{"id":"a142bf22-06fa-4908-b55f-67722ee49d1c","arxiv_id":"1908.08568","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A scanning tunneling microscopy study of Co3Sn2S2 confirms a ~300 meV minority-spin gap and reports standing waves near step edges that the authors interpret as signatures of nontrivial surface states around 50 meV.","lead":"Scanning tunneling measurements on the magnetic compound Co3Sn2S2 reveal a roughly 300 meV energy gap in one electron spin channel, supporting its half-metallic character, plus a feature near 50 meV that the authors link to topological surface states. This is a direct, real-space probe of a material considered a magnetic Weyl semimetal, a candidate platform for topological and spintronic devices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fermi-arc surface-state claim is the weakest link: the 50 mV LDOS hump is assigned by energy coincidence and the step-edge standing waves show a q(V) dispersion with 'considerable deviation' from the calculated nontrivial surface-state scattering; trivial/bulk channels are not excluded.","rationale":"The reader's weakest_assumption matches my reading, so I agree with the conditional verdict. The paper is a careful experimental study, and the minority-spin gap evidence is credible; the spin-resolved tip analysis does include a fitted tip polarization (S = 55%), but that is a secondary parameter and does not undermine the direct observation of a spin asymmetry around the gap. The surface-state claim is different: the authors themselves use hedged language ('signature', 'temporarily propose', 'considerable deviation', 'difficult to precisely determine') and call for more intensive calculation. My concern is routed through identification risk, not through disagreement with the Weyl-semimetal classification of Co3Sn2S2. The proposed QPI/model-comparison test is decisive because it replaces single-q line-scan matching with a full set of scattering vectors and explicitly tests whether the nontrivial surface states are necessary to reproduce the data. Since the reader's conditional verdict already withholds full acceptance of the Fermi-arc claim, my stress-test does not move the verdict; it sharpens the condition that would need to be met for acceptance.","tokens_in":9446,"tokens_out":6015,"duration_ms":64961,"concrete_test":"Compute the full 2D quasiparticle-interference (QPI) spectrum of the S-terminated Co3Sn2S2 slab from the surface-projected Green's function at T = 2 K and V_b = 10-80 mV, including both trivial and nontrivial surface states and bulk states, and convolve with the STM tip. From the simulated QPI maps, extract all peak q vectors as a function of energy and compare with the measured q_s(V) in Fig. 5(d) using a model-comparison statistic (e.g., AIC) that also includes a trivial-surface-only and a bulk-only model. If the measured line-scan dispersion is fit as well or better without the nontrivial surface-state scattering channel, the Fermi-arc assignment in Section C is not supported; if the full model is required, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The half-metallicity part of the central claim is reasonably supported: the spin-resolved dI/dV data and the fitted Pse/Psc comparison in Section B provide direct evidence for a minority-spin gap. The load-bearing weakness is the surface-state signature. In Section A, the ~50 mV hump is described only as 'temporarily propose[d]' to come from Fermi arcs; the calculated slab/bulk DOS is nearly featureless there, so the assignment rests on energy coincidence plus defect sensitivity (Fig. S2), which any surface state would show. In Section C, the standing-wave analysis compares a single q_s(E) extracted from a one-dimensional line scan with one selected calculated scattering path among the nontrivial surface states (black arrow in Fig. 5(f)). The paper states that 'there is considerable deviation' between measured and calculated dispersions and that 'a much intensive calculation is required'; this is an internal acknowledgement that the identification is not secure. The line scan cannot separate Fermi-arc scattering from scattering off trivial surface states (red bands in Fig. 5(e,f)), which are dismissed because they merge into bulk states, or from bulk-interference features. With several candidate cuts and bands, matching one q(V) curve is a curve-selection exercise rather than a falsifiable test. If the hump is a bulk or trivial-surface feature, and the standing waves originate from those channels, the nontrivial-surface-state conclusion is unsupported, leaving only the better-supported half-metallic Weyl-node bulk scenario.