{"id":"fe1784e9-8cdf-4673-b7be-602510e30cbd","arxiv_id":"2507.07409","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"In Sr-doped La3Ni2O7 thin films, ARPES finds 1 to 2 meV leading-edge gaps in both the alpha and beta bands across Tc, with the gamma band 75 meV below EF.","lead":"This paper reports 1 to 2 meV energy gaps opening in two electron bands of superconducting nickel oxide thin films when cooled below the transition temperature. The result offers the first direct spectroscopic sign of a superconducting gap in this nickelate material and narrows down which orbitals are responsible for pairing.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 1–2 meV leading-edge signal sits at the limit of the 5 meV resolution, and the authors concede its value shifts with the EDC normalization; without a normalization-window error budget or a null-sample control, the superconductivity-induced assignment is underdetermined.","rationale":"The reader's conditional verdict rests on the same weakest assumption: the 1–2 meV leading-edge shifts are extracted with ~5 meV resolution and a normalization procedure whose own sensitivity is conceded in the text. My stress-test sharpens this into a concrete, falsifiable requirement: the authors must show that the shift survives a systematic variation of the normalization window and, ideally, that a gapless reference gives a null shift. The paper's internal evidence—temperature cycling, BCS fit, symmetrized EDCs—is suggestive but does not rule out a normal-state pseudogap or a normalization-driven artifact, especially given the conflicting persistent-gap result of ref 24. This is not a fatal objection: the data are plausible and the fermiology section is independent support. The appropriate action is to keep the verdict conditional and ask for the additional analysis or data as a condition. No ad hominem is implied; the concern is entirely about the robustness of the extraction and the uniqueness of the interpretation.","tokens_in":8842,"tokens_out":7453,"duration_ms":83519,"concrete_test":"Re-analyze the 7 K and 60 K EDCs at every reported Fermi momentum (Figs. 4b–d, 4f–h, and 5c–d) with no area normalization and with the normalization window swept from 20 to 50 meV in 5 meV increments, keeping the Fermi–Dirac division and smoothing fixed. If the extracted 7 K–60 K leading-edge shift varies by more than 0.5 meV across the sweep, or if the diagonal-direction shift becomes statistically indistinguishable from zero, then the reported 1–2 meV gap is not robustly separated from the normalization process.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the ~1–2 meV leading-edge shifts are superconductivity-induced rests on the EDC analysis in Fig. 4b–d and Fig. 5c–e. The energy resolution is ~5 meV (Extended Data Fig. 5), and the EDCs are first divided by the Fermi–Dirac distribution and then normalized to the area between 30 and 40 meV binding energy before the leading-edge position is read off. The manuscript concedes that 'the exact value of leading-edge shifts deviates slightly with the EDC normalization process.' Because the signal is only 1–2 meV, this admitted normalization sensitivity matters: if the 30–40 meV area normalization responds to temperature-dependent spectral-weight transfer—from the gap opening itself, from matrix-element variations, or from self-energy changes—the procedure can produce an apparent leading-edge shift in a system with no superconducting gap. No non-superconducting film control, no magnetic-field or disorder-controlled comparison, and no unnormalized or window-scanned extraction is shown. The temperature-cycle check (Extended Data Fig. 7) rules out sample aging, but not a normal-state pseudogap or an analysis artifact. The unresolved conflict with ref 24, which reports a persistent gap above Tc in a related film, makes the present 60 K 'restoration' an insufficient discriminator. Thus the load-bearing assumption—that the residual shift after normalization is intrinsic and tied to the superconducting condensate—is not yet secured.