{"id":"bc78736c-24e3-4303-bc92-3f8f882f4cdf","arxiv_id":"2505.00623","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In FeTe1-xSex films, terahertz conductivity reveals a hidden conduction channel with Planckian linear-in-T scattering, and the superconducting condensate spectral weight comes mainly from this channel.","lead":"The authors used terahertz spectroscopy on two iron-chalcogenide superconductor films and found that the conductivity is best described by two independent conduction channels, one with a scattering rate that grows linearly with temperature at the Planckian slope. The superconducting condensate appears to draw most of its weight from that same Planckian-scattered channel, linking the strange-metal normal state to the pairing mechanism.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Alternative optical decompositions (generalized Drude or two finite-width Drudes) are not tested; the Planckian-channel extraction and condensate-weight assignment rest on the assumed σ0+Drude model.","rationale":"The paper is a careful TDTS study with an internally consistent fitting procedure, and it has one nontrivial independent check: the extracted parameters reproduce the DC resistivity as 1/(a+b/T) over a broad temperature range. The two-fluid, σ2ν-extrapolation, and FGT determinations of Sδ are also roughly consistent. For these reasons I do not regard the central claim as invalid. The risk is underdetermination. The model has about five parameters and the measured window is narrow; a constant real offset is formally the Γ→∞ limit of a Drude, but any finite Γ_fast > 3 THz gives nearly the same low-frequency response, so the \"two channels\" are not uniquely identified by the spectra alone. The Planckian slope is fitted only between 20 K and 40 K and the coefficient is about 3, not the order-unity value often associated with the Planckian bound, so the universal-scale language is a further extrapolation. Finally, the key comparison (Sδ versus S at Tc0+2K) is presented without error bars, so the statement that the background does not contribute to the condensate is stronger than the displayed evidence. All of these issues are addressable by re-analysis with alternative models and proper uncertainty propagation, so the appropriate verdict remains CONDITIONAL.","tokens_in":9188,"tokens_out":10415,"duration_ms":116926,"concrete_test":"Obtain the raw complex conductivity spectra (σ1 and σ2 with uncertainties) behind Fig. 1 for both samples. For each temperature from roughly 5 K to 50 K, fit three models to the same data: (1) the paper's S/(Γ−iν)+σ0 plus superfluid terms; (2) a two-Drude sum S1/(Γ1−iν)+S2/(Γ2−iν) with no σ0 and with Γ2 free or fixed to values ≥3 THz; and (3) an extended-Drude model with frequency-dependent 1/τ(ω) and mass enhancement. Compare reduced χ² and AIC/BIC across models. If model (2) or (3) fits as well as model (1) while the narrow Γ1(T) is no longer linear in T, the Planckian channel is an artifact of the assumed decomposition; if the linear slope near 3 kT/h survives the two-Drude fit, the central claim is robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on decomposition of a single complex-conductivity dataset into two independent channels: a narrow Drude with Γ(T) ≈ (2.4–3) kT/h and a frequency-independent offset σ0. The model, introduced after \"We can understand the spectra...\", is fit separately at each temperature, but no comparison is made to alternative parametrizations: a generalized Drude with frequency-dependent 1/τ(ω), a sum of two finite-width Drudes without a constant offset, or a multiband Drude sum with constrained band weights. Because the data cover only 0.2–3 THz, any Drude with Γ_fast only moderately above 3 THz mimics a real constant with a small linear-in-ν imaginary part that can be absorbed into the other terms; the offset is therefore not pinned to a unique \"fast channel.\" The Planckian slope (α≈3, fitted only over 20–40 K) and the condensate-weight assignment (Fig. 4 comparing Sδ with S at Tc0+2K, with no quoted error bars) are both read out of this assumed decomposition. The DC-resistivity fit is a useful consistency check, but it does not discriminate among optical decompositions because S and Γ are outputs of the same model. If a two-Drude or extended-Drude fit is statistically equally good and yields a non-linear Γ1(T), the central claim fails.