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports a low-temperature (T = 2 K) scanning tunneling microscopy and spectroscopy study of the magnetic Weyl semimetal candidate Co3Sn2S2. The authors assign the observed cleaved surfaces to S-, Sn-, and Co3Sn-terminations on the basis of morphology and comparison of dI/dV spectra with calculated density of states, and they identify a partially opened gap of ~300 meV in the tunneling spectra. Using an antiferromagnetic Cr-coated tip at magnetic fields of ±3 T, they observe field-dependent spin-resolved tunneling spectra and interpret the ~300 meV gap as the gap in the spin-minority channel, i.e., half-metallicity; a quantitative comparison between the measured effective spin polarization Pse and the calculated polarization Psc is presented, using a tip polarization of S = 55%. The paper's third claim is that a ~50 meV hump in the LDOS and standing waves observed near step edges in the bias range 10-80 mV are signatures of nontrivial surface states (Fermi arc-like states) of the Weyl semimetal, supported by comparing the extracted scattering wavevector qs(V) with selected calculated scattering paths. The authors hedge the surface-state assignment qualitatively ('we temporally propose', 'there is considerable deviation'), but the abstract states that the standing waves 'confirmed' the surface-state signature, which overstates the evidence presented.","tokens_in":9726,"tokens_out":15048,"duration_ms":143140,"significance":"The principal strength of this manuscript is the direct spin-resolved tunneling study: the ±3 T field-dependent dI/dV data with an antiferromagnetic Cr-coated tip (Fig. 4(a)), the reproducibility check in Fig. S3, and the agreement of the minority-spin gap with DFT calculations constitute credible spectroscopic evidence for the half-metallic band structure of Co3Sn2S2, complementing transport and photoemission studies of this material. The surface-termination assignments (Figs. 2-3) are careful, and the authors are commendably explicit about where their evidence falls short. If the surface-state interpretation were secured, this would be an important first STM signature of Fermi-arc surface states in a time-reversal-symmetry-breaking magnetic Weyl semimetal, in a competitive and active field (cf. contemporaneous Refs. [25,26]). As it stands, the Fermi-arc claim rests on two observations, each with admitted or plausible alternative explanations (a bulk or trivial-surface origin of the 50 meV hump; trivial-surface or bulk-interference standing waves), so this part of the paper currently contributes a proposed signature rather than a demonstration.","major_comments":[{"comment":"The identification of the ~50 meV LDOS hump with nontrivial (Fermi arc) surface states is load-bearing for the paper's second central claim, but the text itself confines this to a temporary proposal ('Consequently, we temporally propose that the Fermi arcs induce the additional hump'). Because the slab and bulk calculations are 'relatively featureless' in this energy range, the hump is assigned purely by energy coincidence with the predicted Weyl-node energy and by its sensitivity to defects (Fig. S2); neither property is specific to Fermi arcs, as trivial surface states and defect-sensitive bulk resonances would behave similarly. Since this same hump defines the energy window in which the Section III.C standing waves are analyzed, the surface-state conclusion is not independent of this assignment. Please strengthen this point with spatial dI/dV maps of the hump, a quantitative comparison of the measured hump with the computed surface-projected DOS, or an explicit downgrade of the claim; as written, the hump alone does not discriminate Fermi-arc states from other surface or bulk features.","section":"Section III.A, Fig. 2(d)"},{"comment":"The standing-wave analysis is the second pillar of the surface-state claim and, as the paper acknowledges, it is not yet conclusive. A single qs(V) curve extracted from one 1D line scan is compared to a few selected calculated scattering paths (e.g., the black arrow in Fig. 5(f)); the manuscript states that 'there is considerable deviation' between the measured and calculated dispersions and that 'a much intensive calculation is required.' The trivial surface-state bands (red in Figs. 5(e,f)) are dismissed because they merge into bulk states, but a 1D line scan sums over all scattering channels and cannot exclude trivial-surface or bulk-interference origins. With several cuts and several candidate bands available in Fig. 5(e), selecting paths after the fact is a curve-selection exercise rather than a falsifiable test. I request one of the following: (i) standing-wave dispersion measured at additional step edges with known crystallographic orientation, to verify that the q(V) anisotropy matches the calculated scattering path; (ii) a comparison of qs(V) with all candidate scattering vectors (nontrivial, trivial, and bulk) for the relevant cuts; or (iii) a revision of the conclusion and the abstract so that the standing waves are described as consistent with, rather than a confirmation of, Fermi-arc surface states.","section":"Section III.C, Figs. 5(d)-(f)"},{"comment":"The quantitative agreement between Pse and Psc is presented with a fitted parameter. The text states 'Taking S = 55%, Pse fits Psc,' so the tip polarization ratio S is adjusted, and since Pse scales inversely with S (see the unnumbered equation defining Pse, which is missing from the displayed text), the amplitude of Pse is matched by construction; the zero-crossing positions of Pse are independent of S, which gives the comparison some content, but the 'fits nicely' claim overstates the test. The raw spin-resolved dI/dV curves in Fig. 4(a) independently support the minority-spin gap, so the qualitative half-metallicity conclusion is not in question. I ask