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports in-situ angle-resolved photoemission spectroscopy (ARPES) on Sr-doped La3Ni2O7 thin films with ambient-pressure superconductivity. It identifies two Ni-3d_{x^2-y^2}-derived Fermi surface pockets, α and β, whose measured dispersions are renormalized by a factor of 3–4 relative to DFT, and places the Ni-3d_{z^2}-derived γ band about 75 meV below the Fermi level. The central claim is spectroscopic evidence for a superconducting gap: temperature-dependent leading-edge shifts of roughly 1–2 meV in both α and β bands near the Brillouin-zone diagonal, whose temperature dependence is described by a BCS gap function, and which the authors interpret as deviating from the conventional d-wave gap structure.","tokens_in":9112,"tokens_out":3041,"duration_ms":39755,"significance":"If the leading-edge gap claim is robust, this is a milestone result: it would provide the first direct spectroscopic signature of superconductivity-induced gap opening in the bilayer nickelate La3Ni2O7 and would support the view that Ni-3d_{x^2-y^2} orbitals dominate the pairing, while constraining a large body of theoretical work. The fermiology part is solid and well supported by multi-photon-energy Fermi-surface maps combined with DFT comparison; the orbital fillings and the γ-band position are useful and will inform model constructions. The temperature-cycling check (Extended Data Fig. 7) and the explicit discussion of disorder effects are also strengths. The central gap claim, however, is at the resolution limit of the experiment, and the present analysis does not yet secure the assignment to superconductivity.","major_comments":[{"comment":"The load-bearing claim is that ~1–2 meV leading-edge shifts at k_F are superconductivity-induced, but the instrumental energy resolution is ~5 meV (Extended Data Fig. 5), and the authors explicitly concede that 'the exact value of leading-edge shifts deviates slightly with the EDC normalization process.' Because the shifts are smaller than the resolution and comparable to the width of the normalization window effects, the manuscript must provide a quantitative normalization-window scan and a full error budget that includes the 30–40 meV area normalization, the Fermi-Dirac division, and the background subtraction. As written, the admitted normalization sensitivity leaves open the possibility that the apparent low-temperature shift is produced by temperature-dependent spectral-weight redistribution rather than by a superconducting condensate.","section":"Leading-edge shifts measurements; Fig. 4b–d and Fig. 5c–e"},{"comment":"No control experiment is shown that would distinguish a superconductivity-induced gap from a normal-state pseudogap or from an analysis artifact. The temperature-cycle check (Extended Data Fig. 7) rules out sample aging but does not rule out a normal-state gap that persists above T_c, especially because a related film study (ref. 24) reports a persistent energy gap extending well above the superconducting transition. The restoration of the spectral-weight depletion at ~60 K in the present sample is therefore not by itself a sufficient discriminator; a comparison with a non-superconducting film, a magnetic-field or disorder-controlled set of measurements, or an unnormalized extraction is needed.","section":"Leading-edge shifts measurements; Discussion and ref. 24"},{"comment":"The BCS fit used to describe the temperature dependence of the leading-edge shifts is a two-parameter consistency check (gap magnitude and T_c are both free), not an independent validation. Moreover, the authors note that leading-edge shifts in electron-doped cuprates underestimate the actual gap by a factor of 2–3; therefore the 1–2 meV values cannot be read directly as gap magnitudes and cannot, by themselves, support a quantitative comparison against d-wave nodal structure. The manuscript should state clearly that the BCS description is illustrative, and should separate the qualitative observation of a temperature-dependent shift from any quantitative symmetry assignment.","section":"Fig. 4d and h; BCS fit"},{"comment":"The claim that the gap is nonzero both on and slightly away from the zone diagonal and hence 'deviates from the conventional d-wave gap structure' is weakened by the authors' own admission that disorder can fill d-wave nodes and produce a small residual gap. As presented, the momentum-space evidence is limited to two cuts near the diagonal; it does not establish a finite gap precisely at the node versus a disorder-filling effect. A momentum-resolved gap map or at least a measurement across the full nodal region is required before the symmetry statement can be regarded as supported.","section":"Discussion; Fig. 5g"}],"minor_comments":[{"comment":"The main text refers to 'similar leading-edge shifts are observed from symmetric Fermi momenta along the same cut (Fig. 5f)', but in