{"id":"7cc217d2-3192-46b2-9cc1-8b5d40d74e4d","arxiv_id":"2506.11589","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A 38 MHz scan near 10.2 GHz with the QUAX haloscope found no axion signal, setting 90% confidence limits on the axion-photon coupling at the 10^-14 GeV^-1 level.","lead":"The QUAX experiment searched for axion dark matter at a mass near 42 micro-electronvolts using a tunable microwave cavity and a quantum-limited amplifier, and found no signal. The null result sets new limits on the axion-photon coupling and excludes two specific hadronic axion models in a narrow mass window.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exclusion limits rely on an unverified assumption that the TM030 form factor C030 = 0.43 is constant over the clamshell tuning range; a per-step variation would rescale the coupling limits and could break the claimed E/N = 29/3 interval.","rationale":"I read the paper's central result as the narrow exclusion of E/N = 44/3 and E/N = 29/3 hadronic axion models between 41.991 and 42.234 µeV. The reader's strongest claim matches this. The most insecure link in the chain is the conversion from measured excess power to gaγγ, which depends on the cavity form factor. The paper's own text acknowledges the constancy of C030 is assumed from simulations, with only one bead pull at a single opening angle. The Q0 data in Fig. 3(a) demonstrate that the mechanical tuning has non-negligible geometry variation, so assuming C030 is constant is not self-evident. The coupling limit scales as C^{-1/2}; a 20% change in C translates into a 10% shift in the limit curve. Since the E/N = 29/3 exclusion interval is relatively narrow, this systematic could alter the interval boundaries or introduce holes. I therefore agree with the reader's weakest-assumption identification and the CONDITIONAL verdict. I would not escalate to REJECT because the underlying method is sound and the required check is straightforward; the paper just needs to demonstrate C030 stability or include the systematic in the limits. I also note the abstract's \"m_a > 40 µeV\" phrasing overstates the coverage, but that is a presentation issue rather than a load-bearing technical flaw.","tokens_in":15503,"tokens_out":15876,"duration_ms":159686,"concrete_test":"Run a bead pull (or an equivalent finite-element mode reconstruction) at several opening angles spanning the full scan, specifically at the endpoints 10.154 and 10.212 GHz and at tuning points inside the intruder-mode region; apply the measured per-step C030 values to recompute the Fig. 4 limits. If the recomputed limits stay below the E/N = 29/3 model line over the whole quoted 136 neV mass interval, the exclusion claim survives; if not, the interval must be reduced to the sub-ranges where the limits actually cross the model prediction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central exclusion of the E/N = 44/3 and E/N = 29/3 models is computed through the axion power in Eq. (1), where the coupling limit scales as C^{-1/2}. The paper adopts C030 = 0.43 from simulations and states it is \"assumed to be constant throughout the frequency tuning region\", supported by a single room-temperature bead pull at a 0.3 degree opening angle that gave 0.40 ± 0.02. The clamshell tuning changes the gap between the two copper halves up to one degree, and the same mechanical clearances produce the large step-to-step Q0 spread visible in Fig. 3(a); there is no reason to expect the field geometry factor to be strictly constant. If C030 varied by, say, 20% across the scan, the 90% limit curve would shift by about 10% in gaγγ. The quoted E/N = 29/3 exclusion (136 neV) is not a wide-margin result, so even a moderate C030 drift could remove part of the claimed interval. The paper should either measure C030 at multiple tuning steps or quote limits with a systematic band from the assumed constancy.