4σ, whereas Fig. 4 appears to label a threshold of 5.8σ. If a different confidence level or a different definition of N_b is used, that should be stated explicitly. Although the null conclusion is not affected because no event reaches either threshold, the false-positive calibration is a central statistical claim and should be internally consistent.
minor comments (4)
  1. [Abstract and main text] The abstract says a “Bayesian search” is conducted, while the main text correctly describes a frequentist detection threshold combined with a Bayesian upper limit. Please align the wording.
  2. [SI Sec. III.D, III.F, and main text Outlook] Typos: “Dispacement” should be “Displacement”; “emperical” should be “empirical”; “V ector” in the SI heading should be “Vector”; “novel contraints” should be “novel constraints.”
  3. [Main text, Eq. (2) and Fig. 4(c)] The notation “10 n g0/√Hz” is unusual; consider defining ng0 explicitly (1 ng0 = 10^-9 g0) at first use to avoid confusion with “n” as a prefix.
  4. [General] A data-availability or analysis-code statement would strengthen reproducibility. The SI contains a simulated search and detailed derivations, but making the analysis scripts available would be valuable for a null-result paper that emphasizes statistical methodology.

Circularity Check

0 steps flagged

No significant circularity: the measured upper limit is set by noise statistics, not by fitting a dark-matter signal; cited prior work is parameter-free and independent.

full rationale

The central derivation chain is not circular. The UDM signal model and the thermal-noise-equivalent sensitivity estimate in Eq. 2 are taken from the group's earlier peer-reviewed work (Ref. [7]) and from Ref. [12]; these are parameter-free analytic results with stated assumptions, and they do not contain or depend on the present experimental data. The actual search does not fit a dark-matter signal: it measures calibrated acceleration power spectral densities, applies the matched-filter estimator of Eq. 5, and sets the 95% upper limit via the likelihood/posterior construction of Eq. 6 and the analytic approximation in Eq. 7. No fitted constant is renamed as a prediction. The only environmental correction is the chip transfer function r_chip(omega) measured at 300 K in air (SI Sec. III.F, Fig. S6), used as Sa(omega) -> Sa(omega)/r_chip^2(omega). That is an auxiliary, stated assumption that could shift the quantitative g_{B-L} bound if the chip response changes at 4 K, but it is a systematic-uncertainty/measurement-extrapolation concern, not a circular reduction: r_chip is not chosen to reproduce the final limit, and the same analysis with r_chip = 1 (gray curve in Fig. 4c) gives a nearby, independently meaningful result. The thermalization evidence, noise-averaging checks, and shot-noise characterization are self-contained experimental validations. The abundant self-citations are to previously published derivations and methods, which constitute independent support under the reviewing rules and do not raise the circularity score.

Axiom & Free-Parameter Ledger

6 free parameters · 6 axioms · 0 invented entities

The analysis rests on measured device parameters (effective mass from FEM, Q from ringdown, bath temperature from thermistor and cold-damping model, photothermal heating rate from linear fit, empirical variance inflation, room-temperature chip transfer function) rather than on parameters tuned to a DM signal. Theory inputs are the standard vector UDM field model, the B-L differential acceleration, the Derevianko lineshape, and the mechanical dimer transduction model from the group's earlier work. No new particles or entities are introduced.

