{"id":"230db22b-16e7-4ed4-bdab-33eff8ee7337","arxiv_id":"2505.03459","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A modified CPT method that scans the magnetic field instead of the RF frequency allows simultaneous observation of zero-field population redistribution and non-degenerate quantum interference, and enables in-situ three-axis coil calibration.","lead":"This paper demonstrates an experimental method where CPT-like resonances in rubidium vapor are acquired by scanning the magnetic field while keeping the laser frequency separation fixed. This allows simultaneous observation of zero-field population redistribution and high-field quantum interference, and provides a convenient way to calibrate three-axis magnetic coils.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The key separation claim (RF-demodulated CPT contains only interference) relies on an unproven assumption that the zero-field population-redistribution signal is independent of ν_RF; a residual B-dependent baseline could bias coil calibration.","rationale":"The reader's weakest_assumption identifies the same load-bearing point: the paper explicitly relies on the RF-frequency independence of the zero-field population-redistribution signal to justify using the RF-demodulated CPT signal as a clean interference spectrum. This assumption is essential because the subsequent calibration procedure and the claim of circumventing the overlap rely on that clean separation. The concern is concrete: the amplitude of the zero-field optical-pumping feature depends on the detuning of the sidebands, and the sideband separation is proportional to ν_RF, so the RF modulation can in principle convert that dependence into a spurious lock-in signal. The paper provides no derivation, error estimate, or control experiment for this point. The experimental consistency between RF-scanning and magnetic-scanning calibration (Tab. 1) is encouraging empirical evidence, but it does not directly test the RF-independence assumption; the same calibration factors could be recovered if the residual contamination is small. Therefore, the appropriate verdict remains CONDITIONAL: the method is plausible and likely correct, but the central separation claim should be supported by a direct measurement or calculation. I also note a smaller, correctable issue: Eq. 1 gives Ω_L = g_F μ_B B without an explicit factor of ℏ or h, making the equation dimensionally inconsistent as written; although the numerical calibration evidently used the correct atomic constants, the equation should be fixed. This issue does not change the verdict because the empirical agreement supports the method, but it adds to the need for careful revision.","tokens_in":8017,"tokens_out":11429,"duration_ms":126745,"concrete_test":"Set δ_RF to a value such that no nΩ_L resonance lies within the scanned B range (e.g., large |δ_RF|, placing the ±1 resonances outside the scan window), then record the RF-demodulated CPT signal while scanning B through zero. If the output shows a broad feature matching the M_M zero-field resonance, the population-redistribution signal is not independent of ν_RF and the separation claim is falsified. Quantitatively, compare the residual CPT amplitude near B=0 with the M_M signal amplitude; if the residual exceeds a few percent of the narrow resonance heights, the calibration bias becomes non-negligible.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central advantage claimed for the modified method is that the RF-demodulated CPT signal is a clean quantum-interference spectrum, uncontaminated by the zero-field population-redistribution signal. The paper states this in §III after Fig. 2(c): 'Since the population redistribution signal is not dependent on ν_RF, it doesn’t contribute to the demodulated signal.' No derivation or control experiment is provided. The concern is not merely formal: the zero-field resonance arises from optical pumping and Hanle-type coherence among degenerate Zeeman states, and its amplitude and width depend on the pumping rates of both ground hyperfine levels. Those rates depend on the detuning of the two laser sidebands, which is set by 2ν_RF. Thus, when ν_RF is modulated at 440 Hz, the population-redistribution signal can acquire a component at the modulation frequency. Even if small, this component would appear as a B-dependent baseline in the CPT signal, shifting the apparent zero crossings of the narrow CPT resonances and biasing the coil-calibration factors in Tab. 1. The observed clean resonances suggest the residual is small in the present conditions, but the independence is asserted rather than established, and the calibration claim depends on it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript describes an experimental study in which coherent population trapping (CPT) resonances in 85Rb vapor are acquired by scanning the magnetic field at a fixed RF detuning, rather than the conventional RF scan at a fixed magnetic field. The authors observe a zero-field population-redistribution signal and quantum-interference resonances between non-degenerate Zeeman states in both a buffer-gas-filled and an anti-relaxation-coated cell. They propose the method for in-situ calibration of three-axis magnetic coils using the standard two-photon resonance condition, and they compare calibration factors with the conventional RF-scanning method, reporting