{"id":"b7e465e1-745b-411f-9361-81045c6fe4d7","arxiv_id":"2506.18595","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"The CASH haloscope proposal projects that Josephson-junction single-photon detectors can reach axion-photon couplings below 1e-15 GeV^-1 in the 38-54 micro-eV mass range, beating the standard quantum limit.","lead":"This paper proposes a new microwave haloscope experiment, CASH, using Josephson-junction single-photon detectors to search for dark-matter axions in the 38-54 micro-eV mass range. It projects sensitivity to axion-photon couplings down to about 1e-15 GeV^-1, below the DFSZ benchmark, in one year of data taking.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"CASH sensitivity projections assume unity single-photon detection efficiency and field-independent dark counts; neither is supported for the JJ SPD in the 1.7-10 T magnet environment.","rationale":"The central claim—that CASH can reach g_aγγ ~ 1.6e-15 GeV^-1 at fixed mass and scan the 38–54 μeV window to C_γ=0.36–0.25—rests on the SNR formalism of Sec. IV. I checked the algebra of Eqs. (9), (10), (15), and (16); it is internally consistent given R_sig and R_dc. The soft spot is that R_sig is defined as the axion-photon conversion rate in the cavity (P_a/m_a), not the detected photon count rate. A real photodetector has finite quantum efficiency η and possibly background from the magnetic-field environment. No η appears in the formulas, and none is quoted for the Al JJ SPD in the CASH frequency range. The paper's own reference [40] measures switching rates from a thermal photon source, which could in principle be converted into an efficiency, but this is not done. The dark count rate 0.01 Hz is also measured in zero field; Josephson junctions are sensitive to magnetic fields via the flux-dependence of the critical current, so the field environment of the magnet (1.7–10 T) could alter both the dark count rate and the optimum bias point. These are not mere engineering details: they enter the sensitivity formulas as explicit parameters. In the background-dominated regime (which holds for much of the CASH-II scan), C_γ ∝ (R_dc/t)^{1/4}, so a factor-10 increase in dark counts or a factor-10 decrease in efficiency degrades the reach by about 1.8. The proposed concrete check—a bench test of the actual SPD in the magnetic field—would directly validate or falsify the assumptions. Thus the conditional verdict is appropriate: the proposal is promising and arithmetically sound, but the projected reach should not be treated as a firm capability until these detector parameters are measured.","tokens_in":17675,"tokens_out":16600,"duration_ms":168933,"concrete_test":"Bench a JJ SPD of the CASH design in a dilution refrigerator with a calibrated microwave source (e.g., attenuated thermal photons as in Ref. [40]) and measure its single-photon detection efficiency and dark count rate as a function of applied magnetic field from 0 to 10 T, at 10–20 mK and at the operating bias current. If η ≥ 0.5 and R_dc ≤ 0.01 Hz at full field, the CASH sensitivity projections in Eqs. (10)–(11) and Table II remain valid; if either fails, the reach must be re-evaluated using the measured parameters.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section IV (Eqs. 7, 9, 10, 16) takes R_sig = P_a/m_a as the signal event rate and contains no detector efficiency factor. For a real single-photon counter the observed rate is R_det = ηR_sig, so the quoted reach is that of a perfect detector. Since the paper nowhere reports η for the Al JJ SPD at 8.8–13 GHz, and Ref. [40] only demonstrates switching from thermal photons without extracting an efficiency, the headline values (Eq. 11: C_γ=0.22 at 1.7 T; Table II: C_γ=0.36 at 7 T, 0.25 at 10 T) are projections under an implicit η=1 assumption. Additionally, the dark count rate R_dc=0.01 Hz was measured in zero field; the Josephson critical current is strongly flux-dependent, and magnetic shielding for the SPD in the CASH magnet is not described in Sec. III. In the background-dominated regime, which holds for much of the scan, C_γ scales as (R_dc/t)^{1/4}; a tenfold rise in dark counts or a tenfold drop in efficiency would worsen the reach by about a factor of 1.8. Both assumptions require direct experimental validation before the projected sensitivity can be regarded as firm.