{"id":"9446e0b6-0c63-455f-850f-dea879aaee09","arxiv_id":"1908.10978","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"First photoemission measurements of Na-intercalated TiNCl reveal a broad near-Fermi-level feature inconsistent with a simple rigid-band shift, suggesting correlated-electron behavior.","lead":"Researchers used micro-focused photoemission spectroscopy to measure the electronic structure of the superconductor Na-intercalated TiNCl for the first time. They found a broad electronic state near the Fermi level that deviates from rigid-band theory, hinting at strong electron correlations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The near-EF excess that motivates the correlated-metal claim is quantified against a spot-selected XPS composition (Cl:Ti = 0.6) that contradicts the bulk Rietveld value (Cl0.970); if that spot is unrepresentative or surface-depleted, the deviation from the rigid-band calculation is an artifact…","rationale":"This is a carefully reported exploratory µ-PES study, and the parent-compound valence band matching DFT is a solid piece of evidence. The Ti 2p core-level change toward metallic TiN-like character is also credible. The weak point is not the measurement itself but the inference from the Na-TiNCl near-EF spectrum to strong correlation. That inference relies on a quantitative comparison against a rigid-band calculation whose doping level is set by the XPS composition at one intentionally selected spot. Two independent red flags point the same way: the selection criterion (maximum near-EF intensity) introduces a strong bias, and the Cl:Ti ratio used for the case-B simulation (0.6) is inconsistent with the bulk Rietveld composition (0.970). If the Cl deficiency is a surface effect or if the selected spot is an outlier, the observed broad near-EF feature could be explained by a distribution of local doping levels or by surface defects, with no need for an incoherent correlated spectral function. The paper acknowledges position dependence and mentions composition distribution as a possible cause of broadening, but then proceeds to the correlation analogy without an explicit test excluding that simpler explanation. The suggested map-and-correlation test would settle this: if the broad near-EF feature appears even at bulk-composition positions and is independent of spot selection, the concern would be resolved; if not, the central claim would need to be weakened. This does not change the reader's conditional verdict, because the observation is still potentially important but not yet sufficiently supported to claim a correlated metal.","tokens_in":78,"tokens_out":4375,"duration_ms":92347,"concrete_test":"Perform composition-resolved µ-PES mapping across at least 10–20 micro-crystals or positions of the same Na-TiNCl batch. At each position, record the near-EF valence-band spectrum together with quantitative Na 1s/Ti 2p and Cl 2p/Ti 2p core-level ratios. Check whether the broad feature extending to ~1.5 eV appears at positions whose Cl:Ti ratio is close to the bulk Rietveld value (~0.97) and Na:Ti is ~0.16. If the broad feature occurs only at Cl-deficient or maximum-intensity spots, the correlated-metal interpretation collapses; if it persists at near-stoichiometric bulk-like spots, the concern is resolved. As a numerical checkpoint, recompute the case-B simulation using Cl:Ti = 0.97 and Na = 0.16; if the calculated near-EF area is then much smaller than the measured area, the original case-B match depended on a non-bulk Cl deficiency.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that Na-TiNCl shows a correlated-metal spectrum with an incoherent near-EF feature depends on the comparison in Fig. 5, where the calculated rigid-band spectrum is scaled to the doping level inferred from XPS core levels at one selected spot. Section 2 states that spectra were position-dependent and that 'a measured position was selected so that the intensity near EF was at its maximum.' At that spot the core-level composition was estimated as Na:Ti:N:Cl = 0.2:1:1.1:0.6, giving Cl:Ti = 0.6. This contradicts the bulk Rietveld composition Na0.160(3)TiNCl0.970(3) reported in the same section. The difference is numerically large: Cl vacancies at Cl:Ti = 0.6 would dope about 0.4 e/Ti, whereas Cl:Ti = 0.97 gives only about 0.03 e/Ti. The case-B simulation, which reproduces the measured near-EF spectral area, explicitly uses the XPS composition with Cl deintercalation; if the true bulk composition has far less Cl deficiency, the calculated electron count is too high and the apparent excess near EF may simply reflect an incorrect reference spectrum. Additionally, selecting