{"id":"70e03c4c-75ce-4d7e-990b-857775ae240d","arxiv_id":"2412.02599","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"LEED shows no CDW-induced superstructure peaks on the Sb-terminated surface of CsV3Sb5, implying the surface lattice distortion is less than half the bulk value.","lead":"Using a micrometer-sized electron beam, the authors searched for the structural distortion accompanying the charge density wave in the kagome metal CsV3Sb5 and found none on the antimony-terminated surface. The result suggests this surface has a much weaker lattice distortion than the bulk, which matters for interpreting surface-sensitive measurements of these superconductors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Non-detection of sharp superstructure peaks does not by itself prove a reduced PLD amplitude; finite-size or short-range CDW order would broaden the signal into diffuse background that the analysis removes.","rationale":"The most load-bearing step in the paper is the inference from a null sharp-peak measurement to a quantitative reduction in the lattice-distortion amplitude. The reader identified the same weakness: the absence of a sharp superstructure peak can also be explained by finite-size or short-range CDW order, which broadens the superstructure signal and pushes its peak height below the 3σ detection threshold even when the integrated intensity and local displacement amplitude are bulk-like. I agree with that assessment. The paper provides no diffuse-scattering analysis, no temperature dependence of the suspected superstructure positions, and no estimate of the surface CDW coherence length, despite citing STM observations of a 2×2 motif on the same termination. The termination assignment is reasonably supported by the I(V) comparison, and both terminations give comparable predicted superstructure intensities, so termination uncertainty is a weaker objection. The quantitative 'less than half' claim depends on comparing a sharp-peak calculation with a sharp-peak detection threshold, so the unresolved order/coherence question is the primary limitation. This does not mean the conclusion is wrong; it means the central claim is conditional on ruling out a finite correlation length. Since the reader's conditional verdict already reflects this limitation, no change in verdict is needed.","tokens_in":13392,"tokens_out":5344,"duration_ms":65235,"concrete_test":"Re-analyze the existing long-integration LEED images (or acquire new ones) at 30 K and at 120 K (>T_CDW) on the same high-quality Sb-terminated area, using only flat-field correction and no Gaussian background subtraction. Compute a difference map (30 K minus 120 K) and integrate intensity around each expected (0.5,0.5) position as a function of distance from the exact spot. If the difference shows a broad feature (FWHM substantially larger than the main-lattice spots) with integrated intensity comparable to the bulk-predicted superstructure intensity, then the null result in Fig. 4 is due to short-range or small-domain order rather than a factor-of-two amplitude reduction.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central inference—absent sharp superstructure peaks imply a PLD amplitude less than half the bulk value—holds only if all 2×2 structural order on the Sb-terminated surface is long-range and coherent across the ~80–100 µm beam, so the superstructure intensity is concentrated in sharp spots that can be compared with the 3σ noise threshold of Fig. 4c. If the surface CDW instead has a finite correlation length (small domains, stacking disorder, or short-range order), the superstructure scattering spreads into broad peaks or diffuse intensity. The analysis pipeline in Appendix C—15×15 binning followed by subtraction of a slowly varying background with a Gaussian filter of 100-pixel width (6–7× the main-spot width)—would suppress precisely such broad features, even if their integrated intensity corresponds to a bulk-like distortion amplitude. This is not a purely hypothetical alternative: the paper itself notes that STM studies observe a 2×2 motif on the Sb termination, so a CDW-like modulation is present at this surface. The quantitative statement 'less than half its bulk value' also inherits this problem because the quadratic scaling argument in Fig. 2b applies to the peak intensity of a coherent spot, not to integrated diffuse intensity. Without a measurement of diffuse scattering above T_CDW or a spot-profile/domain-size model, the experiment does not uniquely distinguish reduced amplitude from reduced coherence.