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a new haloscope experiment, CASH, to search for dark-matter axions in the 38–54 μeV mass range using an underdamped Josephson-junction single-photon detector coupled to a tunable copper cavity. The projected sensitivity is derived from the standard Sikivie power formula (Eq. 3) and a Poisson-counting SNR formula (Eq. 5), with input parameters from the authors' CST simulations of the cavity (frequency, Q0, form factor) and from their previous measurements of JJ dark counts. The headline projections are Cγ ≈ 0.22 (gaγγ ≈ 1.6×10^-15 GeV^-1) at a single fixed mass after 10^6 s with a 1.7 T magnet (Eq. 11), and a one-year scan down to Cγ = 0.36 (7 T) or Cγ = 0.25 (10 T) over 38–54 μeV (Table II), which would reach below the DFSZ benchmark for most of this range. The paper also discusses the tuning scheme, the optimal quality factor, and the comparison with existing and proposed haloscopes. The central derivation is conventional and internally consistent in its numerics, but several assumptions about detector performance in the final configuration are not fully supported.","tokens_in":17935,"tokens_out":20517,"duration_ms":179474,"significance":"Should the projected performance be realized, CASH would be a major step in axion dark-matter searches: it would cover a mass range that is not yet excluded and would reach couplings below the DFSZ benchmark, going well beyond the standard quantum limit with a single-photon detector. The paper's strength is its concrete design: measured dark-count data, detailed CST simulations of the tuning rod and form factor, and an explicit SNR optimization over Q0. The proposed experiment is falsifiable in the sense that the sensitivity curves in Figs. 7 and 8 make specific, testable predictions. However, the significance is conditional on two detector assumptions—near-unity single-photon detection efficiency and field-independent dark counts—that are not demonstrated in the present manuscript.","major_comments":[{"comment":"The sensitivity calculation uses the axion photon arrival rate R_sig directly in the SNR formula with no detector efficiency factor. For a real single-photon counter the observed signal rate is η R_sig, and the projected Cγ values in Eq. (11) and Table II are therefore those of a perfect detector (η=1). The manuscript reports dark counts and switching due to thermal photons in Ref. [40], but it does not report or reference a measured detection efficiency for the Al JJ SPD in the 8.8–13 GHz band. In the background-dominated regime, fixing SNR and time gives Cγ ∝ R_dc^{1/4}/√η, so η=0.1 would weaken the quoted reach by about a factor of 3, and even η=0.5 by about 40%. Please add an explicit efficiency factor to the signal rate in the SNR equations and either provide a measured value or clearly label η=1 as a working assumption.","section":"Sec. IV, Eqs. (5), (7), (10) and Eq. (11)"},{"comment":"The dark-count rate R_dc=0.01 Hz used in Eqs. (9)–(10) and Table II was measured in a stand-alone cryogenic test (Sec. III.A, Fig. 2) with no magnetic field. In the proposed experiment the SPD is connected by a coaxial line to a cavity in a 1.7–10 T magnet, and the paper does not discuss the effect of the magnetic field on the Josephson-junction critical current and switching statistics, nor any magnetic shielding for the detector. Since the Josephson critical current is flux-sensitive, the in-field dark-count rate is an open experimental question. Because Cγ scales as R_dc^{1/4} in the background-dominated regime, an order-of-magnitude increase in dark counts would weaken the projected sensitivity by about a factor of 1.8; this should be either validated with a measurement in the magnet bore or explicitly presented as a key technical risk.","section":"Secs. III.B and IV.A"},{"comment":"As printed, this equation does not reproduce the numerical results quoted in Eq. (11). Solving Eq. (9) for Cγ gives Cγ = SNR/(B0√(G Q0)) × [ (1+√(1+4tR_dc/SNR^2))/(2t) ]^{1/2}, whereas Eq. (10) has √(SNR/(2t)) in place of SNR/√(2t). Evaluating the printed formula with the parameters stated for Eq. (11) does not yield Cγ=0.22, while the corrected expression does (for Cγ=0.22 it gives SNR≈1.65). Please correct Eq. (10) and also state the unit convention for B0, since Eq. (7) is written with B0 in Tesla while the natural-unit derivation requires B0 in GeV^2.","section":"Sec. IV.A, Eq. (10)"}],"minor_comments":[{"comment":"The unit convention for