{"id":"a33f90d5-6873-42e3-9cb7-d341043a735e","arxiv_id":"2607.18740","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Modulated AC auxiliary fields and quadratic Stark mixing let Rydberg atoms sense 2.5–5 kHz electric fields without internal electrodes, with measured sensitivity 7.5 μV cm⁻¹ Hz^-1/2 at 5 kHz.","lead":"Rydberg atoms can detect weak low-frequency electric fields by mixing them with a stronger alternating auxiliary field, using the atoms' quadratic Stark shift. The demonstrated scheme avoids internal electrodes and reaches 7.5 μV/cm/√Hz at 5 kHz externally (0.52 μV/cm/√Hz after correcting for cell-wall screening).","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Corrected sensitivity claims rest on a single unvalidated 6.9% cell-screening factor; if this calibration is biased, the headline 0.52 μV/(cm·Hz^{1/2}) and 18 nV/cm values shift by the same factor, with no propagated uncertainty.","rationale":"The reader's weakest assumption—the cell-screening calibration—is precisely the most load-bearing concern for the paper's headline corrected numbers. The QFM algebra in Eq. (3) is correct, and the external sensitivity is measured, so the core mechanism is not in doubt. The risk is that the intrinsic sensitivity values, which are emphasized in the abstract and conclusion, depend entirely on a single factor η=6.9% with no error bar and no in-manuscript verification. If η is wrong, the corrected numbers are wrong by the same factor, which is a direct, quantitative impact on the central performance claim. The missing supplement and the group's prior readout model are secondary: the transduction coefficient k is fitted, so the readout model's absolute coefficients do not affect the sensitivity calculation. The screening calibration, however, is an external-to-internal conversion that cannot be absorbed into k for the intrinsic claims. Therefore the reader's CONDITIONAL verdict is appropriate, and an independent calibration would settle the concern. I see no reason to change the verdict; the concern is already captured.","tokens_in":11780,"tokens_out":10347,"duration_ms":126751,"concrete_test":"Independently determine the cell's 5 kHz screening factor η using two methods: (i) reproduce the Ref. [53] time-averaged Stark-shift calibration, and (ii) use a second cell with calibrated internal electrodes (or a directly embedded electrode) to generate a known 5 kHz field at the atoms and compare the QFM signal to that from the external plate field, at the same E_a=0.32 V/cm operating point. If the two η values disagree by more than ~20%, the corrected sensitivity (0.52 μV/(cm·Hz^{1/2})) and E_min (18 nV/cm) must be revised and quoted with an uncertainty propagated from η. Also repeat the calibration after extended operation to check drift of surface screening.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central algebraic claim, Eq. (3), is sound: the quadratic Stark shift of a two-tone field contains a cross term linear in E_s and amplified by E_a, and the external sensitivity 7.5±2.6 μV/(cm·Hz^{1/2}) at 5 kHz is an experimental result. However, the paper's headline intrinsic performance is obtained by scaling the external numbers by a single screening factor η=6.9% (Fig. 3d), measured with a procedure cited to Ref. [53] and reported without an uncertainty. The corrected values are linearly proportional to η: 0.52 = 7.5×0.069 and 18 nV/cm = 0.26×0.069. If the screening calibration is biased—from stray DC fields shifting the time-averaged Stark-shift baseline, nonuniform plate fields, an unverified frequency-dependent transmission model, or slow changes in surface adsorbates—every corrected number in the abstract shifts by the same factor. The external calibration k=1.698 cm does not protect against this, because k is defined relative to the externally applied plate field; converting to 'field seen by the atoms inside the cell' necessarily requires η. The calibration details and the reconstruction/MRC procedures are relegated to the unavailable supplement [44], so the correction cannot be independently checked from the manuscript. This does not undermine the existence of the QFM effect or the external sensitivity, but it means the most impressive corrected performance claims are not yet established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes and experimentally demonstrates a low-frequency electric-field sensing scheme for Rydberg atoms based on a modulated AC auxiliary field (quantum frequency mixing, QFM). The core idea is Eq. (2)-(3): when a weak signal field E_s cos(ω_s t + φ_s) is superimposed on an auxiliary field E_a cos(ω_a t + φ_a), the quadratic Stark shift contains a cross term δ_cr = -α E_a E_s [cos(ω_+ t + φ_+) + cos(ω_- t + φ_-)]/2 that is linear in E_s and amplified by E_a. This shift is read out through EIT with a weak-measurement polarization postselection, and the signal is recovered by demodulating at ω_a with an optimal phase. The authors verify the E_a^2 and E_a scalings in Fig. 2, report