the brightest near-EF spot biases the measurement toward the most heavily doped or most Cl-deficient micro-region; the paper itself mentions sample composition distribution as a possible origin of the broad near-EF shape but does not exclude it. Under a rigid-band picture, a spatial distribution of doping levels would also broaden the near-EF spectrum by summing spectra with different chemical-potential shifts. Thus the deviation from the rigid-band calculation, and the subsequent analogy to correlated transition-metal oxides, rests on an unvalidated assumption that the selected spot is both representative of Na-TiNCl and that its XPS-determined Cl deficiency is intrinsic rather than a surface or grain-boundary artifact.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports micro-photoemission spectroscopy (µ-PES) measurements of pristine TiNCl and Na-intercalated TiNCl (Na-TiNCl). The authors find that the valence band of TiNCl is reproduced by DFT calculations after a 1.3 eV shift, whereas Na-TiNCl shows a broad metallic feature near EF extending to about 1.5 eV. They compare this near-EF spectrum with simulated photoemission spectra based on DFT, considering Na intercalation only (case A) and Na intercalation plus Cl deintercalation (case B). The case-B simulation produces a near-EF peak area of the same order as the measurement, but the measured peak is much broader. The authors interpret the similarity of this broad feature to spectra of early transition metal oxides as evidence that Na-TiNCl may be a correlated metal with an incoherent part of the spectral function. They also report Ti 2p core-level changes indicating increased metallic character upon Na intercalation.","tokens_in":8313,"tokens_out":4087,"duration_ms":41994,"significance":"If the central claim holds, this would be the first photoemission study of the a-form TiNCl and its doped superconductor, and it would provide direct experimental information relevant to discussions of exotic superconductivity and possible correlation effects in this system. The work is also valuable for demonstrating that µ-PES can be applied to small, inhomogeneous samples that are difficult to prepare in large sizes. The authors are explicit about several limitations, including position-dependent spectra, possible composition distribution, and the possible role of Cl deintercalation, which is commendable. However, the main conclusion that Na-TiNCl shows a correlated-metal spectrum depends critically on the comparison in Fig. 5, and that comparison relies on a composition estimate that conflicts with the bulk Rietveld result. The paper does not currently resolve this tension, so the central claim is not yet solidly established.","major_comments":[{"comment":"The XPS composition used for the case-B simulation, Na:Ti:N:Cl = 0.2:1:1.1:0.6, is inconsistent with the bulk Rietveld composition Na0.160(3)TiNCl0.970(3) reported in the same section. Because Cl vacancies contribute electron carriers (about 0.4 e/Ti for Cl:Ti = 0.6 versus about 0.03 e/Ti for Cl:Ti = 0.97), the case-B spectrum is computed with a doping level that may be much higher than the actual bulk value. The paper does not resolve this discrepancy, and the central comparison in Fig. 5 depends on it. Additionally, the measured position was selected because it maximized the near-EF intensity, so the XPS composition may be biased toward a Cl-deficient, heavily doped micro-region that is not representative of the intrinsic electronic structure.","section":"Sec. 2 and Fig. 5"},{"comment":"The statement that the near-EF spectrum 'deviates from band calculations' is weakened by the construction of the comparison. The calculated spectrum uses a 1.3 eV band shift, the case-B Cl deficiency, and normalization to the total valence-band intensity; these choices are adjusted to match the measured spectral weight. After such matching, only the width discrepancy remains, and the paper itself lists composition distribution as a possible origin of the broadening. This alternative explanation is not excluded, so the claim that the broad feature indicates strong electron correlation is not yet established.","section":"Sec. 5, Fig. 5"},{"comment":"The Ti 2p fit for Na-TiNCl is obtained by adding component E only after the initial components were found insufficient to reproduce the spectrum. This additional component is not independently constrained, and the paper does not provide a physical justification beyond the need to achieve a fit. The presence of multiple Ti and Cl components supports chemical heterogeneity, but that same heterogeneity provides a plausible non-correlation explanation