B0 is not stated; inserting B0=1.7 T and the G of Eq. (8) into Eq. (7) requires an implicit Tesla-to-GeV^2 conversion, and the text should say so explicitly.","section":"Sec. IV.A, Eq. (7)"},{"comment":"The '10^4 s' dark-count time and the associated '0.001 Hz' rate are extrapolations from a dashed fit, not measured values; please label them as extrapolations in both the text and the figure caption.","section":"Sec. III.A and Fig. 2"},{"comment":"The statement that 'all theoretically favored models of the Peccei-Quinn axion could be fully explored for this mass' is too strong, because the projection assumes the detector efficiency and magnetic-field compatibility discussed in the major comments.","section":"Sec. IV.A, text after Eq. (11)"},{"comment":"The conclusion repeats the 'not more than a single false switching event in 100 s' claim; this should be tied to the measured dark-count data rather than stated as an intrinsic detector property.","section":"Sec. V"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a well-structured proposal with a conventional and numerically consistent sensitivity estimate; the main risk is not the formalism but the unvalidated detector assumptions (efficiency and magnetic-field tolerance) that enter the headline numbers. The typo in Eq. (10) is straightforward to fix. I would encourage the editor to send the manuscript back for a revision in which the authors either provide the missing detector characterization or explicitly frame the projections as conditional on those measurements. The paper fits the scope of Physical Review D and, with these changes, would be a useful contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take: this is a legitimate experimental proposal, not a hype piece. The new part is the integrated CASH design—tunable copper cavity, Josephson-junction single-photon detector, and a scan optimization over Q0 for a 38–54 µeV mass window—and the quantitative claim that it can hit Cγ ~0.36 (7 T) or 0.25 (10 T) in a year. That would be the first haloscope below DFSZ in that window, so the stakes are real. The sensitivity arithmetic is internally consistent; I checked the scaling of Eq. (16) and the numbers in Table II follow from the stated parameters. Credit where due: they use standard haloscope formulas, report measured dark count rates rather than invented ones, and include a genuine optimization over Q0 and scan overhead, which many proposals skip.\n\nThe soft spots are the ones the stress-test flagged, and they are real but not fatal to the proposal's value. First, no detector efficiency factor appears anywhere in Eqs. (7)–(16). R_sig is taken as the observed rate, which is only true for η=1. For a real JJ SPD at 8.8–13 GHz, η is not reported in ref. [40]; the demonstration there is switching from thermal photons, not a calibrated detection efficiency. If η turns out to be 0.1, the reach in Cγ worsens by roughly a factor of 1.8 in the background-dominated regime. That is not a small effect for a paper whose headline numbers depend on it. Second, the dark count rate of 0.01 Hz was measured in zero field; the Josephson critical current is flux-sensitive, and the paper does not describe magnetic shielding for the detector inside the 1.7–10 T environment. If dark counts rise tenfold, the reach again degrades by ~1.8. These are engineering validation issues, not errors in the formalism, and the authors are not hiding them—they simply haven't done those tests yet. My one real complaint is that the abstract and conclusion state the projected sensitivity as if it were a capability, rather than as a sensitivity conditional on detector performance. That overstatement should be fixed.\n\nWho is this for? Anyone tracking haloscope proposals or single-photon detection for axion searches. It deserves a serious referee—the design is concrete, the math is solid, and the claimed reach is important if the detector integration holds up. I would send it to review, with the explicit request that the referee pressure-test the detector efficiency and magnetic-field assumptions.