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a haloscope search for QCD axion dark matter in a ~38 MHz window centered near 10.2 GHz (axion mass ~42 µeV) using the QUAX apparatus: a dielectrically loaded, clamshell-tunable cavity in an 8 T magnet, read out by a TWPA-based chain with system noise temperatures down to ~1.1 K. After 225 hours of integration over 147 measurement points, no signal candidate survives the SNR=4.5 threshold, and 90% C.L. upper limits on the axion-photon coupling are set at the ~10^-14 GeV^-1 level. The authors claim exclusion of hadronic axion models with E/N=44/3 and E/N=29/3 in narrow mass intervals between 41.991 and 42.234 µeV.","tokens_in":15601,"tokens_out":6977,"duration_ms":66923,"significance":"If the headline exclusion withstands scrutiny, this is a milestone: it would be the first cavity-haloscope bound that rules out specific QCD axion models in the post-inflationary mass region above 40 µeV. The experiment itself is well executed: the quantum-limited receiver, the low system noise at these frequencies, the documented Monte Carlo confidence belts, and the explicit correction for the ~30% baseline-filtering underestimation of axion power are all strengths. The paper also gives a transparent error budget (~4%) for the measured cavity and receiver parameters. The central physics claim, however, rests on two assumptions—constant cavity form factor across tuning and a continuous scan without unexcluded gaps—that need to be demonstrated or conservatively folded into the limits.","major_comments":[{"comment":"The abstract claims 'viable hadronic axion models are ruled out in a currently preferred post-inflationary region ma > 40 µeV.' This overstates the result: Section 'Data analysis and Results' shows that only two specific anomaly-coefficient ratios (E/N=44/3 and E/N=29/3) are excluded, and only in narrow intervals of 153 neV and 136 neV within 41.991–42.234 µeV. The limit curve does not reach KSVZ sensitivity over the entire ma > 40 µeV region; the quoted average sensitivity is about 3.4× the KSVZ coupling over a 300 kHz band. Please revise the abstract and conclusion to state precisely which models are excluded in which mass ranges.","section":"Abstract and Conclusion"},{"comment":"The exclusion of the E/N=29/3 model depends on the assumption that the TM030 form factor C030 = 0.43 is constant throughout the tuning range. The only experimental check is a single room-temperature bead pull at a 0.3° opening angle giving 0.40 ± 0.02. Since the clamshell tuning changes the gap between the two copper halves up to 1°, and the same mechanical clearances produce a large step-to-step spread in Q0 (Fig. 3a), there is no direct evidence that the field overlap factor is stable at the percent level. The conversion power in Eq. (1) scales with C, so the derived coupling limit scales as C^{-1/2}; a 20% change in C would shift the limit by about 10%, comparable to the margin of the 136 neV exclusion interval for E/N=29/3. Please either measure C030 at multiple tuning positions or add a systematic uncertainty band from the assumed constancy to the limit curve. This is a load-bearing point for the central claim.","section":"Experimental apparatus and data collection (form factor paragraph)"},{"comment":"The Figure 4 caption states 'The missing band at about 42.017 µeV is due to data removal for rf contamination.' This band lies inside the quoted excluded interval 41.991–42.234 µeV. If no limit is set in that sub-band, the statement that hadronic axion models are excluded over the entire interval is not strictly correct. Please specify the width of the removed band, show how the exclusion intervals were computed in the presence of the gap, and either quote the intervals excluding the gap or provide a conservative treatment that accounts for the missing coverage.","section":"Figure 4 caption and Data analysis and Results"}],"minor_comments":[{"comment":"Equation (1) gives the axion signal power without explicitly showing the cavity form factor, volume, magnetic field, and loaded quality factor that enter the actual conversion power. Please clarify how these measured parameters are folded into the limit calculation, either by writing the full expression or by stating that they are absorbed into the normalization of P_in.","section":"Data analysis and Results, Eq. (1)"},{"comment":"The conclusion says the scanning was performed 'within a ∼ 60 MHz extended region,' while the abstract and the body state the scanned range is 38 MHz. Please reconcile this wording; if 60 MHz refers to the full tuning range of the cavity rather than the scanned region, say so explicitly.","section":"Conclusion"},{"comment":"The phrase 'viable hadronic axion models' should be qualified by naming the specific anomaly coefficient ratios that are actually excluded, to avoid giving the impression that all hadronic models are ruled out in the whole region.","section":"Abstract"},{"comment":"In the lower panel, the lines for E/N=44/3 and E/N=29/3 should be labeled directly on the plot or in the caption so that the excluded intervals are visually identifiable.","section":"Figure 4"},{"comment":"Reference [12] is cited in the text as 'the landscape of qcd axion models'; the title should be capitalized as 'QCD' for consistency.