{"id":"e3cd0409-314e-44bb-b17c-2b2c0d7e2cba","arxiv_id":"2508.17861","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"STM spectra in 4Hb-TaS2 track a filling-controlled Mott transition, with Hubbard bands emerging, a quasiparticle peak collapsing, and nonsuperconducting puddles forming.","lead":"Researchers used a scanning tunneling microscope to watch electrons in layers of 4Hb-TaS2 gradually become more insulating as they approach a Mott transition, and found that these correlated regions also lose superconductivity. The result is a quantitative, doping-controlled view of Mott physics and its antagonism with superconductivity.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Filling axis n=0.2→1 is not independently calibrated; if n is assigned through the Hubbard-model fit itself, the Brinkman-Rice trajectory is partly circular and the CDW/stacking-disorder alternative is not excluded.","rationale":"The reader's weakest assumption is exactly the load-bearing point: the spectral evolution is interpreted as a filling-controlled approach to half filling, with the gap/bands identified as Mott-Hubbard features rather than CDW-gap or stacking/surface effects. I agree that this is the decisive assumption. The abstract's own language supports this reading: it reports a quantitative filling of 0.2 electrons per site and asserts that interlayer charge transfer tunes doping, but no external calibration is presented. Without such calibration, the central claim risks circularity because the same spectral features are used both to define the filling scale and to validate the Hubbard-model trajectory. The CDW alternative is not merely a consensus disagreement; it is a concrete competing explanation for gap-like features in 1T-TaS2 layers, and the abstract does not provide discriminating evidence. Given the garbled full text, the methods and fits cannot be checked, which further supports a conditional rather than unconditional verdict. I do not see an internal inconsistency or an obviously fatal flaw; the paper could be correct. The fix is an external filling/doping measurement. Therefore the reader's CONDITIONAL verdict should stand.","tokens_in":25317,"tokens_out":4708,"duration_ms":65333,"concrete_test":"Measure the same crystal batch with micro-ARPES or soft-x-ray core-level photoemission: determine the 1T-derived band filling and chemical-potential shift as a function of the same control parameter used in the STS series. Concretely, compute n from the lower-Hubbard-band spectral weight relative to the total 1T bandwidth and compare with the n values claimed in the STS series (0.2→1). If the ARPES-derived n does not systematically increase across the series, or if the gap size/peak position tracks the CDW gap energy rather than U/W, the filling-controlled Mott interpretation is falsified. A purely internal check: refit the STS series with n fixed from ARPES; if the extracted U/W is no longer consistent, the original n was circular.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central inference is that the STS spectral series traverses a filling-controlled single-band Hubbard transition from n≈0.2 to half filling. That requires an externally anchored n-axis. The abstract states n=0.2 per site and claims fine tunability via interlayer charge transfer, but gives no independent determination of n. If n is extracted by matching Hubbard-model spectra, then the observation that the spectra follow Brinkman-Rice is, to that extent, an artifact of the calibration rather than a test. In 1T-TaS2 layers, the relevant energy window also contains the known commensurate CDW gap, and stacking faults or surface potential variations can produce spatially varying pseudo-gaps that resemble Mott-Hubbard band evolution. The supplied full text is garbled and contains an unrelated arXiv header, so the fitting procedure, raw dI/dV maps, and comparisons could not be audited. The claim would be secure if an independent probe fixed local filling and ruled out CDW/disorder; without that, the Mott-transition conclusion is underdetermined.