\n\nRecommendation: accept for peer review.","headline":"A serious haloscope proposal whose sensitivity projections rest on two untested detector assumptions; deserves refereeing, but the projected reach should be labeled conditional on detector efficiency and magnetic-field compatibility.","tokens_in":18645,"tokens_out":1691,"would_cite":true,"duration_ms":16885,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A proposed haloscope could reach dark-matter axion couplings of 1.6e-15 GeV^-1 at 38 µeV.","keywords":["axion dark matter","haloscope","single-photon detector","Josephson junction","microwave cavity","axion-photon coupling","standard quantum limit","QCD axion"],"falsifier":"Energize the CASH magnet at $1.7$ T or $7$ T with the cavity cold at about $20$ mK and the Josephson-junction detector biased in its sensitive range, then time switching events with no injected signal: if the dark-count rate exceeds $0.01$ Hz, or if a calibrated photon flux injected into the cavity shows detection efficiency well below unity, the projected $C_\\gamma$ values in Eqs. (11) and (12) and Table II are not reached.","tokens_in":17506,"feed_emoji":"📡","tokens_out":10335,"duration_ms":97720,"temperature":0.7,"pith_summary":"The paper proposes CASH, a cavity haloscope for dark-matter axions in the mass range $38$--$54~\\mu$eV that replaces the usual microwave amplifier with a Josephson-junction single-photon detector. The central claim is that this avoids the standard quantum limit, so the only relevant noise is the detector's dark-count rate of about $0.01$ Hz. With that noise level, a fixed-frequency measurement at $38~\\mu$eV using a $1.7$ T magnet reaches $C_\\gamma = 0.22$ ($g_{a\\gamma\\gamma}\\simeq 1.6\\times 10^{-15}$ GeV$^{-1}$) in $10^6$ seconds, and a one-year scan with a $7$ T or $10$ T magnet reaches $C_\\gamma = 0.36$ or $0.25$, respectively. If these projections hold, CASH would be the most sensitive haloscope in this mass window and would cover all favored QCD axion models there.","feed_headline":"Haloscope proposal reaches axion couplings of 1.6e-15 GeV^-1","feed_subtitle":"Josephson-junction single-photon counters beat the quantum limit to cover the 38-54 µeV axion window.","key_machinery":"The load-bearing object is the underdamped Josephson junction as a threshold single-photon detector: biased near its critical current, an arriving microwave photon switches it from the superconducting to the resistive state, producing a measurable gap-voltage pulse, and the switching rate in the dark sets the background. The argument is carried by two analytic expressions: Eq. (10), which converts a measured dark-count rate $R_{\\rm d.c.}$ into a $C_\\gamma$ sensitivity at fixed observation time, and Eq. (16), which optimizes $Q_0$ against the number of frequency steps needed to scan a mass interval. The cavity design is a cylindrical TM$_{010}$ copper resonator tuned with a movable copper rod, whose simulated form factor stays above about $0.57$ across the tuning range; the detector, the cavity, and the magnet are co-located in a dilution refrigerator at $10$--$20$ mK.","core_discovery":"The discovery claim is quantitative: single-photon counting with an underdamped Josephson junction bypasses the amplifier noise that sets the standard quantum limit, and leaves Poisson-distributed dark counts as the dominant background. Using the signal rate from axion-photon conversion in a TM$_{010}$ copper cavity and the SNR formula for signal plus background, the paper derives a coupling sensitivity $C_\\gamma$ that depends only on magnetic field, cavity quality factor, volume, form factor, observation time, and dark-count rate. The headline numbers are $C_\\gamma=0.22$ at fixed mass $38~\\mu$eV with $1.7$ T in $10^6$ s (Eq. 11), $C_\\gamma=0.05$ with $7$ T at the same fixed mass (Eq. 12), and a scan of $38$--$54~\\mu$eV in one year reaching $C_\\gamma=0.36$ at $7$ T or $0.25$ at $10$ T (Table II). The scan analysis optimizes the unloaded quality factor: too high a $Q_0$ narrows the resonance and costs frequency steps, so the optimal $Q_0$ for a one-year scan is $8\\times 10^4$, requiring about $1.4\\times 10^4$ tuning steps.","pith_inferences":["If the extrapolated dark-count rate of $10^{-3}$ Hz is achieved, the same formulas improve the projected limits roughly as $R_{\\rm d.c.