that the missing equation be restored, that Pse be shown for a range of S values (e.g., 40%-70%) with the calculated Psc for comparison, and that an independent calibration of the Cr-tip polarization be reported if available; otherwise the text should describe the comparison as consistency within the assumed S value rather than as a two-sided confirmation.","section":"Section III.B, Fig. 4(d)"}],"minor_comments":[{"comment":"'This is further confirmed by the observations of standing waves' overstates Section III.C, which reports 'considerable deviation' between the measured and calculated dispersions; please replace 'confirmed' with a weaker verb such as 'corroborated' or qualify the statement.","section":"Abstract"},{"comment":"'Figures 2b and 2c show typical topographies of Co3Sn2S2 for the thick plate-type sample and thin flake-type sample, respectively' is inconsistent with the Fig. 2 caption, which identifies Fig. 2(c) as a height profile; the flake-type topography appears in Fig. 3(b), so the cross-references should be corrected.","section":"Section III.A"},{"comment":"The conflicting termination assignments between Ref. [25] and Ref. [26] for the same surface are noted only implicitly; since the termination identification underlies the later surface-state analysis, a sentence explicitly reconciling the assignment criteria of the two references would strengthen the argument.","section":"Section III.A (termination discussion)"},{"comment":"The residual LDOS definition Delta(dI/dV)(r,V) = HL[dI/dV(r,V), omega] - (1/n) sum_r HL[...] should state the type and cutoff frequency omega of the low-pass filter and should specify which curves enter the average; the text refers to 'the total number of the averaged curves' without defining the set.","section":"Section III.C"},{"comment":"Reference [25] is an arXiv preprint; if the published version is now available, please update the citation and any statements that depend on it.","section":"References"},{"comment":"'Type-IA half-metallic ferromagnetism' is an unusual term; please define or rephrase it, and consider replacing 'semimetal gap' in the Abstract with a term that distinguishes the partially opened minority-spin gap from a conventional semimetal gap.","section":"Introduction and Abstract"}],"recommendation":"major_revision","confidential_remarks":"The recommendation is driven by the two-pillar structure of the surface-state claim: the Section III.A hump assignment is admittedly 'temporary,' and the Section III.C standing-wave comparison admits 'considerable deviation,' so the 'nontrivial surface states' part of the title is not yet demonstrated. The half-metallicity evidence, in contrast, is solid and could support publication on its own. A second editorial point: the manuscript tests predictions from the same group (Refs [20-22]) within the same band-structure framework, and in Section III.C the comparison is made to selected scattering paths; contemporaneous independent work on the same material (Refs [25,26]) bears directly on the Fermi-arc question, and I recommend asking the authors to discuss those results explicitly in the revised version, in addition to reporting whatever additional measurements they can provide."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know before you read. The spin-resolved STS half of the paper is in good shape; the Fermi-arc half is much weaker than the title implies. If you take away one solid result, it's the direct tunneling evidence for a ~300 meV minority-spin gap in Co3Sn2S2.\n\nWhat is actually new: spin-resolved dI/dV spectra with a Cr-coated tip, tracking the minority-spin gap across EF, and a Pse/Psc comparison that follows the calculated spin polarization below EF. The surface-termination assignment for S-, Sn-, and Kagome-terminated surfaces is also reasonable, and the paper is careful about comparing to slab calculations. The 50 mV hump and the step-edge standing waves are new data, but here the analysis starts to bend.\n\nWhere the soft spots are: the surface-state claim rests on two load-bearing assumptions that the paper itself flags. The 50 mV hump is only \"temporarily proposed\" to be Fermi arcs, with no fingerprint that distinguishes it from a trivial surface or defect state. The standing-wave dispersion q_s(E) is compared to one selected calculated scattering path, and the match has \"considerable deviation\" by the authors' own words. They say a much more intensive calculation is required. That is an admission that the identification is not secure. Scattering from trivial surface states or bulk interference is not excluded. Also, the tip polarization S=55% is fitted so that Pse matches Psc; that doesn't sink the half-metallicity claim, but it means the quantitative polarization number is not independently calibrated. The paper cites the competing STM work (Morali, Yin) and the original theoretical predictions (refs 20-22), so the novelty is limited but the citation practice is honest. This is not a case of overclaiming to the point of fatal flaw; the limitations are in the text and the authors are up front about them.\n\nWho this is for: experimentalists working on magnetic Weyl semimetals, especially anyone using spin-polarized STM. The half-metallicity confirmation will be useful; the surface-state part is useful mainly as a baseline for better momentum-resolved measurements.