{"id":"ed27242b-d554-4173-87bc-99d30b048046","arxiv_id":"2412.14958","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A NbN-coated 8.4 GHz cavity loses quality factor above about 1.5 T, so NbN is not beneficial for high-field haloscope cavities.","lead":"This paper measures how a niobium-nitride coated radio-frequency cavity behaves in a strong magnetic field, and finds its quality factor drops as the field rises. The work tells axion-search experiments which superconducting coatings are worth using above 1.5 tesla.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 1.5 T crossover is drawn against a field-independent copper baseline; without a field-dependent Cu measurement, the central conclusion is not established.","rationale":"The reader's weakest assumption was the S11 normalization used to derive beta and Q0. That is a valid systematic concern, but a scalar normalization error would affect the NbN and copper measurements in a similar way if the same readout is used, so it is partially self-canceling in the comparison. The more directly load-bearing issue is the copper baseline itself: the paper compares field-dependent NbN data against what appears to be a zero-field, field-independent copper line. At 4 K and 1.5 T, copper magnetoresistance can reduce the cavity Q0 substantially, so the reported crossover at 1.5 T may be an artifact of an unfair baseline. The proposed control measurement is straightforward and would settle whether the crossover persists. The rest of the analysis is internally consistent, and the paper's conclusion is appropriately modest, so the conditional verdict remains appropriate.","tokens_in":4323,"tokens_out":9636,"duration_ms":71730,"concrete_test":"Measure Q0 of the identical uncoated copper cavity at 4 K at B = 0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0 T using the same cryostat, antenna configuration, and normalization procedure. If Q0_Cu(B) decreases by more than the roughly 2% per-point uncertainty over 0-4 T, recompute the crossover using a field-interpolated copper baseline. If the NbN/Cu crossing moves by more than about 0.2 T, the 1.5 T claim needs revision. Reporting the copper RRR would also bound the expected magnetoresistance effect.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is comparative: NbN is 'not beneficial' above 1.5 T because the measured NbN Q0 crosses the copper reference near that field. The only textual support for the copper reference is the zero-field statement that the NbN Q0 at 4 K is a factor of five above the non-coated copper value, and Fig. 6 shows a single horizontal 'Cu reference' line with no uncertainty band. A normal-metal copper cavity at 4 K is not expected to have a field-independent Q0: magnetoresistance increases its surface resistance, so Q0_Cu(B) generally decreases with B. If the plotted Cu line is the zero-field value, then at 1.5 T the true copper baseline is lower than drawn, and the crossover field could shift upward. The paper reports no field-dependent copper measurement and no magnetoresistance estimate, so the conclusion 'not beneficial above 1.5 T' is not supported unless the copper baseline was measured at the same applied fields. This is a missing control rather than an internal inconsistency in the NbN data, but it is directly load-bearing for the engineering conclusion.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper characterizes a 2 µm NbN-coated copper cavity (TM010 mode at 8.4 GHz) built for the Supax axion-haloscope testbed in Mainz. The authors measure the quality factor versus temperature, obtaining a transition temperature Tc = 10.34 ± 0.24 K and a maximum unloaded quality factor Q0 = 3.46 × 10^5 at 4 K, about five times the value of the uncoated copper cavity. They then measure Q0 as a function of applied static magnetic field up to 4 T, using a protocol in which the cavity is heated above Tc, the field is set, and the cavity is cooled back to 4 K. The data show a monotonic decrease of Q0 with increasing field, crossing the copper reference value near B = 1.5 T, and a trend similar to the published Nb3Sn data of Ref. [7]. The paper concludes that this NbN coating is not beneficial for cavity operation in static magnetic fields above 1.5 T.","tokens_in":4524,"tokens_out":21938,"duration_ms":172002,"significance":"If fully supported, the result is a useful data point for the axion-haloscope community: NbN thin films, like Nb and Nb3Sn, suffer degrading Q0 in multi-tesla fields, so they do not offer a field-robust alternative to HTS coatings. The manuscript's strengths are that Q0 is measured directly with stated statistical uncertainties, the field-cycling protocol matches the haloscope operating condition of cooling in an applied field, and the comparison to external Nb3Sn data is transparent. The qualitative trend of decreasing Q0 with field is internally consistent. However, the quantitative engineering conclusion, namely the crossover against the copper baseline at 1.5 T, depends on a field-independent copper reference that is not measured, and the quoted 2% error bars appear inconsistent with the stated FWHM resolution at the highest-Q points. This is a