}^{1/4}$ in the dark-count-limited regime, pushing the fixed-mass limit at $1.7$ T toward $C_\\gamma\\simeq0.12$.","The tuning-rod simulations keep the form factor above $0.57$ across the whole $9.2$--$13.1$ GHz band, suggesting the cavity is not the bottleneck; a multi-cell or multi-resonator version could scan the same mass window faster with the same detector technology.","The same detector-noise-limited SNR analysis applies to other fixed-frequency dark-matter searches, such as hidden-photon searches in different mass windows, whenever a threshold single-photon detector is available.","A direct measurement of detection efficiency versus frequency with a calibrated injected tone would convert the quoted projected limits into a testable exclusion curve, since the sensitivity equations currently assume unit efficiency."],"forward_implications":["At the fixed mass $38~\\mu$eV, CASH I with $1.7$ T and $10^6$ s would exclude $C_\\gamma>0.22$ at 95% CL, covering both standard benchmark QCD axion models at that mass.","With the same non-tunable cavity at $7$ T, the fixed-mass exclusion improves to $C_\\gamma=0.05$ ($g_{a\\gamma\\gamma}=3.9\\times10^{-16}$ GeV$^{-1}$).","CASH II's one-year scan of $38$--$54~\\mu$eV would reach $C_\\gamma=0.36$ at $7$ T or $0.25$ at $10$ T, below the benchmark coupling of $0.75$ and below all favored QCD axion couplings.","Because the optimal unloaded quality factor for the one-year scan is $Q_0=8\\times10^4$, the experiment's reach is limited mainly by dark counts and tuning-step overhead, not by maximizing cavity quality.","Without the magnetic field, the same single-photon detector and cavity can search for dark-photon dark matter, extending the setup's physics reach beyond axions."],"supporting_citations":[{"why":"Establishes the resonant-cavity haloscope conversion signal used throughout the proposal.","marker":"[34]"},{"why":"Supplies the form-factor and quality-factor formalism for the axion conversion power.","marker":"[37]"},{"why":"Gives the Poisson statistics of coherent signal photons and distinguishes thermal Bose-Einstein noise.","marker":"[39]"},{"why":"Demonstrates the Josephson-junction single-photon detector and the 0.01 Hz dark-count rate used in the sensitivity estimates.","marker":"[40]"},{"why":"Provides the signal-to-noise formula for Poisson signal and dark counts used in Eqs. (5), (9), and (10).","marker":"[41]"},{"why":"Provides the movable metallic-rod tuning approach adopted for the CASH cavity.","marker":"[52]"},{"why":"Analyzes single-photon versus linear-amplifier detectors for haloscope searches, motivating the design.","marker":"[67]"},{"why":"Reports development of the microwave single-photon counter for axion searches that underlies the detector.","marker":"[69]"}],"fun_headline_variants":["Single-photon haloscope beats quantum limit for axions","CASH haloscope design targets 10^-15 axion coupling","CASH: most sensitive haloscope for axions in 38-54 µeV","Haloscope with Josephson-junction counters to beat quantum limit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The sensitivity estimates assume that the Josephson-junction detector maintains a dark-count rate at or below $0.01$ Hz while operating inside the CASH cryostat in the presence of a $1.7$--$10$ T magnetic field, and that it records essentially every signal photon; the demonstrated $0.01$ Hz rate comes from a separate cryogenic measurement without the final magnetized configuration.","fun_headline_variants_meta":{"raw":{"variants":["Single-photon haloscope beats quantum limit for axions","CASH haloscope design targets 10^-15 axion coupling","CASH: most sensitive haloscope for axions in 38-54 µeV","Haloscope with Josephson-junction counters to beat quantum limit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001267,"raw_usage":{"total_tokens":5214,"prompt_tokens":1003,"completion_tokens":4211,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":619,"completion_tokens_details":{"reasoning_tokens":4135}},"tokens_in":619,"tokens_out":4211,"duration_ms":30994,"temperature":1.0,"reasoning_tokens":4135,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:47:07.087029+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Energize the CASH magnet at $1.7$ T or $7$ T with the cavity cold at about $20$ mK and the Josephson-junction detector biased in its sensitive range, then time switching events with no injected signal: if the dark-count rate exceeds $0.01$ Hz, or if a calibrated photon flux injected into the cavity shows detection efficiency well below unity, the projected $C_\\gamma$ values in Eqs. (11) and (12) and Table II are not reached.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports development of the microwave single-photon counter for axion searches that underlies the detector."}],"review_version":2}