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports scanning tunnelling spectroscopy of the van der Waals heterostructure 4Hb-TaS2, whose 1T layers are proposed to realize a filling-controlled Mott transition tuned by interlayer charge transfer. The authors claim to continuously track the spectral function from a depleted narrow band at approximately 0.2 electrons per site toward half filling, observing the gradual emergence of Mott-Hubbard bands and the Brinkman-Rice collapse of the central quasiparticle peak. They further connect these Mott renormalization effects to spatially resolved suppression of superconducting pairing, leading to 'nonsuperconducting, paramagnetic puddles.' The central claims are that the single-band Hubbard model quantitatively organizes the low-energy spectra and that Mottness locally destroys superconductivity.","tokens_in":25442,"tokens_out":2332,"duration_ms":29115,"significance":"If the central claim holds, the paper would provide a rare, direct spectroscopic visualization of a filling-controlled single-band Hubbard transition in a bulk van der Waals heterostructure, and would place 4Hb-TaS2 as a tunable platform with predictive Hubbard-model behavior. The connection between local Mottness and the spatial suppression of superconductivity is also potentially important for understanding coexisting correlated and superconducting orders in natural heterostructures. The manuscript has attractive ideas: the use of interlayer charge transfer as a continuous filling knob is creative, and the energy resolution claim is interesting. However, at present the evidence is not fully auditable. The filling axis lacks an independent calibration, no alternative-model comparison (e.g., CDW-gap or disorder-driven pseudogaps) is provided, and the supplied full text is corrupted, preventing verification of raw spectra, fitting details, and error analysis. The significance is therefore prospective rather than demonstrated.","major_comments":[{"comment":"The central claim that the STS series tracks a filling-controlled Mott transition from n≈0.2 to half filling rests on assigning the local per-site filling. No independent calibration is presented; if n and U/W are obtained by fitting the same dI/dV spectra that are later shown as Brinkman-Rice evolution, the agreement is partly built into the input. Please provide an external anchor for n (e.g., ARPES-derived band filling, core-level shifts, or Hall data), and state explicitly what is fitted and what is predicted. Without this, the phrase 'predictive power of the Hubbard model' is not justified.","section":"Abstract (filling axis)"},{"comment":"The spectra are said to 'unambiguously demonstrate' Mott-Hubbard bands, but the 1T-TaS2 layer hosts a known commensurate CDW gap, and stacking disorder or surface potential variations can produce pseudogap-like spatial variations. No comparison of the observed gap evolution against CDW or disorder models is shown. The paper would need alternative-model fits, or at least a clear argument for why CDW features are ruled out, before the Mott assignment is load-bearing. The full text supplied does not contain these fits, so this cannot currently be audited.","section":"Abstract (Mott vs CDW)"},{"comment":"The conclusion that Mott renormalization leaves 'nonsuperconducting, paramagnetic puddles' introduces a new real-space entity. I find no quantitative definition or statistical evidence in the provided text: how is the local superconducting pairing potential determined from STS maps, what threshold defines a nonsuperconducting puddle, and what are the error bars? As stated, the nanoscale-puddle claim is not supported.","section":"Abstract (puddles)"}],"minor_comments":[{"comment":"The supplied full text is garbled and contains an unrelated arXiv header (2508.17855 [cs.CL]) along with many unreadable sections. Please provide a clean, correctly compiled manuscript so that figures, tables, and methods can be evaluated.","section":"Full text"},{"comment":"The word 'unambiguously' is too strong given that raw dI/dV spectra, fitting details, and alternative-model comparisons are not shown; consider softening the claim unless all supporting data are presented in the main text or supplement.","section":"Abstract"},{"comment":"Figure and table captions are largely unreadable in the supplied version; axis labels, color scales, and error bars need to be clearly visible and described.","section":"Figures"}],"recommendation":"major_revision","confidential_remarks":"The full text provided to me is corrupted, so I have based this report primarily on the abstract and the accompanying reviewer materials. If the actual submission is clean, the three major comments still stand: independent filling calibration, CDW/disorder alternative modeling, and quantitative support for the puddle claim. I would ask the editor to ensure the manuscript version sent to referees is complete and readable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my take. The core idea is appealing: using interlayer charge transfer in 4Hb-TaS2 to sweep the 1T layers across a filling-controlled Mott transition while tracking the spectral function with STM, and showing that Mott renormalization locally kills superconductivity. If the data support it, that's a real advance—continuous, filling-resolved spectral evolution through a Mott transition in a bulk heterostructure, plus a nanoscale map of Mottness suppressing pairing. The abstract reads coherently, and the experiment is clever. Credit is due for that framing.