close agreement.","tokens_in":8309,"tokens_out":4005,"duration_ms":39429,"significance":"If the central separation claim is correct (that the RF-demodulated CPT signal contains only quantum-interference resonances), the method offers a convenient atom-based calibration of magnetic coils using fundamental atomic constants, together with simultaneous acquisition of zero-field and high-field resonances. The use of the standard resonance condition, the demonstrated consistency between RF-scan and B-scan calibration factors, and the operation in two complementary cell types are strengths. With proper uncertainty analysis, this would be a useful experimental tool for developing vector magnetometers with large dynamic range.","major_comments":[{"comment":"The claim that the RF-demodulated CPT signal is a clean quantum-interference spectrum rests on the statement that the population-redistribution signal is independent of ν_RF. No derivation or control experiment is given. Since the zero-field redistribution signal depends on optical pumping rates, which in turn depend on the detuning of the two laser sidebands (set by 2ν_RF), RF modulation at 440 Hz could in principle produce a B-dependent baseline in the demodulated signal. Such a baseline would shift apparent zero crossings and bias the coil-calibration factors in Tab. 1. Please provide an experimental test (e.g., a scan far from CPT resonance or with RF modulation disabled) or a quantitative estimate showing that the residual is negligible within the claimed calibration accuracy.","section":"§III (after Fig. 2(c))"},{"comment":"No uncertainties are reported for the measured bias fields or the calibration factors. The agreement between the RF-scanning and magnetic-scanning calibration factors (e.g., 0.6 vs 0.6, 0.6 vs 0.5, 3.1 vs 3.1 µT/V) cannot be assessed without error bars. Because the central application is accurate, atom-based coil calibration, please report uncertainties that include line-position fitting error, magnetic-field modulation amplitude, lock-in phase error, and any systematic from the assumed independence of the population-redistribution signal.","section":"Tab. 1 and Tab. 2"},{"comment":"The explanation that the MM-signal polarity is preserved against the sign of nΩ_L because of 'an additional change in the phase of the energy level oscillation (arising from the magnetic field modulation)' is asserted without derivation or reference. This is a nontrivial physical claim and is used to interpret the MM signals in Figs. 3 and 4. Please provide a derivation or a supporting citation, or rephrase the statement as an empirical observation if no microscopic model is intended.","section":"§III (polarity discussion)"}],"minor_comments":[{"comment":"The phrase 'topsy-turvy in the position of the resonances' should be replaced by a quantitative statement, such as an explicit description of how the sign of nΩ_L in Eq. (2) changes when scanning B rather than ν_RF.","section":"§I"},{"comment":"The procedure for extracting the center of the zero-field resonance from the dispersive MM signal, used for the entries in Tabs. 1 and 2, is not described; please specify the fitting or zero-crossing criterion.","section":"§III"},{"comment":"The hyperfine splitting symbol appears garbled as 'Δ𝑕𝑓𝑠'; please ensure correct typesetting of Δ_hfs in the equations and text.","section":"Eqs. (2)-(3)"},{"comment":"There are stray spaces in '~250 C' and '~48 0 C' that should be corrected to '~250 °C' and '~48 °C'.","section":"ARC-cell paragraph"},{"comment":"The caption states the blue curve is 5 times magnified, but it is not clear whether the insert uses the same magnification; please clarify the scaling and add axis labels to the insert.","section":"Fig. 2(b)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the journal's scope and the experimental work appears genuine. The principal risk is the unproven independence of the population-redistribution signal from ν_RF; if the authors can supply a control or estimate, the paper is publishable. The self-citation of Ref. [15] for the +5Ω_L enhancement is from the same group; independent confirmation or additional data would be useful but is not strictly required."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper does something concrete and useful: it shows that you can scan the magnetic field at fixed RF detuning and see both the zero-field population-redistribution feature and the CPT resonances between non-degenerate states, in buffer-gas and anti-relaxation-coated cells, and it uses that to calibrate three-axis coils. That dual-signal view is genuinely not in the earlier references, and the calibration factors from B-scanning and conventional RF-scanning agree to the precision shown. The practical contribution is real.\n\nThe physics is standard CPT plus linear Zeeman shifts, and the authors don't overclaim that. The new part is the procedure and the signal-separation argument.\n\nThe main soft spot is the one the stress-test flags: they assert, after Fig. 2(c), that the population-redistribution signal doesn't depend on ν_RF and therefore doesn't contaminate the RF-demodulated CPT trace. That's plausible but not derived, and it's load-bearing for the claim that the CPT spectrum is a clean interference spectrum. A residual B-dependent baseline at the RF modulation frequency could shift apparent zero crossings and bias the calibration factors. The agreement with RF scanning suggests the effect is small under their conditions, but a control experiment—e.g., detuning ν_RF and checking the baseline—would make the claim solid.