free parameters (6)
  • Effective mass m = 12 ng
    Effective mass of the fundamental trampoline mode from a finite element model; enters the thermal noise calibration and sensitivity formula Eq. 2.
  • Intrinsic quality factor Q_m = 6x10^7 (measured ringdown)
    Inferred from energy ringdown at 4 K (Fig. 2e); sets thermal acceleration noise and the sensitivity estimate.
  • Loaded damping rate gamma = 2 pi x 1 mHz (autocorrelation fit)
    Extracted from instantaneous power autocorrelation (Fig. S3c); used with the cold damping model to infer bath temperature.
  • Photothermal heating rate dT0/dPout = 0.4 K/uW
    Linear fit of bath temperature versus probe power (Fig. S4b); determines thermal noise during the frequency scan.
  • Empirical variance inflation factor zeta = approximately 1.3
    Measured ratio of acceleration PSD estimator variance to ideal periodogram averaging variance; used in the matched-filter variance.
  • Chip transfer function r_chip = 0.5 to 1, frequency dependent (300 K air measurement)
    Measured room-temperature response (Fig. S6c); applied to correct the acceleration PSD in the UDM search.
axioms (6)
  • domain assumption Vector UDM field model with lineshape GDM from Maxwell-Boltzmann velocity distribution (Eq. S2-S4, main text Eq. 4).
    Standard wave-like dark matter model; assumes local density 0.4 GeV/cm^3, velocity dispersion xi_DM about 10^-3, and coherence Q_DM about 10^6.
  • domain assumption B-L differential acceleration a_DM = g Delta12 a0 with Delta12 about 0.04 for Si3N4/Si versus Cu (Eq. 1, Eq. S11).
    Transduction model from Ref. [7]; presumes the detector behaves as a heterogeneous mechanical dimer with two bodies of different B-L content.
  • domain assumption Mechanical dimer model: two membranes of different stiffness on a rigid chip transduce differential base acceleration into relative displacement (Fig. 1, Ref. [7]).
    The readout measures relative motion between trampoline and square membrane; the model maps this to chip acceleration.
  • domain assumption Room-temperature chip transfer function applies at 4 K in vacuum (SI Sec. III.F).
    The chip response r_chip was measured at 300 K in air and assumed similar at cryogenic temperature; the authors state this and plan a cryogenic measurement.
  • standard math Gaussian likelihood for matched-filter estimator and uniform prior for D >= 0 (Eq. S18, S24).
    Standard statistical assumptions for periodogram-averaged matched filtering; validated empirically by the near-Gaussian histogram in Fig. 4b.
  • standard math Thermal noise model S_a = 4 k_B T omega_m / (m Q_m), and cold damping does not change thermal acceleration noise (Eq. 3, Ref. [28]).
    Standard fluctuation-dissipation result for a mechanical oscillator; used to interpret the measured acceleration noise and derive sensitivity projections.

pith-pipeline@v1.3.0-alltime-deepseek · 24816 in / 23544 out tokens · 264849 ms · 2026-08-04T16:40:12.699633+00:00 · methodology

0 comments
read the original abstract

Cavity optomechanical systems have recently been proposed as detectors for ultralight dark matter, leveraging their ability to cool and probe mechanical oscillators at the quantum limit. Here we present a resonant search for ultralight dark matter using a cavity optomechanical accelerometer. The detector consists of a cryogenic Si$_3$N$_4$-membrane cavity mounted to a 4 K copper plate, with photothermal tuning used to scan its 39 kHz mechanical resonance. Shot-noise-limited displacement readout and radiation-pressure feedback cooling yield an acceleration sensitivity of $\sim 10\;\text{n}g_0/\sqrt{\text{Hz}}$ over 30 Hz near resonance. The detector's material inhomogeneity gives access to direct vector coupling to the dark-matter field. We conduct a resonant search based on matched-filter statistics, yielding upper bounds consistent with thermal noise and above those set by equivalence principle tests. No signal is observed, but the experiment demonstrates stable, quantum-limited operation and validates a scalable approach to resonant detection. With optimized test masses, lower temperature, and multiplexed arrays, the platform offers a path toward competitive constraints on vector-mediated dark-matter interactions.

Figures

Figures reproduced from arXiv: 2509.12175 by A. R. Agrawal, C. A. Condos, D. J. Wilson, J. P. Manley, M. Dey Chowdhury.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Cryogenic operation enabled by vibration isolation. (a) Photos of Bluefors LD-4K cryostat (left) and custom vibration isolation system [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Displacement and acceleration measurements. (a) Calibrated [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Search for B-L UDM. (a) Estimates of UDM acceleration power [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Projected performance of contemporary optomechanical ac [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Searching for Ultralight Dark Matter with MOLeQuTE: a Massive Optically Levitated Quantum Tabletop Experiment

    hep-ph 2025-11 conditional novelty 6.0

    A proposed optically levitated milligram-scale plate sensor could reach the standard quantum limit and probe new parameter space for ultralight B-L vector dark matter.