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a strong experimental result, but the two main issues—unverified constancy of the form factor and the data gap inside the claimed excluded interval—are directly tied to the headline physics claim. Both can be addressed either by additional measurements or by conservative changes to the quoted intervals, so I do not recommend rejection at this stage. The overstatement in the abstract should also be corrected. If the authors can demonstrate that C030 is stable to a few percent across the tuning range, or if they quote limits with an appropriate systematic band, the central result could become publishable as is."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a competent, incremental haloscope search that delivers a genuinely new exclusion region near 42 µeV and demonstrates a tunable dielectrically loaded cavity with a TWPA readout reaching 2.3 photons of noise above 40 µeV. That is real progress, and the analysis follows standard practice: Monte Carlo confidence belts, a documented 30% baseline-filtering correction, and a ~4% error budget. There is no circularity; the limits come from measured noise and cavity parameters with external model predictions.\n\nThe soft spots are real but not fatal. The abstract claims hadronic axion models are ruled out in the post-inflationary region m_a > 40 µeV, but the body text correctly limits that to two specific E/N values (44/3 and 29/3) in narrow windows totaling 136–153 neV. That is an overstatement and should be fixed in revision. More substantively, the form factor C030 = 0.43 is assumed constant across the tuning range, based on one room-temperature bead pull at 0.3° opening, while the clamshell mechanism opens up to 1° and the paper itself shows significant step-to-step Q0 spread from mechanical clearances. If C030 varies by even 20%, the coupling limit shifts by ~10%, which could erode part of the E/N = 29/3 exclusion window. The authors should either measure C030 at multiple tuning steps or quote a systematic band from this assumption. This is a legitimate referee request, not a calculation error.\n\nThe central result—new 90% C.L. limits around 10^-14 GeV^-1 in a previously poorly covered region—is probably robust, and the apparatus is a useful step toward higher-mass axion searches. The paper deserves a serious referee: it is a new measurement, the technical work is sound, and the flaws are fixable. My recommendation: send to peer review, ask the authors to tone down the abstract and address the form-factor systematic explicitly. A reader in the axion community gets clear value; someone outside the field can skip the details.","headline":"Solid incremental haloscope result around 42 µeV, with an overstated abstract and one unquantified systematic (form-factor constancy) that a serious referee should push on.","tokens_in":687,"tokens_out":768,"would_cite":true,"duration_ms":38220,"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":"A quantum-limited scan around 10.2 GHz finds no QCD axion signal and excludes two hadronic axion models near 42 µeV.","keywords":["QCD axion","axion dark matter","haloscope","axion-photon coupling","post-inflationary axions","traveling wave parametric amplifier","tunable microwave cavity","quantum-limited detection"],"falsifier":"A direct measurement of $C_{030}$ as a function of clamshell opening angle, by bead-pull or cold-cavity perturbation at several frequencies across 10.154–10.212 GHz, would falsify the constancy assumption if it deviates by more than the few-percent level; such a deviation would rescale the coupling limits and could move or erase the excluded $E/N = 44/3$ and $E/N = 29/3$ mass intervals.","tokens_in":15222,"feed_emoji":"📡","tokens_out":14928,"duration_ms":132457,"temperature":0.7,"pith_summary":"The paper reports a microwave haloscope search for axion dark matter around $m_a \\simeq 42\\,\\mu\\mathrm{eV}$, scanning 38 MHz near 10.2 GHz with a quantum-limited receiver. No axion signal candidate was observed, and the experiment sets 90% confidence upper limits on the axion-photon coupling at the $10^{-14}\\,\\mathrm{GeV}^{-1}$ level. On that basis the paper claims that hadronic