\n\nAlso, there are no uncertainties anywhere. The tables give fields to 0.1 μT and calibration factors to one decimal, but no error bars. For a calibration method that's a real gap. The polarity-change explanation for the M_M signal slope is asserted rather than derived; it's probably right, but a line or two of algebra would help.\n\nMinor: the paper says the modified method allows simultaneous acquisition and then later notes the +5Ω_L resonance overlaps with the population-redistribution signal in the M_M trace, which is why they use the CPT trace. So the simultaneity partly depends on the same unproven separation.\n\nWho this is for: experimentalists doing CPT magnetometry or coil calibration with alkali cells. They'll get a workable recipe and a cautionary tale about the baseline assumption. I'd send it to peer review; it deserves a serious referee, and the fixes are not deep: add uncertainty analysis, a derivation or control for the separation claim, and a brief derivation of the polarity rule. The central method holds up as far as I can tell from the text.","headline":"A practical B-field-scanning variant of CPT that shows simultaneous zero-field and interference resonances; the central separation claim is plausible but under-derived.","tokens_in":8751,"tokens_out":2369,"would_cite":true,"duration_ms":23939,"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":"Scanning the magnetic field, not the RF frequency, captures two kinds of quantum resonance in one rubidium cell and calibrates the coils in place.","keywords":["coherent population trapping","magnetic field scanning","rubidium atomic magnetometry","in-situ coil calibration","zero-field population redistribution","buffer gas cell","anti-relaxation coated cell","vector magnetic field sensing"],"falsifier":"The decisive check is to record the RF-demodulated CPT signal in zero transverse field while sweeping $\\nu_{\\mathrm{RF}}$ at fixed $B$: any $\\nu_{\\mathrm{RF}}$-dependent structure where the condition $2\\nu_{\\mathrm{RF}} = \\Delta_{\\mathrm{hfs}} \\pm n\\Omega_L$ permits no Zeeman resonance would disprove the claimed separation. A complementary check is to repeat the coil calibration at several fixed detunings such as $-58$, $-316$, and $-434$ kHz and verify that the calibration factors are independent of $\\delta_{\\mathrm{RF}}$.","tokens_in":7853,"feed_emoji":"🧲","tokens_out":12028,"duration_ms":98018,"temperature":0.7,"pith_summary":"The paper reports an experimental method for recording quantum-interference resonances in rubidium vapour by scanning the magnetic field while holding the frequency separation of a bichromatic light field fixed, the reverse of the usual CPT procedure. A single scan then contains two kinds of physics at once: a zero-field population-redistribution feature among degenerate Zeeman states and coherent-population-trapping resonances between non-degenerate states, which previously required different experimental arrangements. The method is demonstrated in a nitrogen-buffer-gas cell and in an anti-relaxation-coated cell, and the resonance positions, set by the Larmor frequency $\\Omega_L = g_F \\mu_B B$, are used to estimate ambient field components and to calibrate three-axis magnetic coils in situ. The practical payoff is that the coil calibration factors obtained this way agree with the conventional RF-scanning method, giving a simpler atom-based calibration route.","feed_headline":"Field scan pulls two resonance types from one rubidium cell","feed_subtitle":"Fixing the RF and sweeping B calibrates three-axis coils in place from atomic constants.","key_machinery":"The load-bearing machinery is the inverted resonance condition $2\\nu_{\\mathrm{RF}} = \\Delta_{\\mathrm{hfs}} \\pm n\\Omega_L$ together with the Larmor relation $\\Omega_L = g_F \\mu_B B$; fixing $\\nu_{\\mathrm{RF}}$ and sweeping $B$ makes the magnetic field the spectroscopic axis, so each Zeeman resonance appears at a computable field value rather than a frequency. Two demodulation channels separate the physics: demodulating the photodiode signal at the 440 Hz modulation applied to the RF gives the CPT signal, while demodulating at the 39 Hz modulation applied to the magnetic field gives the $M$ signal, which carries the zero-field population-redistribution feature. The sign structure of the CPT resonances changes between the negative and positive sides of the field scan because the demodulated signal is effectively the field derivative of the line shape, an effect the paper uses to explain the spectra and to propose vector-field readout.","core_discovery":"On its own terms, the central claim is that scanning $B$ at fixed RF detuning $\\delta_{\\mathrm{RF}} = 2\\nu_{\\mathrm{RF}} - \\Delta_{\\mathrm{hfs}}$ is not a variant of CPT spectroscopy but a way to acquire, in one trace, resonances of different physical origins. The two-photon resonance condition $2\\nu_{\\mathrm{RF}} = \\Delta_{\\mathrm{hfs}} \\pm n\\Omega_L$ turns each Zeeman resonance into a specific magnetic-field value; as the field crosses those values, derivative-like signals appear, and the $\\pm n$ pairs sit symmetrically about the zero of the scanned field component even when orthogonal