\n\nRecommendation: send it to a serious referee. The half-metallicity data deserve to be published, and the Fermi-arc claim is the kind of thing a referee can push to be softened or better supported. If the authors address the standing-wave discrepancy and calibrate S independently, the paper could be solid.","headline":"The spin-resolved STS evidence for half-metallicity is solid, but the Fermi-arc surface-state claim rests on a selective comparison the authors themselves admit deviates considerably.","tokens_in":10333,"tokens_out":2533,"would_cite":false,"duration_ms":26671,"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 spectroscopy of the magnetic Weyl semimetal candidate Co3Sn2S2 finds a ~300 meV minority-spin gap and attributes 50 meV standing waves to nontrivial surface states.","keywords":["magnetic Weyl semimetal","Co3Sn2S2","Kagome lattice","half-metallicity","spin-resolved scanning tunneling microscopy","Fermi arc surface states","standing waves","anomalous Hall effect"],"falsifier":"A momentum-resolved photoemission map of the (001) surface at about 50 meV would settle the claim: if the constant-energy contours do not show non-closed Fermi arcs connecting the projected Weyl points, then the signature the paper assigns to nontrivial surface states is something else.","tokens_in":9232,"feed_emoji":"🧲","tokens_out":11625,"duration_ms":104625,"temperature":0.7,"pith_summary":"This paper sets out to show, with low-temperature scanning tunneling microscopy and spectroscopy, that the Kagome-lattice ferromagnet Co3Sn2S2 carries the electronic-structure signatures expected of a magnetic Weyl semimetal. The authors identify the surface terminations exposed by cleaving, measure a partially open gap near 300 meV in the local density of states, and use spin-resolved tunneling with a chromium-coated tip to show that this gap sits in the minority-spin channel, confirming half-metallicity. They also find an extra hump in the density of states near 50 meV, the energy where earlier calculations place Weyl nodes, and observe standing waves near step edges with a dispersion they attribute to scattering between nontrivial surface states. The payoff, if the interpretation is right, is a concrete material in which broken time-reversal symmetry, a gapped spin channel, and topological surface states coexist and can be probed locally.","feed_headline":"Co3Sn2S2 shows a 300 meV spin gap and Weyl surface-state signatures","feed_subtitle":"Spin-resolved tunneling places the gap in the minority band and ties the 50 meV hump to Weyl surface states.","key_machinery":"The load-bearing object is Co3Sn2S2, a Kagome-lattice ferromagnet predicted to be a half-metallic magnetic Weyl semimetal: in the majority-spin channel, nodal rings gapped by spin-orbit coupling leave three pairs of Weyl nodes about 50–60 meV above the charge neutrality point, while the minority-spin channel is gapped at the Fermi level. The measurement machinery is scanning tunneling spectroscopy: normal tungsten tips give the local density of states, chromium-coated antiferromagnetic tips give spin-resolved spectra, and step edges act as scattering centers that produce Friedel oscillations. The comparison quantities that carry the argument are the effective spin polarization of the tunnel junction versus the calculated band spin polarization, and the measured scattering wavevector versus the calculated scattering paths between adjacent nontrivial surface states.","core_discovery":"On its own terms, the central discovery is that the bulk and surface electronic structure of Co3Sn2S2, as calculated, shows up in tunneling spectra point by point. On S-terminated surfaces the dI/dV curves display a partially opened gap of roughly 300 meV, with additional peaks that match the projected density of states; the Kagome-lattice surface shows the same features with a reduced gap near 250 meV. Spin-resolved spectra taken at ±3 T with an antiferromagnetic Cr tip show the gap in one spin channel and a V-shaped metallic density of states in the other, and the extracted junction spin polarization tracks the calculated spin polarization for a tip polarization near 55%. Around 50 meV, where the projected bulk density of states is featureless, the measured LDOS shows a hump; the authors propose it comes from nontrivial (Fermi-arc) surface states, and they support this with standing-wave oscillations between 10 and 80 meV near step edges whose scattering wavevectors have similar behavior to calculated scattering paths between nontrivial surface states, though with considerable deviation.","pith_inferences":["Beyond the paper, a momentum-resolved photoemission experiment in the same 50 meV window could directly image the Fermi arcs and confirm or refute the surface-state assignment proposed here.","Beyond the paper, repeating the standing-wave analysis on step edges along several crystallographic directions, and in magnetic fields that shift the Weyl nodes, would test whether the measured scattering wavevectors track the nontrivial surface-state bands.","Beyond the paper, if the half-metallic gap is robust, Co3Sn2S2 becomes a natural testbed for