{"id":"8fd027b5-44e1-4792-a6b5-050a2d6ec432","arxiv_id":"2607.20096","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"NbN films 2.8-25 nm thick show a power-law (T/T0)^b suppression of microwave superfluid stiffness at low T, with a crossover to Mattis-Bardeen behavior set by Θ(0)/Tc.","lead":"Ultrathin NbN superconducting films show a low-temperature loss of phase rigidity that follows a power law in temperature, not the exponential form of standard BCS theory. The findings propose the ratio of phase stiffness to pairing energy as the design parameter that decides when a high-kinetic-inductance film behaves conventionally or anomalously.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absolute stiffness scale for Θ(0)/Tc rests on BCS gap ratio; a direct resonance-frequency cross-check is needed to validate the crossover claim.","rationale":"The reader's weakest_assumption matches the concern I find most load-bearing. The empirical power law and the b≈2.3-2.8 range are directly supported by the log-log fits and multiple modes, so I would not lower the verdict on those grounds. However, the paper's most novel physical claim—that the crossover is organized by Θ(0)/Tc and that Eq. (5) has an offset T_off≈10 K comparable to Tc—depends on an absolute stiffness scale fixed by the BCS dirty-limit formula Eq. (1). Since that formula is explicitly labeled a baseline, using it as the 'zero-temperature superfluid stiffness' in the ratio and fit is a hidden calibration step. A direct f0-based extraction of Lk,□ from the resonator itself would either validate or falsify this scale. The authors' own caveat about needing a broader sample set does not cover this issue, because it is not statistical but systematic. Thus the verdict should remain CONDITIONAL pending the test.","tokens_in":28801,"tokens_out":13085,"duration_ms":120099,"concrete_test":"For each thickness, extract L_k,□(0) from the measured base-temperature f0 of the λ/2 or λ/4 resonator using the known physical length ℓ, L_g≈4.39 pH/□, and C≈150 pF/m (e.g., for λ/2: L_k = [1/(4ℓ^2 C f0^2)] − L_g). Compare with the transport-derived values in Table I. If any film disagrees by >20%, the absolute Θ(0) scale entering Fig. 3(c) and Eq. (5) is not validated and the T_off interpretation needs refitting; agreement within ~10% would confirm the scale.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing step is the absolute calibration of Θ(0). In Table I and Fig. 3(c), Θ(0) is obtained from Eq. (1), Lk,□(0) = ℏR□/[πΔ(0)] with Δ(0)=1.764 kBTc, via Θ(0)=(ℏ/2e)^2/[kB Lk,□(0)]. The paper itself states Eq. (1) 'serves not as a description of the films, but as the Mattis–Bardeen (mean-field BCS) baseline.' Nevertheless, this baseline is then used as the physical zero-temperature superfluid stiffness entering the ratio Θ(0)/Tc and the linear relation T0≈0.18Θ(0)+T_off (Eq. 5). If NbN has a strong-coupling gap ratio different from 1.764, or if quantum phase fluctuations already renormalize the condensate stiffness at T=0, every Θ(0) value is systematically shifted. Lk,□ is in the denominator, so a gap ratio of ~2.0 rather than 1.764 changes Θ(0) by ~13%; if the correction varies with thickness/disorder, the intercept T_off≈10 K could move by several kelvin, undermining the identification of T_off with the pairing scale. The measured resonator f0 provides an independent, geometry-based measurement of Lk,□ via the known L_g and C, but the paper does not report this cross-check. Without it, the central claim that Θ(0)/Tc governs the crossover is not quantitatively established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a thickness series of NbN superconducting microwave resonators (25–2.8 nm) with sheet kinetic inductance up to 300 pH/□. The central experimental claim is that the low-temperature fractional frequency shift follows a robust power law δf/f = −(T/T0)^b with b ≈ 2.3–2.8, in contrast to the exponential Mattis–Bardeen (M–B) prediction, and that this anomalous regime crosses over to conventional M–B electrodynamics at higher temperatures. The authors associate the power law with phase-fluctuation/collective-mode