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports low-energy electron diffraction (LEED) measurements on the Sb-terminated surface of the kagome metal CsV3Sb5 in its charge-density-wave (CDW) phase, using a µm-sized beam to locate high-quality areas on several cleaved crystals from two batches. The authors compare the measured diffraction patterns with dynamical LEED simulations that incorporate the bulk 2×2×2 periodic lattice distortion (PLD) determined by x-ray diffraction. No superstructure peaks are observed at the positions and intensities expected from the simulations, despite a measured 3σ noise floor that lies below the predicted signal at two electron energies (80 eV and 130 eV), with a factor-of-four margin at 80 eV. From the quadratic dependence of superstructure intensity on distortion amplitude, the authors conclude that the PLD amplitude in the surface layer is less than half its bulk value, and they discuss possible electronic and structural origins for this suppression.","tokens_in":13618,"tokens_out":4923,"duration_ms":51717,"significance":"If the conclusion is correct, the paper provides an important and surprising observation: the lattice component of the CDW can be strongly suppressed at a nominally clean surface, even though the electronic 2×2 modulation is still observed by STM. This would have implications for the interpretation of surface-sensitive probes (STM, ARPES) in kagome metals and for the general question of how bulk and surface CDW order are coupled. The paper's strengths are its use of externally determined bulk distortion parameters from x-ray diffraction, the systematic screening of multiple cleaves and crystals from two growth batches, and a careful, quantitative noise analysis with a clearly defined detection threshold. The central inference, however, depends on the assumption that any 2×2 surface order is long-range and coherent across the probe area; the alternative of finite-size or short-range order, which would broaden the superstructure signal into a diffuse background, is not experimentally excluded.","major_comments":[{"comment":"The central quantitative claim—that the absence of sharp superstructure peaks implies a PLD amplitude less than half the bulk value—holds only if all 2×2 structural order on the Sb-terminated surface is long-range and coherent over the ~80–100 µm beam, so that the superstructure intensity is concentrated in sharp spots that can be compared with the 3σ threshold of Fig. 4c. The analysis pipeline in Appendix C (15×15 binning and subtraction of a slowly varying background using a Gaussian filter with 100-pixel standard deviation, 6–7× the main-spot width) would suppress precisely the broad features that would arise from finite correlation lengths, small domains, or stacking disorder. This is not a hypothetical concern, because the paper itself notes that STM studies observe a 2×2 motif on this termination. The quadratic scaling argument in Fig. 2b applies to the peak intensity of a coherent spot, not to integrated diffuse intensity. To support the amplitude interpretation, the authors should either measure the diffuse scattering directly (for example, above T_CDW or as a function of temperature) or provide a spot-profile/domain-size model that converts the observed upper limit into an amplitude constraint under finite-coherence assumptions.","section":"Section 3 (Fig. 3)"},{"comment":"The termination assignment relies on a qualitative visual comparison of the (00) and (01) spot intensities with dynamical LEED simulations over the range 66–150 eV, without a quantitative metric such as an R-factor or a residual sum. The quantitative comparison in Fig. 4d assumes a pure Sb termination for the predicted superstructure intensities. If the probed area contains a non-negligible fraction of the (√3×√3)R30° Cs termination or other defects, the expected superstructure peak strengths could be modified. Please provide a quantitative termination analysis (e.g., an R-factor or a comparison of several spots over a wider energy range) or a sensitivity study showing that the predicted superstructure intensities are robust against plausible fractions of Cs-terminated patches.","section":"Section 4 (Fig. 4)"},{"comment":"The quantitative comparison between predicted superstructure intensity and the measured detection threshold is performed at only two electron energies (80 eV and 130 eV) and on a single high-quality surface position of sample K2, although the screening described in the text covered multiple cleaves and several positions. The claim of robustness would be considerably strengthened by repeating the noise analysis and detection-threshold determination at additional energies (for example, at the energies where the simulations predict the highest superstructure intensity) and on at least one additional high-quality surface position from a different batch, to confirm that the factor-of-four margin at 80 eV is not specific to one location or one energy.","section":"Section 4 (Fig. 4)"}],"minor_comments":[{"comment":"The sample labels (sample D, sample K1, sample K2) are introduced in the text and figures but never summarized in a table. Please provide a table listing the sample label, growth batch, which cleave was used, and which measurements (spatial scan, LEED-I(V), long-exposure search) were performed on each.","section":"Section 1"},{"comment":"Please specify the exact region from which the noise histogram in