single-sample feasibility study; once the baseline issue is addressed, it would be a compact and useful experimental report.","major_comments":[{"comment":"The central claim that the NbN coating is not beneficial above 1.5 T rests on comparing the measured NbN Q0(B) curve with a single, field-independent copper reference line, yet the paper reports no field-dependent measurement of the uncoated copper cavity and no estimate of copper magnetoresistance at 4 K. Since the surface resistance of a normal metal generally increases with applied field, the true copper baseline at B = 1.5 T is expected to lie below the drawn line, meaning the crossover would occur at a higher field than 1.5 T and the quantitative conclusion is not established. The authors should either measure Q0(B) of the same cavity geometry in the uncoated state under the same protocol or provide a quantitative magnetoresistance estimate and state how the crossover field shifts; if the copper line is meant only as a zero-field reference, that must be stated explicitly.","section":"Sec. 4, Fig. 6"},{"comment":"The coupling coefficient beta is extracted from S11 after rescaling the measured reflection circle so that its maximum magnitude equals 1. This rescaling assumes that the cryogenic amplifier gain and cable losses are frequency-independent and fully captured by a single scalar normalization. The paper neither validates this normalization against an independent calibration nor assigns it a systematic uncertainty. Because beta enters Q0 directly through Q0 = QL(1+beta), a normalization error biases all Q0 values and shifts the crossover in Fig. 6; the field trend is likely robust to a scalar error, but the absolute scale and the crossover are not. Please cross-check the normalization, for example with a calibrated through or known-load measurement, or bound its effect on the reported Q0 values.","section":"Sec. 3, Eqs. (3)-(4), Fig. 4"},{"comment":"The stated uncertainty budget appears internally inconsistent. With an FWHM uncertainty of ±7.1 kHz at f0 = 8.4 GHz, the relative uncertainty on QL is about 7.1 kHz divided by the FWHM; for the highest-Q points (Q0 around 3.5e5, FWHM below about 100 kHz depending on beta) this gives Delta-QL/QL of order 7-15%, far larger than the roughly 2% error bars quoted for Fig. 6. Either the VNA bin width for the high-Q measurements was smaller than the stated ±5 kHz half-bin value, or the error bars in Fig. 6 are underestimated; the text should be corrected accordingly. The 1.5 T crossover also needs an uncertainty once the baseline issue is addressed, as it is currently presented without any error budget.","section":"Sec. 4, uncertainty discussion"}],"minor_comments":[{"comment":"Nb3Tn should read Nb3Sn; the typo appears in both places.","section":"Abstract and Fig. 6 caption"},{"comment":"FeSn is not an iron-based superconductor but an antiferromagnet; if an iron-based superconductor example is intended, name a compound such as FeSe or a pnictide, or drop the example.","section":"Sec. 4, last sentence"},{"comment":"The abstract refers to operation in a 2.5 T field while Fig. 6 shows measurements up to 4 T; please clarify the relation between the haloscope operating field and the field range of these data.","section":"Abstract vs. Sec. 4"},{"comment":"The orientation of the static magnetic field relative to the cavity axis and to the TM010 surface currents is not stated; the field-dependent surface resistance depends on this orientation and it affects comparability with the Nb3Sn data of Ref. [7].","section":"Secs. 2-4"},{"comment":"Please state explicitly whether the non-coated copper cavity reference value (Q0 of about 6.9e4 at 4 K) was measured on the same cavity geometry before deposition and whether it refers to zero field only; Fig. 6 would also benefit from an uncertainty band on the Cu and Nb3Sn curves.","section":"Sec. 4"},{"comment":"The conclusions rest on a single NbN-coated cavity, and the measured Tc = 10.34 ± 0.24 K is well below the bulk NbN value of about 16 K; a sentence noting the single-sample character and the likely role of film microstructure (cf. Refs. [10, 11]) would help the reader judge the generality of the field-dependence result.","section":"Sec. 4"},{"comment":"The relation Q0 = QL(1+beta) between unloaded and loaded quality factors is used implicitly but never written out; please state it explicitly when the unloaded quality factor is first introduced.","section":"Sec. 3"},{"comment":"Typos: Fig. 5 caption 'is measure with' should be 'is measured with' and 'it's max. value' should be 'its max. value'; 'Helmholz-Institute' should be 'Helmholtz-Institut'.","section":"Figs. 1, 5, and Sec. 3"}],"recommendation":"major_revision","confidential_remarks":"This is a short, direct experimental report. The main obstacle to publication is the missing field-dependent copper baseline, which supports the paper's only quantitative conclusion; a field scan of an identical uncoated copper cavity (or a quantitative magnetoresistance argument) would resolve it. The uncertainty-budget inconsistency in Sec. 4 should also be fixed. Regarding scope: the manuscript is a specialized measurement