physics and organize the crossover using the ratio Θ(0)/Tc, where Θ(0) is the zero-temperature superfluid stiffness derived from the M–B kinetic-inductance formula. A linear relation T0 ≈ 0.18Θ(0) + 10 K is presented, with the offset tentatively interpreted as a pairing scale. Transport data on the 2.8 nm film show BKT-like excess conductivity, and STEM shows twin-domain nanocrystalline structure. The paper also reports microwave loss analysis separating TLS-like, quasiparticle, and residual channels.","tokens_in":29211,"tokens_out":4948,"duration_ms":51484,"significance":"If correct, the results would be an important experimental step: they identify a robust power-law suppression of superfluid stiffness in a technologically relevant high-kinetic-inductance material and propose a dimensionless ratio, Θ(0)/Tc, as a controlling parameter for the crossover between phase-fluctuation-dominated and gap-dominated electrodynamics. The strengths are the clean resonator platform, the systematic thickness series, the internal consistency of microwave and transport Tc values, and the inclusion of multiple modes for some films. These features make the anomalous power-law observation credible. The main weakness is that the absolute stiffness scale Θ(0)—and therefore the quantitative form of Eq. (5) and the Θ(0)/Tc classification—rests on the weak-coupling BCS gap ratio 1.764 in Eq. (1), which is simultaneously acknowledged not to describe the films. Without an independent calibration of Lk,□, the crossover parameter is not quantitatively established.","major_comments":[{"comment":"The absolute scale Θ(0) is obtained from Lk,□(0) = ℏR□/[πΔ(0)] with Δ(0)=1.764 kBTc, and Eq. (1) is explicitly called a baseline, not a description of the films. Yet the same Θ(0) is then used as the physical superfluid stiffness entering Θ(0)/Tc and the linear fit of Eq. (5). A strong-coupling gap ratio different from 1.764, or zero-temperature phase-fluctuation corrections to Lk,□, would systematically shift all Θ(0) values; because Lk,□ enters inversely, a gap ratio of ~2.0 would change Θ(0) by ~13%, and if the correction varies with thickness/disorder, the intercept Toff≈10 K could move by several kelvin. The measured resonator f0 and the known Lg, C provide an independent geometry-based cross-check of Lk,□, but this is not reported. Please provide this cross-check or otherwise justify the gap ratio and the use of the M–B baseline as the physical stiffness scale.","section":"Main text, Eq. (1) and after; Table I; Fig. 3(c); Eq. (5)"},{"comment":"Table I reports T0, b, and Θ(0) without uncertainties, and Eq. (5) is fitted to only four thickness points. The statement that Toff≈10 K is 'comparable to the average critical temperature' is post hoc without confidence intervals or scatter from the nominally identical films mentioned in the text. Please report fit uncertainties, sample-to-sample scatter, and a goodness-of-fit/confidence analysis for the intercept. As it stands, the identification of the offset with a pairing scale is not quantitatively supported.","section":"Table I; Eq. (5)"},{"comment":"The main text states that b remains within ≈2.3–2.8 with no clear systematic thickness dependence. However, the SM reports b=3.2490 for an additional mode of the 5 nm film (Fig. S8(a)), outside that range, and the sister-sample 2.8/3 nm modes in Fig. S7 give b≈2.43–2.46, whereas Table I lists b=2.79 for the 2.8 nm film. This internal inconsistency suggests mode-to-mode and sample-to-sample variability that is not reflected in the claim of a robust common exponent. Please clarify selection criteria, report all fitted modes, and discuss whether the exponent is truly thickness-insensitive.","section":"Main text, 'Fig. 3(b)' and 'Table I'; SM Fig. S8(a)"},{"comment":"The decomposition into a power-law term plus an M–B term is validated by the independent microwave Tc agreeing with transport Tc. However, the M–B term in Eq. (4) itself contains a free amplitude and gap scale, and the power-law term has its own amplitude and exponent, so the two components could partially compensate. The reported