combining topological surface-state transport with spin-polarized bulk conduction, possibly enabling spin injection and detection in a single material."],"forward_implications":["Co3Sn2S2 can be treated as a half-metallic ferromagnet around the Fermi level, so the spin-polarized transport and the observed anomalous Hall effect have a microscopic origin in the majority-spin channel.","Because the Weyl nodes sit only 50–60 meV above the charge neutrality point, transport signatures such as negative magnetoresistance from the chiral anomaly should appear at relatively accessible doping or bias energies.","The proposed Fermi-arc surface states near 50 meV can be localized with STM, making step edges and defects usable local probes of topological surface bands in magnetic Weyl systems.","The Kagome lattice in this material hosts both flat-band and Dirac physics and, if this paper is right, magnetic Weyl fermions as well."],"supporting_citations":[{"why":"It predicts the half-metallic magnetic Weyl state and reports the anomalous Hall effect, setting the band-structure expectations this paper tests.","marker":"[20]"},{"why":"It performs band-structure calculations placing Weyl nodes 50–60 meV above the charge-neutrality point and provides the calculated surface-state dispersions used for comparison.","marker":"[21]"},{"why":"It independently predicts Weyl nodes and Fermi arcs in Co3Sn2S2, benchmarking the ~50 meV energy window chosen for the surface-state analysis.","marker":"[22]"},{"why":"It identifies the Sn-terminated surface in prior STM work, providing the basis for the termination assignments made here.","marker":"[25]"},{"why":"It is a concurrent STM study of the same Kagome material whose flat-band peak comparison supports the spectral assignments.","marker":"[26]"},{"why":"It reviews spin-resolved STM and STS techniques, supplying the method and practical spin-polarization limits of the Cr-coated tip used here.","marker":"[31]"},{"why":"It derives the spin-valve effect and the effective spin-polarization formula used to compare experimental and calculated spin polarization.","marker":"[34]"},{"why":"It demonstrates that step-edge Friedel oscillations map surface-state dispersion, motivating the standing-wave analysis around step edges.","marker":"[38]"}],"fun_headline_variants":["Co3Sn2S2 tunneling reveals 300 meV spin gap and Weyl surface states","Spin-resolved STM spots half-metallic gap and nontrivial surface states in Co3Sn2S2","Magnetic Weyl semimetal Co3Sn2S2: 300 meV minority-spin gap, 50 meV surface states","Tunneling into Co3Sn2S2: minority spin gap at 300 meV and Weyl surface hump"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim stands or falls on the interpretation that the ~50 meV hump and the step-edge standing waves come from nontrivial surface states rather than from trivial surface states or bulk interference; the paper itself notes a considerable deviation between the measured dispersion and the calculated scattering paths.","fun_headline_variants_meta":{"raw":{"variants":["Co3Sn2S2 tunneling reveals 300 meV spin gap and Weyl surface states","Spin-resolved STM spots half-metallic gap and nontrivial surface states in Co3Sn2S2","Magnetic Weyl semimetal Co3Sn2S2: 300 meV minority-spin gap, 50 meV surface states","Tunneling into Co3Sn2S2: minority spin gap at 300 meV and Weyl surface hump"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000838,"raw_usage":{"total_tokens":3672,"prompt_tokens":984,"completion_tokens":2688,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":600,"completion_tokens_details":{"reasoning_tokens":2569}},"tokens_in":600,"tokens_out":2688,"duration_ms":17142,"temperature":1.0,"reasoning_tokens":2569,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:35:16.602562+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A momentum-resolved photoemission map of the (001) surface at about 50 meV would settle the claim: if the constant-energy contours do not show non-closed Fermi arcs connecting the projected Weyl points, then the signature the paper assigns to nontrivial surface states is something else.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It predicts the half-metallic magnetic Weyl state and reports the anomalous Hall effect, setting the band-structure expectations this paper tests."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It independently predicts Weyl nodes and Fermi arcs in Co3Sn2S2, benchmarking the ~50 meV energy window chosen for the surface-state analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It is a concurrent STM study of the same Kagome material whose flat-band peak comparison supports the spectral assignments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It reviews spin-resolved STM and STS techniques, supplying the method and practical spin-polarization limits of the Cr-coated tip used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It derives the spin-valve effect and the effective spin-polarization formula used to compare experimental and calculated spin polarization."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It demonstrates that step-edge Friedel oscillations map surface-state dispersion, motivating the standing-wave analysis around step edges."}],"review_version":1}