QCD axion models with anomaly coefficient ratio $E/N = 44/3$ and $E/N = 29/3$ are excluded in narrow mass intervals between 41.991 µeV and 42.234 µeV. This matters because lattice-based arguments favour the post-inflationary axion mass region above 40 µeV, where no previous cavity haloscope had reached QCD-axion sensitivity. A sympathetic reader would take the central achievement to be the first tunable haloscope scan at these masses with a readout noise of about 2.3 photons.","feed_headline":"Quantum-limited scan rules out hadronic axions near 42 µeV","feed_subtitle":"A tunable 10.2-GHz haloscope with a 2.3-photon receiver sets coupling limits at the 10^-14 GeV^-1 level.","key_machinery":"The load-bearing object is the haloscope: a 1-liter sapphire-loaded cylindrical copper cavity operated in the TM030 mode inside an 8 T magnetic field, whose frequency is tuned by a clamshell mechanism that opens the two copper halves by up to one degree. Axion conversion power is governed by Eq. (1) with the form factor $C_{030} = 0.43$, a value from finite-element simulations checked by one room-temperature bead pull at a 0.3 degree opening angle and assumed constant across the tuning range. The readout chain, a traveling-wave parametric amplifier at about 120 mK followed by HEMT amplifiers, gives the lowest noise temperature reported for a tunable haloscope above 40 µeV, with a minimum of about 1.1 K, or 2.3 photons. Signal searches use Fano fits to track the cavity resonance, a Savitzky-Golay filter for baseline estimation, and Monte Carlo confidence belts with a signal-to-noise threshold of 4.5 to set 90% C.L. limits.","core_discovery":"The central claim is that, after 225 hours of integrated acquisition across 124 nominal tuning points between 10.154 GHz and 10.212 GHz, no axion signal appeared, and the resulting 90% C.L. limits on the axion-photon coupling reach on average about $3.4$ times the KSVZ coupling over a 300 kHz band with one hour of integration. In the mass interval from 41.991 µeV to 42.234 µeV, the limits dip below the coupling predicted for hadronic axion models with $E/N = 44/3$ and $E/N = 29/3$, excluding those models over a width of 153 neV and 136 neV respectively. The paper states this as a direct exclusion of viable hadronic axion models in a currently preferred post-inflationary region with $m_a > 40\\,\\mu\\mathrm{eV}$. The discovery, on the paper's terms, is that a dielectric-loaded tunable haloscope with a traveling-wave parametric amplifier can operate at 10.2 GHz with a system noise temperature as low as 1.1 K, corresponding to 2.3 photons, and convert that sensitivity into model exclusion.","pith_inferences":["Extension: the assumption of constant $C_{030} = 0.43$ is tested only at one opening angle; measuring the form factor at several clamshell positions would harden the exclusion, and a few-percent variation would rescale the coupling limits and could shrink or shift the excluded mass intervals.","Extension: if lattice simulations relocate the preferred post-inflationary mass window, the same tuning-plus-TWPA design can be re-scanned over its wider 9.5–11 GHz range, so the technique's value does not depend on the specific 42 µeV window.","Extension: the reported noise performance suggests that quantum-limited amplifiers could also benefit other high-mass haloscope concepts, such as dielectric or plasma haloscopes, where receiver noise currently limits sensitivity."],"forward_implications":["If the exclusion holds, hadronic QCD axion models with $E/N = 44/3$ and $E/N = 29/3$ cannot account for the local dark matter density at axion masses between 41.991 µeV and 42.234 µeV.","The demonstration that a tunable cavity haloscope can reach near-KSVZ coupling sensitivity at 10.2 GHz opens a practical path for covering the full 9.5–11 GHz range the apparatus was designed for.","The average one-hour sensitivity of about $3.4$ times the KSVZ coupling over a 300 kHz band means a modest increase in magnet strength and duty cycle would let the same technique probe weaker couplings, including DFSZ-type models in the post-inflationary window.","Since no candidate was found at the chosen threshold, the data place a direct experimental constraint on QCD axion dark matter in the post-inflationary sub-window around 42 µeV."],"supporting_citations":[{"why":"Supplies the hadronic axion model landscape and the $E/N = 44/3$ and $29/3$ benchmarks whose exclusion is the headline