for the broad near-EF structure, so this fitting procedure does not support the correlation hypothesis.","section":"Sec. 3, Fig. 3(a)"},{"comment":"The comparison between the measured near-EF peak and the case-B calculation is qualitative: the observed peak is described as 'much broader' than the calculated one, without a quantitative measure such as full width at half maximum, integrated residual, or a fit of the coherent and incoherent contributions. A quantitative analysis is needed to distinguish broadening from a spatial distribution of doping (which the authors mention) from genuine incoherent spectral weight due to electron correlation.","section":"Sec. 5"}],"minor_comments":[{"comment":"Please state whether the position with maximum near-EF intensity was selected before or after the core-level composition was measured; this order is relevant for assessing possible selection bias.","section":"Sec. 2"},{"comment":"Reference 28 and reference 49 appear to be the same article (Taguchi et al., Phys. Rev. B 70, 104506 (2004)); duplicate references should be consolidated.","section":"References"},{"comment":"The figure caption does not define the component labels A–H or indicate whether the same labels are used for TiNCl and Na-TiNCl; please clarify.","section":"Sec. 3, Fig. 3(a)"},{"comment":"The figure would be easier to interpret with a vertical line marking EF and with error bars or representative multiple spectra to show reproducibility.","section":"Sec. 5, Fig. 5"},{"comment":"The word 'substracting' appears to be a typo for 'subtracting' in the description of the Shirley background subtraction.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The composition discrepancy between the Rietveld bulk value and the XPS value at the selected spot is the key issue. If the authors can provide spatially resolved composition measurements or a convincing argument that the XPS spot is representative of the bulk, the paper could be acceptable after revision. The central claim is interesting, but the current comparison in Fig. 5 rests on a baseline that may be quantitatively wrong. The paper would also benefit from a quantitative width analysis and from a more cautious discussion of the correlation interpretation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline: this is the first photoemission study of the a-form TiNCl and its Na-intercalated superconductor. The parent compound's valence band matches DFT, and the near-EF spectral weight in Na-TiNCl is a genuinely new observation. But the comparison that motivates the correlated-metal claim is shakier than the text lets on.\n\nWhat's good: they got micro-PES working on 30-micron crystals, which is honest experimental work. The Ti 2p core-level change upon Na intercalation is clean evidence for electron transfer. The parent valence band matching DFT after a 1.3 eV shift is a solid benchmark. They also explicitly report that Na-TiNCl spectra were position-dependent and that they selected the spot with maximum near-EF intensity. That is more candid than most papers.\n\nWhere it gets soft: Fig. 5. The simulated spectra are scaled to a composition measured by XPS at that same spot, with Cl:Ti = 0.6. The bulk Rietveld refinement gives Cl0.970. Those imply very different electron counts. If the selected spot is Cl-depleted or heavily doped, then the apparent excess near EF is exactly what a rigid-band calculation would predict for that local doping. The paper attributes the residual width to a distribution of doping concentrations, which is plausible, but that admission undercuts the correlation interpretation. The extra Ti 2p component added to force the fit is a minor issue, but it adds to the sense that the analysis is bending to the data.\n\nThe correlation conclusion is explicitly tentative (\"may be\", \"suggests\"), so I don't read it as overclaiming. But the evidence—one spot, no k-resolution, no doping series—does not distinguish a correlated metal from an inhomogeneous, locally doped sample. That is a genuine weakness, not a fatal one.\n\nVerdict: the paper deserves a serious referee. It is a first dataset that will be cited. For publication, I would want them to either reconcile the XPS and Rietveld compositions, show a spatial map, or at minimum reframe the near-EF feature as preliminary and soften the correlated-metal suggestion until the composition issue is resolved.\n\nBring it to a reading group if you want to discuss what counts as evidence for correlation in photoemission on inhomogeneous samples.