the Fig. 5 caption panel f is the calculated Fermi surface overlay and panel g is the schematic. The reader cannot locate the comparative EDC data; the reference should be to Extended Data Fig. 10 or the figure panels should be renumbered.","section":"Fig. 5 caption and main text"},{"comment":"The error bars are described as coming from the fitting procedure and Fermi-energy calibration, but the dominant uncertainty in the present context is likely the EDC normalization choice; the figure captions should state whether the error bars include a normalization-window variation.","section":"Fig. 4d and h"},{"comment":"The phrase 'the exact value of leading-edge shifts deviates slightly with the EDC normalization process' should be supported by a quantitative statement, for example the range of extracted shifts as the normalization window is varied between 20–50 meV and 40–60 meV.","section":"Main text, Leading-edge shifts measurements"},{"comment":"The background subtraction for the γ-band analysis relies on averaging EDCs in the momentum range 1.5–1.7 1/Å, but the caption does not state how sensitive the resulting ~75 meV band-top position is to the chosen background momentum range; a brief sensitivity statement would help.","section":"Fig. 3e–f"},{"comment":"The text says the film thickness is 2 u.c. unless otherwise mentioned, while Fig. 1b shows a 3-u.c. film; this is understandable as a different representative sample, but the figure caption should explicitly note the thickness difference to avoid confusion.","section":"Introduction and Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The fermiology measurements and DFT comparison are strong and likely publishable on their own. The superconducting-gap claim is the paper's headline but currently rests on 1–2 meV shifts at a 5 meV resolution with an admitted normalization sensitivity and no non-superconducting control. I would encourage the authors to either add a normalization-window error analysis and a control measurement, or substantially reframe the central claim as a tentative observation that motivates higher-resolution follow-up. The conflict with ref. 24 should also be addressed explicitly, as the persistence of a gap above T_c in similar films is a direct challenge to the interpretation offered here."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe thing to know about this paper: it is the first ARPES report of a temperature-dependent leading-edge shift across Tc in ambient-pressure La3Ni2O7 films, and it does a good job on the fermiology. The multi-photon-energy Fermi surface maps are convincing, the α and β pockets agree with DFT under 2% strain, and the γ band sits ~75 meV below EF, which aligns with one prior study and reasonably settles that debate. The measured orbital fillings and band renormalization of 3-4 are useful constraints. The temperature-cycle check rules out sample aging, and the shifts evolve roughly like a BCS gap with Tc ~40 K. That is real evidence.\n\nThe soft spot is the one the authors themselves flag: the gap signal is 1-2 meV against a 5 meV resolution, and the exact leading-edge shift changes with the EDC normalization. The stress-test note worries this could be an artifact of the normalization procedure or a normal-state pseudogap, and I think that concern is fair, though not damning. The symmetrized EDC depletion at EF that restores above Tc is harder to explain away than a single number, but the absence of a coherence peak, the admitted disorder, and the analysis being limited to two momentum cuts mean the pairing symmetry conclusion is not yet secured. The unresolved discrepancy with ref 24, which sees a persistent gap above Tc, also needs a direct response rather than a mention.\n\nThere are no fabricated results here. The circularity burden is low; the BCS fit is a consistency check, not a prediction. The paper is honest in its limitations and does not overclaim beyond leading-edge shifts.\n\nWho gets value: anyone working on bilayer nickelate superconductivity, especially on the orbital character of the Fermi surface and the debate over the γ band. The gap claim will be tested quickly by other groups, and the data as presented are enough to warrant a serious referee.