agreement of Tc helps, but a residual-plot or parameter-covariance analysis would strengthen the claim that the two-regime form is not merely a flexible fitting function. Please show the fit residuals and, if possible, the cross-correlation between T0, b, and the M–B amplitude.","section":"Main text, Eq. (4) and Fig. 3(c), upper panel"}],"minor_comments":[{"comment":"For the 2.8 nm film, Tc is taken from the Aslamazov–Larkin fit (7.72 K), while the text says that elsewhere Tc denotes the resistive-transition midpoint. Please state explicitly which definition of Tc enters Table I for each film and why the AL value is used for the 2.8 nm film.","section":"Fig. 1(b) and Table I"},{"comment":"The AL paraconductivity expression appears to contain a possible typographical error: the denominator should likely be ε (or sinh(ε)) with the appropriate coefficient, not the form printed. Please check the formula and its derivation.","section":"SM, Eq. (2)"},{"comment":"The loss model has four free parameters for each mode. The text already cautions that the TLS-like temperature dependence does not uniquely identify the microscopic origin. This is fine, but the abstract and conclusions would benefit from a similarly cautious phrasing regarding the 'TLS' label.","section":"Main text, Eq. (6) and End Matter"},{"comment":"The caption says the sister samples are 3 nm, while the text refers to 2.8 nm NbN films and discusses 'sister samples of 2.8 nm'. Please reconcile the thickness labeling.","section":"SM, Fig. S7"},{"comment":"The error bars on T0, b, and Θ(0) are absent from Table I and Fig. 3(c). At minimum, add error bars to the Θ(0) axes and T0 values, even if they are fit-only uncertainties.","section":"Main text, Fig. 3(c)"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of the journal and the power-law observation is likely robust. The core concern is quantitative: the absolute stiffness scale and the Eq. (5) offset are not currently established by an independent calibration, and the reported exponent range is contradicted by some of the SM data. These issues are addressable with data already in hand (f0-based Lk extraction, multi-mode fits, uncertainty propagation), so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. The core observation is solid: these NbN resonators show a clean low-temperature power law δf/f ≈ −(T/T0)^b that is inconsistent with Mattis–Bardeen, and the authors back it with multiple modes, sister samples, and log-log fits. The internal consistency check—transport Tc matching the microwave M-B component—is a good sign. The novelty is real: power-law suppression was known in strongly disordered amorphous superconductors, but a thickness-tuned NbN series with b roughly constant while T0 moves with superfluid stiffness is new, and the twin-domain nanocrystalline microstructure gives a plausible reason why b is larger here than in the amorphous case. I believe the observation.\n\nThe soft spots are where the reader's report and the stress-test both point. The absolute Θ(0) scale is computed from the same Mattis–Bardeen relation (Eq. 1) that the paper argues fails at low T. The paper even calls Eq. (1) a 'baseline,' but then uses Θ(0) from it to build Θ(0)/Tc and the T0–Θ(0) fit. If the gap ratio in these NbN films is not the BCS 1.764 value, or if zero-point phase fluctuations already renormalize the condensate stiffness, every Θ(0) in Table I is systematically shifted. Since Lk sits in the denominator, a ~13% change in the gap ratio is easy. The authors could settle this by comparing the measured f0 to a Sonnet prediction using geometry and Lk; they don't report that cross-check. This doesn't threaten the power-law observation itself, but it does put the Θ(0)/Tc crossover parameter on shakier ground than the abstract implies. Table I also carries no uncertainties on T0 or b, and Eq. (5) is a linear fit to four thickness points; the T_off ≈ 10 K reading is post hoc, though the authors say so and call for more samples. Minor: the SM shows b ≈ 3.25 for one 5 nm mode, outside the main text's 2.3–2.8 range; 'roughly constant' is doing some work.