Reference graph

Works this paper leans on

108 extracted references · 6 linked inside Pith · cited by 1 Pith paper

  1. [1]

    Bertone and T

    G. Bertone and T. M. Tait, A new era in the search for dark matter, Nature 562, 51 (2018)

    2018

  2. [2]

    op- tomechanics

    In conceiving such a system, we first emphasize the challenging acceleration sensitivity re- quirements implied by current UDM constraints in the “op- tomechanics” band, fDM = ωDM/2π ∼ 1 Hz to 1 MHz. For B-L UDM, for example, g < ∼ 10−22 (see Fig. 4) corresponds to an acceleration sensitivity of √ Sa ∼ ga0∆12 √ ωDM/QDM < ∼ ∆12 × 10−11g0/ √ Hz, where g0 = 9...

    arXiv

  3. [4]

    Carney, G

    D. Carney, G. Krnjaic, D. C. Moore, C. A. Regal, G. Afek, S. Bhave, B. Brubaker, T. Corbitt, J. Cripe, N. Crisosto, et al., Mechanical quantum sensing in the search for dark matter, Quantum Sci. Technol. 6, 024002 (2021)

    2021

  4. [5]

    Antypas, A

    D. Antypas, A. Banerjee, C. Bartram, M. Baryakhtar, J. Betz, J. Bollinger, C. Boutan, D. Bowring, D. Budker, D. Carney, et al. , New horizons: Scalar and vector ultralight dark matter, arXiv preprint arXiv:2203.14915 (2022)

    arXiv 2022

  5. [6]

    Arvanitaki, S

    A. Arvanitaki, S. Dimopoulos, and K. V an Tilburg, Sound of dark matter: searching for light scalars with resonant-mass de- tectors, Phys. Rev. Lett. 116, 031102 (2016)

    2016

  6. [7]

    Manley, D

    J. Manley, D. J. Wilson, R. Stump, D. Grin, and S. Singh, Searching for scalar dark matter with compact mechanical res- onators, Phys. Rev. Lett. 124, 151301 (2020)

    2020

  7. [9]

    Hirschel, V

    M. Hirschel, V . V adakkumbatt, N. Baker, F. Schweizer, J. Sankey, S. Singh, and J. Davis, Superfluid helium ultralight dark matter detector, Phys. Rev. D 109, 095011 (2024)

    2024

  8. [10]

    C. G. Baker, W. P . Bowen, P . Cox, M. J. Dolan, M. Goryachev, and G. Harris, Optomechanical dark matter instrument for di- rect detection, Phys. Rev. D 110, 043005 (2024)

    2024

  9. [11]

    Deshpande, A

    T. Deshpande, A. Ionescu, N. Miller, Z. Wang, G. Gabrielse, A. A. Geraci, and T. Kovachy, Demonstration that differential length changes of optical cavities are a sensitive probe for ul- tralight dark matter, arXiv preprint arXiv:2412.20623 (2024)

    arXiv 2024

  10. [12]

    Carney, S

    D. Carney, S. Ghosh, G. Krnjaic, and J. M. Taylor, Proposal for gravitational direct detection of dark matter, Phys. Rev. D 102, 072003 (2020)

    2020

  11. [13]

    Carney, A

    D. Carney, A. Hook, Z. Liu, J. M. Taylor, and Y . Zhao, Ultra- light dark matter detection with mechanical quantum sensors, New J. Phys. 23, 023041 (2021)

    2021

  12. [14]

    Blanco, B

    C. Blanco, B. Elshimy, R. F. Lang, and R. Orlando, Models of ultraheavy dark matter visible to macroscopic mechanical sens- ing arrays, Phys. Rev. D 105, 115031 (2022)

    2022

  13. [16]

    A. J. Brady, C. Gao, R. Harnik, Z. Liu, Z. Zhang, and Q. Zhuang, Entangled sensor-networks for dark-matter searches, PRX Quantum 3, 030333 (2022)

    2022

  14. [17]

    A. J. Brady, X. Chen, Y . Xia, J. Manley, M. Dey Chowd- hury, K. Xiao, Z. Liu, R. Harnik, D. J. Wilson, Z. Zhang, et al. , Entanglement-enhanced optomechanical sensor array with application to dark matter searches, Commun. Phys. 6, 237 (2023)