\n\nMy recommendation: send it to review. A referee should ask for a normalization-window error budget, more momentum cuts, and ideally a comparison with a non-superconducting film or a magnetic-field/impurity control. But the core experiment is competently done and the question is important enough that this deserves refereeing, not a desk reject.","headline":"A serious ARPES study of bilayer nickelate films whose fermiology is solid and whose 1-2 meV gap claim is plausible but not yet airtight.","tokens_in":9716,"tokens_out":2106,"would_cite":true,"duration_ms":22384,"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":"Superconductivity in strained nickelate films leaves a 1–2 meV leading-edge gap in both Ni-3d_{x^2-y^2} Fermi pockets, pointing to orbital-selective pairing and away from a simple d-wave gap.","keywords":["nickelate superconductors","La3Ni2O7 thin films","angle-resolved photoemission spectroscopy","superconducting gap","leading-edge shift","orbital-selective pairing","Fermi surface","bilayer nickelate"],"falsifier":"A decisive test would be to measure the same films with sub-meV energy resolution and check whether a coherence peak develops and the leading-edge shift closes exactly at the superconducting transition; alternatively, re-analyzing the present data with different normalization windows and seeing the shifts vanish would show the claimed gap is not intrinsic.","tokens_in":8623,"feed_emoji":"⚛️","tokens_out":9929,"duration_ms":103944,"temperature":0.7,"pith_summary":"This paper reports a direct spectroscopic signature of the superconducting gap in ambient-pressure bilayer nickelate thin films. Using in-situ angle-resolved photoemission on Sr-doped $\\mathrm{La}_3\\mathrm{Ni}_2\\mathrm{O}_7$ films, it shows that the two Ni-$3d_{x^2-y^2}$-derived Fermi pockets, an electron pocket $\\alpha$ and a hole pocket $\\beta$, both develop leading-edge shifts of about 1–2 meV when cooled through the 37 K superconducting transition. The gap appears both at and slightly away from the Brillouin-zone diagonal, so it does not follow the conventional $d_{x^2-y^2}$-wave nodal structure of cuprates. The authors take this as evidence that Ni-$3d_{x^2-y^2}$ orbitals dominate the pairing, while the Ni-$3d_{z^2}$-derived $\\gamma$ band sits about 75 meV below the Fermi level and plays no direct role.","feed_headline":"A 1–2 meV gap opens at Tc in nickelate films","feed_subtitle":"Both nickelate Fermi pockets shift below 37 K, signaling that 3d(x²−y²) orbitals drive pairing.","key_machinery":"The central mechanism is leading-edge shift analysis of ARPES energy distribution curves: EDCs taken at Fermi momenta are divided by the Fermi-Dirac function, symmetrized, normalized to the 30–40 meV binding-energy window, and compared between temperatures below and above $T_\\mathrm{c}$; the resulting shifts are then fitted with the BCS gap function. This converts a small spectral movement of order 1–2 meV into a gap estimate. The $\\alpha$ and $\\beta$ Fermi pockets, with their Ni-$3d_{x^2-y^2}$ orbital character, are the objects whose temperature evolution carries the argument.","core_discovery":"On superconducting $\\mathrm{La}_{2.79}\\mathrm{Sr}_{0.21}\\mathrm{Ni}_2\\mathrm{O}_7$ thin films, in-situ ARPES resolves a Fermi surface made of an electron pocket $\\alpha$ centered at $(0,0)$ and a hole pocket $\\beta$ centered at $(\\pi,\\pi)$, both Ni-$3d_{x^2-y^2}$-derived, with orbital fillings of $0.11\\pm0.02$ electron/Ni and $0.66\\pm0.03$ hole/Ni. The bands show moderate electron correlations with a renormalization factor of 3–4 relative to density-functional theory. Upon cooling through the superconducting transition, Fermi-Dirac-divided and symmetrized energy distribution curves at individual Fermi momenta lose spectral weight at $E_\\mathrm{F}$, and the leading edges shift by about 2 meV near the zone diagonal and about 1 meV on the diagonal. The temperature evolution follows a BCS energy-gap function with $T_\\mathrm{c}\\approx 40$ K and gap size about 2 meV. Because the shift is present along the diagonal, the gap deviates from the conventional $d_{x^2-y^2}$-wave structure; the authors attribute the residual nonzero value there to disorder filling the nodes. The $\\gamma$ band, mainly Ni-$3d_{z^2}$, lies $75\\pm5$ meV below $E_\\mathrm{F}$, contrary to predictions that it crosses the Fermi level under pressure.","pith_inferences":["If disorder-filling is the correct explanation, then cleaner films with sharper spectra should reveal nodal zeros, making the measured diagonal shift an upper bound on the intrinsic nodal gap.","Because the authors note the exact shift value varies with the EDC normalization, the 30–40 meV normalization window could be varied to check which part of the