\n\nBut I can't verify anything from what we have. The full text we received is garbled and includes an unrelated arXiv header, so methods, fits, and figures are out of reach. On the abstract alone, two concerns stand out, and neither is minor. First, the filling axis: the claim of 0.2 electrons per site and the approach to half filling need independent calibration. If n is assigned by matching Hubbard-model spectra to the very data being displayed, then the Brinkman-Rice trajectory is partly an output of the fitting, not an independent test. The abstract calls this predictive power, which is circularity risk. Second, the energy scale in question in 1T-TaS2 layers is exactly where the known commensurate CDW gap sits. Without an explicit comparison against CDW-gap or disorder/stacking-fault models, 'Mott-Hubbard bands' is underdetermined. These are load-bearing concerns, not cosmetic—they are exactly what would convert raw dI/dV maps into a Mott-transition narrative.\n\nTo be fair, I can't rule out that the full paper addresses both. A well-executed version would include an external filling probe (ARPES band positions or core-level shifts), a CDW-gap fit as a competing model, and raw spectral series with error bars. If those exist, the paper could be strong. As it stands, the abstract overclaims—'unambiguously demonstrate' is not justified by the evidence shown.\n\nI'd still send it to peer review: the claim is important enough and plausible enough that referees should see the full data. But I'd tell them to insist on the independent calibration and model comparison. I wouldn't cite it yet, and I don't think we can judge soundness until we see a readable manuscript.","headline":"Potentially important experiment, but as submitted the evidence is not auditable, and the filling axis plus Mott-vs-CDW identification look under-constrained.","tokens_in":26170,"tokens_out":2587,"would_cite":false,"duration_ms":32795,"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":"Scanning tunneling spectra of 4Hb-TaS2 trace a filling-controlled Mott transition and show that Mottness suppresses superconducting pairing at the nanoscale.","keywords":["Mott transition","Brinkman-Rice scenario","scanning tunneling spectroscopy","4Hb-TaS2","van der Waals heterostructure","Hubbard model","nanoscale puddles","superconductivity suppression"],"falsifier":"Measure the local 1T-layer electron filling independently, for example via core-level binding-energy shifts or angle-resolved photoemission band positions on the same crystals, and overlay it on the dI/dV maps: if the spectral gap and the collapse of the quasiparticle peak occur at constant filling or track the known charge-density-wave order instead of the filling axis, the Mott-transition interpretation fails.","tokens_in":25040,"feed_emoji":"🔬","tokens_out":9649,"duration_ms":109434,"temperature":0.7,"pith_summary":"This paper uses scanning tunneling spectroscopy on 4Hb-TaS2, a naturally stacked van der Waals crystal that alternates strongly correlated 1T-TaS2 layers with superconducting 1H-TaS2 layers, to make a quantitative case for a filling-controlled Mott transition. It claims that interlayer charge transfer tunes the 1T layer's electron filling from about 0.2 electrons per site toward half filling, and that the measured spectra evolve exactly as the Brinkman-Rice picture of the Hubbard model predicts: Mott-Hubbard bands appear while the central quasiparticle peak sharpens and then vanishes. The paper further claims that where this renormalization is strongest, the superconducting pairing potential is destroyed, leaving nanoscale nonsuperconducting, paramagnetic puddles. If correct, this would make the single-band Hubbard model a quantitatively predictive description of a bulk correlated material and would show that Mottness acts locally to suppress superconductivity.","feed_headline":"Spectra trace Mott transition then pairing collapse in 4Hb-TaS2","feed_subtitle":"Charge transfer tunes the 1T layer toward half filling, where superconductivity fades to nanoscale puddles.","key_machinery":"The central object is the local differential conductance spectrum dI/dV measured by the scanning tunnelling microscope, used as a direct window into the single-particle