{"id":"ab43134f-b8b4-429a-a2a1-87976f971eab","arxiv_id":"1908.04797","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Ultralight dark matter with masses near 10^-8 eV could be detected by milligram-scale quantum-limited optomechanical sensors operating at kHz frequencies, with arrays improving the reach.","lead":"This paper shows that tiny mechanical force sensors, already available in quantum labs, could detect ultralight dark matter by feeling its extremely weak oscillating push. It maps out the sensitivity of milligram-scale optomechanical devices and argues that arrays of them can cover new dark matter mass ranges that other experiments miss.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"","rationale":"","tokens_in":161,"tokens_out":16145,"duration_ms":187547,"concrete_test":"","verdict_should_be":"CONDITIONAL","load_bearing_attack":"","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes using quantum-limited optomechanical and electromechanical force sensors to search for ultralight dark matter candidates that couple coherently to standard-model charges, specifically vector B-L and scalar-neutron couplings. It derives the force sensitivity of a single-sided cavity optomechanical sensor from first principles in Appendix B, obtains the standard quantum limit (SQL) noise PSD, and uses it to project detection reach for three search strategies: fixed laser power, SQL scanning at fixed mechanical frequency, and resonant SQL scanning. The central claim is that milligram-scale sensors operating near the SQL, possibly arranged in arrays, could probe new coupling regimes for dark matter masses below about 10^-8 eV, complementing torsion-balance and atom-interferometer experiments. The paper also discusses how arrays of sensors with different materials enable differential measurements to suppress correlated technical noise such as seismic vibrations.","tokens_in":19269,"tokens_out":20880,"duration_ms":206061,"significance":"If the projections are correct, this is a valuable contribution to the ultralight dark matter detection program. It provides a clean, first-principles derivation of the SQL-limited force sensitivity in the appendices, and it makes explicit, falsifiable predictions with clearly stated experimental parameters (mass, mechanical frequency, damping, temperature, laser power). The paper does not fit any quantities to data; the sensitivity curves follow from standard quantum noise formulas. The array-scaling discussion is a useful addition to the prior accelerometer-based proposals. However, as detailed below, the quantitative reach curves contain a significant error in the treatment of the differential measurement and a smaller inconsistency in the thermal noise floor, so the headline numbers need correction before the claimed reach can be taken at face value.","major_comments":[{"comment":"The reach formula omits the differential acceleration coefficient Δ defined in Eq. (4). The paper explicitly states that the detection method relies on a differential measurement between two materials (see Fig. 1 and the Fig. 4 caption, which fixes Δ≈0.03 for an iron/silicon pair). For such a measurement, the signal force is proportional to the difference in charge numbers, i.e., approximately 2Δ N_g g F0, not N_g g F0. Consequently, the denominators in Eqs. (18) and (20) should contain (2Δ N_g)^2, and the sensitivity curves in Fig. 4 are optimistic by roughly 1/(2Δ)≈15 in coupling. Relatedly, Eq. (4) as written is dimensionally inconsistent: a_s is called a differential acceleration, but F0 has dimensions of force, so a correct expression must involve an additional factor of 1/m_n (or an equivalent definition of F0). This issue is load-bearing for every quantitative projection in the paper and must be corrected and re-plotted.","section":"§IV, Eqs. (18)–(20), and Fig. 4"},{"comment":"The main-text formula for thermal force noise, S_FF^T = γ m_s kT, disagrees with the derivation in Appendix B, Eq. (B22), which gives S_FF^T = 4γ m_s kT (the standard fluctuation-dissipation result for the two-sided spectral density used in the paper). Equation (20) inherits the smaller value, understating the thermal floor by a factor of 4 in PSD and a factor of 2 in coupling reach in the thermal-dominated regime. The authors should reconcile Eq. (14) with Eq. (B22) and update the figures accordingly.","section":"§III A, Eq. (14), and Appendix B, Eq. (B22)"},{"comment":"The projections assume that correlated technical noise, especially seismic noise, is fully suppressed: the text states, 'we assume that these correlated noise sources have been sufficiently controlled so that thermal and measurement-added noise are dominant.' This assumption is load-bearing for the low-frequency part of Fig. 4, where the plotted seismic PSD is many orders of magnitude above the thermal and SQL floors. The manuscript does not provide a quantitative estimate of the required common-mode rejection (in dB) or a