\n\nWho gets value: people working on high-kinetic-inductance circuits or disordered thin-film superconductors. The writing is clear and the supplementary is careful. It deserves peer review, with requests for the calibration cross-check, error bars on the fit parameters, and a more guarded statement about what four points can support.","headline":"Real, well-evidenced power-law stiffness suppression in NbN; the Θ(0)/Tc crossover claim needs a direct absolute-calibration cross-check.","tokens_in":29738,"tokens_out":3981,"would_cite":true,"duration_ms":37407,"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":"Ultrathin NbN films show a power-law suppression of superfluid stiffness that conventional Mattis–Bardeen theory cannot explain; the crossover is set by the ratio of stiffness to pairing energy, Θ(0)/T_c.","keywords":["superfluid stiffness","kinetic inductance","NbN thin films","Mattis–Bardeen","power-law response","Berezinskii–Kosterlitz–Thouless transition","phase fluctuations","microwave resonators"],"falsifier":"Measure the actual pairing gap and an independent superfluid stiffness (for example, through tunneling or optical conductivity for the gap, and lower-critical-field or mutual-inductance screening for the stiffness) on the same 2.8 nm and 25 nm NbN films, then re-plot T0 versus Θ(0); if the BCS gap factor is wrong, the linear relation and the Θ(0)/Tc crossover would fail or the offset would change.","tokens_in":28707,"feed_emoji":"❄️","tokens_out":10022,"duration_ms":73927,"temperature":0.7,"pith_summary":"The paper sets out to show that the low-temperature electrodynamics of disordered niobium-nitride films, from 25 nm down to 2.8 nm thick, is not the exponentially activated response of ordinary BCS superconductors. Instead, the resonant-frequency shift follows a power law δf/f = -(T/T0)^b with b ≈ 2.3–2.8, which points to low-energy collective excitations that conventional Mattis–Bardeen theory cannot account for. The same films recover conventional gap-dominated behavior at higher temperatures, and the crossover is governed by the ratio of the zero-temperature superfluid stiffness to the pairing temperature, Θ(0)/Tc. If this is right, very large kinetic inductance in NbN is inherently tied to a phase-fluctuation-dominated condensate response, and the stiffness-to-pairing ratio—not disorder alone—becomes the parameter that decides which electrodynamics a film exhibits.","feed_headline":"Power law, not BCS, sets superfluid stiffness in thin NbN films","feed_subtitle":"From 25 nm to 2.8 nm, superfluid response follows (T/T0)^b; a stiffness-to-pairing ratio sets the BCS crossover","key_machinery":"The central object is the superfluid-stiffness-to-pairing ratio Θ(0)/Tc, where Θ(0) is the two-dimensional superfluid stiffness obtained from the measured sheet kinetic inductance via Θ(0) = (ħ/2e)^2/[kB Lk,□(0)]. The kinetic inductance itself is set through the Mattis–Bardeen relation Lk,□(0) = ħR□/[πΔ(0)] with Δ(0) = 1.764 kBTc, which serves as the baseline against which the anomalous response is measured. The power-law form δf/f = -(T/T0)^b, combined with the Mattis–Bardeen term for higher temperatures, is the mechanism that separates the anomalous low-energy depletion of stiffness from ordinary thermal quasiparticles. The ratio Θ(0)/Tc ranges from 3.4 to 48 across the series and is used","core_discovery":"Across a thickness series from 25 nm to 2.8 nm, NbN microwave resonators show a low-temperature fractional frequency shift δf/f = -(T/T0)^b with b ≈ 2.3–2.8, departing from the exponential Mattis–Bardeen form. The thinnest film, with sheet kinetic inductance up to 300 pH per square, also shows transport signatures of a Berezinskii–Kosterlitz–Thouless transition—the two-dimensional vortex-unbinding mechanism by which local pairing persists while long-range phase order is lost. As thickness increases, the power-law regime narrows (from T ≲ 0.35 Tc to T ≲ 0.15 Tc) and the response crosses continuously into conventional, quasiparticle-dominated electrodynamics. The authors interpret the power la","pith_inferences":["If the linear relation T0 ≈ 0.18 Θ(0) + 