note that fits an applied-superconductivity or instrumentation venue; the editors may wish to consider whether the result is of sufficient general interest for this journal. The single-sample nature is acceptable for a feasibility study but should be stated as a limitation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this if you care about coating choices for haloscope cavities. The genuinely new piece is the measurement of the quality factor of a 2 µm NbN-coated copper cavity as a function of applied magnetic field. No one has published that curve before, and it is a useful data point: Q0 drops steadily with field, and the paper compares the trend with earlier Nb3Sn data. The measurement itself is straightforward and carefully described. QL comes from the S21 bandwidth, beta from a rescaled S11 circle fit, and the propagated uncertainty on Q0 is around 2%. The Tc fit is descriptive and not load-bearing. There is no circularity problem; the central result is directly measured.\n\nThe soft spot is the copper baseline. The paper's main quantitative claim is that NbN 'surpasses the value of the reference copper cavity at B = 1.5 T' and is 'not beneficial above 1.5 T.' But Fig. 6 shows a single horizontal line for copper, with no field dependence and no uncertainty. Copper will have magnetoresistance at 4 K, so its Q0 should fall as B increases. Without a field-dependent copper measurement or at least an estimate of the magnetoresistance, the crossover field is not actually established—it could shift upward (or down) if copper in field is worse than the zero-field value. This is a missing control, not an internal inconsistency; the NbN degradation is real. But it is load-bearing for the engineering conclusion.\n\nThere is a second, minor issue: the S11 normalization assumes the amplifier gain and cable losses are frequency-independent and fully captured by a single scalar rescale. The paper does not cross-check that against an independent calibration. That is a reasonable assumption in this kind of setup, so I would call it minor. The lack of error bars on the Nb3Sn comparison points is also worth fixing but doesn't change the message.\n\nWho this is for: the haloscope cavity community, and anyone choosing coating materials for RF cavities in static fields. It is a small paper, not a breakthrough, but it is honest and the result is new. A serious referee should see it, not a desk reject. The requested revision is straightforward: measure the copper cavity at the same fields (or estimate magnetoresistance), add uncertainty bands to the comparison curves, and soften the crossover claim until that is done. With that, the paper would be solid.\n\nRecommendation: send to peer review; it likely needs moderate revision.","headline":"Useful new NbN Q(B) data, but the 1.5 T crossover rests on a field-independent copper baseline that was never measured, so the quantitative conclusion needs an extra control before it is trusted.","tokens_in":5040,"tokens_out":3211,"would_cite":true,"duration_ms":25602,"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 2-micron NbN coating loses its quality-factor edge above 1.5 T.","keywords":["niobium nitride","superconducting radio frequency cavity","unloaded quality factor","surface resistance","magnetic field dependence","axion haloscope","thin film superconductor","TM010 mode"],"falsifier":"Re-measure the same cavity with an independent calibration of the reflection path, for instance a through-connection at the cavity flange at 4 K or a vector-network-analyzer power sweep, and recompute $\\beta$ and $Q_0$ at 0, 1, 1.5, and 2 T. If the crossover field moves by more than the few-percent quoted uncertainty, the reported 1.5 T boundary is an artifact of the scalar rescaling.","tokens_in":4129,"feed_emoji":"🧲","tokens_out":10820,"duration_ms":84663,"temperature":0.7,"pith_summary":"This paper asks whether a thin superconducting NbN coating can raise the quality factor of an axion-haloscope cavity without losing that advantage when a strong static magnetic field is applied. It reports that a 2-micrometre NbN film on copper, measured at 8.4 GHz and 4 K, reaches a maximum unloaded quality factor of $3.46\\times10^5$, about five times that of the bare copper cavity, with a critical temperature of $10.34\\pm0.24$ K. The quality factor falls as the external magnetic field rises, and the paper finds the crossover to the copper reference at 1.5 T, concluding that this coating is not beneficial in static fields above that strength. The main contribution is a quantitative field ceiling for NbN-coated cavities, which matters for choosing materials for high-field axion searches.","feed_headline":"NbN-coated cavity beats copper only below 1.5 tesla","feed_subtitle":"A 2-micron NbN coating beats copper only below 1.5 T; above that, the quality factor falls away.","key_machinery":"The central object is the unloaded quality factor $Q_0$, derived from the loaded quality factor $Q_L=f_0/\\mathrm{FWHM}$ and the coupling coefficient $\\beta$. The paper extracts $\\beta$ from the S11 reflection curve plotted in the complex plane: after rescaling the measured circle so its fitted