concrete experimental scheme demonstrating that such rejection is achievable with existing isolation and differential techniques. If this cannot be established, the reach curves below roughly 10^-13 eV should be omitted or explicitly labeled as contingent on unproven technical noise suppression.","section":"§III C and Fig. 4"}],"minor_comments":[{"comment":"The mechanical susceptibility is written with a damping term -i γ ω_s, whereas the time-domain equation (B9) contains -γ p, which yields -i γ ω in the Fourier domain. The discrepancy is negligible exactly on resonance but can matter for off-resonant scans; please either use the exact expression or state the high-Q approximation explicitly.","section":"Eq. (B18)"},{"comment":"The text says 'due to the viral theorem'; this should be 'virial theorem'.","section":"§II"},{"comment":"There are several typographical errors, including 'persisent' in the Introduction, 'Fundamanetal' in the affiliation, and 'intergration' near Eq. (18). These should be corrected in a revised version.","section":"Throughout"},{"comment":"The definition of Ttot uses a piecewise expression with Tint<Tcoh and Tint>Tcoh. The text explains the scaling, but it would help to state explicitly that the second line arises from incoherently combining N_bins=Tint/Tcoh independent coherent segments, which is a standard but nontrivial statistical step.","section":"Eq. (19)"}],"recommendation":"major_revision","confidential_remarks":"The omission of the Δ factor in Eq. (18) is a serious quantitative error that materially changes the paper's headline sensitivity curves; it should be caught and fixed before publication. The qualitative proposal is still of interest, and the derivation framework in Appendix B is sound, so I recommend major revision rather than rejection. The authors should also be asked to reconcile the thermal-noise normalization between the main text and the appendix."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a look. This is a clean, well-scoped sensitivity study that applies the standard optomechanical force-sensing formalism to ultralight dark matter searches in the kHz-MHz band, a genuine gap between torsion balances and atom interferometers. It does not produce a discovery or a breakthrough formalism; it produces concrete reach projections with transparent scaling.\n\nThe strongest part is the derivations in Appendix B. The SQL noise PSD in (B26) and the reach formula in (20) are internally consistent, the assumptions are stated, and the curves follow from the declared parameters. There is no fitting to data, and the authors are careful to flag the main caveat: correlated technical noise must be suppressed via differential measurements between materials. They don't pretend to have solved that problem.\n\nThe soft spots are modest. The technical-noise assumption is real, but the paper's best reach is at kHz-MHz, where seismic noise is less dominant than at sub-Hz frequencies; the 1/f seismic curve they plot matters at the low-frequency end, not at the peak sensitivity. A more concrete limitation is the laser power needed to reach the SQL: they cap at 1 W and admit that above roughly 10^-8 eV the readout runs above the SQL. They handle this honestly. The 'resonant scan' strategy depends on tuning the mechanical frequency by dynamical stiffening, which is plausible but not demonstrated. These are normal caveats for a proposal paper.\n\nI found the array discussion useful, especially the sqrt(N) improvement with incoherent readout and the point that differential measurements reject equivalence-principle-violating backgrounds. The citation pattern looks fair; they credit Graham et al. and Clerk et al. properly, and the self-citation to [31] is relevant.\n\nWho is this for? Experimentalists in cavity optomechanics and levitated sensors who want near-term DM targets, plus theorists needing a reference for mechanical-sensor reach. It is a solid proposal paper, not a paradigm shift. It deserves a serious referee. I would send it out; minor comments could ask for a quantitative estimate of how well differential measurement suppresses seismic noise at the relevant frequencies, and a sentence on the practicality of scanning the mechanical frequency.","headline":"A clean, well-scoped sensitivity study applying standard optomechanical force sensing to ultralight DM; the assumptions are stated and the reach curves are transparent, so it deserves serious refereeing.","tokens_in":19789,"tokens_out":2699,"would_cite":true,"duration_ms":27184,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["95.35.