\n\nBest.","headline":"First photoemission data on a-form TiNCl and Na-intercalated TiNCl, with a near-EF excess that is real but whose interpretation as a correlated metal rests on a spot-selected XPS composition that contradicts the bulk Rietveld value.","tokens_in":8869,"tokens_out":1922,"would_cite":true,"duration_ms":20048,"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":"Using micro-focused photoemission, this paper argues that Na-intercalated TiNCl develops a broad metallic state reaching about 1.5 eV below the Fermi level that rigid-band shifts cannot explain, and reads this as evidence that the doped…","keywords":["micro-photoemission spectroscopy","TiNCl","Na intercalation","valence band structure","strong electron correlation","rigid band model","layered superconductor","Cl deintercalation"],"falsifier":"Spatially resolve the near-Fermi spectrum across many crystallites of the same Na-TiNCl batch and compare it with locally measured Na 1s and Cl 2p intensities: if the broad 1.5 eV feature disappears wherever the Cl:Ti ratio approaches the Rietveld value of 0.970, or appears only in Na-rich or Cl-poor spots, the interpretation of an intrinsic correlated-metal spectrum is falsified.","tokens_in":7756,"feed_emoji":"🔬","tokens_out":8890,"duration_ms":84826,"temperature":0.7,"pith_summary":"This paper reports the first direct photoemission view of the electronic structure of the layered superconductor Na-intercalated TiNCl and its parent compound TiNCl, made possible by focusing the probe beam to ten micrometres so that small, air-sensitive crystals could be measured. The valence band of undoped TiNCl matches density-functional calculations, but doping with sodium produces a broad metallic feature near the Fermi level that extends roughly 1.5 eV into the occupied states, far wider than any rigid-band shift of the parent bands predicts. The paper interprets that extra spectral weight as a sign of strong electron correlation, comparing it to spectra of early transition metal oxides, and suggests Na-TiNCl may be a correlated metal. That matters because TiNCl superconducts with a large gap ratio and suspected exotic pairing, and knowing whether correlations shape the normal-state spectrum constrains the pairing mechanism.","feed_headline":"Sodium doping adds a broad metallic state to TiNCl","feed_subtitle":"Micro-photoemission finds a near-Fermi spectral shape that rigid-band theory cannot explain.","key_machinery":"The central object is the near-Fermi spectral function of Na-TiNCl, measured by micro-photoemission spectroscopy (μ-PES) with a 10 µm focused 1200 eV photon beam, which lets the authors select a single microcrystal of an air-sensitive sample that cannot be made large or homogeneous. The argument is carried by comparing this spectrum with density-functional band-structure calculations (WIEN2k with the PBE functional) and with simulated photoemission spectra for two doping scenarios, Na intercalation alone versus Na intercalation plus Cl deintercalation. Core-level curve fitting of Ti 2p and Cl 2p supplies the chemical-state and stoichiometry information that distinguishes metallic charge transfer from simple rigid-band filling.","core_discovery":"On the paper's own terms, the discovery is that the valence-band electronic structure of the parent compound TiNCl can be reproduced by density-functional band calculations, while the Na-intercalated superconductor cannot be described by a simple rigid-band shift. Ti 2p core-level spectra show a new feature near 455 eV, close to metallic TiN, indicating electron transfer into the TiNCl layers, and Cl 2p spectra indicate Cl deficiency consistent with additional \"Cl deintercalation\" doping. The near-Fermi region of Na-TiNCl shows a Fermi edge plus a broad structure centered around 0.5 eV that extends to about 1.5 eV. Simulated photoemission spectra that include only Na intercalation produce too little spectral area, while including Cl deintercalation matches the area but predicts a much narrower peak. Because the broad experimental shape resembles spectra of early transition metal oxides attributed to strong electron correlation, the paper concludes that Na-TiNCl may be a correlated metal whose spectral function has an incoherent part beside the coherent quasiparticle peak.","pith_inferences":["A doping-series study across the superconducting dome would be a natural extension: if the broad 1.5 eV feature is genuinely correlation-induced, its spectral weight and width should evolve smoothly with Na content and track Tc; the paper does not report such a series.","The analogy with doped a-HfNBr, which behaves as a Mott-like localized system, suggests the wider a-MNX family may span a crossover from localized to correlated-metallic behaviour; this