a linear transduction k = 1.698 cm (Fig. 3a), an external sensitivity of 7.5 ± 2.6 μV/(cm·Hz^{1/2}) at 5 kHz, and a minimum detectable field of 0.26 ± 0.04 μV/cm at 1000 s. Using a measured cell-screening factor η = 6.9% at 5 kHz, they quote corrected intrinsic values of 0.52 μV/(cm·Hz^{1/2}) and 18 nV/cm. Multi-frequency auxiliary fields are demonstrated with maximal ratio combining (MRC).","tokens_in":12069,"tokens_out":8494,"duration_ms":81275,"significance":"The central algebra of Eq. (3) is exact, and the scaling checks in Fig. 2, together with the linear transduction in Fig. 3(a), provide strong evidence that the proposed QFM mechanism works. The external sensitivity numbers are an experimental result and, if reproducible, are already useful for LF Rydberg electrometry. The significance of the paper, however, depends in part on the corrected intrinsic numbers, which are obtained by a single screening factor with no propagated uncertainty and with calibration details deferred to the supplemental reference [44]. With those details supplied, this would be a solid contribution; in its current form the strongest performance claims are not self-contained.","major_comments":[{"comment":"The quoted intrinsic sensitivity 0.52 μV/(cm·Hz^{1/2}) and minimum detectable field 18 nV/cm are obtained by multiplying the externally calibrated values by η = 6.9% (Fig. 3d). The screening calibration is described only as 'a procedure similar to Ref. [53]' and deferred to [44]; no uncertainty or independent validation is given. Since all corrected numbers are exactly proportional to η, any bias (stray DC fields, nonuniform plate fields, frequency-dependent transmission, or surface-adsorbate drift) changes them by the same factor. The calibrated transduction coefficient k = 1.698 cm does not constrain η because k is defined relative to the externally applied plate field. Please provide a self-contained calibration with propagated uncertainty, or present the external values as the headline performance and treat the η-corrected numbers as illustrative.","section":"Abstract and the paragraph after Fig. 3(d)"},{"comment":"Several load-bearing procedures are only in the supplemental reference [44]: the reconstruction algorithm and analytical phase θ_opt, the screening calibration, the amplitude/phase optimization and MRC combination for the multi-frequency data, and the FM-compensation scheme. As the supplement is not available in this manuscript, the reader cannot independently verify Eq. (1), the phase formula, or the SNRs in Fig. 4. Please include the essential formulas, definitions (SNR, integration time, noise floor), and calibration geometry in the main text or an accessible supplement.","section":"Signal reconstruction, screening calibration, and multi-frequency methods"},{"comment":"The claim that MRC 'consistently yields a higher SNR' is based on five measurements with amplitudes and phases optimized '[44]'. If the optimization and the evaluation are performed on the same data, the comparison can be optimistic. Please state the SNR definition, whether the parameters were selected on separate training data, and give a statistical significance test or cross-validated result. This is secondary to the single-frequency demonstration but is load-bearing for the generalized 'systematic framework' claim.","section":"Fig. 4 and the multi-frequency paragraph"}],"minor_comments":[{"comment":"Please clarify the units of δ (angular frequency vs. Hz) and the derivation/dimensions of β. As written, β is defined through an expression with A and Γ_p but no derivation; this makes the transfer function hard to check.","section":"Eq. (1) and the definition of β"},{"comment":"The text says the T^{-0.48} exponent agrees with 'quantum-noise-limited scaling' T^{-0.5}, but also states the system is 'well above the standard quantum limit'. This wording is misleading; 'shot-noise-like 1/√T scaling' would be more accurate.","section":"Fig. 3(b)"},{"comment":"The paper states that the transmission readout 'exhibits the same qualitative behavior' and is substantially worse, but does not clearly point to a quantitative comparison. Please indicate which panels support this statement.","section":"Conventional transmission configuration"},{"comment":"Ref. [50] is an arXiv preprint from the same group; if a published version exists, please cite it. Also, the supplemental material [44] currently has no URL.