+d"],"model":"deepseek-v4-flash","headline":"A milligram-scale force sensor at the standard quantum limit could detect ultralight dark matter that couples to ordinary matter, and arrays of such sensors would extend the reach.","keywords":["ultralight dark matter","optomechanics","force sensing","standard quantum limit","equivalence-principle violation","sensor array","B-L vector dark matter","scalar-neutron coupling"],"falsifier":"Measure the force noise of a 1 mg, 1 Hz pendulum in a 10 mK dilution refrigerator, with a silicon mirror referenced against an iron mirror, and compare the differential noise spectrum with the paper's thermal-plus-SQL floor: if the seismic or other correlated background stays above roughly $10^{-20}$ N/$\\sqrt{\\rm Hz}$ near 1 Hz, the claimed exclusion reach at low $m_\\varphi$ is not achievable with that sensor. A null test for the dark-matter signal itself would be a long integration showing no coherent peak at the expected frequency with amplitude above the quoted sensitivity.","tokens_in":19113,"feed_emoji":"⚛️","tokens_out":5375,"duration_ms":51455,"temperature":0.7,"pith_summary":"This paper proposes that quantum-limited mechanical force sensors—small suspended mirrors or levitated objects around a milligram in mass—can directly detect ultralight dark matter fields with masses below about $10^{-8}$ eV. The dark matter acts as a nearly monochromatic, coherent force on every atom in the sensor, and an optomechanical readout at the standard quantum limit can resolve the tiny accelerations this force produces. The authors show that a single such sensor can probe new couplings in vector $B-L$ and scalar-neutron models beyond current torsion-balance limits, and that arrays of $N_{\\rm det}$ sensors improve the reach by at least $\\sqrt{N_{\\rm det}}$, with coherent readout potentially giving the full $N_{\\rm det}$ Heisenberg scaling. This matters because it maps ultralight dark matter detection onto an already-demonstrated metrology platform and identifies the kHz–MHz band as a promising search window.","feed_headline":"Milligram sensors at quantum limit can probe ultralight dark matter","feed_subtitle":"One 1-mg pendulum at the standard quantum limit reaches new couplings below 10^-8 eV, and arrays go further.","key_machinery":"The load-bearing object is the optomechanical force sensor: a high-finesse cavity whose movable mirror is a mechanical oscillator of mass $m_s$, frequency $\\omega_s$, and damping $\\gamma$, monitored by laser light. The paper reduces its sensitivity to a closed-form noise power spectral density $S_{FF} = 4\\gamma m_s kT + S_{FF}^{\\rm M, SQL}(\\omega_\\varphi)$ at the standard quantum limit, with $S_{FF}^{\\rm M, SQL}(\\omega_\\varphi)=2m_s\\sqrt{(\\omega_\\varphi^2-\\omega_s^2)^2+\\gamma^2\\omega_s^2}$, obtained by balancing shot noise against backaction at the target frequency. The dark-matter signal is modeled as a sinusoidal force $F(t)=gN_g F_0\\sin(\\omega_\\varphi t)$ with $F_0\\simeq 10^{-15}$ N and $N_g$ the number of coupled charges in the sensor; equating this signal to the noise floor yields the coupling reach. The array generalization uses the correlated signal across sensors to average down uncorrelated thermal and measurement noise, and the material-dependent coupling enables differential measurements that cancel common-mode backgrounds.","core_discovery":"The paper's central claim is that a mechanical sensor with mass around or below 1 mg, mechanical frequency near 1 Hz, damping near $10^{-6}$ Hz, and temperature near 10 mK, operated with the readout laser tuned to reach the standard quantum limit at each target frequency, could detect or exclude vector $B-L$ and scalar–neutron ultralight dark matter for masses below about $10^{-8}$ eV, corresponding to signal frequencies in the kHz range and below. Using equations (20) and (21), the smallest detectable coupling scales as $g \\sim \\sqrt{S_{FF}}/(N_g F_0 \\sqrt{N_{\\rm det} T_{\\rm tot}})$; because the signal force grows with sensor mass, the sensitivity improves roughly as the inverse square root of total mass. The paper further claims that making the sensor and its reference from different materials turns the dark-matter signal into a differential, equivalence-principle-violating acceleration that can be distinguished from common-mode seismic noise, and that an array of $N_{\\rm det}$ sensors improves the coupling reach by at least $\\sqrt{N_{\\rm det}}$.","pith_inferences":["If the differential-material noise cancellation works as assumed, the same array could be repurposed as a broadband equivalence-principle test, since the technique is essentially an equivalence-principle-violating force search with a variable source frequency.","A two-site or multi-site version of the array with sensors made of different materials could exploit the predicted Earth-size coherence length for $m_\\varphi \\lesssim 10^{-9}$ eV to reject local noise and confirm a common signal.","This points to a concrete near-term engineering target: demonstrate that a cryogenic mg-scale pendulum pair with silicon and iron mirrors achieves a differential noise floor below $10^{-20}$ N/$\\sqrt{\\rm Hz}$ near 1 Hz, which is the decisive step for the proposed reach."],"forward_implications":["A single 1 mg sensor at the SQL can reach unexplored values of the $B-L$ and scalar–neutron couplings for $m_\\varphi \\lesssim 10^{-8}$ eV, as shown in Fig. 4.","An array of $N_{\\rm det}$ independently read sensors improves the coupling reach by at least $\\sqrt{N_{\\rm det}}$; a single coherent readout can approach the $N_{\\rm det}$ Heisenberg scaling.","The natural search band is kHz–MHz signal frequencies, complementing torsion balances and atom interferometers at sub-kHz frequencies.","Resonant operation, $\\omega_s = \\omega_\\varphi$, gives the best sensitivity, and dynamical stiffening is one route to scan the mechanical frequency.","Post-SQL techniques such as squeezed light or backaction evasion would extend the reach beyond the curves plotted here, especially at frequencies above $kT \\sim 1$ GHz."],"supporting_citations":[{"why":"Supplies the key observation that ultralight dark matter creates equivalence-principle-violating accelerations detectable with accelerometers, and the differential-material method.","marker":"[11]"},{"why":"Provides the demonstrated mg-scale pendulum parameters and displacement sensing used for the benchmark sensitivity curves.","marker":"[19]"},{"why":"Supplies the quantum noise, input-output, and optomechanical force-sensing formalism from which the SQL noise PSD is derived.","marker":"[44]"},{"why":"Gives the torsion-balance constraints that the proposed reach must beat in the low-frequency regime.","marker":"[9]"},{"why":"Provides the modern torsion-balance equivalence-principle limits used as the excluded region in Fig. 4.","marker":"[10]"},{"why":"Demonstrates zeptonewton-force sensitivity in levitated nanospheres, supporting the claimed force-sensitivity scale.","marker":"[40]"},{"why":"Underwrites the Heisenberg-limit scaling claim for coherent readout of the array.","marker":"[30]"}],"fun_headline_variants":["Milligram quantum sensors probe dark matter below 10^-8 eV","Quantum-limit sensors reach new ultralight dark matter couplings","Sub-10^-8 eV dark matter within reach of milligram sensors","Milligram sensor arrays boost ultralight dark matter reach"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results assume that correlated technical noise—above all seismic vibration—can be cancelled well enough by differential measurements between different materials that only thermal and quantum measurement noise remain; if that cancellation falls short, the low-frequency reach is overstated.","fun_headline_variants_meta":{"raw":{"variants":["Milligram quantum sensors probe dark matter below 10^-8 eV","Quantum-limit sensors reach new ultralight dark matter couplings","Sub-10^-8 eV dark matter within reach of milligram sensors","Milligram sensor arrays boost ultralight dark matter reach"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001175,"raw_usage":{"total_tokens":4803,"prompt_tokens":838,"completion_tokens":3965,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":454,"completion_tokens_details":{"reasoning_tokens":3892}},"tokens_in":454,"tokens_out":3965,"duration_ms":29810,"temperature":1.0,"reasoning_tokens":3892,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:32:42.825885+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the force noise of a 1 mg, 1 Hz pendulum in a 10 mK dilution refrigerator, with a silicon mirror referenced against an iron mirror, and compare the differential noise spectrum with the paper's thermal-plus-SQL floor: if the seismic or other correlated background stays above roughly $10^{-20}$ N/$\\sqrt{\\rm Hz}$ near 1 Hz, the claimed exclusion reach at low $m_\\varphi$ is not achievable with that sensor. A null test for the dark-matter signal itself would be a long integration showing no coherent peak at the expected frequency with amplitude above the quoted sensitivity.","supporting_citations":[{"cited_title":"Demonstration of displacement sensing of a mg-scale pendulum for mm-and mg-scale gravity measurements,","cited_arxiv_id":null,"evidence_quote":"Provides the demonstrated mg-scale pendulum parameters and displacement sensing used for the benchmark sensitivity curves."},{"cited_title":"Introduction to quantum noise, measurement, and amplification,","cited_arxiv_id":null,"evidence_quote":"Supplies the quantum noise, input-output, and optomechanical force-sensing formalism from which the SQL noise PSD is derived."},{"cited_title":"Torsion balance experiments: A low-energy frontier of particle physics,","cited_arxiv_id":null,"evidence_quote":"Gives the torsion-balance constraints that the proposed reach must beat in the low-frequency regime."},{"cited_title":"Review of Particle Physics,","cited_arxiv_id":null,"evidence_quote":"Provides the modern torsion-balance equivalence-principle limits used as the excluded region in Fig. 4."},{"cited_title":"Optical levitation of 10-ng spheres with nano-g acceleration sensitivity,","cited_arxiv_id":null,"evidence_quote":"Underwrites the Heisenberg-limit scaling claim for coherent readout of the array."}],"review_version":1}