Fig. 4c was extracted (for example, the stripe between which main-lattice spots) and state explicitly that this region does not contain any expected superstructure position. It would be useful to also report the noise level measured in the immediate vicinity of the (0.5 0.5) position.","section":"Section 4"},{"comment":"The statement 'less than half its bulk value' is an upper limit derived from the chosen 3σ threshold and the 4× margin at 80 eV. Please state explicitly that this is an upper bound and not a measured value with an uncertainty, and consider providing the corresponding confidence level.","section":"Section 4"},{"comment":"The text says the LEED-I(V) scans cover '66 eV to 150 eV', but the horizontal axis in Fig. 3 appears to start at 60 eV. Please align the text and figure axes, and define the energy step used in the scans.","section":"Section 2"},{"comment":"Please specify which main-lattice peaks are summed to normalize the experimental intensities in Fig. 4d and whether the (00) spot is included in that sum. This is needed to reproduce the quantitative comparison.","section":"Section 4"},{"comment":"The conclusion is phrased as applying to 'the Sb-terminated surface layer', but at the electron energies used (80–130 eV) the probing depth includes several atomic layers below the surface. Consider wording such as 'the near-surface region' or 'the Sb-terminated surface and adjacent layers' to avoid overstating the depth resolution.","section":"Title and Abstract"},{"comment":"References 24, 26, and 52 are cited as arXiv preprints. If any of these have been published in peer-reviewed journals by the time of resubmission, please update the citations.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the diffuse-scattering alternative to a reduced PLD amplitude. This is a standard concern in any LEED search for superstructure peaks, and the paper currently does not provide a quantitative way to exclude it. The authors have the tools to address it—for example, by acquiring LEED patterns above T_CDW and subtracting them from the low-temperature patterns, or by modeling the spot profile expected for finite domains and comparing the integrated intensity. If that can be done, the paper would be a solid contribution. The termination assignment is also qualitative; a quantitative R-factor analysis would be appropriate. The paper fits the scope of a surface-science or correlated-electron journal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague, read this one. It is the first LEED study targeting the CDW periodic lattice distortion at a cleaved surface of a kagome metal, and it is cleaner than most surface-structure papers in this subfield. The experiment is a real new measurement, not a reanalysis: µm-sized beam, six cleaves from two growth batches, long integrations, and a careful 3σ noise analysis. The dynamical LEED simulations use the bulk 2×2×2 distortion from external x-ray data as input, so there is no circularity between prediction and comparison. The termination assignment by LEED-I(V) is qualitative but gets support from the Cs-termination spot extinctions that are absent in the data. Credit where due: this settles the obvious question of whether STM's 2×2 motif on the Sb termination comes with a bulk-like atomic displacement. Apparently it does not, and that is a useful, surprising result.\n\nThe soft spot is exactly the one the stress-test flags, and I think it lands. The quantitative claim—PLD amplitude less than half the bulk value—holds only if the 2×2 structural order on this surface is long-range and coherent over the ~100 µm beam, so its scattering is concentrated in sharp spots. If the surface CDW forms small domains or has short-range order, the superstructure intensity spreads into broad or diffuse features, and the Appendix-C background subtraction (Gaussian filter of 100-pixel width) would remove precisely those features. The paper's own observation that STM sees a 2×2 motif on this termination makes the short-range-order alternative concrete: the electronic modulation exists, and the structural response may be present but disordered. The quadratic scaling argument in Fig. 2b applies to peak intensity of a coherent spot, not to integrated diffuse intensity, so the factor-of-two bound inherits that limitation. This is not a fatal flaw; the central statement can be softened to \"no long-range PLD detectable, consistent with surface suppression or short-range order.\" Two other minor issues: only two energies are used for the quantitative comparison, and the detection threshold is measured in one region of the pattern. Neither changes the qualitative conclusion.