maximum lies at magnitude one, $\\beta$ is computed from the minimum reflection through Eqs. (3)-(4). The rescaling step is the load-bearing part of the machinery, because it is what converts the pre-amplified and cable-attenuated reflection data into a quantity comparable across fields.","core_discovery":"The paper claims that its NbN-coated copper cavity has a critical temperature of $10.34\\pm0.24$ K and a peak unloaded quality factor of $3.46\\times10^5$ at 4 K, a factor of five above the uncoated copper value. It claims that the unloaded quality factor decreases with increasing static magnetic field and that at $B=1.5$ T the value crosses the reference copper value; above that field the coating no longer provides an advantage. The authors interpret this as evidence that a 2-micrometre NbN layer is not beneficial for operating a cavity in a static magnetic field above 1.5 T, and they note the trend is similar to Nb3Sn.","pith_inferences":["The paper does not say this, but its 1.5 T ceiling is tied to a single 2-micrometre film; a thickness and deposition scan would show whether the field limit is a film-quality effect or an intrinsic material limit.","A testable extension the authors leave implicit is to measure $Q_0$ versus field at several temperatures; if the field dependence sharpens near $T_c$, the loss is likely vortex-related, whereas a temperature-independent drop would point to a static field effect on the surface.","For the haloscope programme, the practical consequence not drawn in the paper is that the magnet should be tuned below 1.5 T whenever the NbN cavity is in place, or the cavity should be swapped above that field."],"forward_implications":["Below 1.5 T, the NbN coating retains a real quality-factor advantage over bare copper, so low-field axion-haloscope operation is still a use case.","Above 1.5 T, the 2-micrometre NbN coating gives no benefit over copper at 4 K, so high-field searches would need a different material.","The magnetic-field dependence resembling Nb3Sn suggests conventional superconducting thin-film coatings share a common field-loss mechanism.","The measured critical temperature and peak quality factor are specific to this film, so other thicknesses or deposition conditions could behave differently."],"supporting_citations":[{"why":"Supplies the Nb3Sn comparison points and the non-coated copper-cavity reference line used in the magnetic-field plot.","marker":"[7]"},{"why":"Shows the high-field performance of high-temperature superconducting cavities that motivates searching for alternative coating materials.","marker":"[8]"},{"why":"Describes the axion-haloscope programme for which this cavity is intended.","marker":"[9]"},{"why":"Documents the NbN thin-film deposition procedure used to coat this cavity.","marker":"[10]"}],"fun_headline_variants":["NbN beats copper only below 1.5 T","NbN coating loses its edge above 1.5 tesla","Superconducting NbN cavity fails above 1.5 T","NbN's quality factor advantage ends at 1.5 T","Magnetic field above 1.5 T erases NbN benefit"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The field ceiling rests on the assumption that the cryogenic amplifier and cables only rescale the reflection signal by one fixed number, so that the measured S11 circle can be normalized to magnitude one; if that rescaling misses frequency-dependent effects, every coupling coefficient and the 1.5 T crossover would be biased.","fun_headline_variants_meta":{"raw":{"variants":["NbN beats copper only below 1.5 T","NbN coating loses its edge above 1.5 tesla","Superconducting NbN cavity fails above 1.5 T","NbN's quality factor advantage ends at 1.5 T","Magnetic field above 1.5 T erases NbN benefit"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000425,"raw_usage":{"total_tokens":2096,"prompt_tokens":778,"completion_tokens":1318,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":394,"completion_tokens_details":{"reasoning_tokens":1242}},"tokens_in":394,"tokens_out":1318,"duration_ms":7998,"temperature":1.0,"reasoning_tokens":1242,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:44:38.502464+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-measure the same cavity with an independent calibration of the reflection path, for instance a through-connection at the cavity flange at 4 K or a vector-network-analyzer power sweep, and recompute $\\beta$ and $Q_0$ at 0, 1, 1.5, and 2 T. If the crossover field moves by more than the few-percent quoted uncertainty, the reported 1.5 T boundary is an artifact of the scalar rescaling.","supporting_citations":[{"cited_title":"Schmieden and M","cited_arxiv_id":null,"evidence_quote":"Describes the axion-haloscope programme for which this cavity is intended."},{"cited_title":"Leith, M","cited_arxiv_id":null,"evidence_quote":"Documents the NbN thin-film deposition procedure used to coat this cavity."},{"cited_title":"Thin Film (High Temperature) Superconducting Radiofrequency Cavities for the Search of Axion Dark Matter","cited_arxiv_id":"2110.01296","evidence_quote":"Supplies the Nb3Sn comparison points and the non-coated copper-cavity reference line used in the magnetic-field plot."}],"review_version":1}