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main concern for the editor is that the paper's headline intrinsic sensitivity numbers are not self-contained: a single screening factor without uncertainty, and multiple key procedures deferred to [44]. The effect itself appears real and the external sensitivity is credible, so I recommend major revision rather than rejection. I would also note that the novelty over Ref. [50] should be made clearer: the weak-measurement readout is taken from that work, and the new element is the AC auxiliary-field QFM. The screening calibration issue should be resolved before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nIf you work on Rydberg electrometry, this one is worth a look. The idea: replace the usual DC bias field with an AC auxiliary field and exploit the quadratic Stark cross term to get a signal linear in E_s and amplified by E_a — essentially quantum frequency mixing transplanted from NV centers to Rydberg LF sensing. The central algebra is exact, and the experiments confirm the expected scalings (quadratic in E_a at 2ω_a, linear at ω_-). The external sensitivity of 7.5 ± 2.6 μV/(cm·√Hz) at 5 kHz and 0.26 μV/cm at 1000 s is a solid, reproducible-looking result, and the multi-tone MRC extension is a nice bonus. The phase-preserving readout that uses both sidebands is a genuine improvement over the standard heterodyne loss.\n\nThe soft spots are not in the physics but in the performance layer. The corrected 'intrinsic' numbers — 0.52 μV/(cm·√Hz) and 18 nV/cm — are obtained by dividing the external numbers by a cell-screening factor η = 6.9%, measured with a procedure cited to Ref. [53] and reported without an uncertainty. If that calibration is off, the corrected numbers shift by the same factor. Since the abstract reports only the external numbers, the paper doesn't overclaim at the headline, but the summary touts the intrinsic values as the achievement. That needs a caveat and a full error budget. Also, the reconstruction algorithm, the optimal phase formula, and the MRC details are all in the supplement, which isn't included in the arXiv version; an independent check is impossible without it. The readout model is imported from the same group's weak-measurement paper, which is acceptable but does mean the transduction coefficient k carries some model dependence.\n\nThe paper is honest about its limits: it explicitly says the performance is above the SQL and identifies the EIT bandwidth as the bottleneck. I see no fabricated results; the fitting and scaling checks look consistent.\n\nBottom line: this is a solid, useful contribution to the Rydberg-sensing subfield. It deserves a serious referee. I'd push the authors to either move the screening calibration details into the main text or at least provide a propagation of uncertainty for η, and to make the supplement material available. If the screening factor holds up, the intrinsic sensitivity claim is meaningful. If not, the external numbers still stand. Worth engaging.","headline":"A clean QFM demonstration for low-frequency Rydberg sensing; the external sensitivity holds up, but the corrected 'intrinsic' numbers rest on a single screening factor with no propagated uncertainty.","tokens_in":12660,"tokens_out":2472,"would_cite":true,"duration_ms":21804,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"An AC auxiliary field, not a DC bias, is what this paper uses to make Rydberg atoms sense low-frequency electric fields: the quadratic Stark shift produces a cross term linear in the signal amplitude and amplified by the auxiliary field, yi","keywords":["Rydberg atoms","quantum sensing","low-frequency electric field","Stark shift","quantum frequency mixing","heterodyne detection","weak measurement","electromagnetically induced transparency"],"falsifier":"Apply a precisely known 5 kHz field to the cell using an independent, traceable method (e.g., a calibrated loop or a second cell with internal electrodes) and compare the recovered amplitude to the QFM prediction; a disagreement beyond the quoted errors would invalidate the screening correction. Alternatively, measure the reconstructed signal's phase at several signal frequencies and check that the recovered phase is consistent with the known drive phase after the constant offset.","tokens_in":11594,"feed_emoji":"⚛️","tokens_out":3941,"duration_ms":34832,"temperature":0.7,"pith_summary":"This paper proposes and demonstrates a method for sensing low-frequency electric fields with Rydberg atoms by replacing the usual constant bias field with a modulated AC auxiliary field. The AC field mixes with the incoming signal through the atoms' quadratic Stark shift, producing a cross term that is linear in the signal amplitude and amplified by the auxiliary field's amplitude. This effectively transfers the low-frequency information to higher-frequency sidebands, where it can be read out with a weak-measurement-enhanced optical probe. The authors report a sensitivity of 7.5 μV/(cm·Hz^1/2) at 5 kHz and a minimal detectable field of 0.26 μV/cm after 1000 s, with intrinsic (cell-corrected) values of 0.52 μV/(cm·Hz^1/2) and 18 nV/cm. They further show that using multiple auxiliary frequency components and combining the reconstructed signals improves sensitivity.","feed_headline":"AC mixing lets Rydberg atoms sense 18 nV/cm fields","feed_subtitle":"A modulated AC bias replaces fragile DC electrodes, making low-frequency electric-field sensors stable and calibration-friendly.","key_machinery":"The load-bearing object is Eq. (3), the QFM decomposition of the quadratic