    2023

  15. [19]

    D. F. Jackson Kimball and K. V an Bibber, The search for ultra- light bosonic dark matter (Springer Nature, 2023)

    2023

  16. [20]

    [ 54–70]

    See Supplemental Material, which includes Refs. [ 54–70]

  17. [21]

    M. D. Chowdhury, A. R. Agrawal, and D. J. Wilson, Membrane-based optomechanical accelerometry, Phys. Rev. 6 Appl. 19, 024011 (2023)

    2023

  18. [22]

    Note that xDM is also phase shifted by 90 ◦ relative aDM. Eq. 2 precedes from more precise linear response treatment [ 7]

  19. [23]

    Here and throughout, we use ‘ ∼‘ to mean ’roughly’ or ’on the order of.’

  20. [24]

    T. A. Wagner, S. Schlamminger, J. H. Gundlach, and E. G. Adelberger, Torsion-balance tests of the weak equivalence prin- ciple, Class. Quantum Grav. 29, 184002 (2012)

    2012

  21. [25]

    A direct B-L UDM search has been made with the EW balance at sub-Hz frequencies [ 71]

    We note that the Eöt-Wash (EW) torsion-balance experi- ment [ 23] and other equivalence-principle tests such as MI- CROSCOPE [34] probe composition-dependent, DM-mediated forces between SM test bodies, but do not access the tempo- ral coherence of the UDM field. A direct B-L UDM search has been made with the EW balance at sub-Hz frequencies [ 71]

  22. [29]

    Pontin, M

    A. Pontin, M. Bonaldi, A. Borrielli, F. S. Cataliotti, F. Marino, G. A. Prodi, E. Serra, and F. Marin, Detection of weak stochas- tic forces in a parametrically stabilized micro-optomechanical system, Phys. Rev. A 89, 023848 (2014)

    2014

  23. [30]

    Gisler, D

    T. Gisler, D. Hälg, V . Dumont, S. Misra, L. Catalini, E. C. Lang- man, A. Schliesser, C. L. Degen, and A. Eichler, Enhancing membrane-based scanning force microscopy through an optical cavity, Phys. Rev. Appl. 22, 044001 (2024)

    2024

  24. [31]

    Cold damping does not improve sensitivity [ 28, 72], but re- duces nonlinearities and mitigates the effect of spectral diffu- sion over the course of the measurement [ 73], especially rele- vant for high-Q oscillators

  25. [32]

    The choice of statistical method reflects differing priorities: fre- quentism naturally incorporates the look-elsewhere effect when controlling false positive rates (detection threshold); Bayesian analysis incorporates prior knowledge and avoids unphysical outcomes when interpreting null results (upper limits) [ 74]

  26. [33]

    H.-K. Guo, K. Riles, F.-W. Y ang, and Y . Zhao, Searching for dark photon dark matter in ligo o1 data, Commun. Phys. 2, 1 (2019)

    2019

  27. [34]

    Abbott, T

    R. Abbott, T. Abbott, F. Acernese, K. Ackley, C. Adams, N. Ad- hikari, R. X. Adhikari, V . Adya, C. Affeldt, D. Agarwal, et al., Constraints on dark photon dark matter using data from ligo’s and virgo’s third observing run, Phys. Rev. D 105, 063030 (2022)

    2022

  28. [35]

    Bergé, P

    J. Bergé, P . Brax, G. Métris, M. Pernot-Borràs, P . Touboul, and J.-P . Uzan, Microscope mission: first constraints on the viola- tion of the weak equivalence principle by a light scalar dilaton, Phys. Rev. Lett. 120, 141101 (2018)

    2018

  29. [36]

    D. W. Amaral, D. G. Uitenbroek, T. H. Oosterkamp, and C. D. Tunnell, First search for ultralight dark matter using a magneti- cally levitated particle, Phys. Rev. Lett. 134, 251001 (2025)

    2025

  30. [37]

    Fileviez Pérez and M

    P . Fileviez Pérez and M. B. Wise, Baryon and lepton number as local gauge symmetries, Phys. Rev. D 82, 011901 (2010)

    2010

  31. [38]