\n\nThe paper deserves a serious referee. I would send it out. The authors should be asked to either measure diffuse scattering or provide a spot-profile/domain-size estimate before the amplitude bound is stated in the abstract. For someone working on kagome superconductors or surface CDW physics, this is a citation-worthy new data point even in its current conditional form.","headline":"First surface-sensitive structural probe of the CDW lattice distortion in CsV3Sb5; the non-detection is real, but the 'less than half bulk amplitude' bound assumes coherent spots rather than short-range order.","tokens_in":14220,"tokens_out":1032,"would_cite":true,"duration_ms":14102,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Low-energy electron diffraction finds no CDW superstructure peaks on the Sb-terminated surface of CsV3Sb5, implying the periodic lattice distortion there is less than half its bulk value.","keywords":["kagome metal","CsV3Sb5","charge density wave","periodic lattice distortion","low-energy electron diffraction","LEED","surface termination","antimony termination"],"falsifier":"A decisive test would be a surface-sensitive diffraction experiment that specifically looks for broad diffuse intensity around the expected CDW positions: if the integrated diffuse signal matches the total scattering predicted from the bulk distortion, then the reduced-amplitude conclusion would be wrong. Alternatively, finding the predicted (0.5, 0.5) superstructure peak on a freshly cleaved, predominantly antimony-terminated surface below the transition temperature, with a noise floor comparable to this experiment's, would directly contradict the claim.","tokens_in":13193,"feed_emoji":"🔬","tokens_out":11987,"duration_ms":109310,"temperature":0.7,"pith_summary":"This paper tries to establish that in the charge density wave (CDW) phase of the kagome metal CsV3Sb5, the atomic lattice distortion on the antimony-terminated surface is much weaker than in the bulk. Using low-energy electron diffraction (LEED) with a micrometer-sized beam on multiple cleaved crystals, the authors looked for the superstructure spots that a 2 × 2 × 2 bulk distortion should produce, and found none at the intensities predicted by dynamical LEED calculations. Because superstructure intensity grows with the square of the displacement amplitude, the non-detection corresponds to a surface periodic lattice distortion below half the bulk value. If correct, this means structural and electronic signatures of the CDW can decouple at the surface, and surface measurements of charge order cannot be used to infer the bulk lattice response.","feed_headline":"Kagome metal's surface distorts less than half as much as its bulk","feed_subtitle":"Electron diffraction finds no charge-order lattice spots on the antimony surface, so its atomic response is weak.","key_machinery":"The argument is carried by low-energy electron diffraction with a micrometer-sized beam, used to measure sharp diffraction patterns on high-quality patches of cleaved surfaces, together with dynamical LEED simulations of the bulk 2 × 2 × 2 distortion. The central quantitative lever is the quadratic relation between superstructure spot intensity and distortion amplitude: the measured noise floor sets an upper bound on any undetected peak, and that bound translates directly into an upper bound on the surface atomic displacement. The termination assignment comes from comparing measured and simulated spot intensities as a function of electron energy, a LEED intensity-versus-energy analysis.","core_discovery":"The central discovery is that the Sb-terminated surface of CsV3Sb5 does not show the diffraction spots that the bulk CDW's periodic lattice distortion would force it to show. In multiple samples and at several electron energies, no (0.5, 0.5) or (0, 0.5) superstructure reflexes appear above the noise floor, even though dynamical LEED simulations built from the x-ray-determined 2 × 2 × 2 bulk distortion predict those spots at two to four times the detection threshold. Since spot intensity scales quadratically with atomic displacement, the authors estimate the surface-layer periodic lattice distortion is less than half the bulk value. They also verify that the measured main-spot intensity-versus-energy curves match the antimony termination rather than the caesium termination, so the missing peaks are tied to the Sb surface itself.","pith_inferences":["The paper leaves open the possibility that what is reduced at the surface is not the atomic displacement but the CDW domain size or ordering quality: short-range or small-domain order would broaden the superstructure signal into diffuse scattering that the background subtraction removes, and a dedicated diffuse-scattering measurement would decide between these readings.","If the hole-doping explanation generalizes, a testable extension is that the same suppression should appear at the Sb-terminated surfaces of RbV3Sb5 and KV3Sb5, and it might be tunable by surface electron doping.","The micrometer beam's spatial mapping could be used to correlate local structural suppression with cleavage steps, strain, or caesium coverage, turning a null result into a map of where the CDW lattice coupling fails."],"forward_implications":["An STM-visible 2 × 2 charge modulation on the Sb surface is not reliable evidence for a bulk-like periodic