10 K with the offset near Tc holds beyond this series, then the anomalous suppression would persist even for very stiff films, and the pairing scale would ultimately set the floor for T0; extending the thickness series above 25 nm would test this.","Because Θ(0) is inferred through the same Mattis–Bardeen relation containing the BCS gap factor, an independent measurement of the pairing gap (tunneling or optical conductivity) or of the stiffness (lower critical field or two-coil mutual inductance) on identical films would either confirm the crossover parameter or reveal a systematic offset.","The twin-domain microstructure suggests a growth-control handle: varying domain size or crystalline texture should change the exponent b and the offset if the pairing-amplitude distribution is the underlying mechanism, which would separate microstructural from purely compositional disorder.","The Θ(0)/Tc classification may transfer to other high-kinetic-inductance superconductors, meaning the ratio could serve as a general design guide for compact high-impedance quantum circuits."],"forward_implications":["All films in the series, including the 25 nm film whose resonance is still mostly kinetic-inductance dominated, show the same two-regime response, so high kinetic inductance in NbN comes with an intrinsic anomalous low-temperature channel.","The exponent b stays nearly constant (2.3–2.8) while T0 varies by a factor of five, so the shape of the low-energy excitation spectrum is essentially thickness-independent while its energy scale tracks the superfluid stiffness.","The temperature window of the anomalous regime shrinks with thickness, from about 35% of Tc in the 2.8 nm film to about 15% of Tc in the 25 nm film, making thicker films progressively more conventional.","The critical temperatures extracted independently from transport and from the Mattis–Bardeen component of the microwave fit agree, indicating that the two-regime fit reflects a real crossover rather than a fitting artifact.","At intermediate thickness (25 nm), resonators retain internal quality factors around 10^5 with no strong low-temperature power dependence, so large kinetic inductance and low loss can coexist away from the extreme ultrathin limit."],"fun_headline_variants":["Superfluid stiffness in NbN films follows power law, not BCS","Thin NbN: superfluid stiffness drops as power law, BKT at edge","Power-law superfluid response in NbN, BKT in 2.8 nm films","From BKT to BCS: power-law stiffness in NbN nanofilms","NbN films: phase-fluctuation regime scales as (T/T0)^b"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the absolute stiffness scale Θ(0) is accurately given by the Mattis–Bardeen relation with a BCS gap Δ(0) = 1.764 kBTc; if NbN has a different gap ratio or if zero-point phase fluctuations already suppress the zero-temperature condensate, then the Θ(0)/Tc values and the 10 K offset in the fit would shift.","fun_headline_variants_meta":{"raw":{"variants":["Superfluid stiffness in NbN films follows power law, not BCS","Thin NbN: superfluid stiffness drops as power law, BKT at edge","Power-law superfluid response in NbN, BKT in 2.8 nm films","From BKT to BCS: power-law stiffness in NbN nanofilms","NbN films: phase-fluctuation regime scales as (T/T0)^b"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000324,"raw_usage":{"total_tokens":1682,"prompt_tokens":797,"completion_tokens":885,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":790}},"tokens_in":541,"tokens_out":885,"duration_ms":62077,"temperature":1.0,"reasoning_tokens":790,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T10:47:55.825668+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the actual pairing gap and an independent superfluid stiffness (for example, through tunneling or optical conductivity for the gap, and lower-critical-field or mutual-inductance screening for the stiffness) on the same 2.8 nm and 25 nm NbN films, then re-plot T0 versus Θ(0); if the BCS gap factor is wrong, the linear relation and the Θ(0)/Tc crossover would fail or the offset would change.","supporting_citations":[],"review_version":1}