    Bauer, P

    M. Bauer, P . Foldenauer, and J. Jaeckel, Hunting all the hidden photons, J. High Energy Phys. 2018 (7), 1

    2018

  32. [39]

    Agrawal, J

    A. Agrawal, J. Manley, D. Allepuz-Requena, and D. Wilson, Focusing membrane metamirrors for integrated cavity optome- chanics, Optica 11, 1235 (2024)

    2024

  33. [40]

    J. P . Moura, R. A. Norte, J. Guo, C. Schäfermeier, and S. Gröblacher, Centimeter-scale suspended photonic crystal mirrors, Opt. Express 26, 1895 (2018)

    2018

  34. [41]

    Norder, S

    L. Norder, S. Yin, M. H. de Jong, F. Stallone, H. Aydogmus, P . M. Sberna, M. A. Bessa, and R. A. Norte, Pentagonal pho- tonic crystal mirrors: scalable lightsails with enhanced acceler- ation via neural topology optimization, Nat. Commun. 16, 2753 (2025)

    2025

  35. [42]

    N. J. Engelsen, A. Beccari, and T. J. Kippenberg, Ultrahigh- quality-factor micro-and nanomechanical resonators using dis- sipation dilution, Nature Nanotechnology 19, 725 (2024)

    2024

  36. [43]

    Cupertino, D

    A. Cupertino, D. Shin, L. Guo, P . G. Steeneken, M. A. Bessa, and R. A. Norte, Centimeter-scale nanomechanical resonators with low dissipation, Nat. Commun. 15, 4255 (2024)

    2024

  37. [44]

    J. R. Pratt, A. R. Agrawal, C. A. Condos, C. M. Pluchar, S. Schlamminger, and D. J. Wilson, Nanoscale torsional dis- sipation dilution for quantum experiments and precision mea- surement, Phys. Rev. X 13, 011018 (2023)

    2023

  38. [45]

    F. Zhou, Y . Bao, R. Madugani, D. A. Long, J. J. Gorman, and T. W. LeBrun, Broadband thermomechanically limited sensing with an optomechanical accelerometer, Optica 8, 350 (2021)

    2021

  39. [46]

    Bawden, B

    N. Bawden, B. J. Carey, P .-M. Y eo, N. Arora, L. Semen- tilli, V . M. V alenzuela, E. Romero, G. I. Harris, M. Wegener, and W. P . Bowen, Precision optomechanical accelerometer via hybrid test mass integration, arXiv preprint arXiv:2508.16088 (2025)

    Pith/arXiv arXiv 2025

  40. [47]

    Condos, J

    C. Condos, J. Pratt, J. Manley, A. Agrawal, S. Schlamminger, C. Pluchar, and D. Wilson, Ultralow loss torsion micropendula for chipscale gravimetry, Phys. Rev. Lett. 134, 253602 (2025)

    2025

  41. [49]

    D.-C. Shin, T. M. Hayward, D. Fife, R. Menon, and V . Sudhir, Active laser cooling of a centimeter-scale torsional oscillator, Optica 12, 473 (2025)

    2025

  42. [50]

    L. Sun, B. J. Slagmolen, and J. Qin, Differential torsion sensor for direct detection of ultralight vector dark matter, Phys. Rev. D 111, 063064 (2025)

    2025

  43. [51]

    Y . Xia, A. R. Agrawal, C. M. Pluchar, A. J. Brady, Z. Liu, Q. Zhuang, D. J. Wilson, and Z. Zhang, Entanglement- enhanced optomechanical sensing, Nat. Photonics 17, 470 (2023)

    2023

  44. [52]

    C. Ge, D. Dominguez, A. Rubenok, M. Miller, and M. Eichen- field, Towards navigation-grade and deployable optomechani- cal accelerometry, arXiv preprint arXiv:2505.11751 (2025)

    Pith/arXiv arXiv 2025

  45. [53]

    R. N. Clark, B. Puzio, O. M. Green, S. T. Pradyumna, O. Trojak, A. Politi, and J. C. Matthews, Integrated pho- tonics for continuous-variable quantum optics, arXiv preprint arXiv:2506.04771 (2025)

    Pith/arXiv arXiv 2025

  46. [57]