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports time-domain terahertz spectroscopy (TDTS) measurements from 0.2 to 3 THz on two FeTe1−xSex thin films with x = 0.45 and x = 0.35. The authors fit the complex conductivity with a model consisting of a Drude term, a frequency-independent offset σ0, a superfluid delta function, and a lattice background. From these fits they extract a Drude scattering rate Γ that grows linearly with temperature between 20 K and 40 K, with slopes α ≈ 3.02 (x = 0.45) and α ≈ 2.42 (x = 0.35) in units of kT/h. They determine the superfluid density via three methods—two-fluid fits, quadratic σ2ν extrapolation, and a Ferrell-Glover-Tinkham sum rule—and show that the DC resistivity is well described by ρ = 1/(a + b/T). The central claim is that the THz response contains two parallel conduction channels: a fast-relaxing channel giving the frequency-independent σ0 and a slower Drude channel whose scattering rate is Planckian, Γ ≈ 3 kT/h. The authors further claim that the superconducting condensate spectral weight is drawn mainly from the Planckian channel rather than from the fast background channel.","tokens_in":9390,"tokens_out":6407,"duration_ms":69558,"significance":"If the proposed decomposition is unique, this work would provide a striking observation of a resolved Planckian scattering channel in an iron-chalcogenide superconductor and a direct correlation between the spectral weight of that channel and the superfluid density. The paper has several strengths: it uses phase-sensitive TDTS to obtain both components of σ without Kramers-Kronig processing; it checks the optical model against measured DC resistivity; and it obtains Sδ by three independent methods whose temperature dependences are broadly consistent. However, the significance is currently conditional on model uniqueness. Because the spectral window is only 0.2–3 THz, a Drude term with a scattering rate moderately above 3 THz closely mimics a real constant with a small imaginary part, so the distinction between a true frequency-independent offset and a broad, finite-width conduction channel is not demonstrated. The paper provides no alternative fits, no statistical model comparison, and no uncertainty estimates for the extracted parameters, all of which are load-bearing for the Planckian and condensate-weight claims.","major_comments":[{"comment":"The additive decomposition σ~(ν) = S/(Γ−iν) + σ0 + (π/2)Sδ δ(ν=0) + iSδ/ν − iϵ0(ϵ∞−1)ν is the load-bearing assumption of the paper, but the authors do not test whether the 0.2–3 THz data require this form over alternative models. A Drude term with a scattering rate only moderately above the upper edge of the window (for example, 5–10 THz) approximates a real constant with a small imaginary part, and that imaginary part can be absorbed into the other terms, so Γ(T), σ0(T), and S(T) are not pinned uniquely. I request fits with (i) two finite-width Drude terms and no offset, (ii) a generalized Drude model with frequency-dependent 1/τ(ω), and (iii) a multiband Drude sum with band weights constrained by independent measurements, together with a comparison of residuals or an information criterion. The claimed linearity of Γ(T) and the assignment of condensate weight to the 'slow' channel should then be shown to be robust across these competing models.","section":"Model equation after Fig. 2; fits in Fig. 2"},{"comment":"The extracted parameters α, Γ0, S, σ0(T), and Sδ are reported as point values without uncertainties or goodness-of-fit statistics. This is especially problematic for the central comparison in Fig. 4b,d, where the claim that Sδ at low temperature 'just overshoots' S at Tc0+2K is made without error bars on either quantity. The Planckian slope α is also a fit parameter correlated with Γ0 and with the choice of the 20–40 K fitting interval. Please propagate uncertainties from the complex-conductivity fits, including systematic uncertainties in film thickness and substrate index, and report confidence intervals for α, Γ0, and Sδ.","section":"Fig. 3 and Fig. 4"},{"comment":"The authors present the successful fit of the measured DC resistivity to ρ = 1/(a+b/T) as evidence for parallel conduction channels, but this test is not independent: a and b are derived from the same fitted S and Γ, and any optical model that gives σDC = σ0 + S/Γ with Γ ∝ T will reproduce the same DC functional form. To make the DC constraint meaningful, the authors should show that the measured ρ(T) is reproduced by the optical parameters without re-fitting, and that alternative optical decompositions fail this extrapolation test.","section":"DC resistivity section ('This perspective of parallel conduction channels')"},{"comment":"The Planckian scaling is extracted from a linear fit over only 20–40 K, although Γ is plotted from 50 K down to the measurement limit. With a narrow fitted interval and no residual analysis or comparison with alternative forms (T^2, T/(T+Θ), or saturating behavior), the claim that the scattering rate is linear in T over the measurable range is not fully established. Please show the complete Γ(T) dependence, mark the fit range, and report residuals or an F-test against competing temperature dependences.","section":"Fig. 3a,c and 'Notably a clear linear dependence...' paragraph"}],"minor_comments":[{"comment":"The phrase 'the thickness difference between the the substrate the sample was grown on' contains a duplicated 'the'; it should read 'between the substrate...'