lattice distortion at that surface.","The structural response to the CDW can be termination-dependent, so bulk structural models from x-ray diffraction should not be applied blindly to surface-sensitive electronic measurements.","The surface suppression strengthens the case that carrier concentration or surface polarity controls the CDW's coupling to the lattice, since the Sb termination is effectively hole doped and hole doping suppresses long-range CDW order in bulk.","A controlled experiment depositing caesium on the Sb surface should restore the anticipated superstructure peaks if the suppression is truly termination-specific."],"supporting_citations":[{"why":"Provides the x-ray-determined atomic displacements of the bulk 2 × 2 × 2 CDW structure that seed the dynamical LEED calculations.","marker":"[11]"},{"why":"Documents the Sb-terminated and partial caesium-terminated surfaces produced by cleaving, used to identify the surface termination.","marker":"[34]"},{"why":"Describes the micrometer-sized-beam LEED apparatus that makes spatially resolved diffraction on small high-quality surface patches possible.","marker":"[37]"},{"why":"Shows the prior TaS2 surface case where the CDW lattice distortion differs moderately from the bulk, serving as the comparison benchmark for the stronger suppression seen here.","marker":"[41]"},{"why":"Finds CDW-related electronic splitting only on alkali-terminated surfaces, supporting the termination-dependent picture into which the structural result is placed.","marker":"[42]"},{"why":"Extends the termination dependence of electronic states to all AV3Sb5 compounds, used to frame the surface-bulk dichotomy.","marker":"[43]"},{"why":"Shows that hole doping suppresses long-range CDW order in bulk CsV3Sb5, the analogy the paper invokes for the effectively hole-doped Sb surface.","marker":"[44]"}],"fun_headline_variants":["Kagome surface misses charge-order lattice spots","Sb surface of kagome metal barely distorts","Kagome metal's surface distortion less than half bulk","No CDW superstructure on Sb surface of CsV3Sb5","Surface of kagome metal resists periodic lattice distortion"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reasoning depends on treating the missing superstructure spots as evidence of small atomic displacements rather than of CDW domains too small or disordered to give sharp peaks above the detection threshold.","fun_headline_variants_meta":{"raw":{"variants":["Kagome surface misses charge-order lattice spots","Sb surface of kagome metal barely distorts","Kagome metal's surface distortion less than half bulk","No CDW superstructure on Sb surface of CsV3Sb5","Surface of kagome metal resists periodic lattice distortion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000856,"raw_usage":{"total_tokens":3685,"prompt_tokens":880,"completion_tokens":2805,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":496,"completion_tokens_details":{"reasoning_tokens":2736}},"tokens_in":496,"tokens_out":2805,"duration_ms":21386,"temperature":1.0,"reasoning_tokens":2736,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T23:16:21.028591+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive test would be a surface-sensitive diffraction experiment that specifically looks for broad diffuse intensity around the expected CDW positions: if the integrated diffuse signal matches the total scattering predicted from the bulk distortion, then the reduced-amplitude conclusion would be wrong. Alternatively, finding the predicted (0.5, 0.5) superstructure peak on a freshly cleaved, predominantly antimony-terminated surface below the transition temperature, with a noise floor comparable to this experiment's, would directly contradict the claim.","supporting_citations":[{"cited_title":"Stahl, D","cited_arxiv_id":null,"evidence_quote":"Provides the x-ray-determined atomic displacements of the bulk 2 × 2 × 2 CDW structure that seed the dynamical LEED calculations."},{"cited_title":"Vogelgesang, G","cited_arxiv_id":null,"evidence_quote":"Describes the micrometer-sized-beam LEED apparatus that makes spatially resolved diffraction on small high-quality surface patches possible."},{"cited_title":"von Witte, T","cited_arxiv_id":null,"evidence_quote":"Shows the prior TaS2 surface case where the CDW lattice distortion differs moderately from the bulk, serving as the comparison benchmark for the stronger suppression seen here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Finds CDW-related electronic splitting only on alkali-terminated surfaces, supporting the termination-dependent picture into which the structural result is placed."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Extends the termination dependence of electronic states to all AV3Sb5 compounds, used to frame the surface-bulk dichotomy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows that hole doping suppresses long-range CDW order in bulk CsV3Sb5, the analogy the paper invokes for the effectively hole-doped Sb surface."}],"review_version":1}