fields are present. In the same scan a broad zero-field population-redistribution feature appears, and the authors argue that it is excluded from the RF-demodulated CPT channel because it does not depend on $\\nu_{\\mathrm{RF}}$. They verify the method in buffer-gas-filled and anti-relaxation-coated cells, observe the expected even/odd harmonic selection for longitudinal versus transverse field scans, and report bias-field and calibration results consistent with the conventional method.","pith_inferences":["An implication left implicit is that the same fixed-RF scan could serve as a two-instrument magnetometer, reading a Hanle-type zero-field response and a CPT-type high-field response from one beam path, which would enlarge the usable dynamic range without a second apparatus.","A testable extension is to monitor the zero-field resonance centre in the $M$ channel as a long-term drift reference and compare it with the CPT-derived coil calibration; the anti-relaxation-coated cell comparison in the paper already suggests agreement at the level of about a microtesla.","The slope-asymmetry effect suggests a direction-discriminating vector magnetometer could be built by registering the sign of CPT resonance derivatives as the field is scanned across zero, though the quantitative dependence on scan rate, modulation amplitude, and RF detuning is not developed in the paper."],"forward_implications":["The magnetic-field-scan method yields the same bias-field estimates and coil calibration factors as the conventional RF-scan method, so it can serve as an atom-based in-situ calibration of three-axis coils without a reference magnetometer.","Because the $\\pm n\\Omega_L$ resonances remain symmetric about the zero of the scanned field component even when orthogonal field components are present, one scan along an axis directly gives the background field along that axis.","In the CPT channel the population-redistribution feature is absent, so quantum-interference resonances can be read out cleanly even when they would overlap the zero-field signal in the $M$ channel.","The opposite slope of CPT resonances on the negative and positive sides of the scan provides a polarity-dependent signature that can be used for vector-field measurement.","The two cell types complement each other: buffer gas makes the zero-field resonance strong, while anti-relaxation coating makes the quantum-interference resonances prominent even in the magnetic-demodulation signal."],"supporting_citations":[{"why":"Supplies the conventional CPT magnetometer configuration that the modified magnetic-field-scan method is compared against.","marker":"5"},{"why":"Provides the near-zero-field atomic magnetometer approach whose zero-field resonance is merged into the same scan.","marker":"6"},{"why":"Establishes the monochromatic-light zero-field population-redistribution resonance that the paper extends to the bichromatic case.","marker":"12"},{"why":"Reports the enhanced $+5\\Omega_L$ resonance under transverse bias in conventional RF scanning, which the modified method reproduces in field scanning.","marker":"15"},{"why":"Gives the buffer-gas versus anti-relaxation-coated cell comparison that the paper uses to explain differing resonance amplitudes.","marker":"21"},{"why":"Supplies the rubidium atomic constants that make the in-situ coil calibration traceable to fundamental values.","marker":"23"}],"fun_headline_variants":["Field sweep, not frequency, yields twin resonances in rubidium","Sweeping B, fixed RF: two interference signals in one trace","Magnetic scan calibrates three-axis coils from atomic lines","One B-scan sees population and superposition resonances together","Scan the field, keep RF: dual quantum interference in vapors"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the zero-field population-redistribution signal is independent of the RF frequency $\\nu_{\\mathrm{RF}}$, so that demodulating at the RF modulation leaves only the quantum-interference resonances; the paper states this without derivation.","fun_headline_variants_meta":{"raw":{"variants":["Field sweep, not frequency, yields twin resonances in rubidium","Sweeping B, fixed RF: two interference signals in one trace","Magnetic scan calibrates three-axis coils from atomic lines","One B-scan sees population and superposition resonances together","Scan the field, keep RF: dual quantum interference in vapors"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000202,"raw_usage":{"total_tokens":1345,"prompt_tokens":868,"completion_tokens":477,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":389}},"tokens_in":484,"tokens_out":477,"duration_ms":5269,"temperature":1.0,"reasoning_tokens":389,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:51:02.421040+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The decisive check is to record the RF-demodulated CPT signal in zero transverse field while sweeping $\\nu_{\\mathrm{RF}}$ at fixed $B$: any $\\nu_{\\mathrm{RF}}$-dependent structure where the condition $2\\nu_{\\mathrm{RF}} = \\Delta_{\\mathrm{hfs}} \\pm n\\Omega_L$ permits no Zeeman resonance would disprove the claimed separation. A complementary check is to repeat the coil calibration at several fixed detunings such as $-58$, $-316$, and $-434$ kHz and verify that the calibration factors are independent of $\\delta_{\\mathrm{RF}}$.","supporting_citations":[],"review_version":1}