.","section":"Methods, Eq. (1)"},{"comment":"The sentence 'this spectral weight is mainly drawn from of channel that displays the NFL Planckian behavior' contains a typo; it should likely read 'drawn from the channel that displays...'.","section":"Paragraph after Fig. 4"},{"comment":"The text says 'the fitting parameter Tc0 for the x = 0.45 sample needed to be 18K and for the x = 0.45, Tc0 = 16K'; the second doping value should presumably be x = 0.35, not x = 0.45.","section":"Paragraph on dirty s-wave fit"},{"comment":"The model equation after Fig. 2 is not numbered, although it is central to the paper; numbering it and referring to it explicitly in the discussion would improve readability.","section":"General notation"},{"comment":"The caption refers to 'the spectral weight of the Drude peak, S, at 2K above the onset of superconducting fluctuations', while the text uses 'Tc0 + 2K'; please clarify whether these are the same temperature and define both symbols consistently.","section":"Fig. 4 caption"}],"recommendation":"major_revision","confidential_remarks":"To the editor: this paper addresses an important question and the raw terahertz data appear to be of good quality. However, the central claim is not yet established because the two-channel decomposition is not tested for uniqueness, and the reported fit parameters lack uncertainties. The authors should be asked to perform alternative fits, report statistical comparisons, and provide error bars before publication. I also recommend that the authors carefully compare with earlier optical studies of Fe(Te,Se) that may have used different model decompositions, as this could affect the novelty assessment. With these additions, the paper could become a strong contribution to the field."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth your time. This is a careful TDTS study of two FeTe1-xSex films that decomposes the THz conductivity into a broad, weakly temperature-dependent offset plus a narrow Drude whose scattering rate is linear in T with slope ~2.4–3 kT/h. The claim that the superfluid condensate is drawn mainly from this Planckian channel is genuinely new for FTS, even though the parallel-channel framework and Planckian scattering have been reported elsewhere. What the paper does well: the data are clean, the fits look consistent, and the superfluid density is estimated three independent ways (two-fluid, quadratic σ2ν extrapolation, FGT sum rule). The DC resistivity fitting ρ = 1/(a+b/T) is a useful consistency check, not just a rhetorical flourish. The comparison with nematic-fluctuation theory is a nice touch and the authors appropriately caution that lower-frequency probes like µSR are needed to nail the T dependence.\n\nThe soft spots are real but addressable. The load-bearing assumption is the model itself: σ(ν) = Drude + constant σ0 + superfluid terms. The data span only 0.2–3 THz, so a second Drude with a moderate width or a generalized Drude with frequency-dependent scattering could plausibly mimic the same spectra. The offset is not pinned uniquely, and the extracted Γ(T) and the condensate-weight assignment depend on that choice. The paper does not test alternative decompositions. Also, the fitted parameters (α, Γ0, S, σ0) and the Sδ vs S comparison lack error bars, and the linear-T fit covers only 20–40 K. The statement that σ0 does not contribute to superconductivity is slightly stronger than what the overshoot actually shows. None of these are fatal; they are what referees should press on.\n\nReproducibility is limited because raw data and code are not provided, so confidence in the numerical extraction is moderate. Still, the paper is coherent, honest about its limitations, and cites the relevant literature. It is a useful data point for iron-based superconductors and strange-metal phenomenology, not a paradigm shift. I would send it to a serious referee: the question of alternative decompositions should be settled by fitting, and that is exactly the kind of work peer review exists to demand.