signal is intrinsic; a genuine gap feature should survive renormalization.","The same in-situ ARPES measurement could be extended across the Sr-doping series and to other layered nickelate phases to see whether the leading-edge gap and $\\gamma$-band position track $T_\\mathrm{c}$ or the Fermi-surface volume.","A quantitative comparison of the momentum-dependent gap shape with theoretical pairing candidates, such as $s_\\pm$-wave proposals, would use these data as a constraint, though such a comparison is not made here."],"forward_implications":["A leading-edge gap in both $\\alpha$ and $\\beta$ bands implies that Ni-$3d_{x^2-y^2}$ orbitals participate in the superconducting condensation, so minimal models of bilayer nickelates should include both pockets.","The small measured shift and the analogy with electron-doped cuprates imply the true superconducting gap may be several meV, which higher-resolution photoemission or tunneling could test.","The $\\gamma$ band sitting about 75 meV below $E_\\mathrm{F}$ rules out a Fermi-surface-crossing role for Ni-$3d_{z^2}$ in these strained films.","A nonzero gap near the zone diagonal constrains the pairing symmetry away from a simple $d$-wave; either the gap is nodeless or disorder fills the nodes.","The BCS-like temperature dependence of the shift supports the interpretation that the feature is the superconducting gap rather than a normal-state pseudogap."],"supporting_citations":[{"why":"Established that ambient-pressure superconductivity can be stabilized in compressively strained La3Ni2O7 thin films, the material platform measured here.","marker":"4,5"},{"why":"Bilayer two-orbital model that predicts a Ni-3d_{z^2} γ band near EF under pressure; the theoretical reference for orbital contributions.","marker":"8"},{"why":"Prior ARPES claiming the γ band crosses EF in (La,Pr)3Ni2O7 films, the conflicting result this work addresses.","marker":"22"},{"why":"Prior ARPES on superconducting La2PrNi2O7 films finding the γ band about 70 meV below EF, providing the comparison for band positions and renormalization.","marker":"23"},{"why":"Earlier laser-ARPES report of a persistent nodeless gap above Tc, the anomaly this work helps interpret.","marker":"24"},{"why":"Bulk ARPES on La3Ni2O7 giving orbital-dependent renormalization baselines.","marker":"25"},{"why":"X-ray absorption data indicating d_{x^2-y^2} orbital dominance, supporting the orbital interpretation of pairing.","marker":"34"},{"why":"Electron-doped cuprate studies showing leading-edge shifts underestimate the superconducting gap by a factor of 2–3, used to extrapolate the true gap size.","marker":"35,36"}],"fun_headline_variants":["Nickelate films reveal 1–2 meV superconducting gap","ARPES sees 1–2 meV gap open in nickelate films","Nickelate gap at Tc defies conventional d-wave","Direct ARPES gap evidence in Sr-doped nickelate films","Nickelate superconductivity gap measured directly"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The superconductivity-gap conclusion rests on the assumption that the ~1–2 meV leading-edge shifts, measured with about 5 meV energy resolution and normalized to the 30–40 meV window, are intrinsic spectral changes at $T_\\mathrm{c}$ rather than artifacts of normalization, Fermi-Dirac division, background subtraction, or temperature-dependent matrix elements.","fun_headline_variants_meta":{"raw":{"variants":["Nickelate films reveal 1–2 meV superconducting gap","ARPES sees 1–2 meV gap open in nickelate films","Nickelate gap at Tc defies conventional d-wave","Direct ARPES gap evidence in Sr-doped nickelate films","Nickelate superconductivity gap measured directly"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00125,"raw_usage":{"total_tokens":5283,"prompt_tokens":1263,"completion_tokens":4020,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":879,"completion_tokens_details":{"reasoning_tokens":3935}},"tokens_in":879,"tokens_out":4020,"duration_ms":31634,"temperature":1.0,"reasoning_tokens":3935,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:41:56.425673+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be to measure the same films with sub-meV energy resolution and check whether a coherence peak develops and the leading-edge shift closes exactly at the superconducting transition; alternatively, re-analyzing the present data with different normalization windows and seeing the shifts vanish would show the claimed gap is not intrinsic.","supporting_citations":[],"review_version":1}