generalisation goes beyond the paper's TiNCl-specific data.","Momentum-resolved photoemission on oriented films could separate disorder broadening from intrinsic incoherent weight—Cl-vacancy disorder should be angle-independent, whereas a correlation-induced incoherent component should carry a characteristic band-structure signature.","If Cl deintercalation dopes locally while Na doping fills bands rigidly, the same nominal composition made by different routes could show different near-Fermi spectra; this is a testable consequence of the two-channel doping picture."],"forward_implications":["The normal-state electronic structure of Na-TiNCl cannot be obtained by rigidly shifting the parent TiNCl bands; any quantitative model of the doped compound must add spectral weight beyond the band picture.","Carrier concentration in these samples is set by both Na intercalation and Cl deintercalation, so nominal Na content alone does not determine the doping level relevant to superconductivity.","If the correlated-metal interpretation holds, theories of the superconductivity in TiNCl—including proposals based on charge fluctuations—should treat the incoherent spectral weight as part of the pairing problem.","Micro-photoemission spectroscopy is a workable route to electronic-structure data for air-sensitive, inhomogeneous, small-crystal layered superconductors that have previously resisted photoemission study."],"supporting_citations":[{"why":"Reports the synthesis and superconductivity of TiNCl and Na-TiNCl and contains the earlier claim that the valence and conduction band shapes are similar, which this paper's near-Fermi spectrum contradicts.","marker":"3)"},{"why":"Provides the Na-intercalation synthesis route and the relation between basal spacing and Tc, plus the lattice parameters used to characterise the present samples.","marker":"23)"},{"why":"Supplies the band-structure calculations and k-mesh used to simulate the valence band and to discuss charge-fluctuation-mediated superconductivity.","marker":"31)"},{"why":"Identifies the WIEN2k code used for the density-functional calculations that reproduce the parent TiNCl valence band.","marker":"36)"},{"why":"Gives the Perdew-Burke-Ernzerhof exchange-correlation functional used in those calculations.","marker":"37)"},{"why":"Provides the Ti 2p binding energy of metallic TiN used to identify the increased metallic character of Na-TiNCl.","marker":"41)"},{"why":"Documents Cl deintercalation in related b-MNCl compounds, the basis for treating Cl deficiency as an additional doping channel.","marker":"45)"},{"why":"Attributes the near-Fermi spectral shape in early transition metal oxides to strong electron correlation, the comparison at the heart of the correlated-metal interpretation.","marker":"51)"}],"fun_headline_variants":["Na-doping turns TiNCl into a correlated metal, μ-PES finds","μ-PES shows Na-TiNCl is not a rigid-band superconductor","Electron doping exposes correlated behavior in TiNCl","Micro-photoemission challenges rigid-band model for Na-TiNCl","Na-intercalated TiNCl: correlated metal with broad spectral peak"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the small region of Na-TiNCl chosen for measurement because it showed the strongest intensity near the Fermi level represents the material's intrinsic electronic structure, rather than an atypical Na-rich or Cl-poor patch that happens to look metallic.","fun_headline_variants_meta":{"raw":{"variants":["Na-doping turns TiNCl into a correlated metal, μ-PES finds","μ-PES shows Na-TiNCl is not a rigid-band superconductor","Electron doping exposes correlated behavior in TiNCl","Micro-photoemission challenges rigid-band model for Na-TiNCl","Na-intercalated TiNCl: correlated metal with broad spectral peak"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000303,"raw_usage":{"total_tokens":1767,"prompt_tokens":996,"completion_tokens":771,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":612,"completion_tokens_details":{"reasoning_tokens":673}},"tokens_in":612,"tokens_out":771,"duration_ms":8289,"temperature":1.0,"reasoning_tokens":673,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:27:50.299031+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Spatially resolve the near-Fermi spectrum across many crystallites of the same Na-TiNCl batch and compare it with locally measured Na 1s and Cl 2p intensities: if the broad 1.5 eV feature disappears wherever the Cl:Ti ratio approaches the Rietveld value of 0.970, or appears only in Na-rich or Cl-poor spots, the interpretation of an intrinsic correlated-metal spectrum is falsified.","supporting_citations":[],"review_version":1}