Stark shift into a quasi-static term, a second-harmonic term, and the cross term δ_cr. That cross term is a product of the auxiliary and signal amplitudes, making it a linear amplifier for the signal whose gain is set by the locally generated auxiliary field. The readout is a weak-measurement protocol: the Stark shift is mapped onto the polarization rotation of a probe laser via EIT, and a post-selection polarizer at angle -π/4 + ε converts the shift into an intensity change with technical-noise suppression. The reconstruction algorithm uses lock-in demodulation at ω_a and an optimal quadrature combination to recover","core_discovery":"The central discovery is the quantum frequency mixing (QFM) identity for Rydberg Stark shifts: when the total field is the sum of an auxiliary AC field E_a cos(ω_a t+φ_a) and a weak signal E_s cos(ω_s t+φ_s), the quadratic Stark shift δ = -αE^2/2 decomposes into a cross term δ_cr = -α E_a E_s [cos(ω_+ t + φ_+) + cos(ω_- t + φ_-)]/2, with ω_± = ω_a ± ω_s. This term is linear in E_s and scaled by E_a, so the auxiliary field acts as a parametric amplifier that imprints the amplitude and phase of the signal onto sidebands of the shift. The paper shows experimentally that demodulating the EIT-encoded probe at the sideband frequency and optimally recombining the quadratures recovers the original s","pith_inferences":["If an independent, traceable calibration of the cell screening factor replaced the current 6.9% measurement, the quoted intrinsic sensitivity could be directly tested; the external calibration constant k = 1.698 cm is already robust, so the method's core linearity claim does not hinge on the screening correction.","Because the cross term is linear in the signal, the demodulated stream at ω_a can in principle support simultaneous broadband reconstruction of many low-frequency components, resembling a software-defined atomic receiver.","The same modulated-bias, quadratic-coupling strategy could be adapted to sense magnetic or mechanical fields that couple to Rydberg levels through a quadratic term, using an appropriate AC auxiliary field.","The preservation of both sidebands, unlike standard heterodyne mixing that discards one, suggests a general 3 dB advantage for any quadratic sensor that reconstructs the waveform directly at the signal frequency."],"forward_implications":["The method yields a calibration-friendly low-frequency electric-field sensor that needs no intra-cell electrodes, avoiding the drift and instability of DC bias fields.","Sensitivity can be raised by increasing the auxiliary amplitude up to the EIT bandwidth limit; frequency-modulating the coupling laser pushes that boundary further.","Multi-tone auxiliary fields allow frequency-diverse sensing and SNR gain through maximal ratio combining.","The recovered waveform preserves both amplitude and phase, enabling coherent detection of low-frequency signals.","The screening-corrected intrinsic numbers (0.52 μV/(cm·Hz^1/2) and 18 nV/cm) suggest nV/cm-level sensing is reachable in cells with less field attenuation."],"fun_headline_variants":["Rydberg AC mixing senses 260 nV/cm fields","AC auxiliary field stabilizes Rydberg E-field sensing","Quantum frequency mixing boosts Rydberg sensor sensitivity","Modulated AC field replaces DC bias in Rydberg receivers","Rydberg atoms detect 0.26 μV/cm via AC sidebands"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The quoted intrinsic sensitivity and minimal detectable field assume that the measured 6.9% cell screening factor at 5 kHz correctly represents the fraction of an external field that reaches the atoms; if that calibration is biased, all corrected numbers shift by the same factor.","fun_headline_variants_meta":{"raw":{"variants":["Rydberg AC mixing senses 260 nV/cm fields","AC auxiliary field stabilizes Rydberg E-field sensing","Quantum frequency mixing boosts Rydberg sensor sensitivity","Modulated AC field replaces DC bias in Rydberg receivers","Rydberg atoms detect 0.26 μV/cm via AC sidebands"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00085,"raw_usage":{"total_tokens":3610,"prompt_tokens":898,"completion_tokens":2712,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":642,"completion_tokens_details":{"reasoning_tokens":2626}},"tokens_in":642,"tokens_out":2712,"duration_ms":18721,"temperature":1.0,"reasoning_tokens":2626,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T14:29:55.899510+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Apply a precisely known 5 kHz field to the cell using an independent, traceable method (e.g., a calibrated loop or a second cell with internal electrodes) and compare the recovered amplitude to the QFM prediction; a disagreement beyond the quoted errors would invalidate the screening correction. Alternatively, measure the reconstructed signal's phase at several signal frequencies and check that the recovered phase is consistent with the known drive phase after the constant offset.","supporting_citations":[],"review_version":1}