spectral function of the 1T layers. The mechanism that makes the experiment possible is interlayer charge transfer: the metallic 1H layers donate or withdraw electrons from the 1T layer, letting the filling move along the Hubbard-model phase diagram without external gating. The theoretical spine is the Brinkman-Rice scenario, in which quasiparticle weight collapses as the Mott transition is approached and spectral weight is transferred into lower and upper Hubbard bands. Spatial maps of these spectral features are what connec","core_discovery":"The paper reports scanning tunnelling spectroscopy on 4Hb-TaS2, a naturally stacked van der Waals crystal in which strongly correlated 1T-TaS2 layers alternate with metallic, superconducting 1H-TaS2 layers. It claims that interlayer charge transfer tunes the effective electron filling of the 1T layer from about 0.2 electrons per site toward half filling, and that the measured dI/dV spectra follow the evolution predicted by the Brinkman-Rice scenario of the Hubbard model: a narrow band depletes, Mott-Hubbard bands emerge, the central quasiparticle peak sharpens and then collapses at the Mott transition. The same data are read as showing that in regions where low-energy electrons are most stro","pith_inferences":["An independent calibration of the local filling axis, for instance core-level binding-energy shifts or angle-resolved photoemission on the same crystals, could confirm that the spectral evolution is doping-driven rather than caused by the known charge-density-wave gap of 1T-TaS2.","If the filling-controlled picture is right, tuning the interlayer coupling by chemical pressure, stacking sequence, or twist angle should move the system toward or away from the Mott boundary, with superconductivity weakening as half filling is approached.","A similar spectral analysis applied to other alternating van der Waals heterostructures could test whether nanoscale Mott puddles are a generic way for strong correlations to suppress superconductivity."],"forward_implications":["4Hb-TaS2 becomes a bulk, gate-free testbed for the Hubbard-model spectral function, with local tunneling maps serving as a spatially resolved phase diagram across the filling axis.","Superconductivity in this material should be spatially inhomogeneous: nanoscale Mott puddles suppress pairing, so the current path is likely percolative.","The observed local correlation between spectral renormalization and the disappearance of the gap means superconducting order is set by local proximity to half filling, not only by the average doping.","The heavily doped Mott regime in such heterostructures is identified as a place to search for new correlated ground states."],"supporting_citations":[],"fun_headline_variants":["Spectra map Mott transition and pairing collapse in 4Hb-TaS2","Nano puddles reveal superconductivity dies at Mott threshold in 4Hb-TaS2","Mott transition seen in spectra as superconducting puddles vanish","Doping tunes 4Hb-TaS2 toward Mott, erasing local superconductivity","From depleted band to Mott collapse: tunneling spectroscopy in 4Hb-TaS2"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The spectral evolution is assumed to be a filling-controlled approach to half filling driven by interlayer charge transfer, rather than a charge-density-wave gap, stacking disorder, or surface potential variation.","fun_headline_variants_meta":{"raw":{"variants":["Spectra map Mott transition and pairing collapse in 4Hb-TaS2","Nano puddles reveal superconductivity dies at Mott threshold in 4Hb-TaS2","Mott transition seen in spectra as superconducting puddles vanish","Doping tunes 4Hb-TaS2 toward Mott, erasing local superconductivity","From depleted band to Mott collapse: tunneling spectroscopy in 4Hb-TaS2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000827,"raw_usage":{"total_tokens":3500,"prompt_tokens":844,"completion_tokens":2656,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":2551}},"tokens_in":588,"tokens_out":2656,"duration_ms":24364,"temperature":1.0,"reasoning_tokens":2551,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T16:44:45.586114+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the local 1T-layer electron filling independently, for example via core-level binding-energy shifts or angle-resolved photoemission band positions on the same crystals, and overlay it on the dI/dV maps: if the spectral gap and the collapse of the quasiparticle peak occur at constant filling or track the known charge-density-wave order instead of the filling axis, the Mott-transition interpretation fails.","supporting_citations":[],"review_version":1}