result.","marker":"[12]"},{"why":"Lattice simulation motivating the post-inflationary axion mass region above 40 µeV that this search targets.","marker":"[21]"},{"why":"Updated adaptive-mesh-refinement lattice prediction that keeps the post-inflationary mass window above 40 µeV.","marker":"[22]"},{"why":"Describes the predecessor high-frequency haloscope run whose apparatus and procedures this work extends.","marker":"[28]"},{"why":"Characterizes the clamshell-tuned dielectric resonator at about 11 GHz that provides the cavity tuning used here.","marker":"[29]"},{"why":"Gives the bead-pull technique used to check the simulated form factor $C_{030}$ at a single opening angle.","marker":"[31]"},{"why":"Reports the traveling-wave parametric amplifier used as the first-stage quantum-limited readout device.","marker":"[32]"},{"why":"Defines the phenomenologically preferred axion-model band $E/N \\in (5/3, 44/3)$ that brackets the excluded models.","marker":"[67]"},{"why":"Provides the velocity-integrated axion lineshape with width about 10 kHz used to set search bin spacing.","marker":"[71]"},{"why":"Supplies the unified confidence-belt construction used to correct the 90% C.L. limits for non-negative excess power.","marker":"[72]"}],"fun_headline_variants":["Axion hunt at 42 µeV: no signal, hadronic models ruled out","Quantum-limited haloscope excludes hadronic axions at 42 µeV","No axions seen: quantum receiver tightens limits at 42 µeV","42 µeV axion search yields empty band, hadronic models ruled out","Hadronic axion models excluded at 42 µeV by quantum-limited scan"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The limits rest on the assumption that the cavity form factor $C_{030} = 0.43$ stays constant throughout the tuning range, a value checked by simulation and a single bead-pull measurement at one opening angle rather than at each tuning step.","fun_headline_variants_meta":{"raw":{"variants":["Axion hunt at 42 µeV: no signal, hadronic models ruled out","Quantum-limited haloscope excludes hadronic axions at 42 µeV","No axions seen: quantum receiver tightens limits at 42 µeV","42 µeV axion search yields empty band, hadronic models ruled out","Hadronic axion models excluded at 42 µeV by quantum-limited scan"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000777,"raw_usage":{"total_tokens":3426,"prompt_tokens":928,"completion_tokens":2498,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":544,"completion_tokens_details":{"reasoning_tokens":2396}},"tokens_in":544,"tokens_out":2498,"duration_ms":15353,"temperature":1.0,"reasoning_tokens":2396,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T04:03:39.770208+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct measurement of $C_{030}$ as a function of clamshell opening angle, by bead-pull or cold-cavity perturbation at several frequencies across 10.154–10.212 GHz, would falsify the constancy assumption if it deviates by more than the few-percent level; such a deviation would rescale the coupling limits and could move or erase the excluded $E/N = 44/3$ and $E/N = 29/3$ mass intervals.","supporting_citations":[{"cited_title":"Di Luzio, M","cited_arxiv_id":null,"evidence_quote":"Supplies the hadronic axion model landscape and the $E/N = 44/3$ and $29/3$ benchmarks whose exclusion is the headline result."},{"cited_title":"Buschmann, J","cited_arxiv_id":null,"evidence_quote":"Updated adaptive-mesh-refinement lattice prediction that keeps the post-inflationary mass window above 40 µeV."},{"cited_title":"Alesini, D","cited_arxiv_id":null,"evidence_quote":"Describes the predecessor high-frequency haloscope run whose apparatus and procedures this work extends."},{"cited_title":"Di Vora, A","cited_arxiv_id":null,"evidence_quote":"Characterizes the clamshell-tuned dielectric resonator at about 11 GHz that provides the cavity tuning used here."},{"cited_title":"Sikivie, Invisible axion search methods, Reviews of Modern Physics 93, 015004 (2021)","cited_arxiv_id":null,"evidence_quote":"Gives the bead-pull technique used to check the simulated form factor $C_{030}$ at a single opening angle."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports the traveling-wave parametric amplifier used as the first-stage quantum-limited readout device."},{"cited_title":"Braggio, L","cited_arxiv_id":null,"evidence_quote":"Defines the phenomenologically preferred axion-model band $E/N \\in (5/3, 44/3)$ that brackets the excluded models."}],"review_version":1}