\n\nRecommendation: accept into review with a request for alternative-model fits, error bars, and a broader temperature window if possible. The central idea deserves a fair hearing.","headline":"Solid THz study finding a hidden Planckian channel in FTS, but the central decomposition is not tested against alternatives; deserves serious refereeing.","tokens_in":10033,"tokens_out":1567,"would_cite":true,"duration_ms":19286,"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":"In FeTe1−xSex films, a slow conduction channel scatters at the Planckian rate ~3kT/h and supplies the superfluid condensate.","keywords":["iron chalcogenide superconductors","FeTe1-xSex","Planckian scattering","terahertz spectroscopy","parallel conduction channels","superfluid density","non-Fermi liquid","linear-in-T resistivity"],"falsifier":"A concrete falsifier would be a re-analysis of the same raw σ(ν) data using a generalized Drude model with a frequency-dependent scattering rate 1/τ(ω) but no constant offset. If such a fit reproduces the spectra at all temperatures as well as the two-channel model does, then the claim that there are two parallel channels, and hence the assignment of the condensate to a Planckian channel, would be called into question. Alternatively, extending the measurement window to higher frequencies where σ0's own Drude roll-off would appear (if it has a finite scattering rate) would directly confirm or rule out the fast channel.","tokens_in":8889,"feed_emoji":"⚡","tokens_out":5618,"duration_ms":49413,"temperature":0.7,"pith_summary":"This paper reports terahertz conductivity measurements on two FeTe1−xSex films and decomposes the spectra into two parallel conduction channels. One channel is broad in frequency and nearly temperature-independent; the other is a sharper Drude channel whose scattering rate grows linearly with temperature, Γ ≈ 3 kT/h, the Planckian limit. By comparing the spectral weight of the Drude term with the superfluid density from the Ferrell–Glover–Tinkham sum rule, the authors show that the superconducting condensate draws almost entirely from this Planckian-scattered channel. The finding ties the anomalous linear-in-T scattering to the pairing condensate in an iron-chalcogenide superconductor.","feed_headline":"Planckian channel carries the superfluid in FeTe1-xSex","feed_subtitle":"A slow Drude channel with Γ≈3kT/h supplies the superconducting condensate in FeTe1-xSex films.","key_machinery":"The central object is the additive two-channel conductivity model $\\tilde{\\sigma}(\\nu)=S/(\\Gamma-i\\nu)+\\sigma_0+\\frac{\\pi}{2}S_\\delta\\delta(\\nu=0)+iS_\\delta/\\nu-i\\epsilon_0(\\epsilon_\\infty-1)\\nu$. The Drude term supplies the slow channel, with spectral weight $S$ and scattering rate $\\Gamma$; the frequency-independent $\\sigma_0$ encodes the fast channel with $\\Gamma\\gg$ the measured range. Combining the parallel channels gives $\\sigma_{DC}=\\sigma_0+S/\\Gamma$, which with $\\Gamma=\\alpha kT/h$ yields $\\rho_{DC}=1/(a+b/T)$, matching DC data from 200 K to 25 K. The superfluid weight $S_\\delta$ is computed via the FGT sum rule $S_\\delta=S_N-S_U$ and compared with $S$ measured just above the transition.","core_discovery":"The authors find that the measured THz conductivity of FeTe0.55Se0.45 and FeTe0.65Se0.35 films cannot be captured by a single Drude response across the whole temperature range. They fit the spectra with a model consisting of a Drude term, a constant (frequency-independent) offset σ0, a superfluid delta-function term, and a lattice polarizability ϵ∞=4, constrained by DC resistivity. The Drude scattering rate extracted from fits between 20 and 40 K is linear in temperature, Γ ≈ 3.0 kT/h for the optimally doped sample and Γ ≈ 2.4 kT/h for the overdoped sample, i.e., of the Planckian form. The constant offset represents a fast relaxation channel with scattering rate far above the measurement window. The superfluid density, obtained three independent ways (two-fluid fits, σ2ν extrapolation, and the FGT sum rule), saturates at a level that just matches the normal-state spectral weight of the Drude channel, showing that the charge carriers undergoing Planckian scattering are the principal contributors to the condensate, while the fast channel stays largely unaffected by the transition.","pith_inferences":["If the two-channel picture is correct, then the linear-in-T scattering rate is a property of only a subset of carriers; a single generalized-Drude analysis that assumes a common scattering rate for all carriers would average the two channels and might mistakenly conclude the material is a conventional Fermi liquid at low T.","The identification of the slow channel with the condensate suggests a testable prediction: in FTS samples where the Planckian channel's spectral weight is suppressed (e.g., by disorder), the superfluid density should decrease proportionally.","The parallel-channel decomposition could be extended to other iron-based superconductors and to the cuprates; looking for a σ0-like fast channel in their THz spectra may reveal whether the Planckian component is generically paired with the condensate across strange metals.","A lower-frequency probe, such as microwave or muon spin rotation, could directly resolve the fast channel's own scattering rate and its temperature dependence, sharpening the claim that it is temperature-independent."],"forward_implications":["The linear-in-T scattering