    P . W. Graham, J. Mardon, S. Rajendran, and Y . Zhao, Parametri- cally enhanced hidden photon search, Phys. Rev. D 90, 075017 (2014). 7

    2014

  47. [71]

    E. Shaw, M. Ross, C. Hagedorn, E. Adelberger, and J. Gund- lach, Torsion-balance search for ultralow-mass bosonic dark matter, Phys. Rev. D 105, 042007 (2022)

    2022

  48. [72]

    Vinante, M

    A. Vinante, M. Bonaldi, F. Marin, and J. Zendri, Dissipative feedback does not improve the optimal resolution of incoherent force detection, Nat. Nanotechnol. 8, 470 (2013)

    2013

  49. [73]

    Gavartin, P

    E. Gavartin, P . V erlot, and T. Kippenberg, Reply to’dissipative feedback does not improve the optimal resolution of incoherent force detection’, Nat. Nanotechnol. 8, 692 (2013)

    2013

  50. [74]

    Palken, B

    D. Palken, B. Brubaker, M. Malnou, S. A. Kenany, K. Backes, S. Cahn, Y . Gurevich, S. Lamoreaux, S. Lewis, R. Maruyama, et al. , Improved analysis framework for axion dark matter searches, Phys. Rev. D 101, 123011 (2020)

    2020

  51. [75]

    Xiong, L

    F. Xiong, L. Guo, Z. Chen, A. Du, C. Li, and T. Wu, Achieve- ment of a vacuum-levitated metal mechanical oscillator with ultra-low damping rate at room temperature, Commun. Phys. 8, 80 (2025)

    2025

  52. [76]

    Hofer, R

    J. Hofer, R. Gross, G. Higgins, H. Huebl, O. F. Kieler, R. Kleiner, D. Koelle, P . Schmidt, J. A. Slater, M. Trupke,et al., High-q magnetic levitation and control of superconducting mi- crospheres at millikelvin temperatures, Phys. Rev. Lett. 131, 043603 (2023)

    2023

  53. [77]

    Manley, C

    J. Manley, C. Condos, S. Schlamminger, J. Pratt, D. Wilson, and W. Terrano, Microscale torsion resonators for short-range gravity experiments, Phys. Rev. D 110, 1220 (2024)

    2024

  54. [78]

    Hines, A

    A. Hines, A. Nelson, Y . Zhang, G. V aldes, J. Sanjuan, J. Stoddart, and F. Guzmán, Optomechanical accelerometers for geodesy, Remote Sens. 14, 4389 (2022)

    2022

  55. [79]

    Guzman, L

    F. Guzman, L. M. Kumanchik, R. Spannagel, and C. Braxmaier, Compact fully monolithic optomechanical accelerometer, arXiv preprint arXiv:1811.01049 (2018)

    Pith/arXiv arXiv 2018

  56. [80]

    Nelson, M

    A. Nelson, M. Warrayat, J. Dahn, J. Sanjuan, and F. Guzman, A six-axis optomechanical inertial sensing and navigation sys- tem for future mass change missions, in AGU Fall Meeting Ab- stracts, V ol. 2024 (2024) pp. G11C–3504. END MATTER In Table 1 and Fig. 5, we compare various optomechanical accelerometers and their projected performance as B-L UDM detec...

    2024

  57. [81]

    A 22 mg, 20 cm square Si 3N4 membrane as originally proposed in [ 7]

  58. [82]

    A 0.5 mg, 10 cm version of the membrane in [ 7]

  59. [83]

    The 12 ng Si 3N4 trampoline reported in this work

  60. [84]

    See [ 19] for preliminary experimental results

    A 1 µg, centimeter-scale “sail” trampoline under explo- ration as the second generation device for our experi- ment. See [ 19] for preliminary experimental results

  61. [85]

    The acceleration noise reported in [35] is 100 times higher than projected, highlighting challenging vibration isolation requirements

    The 0 .4 mg permanent magnet levitated in a supercon- ducting trap, used in the first generation POLONAISE UDM experiment [ 35]. The acceleration noise reported in [35] is 100 times higher than projected, highlighting challenging vibration isolation requirements

  62. [86]

    An alternative to superconducting levitated setups, this system enables low eddy current damping in a levitated diamagnetic platform