rate of the slow channel implies that the normal-state resistivity of FTS is non-Fermi-liquid, even though the total resistivity looks curved; the T-linear component is hidden beneath a temperature-independent offset.","The match between the condensate spectral weight and the Drude channel weight implies that the same degrees of freedom producing Planckian scattering are the ones that pair to form the superfluid.","The fast channel, represented by σ0, is essentially unaffected by superconductivity and does not contribute to the condensate, pointing to two separate conduction populations.","The observed superfluid density ratio Sδ/SN of ~0.53 (x=0.45) and ~0.30 (x=0.35) is consistent with the nematic-fluctuation-mediated superconductivity prediction with impurity scattering.","The presence of in-gap spectral weight at low temperatures is reminiscent of cold-spot excitations predicted for a nematic quantum critical point."],"supporting_citations":[{"why":"sets the high-frequency dielectric constant ε∞=4 used in all fits.","marker":"[17]"},{"why":"provides the conventional dirty-limit scenario where in-gap spectral weight is absent, the contrast case for this data.","marker":"[19]"},{"why":"supplies the σ2ν quadratic-extrapolation method for extracting superfluid density.","marker":"[23]"},{"why":"together with [27], gives the optical sum rule used to compute Sδ from normal-state and superconducting-state spectral weights.","marker":"[26]"},{"why":"defines the Ferrell–Glover–Tinkham sum rule used to compute the condensate weight.","marker":"[27]"},{"why":"predicts Sδ/SN≤0.5 for nematic-fluctuation-mediated superconductivity with impurities, the target comparison for the measured ratios.","marker":"[29]"}],"fun_headline_variants":["Planckian scattering channel supplies FeTeSe superfluid","Superfluid weight traces to Planckian Drude channel in FeTeSe","FeTeSe condensate borrowed from Planckian-scattered carriers","Two channels, but superfluid follows Planckian line in FeTeSe"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The assumption that the measured spectra are exactly the sum of a Drude term, a constant σ0, and superfluid terms—so the two channels are truly independent—is load-bearing; if a single frequency-dependent scattering rate could reproduce the same data, the extracted Planckian scaling would be an artifact of the fit.","fun_headline_variants_meta":{"raw":{"variants":["Planckian scattering channel supplies FeTeSe superfluid","Superfluid weight traces to Planckian Drude channel in FeTeSe","FeTeSe condensate borrowed from Planckian-scattered carriers","Two channels, but superfluid follows Planckian line in FeTeSe"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000554,"raw_usage":{"total_tokens":2671,"prompt_tokens":1006,"completion_tokens":1665,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":622,"completion_tokens_details":{"reasoning_tokens":1591}},"tokens_in":622,"tokens_out":1665,"duration_ms":12503,"temperature":1.0,"reasoning_tokens":1591,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:38:08.311908+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete falsifier would be a re-analysis of the same raw σ(ν) data using a generalized Drude model with a frequency-dependent scattering rate 1/τ(ω) but no constant offset. If such a fit reproduces the spectra at all temperatures as well as the two-channel model does, then the claim that there are two parallel channels, and hence the assignment of the condensate to a Planckian channel, would be called into question. Alternatively, extending the measurement window to higher frequencies where σ0's own Drude roll-off would appear (if it has a finite scattering rate) would directly confirm or rule out the fast channel.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"sets the high-frequency dielectric constant ε∞=4 used in all fits."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the conventional dirty-limit scenario where in-gap spectral weight is absent, the contrast case for this data."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the σ2ν quadratic-extrapolation method for extracting superfluid density."},{"cited_title":"& Ferrell, R","cited_arxiv_id":null,"evidence_quote":"together with [27], gives the optical sum rule used to compute Sδ from normal-state and superconducting-state spectral weights."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"defines the Ferrell–Glover–Tinkham sum rule used to compute the condensate weight."},{"cited_title":"Unconventional Superconductivity Mediated by Nematic Fluctuations in a Multi-Orbital System -- Application to doped FeSe","cited_arxiv_id":"2412.07008","evidence_quote":"predicts Sδ/SN≤0.5 for nematic-fluctuation-mediated superconductivity with impurities, the target comparison for the measured ratios."}],"review_version":1}