    A 1 .8 mg diamagnet levitated in a permanent magnet trap, reported in [75]. An alternative to superconducting levitated setups, this system enables low eddy current damping in a levitated diamagnetic platform

  63. [87]

    An advan- tage of this magnetic levitation platform is suppression of dissipation from hysteresis and eddy currents

    A 5 .6 µg superconducting test mass levitated in a static magnetic trap (superconducting coil) [ 76]. An advan- tage of this magnetic levitation platform is suppression of dissipation from hysteresis and eddy currents

  64. [88]

    An advantage of this plat- form is its arrayability and potential enhanced vibration isolation if heterogenously mass-loaded [ 50, 77]

    A 0.3 mg, lithographically defined torsion pendulum re- ported in [ 47], formed by suspending a Si microchip from a Si 3N4 nanoribbon. An advantage of this plat- form is its arrayability and potential enhanced vibration isolation if heterogenously mass-loaded [ 50, 77]

  65. [89]

    A 95 mg, flipchip mass-loaded Si 3N4 trampoline re- cently reported in Ref. [ 46]. An advantage of this plat- form is the potential for heterogenous integration of test masses with high neutron density (e.g., Pt in [ 46])

  66. [90]

    A 10 mg, dielectric-mirror-loaded Si 3N4 trampoline re- ported in Ref. [ 45]. An advantage of this platform is compatibility with integration in to a high finesse cav- ity, extending thermal-noise-limited bandwidth. 8 Test mass m [µg] ωm/2π [Hz] Qm [106] √ Stha [ng0/Hz1/2] 300 K 4 K 10 mK (a) 300 K 4 K 10 mK (a)

  67. [91]

    20 cm membrane [ 7] 2 .2 × 104 2 × 103 103 6 × 10−5

  68. [92]

    10 cm membrane [ 7] 5 .4 × 103 3.8 × 103 103 2 × 10−4

  69. [93]

    Trampoline [this work] 1 .2 × 10−2 39.7 × 103 11 60 60 (b) 6 × 102 28 1 (b)

  70. [94]

    “Sail” resonator [ 19] 1 .2 7 .3 × 103 2 2 70 (c) 50 6 (b) 0.3(c)

  71. [95]

    Levitated permanent magnet [ 35] 4 .3 × 102 27 9 .3(b) 5 × 10−4(b)

  72. [96]

    Levitated diamagnet [ 75] 1 .8 × 103 18 0 .1 1 0 .3 6 × 10−4(b)

  73. [97]

    Levitated superconductor [ 76] 5 .6 240 26 (b) 8 × 10−3(b)

  74. [98]

    Mass-loaded micro-pendulum [ 47] 3 .3 × 102 98 5 25 0 .4 9 × 10−4

  75. [99]

    Mass-loaded trampoline (flipchip) [ 46] 9 .5 × 104 117 2 × 10−3(d) 10−2 0.8 2 × 10−3

  76. [100]

    Mass-loaded trampoline (integrated) [ 45] 1 .0 × 104 9.9 × 103 1 × 10−4(d) 1 100 6 × 10−3

  77. [101]

    Photonic-integrated bulk silicon [ 52] 1 .7 × 104 93 × 103 5 × 10−5(d) 10−2 3 × 102 0.1

  78. [102]

    Bulk fused silica [ 78] 2 .6 × 106 4.7 0 .5 0 .5(b) 2 × 10−3 1 × 10−5(b)

  79. [103]

    Comparison of optomechanical accelerometer test masses and projected thermal noise at cryogenic temperatures

    Bulk fused silica [ 79] 1 .1 × 103 690 0 .5 1 0 .1 5 × 10−4 (a) projection (b) using best measured Qm (c) using Qm from simulation (d) measured in air TABLE I. Comparison of optomechanical accelerometer test masses and projected thermal noise at cryogenic temperatures

  80. [104]

    An advantage of this platform is its compatibility with proposals for cryogenic sensor arrays probed with distributed squeezed light [ 16]

    A 17 mg, photonic-integrated, bulk Si accelerometer re- cently reported in [ 52]. An advantage of this platform is its compatibility with proposals for cryogenic sensor arrays probed with distributed squeezed light [ 16]

Showing first 80 references.